Mikhalchishin/Stetsko

Fighting Chess with Magnus Carlsen

ProgressinCiiess

Volume 31 of the ongoing series

Editorial board
GM Victor Korchnoi
GM Helmut Pfleger
GM Nigel Short
GM Rudolf Teschner

2012
EDITION OLMS

m
Adrian Mikhalchishin Oleg Stetsko

FIGHTING CHESS WITH

MAGNUS CARLSEN

Translated and edited by Ken Neat

2012
EDITION OLMS

m
The Authors

Adrian Mikhalchishin (born 1954) has an excel lent i nternational reputation, above all as a fine chess trainer.
T h e Ukrainian-born grandmaster worked from 1980 to 1986 with the 12th world champion Anatoly Karpov,
and he has also assisted Zsuzsa Polgar, M aya Ch iburdanidze, Vassily l vanchuk and Arkady Naid itsch. As
chairman of the FIDE Trainers' C o m m ission, M i khalch ishin can also call on on h i s wealth of experience as
an active player. He participated in several USSR national championships, 4th place in 1984 in the 51st
championship in his home city of Lvov bei ng his greatest competitive success. He has also taken part i n
four Chess Olympiads, three t i m es f o r his new homeland o f Slovenia (2000, 2002 a n d 2004).

Oleg Stetsko (born 1936) was a professional ai rforce pilot and a chess player of Soviet master standard,
who played several times i n the strong USSR Armed Forces championships. After h i s ret i rement i n 1984 he
was appointed to the trai n i n g staff of the USSR national team, which in 1989 he helped to win the second
World Team Championship i n Lucerne. Stetsko then worked as a journalist for the prestigious 64 magazine,
of which Anato ly Karpov was the editor-in-chief. In the 1990s together with the well-known grandmaster
Eduard Gufeld he wrote several books, includ i n g The Complete Dragon and Winning with the Torre Attack.

Bibliograph ical Information published by T h e Deutsche

Nationalbibliothek

The Deutsche Nationalbibliothek l i sts this publ ication in

the Deutsche Nationalbibliografie; detailed bibliographic
data i s available on the Internet at http://d nb.d-nb.de.

Copyright 2012 Edition Olms AG

W i l l i konerstr. 10a CH-8618 Oetwi l a.S./ZOrich, Switzerland
E-mail: info@ed ition-olms.com
Internet: www.ed ition-o l m s.com

All rig hts reserved. T h i s work is protected by copyright. No part of this publ ication may be reproduced,
stored i n a retrieval system or transm itted i n any form or by any means, electronic, electrostatic, magnetic
tape, photocopying, record ing or otherwise, without prior perm ission of the publ isher.

Printed in Germany
Printed on acid-free and ageing-resistant paper

Editor: Ken Neat

Photog raph: Ray M o rris-Hi l l Photog raphy

Translator: Ken Neat
Typesetti n g by: Art & Satz Ulrich Di rr, D-80331 Munich
Printed by: Druckerei Friedr. Sch mucker GmbH, D-49624 Lon i ngen
Cover: Ni klas WeiB, D-79104 Freiburg

ISBN13: 978-3-283-01020-1
 Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Mag nus Carlsen - Hero of the Computer Era . . . . . . . . . . . . . . . . . . . . . . . . 10
G. Kacheishvili-M. Carlsen 14 M. Maki Uuro- M. Carlsen 14 M. Carlsen-D. Kedik 15

F. Bindrich- M. Carlsen 15 D. Pavasovic- M. Carlsen 16 E. Rozentalis-M. Carlsen 16

M. Carlsen- L. Aronian 17 A. Onischuk- M. Carlsen 17 C. Pedersen- M. Carlsen 18

M. Carlsen-G. Kamsky 18 M. Carlsen-A. Adly 19 M. Carlsen-A. Naiditsch 19

L. Aronian-M. Carlsen 20

1 From Master to Grandmaster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Game 1 M. Carlsen- H. Harestad 23 Game 2 M. Carlsen-S. Zhigalko 26

Game 3 M. Carlsen-S. Ernst 31 Game 4 M. Carlsen-S. Dolmatov 34

Game 5 M. Carlsen- E. Shaposhnikov 35 M. Carlsen- G. Kasparov 39

Game 6 M. Carlsen- E. V ladimirov 41 Game 7 D. Palo- M. Carlsen 43

Game 8 M. Carlsen- N. Short 46

2 The Way to the Top . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

P. H. Nielsen- M. Carlsen 50 Game 9 M. Carlsen-A. Shirov 51

Game 10 I. Cheparinov- M. Carlsen 55 Game 11 M. Carlsen- P. Nikolic 57

Game 12 M. Carlsen-K. Lie 62 Game 13 B. Jobava- M. Carlsen 66

Game 14 M. Carlsen- F. Amonatov 71 Game 15 M. Carlsen- G. Kamsky 73

Game 16 M. Carlsen-A. Beliavsky 77 M. Carlsen -A. Beliavsky 80

Game 17 M. Carlsen-J. Nunn 81 Game 18 M. Carlsen- M. Adams 84

Game 19 M. Carlsen-A. Morozevich 90 Game 20 A. Morozevich- M. Carlsen 95

M. Carlsen-A. Shirov 99 Game 21 M. Carlsen- A. Morozevich 101

M. Carlsen-V. Topalov 107 Game 22 M. Carlsen-V. lvanchuk 107

Game 23 M. Carlsen- L. Aronian 112 Game 24 M. Carlsen- L. Aronian 115

Game 25 L. Aronian-M. Carlsen 118 M. Carlsen- L. Aronian 123

M. Carlsen-Y. Pelletier 125 Game 26 M. Carlsen-A. Onischuk 125

Game 27 M. Carlsen- T. Radjabov 129 Game 28 V. Kramnik- M. Carlsen 133

Game 29 D. Jakovenko-M. Carlsen 135 Game 30 M. Carlsen- L. Dominguez 140

Game 31 M. Carlsen- M. Adams 143 Game 32 M. Carlsen -I. Cheparinov 149

5
3 Life at the Top . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Game 33 M. Carlsen- P. Eljanov 155 Game 34 V. Kramnik- M. Carlsen 158

Game 35 V. Topalov- M. Carlsen 163 Game 36 M. Carlsen-L. Aronian 166

Game 37 M. Carlsen-V. Topalov 169 Game 38 T. Radjabov- M. Carlsen 174

Game 39 M. Carlsen- V. lvanchuk 178 Game 40 Y. Pelletier- M. Carlsen 184

Game 41 L. Dominguez- M. Carlsen 187 Game 42 M. Carlsen- T. Radjabov 191

Game 43 L. Aronian- M. Carlsen 195 Game 44 M. Carlsen- L. Aronian 198

Game 45 M. Carlsen -A. Beliavsky 201 Game 46 M. Carlsen- L. Dominguez 205

Game 47 M. Carlsen- V. Anand 208 Game 48 L. Dominguez- M. Carlsen 212

Game 49 M. Carlsen-A. Grischuk 215 Game 50 M. Carlsen-V. Topalov 218

Game 51 M. Carlsen-Wang Yue 221 Game 52 M. Carlsen-D. Jakovenko 226

Game 53 M. Carlsen- P. Leko 230 Game 54 M. Carlsen-V. Topalov 233

Game 55 M. Carlsen-T. Radjabov 237 Game 56 M. Carlsen- V. Anand 240

Game 57 M. Carlsen -R. Ponomariov 244 Game 58 M. Carlsen- V. Kramnik 247

Game 59 S. Karjakin- M. Carlsen 252 L. Aronian -M. Carlsen 256

Game 60 L.-D. Nisipeanu- M. Carlsen 258 Game 61 R. Ponomariov- M. Carlsen 261

Game 62 M. Carlsen- E. Bacrot 265 Game 63 M. Carlsen-H. Nakamura 269

Game 64 M. Carlsen- L.-D. Nisipeanu 271

4 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
Tournament and Match Record 274

Index of Carlsen's Opponents 279

Index of Openings 280

6
7
 Preface
The swift rise of the young Norwegian Mag
nus Carlsen to the top levels of chess has
literally overturned impressions about the
possibilities of teaching talented children.
Indeed, to progress at a young age from a
run-of-the-mill master to a real contender
for the chess crown in just five years is
something that only an outstanding tal
ent is capable of doing. In the history of
chess only Garry Kasparov has achieved
any similar. In this occurrence there is
much that is surprising: both the fact that
such an extraordinary talent appeared in
'non-chess' Norway, and the way itself in
which this chess mastery was acquired.
For graduates and successors of the So
viet Chess School, nurtured in a 'chess'
foundation established over decades, and
fostered from childhood in clubs at Pio
neers Palaces and then in chess schools
run by famous mentors, it is hard to imagine
a boy grasping the wisdom of chess virtu
ally on his own.
This is correct only in the sense that
Magnus likes to study chess on his own.
But there were also attendant circumstanc
es: Magnus's father, Henrik Carlsen, was
well known in Norway as a chess organiser,
and it was he who noticed his son's talent,
observing that already at the age of five
he had good analytical ability. And, not
surprisingly, knowing the precedent of the
famous Hungarian Judit Polgar, from an
early age the father began assisting his son
to become a chess professional, hoping
8
for great achievements. Another impor
tant factor was the boy's home situation.
Magnus has a secure basis-a wholesome
family, in which his sisters virtually idolise
him, and the attention and support of rela
tives and of society. Already at the age of
ten he had the help of a trainer, and as his
prowess developed he also acquired the
support of financial bodies, including the
Microsoft company.
By present-day junior standards, Mag
nus began a serious study of chess rather
late-at the age of eight. The initial motiva
tion was a promise by his father to include
him, if he played well, in the Norwegian
under-11 championship. This stimulated
the boy to study chess for 2-3 hours a
day over a period of six months. The rest
was accomplished by his all-consuming
interest in the game. Many point to the
prevalence of computer preparation, even
calling Magnus a computer boy. He does
not deny his enthusiasm for the computer;
indeed it would be strange not to make
use of it in the age of universal comput
erisation. For modern children, using a
computer does not present any particular
problems, and a grasp of the accompa
nying programs comes quite naturally to
them. As for purely chess preparation,
Carlsen uses the computer as an instru
ment, and he relates sceptically to con
tentions that it is emasculating live chess.
The main thing for Magnus is improving
his mastery and achieving results. Here
 Preface
it is appropriate to recall the pithy saying
of Mikhail Botvinnik: 'Chess cannot be
taught, one can only learn it!' It is unlikely
that Magnus is familiar with the Patriarch's
claim, but he has patiently learned, pro
ceeding with the help of the computer,
picturesquely speaking, in seven-league
steps.
Another very important factor in his im
provement has been the reading of chess
literature. In the words of his father, he
literally swallowed all Kasparov's series
My Great Predecessors and from them he
remembers practically everything. These
books are a kind of academic course on
the classical heritage of chess. And the
fact that he prefers to work on chess on
his own ('Magnus is very independent
as regards preparation, I only help him',
testifies his trainer Simen Agdestein), is
merely an indication of his character. In
short, we see an example of the effective
study of chess by correspondence, based
on fertile ground. In the opinion of one
of the best trainers of the present time,
grandmaster Yuri Razuaev, the boy pos
sessed the gift of self-instruction.
It would be incorrect to talk about the
ineffectivness of this type of study, with
references to a lack of 'schooling'. There
9
have been countless examples of success
ful self-realisation. His way is familiar to
me not just by hearsay, since at a mature
age I managed to achieve chess master
standard and complete a correspondence
course at the Aviation Institute.
By his 19th year Carlsen had completed
his chess Academy, he had received high
marks from his examiners by victories in
very strong tournaments, and he was con
tinuing to improve. This is indicated by his
attainment of the high point 2800 in the
world rating list, which allows him to be
regarded as the strongest player in the
West. Carlsen has made his 'way to the
top' in chess high society. One of the many
ways available to a talented young person.
We would like to wish that his 'life at the
top' does not collapse, as in the novel of
the same name by the well-known English
writer John Braine.
In conclusion, we would like to thank
grandmaster Andrey Devyatkin for help in
describing the childhood stage of Magnus
Carlsen's chess career.
Oleg Stetsko
USSR Master of Sport
Senior Trainer of the USSR
national team 1985-89
 Magnus Carlsen - Hero of the
Computer Era
When we beg i n examining the chess phe
nomenon Magnus Carlsen and his way
to the top, the fi rst th i ng that comes to
m i n d is whether we can compare the tal
ented Norweg ian with the bri l l iant chess
gen iuses of the second half of the 20th
centu ry, Robert Fischer and Garry Kasp
arov. What is it that unites him with these
great champions? Above all - a bri l l i ant
m e m o ry. H i s fi rst tra i ner, g ran d m aster
Si men Agdestei n , to ld me that he gave
the you n g Magnus an open i n g book to
read , and the next day, as if noth ing had
happened , he performed the new open
ing to a h i g h standard . True, a striving for
sharp positions i m med iately on emerg ing
from the open i n g , as possessed by the
young Fischer and Kasparov, was not very
apparent, but already at the age of fifteen
he was confidently playi ng complicated
open ings such as the Sveshnikov Variation
and the Botvi n n i k Variation.
The greatest influence on the develop
ment of the you n g talent was u ndoubt
ed ly made by the Norweg ian grandmaster
Si men Agdestein . Th is showed itself not
only in chess, but also in the fact that Mag
nus is a good footbal ler. After al l , his trainer
played for his cou ntry at both chess and
football , and he even scored a goal against
Italy! They met when Magnus was 1 0 years
old. Initially the boy studied with the Norwe
g ian master Torbj0rn Ringdal Hansen , who
was doing his m i l itary service at a Col lege
for Top Athletes, where Agdestein ran the
chess section. And with in a year he made
i ncred i ble progress. This is not surprisi ng:
M ag n u s was l iterally gripped by c hess,
playing i n all possible tournaments and
read ing nu merous chess books in English.
His fi rst book was Find the Plan by Bent
Larsen. At that time Agdestein correctly
thought that, for lessons with a boy who
was at the stage of masteri ng the basics
of chess, a computer was not needed .
And only later, when the open ing became
an i mportant element of the play in tour
naments, M ag n u s began work i n g with
the analytical modu les Fritz and Rybka.
Agdestein worked with Magnus several
times a month, spending two or three hours
on the analysis of a game. I n Agdestein's
opinion, children cannot be forced to study
chess. Of cou rse, it is not easy to teach a
youngster to play, but if you can arouse his
ind ividual interest, miracles occur. Th is is
how it was with Magnus - motivation was
the main key to success!
M ag n u s t h o u g ht that on the way to
hig her achievements he should become
N o rweg ian cham p i o n , but in 2004 and
2005 the road was blocked by his expe
rienced teacher, and it was only in 2006
that he managed to su rpass h i m . Surpris
ing ly, Magnus was also unable to win i n
t h e young age categories o f t h e European
10
 M ag nus Carlsen - Hero of the Computer Era

Cham pionsh i ps. But where now are those by move', so to speak. Although he was
who fin ished ahead of him? In the history sti l l a long way from the best players of this
of chess this has also occurred with other type, such as Fischer and Kasparov, he
players. For example, up to the age of 1 6 was nevertheless notable for his sensible
the great Garry Kasparov was twice unable and logical play. True, at ti mes his play
to demonstrate his superiority in class. I n looked somewhat simplified , sometimes
the fi rst stage o f h i s work with M agnus, automated , which was a consequence of
Agdestein set the objective of reach i n g fascination with computer preparation. Not
master level o f play a n d he regarded indi without reason , Garry Kasparov warned
vidual tournaments only as stages towards about the poi ntlessness of such play. Th is
the ach ievement of this aim. But al ready is how he descri bes the essence of the
in 2004 he realised that he was u nable to play in so-cal led post-modern ism chess:
raise the youngster to a higher level , and ' Nowadays most you n g p l ayers s i m ply
the search for a new trai ner began. move the pieces i n antici pation of a mis
Contact with Garry Kasparov early in take, l i ke an ice-hockey team knocking
2005 was not developed any further. But the puck into the opponent's zone, hoping
a mentor was found on Scandi navian soi l . that someth ing will come of this. The pure
After contacts at a number of joint tourna prag matism of the young shows itself in
ments in Scandinavia, the strong Danish many games. Carlsen sti l l needs to raise
grandmaster Peter Heine N ielsen began his u nderstand i n g to m atch u p with h i s
working with the you n gster. For a long energy a n d optimism'.
time he had been keeping an eye on the But even so, Magnus did not develop
young Norweg ian and a year earlier he had l i ke a computer player, although the work
com mented in the press: the West had with N ielsen , who made active use of the
not known such a great talent si nce the computer, seriously developed the ana
time of Fischer, and at such a young age lytical aspect of his talent, which is often
Carlsen was not inferior to either Karpov or called playing computer chess. Is this good
Kasparov. At that time N ielsen was actively or bad? N owadays chess has changed ,
col laborating with Vishwanathan Anand, and the top players are not inclined to phi
and later, when he was having to work with losophise about the position; they need
both of them , he wou ld sometimes enlist quickly and efficiently to find not so much
Mag n u s to p repare Vishy for i m portant the best move, as the optimal decision.
events, such as, for example, the match This enables them for a lengthy time to
for the world championship with Kram nik maintain the tension on the board and to
in 2008. set the opponent the maximum discomfort
Working with N ielsen , Carlsen substan in the solving of constant problems. The
tially expanded h i s o pen i n g repertoi re, plans for conducting the game in this type
which had to be greatly modern ised . For of chess are far less concrete than during
example, with Black he stopped playing the the ti mes of the g reat world champions,
King's I ndian Defence, which did not suit who ad hered to the classical style of play,
his style. At that time Magnus was most with the possible exception of M i khail Tal .
successful in tense positions, in play 'move But these plans are based on the i nten-

11
 M ag nus Carlsen - Hero of the Computer Era
..;;;..
_______
sity level of the position and are solved
exclusively by concrete measures, not by
general reasoning.
The well-known trainer Arshak Petro
sian, observing Carlsen's play at the C
Tournament in Wijk aan Zee in 2004, was
surprised at how confidently the 13-year
old boy could 'serve with either hand', i.e.
play both 1. e4 and 1. d4 with equal suc
cess. At so young an age, such a thing was
typical only of Boris Spassky, the most uni
versal player of the Soviet Chess School,
who (stretching the point somewhat) can
be regarded as the forerunner of the mod
ern computer generation. Assessing Mag
nus's play, Arshak Petrosian quite seriously
warned his pupil, Peter Leko: ' Here is your
future rival, you should observe his play
very carefully. He is already quite a strong
positional player, and he places his pieces
correctly - a rare quality for such an age'.
A new stage in Carlsen's chess biogra
phy began quite recently - in the summer
of 2009, when he began working with Garry
Kasparov. Their first meeting occurred in
2004 at the festival in Reykjavik, when the
prodigy's play impressed Kasparov in their
rapid game, which the maestro saved only
with great difficulty. A year later Magnus
was invited by Kasparov to Moscow for
lessons with the famous trainers Yuri Ra
zuvaev and Alexander Nikitin. But this work
was not continued. Carlsen declined to
follow the program developed by the Rus
sian trainers, and decided to improve his
chess independently.
However, without highly-qualified help
it proved beyond Carlsen's powers to im
prove on the 'star' level achieved in 2008
(his official rating reached 2776), and the
assistance of Kasparov proved most op
portune. For Garry the field of activity was
12
very fruitful. Magnus possesses a natural
gift for correctly evaluating a position, a
very rare quality, one which at a young
age has been possessed by many world
champions. In Kasparov's opinion, in style
of play the Norwegian youngster is close
to Karpov, Smyslov and Capablanca. But
since Magnus's style of play is markedly
different from that of Kasparov, his contact
with the famous champion will undoubt
edly expand his chess views. Usually a
trainer works with his pupil, but it is unlikely
that this is what Kasparov does. His role
is rather that of a mentor; there are things
he can pass on to his protege. There is no
doubt that Carlsen's opening repertoire will
be strengthened -this is a strong compo
nent of Garry. Especially since as yet the
level of the youngster's opening repertoire
does not correspond to his strength of play.
At the same time it is important to learn a
more productive organisation of the work
ing process, in order to make maximum
use of his natural talent.
I n the opinion o f Arshak Petrosian,
working with Kasparov should not so much
change the young Norwegian's opening
repertoire, so much as shift his motiva
tion to the maximum level. This is the ap
proach of real champions to the play in
every game, and the accompanying feeling
of one's own superiority can be instilled
only by a great champion. This approach
to the fight was displayed in October 2009
at the super-tournament in Nanjing. It was
there that it became apparent what Mag
nus has in common with the champions
- he began to be feared!
I saw Magnus at many Olympiads and
European Championships during his de
velopment period. And the first thing that
strikes you (incidentally, this is also the
----------------
Magnus Carlsen - Hero of the Computer Era
opinion of the well-known manager Silvio
Danailov) is the feigned lack of energy, both
in external appearance, and in play. Al l the
champions, apart from Karpov and Pet
rosian , looked very energetic, sometimes
even frightening. As for Karpov, he had
enormous i nternal energy, which he was
able to transfer to the chess board . The
drive of Kasparov and Fischer was felt as
soon as their opponent arrived at the chess
board . Carlsen conducts certai n parts of
a g ame as though i m percepti bly, i n the
style of Karpov or Adams, but, in contrast
to them , he not only does not avoid sharp
variations, but happily goes in for them.
An indication of the youngster's thorough
self-confidence is that he ventures to play
very sharp variations, without working out
all thei r details, which Kasparov always
aimed to do. Yes, there are instances when
he plays superficially, but this is typical of
young players, who do not always u nder
stand and sense the dangers associated
with the opponent's counterplay. As for
Carlsen, he possesses such a feel ing, and
this is very i m portant i n high-level play. It
is another m atter that his power begins
to display itself only from a certain point
in the game, but here he can already be
irresistible. Agai n , this is a vivid charac
teristic, typical of Karpov and to a certai n
extent of Fischer.
I n recent t i m es M ag n u s h as g reatly
added to his depth of understanding of
chess, he has beg u n to display an abil
ity to find min imal and latent chances i n
the position and he has clearly become
more forcefu l i n his play. That is, he has
mastered the transition to total chess, the
readiness to fight to the end. We are seeing
the development of a fig hter. It should be
said that, i n their youth, neither Fischer nor
----------------
t2J
---
Kasparov was an uncompromising fighter.
They became this with the g rowth i n the
confidence and power of their play and,
let's make no secret of the fact, an aware
ness of their mission in the chess world. We
are also observi ng this phenomenon with
the increase i n the mastery of the young
Norwegian, which is reflected i n an abil ity
to find the best decisions in any position.
But Carlsen is insepara b l e from h i s
era. He is sim u ltaneously also a computer
player - it is all the same to h i m what he
plays, as he looks for the best or optimal
solution on the basis of his own i n d ividual
algorith m . Carlsen is not a strateg ist and
not a thin ker - he is a performer, l i ke an
eng i neer finding the sol utions to specific
problems . And his play rese m bles the
output g raph of an atomic power station
- constant and h i g h . Although h i s play
lacks fantastic moves and d iscoveries, it
is staggering in its logic and composure. It
is as though he has no nerves - the typical
character of a northerner. But perhaps it
is not just a question of this? N ot without
reason , at the conclusion of the Nanjing
tou rnament of 2009, the 2nd-placed To
palov declared : 'I was the win ner among
people, but Carlsen is simply from another
planet ! ' There you have an explanation,
which does not make the Carlsen enig m a
a n y more understandable.
All these best q ualities of the you n g
Norweg ian are what w e have tried to pre
sent in his games. Through the chronol
ogy of the games we wanted to show the
development of the most talented player
from the West, and to compare Magnus
at the formation stage with the modern
M agnus. Of cou rse, there are many as
pects to his play and in the framework of
one book it is i m possible to encom pass
13
 M agnus Carlsen - Hero of the Computer Era

the unbounded . Therefore we will restrict G. Kacheishvi li - M. Carlsen

o urselves to just one com ponent of h i s Bad Wiessee 2001
play, a very i mportant ind icator o f play a b c d e f g h
ing strength - the abil ity to play the end 8
game, which has i nvariably d isti nguished
7 7
the great champions of the past.
6
It is wel l known that, espec ially with
5 5
the modern strict time control, the end
game can be d ifficult for you n g players, 4

and here they com m it far more mistakes 3 3

than in the midd legame. There are vari 2 2
ous reasons for this: a lack of knowledge 1 J.
of standard positions, a poor mastery of a b c d e f g h
typical manoeuvres and, it goes without
saying, i g n o rance of classic exam ples. 57. .. a5+?
These gaps are not easy to el i m i n ate, and
If possible, one should normally beg i n by
here the serious assistance of a trai ner
advancing a passed pawn - 57. . .f2!, and
is needed . Therefore it w i l l be interest
only then decide what to do next. 58. l:!g8+
i n g for students of chess to see not only
<;!> h5 59. l:!f8 l:!a5+ 60. <;!> d4 l:!xa6 61 . b7
Magnus's mistakes when he was a young
l:!b6 would have led to a d raw.
ster, but also how he was able to correct
them, and on what he is sti l l conti n u i n g to 58. d4 xa6?
work. At the present time his technique of
It was not yet too late for 58 . . . f2 .
converting an advantage in the endgame
resembles that of Fischer, and it rarely 59. b7 b6 60. b8'W xb8 61 . xb8
misfi res - an indication of the serious work f4 62 . f8+ g3 63. e3
he has done. The defence of i nferior posi
tions is another matter - accord ing to the Black resigned .
defi nition of Botvi n n i k, it is this q ual ity of M. Maki Uuro - M. Carlsen
a player that i n d icates his strength of play Helsinki 2002
i n the endgame. And here, as was once
the case with Fischer, M ag n u s has had a b c d e t g h

many problems. But, as the examples will 8 8

show, he is wel l aware of this deficiency 7 7

and is working actively to elimi nate it. It 6 I. 6
can now be said that, whereas i n the initial
5

stage of Carlsen's career the endgame 4 4

was the weak aspect of his play, now we
3 3
can talk about h i s su periority over other
2 2
grandmasters who have crossed the 2700
rating mark. H is l atest tournaments have
demonstrated this. a b c d e f g h

14
 Mag nus Carlsen - Hero of the Computer Era

56. f4 Ja a1 57. Ja b4 Jaf1 + 58. g4 It is usefu l to make use of the 'desperado'

Ja e1 59. e4 f6 60. Ja b6+ g7 61 . f4
f1 +?
Anything, except release the attack on the
passed pawn! 61 . . . h6 or 61 . . . f7 would
have enabled Black to fight for a draw.
62. g5 f3 63. g4 e3 64. xg6+
f7 65. f6+ e7 66. f4 e6 67. g6
e5 68. f8 a3 69. g5 xe4 70. g7
e5 71 . g6 Jag3 72. h7 Ja h3+ 73. g8
e 6 74. g7 Jag3 75. Jaf1 e7 76. Ja e1 +
d7 77. Ja e4
And by constructing a ' bridge', Wh ite won .
M . Carlsen - D . Kedik
Norwegian U nder-1 8 Championship 2002
a b c d e f 9 h
8 8
7 7
6
rook to drive the king away from the pawns.
With the king cut off, it is easier to blockade
the pawns.
56. e8 Ja f4! 57. g5+ h5 58. Ja h7+
g6 59. Ja h 6+ g7 60. h5 Ja e4+
61 . d7 Ja e5 62 . Jag6+ h7 63. d 6
Ja a5 64. c7 Ja e5 65. d7 Ja a5 66. e6
Ja a 6+ 67. f5 Ja xg6
Draw.
After the lesson in this game, all Magnus
could do was to remember for ever the
method of ' h itching' the rook to the king
and the passed pawns. A year l ater he
hi mself saved a similar end ing.
F. Bindrich - M. Carlsen
European U nder-1 4 Championship
Budva 2003

5 5

4 [j, 4

3 [j, 3

2 2

a b c d e f 9 h

48. g4?

First the way for the king should have been

secured - 48. fla4. a b c d e f 9 h

48 ... Ja b3+ 49. f4 Ja b4+ 50. f5 Ja bS+
51 . e6 Ja b4!
It is important not to allow the pawns to ad
vance! Subsequently Black demonstrates
the same mechan ism.
52. Ja a7+ h 6 53. f5 labS+ 54. f6
l'l b6+ 55. f7 f6+!
76. Ja f3 Ja h1 77. g4 Jag1 + 78. Jag3 Ja a1
79. Ja c3 Jag1 + 80. Jag3 Ja a1 81 . d3
Jag1 + 82. f4 f1 + 83. g4 g1 +
84. f3 Ja f1 + 85. g2 Ja f5 86. Jag3 a5
87. h3 Ja a1 88. b3 Ja h1 + 89. g4
g1 + 90. f4 f1 + 91 . f3 h1 92. f5
xh5 93. a3 h1 94. a7+ f8
95. g6 Jag1 96. a8+ e7 97. h6?!

15
 M agnus Carlsen - Hero of the Computer Era
------

There was a win by 97. g8!, intending h7 58. d 6 b6 59. f5

and g5-g6, a manoeuvre which is useful
Black resigned .
to remember.
Bindrich d i d not in fact find this m a
noeuvre (although the possibility presented E. Rozenta lis - M. Carlsen
itself several times), and the game ended Malmo/Copenhagen 2004
in a draw on the 1 1 5th move. a b c d
8 8

7
D. Pavasovic - M. Carlsen
6 6
Wijk aan Zee C 2004
5 5
a b c d e f g h
4 4
8 8
3 3
2
7 7
2
6 6

5 5

4 4
a b c d e f g h

3 .I 3 29 h 6?

2 2 The offer to exchange rooks - 29 . . . f7 !

3 0 . xf7+ xf7 3 1 . fxg4 - wou l d h ave
a b c d e f g h enabled Black to hold the position. M ag
nus did not work out the elementary pawn
44 ...g2? end ing : 31 . . . e6!? 32. g3 fS 33. hS fxg4
Surely it was obvious that Black should aim 34. hxg6 f6, since in the event of 34. h6?
f7 35. xg4 g8 it is Black who wins.
for the exchange of the kingside pawns.
The simple 44 . . . c6! 45. b4 g5 would 30. fxg4 c4 31 . f3 cxd3 32. cxd3 c8
have enabled him to draw. 33. f4 g5+ 34. hxg5+ fxg5+ 35. e4
c1 36. xd4 g1 37. e4 a1
38. c5 xa3 39. d4 g6
The exchange of rooks leads to a lost pawn
Th ings would not have been changed by
end ing .
39 . . . as 40. dS c3+ 41 . b6 d3 42. c6
46...h2 47. xg6+ b5 48. g5+ b6 C3+ 43. d7.
49. h 5 g2 50. a4 f2 51 . h 6+
40. d 5 c3+ 41 . b5 f6 42. d 6 d3
b7 52 . b4
43. c6 aS 44. d7 c3+ 45. b7 b3+
The winning technique is sim ple and in 46. c7 c3+ 47. d8 f7 48. e7+
structive. f8 49. e5

52 . f5 53. h5 e5 54. b5 e1 55. f6 Black resigned . He has no defence against

a1 + 56. b4 b1 + 57. c5 c1 + the check on f5 and then e7.

16
 M ag nus Carlsen - Hero of the Computer Era
M. Carlsen - L. Aronian
FIDE World Cham pionship
Tripoli 2004
a b c d e f 9 h
8 8
a b c d e f 9 h
Th is position occurred in the first game of
the tie-break. It is well known that rook end
ings with f- and h-pawns offer the weaker
side real saving chances. And the fact that
this was a rapid-play game is no excuse for
Carlsen bei ng u nable to save the game. A
top-class g randmaster is obliged to know
key endgame positions.
In end ings of this type the best squares
for the king are considered to be f2 and
g2. Here 66. g2 suggests itself, aiming
to block the h-pawn , and leaving the rook
the option of checking along the ran k and
along the file (the best square for it is a8).
But there fol lowed :
66. mg4?! h a
a n d it tran s p i red that, s i n ce t h e w i n of
the h3- pawn leads to a l ost pawn end
ing, it has become more d ifficult to hold
the position.
67. g5+?
Th is trick, based on stalemate - 67. . . f xgS ?.
- leads to defeat. 67. ga2 fS+ 68. f4 would
have held the posit i o n , i nten d i n g after
68 . . ,gb8 to play 69. ga6+ hS 70. g3
gb3+ 71 . h2 or immediately 69. g3 gh8
70. h2.
67. . . mf7 68. h5 me6 69. h8 c3
10. mt4 mt7 (70 . . Jc2!) 11 . m94 mg6
72. mf4 c2!
The conversion technique is i n structive.
Since 73. gxh3 gf4+ leads to the exchange
of rooks, the pawn contin ues its advance,
and the king makes an outflanking m a
noeuvre.
1a. mga h2 74 . mta mt5 75. h 5 +
m e 6 7 6. h 8 m e 5 7 7. e8 + m d 4
7 8 . d 8 + m c 3 7 9 . h 8 d 2
8 0 . mg3 m d 3 8 1 . mf3 m c 2 82 . mg3
md1
White resigned .
A. Onisch uk - M. Carlsen
Olympiad , Calvia 2004
a b c d e
8 8
7 7
6 j.s
5 5
4 4
fj,
3 !j, cJ;; 3
2 fj, 2
a b c d e f 9 h
Black should be able to hold such an end
ing a pawn down , but Magnus has no clear
impression of how to defend.
44 ...mt8 45. h4 g6 46. mt4 t6?
Such a move is sim ply inconceivable for a
grandmaster. An obvious gap in Magnus's
'schooling'. 46 . . . e7 47. gS hxg5+ 48. hxgS
 Magnus Carlsen - Hero of the Computer Era
f6! , reducing the nu mber of pawns, was
correct.
47. h 5!
N ow holes appear in Black's position.
47 mf7 48.E[b7+ mg8 49. me4 gxh 5
50. gxh 5 mf8 51 .E[h7 E[e1 + 52. mf5
m9a 53.E[d7 E[f1 54. t4 mta 55. d 6
mea 56.E[h7 E[d1 57. mea
Black resig ned.
C. Ped ersen - M. Carls en
Gausdal 2005
a b c d e f g h
8 8
7 7
6 6
5 5
4 playing 45. e3+ e4 (45 . . . e5 46. xh7)
3 3
2 2
a b c d e f g h
This is a theoretically drawn endgame, but
you have to know the defensive proced ure.
91 . Irc2 J.b4 92.E[c4 J. e7 93. mf4 mh 5
I t is correct t o head for t h e corner o f the
opposite colour to the bishop - 93 . . . h3.
94. f5 AdS??
If c h ess n otation a l l owed it, this m ove
would deserve a g reater n umber of ques
tion m arks. The simple 94 . . . h6 would
have enabled Black to draw.
95.E[c8
Black resigned.
18
M. Carls en - G. Kamsky
World Cup, Khanty-Mansiysk 2005
a b c d e f g h
8 8
7
6 6
5 5
4 B .\ .!. 4
B
3 3
2 B B 2
a b c d e f g h
White is a pawn up with a winning position.
All that is requ i red is a certain accuracy.
41 .E[a6+ me5 42 .E[h 6E[d4 43.1tlc4+
mt5 44. h4 ltlt7 45.E[b6?!
45. xh7? was not possible because of
the loss of the rook after 45 . . . g 6 ! , but by
46. e6+ e5 47. c4 White would have
won easily.
45...E[xh4 46.E[xb4
And here 46. g3! h1 47. xf3 would
have led to the win of the f3-pawn.
4 6...E[h3 47. a 5/tlg5 48. a 6??
This elementary oversight changes the
picture. N ow W h ite i s losi n g , altho u g h
a simple knight manoeuvre - 48. e3+
and 49. g4 - would have enabled h i m
t o q ueen his a-pawn i n comfort.
48 . ltle4+ 49. me3 f2+ 50. me2 E[c3?
50 . . . f3! would have won i mmediately.
51 . d2?
The comedy of e rrors conti n u es . After
51 . f1 ! Wh ite wou l d agai n have won :
 Mag nus Carlsen - Hero of the Computer Era

51 . . JU3 52. d6+ xd6 53. a?.
51 . . .J;lxc2 52. J;ld4 xd2 53. J;ld5+ e6
White resigned .
By attacking the pawn, Black releases his
rook. White is short of one move, which
could have been provided by the far-sight
ed 55. d3.

58. J;la6?

M. Carlsen - A. Adly Wh ite should have exchanged rooks, after

Reykjavik 2006 which his knight blockades the pawn pair,

I.
guaranteeing a draw. Now the black pawns

.i.
a b c d e f 9 h
become dangerous .
8 8

7 7 58 J;if8 59. d2 e4+! 60. xe4 J;lf3+

. .

6 M 6 61 . c2 xe4 62. J;la8 J;lc3+ 63. d1

5 .t. 5 J;lb3 64. c2

4 .t. tiJ 4 After 64. f!c8 Black casts a m ati n g net

3 3 around the enemy king: 64 . . . Af4 65. c7
2
2 f!xb7 66. f!e8+ d3.

64 . . . J;lb6 65. lieS d3+ 66. d1 J.f4

a b c d e f 9 h 67. J;le8+ f3 68. b8W J;lxb8

Wh ite's pair of passed pawns i s obv i White resigned .

ously more promising than Black's. But
with which pawn should he beg i n? It is not
hard to calculate that in the event of 55. c7 M. Carlsen - A. Naiditsch
A xe? 56. bxc7+ d7 57. f!a7 c6 58. d3 Sarajevo 2006
cJ? b6 White loses his pawns, and it will be
a b c d e
a draw. Therefore M ag nus begins with the 8 8
other pawn, but he reaches an i mpasse.
7 M
6 .f.
The truth is revealed within three moves.
6
55. b7?!
5 5
This advance should have been prepared 4 4

by bringing up the king - 55. d3! . Now, 3 I. 3

however, Black succeeds i n neutralising 2
2
the pawn pair.
M
a b c d e f 9 h
55 . . . .tc7 56. J;la8 J;lb8 57. d3

After 57. c5+ d5 58. a6 Wh ite loses W h ite's p o s i t i o n is c l ose to w i n n i n g ,

his passed pawns: 58 . . . xc6! 59. xb8+ but he sti l l has to make a few accu rate
cJ? xb7 and the result is a draw. m oves .

57. . . d5 29. J;Icc7?!

19
 Magnus Carlsen - Hero of the Computer Era
Material would have been won by 29. gxa6
g5 30. A c4 e4+ 31 . f3 d2+ 32. e2
gb2 (after the exchange of the minor pieces
Wh ite has a won rook end ing) 33. A xe6+
h8 34. gb6.
29 . . . b2 30. h4?!
A poi ntless move. The king should have
been b ro u g ht i nto p l ay : 3 0 . e1 g b1 +
31 . d2 gb2+ 32. gc2, retaining winning
chances. Now Black succeeds in begin
ning cou nterplay.
30 ...g7!
Before the second rook is activated , the
king m u st be m oved out of the danger
zone. In the event of 30 . . . gda 31 . gxf7
gdd2 32. gfd7 a l i near mate is threatened ,
and Black can not play 32 . . . gxe2+ 33. f1
gf2+ 34. e1 gfe2+ 35. d1 .
31 . xa 6 d8 32 .aa7 dd2 33. xf7+
g 6 34. h5+
The c o n seq uences of 34. f3 g xe2
35. f4 g b4+ 36. e5 g xe3+ 37. d 6
gd4+ 3 8 . e7 gxh4 are unclear.
34 . . . g5 35. g7+ h4! 36. g1?!
I n rook endings, king activity is of g reat
i m portance, but it wou ld appear that the
black rooks are al ready fu lly operational .
If 36. f3 there can follow 36 . . . gxe2 (not
36 . . . e5? 37. g4! with advantage to White)
37. f4 gb5! and Black holds the position.
The exchange of rooks 37. . ,gxg2 38. gxg2
gxg2 39. e5 would merely com plicate
his defence.
36. .. xe2 37.ab7?
As a result, after bei ng a pawn up, Wh ite
loses two. But also after 37J!ad7 ga2! he
can not hope for anyth ing.
20
37. . .xg2+ 38. xg2 xb7 39. a2
39. h6 ga7 40. gg7 gxa5 41 . gxh7 h5
also leads to a draw.
39 . . . xh5 40. a6 a7 41 . f2 g4
42. a4+ g5 43. f3 e5 44. a5 f6
45. g3 e6 46. h4 f6
Draw.
L. Aronian - M. Carlsen
Tal Memorial Tou rnament
Moscow 2006
a b c d e f 9 h
8 8
7 7
6 6
5 5
4 4
3 3
2 2
a b c d e f 9 h
A theoretically d rawn position. Magnus
knew - and had em ployed earlier i n simi
lar positions - the method of keeping the
pawn under fire: 69 . . . ge2 ! . Instead of this
he prefers another plan of defence: attack
ing with the rook from the long side, which
proves to be more complicated .
69 ...a1 70. e7 a5 71 . e6 a7+
12 . d7 as 73. d 6 a7+?
In the flank attack one must not give u p the
8th ran k. 73 . . . g6 was the only defence.
74 eS
Black resigned .
 Magnus Carlsen- Hero of the Computer Era
It has to be said that Carlsen drew the con
clusions from the deficiencies of his play in
endings, and the consequences of the work
he put it soon became apparent. Already
at the Olympiad in Turin (2006) he dem
onstrated endgame play on a grand scale
against Adams (No. 1 8), where he converted
the advantage of the two bishops.
The year 2007 proved exceptionally pro
ductive. There was the impressive conver
sion of an extra pawn in an endgame with
rooks and opposite-colour bishops i n a
game with Morozevich (No. 2 1 ) . There was
an ad mirable depth of ideas in the endings
from the 3rd and 5th games of his Candi
dates match with Aronian (Nos. 23 and 24),
in which with subtle manoeuvres Magnus
was able to confine the enemy king in a
'box'. The finish to his game with Onisch uk
(No. 26), where his knights restricted the
mobi lity of a rook, also i nvites inclusion i n
the books. At t h e World Cup i n Khanty
Mansiysk, in his game with Adams (No. 31)
Magnus again showed brill iant technique in
converting the advantage of the two bish
ops in a complicated multi-piece endgame.
Also instructive is the fi nish to the game
with Chepari nov (No. 32), where l iterally
by nuances Carlsen was able to outplay
an active rook with his bishop and knight.
Also impressive were some masterpiec
es by Carlsen early in 2008. In Wijk aan Zee
there was a nice fi nish to his game with El
janov (No. 33), where he subtly coordinated
his rooks and knight. A special place goes
to his win over the world champion Kramnik
(No. 34) - Magnus was able to construct
another ' box' for the king. At the end of the
year in Bilbao he added to the textbook
col lection of heavy- piece endings: he was
able to convert an outside passed pawn
in a heavy piece endgame with Aronian
(No. 43). Of the games from 2009 mention
should be made of his subtle hand ling of
the endgame against Jakovenko (No. 52),
in which Magnus was able to exploit some
im perceptible errors by the opponent in an
eq ual position.
21
 Chapter 1

From Master to Grandmaster

G ifted young players have always drawn the attention of both ad mirers of the ancient
game and of recog nised experts. During the active development of chess over the last
century and a half, young talents who have gone on to achieve an outstanding level of
play can be counted l iterally in single figures. As for recent ti mes, the record belongs
to the U krai ne-born Sergey Karjaki n , who became a grandmaster at the age of 1 2 .
Such is the trend of the times, the age of techn ical progress. Universal computerisation
has become a kind of accelerator in the development of the new chess generation.
Chess databases and modern analytical modu les enable young talents to acq u i re the
necessary knowledge far more quickly, and lack of experience i n the development of
tactical and strateg ic understanding is com pensated by a computer- l i ke calculation
of variations. Whereas 30 years ago, accord ing to the picturesque expression of Ti
gran Petrosian , young players emerg ing into the chess arena were cal led 'children of
l nformator', now they are cal led 'ch ildren of the computer'. Therefore rumours about
the young Norweg ian boy Magnus Carlsen , who displayed outstanding chess talent
at an eq ually early age as Sergey Karjaki n , quickly spread around the chess world .
But in 2002 i t was t h e U krai nian youngster w h o reigned supreme. The height of
his fame came at a tournament of the Grand Prix series, held in Moscow, where one
of the authors of this book partici pated in the capacity of arbiter. The spectators who
arrived at the Mezhdu narodnaya Hotel were able to observe a curious picture. On an
em pty stage the last pair were continuing to battle: l iterally impending over a frail lad
with his touching l ittle q uiff was a bearded hunk of a man . The experienced l lya Smirin
needed 1 40 (!) moves to overcome the resistance of the 1 2-year-old Seryozha Karjaki n .
Th is was their second game with t h e classical time control; in t h e first t h e grand master
had managed to save a hopeless position a rook down by means of perpetual check.
At that time Seryozha's contem porary - Magnus - was sti l l at the start of his career,
although at the age of eleven he was able to share 1 st-2nd places in the N orweg ian
U nder-1 8 Championship. H owever, he did not d isti nguish h i mself at the European
Under-1 2 Championsh ip in Spai n , where he finished only sixth. But in the world cham
pionship among his contemporaries in Greece, Carlsen displayed his character and
shared 1 st-2nd places with lan Nepom niachtchi, who on the tie-break was proclai med
world champion. Junior events at such a tender age do not usually cause much of
a stir. This makes the almost i m probable episode with Nepomn iachtchi all the more
curious. The emergence of the 1 2-year-old world champion made such an impres
sion on one of the major Russian Newspapers lzvestiya, that it decided to sponsor
h i m , overlooking that he was merely the best player among his contem poraries. True,

22
 M . Carlsen - H. Harestad
------

soon the sponsorship came to nothing, since lan Nepomn iachtchi d isappeared i nto
the general mass of Russian players.
It is possible, therefore, that i n post-Soviet chess they learned about the Norwegian
prodigy only i n 2003, when the magazine 64 - Shakhmatnoe obozrenie reported that in
Scandinavia there was a shining new 1 2-year-old master by the name of Carlsen. First
he shared 3rd-7th places in the N orweg ian Championship, and then he distinguished
hi mself in the Politiken Cup tournament in Copen hagen . This was the first prestigious
tournament i n the world chess calendar i n which Carlsen took part. The i nfluential
Dan ish Politiken newspaper has al ready staged it 25 times, trad itionally providing a
sol id prize fund. Th is attracts strong players, and it is sufficient to recall that among
the winners have been such famous names as Smyslov, Korchnoi and Vaganian. The
young Norwegian also distingu ished h imself i n Copen hagen , scori ng 8 poi nts out of 1 1
games and fi nishing only a point behind the win ner Krishnan Sasikiran. Starting with
a rating of 2385, Carlsen demonstrated a strength of play corresponding to a perfor
mance of 2500 (it should be mentioned that Karjakin already had this official rati ng).
In that period the fou ndation was laid of the mastery that Magnus demonstrated both
in attack on the king, and i n positional play.
Carlsen's 'visiting card ' in Copenhagen was a spectacular m ating attack in his
game with Harestad .

Game 1 b2-b4 is possi ble.

M. Carlsen - H. Harestad
a b c d e f 9 h

..t K.
Copenhagen 2003
8 8

7 K 'if ..t
Ruy Lopez [ C98]
7
1 . e4 e5 2. f3 c6 3 .tb5 a6 4 .ta4

6 . 6

f6 5. 0-0 b5 6 .tb3 .te7 7. Ete1 d 6

5 5

4
_,-
8 . c 3 0-0 9. h3 a5 1 0 .tc2 c5 1 1 . d 4

4
Y!/c7
3 ttJ 3

The classical Chigorin Variation of the Ruy 2 2
Lopez.
1 .:1. .:l.tt:J 1

12. bd2 c 6 1 3 . d5 d8 14. a4 Ela7 a b c d e f 9 h

Black does not want to concede the a-fi le.

1 6 .th 6

However, the presence of the rook on a7

allows White to play b2-b4 in a more fa 1 6. b4!? cxb4 1 7. cxb4 Ad7 1 8. Ae3 .b7
vourable situation. The main continuation 1 9. axb5 A xb5 20. Y!/d2 is more in the spirit
here is 1 4 . . . f!b8. of the position, with the better game for
15. f1 g6 White.

If 1 5 . . . h6, then 1 6. Ae3 with the idea of 1 6...Ete8 1 7. g3

23
 Chapter 1
From M aster to G randmaster

The preparatory 1 7. g4 should have been hopes i nvolving tD h6 and transferri ng his
incl uded. own knight to c5.

17 . ..thd7 1 8. thh2 29. f1 thc5

And here 1 8 . b4 Af8 1 9. A e3 was more a b c d e f 9 h

logical, mai ntain i ng the pawn tension. 8 .1 8

1 8 ...f6 1 9. Ae3 thb6 20. axb5 axb5 7

-*--*-*" 7

21 . Ad3 Ad7 22. Yi'd2 6 'iif .l .l.l 6

5 .l .l 5

.l ttJ
a b c d e
4 4

.I iV tt.J
8 8
3 3

jl
7
2 2
6 6

5 5
1
jl g 1
a b c d e f 9 h
4 4

3 Somehow i m percept i b l y Carlsen has

2 aimed his pieces at the kingside, but Black
does not sense the danger. He continues
a b c d e f 9 h to ignore the move 29 . . . h5, although after
it the black king wou l d have been more
At the young age of 1 2 it is not easy to stay safely placed than in the game. In the event
patient and engage in strateg ic manoeu of 30. tD h6 tD g5 31 . f4 exf4 32. xf4 xh6
vring, trying to find a vul nerable point in (not 32 . . . A xh3? because of 33. tD xh5+!
Black's solid defensive l i nes. But Magnus gxh5 34. e5! f8 35. tDf5+ A xf5 36. xf5
patiently bides his time. and Wh ite wins) 33. h4 f8 Black safely
withdraws his king.
22 . .. thf7 23. xa7 Wxa7 24. We2 Wa 6
25. thg4 c!>g7 30. thh 6! thg5

Black i s n ot tem pted by the p rovoca
t i ve 2 5 . . . h 5? ! , after w h i c h there can
fol l ow 26. tD x h 5 ! gxh5 27. tD h6+ tD x h 6
28. A xh6 Af8 29. xh5 A xh6 30. xh6
e7 31 . e3 g7 32. xf6, when Wh ite
has three pawns for the piece and a con
tinuing attack.
26. Ac1 tha4 27. Ac2 a8 28. We3 c4
Black could have contin ued his waiting
strategy - 2 8 . . . tD b6 2 9 . A d 3 a4, but
28 . . . A xg4!? 29. hxg4 c4 also came into
consideration, nipping in the bud Wh ite's
Black should have restricted h i m self to
the cool- headed exc hange 30 . . . tD x h 6
3 1 . x h 6+ g 8 , w i t h the i ntent i o n of
switching his bishop to g7.
31 . 141 exf4 32. Wxf4 Axh3?
Black is tem pted by the w i n of a paw n ,
underestimating White's latent attacking
resources . It was better to s i m p l ify the
position: 32 . . . xh6 33. h4 f8 34. hxg5+
f x g 5 35. h4+ g7 36. A x g 5 x f 1 +
37. tD xf1 A xg5 38. xg5, although here
too Wh ite's chances are preferable. N ow,

24
 M. Carlsen - H. Harestad
------------------
tt:J

however, Magnus succeeds i n demon a b c d e f 9 h

strating the l atent energy of h i s pieces a J. 8
in all its g l o ry.

33. h4 Ad7

(see next diagram)

34. e5! dxe5 35. t'Llh5+! gxh5

a b c d e f 9 h
Now Wh ite announces mate in 3 moves.
But B l ac k wo u l d also h ave lost after 36. xg5+! fxg5 37. l3f7+ x h 6
35 . . r;!l h8 36. A xg5 fxg5 37. 'Mig3!.
. 38. l3xh7 mate

Carlsen's successful performance in the Politiken Cup tournament greatly raised his
self-esteem , and he was also noticed by the organ isers of other tournaments. True,
his performances in the 'compulsary program' among his contem poraries were not
so successfu l . Competitive fatigue made itself felt. Especially vexing was the set-back
at the European Under-1 4 Championsh i p in Budva (Monteneg ro). When Magnus won
again st his main rival Sergey Zhigalko and took the lead with 6Y2 poi nts after the 7th
rou n d , it appeared that the question of the champion was decided . But he contrived
to lose both his last two games, in which he was winning, and fi n ished only third .
Magnus also started confidently at the world championsh i p i n Halkidiki (Greece),
scorin g 3% points in the fi rst 4 rounds. But u nforeseen circumstances prevented him
from becoming champion. As Agdestein , who was accompanying him, described it, the
air was l iterally buzzing with bacteria and Magnus became i l l (h is tem peratu re reached
40 deg rees). He nevertheless kept in the lead ing group until the 9th round, but on this
occasion he lost to Zhigalko and had to be satisfied with a share of 9th-1 3th places
(?Y2 out of 1 1 ). These set- backs were not accidental : a child's organ ism is sti l l delicate,
and since the 'compulsary prog ram ' for the year was over-generously combined with
the 'free' o ne, by the end of the twelve months Magnus was very tired . This is not
surprising - he had played about 1 50 games!
It is wel l known that i n Norway since long ago they have long shown a great respect
for their heroes, and they try to create the best conditions for the d isclosure of thei r tal
ent. Carlsen was no exception. At the age of ten he drew the attention of the Norweg ian
grandmaster Si men Agdestein, who i n 2002 began working with the prodigy. Despite
the comparative shortness of thei r lessons - twice a month spending 2-3 hours on
the analysis of games played - their collaboration produced resu lts. I n 2003 Carlsen
three times achieved the i nternational master norm (the last time at the Pol itiken Cup
tournament) and by the end of the year he had raised h i s rating to 2450.

25
 Chapter 1
From Master to G randmaster
----------------------

I. I.
Game 2 a b c d e f g h
M. Carlsen - S. Zhiga lko 8 'ii' . 8

1.1\ ..t. .
Eu ropean Under-1 4 Championship 7 7
Budva 2003
6 6

5 . . 5
Sicilian Defence [ 833]

1 . e4 c5 2. f3 c6 3. d4 cxd4
4. xd4 f6 5. c3 e5

Despite its apparently anti-positional na

ture, the Chelyabinsk Variation is sti l l alive
and has not been refuted . Among the el ite a b c d e f g h
players it is constantly employed , for ex
ample, by Radjabov and Shirov. 1 5 . . . e4 1 6. Ac2 'Wc8 1 7. :ae1 0-0
18. Ab3
6. db5 d6 7. Ag5 a6 8. a3 b5

I. 'iV
9. Axf6 a b c d

At the present time Wh ite more often turns 8

to the q u ieter l i ne 9. tb d 5 A e7 1 0. A xf6 7

A xf6, and here 1 1 . c3 or 1 1 . c4. 6

9 . . .gxf6 1 0. d5 f5 1 1 . Ad3 5 VJ!/ 5

4 4
The alternative is 1 1 . c3, to which Black
3
should reply 11 . . . A g7, si nce the capture
on e4 is bad i n view of 1 1 . . .fxe4 1 2 . A xb5!
axb5 13. tb xb5, and against the th reat of
1 4 . tb bc7+ 'i!? d7 1 5. g4+ there is no sat a b c d e f g h

isfactory defence, while 1 1 . . . Ae6? is poi nt
less i n view of 1 2. exf5 A xf5 1 3 . f3 .
1 1 . . . Ae6 1 2. 0-0 Axd5 1 3 . exd5 e7
1 4. c3
The capture 14. tb xb5 Ag7 1 5. tbc3 e4 gives
Black good compensation for the pawn.
14 Ag7
. .
I n his com mentary for ChessBase, Do
rian Rogozen ko rightly remarked that it is
hard to think of a more natural move - the
d5- pawn is defended and the manoeu
vre tb a3-c2 prepared . The d i rect 1 8 . f3
gives Black the opportu nity for rather easy
equal ity: 1 8 . . . b4 1 9. tD b1 bxc3 20. tb xc3
A x c 3 2 1 . b x c 3 x c 3 2 2 . A b1 tb g 6
2 3 . xf5 exf3 24 . xf3 xf3 2 5 . .xf3

(see next diagram) f!ae8 with a d rawn endgame, as i n Karen

1 5. 'Wh5
Theory has grad ually come to the conclu
sion that from the standpoint of fighting
for an advantage 1 5. f!e1 0-0 1 6. h5 e4
1 7. Af1 is more promising.
Asrian's game with Wan g Yue (Khanty
Mansiysk 2005). However, for q u ite a long
time the main conti n u ation was consid
ered to be 1 8 . c.!i h 1 - u nt i l at the tou r
nament i n Dortm und (2002), i n h i s game
with Topalov, Leko employed the strong

26
 M . Carlsen - S. Zhigalko

novelty 1 8 . . . .b8! , ai med at creating q u ick 20. f4 (plan n i ng e3 and g2-g4) or even
cou nterplay on the q ueenside. It remains 1 9. f3!? was more logical.
to add that after the i m med iate 1 8 . . . b4
White has the u n pleasant reply 1 9 . cxb4 1 9 ...1ti g 6 20. ltic2 ltif4 21 . Yh4
A xb2 20 . .e3 ! .
a b c d e
In general, the character of the position
is such that if both sides make the critical
moves (which normally can be achieved
only with the hel p of deep home prepa
ration), the game m ost often concludes
with the mutual el i m i nation of the forces
and a draw.

1 8 ..h8?!
.

Black has confused something - here this

move is poi ntless. Let us briefly exam
ine t h e other poss i b i l ities. Tou rnament
practi ce has shown that after the q u iet
18 . . . g 6 1 9 . c2 W h ite's chances are
preferable (if 1 9 . . . f4 there is a danger
ous exchange sacrifice: 2 0 . 'W g 5 d 3
21 . e3 !). After 1 8 . . . . b 8 Wh ite has time
to consolidate his q ueenside - 1 9. c2 a5
20. a3 ! after which he can cal mly mount
an offensive on the opposite side of the
board .
The only way to equalise is by the ener
getic 1 8 . . . a5! , which occurred a few days
later in the game Carlsen - McShane, which
continued 1 9. xb5 a4 20. Ad1 'll!/c5 21 . d4
W xd5 22 . .e3 'We5 23 . .h3 h6 24. f4 'Wf6
with dynamic equilibrium. If 1 9. 'll!/g 5 the best
is apparently 1 9 . . . 'Wd7! 20. f3 g6! 21 . fxe4
f4 with excellent play for Black (Fernandez
Romero - Eijanov, Andorra 2003).
1 9. h1?!
On the emergence from the open ing, both
players have made decisions that are hard
to understand , which, however, is easily
explai ned i n view of their youth and the
complexity of the position. 1 9. c2 g6
a b c d e f g h
21 ...Ae5?!
It is strange that Black did not in fact carry
out that for which he was obviously aiming:
21 . . . d3, after which White wou ld prob
ably have had to give up a rook for the pow
erfu l knight. H owever, his compensation
would have been sufficient: 22 . .e3 'Wd8!
23. 'W h3 Wg5 24 . .xd3 exd3 25. 'll!/ x d3 f4! ,
and the position is unclear.
22. 1tie3
N ow, as if by sched u le, the wh ite pieces
arrive at the necessary posts.
22 ...9:g8 23. Ac2! b4 24. g3 1tid3
The opponent's in itiative after 24 . . . g6!?
25. 'll!/ h 5 bxc3 26. bxc3 (or 2 6 . b3!? f4
27. f5) 26 . . . A xc3 27. .d1 seemed too
dangerous to Zhigalko, and so he decided
to g ive up a pawn.
2 5 . A x d 3 e x d 3 26. Wxb4
(see next diagram)

27
 Chapter 1
From M aster to G ran d m aster

a b c d e retained the advantage. N ow, however,

Carlsen retains the d5-pawn with the help
of a ' l ittle combi nation'.

6 34. c4! g7
5 5
If 34 . . . A xb2 there follows 35. !!e8+.
4
'f!i 4

8 .t ttJ 8
35. b3
3 3

2 88 8 8 2 b c d e f 9 h
.: .: 1
a

.!
J .J .t .t
8 8

.t .i. -
a b c d e f 9 h 7 7

6 .t_[ l 6

.1 8
White is sti l l a long way from a trouble
5 5
free conversion of his advantage - it is
n o easy m atter to suppress the typical 4 8 1:, 8 4

'Chelyabinsk' activity of the enemy pieces. 3 8 .t 3

2 8 J,
8 2
However, from this moment Magnus begins
demonstrating mature technique beyond 1 ttJ .: 1
his years. a b c d e f 9 h

26 . . 13:b8 27. Wc4!

White has set up a seemingly indestructible
27. Wa3? would have lost to 27. . .f4 28. ti:J g2 monolith, but, strangely enough, Black stil l
fxg3 2 9 . fxg3 A xg3 ! . has some drawing chances.

27 f4! 35 . 13:a5

It is not possible to regain the material: 35 . . . a5 36. !!4e3 d2 37. !!1 e2 was hopeless
27. . . !!xb2 28. W xd3, or 27. . . d2 28. !!b1 . for Black.

28. Wxc8 J3:gxc8 29. /l)d1 ! J3:c5 30. gxf4 36. J3:g1 + f8 37. J3:g2 f!:a3!

30. !!e4! was more accurate: 30 . . . !!xd5 (or
30 . . .fxg3 31 . fxg3 with the same idea of
soon surrounding the d3-pawn) 31 . gxf4.
30 J.xf4 31 . J3:e4 J.e5 32 . f4 J.g7!
Better than 32 . . . Af6.
33. J3:fe1
Or 33. !!f3 !!xd5 34. !!ee3 A h6!.
33 . At6
3 3 . . . h 5 was poss i b l e , altho u g h after
34. cl1 g2 !!xd5 35. cl1 f3 White would have
The two players are as though competing
in inventiveness. Zhigalko persistently tries
to break through the obstructive barriers,
skilfully erected by Carlsen in the path of
the black pieces.
38. J3:e3 J3:e8
If 38 . . . aS, then 39. !!xd3 a4 40. !!dg3 cl1 e7
41 . bxa4 !!xa4 42. !!b3 ! , and Wh ite should
grad ually convert his two extra pawns.
39. J3:d2!
An accurate move, enabling the knight to be

28
 M . Carlsen - S. Zhigalko

activated with gain of tempo. 39. !lxe8+?! Ji x h 6 50. xh 6+ mta 51 . f5 aS

xe8 40. !ld2 aS 41 . lilf2 a4 was weaker.
a b c d e f g h
39 . ..Ji xe3 40. xe3 .id4 41 . f5
8 8

After 41 . !lxd3 A xe3 42. !lxe3 !lxa2 the 7

.i. B tD
win for White is i n question, since his king 6 6
is badly placed . .
B B
5 5

41 . ..Ac5 42 . Jixd3 4 4

B
3 3
42. g2! A b4 43. !lxd3 !lxa2+ 44. g3
2 2
was more method ical , bringing the king
into play.
a b c d e f g h
42 . . .Ji xa2 43. Jih3
52. d 6!
a b c d e f g h
8 8 The timid 52 . d3? a4 53. c3 e8 would
7 7 have delayed the win, although probably
.
.i. B
6 6 would not have thrown it away.

B
5 5
52 ...a4 53. mds Ab4

B
4 4

M If 53 . . . a3 , then 54. xeS e8 (54 . . . a2

B
3 3

2 .I 2
55. d7) 55. til d4 a2 56. tilb3.

1 54. c5 mea 55. c6 Aas 56. mc4

a b c d e f g h
Of course, there was no point in playing
43 . . . mga? 56. c7? A xe? 57. dxc7 d7, since a knight
Blac k s h o u l d h ave p l ayed 4 3 . . . !lf2 ! is bad at combatting a passed rook's pawn.
44. !lxh7 (44. !lh6? !lxf4 45. til xd6? g7) N ow Black is i n zugzwang.
44 . . . g8 4 5 . !l h 3 !lxf4 4 6 . til e?+ g7
s s ...mda 57. e7 hs sa. ts h4 5 9 . h3 ts
(46 . . . f8 47. Cil c6) 47. g2 (or 47. Cil c6),
60. d5 a3 61 . mb3
and although with best play White should
win, he would have to overcome techn ical Black resigned.
difficulties.

44. Ji h 6! This far from faultless but e ntertai n i n g

With the fall of the d6-pawn , Black's entire
position collapses. However, accurate cal
culation is still requ i red of Wh ite.
44 . ..Jia1 + 45. mg2 Jia2+ 46. mf3 Jia3
47. xd 6 Jixb3+ 48. me4 Jib6 49. f5
game is a good i l l u stration o f Magnus's
abil ity from his youngest years to create
positions in which m i stakes are practi
cally inevitable. In so d o i n g , he normally
makes fewer m i stakes i n them than his
opponents.

29
 Ch apter 1 From M aster to G ra n d m aster_________
________._

The lessons of 2003 were not wasted . During this time Carlsen changed l ittle in the
way he worked to improve his standard of play. He sti l l preferred to do this indepen
dently, devoting to it five hours a day. Apart from working at the board , he also read
chess books with great pleasure. True, the range of them was very specific, but here
it is better to hand over to his father, Henri k Carlsen.
' Mag nus has looked at many games comprising the classical heritage of chess,
including all the games from Kasparov's books My Great Predecessors, which he l iter
ally devoured and of which he remem bers practically everything. He reads a lot: Fire on
Board by Shirov, Kram n i k's best games, fundamental works on the endgame and, of
course, many books on the open ings. I should also add the New in Chess magazines
and Yearbooks and other i nteresting chess magazines. Without any d ifficu lty Magnus
combines read ing of chess material and its simu ltaneous analysis. I n the fi rst years
of his chess career he spent a lot of time moving the pieces on the board , repeating
games and remem beri ng ideas he had seen . But in the last two years he has mainly
been analysing i n his head .'
There are those who l i ke to tal k about Carlsen's lack of 'schooling', but I should l i ke
to rem ind them of the possibility of studying by correspondence. As regards its effec
tiveness, everything is decided by motivation. There are n u merous examples of this.
Besides, can one talk about a lack of 'schooling', when you have such a prestigious
trai ner as Simen Agdestein?
Graduates of the Soviet chess school and thei r fol lowers improve their mastery by
studying the classical heritage of the great world champions - Lasker, Capablanca,
Alekhine and Botvinnik, and researchers such as Tarrasch, N i mzowitsch and Reti.
To them it may seem amazing how, without read ing a single book by these classics,
Carlsen independently achieved such a high level of chess u nderstanding. At the time
such a thing seemed i mpossi ble. For example, Tigran Petrosian l i ked to remem ber his
youth, when he kept a copy of N imzowitsch's book My System under his pillow. But,
as we learned earlier, Magnus absorbed the best of the classical heritage from Kasp
arov's five-volume My Great Predecessors. As for the speed of assi m i lating material ,
present-day geniuses have their resources. I n the computer age it is hard to i mag ine a
1 3-year-old boy who does not spend hours on a computer. Carlsen was no different,
using the computer as his main hel per in preparing for tournaments. As he matu red
the efficiency increased , his open ing repertoire expanded , and he perfected his abil ity
to make a deep study of his opponents' play and find their weak poi nts.
Magnus began the year 2004 wel l prepared . One of the pillars of the world chess
calendar is the trad itional festival in the small Dutch town of Wij k aan Zee. It is famous
not only for its main A Tou rnament, which year after year assem bles almost the entire
chess elite, but also its less prestigious B and C all-play-all tournaments, enabl ing young
talents stage by stage to make thei r way into chess high society. And to beg i n with the
C tournament, even a FIDE category 9 - modest by today's standards - is q u ite natural
for a young matador. Carlsen's debut created a genuine fu rore! He not only won with
a score of 1 01/2 from 13 games, but also exceeded the i nternational grandmaster norm

30
 M . Carlsen - S. Ernst

by 1% points. The sharp and energetic attacking style of the 1 3-year-old Norweg ian
testified to the appearance of a new star on the chess horizon. Professionals began
talking about h i m . The highly experienced grandmaster Lj ubomir Kavalek, who in his
time had seen many talents, saw Carlsen as a 'Chess Mozart'. He was echoed by
Evgeny Bebchuk, who in the 1 990s was head of the Russian Chess Federation: 'As a
person closely acq uainted with the play of all the great grandmasters, I can confidently
say that in the history of chess no one has played l i ke this at the age of 1 3 .'
The game with i nternational master Sipke Ernst, played at a key moment of the fight
for the first prize, not only captivated those who value bri l l iancy, but gave a slight insight
into the secrets of Magnus's home laboratory in work with a computer. To a q uestion,
whether the attack on the king was the result of preparation or the fru it of inspiration,
he replied: 'I knew that 1 7. 'We2 was the theoretical move, but I didn't remem ber the
variations, so I thought about it for some 25 minutes, trying to understand the position .
I saw the kn ight move 1 8 . t:D g6 in the game Bologan -Anand (Dortmund 2003). You can
decide for yourself whether it was i mag ination or home preparation ! '.
It wou ld not be out of place to mention: ideas in themselves do not hover i n the air.
The patent on the knight move to g6 in this variation belongs to Alexander Beliavsky,
who twice used it to defeat the famous ' Danish prince' Bent Larsen in 1 981 and 1 984,
at the tournament i n Ti lburg and i n the USSR v. Rest of the World match. (The inqu isi
tive reader is referred to Bel iavsky's book Uncompromising Chess).

1 1 . Af4
Game 3
M. Carlsen - S. Ernst At one time 1 1 . A d2 was also played , since
Wij k aan Zee C 2004 Black's main reaction was considered to be
Caro-Kann Defence [ 8 19] the set-u p with q ueenside castl ing. But in
recent years he has often preferred king
1. e4 c6 2 . d4 d5 3. <tlc3 dxe4 4. <tl xe4 side castling, and in this case the bishop
.ats 5. <tlg3 A9s s. h4 h6 7. <tlf3 <tld7 is more actively placed at f4.
a. h5 Ah7 9. Ad3 Axd3 1 0. Wxd3 e6
1 1 . . . <tlgf6
a b c
a !, If he wants to carry out the plan with queen
side castling, Black can play 1 1 . . . 'W a5+,
provoking 1 2. A d2 (1 2 . c3 makes it easier
6
for him to develop counterplay with . . . c6-
5
c5), and then retreat 1 2 . . . 'Wc7.
4 4

3 VJg tt:Jtt:J 3 1 2 . 0-0-0 Ae7 1 3 . <tle4

2 2
: 1
A popular move, which has the aim in the
1 : j, event of 13 . . . 0-0 of attacking the king's
a b c d e f g h position with the g-pawn - 1 4 . t:D xf6+ t:D xf6

31
 Chapter 1
1 S. g4, as in the game Shirov - Dreev (Dos
Hermanas 2003).
Before this 13. tD eS used to be played .
The problems that Black encou ntered are
wel l demonstrated by the game Beliavs
ky - Tal (Moscow 1 981), which contin ued
13 . . . 0-0 14. e2 aS 1 S. b1 ad8 1 6. c4
tD xeS 1 7. dxeS tD d7 1 8 . d2 A gS (with the
aim of weakening the eS-pawn, but 1 8 . . .
bS!? 1 9. hd1 tD b6 came i nto considera
tion) 1 9. A xgS hxgS 20. h6 (20. hd1 tD xe5!)
20 . . .tD xe5 .
.tJl- .! .. .
' '
... ...
'iV 1.1 ...
!::,
4J
!::, !::, : !::,
:
Analysis diagram
21 . d5! ! . This is now the last romantics
played ! The rook places itself under a four
fold attack , disrupting the coord i n ation
of Black's q ueen and knight. I n the event
of 2 1 . . . cxd5? 2 2 . xeS g x h 6 23. x h 6
f 6 24. xe6+ f7 2 5 . tD h5 Black cannot
hold out. Tal preferred a prosaic solution:
21 . . . xdS! 22. cxdS xdS 23. hxg7 xg7
Y2-Y2. Most of the spectators (and in the
concert hall of the International Hotel there
were more than a thousand) were puzzled
by the peace agreement. This is not sur
prising, since a pretty variation remai ned
off-stage: 24. tD hS+! g6! 25. tDf4+! gxf4
26. h5+ f6 27. h4+ fS 28. hS+ and
Black has to reconcile h i mself to perpetual
check, s i n ce 28 . . . e4? ! 29. e2+ d4
30. d1 + cS 31 . xd5+! xdS 32. d2+
is risky.
From M aster to G rand master
The position reached in this game af
ter Black's 1 5t h m ove occu rred nearly
a q uarter of a century later i n the game
Bologan - Anand (Dort m u n d 2 003), but
Wh ite played 1 6 . tD g6! and after 16 . . . fe8
1 7. tD xe7+ xe7 1 8 . d3 ee8 1 9 . hd1
d S 20. g1 ! b5 2 1 . d2 aS?! (21 . . . c5!)
22. tD e2 b4 23. g4 his attack was more
real.
But the pioneer of the 'impending' tDeS
g6 (which is the idea of placing the knight
on eS) was again Bel iavsky, who carried
it out in a game with Bent Larsen (Ti lburg
1 981). Black refrained from castling i n fa
vou r of 1 3 . . . a5 1 4 . he1 a4?, after which
!::, he ran i nto 1 5 . tD g6!. Nowadays this is one
of the techn ical devices i n the solving of
such positions, which Carlsen used i n the
present game. There fol lowed 15 . . .tD dS?!
(1 S . . . a3! 1 6 . b3 tD dS was correct, although
here too 1 7. tDf5! secu res White an advan
tage) 1 6. tDf5! Af8? (1 6 . . . exf5 really was
better) 1 7. Ad6 g8 1 8 . c4 tD b4 1 9. h3!
fxg6 20. xe6+ f7 2 1 . h x g 6 + xe6
22. e1 + tDeS 23. A xeS 1 -0.
13 . . . 'Wa5
I n the g a m e Bo l o g a n - D reev ( Dago
mys 2006) Black simpl ified the position:
13 . . . tD xe4 14. xe4 tD f6 1 S . e2 d S
1 6 . b1 'We4 and obtained a roughly eq ual
endgame.
14. mb1 0-0 1 5. lt\xf6+
1 5 . g4 tD xg4 1 6. tDe5 tDdxe5 1 7. dxeS ad8
1 8. W h3 with sharp play has also occu rred .
1 5 . . . lt\xf6 1 6. lt\e5 adS 1 7. 'We2
The th reat of . . . c6-c5 could also h ave
been parried by 1 7. 'W b3, but after 17 . . . W b6
1 8 . W xb6 axb6 a practically eq ual end
game is reached .
32
 M. Carlsen - S. Ernst

17. . . c5?! destroying his king's protection.

Ernst proves to be u nfami l iar with the var 20 . . g8

iation, otherwise he wou l d have played
17 . . . VN b6! (Anand - Bareev, team m atch
Russia v. Rest of the World, Moscow 2002),
where the th reat 1 8 . til g6? is parried by
18 . . . Aa3 .
If 20 . . . de8 or 20 . . . d7 there is the deci
sive 21 . xh6+! gxh6 22. A xh6 fol lowed
by g6-g7. But the rei nforcement of the
h6-pawn does not change things.

a b c d e f 9 h a b c d e f 9 h

8 .! . . 8 8 .! . . 8
7 1i .. 7 .1. 7
6 6 6 VIii 6
5 'if r--.J tb 5 5
4
.,L--.J 4 4 4

3 3 3 3

Vjj 2

g 1
2 2

1 g g 1
a b c d e f 9 h a b c d e f 9 h

21 . J.xh6! gxh6 22. xh6+! xh6

18. g6! fxg6?
And this shows a lack of elementary knowl
edge (Bologan -Anand, Dortm und 2003).
18 . . . fe8 1 9. til xe7+ xe7 20. dxc5 ed7!
was essential , when it is not easy for Wh ite
to demonstrate his su periority, si nce the
c5-pawn is regai ned .
23. Wxe7 f7 24. g xf7 g7 25. d3
d6
Mate could have been avoided by playing
25 . . . 'W b6 26. g3+ 'Wg6 27. xg6+ 'i!i> xg6,
but all the same the endgame after 28. d5
is hopeless.

19. Wxe6+ h8 20. hxg6! 26. g3+ g6 27. We5+ xf7 28. Wf5+
f6 29. Wd7 m ate
Creating a m at i n g net, s i n ce Black has
no defence against the sacrifice on h 6 , A spectacular epau lette mate!

Inspired by his bri l l iant performance in Wijk aan Zee, Carlsen decided to test his powers
in Moscow at a far higher level in the Aeroflot Open , which was then becoming increas
ingly strong. By that time in Russia the affectionate n ickname Malysh had become
firmly attached to his name, by analogy with the popular hero of the c h i l d ren's stories
Karlson on the Roof by the well-known Swed ish author Astrid Lindgen . The attention
of the experts in the press to the Norweg ian prod igy was assured .
And so, the Rossiya Hotel, the A Tou rnament, where the consistently strong field
did not afford any opportun ities for a respite. I n itially there was a set-back - in the fi rst

33
 C h a pter 1
From M aster to Grandmaster

round Magnus lost to Shu lskis, after overlooking a simple tactic in ti me-trou ble. But
the youngster did not lose heart and he prod uced a series of three wins, gai ned in
such dynam ic style, that it became clear: he did not i ntend to yield to successors of
the celebrated Soviet chess school.
' I will never forget the veneration', remem bers the wel l-known chess journal ist Evg
eny Atarov, 'with which Kasparov's famous trainer Alexander N i kitin produced a copy
of the Carlsen - Dol matov score sheet: "Th is is the play of a genius!"' Indeed , we don't
recall anyone defeating the experienced Sergey Dolmatov in 19 moves. The secret
lay i n Magn us's thorough preparation for a specific opponent. His open ing repertoire
i ncorporated play with Wh ite, so to speak, from either hand - 1 . e4 and 1 . d4, but for
Dolmatov he made an exception - 1 . &Df3! . And this is why. I n his computer database he
d iscovered 1 0 games with the reply 1 . . . f5 . It was here that the mine was laid - Magnus
prepared a new plan in what was a seem ingly harmless variation for Black.
There was a past history to the events which occu rred i n this game. The position
arising after Wh ite's 8th move occu rred in the game Lerner- Bareev from the 1 986
USSR Championship Prem ier League in Kiev, where Dolmatov was also playi ng. Evgeny
Bareev recal led (in a conversation with Maxim Notki n , the notable chess com menta
tor) that when this position was reached he became i l l at ease, and, recog nising the
dangers facing Black, he managed to find a move order with which he did not lose
(8 . . . c6! 9. A d3 A xd3 1 0 . '!W xd3 &Df6 1 1 . 0-0 Ae7 1 2 . &D e2 d5). The confident defence
found by Bareev, who did not allow the bishop to come out to c4, lul led the vigi lance
of Dutch players, who perm itted themselves a seem ingly insign ificant transposition of
moves, and this was noticed by Carlsen .

.I .i. 'iW .i. l.&\ .I

Game 4 a b c d e f g h

M. Carlsen - S. Dolmatov 8 8

Moscow 2004 7 .t .t .t .t .t 7
Reti Opening [A04] 6
1.1\ .t 6

1 . tilf3 f5 2. d3!? 5 .t 5

4 4

t:jj t:jj
A kind of anti-Dutch variation, which mainly
3 3

2 2
has a psycholog ical effect.

1 1:i ii, VJg ii, 1:i 1

2 . d6 3. e4 e5 4. tilc3
. .

The position has acquired contours typical a b c d e f g h

of the open games.

5. exf5! xf5 6. d4 tilxd4
4 . . . tilc6

Evgeny Bareev evaluates this position as Now if 6 . . .e4 there is the unpleasant 7. d5!.
favourable for Wh ite, and thinks that in this
set- up Black should play 4 . . . &Df6. 7. til x d 4 e x d 4 8. 'Wxd4 tilf6?

34
 M. Carlsen - E. Shapos h n i kov
-------

The capture on c2 is dangerous for Black,
who is clearly beh ind in development, but
8 . . . c6 is stronger, hindering the develop
ment of the bishop at c4.
9. i.c4!
Far more energetic than 9. A d 3 A xd3
10. W xd3 c 6 1 1 . 0 0 A e7 1 2 . tti e4 tti xe4
13. Wxe4 0-0 with equal ity, Romanishin
Malaniuk (Tal linn 1 987).
9 . . c6 1 0. Ag5!
.
cxd5 14. A xf6 g xf6 1 5. Wf4 or 12 . . .Wd6
13. e5 Ae6 14. A xf6 gxf6 1 5. xe6! W xe6
1 6. tLl xdS are convi ncing enough.
1 1 . J.b3 J.e7?
The decisive mistake. Black's position was
still hang ing by a thread : 1 1 . . . We7+! 1 2. f1
(or 1 2. d2 0-0-0) 1 2 . . . 0-0-0 1 3 . a4 b4!
1 4 . W xb4 d5, and at the cost of a pawn
he would have avoided an im mediate loss.
1 2 . 0-0-0 Wd7 13. !a he1 d8

The position is an open one, and, as is a b c d e f 9 h

.! 8
customary i n such a situation, Carlsen

iV ..t 7
develops his pieces i n their most active
positions.
6 6

.t Jt
a b c d e f 9 h
5 5
'ii' e ..t .! 8
7
6 J!a\ 6

5 ..t Jt 5

a b c d e f 9 h

This position recalls the times of the ro

mantic 1 9th century. Simi lar positions can
a b c d e f 9 h be fou n d in the games of Pau l M o rphy
and Adolf Andersse n . 'There fol l ows a
1 0 . . b5
.
combi nation, as natural as the smile of a
1 3-year-old youth' ( M . N otkin).
This move does not change the assess
ment of the position as dangerous for 1 4. !a xe7! Wxe7 1 5. Wf4 i.d7 1 6 . tbe4
Black. If 1 0 . . . d5 there follows 1 1 . 0-0-0!, d5 1 7. tbxf6 h6 18 . .lh4 g5 1 9. Wd4
and it only remains to include the second Black resig ned . An impressive rout!
rook in the play, after which the king will
be doomed . Especially sinc e the bishop
on c4 is i n d i rectly defended: 1 1 . . . d xc4 Game S
12. We5+ or 1 1 . . . Ae7 1 2 . he1 0-0 (1 2 . . .
M. Carlsen - E. Shaposhnikov
dxc4 1 3 . A xf6) 1 3 . We5. The king can be
Moscow 2004
moved off the dangerous file - 1 1 . . .c!? f7,
Nimzo-/ndian Defence [ E38}
although here too after 1 2 . he1 Black
is lost. The variations 12 . . . h6 13. tLl xdS! 1 . d4 tbf6 2. c4 e6 3. tbc3 i.b4 4. Wc2

35
 Chapter 1
From M aster to G randmaster
This q ueen m ove, which has the aim of
keeping the q ueenside pawn structure in
tact, is one of the most common reactions
to the N i mzo- l nd ian Defence.
4 . . . c5
Other popular replies are 4 . . . 0-0 and 4 . . . dS.
5. dxc5 AxeS
S . . . tb a6 6. a3 A xc3+ 7. 'l!V xc3 tL! xcS is
also played , but most often Black repl ies
S . . . 0-0 and captures on cS with his bishop
only after 6. a3.
6. tDf3
It is curious that in the event of 6. Af4 the
p l ay m ay u n ex pected ly transpose i nto
a l i ne of the c l assical Queen's Gambit:
6 . . . 0-0 7. tLlf3 tLl c6 8. e3 dS 9. !!d1 'WaS
1 0. a3 etc.
6 . 'Wb6
Black p rovokes e2-e3 , to p revent the
development of the bishop at g S or f4.
However, as the further course of the game
will show, it is just as wel l placed on the
long d iagonal .
7. e3 a6
a b c d e f g h
Of course, Black's queen is obviously m is-
36
placed , and so he prepares to retreat it to
the more appropriate square c7, by taking
control of bS.
8. b3 'Wc7 9. Ab2 b6 1 0. Ae2 Ab7
1 1 . g4!?
As usual, Carlsen resolutely tries to exploit
the opponent's slightly retarded develop
ment, especially as i n the previous round
he had gained a powerful charge of positive
emotions. The alternative was the q u iet
1 1 . 0-0, but in this case Black can beg in
rather unpleasant (by h u man standards)
' Paulsen-like' cou nterplay with 1 1 . . . tLl g4!
(there is also another, more solid reply -
1 1 . . . Ae7!? with the i ntention of setting up
a ' hedgehog' by . . . d7-d6 and . . . tLl bd7).
1 1 . !!d1 !?, recommended by Vitali Golod,
deserves consideration.
11 . . tD xg4
The sacrifice has to be accepted , as oth
erwise Wh ite's play is too simple: g4-gS ,
tLl e4, 0-0-0, !! d g 1 a n d s o on.
1 2 . lag1 tD x h 2 1 3. tDgS!
For what has White given u p two pawns?
All his pieces are in play, apart from the
rook on a1 , which needs only one move
to be incl uded . B lack, on the other hand,
has to solve the problems of both his un
castled king, and, no less i m portant, the
knight on h2 which is cut off from its main
forces. The fearless Rybka assesses the
position as eq ual, but in practice it is far
more pleasant playing Wh ite.
13 . . . tDc6! 14. 0-0-0
Th i n g s are u n clear after the tem pti n g
14. tLl xf7!? xf7 1 S . !!xg7+ xg7 1 6. tLldS+
WeS! - Black obtains too much material for
the q ueen . For example, 1 7. A xeS+ tLi xeS
 M. Carlsen E. Shaposh n i kov

18. b2 A d6 1 9 . tD xb6 A b4+ 20. d 1 1 5. ce4 0-0-0 1 6. xc5 bxc5

tD hf3 21 . tD xa8 l;!.xa8 with chances for both
sides. 1 6 . . . xc5 is bad in view of 1 7. tDe4! A xe4
a b c d e f g h
(otherwise 1 8 . tD d6+) 1 8 . xe4, and the
J. J. 8 knight on h2 is lost.

.i. 'it' .t.

8

7 1. 1. 7
.t. .t. .t.
1 7. f4

.i.
6 6
1 7. tD e4! b8 1 8 . l;!.xg7 was even better,
5 tjj 5
with a very strong i n itiative. After the move
4 [j, 4 in the game, strangely enough, White does
3 [j, tjj [j, 3 not have a forced win.
2 [j, VJii [j, 2 1 7. . . ef3
a a
a b c d e f g h
a b c d e f g h

1 !.1 I!.
.i. 'it' .t. :t. l.t. .t. 7
8 8
14 e5?

.t. .t.
. . .

l .t.
This natural move, defend ing the f7-pawn 6 6
and prepari n g q ueenside castl i n g , is a 5
tjj 5
mistake. I n our computer age, concrete
variations in chess increasingly take prec
4 [j, [j, 4

3 [j, [j, 3

2 [j, VJil 2
edence over general considerations. Very
many players (especially of the older gener
ation) find it hard to accept this, especially 1 a a 1
a b c d e f g h
since the variations are often such that they
can hardly been found at the board when
there is restricted time for thought. 1 8. xf7?!
Only 1 4 . . . d5!! - a solution which at first
What could be more natural than this cap
sight looks l i ke an attempt to exti nguish
ture? However, the most obvious conti nu
a fi re with petro l , and therefore without
ation again proves to be not the strongest.
analysis will be rejected by 80% of play
It was preferable to play 1 8 . tD xf3!? tD xf3
ers - enables Black to maintain a dynam ic
1 9. l;!.xg7 J;!.df8 20. J;!.f1 ! (the m ost accu
balance. Here are these variations, which
rate; there is no clear win after 20. l;!.xh7
conclude with the assessment ' u nclear' :
J;!. h g 8 ! , w h i le if 2 0 . J;!.g3 Black i s saved
a) 1 5 . A h5 g6 1 6 . tD xd5 exd5 1 7. A xh8
by 20 . . .tD d4! 2 1 . exd4? xf4+) 20 . . . tD h4
gxh5 18. xh7 0-0-0;
2 1 . J;!.g4 tD f5 2 2 . A x h 8 f! x h 8 23. c3.
b) 15. tD xd5 exd5 1 6 . A xg7 l;!.g8 1 7. xh7
White succeeds i n transforming his ini
0-0-0;
tiative i nto something more tang ible, and
c) 1 5. cxd5 tD b4 1 6 . e4 0-0-0;
he should g radually convert his exchange
d) 1 5. tD a4 tD b4! 1 6. d2 0-0-0 1 7. tD xc5
advantage.
bxc5 1 8 . A xg7 d xc4 1 9. A x h 8 l;!.xd2
20. l;!.xd2 as. 18 . . . xg1 1 9. !;lxg1 f3 20. !;lxg7?!

37
 C hapter 1
From Master to G randmaster

The midd legame is of a purely calculating 24. 'Wg6

nature, and is not easy to annotate. The
i m petuous rook capture could have lost
Wh ite the g reater part of his advantage,
which wou ld have been retained by 20. gf1
tD h4 21 . tD xh8 gxh8 22. A xg7 gea 23. A h5
(or 23. x h7). After 20. gg3 Black is again the endgame.
saved by 20 . . .tD d4! .
20 . . J ihg8 21 . xd8
Wh ite should h ave considered 24. gf7!?
g x h7 2 5 . g x f8+ c7 26. g h a g x h 8
27. A xh8 A e4, when he has q u ite good
chances of converting his extra pawn in
24 . . . 1ih1+ 25. b2 !ig1 26 . .tg5
a b c d e f 9 h

a b c d e f 9 h 8 'if 8

. QJ !. 1:(
.i. 'it' .t J:( .t
8 8 7 7

VJH
.t .t
7 7 6

.t
6 6 5 5

5 5 4 [3J 4

4 [3J [3J 4 3 [3J 3

3 [3J [3J 3 2

2 [3J V/if 2 !.

!.--- a b c d e f 9 h

a b c d e f 9 h
26 . . . !ixg5?

21 . . . 'Wxd8?
B l ac k m i sses h i s c h a n c e : 2 1 . . , g xg7!
22. A xg7 xd8, and if 23. xh7 there is
the u n p l easant 23 . . . aS! . N ow Wh ite's
position is again almost won .
2 2 . .tf6! 'Wf8 23. Wxh7 !i h8!
23 . . . g xg7 is hopeless: 24. A xg7 f7 (if
24 . . . e8 25. A c3, and the detached knight
is soon won) 25. h8+ c7 26. Af6 A c6
27. dB+ b7 28. A xf3 A xf3 29. Ae7! or
29. A e5.
S haposh n i kov a p p arently despai rs of
com peting with h i s you n g opponent i n
the calculation of variations and he does
not notice 26 . . . tD h4! 27. f6 (if 27. h7,
then 27 . . . A e4! 28. gf7 Axh7 29. gxf8+
c7) 27. . . xf6+ 28. A xf6 ge1 ! 29. A g4
tD g2 - thanks to his active rook Black sti l l
has saving chances.
27. fxg5 e5 28. 'Wf6 'Wd6 29. !ig8+
c7 30. c3!
Black resigned.

Carlsen played the second half of the tournament with alternating success, but never
theless, by scori ng 5% poi nts in 9 rounds, for the second time in succession (after Wij k
aan Zee) he achieved the g randmaster norm. He was developing l iterally before one's
eyes, and a d i rect acq uai ntance with the Norweg ian prod igy allowed certain observa
tions to be made. In particular, pleasantly surprising was his modest behaviour and

38
 M. Carlsen - G. Kasparov

the amazing com posure with which Magnus conducted his games with experienced
grand masters. I n his style of play one senses a classical approach to the game. He
com petently deploys his pieces, displaying his read iness to engage i n a large-scale
strategic battle. At the same time, his age often tel ls. Magnus is not averse to taki ng
tactical decisions, and risky sacrifices of material occur for the sake of maintaining the
initiative. Also, he constantly keeps the enemy king in view, and he displays exceptional
imag ination when attacking it. And all this is supported by wonderfu l calculating abil ity,
not only with speed of thinking, but also depth of penetration into the position.
A surprising thing was Carlsen's striving for complete independence in his work on
chess: 'I l i ke to prepare on my own , and I don't understand how some outsider can
impose on me his choice of open ing or his vision of the fight.' A unique view for such a
young age. At the same time he spoke with great respect about his fi rst trainer Simen
Agdestein, who responded i n simi lar terms (Agdestein wrote about Magnus the book
Wonderboy, publ ished in 2004 by New in Chess).
After his successfu l performance in Moscow, Carlsen played less wel l at the trad i
tional tournament in Reykjavik, where he scored only 50%. But the main result of the
visit to Reykjavik was Carlsen's meeting with Garry Kasparov, who had been i nvited to
take part in a bl itz and rapid tournament. Garry easily beat Magnus at bl itz, but in the
rapid tournament, held on the knock-out system , an embarrassment almost occu rred .
In the first game, after playi ng the open ing su perficial ly, the strongest player i n the
world was thoroughly outplayed by the youngster, and it was only a lack of experience
that prevented the latter from winning the game. In the second game Kasparov was
extremely com posed and he won the m i n i -match, but the impression remained and
it had im portant consequences for Carlsen. When a year later Kasparov gave u p top
level chess, he decided to share his experience with the one he considered the most
worthy. His choice fel l on the strongest young player from the West .

M. Carlsen - G. Kasparov risky for Black) 21 . x h 2 x e4+

22. Ae5 d6
1 . d4 d5 2. c4 c6 3. f3 f6 4. c3 e6
5. Ag5 bd7 6. e3 'WaS 7. d2 Ab4 a b c d e
8. 'Wc2 0-0 9. Ae2 e5 1 0. 0-0 exd4 a !. ..t 8
1 1 . b3 'Wb6 7 .\ 'if 7

11 .. .'c7 was much better. Now somehow 6 .\

imperceptibly Garry sl ides into a positional 5 .t 5
mire. 4 B B 4

1 2 . exd4 dxc4 1 3 . Axc4 a5 14. a4

3
ttJ B 3

Wc7 1 5. ae1 h 6 1 6. Ah4 Ad6 1 7. h3

2 B VJJ! B B 2

b6 1 8. A xf6 xc4 1 9. e4 Ah2+ ::

20. h1 d6 (20 . . . b5 2 1 . tLl bc5 was a b c d e f 9 h

39
 C hapter 1
From Master to G randmaster

23. 'Wc5 32. 'Wg4

White acts as though in the Russian say As it is fashionable to say nowadays, up to

ing: 'You can't have too much of a good
thing', but here it does not apply. Although
Black's position is sti l l u nenviable - it is not
easy for h i m to complete his development
under fi re by the opponent - Kasparov is
g iven a respite.
The i m mediate 23. d5! would have set
Garry serious problems.
23 . . . d8 24. d5
It was also possible to transpose i nto a
favou rable endgame: 24. A xd6 xd6+
25. xd6 xd6 26. e8+ h7 27. fe1
b5 28.1 e7 bxa4 29. &D c5 xd4 30. xf7.
24 . . . 'Wd7 25. ttld4 ttlf5 26. dxc6 bxc6
27. til xc6 e8 28. d1 'We6 29. fe1
.tb7 30. ttld4 etlxd4 31 . 'Wxd4 'Wg6
here Magnus has played like Fritz or Rybka.
But the problem with com puter programs
is that they often evaluate endings with
an extra pawn as though their 'brains' are
switched off. They fol l ow the principle: I
have a pawn , so there is no need to think!
And although, com pared with other drawn
positions, here there are opposite-colour
bishops, not a single computer g ives a
drawn assessment.
Winning chances would have been re
tained by 32. f3 , removing the threat from
the bishop on b7, whereas Wh ite's dark
square bishop can attack both along the
long diagonal, and - especially i m portant
- the a5-pawn .
32 . . .'Wxg4 33. hxg4 .tc6 34. b3 f6
35 .tc3 xe1 36. xe1 .td5 37. b1

a b c d e f g h
a J.
'i!lf7 38. 'i!lg3 b8 39. b4 axb4
I.
.i.
8
40 . .txb4 .tc4 41 . a5 .ta6 42. f3 'i!lg6
7 7 43. 'i!lf4 h5 44. gxh5+ 'i!lxh5 45. h 1 +
6 'ti' . 6 'i!lg6 4 6. .tc5 b2 47. 'i!l g 3 a2
5 . Jt 5 48. Ab6 'i!lt7 49. c1 g5 50. c7+ 'i!lg6

4 [j, 4
51 . c6 .tf1 52. Af2

3 [j, 3 Draw.
2 [j, [j, [j, cJ;; 2

1 .: .: 1
a b c d e f g h

A month later, at the Open in Du bai , Magus achieved the g randmaster norm for the
third time (61/z poi nts from 9 rounds), and then he confirmed his g rowi ng class in
Scand i n avia at the Sigeman Tou rnament (FI D E category 1 3). The tournament was
held in Malmo and Copenhagen, but whereas in the Swed ish part Magnus's play was
rather restrained (21/z poi nts from 5 rounds), in Denmark he defeated three g rand
masters and fi nished i n 3 rd place with 51/z poi nts out of 9, once again ach ievi ng the
g randmaster norm .

40
 M . Carlsen - E. Vlad i m i rov

The 1 3-year-old Norweg ian's achievements were recognised by his inclusion i n the
World Championsh i p i n Li bya, held on the knock-out system i n May 2004. However,
Carlsen's appearance went practically unnoticed, since in the very fi rst round he lost
to the rapidly progressing Levon Aronian. But it should be mentioned that the latter
was able to overcome his young opponent only in the second rapid tie-break game, in
which Magnus lost a theoretically d rawn rook end ing with f- and h-pawns.
After the set-back in Libya, Carlsen played without interruption, one tournament after
another, and his rating also improved . By the time of the Olympiad in Calvia (Mallorca)
the youngster had raised it by another hundred points and reached 2581 , finally catch
ing his more successful contemporary Sergey Karjakin. I n the Norwegian team Magnus
immediately made his debut on board 1 , which was conceded to him by the permanent
team leader, grandmaster Agdestein. Despite his higher rating, M ag n us's trainer did
not hesitate - the future belonged to him. And Carlsen did not let him down. He scored
3% points from 5 games, and the Norwegian team fin ished in a respectable 31 st place.
These successes marked an i m portant watershed . The time of Swiss ' l otteries'
was at an end . At the age of 1 4 , with such a high rating, M agnus could hope for more
favourable i nvitations from tournament organisers, and the opportun ity presented itself
to plan a sched ule of performances. After the protracted series of official com petitions,
which had taken much strength , now was the time from him to rest and assess his
capabil ities. But sig ned contracts have to be fulfi l led . And after the Olympiad came a
considerable slump: an ind ifferent performance at the four-player Hoogeveen Essent
Crown tournament in Hol land and a complete fai l u re in the Spanish Team Cham pion
ship (four losses and five d raws). Such a d ismal finish cou l d hardly have improved his
mood before the New Year. But how could he reject taking part in a FIDE category 1 4
tournament i n his homeland?

Game 6 a b c d e f 9 h
.i 'it' .i.. .i
7 .t. .t. "' .t. .t. "'.t. .t. .t. 7
M. Carlsen - E. Vladi m i rov 8 8

Dubai 2004 ..t

Sicilian Defence [ 867] 6 6

1 . e4 c5 2. tllf3 tll c 6 3. d4 cxd4 5 5

4. tll x d4 tllf6 5. tll c 3 d6 6. g5 e6 4 4

7. Wd2 a6 8. 0-0-0 .td7 9. f3 3 3

(see next diagram)

9 . .c8 a b c d e f 9 h

Apart from the move in the game, B lack i s poss i b l e : 9 . . . tD x d 4 1 0 . x d 4 A e7

has a wide choice of alternatives: 9 . . . A e7, 1 1 . g4 A c6 1 2 . A e3 0 - 0 1 3 . g 5 tb d7
9 . . b5 or 9 . . .tb xd4. For example, i n the last
. 1 4 . h4 b5 1 5. b1 b4 1 6. tb e2 d5 1 7. exd5
case the fol lowing development of events A xd5 1 8 . f4 .c8 with good cou nter-

41
 Chapter 1
From M aster to G ran d m aster

p l ay fo r Black (Ta l - Balashov, Tal l inn
1 973).
1 0. J.e3
1 0. g4 is also possible: 1 0 . . . Ae7 1 1 . Ae3
tD e5 (1 1 . . . h 6 1 2. h4 transposes into the
game) 1 2. g5 tD h5 with double-edged play.
10 . . e5
Another development of events involves the
exchange of knig hts: 10 . . .tD xd4 1 1 . A xd4
b5 1 2 . g4 Ac6 1 3 . g5 tD d7 with pawn at
tacks on opposite wings.
A xc2 26. tDf6+ h8 27. tDe8 'Wc6 28. 'We5+
g8 29. d6?? (29. tDf6+ h8 30. tD e8+
wou ld have forced a d raw) 29 . . . A h6+ 0-1 .
1 7. b1 xf3?!
H ow should this m ove be assessed? It
looks very attractive, s ince the knig ht
capt u re 1 8 . tD xf3? m ay lead to m ate:
1 8 . . . tD c3+! 1 9. bxc3 bxc3 20. 'Wc1 A a3,
but Black's i l l usions are soon dispelled . It
would have been more logical to settle on
1 7. . .tD xd3 1 8. 'W xd3 tD xe3 1 9. W xe3 Ae7,
with an inferior but defensible position.

1 1 . g4 h6 1 2 . h4 b5 1 3. Ad3 b4 1 8. 'Wf2 xe3 1 9. 'Wxe3 xd4

14. ce2 d5 1 5. exd5 xd5 1 6. f4
If 19 . . . Ac5 there follows 20. 'W xf3 A xd4
b c d e f h
21 . tD g6!, and the attempt to play for mate
a 9

.!. 'if i. .!.

- 21 . . . e5!? 22. tD xh8 Ae6 - is parried by

.t. .t.
8 8

i.
23. b3 'Wa3 24. A b5+ axb5 25. xd4 exd4

.t. .t.
7 7
26. e1 'Wa7 27. tD g6 .
6

1.& 1.&
5 5
20. 'Wxd4 Wc5

4 4
tij
3
il, il,
ttJ 3
8
2 WJ 2

g g 1
7

a b c d e f 9 h

16 'Wa5

A p p a rently, to the experienced Evg

eny Vlad i m irov, Black's ach ievements in
Gazi k - Popovic (Stary Smokovec 1 99 1 )
seemed inadequate. That game continued
1 6 . . .tD xd3+ 1 7. 'W xd3 tD xe3 1 8. 'W xe3 Ac5
1 9. tD h5 0-0 20. g5 We? 2 1 . d2 (21 . tDf6+
gxf6 22. gxf6 h7 23. 'We4+ h8 24. 'We3
leads to a curious d raw) 21 . . . A a4 22. gxh6
A xc2? (22 . . . g6!? 23. b1 fd8 was cor
rect) 23. hxg7 A g6 24. g xf8'W+ (24. d1 !
A xd4 25. gxf8'W+ xf8 26. 'W xd4 would
have won m aterial) 24 . . . A xf8+ 25. tD c2
a b c d e f 9 h
21 . 'We4
Black is beh ind i n development and Wh ite
builds up the threats (if 21 . . . Ae7 there fol
lows 22. tD h5), completely rul ing out the
possibility of the black king castling. At first
sight the q ueen move seems i m possible
d ue to 2 1 . . . Ac6 22. 'We2 A xh1, but it is

42
 D. Palo - M. Carlsen

.i. .i. .I
here that a 'mine' has been laid: 23. CL!xe6! a b c d e f 9 h

.I
W/e7 (23 . . .fxe6 24. xe6+ leads to mate: 8 8
24 . . .<;t>dB 25. Ae4+) 24. Af5! ! and there is no
7 7
defence against the mate by 25. !!dB+ !::! x dB
6 'iiV 6
26. tile?#. A pretty mate also awaits B lack
after 21 . . . Ae7 22. Cbh5 Ac6 23. CL!xg7+ fB
5 {jj 5
24. /i}xe6+ fxe6 25. !::! h f1 + eB 26. xc6+ /j /j 4
Wfxc6 27. Ag6#! 3 3
2 [j [j [j fH 2
:g :g
21 . . .'Wc6 22. 'We2 aS

.I
a b c d e f 9 h a b c d e f 9 h

8 8
26. 'Wd4!
7 7

6 6 The double attack (27. A b5 with mate at

5 dB is threatened , as wel l as the captu re on
4 {jj [j [j 4
g7) is irresistible.

3 3 26 . . . d7 27. xg7+ .1xg7 28. 'Wxg7

2 /j /j /j VJ!i 2 f8 29. 'Wxh6

:g :g 1 Wh ite is now a pawn u p with a conti n u i ng

a b c d e f 9 h
attack on the king. The game is decided .

Black has an extra pawn and t h e two 29 . . . b3

bishops: what else, it wou l d see m , can
If 29 . . . a3 , then 30. g5 with the th reat of
he want? But the main thing is the insecure
31 . A b5 is decisive, since after 30 . . . A a6
position of his king. If 22 . . . A d6 there fol
31 . A xa6 xa6 there follows 32. f4, as
lows 23 . Cbh5 0-0 24. g5 with a dangerous
occurred in the game.
attack.
30. axb3 axb3 31 . cxb3 .ta6 32 . .txa6
23. h5 a4 24. hf1 ! c7
'Wxa6 33. 'Wf4! a7
Black is obl iged to forg et about c o u n
If 33 . . . b5, then 34. c4 is an adequate
terplay: 2 4 . . . b3 25. c x b3 axb3 26. f2
reply.
bxa2+ 27. a1 and he has to run with his
king - 27. . . d8 (27. . . f6? 2B. CL!xf6+!), al 34. 'Wb8+ e7 35. 'Wb4+
lowing 2B. xf7. Therefore he defends the
f7- pawn with his rook, but Wh ite stri kes a Black resig ned .
blow at its neighbour. I n the event of the
defence by the q ueen 24 . . . c5 25. f3
Game ?
Wie7 there fol l ows a blow from the left -
D. Palo M. Carlsen
26. b7.
-

Gausdal 2004
25. 'Wf2 .tea Nimzo-lndian Defence [E32]

43
 C hapter 1
From Master to G ran d m aster

1 . d4 f6 2. c4 e6 3. c3 Ab4 4. Vc2 1 2 cxd4 13. Vxd4 c5 14. Ac2 e5

0-0 5. a3 .Q.xc3+ 6. 'Wxc3 b6 7. .Q.g5 1 5. 'Wf2

Another plan i nvolves the development a b c d e f g h

of the bishop at b2: 7. !Df3 A b7 8 . e3 d6 8 .I 'ii' .l. 8

9. Ae2 !D bd7 1 0. O-O !De4 1 1 . c2 f5 1 2. b4 7 .t. .t ... ... ... 7
etc. 6 ... ... "' 6

7 J.b7 5 "' ... 5

::, ::,

4 4
a b c d e f g h
::, ::,
i. "' 'if I. 8
3 3

fj, {jj WJ [j, [j, 2

8

7 .f. ..t .t. .t. 1. 1. 1. 7

2

l:t l:t
6 ... "' 6 a b c d e f g h
5
5
4 1 5 . . . g4!
3
This tactical trick enables Black to solve
his opening problems. 1 5 . . .!Dfxe4? did not
work because of 1 6. !D xe4 !D xe4 1 7. A xe4.
a b c d e f g h

8. f3
At the present time fashion gives prefer
ence to 8. e3 d6 9. !D e2 !D bd7 1 0. c2.
For exam ple: 10 . . . h6 1 1 . A h4 c5 1 2 . d1
c7 1 3 . !D c3 a6 1 4 . d2 adS 1 5 . f3 d5
1 6. cxd5 exd5 1 7. Ad3 fe8 1 8 . 0-0 c6
1 9. Ac2, and one can speak of some ad
vantage for White (Dreev - Korotylev, Mos
cow 2009).
8 . . d6 9. d2
The simple 9. e3 !D bd7 1 0. A d3! is also
possible, since the exchange on f3 can only
favour White, who gains the advantage of
the two bishops against two knights and the
half-open g-file for an attack. 1 0. Ae2 c8
1 1 . 0-0?! is less good because of 1 1 . . .!De4! .
9 . bd7 1 0. f 3 c5 1 1 . e 4 Ile8 1 2 . J.d3
1 2 . Ae2 deserves consideration.
1 6. Axd8?!
1 6. h4! f6! 1 7. xg4 fxg5 1 8. 0-0-0 was
stronger, with an unclear game. The Danish
player probably underesti mated the dan
gers awaiting him in the endgame, which
is by no means so harmless for White as
it may seem .
16 . . xf2 1 7. xf2 Ilexd8
A close examination of the position reveals
that the weak d6-pawn is easily defend
ed , and the possession of the d 5 - point
does not play any g reat role. At the same
ti me, Wh ite's light-square bishop is bad ,
and the dark squares in his position are
weakened .
1 8. b4
In the event of 1 8. hd1 !D e6 the black
k n i g ht wou l d have reached d4, w h i c h ,
however, wou ld hardly b e fatal for White.

44
 D. Palo - M. Carlsen

The move in the game is more active, but 27. Eid1 + c7 28. Ei xd8 Eixd8 29. e3
also more committing - the c4- pawn is Eid4 30. Ad3 Ad7 31 . Etb1
weakened .
If 31 . a2!? Carlsen would probably have
18 . . . li:le6 1 9. Ethc1 f8 20. li:lb3 e7
played 31 . . . g6 32. b2 f5 . The difference
21 . Ad3 Ac6
com pared with the game continuation was
that the 6th ran k wou l d rem a i n c l osed ,
a b c d e f 9 h
a J. I. a
and B lack wou ld be denied the possibility

7 . 7
of . . . d6-g6-h6, which could have g iven

6 -*- 6
Wh ite add itional chances.

5 5 a b c d e f 9 h

4 :::, :::, :::, 4

8 8

3 ::, /!iJ
-*-
:::,
7

6 6
3

2 rJ;; !:::, :::, 2 5 . 5

1 .a.

.a. 4 :::, 4
a b c d e f h :::, 1. :::,
:::,
9
3 3

22. a4? 2 :::, :::, 2

This merely agg ravates Wh ite's difficu lties. a h
b c d e f 9
The accu rate 22. tD d2 was better.

22 . . . a5! 23. bxa5 tileS! 31 . . .f5! 32. Eic1

Apparently Palo was only rec ko n i n g on It was not possible to captu re on f5 be

23 . . . bxa5 24. c5! , when the position be cause of 32. exf5 xd3+! 33. 'i!;>xd3 A xf5+,
comes equal . Now another wh ite pawn is but 32. g3 g5 (32 . . . d6 33. A c2) 33. g1
fixed on a light square. was i nteresting.

24. Ac2 32 . . . f4+ 33. e2 Eid6 34. Eib1 Eih6!

After 24. tD xc5 both captures wou ld have It is usefu l to provoke h2-h3, after which
led to a big advantage for Black. Wh ite's pawns are on squares of the colour
of his bishop.
24 . . . bxa5 25. Eicb1 d7 26. tilxc5+
dxc5 35. h3 Etg6!

The position can be assessed as 'between Carlsen demonstrates excellent technique

a win for Black and a d raw'. It is i m portant
only for Black to retain at least one pair of
rooks, to avoid the creation of a fortress. I n
practice the stronger side i s often success
ful in such cases, since it is very difficult
to defend passively.
- it is also not out of place to draw the king
away from the centre.
36. f2 Etd6 37. e2 g5 38. Ac2 h5
39. Eid1 Eid4
It wou l d h ave been a b l u n de r to p l ay

45
 Ch apter 1
From Master to G ran d m aster
__________________.____ ------

39 . . J !xd1 ? 40. xd1 g4 41 . h4, when there a b c d e f g h

is nowhere for B lack to break through. 8 .!. 8

40 . .td3 7 ..t 7

6 6

.l .l .l .l
40. l:!xd4 cxd4 followed by the king march
5 5
to c5 was also hopeless.
4 .l 4

40 .. J;td6 41 . .tc2 9:d4 42 . .td3 .tc6
3 3

The open ing of a second front comes into 2 2

Black's plans, but for the moment 42 . . . g4
is premature because of 43. hxg4 hxg4
4 4 . l:!h1 with the activation of the rook
(44 . . . A x a4 45. l:!a1 ).
43. 9:g1 9:d8 44. g3!?
If 44. l:! b1 , then 44 . . . A x a4! (44 . . . l:!h8 is
less good : 45. A c2 g4 46. h4! and Wh ite
has exc e l lent c h ances of setti n g up a
fortress) 4 5 . l:! a1 A c 6 46. l:! x a5 b6!
( b u t not 4 6 . . . l:!a8 47. l:! x a8 (47. l:!xc5
d 6) 47. . . A xa8 48. d2 b6 49. c3
a5 50. b3=) 47. l:!a1 l:!a8. The point
of B l ack's plan is to break through with
his king via the q u eenside: 48. l:!xa8 (af
ter 48. l:! b1 + c7 Wh ite is probably lost)
4 8 . . . A x a8 4 9 . d 2 a5 50. c3 a4
51 . A c2+ a3 52. A b1 A c6 53. A c2 A a4
54. A d3 a2! and wins.
a b c d e f g h
and, despite h i s two extra pawns, B lack
sti l l has to overcome some techn ical d if
ficulties.
47. . . 9:g7 48 . .te4 .txh3
As usually happens when converting an
advantage, at the critical moment one has
to switch from the systematic strengthen
ing of the position to the calcu l ation of
specific variations.
49. a:h1 g4 so. a:b1 a:9s s1 . a: b7+ ct>da
52. e6 a:es
The e-pawn has been stopped , and Wh ite
is u na b l e to cope with the c o n nected
passed pawns.

44 . . . 9:g8 45. gxf4 exf4 46. e5! 53. 9:f7 g3 54. 9:xf4 9:g5 55 . .td5
.txe6! 56. mt1 .txd5 57. cxd5 9:xd5
Wh ite deprives h i mself of his only trump
58. ct>g2 9:d4 59. 9:f5 h4 60. mh3 c4
the protected passed pawn , but this deci
sion is forced , since Black was threatening Wh ite resig ned .
. . . d6-e5 with a total bind.
46 . . . .td7
Game 8
(see next diagram) M. Carlsen - N . Short
Hoogeveen 2004
47. .th7?!
Queen 's Gambit {037]
47. A e4! was m o re res i l ient: 47. . . A xh3
48. A d 5 l:!g7 49. l:! b1 l:!e7! 50. l:!b7+ d8 1. d4 f6 2. c4 e6 3. f3 d5 4. c3
51 . l:!b8+ A c8 52. l:!a8 l:! xe5+ 53. f2 Ae7 5. Af4 0-0 6. e3 c5 7. dxc5 AxeS

46
 M . Carlsen - N. S hort

a b c d e f g h 10 . . . c6 11 . .td3 .tb6 12. 0-0

1. 1.& 1. 'if I. .
.t. .t. .t. .t. 7
8 8

7 .t.
a b c d e

: J j_ l .t. l.& :
4 i, 4

3
i,
3

2 2
4 4

1 II i, II 1
a b c d e f g h
8. cxd5
A few years later, when he had grown up
a little, at the 2008 B i l bao tournament
Carlsen chose a more agg ressive varia-
tion agai nst lvanchuk: 8 . "Wc2 tb c6 9. a3
a5 1 0 . 0-0-0 .te7 1 1 . h4 l::!. d 8 1 2 . tD d2
a6!? 1 3 . Ae2 b5! 1 4 . cxd5 exd5 1 5 . g4? ! .
Th is attack is too l ate, a n d W h ite
should have restri cted h i mself to central
strategy: 1 5 . tD b3 "Wb6 1 6 . tD d4 tD xd4
1 7. xd4, restra i n i n g B l ack's i n itiative. If
1 7 . . Ac5 there can fo l low 1 8 . A e5 A xd4
. 20. '1Wb3 A xf3 21 . .xb6 Ae2 22. .b1 A c4.
1 9 . Axd4 when the centralised bishop,
which has n o opponent, i s n o weaker
than a roo k .
The game contin ued 1 5 . . . A e6 1 6. tD b3
b6 17. g5 tD e4 1 8 . tD xe4 d xe4 1 9. .xd8+
xd8!? (1 9 . . . "Wxd8 20. b1 Wd5 21 . tD c1
c8 22. Wd 1 "Wc5 23. tD b3 'W f5 is also
possible, retai ning a dangerous i n itiative)
20. 'i!?b1 l::!. c 8 2 1 . "Wd 1 tD c6 22. h5 a5! and
Black's attack proved more real.
8 . . . CLi x d 5 9. x d 5 e x d 5 1 0. a 3
Sharper play resu lts from 1 0. A d3 A b4+
1 1 . 'i!?e2 tb c6 1 2 . "Wc2 h6 1 3 . l::!. h d1 with
strong press u re i n the centre, although
Wh ite's u n s h e ltered k i n g n evertheless
gives him some problems.
3 i, 3
2 2
1 'if II
a b c d e f g h
1 2 . . . d4
Magnus has also had occasion to play this
position with B lack. At the Amber Tourna
ment (N ice 2009) in his blindfold game with
lvanchuk he set Wh ite more com plicated
problems: 1 2 . . . A g4 1 3 . h3 A h5 1 4 . b 4 .e8
1 5 . .c1 a6 1 6 . A xa6 . x a6 1 7. b5 . x a3
1 8 . bxc6 bxc6 1 9. .xc6 .a? and Wh ite
could find nothing better than to go i nto
a position with opposite-colour bishops -
The game is level.
1 3. e4 .tg4 14. h3 .th5 1 5. g4 .tg6
1 6. d2 f6
If 1 6 . . . .e8, then 1 7. A g3 followed by f2-f4
is possi ble.
1 7. lac1
(see next diagram)
I n this variation 1 7. tD c4 A c7 1 8 . A xe?
'1Wxc7 1 9 . f4 is m o re often p l ayed . The
game Pel letier - U bi lava (Spain 2000) con
tinued: 19 . . . Af7! 20. .c1 h8 21 . b4 A xc4
22. l::!. x c4, and here instead of 22 . . . W d6
Black could have played 22 . . ."Wb6! with
strong cou nterplay i n connection with . . .
a7-a5 (Dautov).

47
 Chapter 1
From M aster to G ran d m aster

a b c d e f g h In view of the weakening of the kingside,

it was essential to cover the f6-sq uare -
24 . . . A g6!, preventing White's manoeuvre
which he carried out i n the game. Then
25. A xc6 bxc6 is pointless, since the pawn
5 5
breakthrough 26. f5 Af7 27. e5? A dS is to
Black's advantage. He also has a sound
position after 25. 'I!Ng3 'I!Nc5 or 25. '1Wd3 f!d8.

25. /ixc6 bxc6 26. fxg5 fxg5 27. YJ/f6+!

a b c d e f g h At that time Carlsen was already n oted

for his practical approach. If he sees real
1 7. . . h8 1 8. c4 /ic7 1 9. /ixc7 YJ/xc7 chances in an endgame, he is prepared to
20. f4 ae8 be satisfied with a small advantage.
a b c d e f g h
20 . . . Af7! was stronger. 8

7
21 . YJ/f3 YJ/e7 22. ce1 lif7 23. d2!
6
A subtle move, making way for the bishop. 5 5

a b c d e f g h 4 4
8 .! .! 8 3
7 .t. .t. 'ii' .S.. .t. .t. 7 2
6 I.& .t. 6

5 5 a b c d e f g h

4 l . 4

3 i., VIIi 3
27. . . g8?

2 CiJ 2
1 :: 1
a b c d e f
23 . . . g5!
Sens i b l e p ro p hylaxis. B lack sets u p a
barrier on the distant approaches to his
king. With passive play the pawn offensive
could have become very dangerous: 23 . . .
a6 24. h4! h6 25. g5! fxg5? 26. hxg5 hxg5
27. 'I!Nh3+ <;!> g8 28. e5 g6 29. e6! and Wh ite
wins the bishop (29 . . . A xe6 30. f!xe6).
Short was clearly rattled . Otherwise how
can it be explai ned that he gave up the d4-
g h
pawn, and then 'asked' to go i nto an end
game. After 27. . .'I!Nxf6 28. f!xf6 c5 29. e5!
(29. f!ef1 <;!> g7 30. f!xa6 f!b8 31 . f!a7 <;!> g8
32. e5! f!xb2 33. tb e4 A g6 is not so clear)
29 . . . f!e7 (with the idea of 30. f!xa6 f!fe8,
since otherwise W h ite w i l l p i c k up the
a6-pawn for free) 30. tbf3 .feB 31 . .d6!
a5 32. f!c1 Wh ite fixes all the opponent's
weaknesses.
28. YJ/xd4 c5 29. YJ/f6 YJ/xf6 30. xf6

24. /ibS a6? Wh ite is a pawn up with the better posi-

48
 M . Carlsen - N. Short

tion. Magnus conducts the technical phase 36. Eic1 Ei b3 37. f2 c3

impeccably.
30 J !b8 31 . f3 flxb2 32. xg5 A a2
. . The pawn capture 37. . . flxa3 38. fla1 c3
33. xf8+ xf8 34. xh7+ <tle7 35. g5 c4 39. <tle3 leads to a mortal p i n .
a b c d e f 9 h
38. e5 Etxa3 39. e4 c2 40. Ei xc2 Eixh3
8 8
41 . Eic7+ d8 42. Eia7 Ac4 43. g5 Eih4
7 7 44. d6 Ae6 45. g6
6 ' 6
Here the curtain could have been lowered.
5 iLl 5

4 fj, fj, 4 45 . . . Etg4 46. g7 a5 47. f3 Eig1

3 /j, fj, 3
48. b7+ e8 49. c5 Af7 50. e4

2 .i. l.
Eixg7 51 . d6+ f8 52. Eia8+ e7
2
53. f5+ d7 54. Eta7+

a b c d e f 9 h B lack resigned.

Black's last hope is his passed paw n . But

Magnus is vigilant.

49
 Chapter 2

The Way to the Top

The chess festival in the town of Drammen, not far from Oslo, was a notable event in
Norway. Just think, 500 players, divided into 1 0 tournaments, the main of which was
the Dram men Smartfish Masters (FI DE category 1 4), were invited to greet the New Year
2005 in a friendly atmosphere. How could their idol refuse such an invitation? Of cou rse,
M agnus played , but somehow superficial ly. In the very first round, against the Danish
g randmaster Peter Heine N ielsen , wel l-known for his high-quality open ing knowledge
(he had worked with none other than Anand), he played , so to speak, 'at sight'.

P. H. Nielsen - M. Carlsen 10 . fg4?! 1 1 . f3 g5!?

Dram men 2004/2005
But this is already his own home prepa
1 . f3 f6 2. c4 c6 3. d4 d5 4. c3 ratio n , si nce M o rozevich's cont i n u ation
dxc4 5. a4 .tf5 6. e5 bd7 7. xc4 11 . . .''We7 12. fxg4 ttid3+ 13. 'Wxd3 A xd3
'Wc7 8. g3 e5 9. dxe5 xe5 1 0 . .tf4 14. ttid6+ d7 1 5 . 0-0-0 led to serious
problems. However, the attempt to rehabili
a b c d e f 9 h
tate the idea of the piece sacrifice proves
.t .I 8 unsound .
......... 7
6 6 1 2 . fxg4 gxf4 1 3. xe5 'Wxe5 1 4. gxf5
5 .t 5 .tc5 1 5. 'Wc2
4
!::, t2J 4
By playing 1 5 . 'Wd3! N ielsen could have
3
t2J !::, 3
placed B lack in a critical position . N ow,
2
!::, !::,!::, !::, 2 however, he gains defi nite counterplay.
1: :1
a b c d e f 9 h
1 5 0-0-0 1 6 . 'We4 fxg3 17 . .tg2 'Wd6
.

In this tabiya of the Slav Defence, where 1 8. 'Wd3 'Wf4?

since the 1 930s 1 0 . . J:td8 1 1 . 'Wb3 tDfd7
has been played , Magnus su rprised the
open ing encyclopaed ist with an eccentric
idea, which shortly before he had seen in
an Internet game of Alexander Morozevich.
B lack is ready to sacrifice a knight with the
18 . . .'Wf6 was correct.
1 9. 'Wxg3 'Wd2+ 20. f1 'Wxb2 21 . J::i: b 1
'Wxb1 + 22. xb1 J::i:d 1 + 23. 'We1 J::i: h d8
24 .te4, and Wh ite converted his extra

aim of keeping the wh ite king in the centre. piece.

50
 M . Carlsen - A. Shirov

However, every cloud has a silver lining! N ielsen began taking a closer look at Magnus
(they had al ready played six months earlier in the Sigeman tournament) and he soon
become his mentor. Talking about N ielsen, it should be mentioned that the Dane had
absorbed many of the methods of the Soviet chess school. It was no accident that
Garry Kasparov named him as one of the trainers who wou ld help in the development
of Carlsen's open ing repertoire.
As regards Magnus's further play in the Dram men tournament, the casual start did
not bring any benefit. In the 3rd round he was 'punished ' for his su perficial play in the
opening by the 73-year-old veteran Victor Korchnoi. Then came opening d isasters
against Johan nessen and Macieja, which demonstrated the inadequacy of his ' black'
repertoire. The result was a share of last place with the lady world champion Antoan
eta Stefanova. Some consolation was provided by his only win, over Alexey Shirov.
True, this was not without the help of the el ite grandmaster, who failed to withstand
Magnus's 'cavalier' attack. This win allowed Peter Heine N ielsen to catch Shirov and
share 1 st-2nd places with h i m . The two Scand inavians then travelled together to the
tournament in Wij k aan Zee.

Game 9 Black prevents the d4-d5 advance, which

M. Carlsen A Shirov
-
. is possible after the automatic fianchetto
Drammen 2004/2005 of the bishop - 1 5 . . . Ag7 1 6 . d5 'We? 1 7. c4,
Ruy Lopez [ C95} when White gains a spatial advantage.

16. Ag5 Ag7

1. e4 e5 2. f3 c6 3. Ab5 a6 4. Aa4
lUf6 5. 0-0 Ae7 6. ile1 b5 7. Ab3 0-0 The preparatory 1 6 . . . h6 came into consid
8. c3 d6 9. h3 b8 eration, to d rive the bishop from its active
position .
a b c d e f 9 h

sK.i.'iV K. a 1 7. 'Wd2 'Wc7

7
1. 7 The trad itional ' S pa n i s h ' post for the
J!l\ 6
queen , but i n recent years Black has also
5 5
played 17 . . . 'We?, a keen advocate of which
4 4
is Krishnan Sasikiran. For example, one
3 JL ttJ 3 of his recent games, against Bauer in the
2 2 2008 French Team Championship, con
1 .:t: ttJJL.: @ tinued 1 8. a4 'Wf8 1 9. Ad3 d5!? 20. d xe5
a b c d e f 9 h tbxe5 21 . tDxe5 xe5 22. Af4 ee8 23. e5
tbd7 with double-edged play.
The ultra-solid Breyer Variation .
1 8 . a4
10. d4 bd7 1 1 . bd2 Ab7 1 2 . Ac2
e8 13. f1 Af8 14. g3 g6 1 5. b3 c6 (see next diagram)

51
 Chapter 2 The Way to the Top

a b c d e f 9 h a b c d e f 9 h

a .!. .!. a a .!. .!. a

7 -*-
j_'Yjfl,& 7 7 j_ 1.& -*- 7
6 '-' 6 6 . 'iV .jL 6
5 J jL 5 5 .8 5
4 8 88 4 4 8 4

3
88 tiJtiJ8 3 3
88 tiJ 8 3
2
jLfjf 8 8 2 2
jL 8 8 2

1 .: 1 .: .:
.:
a b c d e f 9 h a b c d e f 9 h

1 8 . dS
and by exploiting the p i n , Wh ite wins a
This leads to the open ing of the position , piece: 27. f4 tbc4 28. f!xe6 tbxd2 29. f!xe8
which allows Wh ite, thanks to the active f!xe8 30. f!xb7.
placing of h i s pieces, to develop an at-
24 f4 Y!/e7 2s. E!ea
tack on the kingside. 1 8 . . . f!ad8 followed
by . . .tbf8 or the i m mediate 1 8 . . . tbf8 was With the threat of 26. ttlf5 gxf5 27. f!g3+.
more sol i d , after which Carlsen was in-
2S . . . f8
tending 1 9. b4 and then b3,
trai ning his
bishop on f7. a b c d e f 9 h
---=
1 9. dxeS xeS .!. l,& . j_ a
After the complete open ing of the centre - 'iiV. 7
1 9 . . . tbxe4 20. Axe4 dxe4 21 . tbxe4 tbxe5
.jL 6
22. tbxe5 Axe5 - Black has to reckon with
23. ttlf6+ Axf6 24. Axf6, when it is not easy
for h i m to defend on the kingside.

20. xeS Y!/xeS 21 . Af4

I n the event of 21 . f4 V!lc7 22. e5 tbd7 the

bishop at g5 proves to be out of play.
26. E!f1 ?!
21 . . Y!/e6 22. eS d7 23. Ah6
.

A natural move, and an indication of the

(see next diagram)
23 . . . Ah8
The attem pt to win a pawn - 23 . . . bxa4
(23 . . .tbxe5? 24. Axg7 rilxg7 25. f4 1oses the
knig ht) 24. bxa4 tbxe5 25. Axg7 r:ilxg7 is
parried by the interposition 26. f!ab1 ! f!ab8,
deficiencies in the young g randmaster's
'schoo l i n g '. The attack should be com
bi ned with prophylaxis agai nst possible
counterplay by the opponent. It was correct
fi rst to restrict Black's play on the queen
side by 26. b4!? with the possible continu
ation 26 . . . c5 27. f!ae1 bxa4?! 28. A x a4

52
--------------------
M. C arlsen - A. Shi rov
flec8 29. Ag5!? (this is stronger than 29. f5
cxb4 30. cxb4 f!.c4) 29 . . . e6 30. bxc5.
But for the moment doubts are something
unfam i l iar to Magnus, and he lau nches an
attack on the king, indicating his read iness
to sacrifice a whole rook.
26 . . . cs 27. f5?!
This direct attack is parried by cou nterplay
in the centre, a fact ad m itted by Carlsen
hi mself, who considers the correct move
to be the prophylactic 27. f!.fe1 !?.
27. . . d4
Without the slightest doubts! The e5-pawn
is 'poisoned ' because of the pin on the bish
op- 27. . . Axe5?! 28. Axf8 c7 (28 . . . f!.xf8?
loses to the i nterposition 29. f6, casting
a mating net around the king: 29 . . .d6
30. gxe5 xe5 31 . h6). But now White
forces the transition i nto a clearly favour
able endgame: 29. f!.xe5 xe5 30. h6
e3+ 31 . h2 xh6 32. Axh6 (variation
by Carlsen).
28. cxd4 cxd4 29. fxg6?!
A continuation of the previously calculated
line. It was not yet too late to 'call a halt' -
29. xd4 Axe5.
29 . . . hxg6?!
The energetic actions of his young oppo
nent so impressed the experienced fig hter
Shirov, who h i mself had gained nu merous
wins with a direct attack on the king, that
he fl inched and could not bring hi mself to
go in for the c ritical continuation 29 . . . dxe3!
30. gxh7+ tDxh7 31 . e2 .
53
------------------
ttJ
.I .I .a
1. 'if. ""
A
8
[::,
[::, l!jjl::,
A V!;V [::,
1:1
Analysis diagram
However, the th reat of 32. g4+ can be
parried, and Wh ite does not have compen
sation for the sacrificed rook, si nce he has
to reckon with the advance of the e3-pawn.
Carlsen gives the fol lowi ng variations:
a) 31 . . . xe5 32. g4+ Ag7 (the pin al
lows Wh ite to regain one of the pieces)
33. A x h7+ h 8 (if 33 . . . x h7? there
is 34. f!.xf7, when 34 . . . x h 6? fai l s to
35. tDf5+), and Wh ite's attack suffices
only to save the game: 34. tDf5 Axh6
35. tDxh6 f!.e7 (or 35 . . . Ad5 36. h4 g7
37. tDxf7+ Axf7 38. Ag6+ g8 39. Axf7+
xf7 40. f!.xf7 xf7 41 . f4+, forcing a
draw by perpetual check) 36. h4 e2
37. tDxf7+ f!.xf7 38. Ag6+ g8 39. Axf7+
f8 4 0 . h 6 + e7 41 . h4+ f8
42 . h6+ with perpetual check;
b) 31 . . . Axe5!? (vacating a square for the
ki ng) 32. Axh7+ (32 . g4+ h8 33. tDf5
f!.g8) 32 . . . xh7 33. h5 e2 34. tDxe2
g8 and W h ite's attack comes to a
standstill.
30. f5?
Carlsen can not stop himself, but this reck
less knight sacrifice should have led to de
feat. Therefore he should have retreated his
rook - 30. f!.e2 (weaker is 30. xd4 Axe5),
i ntending after 30 . . . A xe5 the exchanging
operation 31 . Axf8 xf8 32. f!.xe5 f!.xe5
33. Axg6 Ad5 34. tt:'ih5.
30 . . . gxf5 31 . g3+ g6 32 . Axf5
 Chapter 2 The Way to the Top

a b c d e f 9 h a b c d e f 9 h

.I . .... 8 8 J. 8
7
'ti'.l 7 7
..t.. .I ..t.. 7
J!l\ 6 6 .l g 6
5 5 5 .l 'iV 5
4 4 .l g 4
3 g 3 3 3
2 2 2 2
g
a b c d e f 9 h a b c d e f 9 h

32 . . .Wxe5?!
After 32 . . . AxeS! 33. gg4 d 6 34. d3
Ac8! 35. gxg6+ fxg6 36. Axg6 Ad? White's
hopes would have been dashed . But even
after the text move it is not easy for him to
save the game.
36 Ae4??
In the time scram ble Shirov overlooks an
interposition , lead ing to h i s defeat. The
bishop manoeuvre was good only after
the i n c l usion of 36 . . . gf8 , w h i c h wou l d
have forced the exchange o n g7, since if
37. gxd4 Ae4! 38. ggs Black has 38 . . . gfs.

37. Ei:g5! We6 38. Axg7

33. Ei:g4 Ag7 34. Axg6 fxg6 35. xg6
Ei:e7 36. Ei:f4 Black resigned .

In 2005 the composition of the B Tournament at the famous festival in Wijk aan Zee was
selected especially carefully by the organisers, who were able to ensure a solid FIDE
category 1 3 . More than half of the participants comprised the most promising young
players of that time. There were the recent prodigies Sergey Karjakin and Magnus
Carlsen, who were meeting for the first time, and some slightly older players: Shakhri
yar Mamedyarov, Ivan Cheparinov, Daniel Stellwagen, Jan Smeets, Alejandro Ramirez
and Alexand ra Kosteniuk - all grandmasters. I n short, in such a youth parade a good
performance was a matter of prestige! Magnus arrived at the tournament not only ful ly
prepared in the openings, but also with a more pragmatic approach. Even his entourage
had changed. Alexandra Kosteniuk made this curious observation: 'The numerous snacks
and d rinks had disappeared from his table, and in Wijk aan Zee I never saw him wearing
his favourite trainers.' What told was the favourable influence of Peter Heine N ielsen,
who was playing alongside. Everyone saw how much he chatted with Carlsen during
walks together. Outwardly, because of the d ifference in their weight categories, this pair
looked very eccentric and provoked smiles on the faces of passers-by.
But let's return to the chess itself. Magnus's play was rather restrained: after winning
in the 2nd round against Kosteniuk, he proceeded cautiously, as though awaiting his

54
 I. Cheparinov - M. Carlsen

opportunity. And after a series of seven d raws the Norwegian as though gained his
second wind and powerful ly attacked the kings of Chepari nov and N i kolic. Moreover,
the attack against the experienced Bosnian grandmaster was awarded the spectators'
daily prize. Unfortunately, after joining the leading g roup, Magnus ran out of steam and
lost his two remaining games. The result was only a modest 7th place. First prize and
an invitation to the A Tou rnament the fol lowing year was won by Sergey Karjakin, who
achieved an excellent result - 9% out of 1 3 .
There was reason for h i m to b e distressed - progress t o higher society had to be
deferred for another year.

Game 1 0 im portant than the sacrificed pawn.

I. Cheparinov - M. Carlsen The main continuations are 1 5. Y#f3 and
Wij k aan Zee B 2005 1 5. 'Wh5, forcing Black to relieve the pawn
Sicilian Defence [ 833] tension, which is i mportant in the fight for
the d5-square.
1. e4 c5 2. f3 c6 3. d4 cxd4
4. tnxd4 f6 5. c3 e5 1 5 . . . d5 1 6. exd5 e4 1 7. 'We2 Eib8

The Chelyabinsk Variation assu med an
honourable p l ace in M ag n u s's open i n g
arsenal , a n d he soon became famed as
one of its main experts.
6. tndb5 d6 7. Ag5 a6 8. a3 b5 9. Axf6
gxf6 10. d5 Ag7 1 1 . Ad3 e7 12. xe7
'f!xe7 13. c4 f5 14. 0-0 0-0
a b c d e f
17 . . . 'tlfb4 1 8 . Ac4 'W x b2 1 9. tbc2 axb5
20. f!.ab1 'tlff6 21 . f!.xb5 f4 (Rogozen ko) also
came into consideration. The active pawn
pair gives Black compensation for the sac
rificed material (for exam ple, if 22. 'tlfxe4?!
there is 22 . . . Af5 23. 'tlfe2 f3).
1 8 . Eiab1
a b c d e f 9 h

.!
8
.!.i. .. 8
if ... .i. ... 7
8 8
7
7
6
6 ... 6
5 [3:, 5
5 5
4 4 4
4
3 ti:J 3 3 ti:J 3
2 [3:, [3:, [3:,[3:,[3:, 2
2 /J:,/J:,
1
a b c d e f 9 h
a b c d e f 9 h

1 8 . . . Eib6!
15. cxb5?!

A poor move, allowing Black a pawn centre, A typical trait of the young Carlsen was that
which in such a dynam ic position is more he always had his eye on the opponent's

55
 Chapter 2 The Way to the Top

king. With the vigorous rook manoeuvre to
the kingside he, as they say, takes the bull
by the horns - the king will have no peace.
1 9. 'We3 Jag6 20. J.c2
Wh ite pins his hopes on the b-pawn, since
after 20. bxa6 it is not possible to exploit
the passed a-pawn. I n view of the fact that
the g2-pawn is inadequately defended , in
many variations B lack is practically g uar
anteed perpetual check by a rook sacri
fice: 20 . . . Wh4 21 . a? Ad4 (with the idea
of 22. Wxd4 xg2+) 22. a8W (22 . Wg3
Wxg3 23. hxg3 Axa7 24. Ac2 e3 is risky
for Wh ite) 22 . . . Axe3 23. fxe3 exd3 24. f2
(if 24. Wa7 there is 24 . . . xg2+ 25. 'it>xg2
Wg4+) 24 . . . We? - Black retains the initiative
and is th reatening not only the e3-pawn,
but also the move . . . Ab7.
The passive 20. fd1 We5 21 . Af1 allows
Black to build u p a dangerous attack: 21 . . .
f4 22. Wc3 Wh5.
20 'Wh4
..
a b c d e f g h
8
.. 8
7 .i. 7
6 .! 6 . . . xg3.
5 8 8 5 24
4 'iW 4
3 CLJ '@' 3 a
of his pieces is illustrated by the following
variation: 21 . bxa6? Axa6 22. fd1 Ah6
23. We1 xg2+! (now this sacrifice is ap
propriate) 24. 'it>xg2 Wg4+ 25. 'it>h1 Wf3+
26. 'it>g1 Ae2! and Black has set up a mating
construction. If 21 . fd1 White has to reckon
with the advance of the f-pawn - 21 . . . e8
22. d6 f4. He is also u nable to set u p a
blockade by 21 . f4 on account of 21 . . . exf3
22. Wxf3 Ae5 23. h3 g3 24. We2 Wxh3 etc.
21 . . . h8!?
By including his second rook in the attack
along the g-file, Magnus demonstrates that
he is not thi nking about a draw, which he
could have forced by 21 . . . Ad4 22. Wxd4
xg2+ 23. 'it>xg2 Wg4+.
22. c4 JagS
The murderous . . . Ad4 is th reatened .
23. g3 'Wh3 24. Jafd1
A usefu l move, at the same time vacating
a square for the king. If 24. Wd2 there fol
lows 24 . . . f4! with the threat of 2 5 . . . h6 (the
im mediate 24 . . . h6 is parried by 25. f4) .
24. f4?! im med iately is not good because
of 24 . . . Af8! with the th reats of . . . Ac5 and
Jah6 25. 'Wf4
b c d e f g h

2 8 8 iL 888 2
8
.i... 8

1:1 1:1
7 .i. 7
a b c d e f g h
6 8 .! 6
5 8 5
21 . b6 4 LD '@' 4
This lone pawn hard ly has any future, but
3 8 1V 3
it is not easy for Wh ite to choose a sound 2 8 8 iL 8 8 2
plan, whereas Black is ready to add fuel to 1:1 1:1
his attack. For example, the latent activity a b c d e f g h

56
 M . Carlsen - P. N i kolic

25. . At6? 33. g1 . Now, however, Wh ite is doomed .

26 . Axe5 27. 'Wxe5+ f6 28. 'Wf4 Elg4

A careless move, which could have cost
Black the fruits of his efforts. He should There was a quicker win by 28 . . . aS! 29. b4
have incl uded the other bishop in the at- axb4 30. xb4 Aa6 with the threat of play
tack by playing 25 . . . aS! , when it is doubtfu l ing the bishop to f3 .
whether White can prevent the develop-
29 Wc7 ElgS 30 Ele1
ment of the bishop at a6. For example:
26. b4 Ac3 27. bxaS (27. bS? h4) 27. . . Aa6 There is no way of saving the game. 30. b7
28. Ab3 h4 (in the event of 28 . . . 'Wxh2+ d oes n ot work because of 3 0 . . . A x b7
29. f1 c8 30. e2 xc4 31 . Axc4 Axc4+ 31 . 'Wxb7 e3! with u n avoidable m ate. If
32. e3 'WhS 33. e1 d6 34. bc1 ! Black's 30. 'Wf4, then 30 . . . aS! is decisive.
attack comes to a stand sti l l) 29. 'Wxh4
30 . . . 'Wxh2+ 31 . <!>t1 'Wh1 + 32 . <!>e2
(29. e3? xh2+ leads to mate) 29 . . . xh4,
'Wf3+ 33. <!>d2 'Wxf2+ 34. <!>d1 f4
winning the queen.
35. 'Wc3 fxg3 36. Axe4 Ag4+ 37. <!>c1
Wh ite is o b l i g ed to th row caution to
lieS
the winds: 26. d6 Aa6 27. b3 e3! 28. d3
(evidently the only move, since 28. Ad3? Wh ite resigned .
is bad: 28 . . . e2 29. d7 'Wxh2+ 30. f1 e2+!
31. xe2 e6+ 32. <!>d3 'Wg2 33. 'WxfS
flf6, and Black wins the queen) 28 . . . 'Wxh2+ Game 1 1
29. f1 with sharp play, where Wh ite is M. Carlsen - P. Nikolic
apparently able to defend . Here are some Wijk aan Zee B 2005
sam ple variations: French Defence [COB]
a) 29 . . . e8 30. xe3 'Wh1 + 31 . e2 'Wc6
1 . e4 e6 2. d4 d5 3. d2 c5 4. exd5
32. d2 Axd3 33. Axd3 xd6 34. 'WxfS
exd5 5. gf3
g8 3S. e2, and the wh ite pieces are
excel lently coordinated ; a b c d e t 9 h

b) 29 . . . exf2 3 0 . 'W x f2 (30 . d 7 ? ! 'Wg 1 + 8 .I .S..'ii'tfi.S.. .I 8

31 . e2 e6+ i s dangerous for Wh ite, 7 .. 7

while if 32. e3 or 32. e3 there follows 6 6
32 . . . Ah6) 30 . . . 'Wxf2+ 31 . xf2 h2+ 5 5
32. g1 xc2 33. d7 Af6 34. e1 , and
4 4
White's passed pawns com pensate for
the lost piece.
3 tt:J 3
2 /j, /j, [j, tt:J fj, /j, /j, 2
26. e5?
1 It It 1
After 26. b7! Axb7 27. d6 'Wxh2+ 28. f1 a b c d e f 9 h

Black's attack would have come to a stand

sti l l , for example: 28 . . . g7 29. xf5 (or 5 c4!?
29. xfS Ae7! 30. xb7?! f6) 29 . . . 'Wh1 + The experienced N i kolic aims to neutralise
30. e2 'Wh5+ 31 . g4 xg4 32. 'Wxh6 WxfS the fruits of his young opponent's home

57
 Chapter 2
preparation and he avoids positions with an
isolated d5-pawn . Another version of this
idea - 5 . . .tt:lc6 6. Ab5 c4 - has its special
features, since Wh ite has determ ined the
position of his light-square bishop.
6. b3 cxb3 7. axb3
A prag matic approach to the choice of
m ove, t y p i c a l of p l ayers of the c o m
p u t e r g e n e rat i o n . M a g n u s p refe rs a
s o u n d m ove, a n d d e c l i n e s to enter a
d i s c u s s i o n i nvo l v i n g t h e i nterm e d i ate
exc h a n g e of the l i g h t- s q u a re b i s h o ps
after 7. Ab5+ Ad7 8. We2+ We7 9. Axd7+
ti:lxd7 1 0 . a x b 3 , w h i c h had occu rred i n
seve ral o f N i ko l ic's g a m e s . For exa m
p l e , t h e g a m e Ad a m s - N i ko l i c ( N e u m
20 02) conti n u ed 1 0 . . . f 6 1 1 . tLlf1 Wxe2+
1 2 . xe2 A d 6 1 3 . Cb e 3 Cb e7 1 4 . Aa3
Axa3 1 5 . Z!xa3 f7 1 6 . tt:le1 tLlf8 1 7. tLld3
tt:le6 1 8 . c3 Z!hd8 1 9 . d2 tt:lc6 2 0 . Z!d 1
Z!ac8 %-%.
Apart fro m t h e exc h a n g e of t h e l i g ht
square bishops, also of i nterest i s the
p ract i c a l l y u n ex p l o re d 7 . . . tt:l c 6 ! ? , not
fearing the swift attack suggested half
a c e n t u ry a g o by I g o r B o n d arev s k y :
8 . We2+ Ae7 9 . tt:le5 ! (not h i n g is g iven
by 9 . a x b3 tt:lf6 1 0 . Aa3 0-0 1 1 . Axc6
A x a3 1 2 . Z! x a3 Wd6 1 3 . Z!a5 Wxc6=) .
The best h e re i s 9 . . . We? ! ( i t i s i m p o r
tant to d efend the knight, not forgetti n g
about the defence o f the bishop on e7)
1 0 . axb3 (th i s is stronger than 1 0 . c4 tLlf6
1 1 . Axc6+ bxc6 1 2 . axb3 0-0 1 3 . 0-0
c 5 = o r 1 0 . ti:ldf3 f6 ! ? 1 1 . Axc6+ bxc6
12. Cbd3 bxc2=) 10 . . . ti:lf6 1 1 . Axc6+!? (or
1 1 . Aa3 0-0 1 2 . Axc6 A x a3 1 3 . Z!xa3
bxc6 1 4 . 0-0 c5=) 1 1 . . . bxc6 1 2 . 0-0 0-0
1 3 . Z!e1 Z!e8 , and B l ack, alternating the
i deas . . . Ae7-b4 and . . . c6-c5, o bta i n s
g o o d play.
The Way to the Top
a b c d e f 9 h
8 I. 1.11 ..t 'Vi ..t 1.11 I. 8
7 .t.t .t .l ?
6 6
5 5
4 4
3 [j, tb 3
2 [j, tiJ [j, [j, [j, 2
1 1:, .: 1
a b c d e f 9 h
7. . . .tb4
This bishop manoeuvre was actively em
ployed by David Bronstein. 7. . . Ad6 8 . Ad3
tt:le7 9. 0-0 0-0 1 0. Z!e1 tt:lbc6 1ooks more
natural . For exam ple, the game Lebedev
Bronstein (Moscow Championship 1 953)
contin ued 1 1 . c4 Ab4! (here this bishop
manoeuvre effectively cou nters the rook
move to e1 ) 1 2 . \Wc2 ti:lg6 1 3 . Ab2 Ae6
1 4 . Z!ad1 Wf6 1 5 . Z!e3 Ad6! 1 6. Wc3 ti:lf4
with the initiative for Black.
After the modest move 1 1 . c3 B l ac k
c o m p l etes h i s d eve l o p m ent: 1 1 . . . Af5
1 2 . tLlf1 Z!e8, retaining a sound position,
for example: 13. Aa3 Axa3 14. Z!xa3 Wd6
1 5. Z!a1 Wf6 1 6 . tt:le3 Axd3 1 7. Wxd3 a6
1 8 . b4 eLla? 1 9. tt:le5 tLlb5 with the idea of
. . . Z!ac8 .
8. e5
It is hard to catch Carlsen unawares, even
in such a rare variation. He is fol lowing the
game Svidler- Korchnoi (G roningen 1 996).
Black is not caused any particular prob
lems by 8. Ad3 tt:le7 9. 0-0 ti:lbc6 1 0. Aa3
0-0 or 8. Ab5+ Ad7 9. We2+ tt:le7 1 0. 0-0
0-0 1 1 . Aa3 Axa3 1 2 . Z!xa3 ti:lbc6.
8 . . . e7
58
 M. Carlsen - P. N i kolic

Korchnoi played 8 . . . ttlf6 and after 9. Ad3
0-0 1 0. 0-0 Ac3 1 1 . f!a4 Ad? 1 2. ttlxd7
tDbxd7 1 3 . ttlb1 Aa5 he maintained the bal
ance. But 1 0. Ab2!? was possibly stronger.
9. Ad3 bc6 1 0. 0-0
One gains the impression that Magnus had
analysed this position at home, since he
ignores the th reat of . . . Ab4-c3 with the
win of a pawn (otherwise he would have
played 1 0. Ab2).
1 0 . . . Ac3
Black accepts the challenge, although his
retarded development undoubted ly puts
his king in a dangerous position . It was not
yet too late to castle - 1 0 . . . 0-0 1 1 . Ab2
with somewhat the better game for Wh ite.
11. !::[ a 4!
a) 1 3 . . . ttlg6 1 4 . f! h 5 0-0 1 5 . ttlb1 Ab4
1 6 . c3 Ae7 1 7. f4 Wc8 1 8. f5! with an
irresistible attack in view of 1 8 . . .CL!xe5
1 9. f6 . The trick 1 8 . . . Wc6, with the idea
after 1 9 . fxg6 fxg6 of trapping the rook,
also does not save Black in view of the
simple 1 9. e6;
b) 13 . . .ttlc6 14. ttlf3 h6 1 5 . Ad2 Axd2 (weak
er is 1 5 . . . Axe5 1 6 . l:!h5 Af6 1 7. f!xd5
Wc8 1 8 . We2+ Ae6 1 9. We4 g6 20. Ac4
with advantage to Wh ite) 1 6. Wxd2 0-0
1 7. f!e1 f!e8 1 8 . Wf4 (1 8 . f!xh6 tbxe5)
1 8 . . . Wa5 1 9. b4 Wb6 20. f! x h 6 ! gxh6
21 . Wxh6 ttld8 22. Wh7+ (it was not for
the sake of perpetual check - 22. Ah7+
h8 23. Ag6+ - that Wh ite sacrificed a
rook) 22 . . . f8 23. Wh8+ e7 24. Wh4+
f8 25. ttlg5 with a strong attack for
Wh ite.

1 2 . xc6 xc6 1 3. Aa3 Ae6 14. f3

a b c d e f
Ab6?!
a J. .t.'ii'
7 .f..f. It was essential to neutralise the th reats
on the a3-f8 diagonal , which would have
6 6
been achieved by 1 4 . . . Af6 1 5. l:!e1 Ae7.
5 .l tD 5

4 2:, 8 4 1 5. 'Wa1 !
3
8-*.iL 3
A subtle manoeuvre, dispelling Black's illu
2
8tD 888 2 sions of hiding his king on its own wing: if
iLvtli 2:. 1 5 . . .f6 there can fol low 1 6 . f!e1 f7 1 7. l:!f4
a b c d e f g h with the threat of ttlf3-g5+.

11 . . . Axd4 15 . . . Wc7 16. b4!

The attack on the king was the 1 4-year-old

It looks p rovocative to play 1 1 . . .CL!xe5
Carlsen's natural element. He did not play
12. dxe5 Ad? (or 12 . . . Axe5 1 3 . f!e1 Af6
1 5. Wa1 in order to limit himself to 1 6. Wxg7,
14. Wh5 f8 1 5. ttlf3 g6 1 6. Wh6+ g8
although this also g i ves a considerable
17. Wf4 with the better chances for White)
advantage. With the pawn th rust Magnus
13. l:!h4 (after the passive 13. f!a2 ttlc6 Black
warns his opponent about the danger of
has an easy game in view of the threat of
castling queenside .
. . .tDc6-b4), since the black king has not yet
castled. Here are a few variations: (see next diagram)

59
 C apter 2
-------------------h- - _ __
a b c d e f 9 h
6 6
5 5
4 M8 4 preventing castling, which is possible, for
3 exam ple, in the event of 1 9. f!xe5? 0-0 ! ,
2
8 and i f 20. l:!xe6 there is 20 . . . Axf2+ 21 . h1
1 'Wd7, forking the rooks.
a b c d e f 9 h
1 6 . . . f6
Black opts for artificial castl ing, realising
that after 16 . . . 0-0 1 7. b5 l!iJe7 1 8. f!e1 ! he
will be unable to withstand Wh ite's dan
gerous i n itiative. For example, 1 8 . . . Ac5
1 9 . Ab2 'Wb6 20. f!e2 f6 2 1 . Ad4 or
18 . . . f!fe8 1 9. Ab2 l!iJf5 20. g4 'Wc5 21 . f!e2
l!iJh6 22. Axg7 l!iJxg4 23. Ad4 'Wd6 24. l:!xe6
with an obvious advantage to Wh ite in both
variations.
Let us exam ine the dangers awaiting
Black after 16 . . . 0-0-0 1 7. b5:
a) 17 . . . l!iJa5 1 8 . A b2 f6 1 9 . Ac3 l!iJc4
20. Ad4 b8 (if 20 . . . 1!iJa5 there follows
21 . 'Wa2 and then 22. f!a1 ) 21 . f!e1 l:!he8
22. f!xa7! with a strong attack;
b) 17 . . .1!iJe7 1 8 . Ac1 (or 1 8 . f!c1 b8 1 9. c4
d xc4 2 0 . A xc4 'Wd7 2 1 . l!iJe5 'WeB
22. Axe6 fxe6 23. l!iJc4 Ad4 24. Ad6+
and Wh ite wins the exchange) 18 . . .l!iJf5
(1 8 . . . 1!iJ g 6 1 9 . Ae3 A x e 3 2 0 . fxe3
'Wc5 2 1 . f! x a7 'Wxe3+ 2 2 . h 1 'Wb6
23. l!iJg5 with the th reats of 24. l!iJxe6
and 24. l!iJxf7) 1 9. Axf5 A xf5 20. Af4
Wc5 21 . Ae3 with an obvious advantage
to Wh ite.
1 7. e1 'i!lf7
It would be wrong to criticise N i kolic for
W y to t he To p
_The a - __________________ __
this natural move. Black's position is so
comprom ised that it is already hard to offer
any good advice. Lev Psakh is considers it
best to return the pawn - 1 7. . .l!iJe5, but even
here White's attack is very strong: 1 8. l!iJxe5
fxe5 1 9. b5! . An im portant i nterposition,
I I
AA'Y! AA
j_ j_
8 AA
J:r
Jt Jt
8 8 88
J:r
Analysis diagram
1 9 . . . Wf7 20. l:!xe5 (if 20. f!e2 there can
fol low 20 . . . e4!? 21 . Axe4 dxe4 22. f!axe4
0-0-0 23. f!xe6 Axf2+ 24. h1 Ad4, and
B l ack somehow h o l d s o n) 20 . . . A x f2+
21 . h1 0-0-0 22. Ab2 with a dangerous
initiative for Wh ite. We wi l l trust the two
times USSR champion Lev Psakh is, an
expert on this opening, but also enquire
what Rybka thinks about it. Its verd ict is
pessi m istic: 20 . . . Axf2+ 2 1 . h1 0-0-0
22. Ac1 ! l:!hf8 23. f!f4 'Wd7 24. Wc3+ b8
25. Wb4! f!f7 26. f!xf7 Wxf7 27. Wf4!?
(27. l:!xe6 Wxe6 28. Wf4+ a8 29. Wxf2
is also good) 27. . . Wxf4 28. Axf4 and Black
cannot avoid loss of material.
19 . . . 0-0-0 20. l:!xe5 'Wf7 21 . We1 f!he8
22. Ac1 ! is also dangerous for Black - on
the q ueenside too his king cannot fi nd
peace: 22 . . .b8 23. Af4 a8 24. 'Wa1 with
the threat of 25. l:!xa7+.
Fleeing with the king to the queenside
60
-------
M . Carlsen - P. Nikolic
---------------------------
ltJ

is realistic after the s i m p l ification rec 19 .. !3:ae8

ommended by And rey Devyat k i n : 1 9 . . .
There is no way of saving the game: 19 . . .
e4!? 2 0 . A xe4 d xe4 2 1 . !!axe4 0-0-0
g 6 (1 9 . . . ttlc4 2 0 . Af5 !) 2 0 . ttlg5+! fxg5
22. xe6 !!d2 23. !!e8+ !!xe8 24. !!xe8+
21 . 'Wf3+ g7 22. Ab2+ d4 23. !!xd4 Axd4
d8 25. !!xd8+ 'Wxd8, but after 26. 'Wc3+
24. Axd4+ h6 25. h4 and Wh ite wins.
b8 27. 'Wf3 he has to play a difficult end
game a pawn down.

18. b5 as

Putting the knight out of play leads to a

rapid defeat . But also after the logical
18 . . .ttld8!? Wh ite retains a dangerous at
tack. Here is a sample variation: 1 9. 'Wd1
d7 20. !!h4 g8 21 . ttld4 Axd4 22. !!xd4.

19. 'Wd1 !
a b d f g h
The queen has done excellent work on the c e

a1 -h8 diag onal and it returns home in order

20. g5+! fxg5 21 . Wf3+ g8 22. l3:xe6
to attack the king on the light squares. I n
the fi rst instance ttlg5+ i s threatened . Black resigned .

The modest performance i n Wij k aan Zee showed that the youngster needed time to
improve his play. Carlsen took a time-out, probably the fi rst in his career. However, the
more than two months rest from tournaments was a wearisome period for Magnus.
Strangely enough, the break did not benefit him, and a warm-up, planned on home
territory at the Gausdal Classic Tou rnament (FI D E category 1 0), was frankly a fai lure.
Mag nus started extremely bad ly - half a point i n the fi rst fou r rounds, and he fi nished
in a share of 6th-8th places. In a class of his own was Sergey Tiviakov, who scored 8%
poi nts out of 9. Mag nus's mood was improved by h i s partici pation in a series of rapid
play tournaments with a 25-minute time control. The most prestigious of these was
the four-player Cuidad de Leon tournament in Spain with the partici pation of Anand,
Kasimdzhanov and Shirov. True, he did not play the latter two, s ince i n the sem i-final
Carlsen lost 1 -3 to Anand, but the fi rst meeting with one of the strongest players in
the world was undoubted ly beneficial.
At the European Championshi p i n Warsaw, Mag n u s justified the expectations of
his supporters, scori ng 8 poi nts from 1 3 games (+5 -2 =6). In such a un iformly strong
'Swiss', this cou ld be considered a successfu l result for a 1 4-year-old grandmaster.
His performance in the double-round tournament in Biel with six g randmasters (FI DE
category 1 6) was more modest. Magnus propped u p the tournament table with 4 poi nts
from 10 games, in which he did not score a single win and suffered two defeats. Before
the next category 1 6 tournament, the Samba Cup in Denmark, Carlsen warmed up at

61
 Chapter 2 The Way to the Top

home in the Gausdal Bygger'n M asters (1 st place - 8 poi nts out of 9) - and 'burned
h i mself out'. There was no time to regain his strength, since the Samba Cup started
the very next day. The tournament in Skanderborg followed the same pattern as i n
Drammen: 8th place o u t o f 1 0 contestants was undoubted ly a fai lure for Carlsen . The
only fai nt consolation was a win over the fi rst prize-winner, Baad u r Jobava.
These set-backs can hardly have been conducive to a good mood before the main
tournament of the year - the World Cup i n Khanty- Mansiysk. If one sums u p the series
of fai l u res, beg i n n i ng with the Spanish team championship on the eve of 2005, where
Magnus suffered fou r defeats, the picture was depressing, and the fai lures in his native
Scandi navia were especially unpleasant: the Drammen, Gausdal Classic and Samba
Cup tournaments. Less than a month remained to prepare for the World Cup.
I n youth , which is typified by optimism, it is easier to heal one's wounds.

Game 1 2 a b c d e f g h
M. Carlsen - K. Lie
8 .i 1.1\.i. .. 8
Gausdal 2005
7 -*. 7
'if JA,I
King's Indian Defence [E62]
6 6

This game is i nteresting for Carlsen's ap 5 .&I 5

proach to the solving of open ing problems 4

BB 4

i n a non -standard situation, and also the 3

B [jj [jjB B 3

conversion of a m i n i mal advantage i n an 2 B BB 2

endgame with opposite-colour bishops 1 -: : 1
with rooks. a b c d e f g h

1 . d4 f6 2. f3 g6 3. g3 .tg7 4 . .tg2
0-0 5. 0-0 d6 6. c4 c6
A rare branch of the King's Indian Defence,
where Black attacks not the d4-pawn , but
the c4-pawn.
7. c3 'Wb6 8. h3 Wa6 9. b3 b5
(see next diagram)
1 0 . .tg5!?
On encountering a comparatively rare vari
ation, Carlsen decides to fight for the initia
tive from the very first moves. He sacrifices
a pawn, hoping to exploit the retarded de
velopment of the black pieces. Especially
since experience of playing this position i n
a routine manner ind icates that Black has
q u ite good cou nterplay. For exam ple, the
attempt to attack the b5-pawn - 1 0. cxb5
cxb5 1 1 . Wd3 - achieves its aim only after
the awkward 1 1 . . . Ad7 1 2. Ag5 ! , when the
bishop on d7 merely hinders Black in com
pleting the development of his q ueenside,
while at the same time the exchange on f6
followed by tDd5 is th reatened. But 1 1 . . .
b4 is far stronger, and the attem pt to get
at the rook on a8 - 1 2. tDe1 - is parried by
1 2 . . . Axh3! 1 3 . Axh3 (the mass e l i m i na
tion 1 3 . Axa8 Axf1 leads to the loss of
a pawn) 13 . . . bxc3 14. Ag2 d5 with good
play for Black.

62
 M . Carlsen - K. Lie

The main conti nuation is considered

to be 1 0. ttld2 bxc4 1 1 . ttlxc4 Ae6, where
Black has q uite good cou nterplay. I n his
time Magnus's trainer Agdestein became
aware of this in a game with Klaus Bischoff
(Dortmund 1 987), which continued 1 2 . ttle3
tDbd7 1 3 . Ab2 .fc8 1 4 . .c1 .ab8 1 5. d2
CL!b6 1 6. f4 c5 1 7. d5 Ad7 1 8 . .a1 as with
active play for Black.

10 . bxc4 1 1 . bxc4 Wxc4 1 2 . !ac1 .ta6

. .

a b c d e f g h
White's plans include e2-e4 followed by
tDc3-d5. Therefore it is useful to clarify the 1 4 . . . e5?
intentions of the active bishop on g5 - 1 2 . . .
Understandably, Black is tired of being un
h6!?, which is usually practised i n King's
der pressure, but if he did not l i ke 14 . . . .!::! d 8
Ind ian set-u ps. Now it is i l logical to play
1 5. d2 with the threat of 1 6. ttld5, then
1 3 . Axf6 Axf6 1 4 . ttle4 xa2 1 5. ttlxf6+
14 . . . h 6 1 5. Ad2 b6 would have enabled
exf6, since the position is simplified , and it is
him to defend. I n returning his extra pawn,
not easy for White to breach the black pawn
Black ends u p i n a strategically i nferior
barricade. Igor Sto h l considers 1 3 . Ad2
position i n view of his weak pawn on c6.
'We6 1 4 . .e1 ttlbd7 (1 4 . . . Aa6 1 5. e4 .c8 is
also possible, but not 15 . . . Ac4?! because 15. dxe5 dxe5 1 6. xe5 Wa5
of 1 6. e5! ttlh7 1 7. exd6 xd6 1 8. tbe4 d5
If 16 . . . .e8? White has the decisive 1 7. tbxf7!
1 9. tbe5) 1 5 . e4 ttlb6 1 6 . e5 tbfd5 1 7. ttlxd5
xf7 1 8. e5 tbh5 1 9. ttld5! b2 20. ttlc7.
cxd5, and here 18 . .c7 retai ns the i n itia
tive for Wh ite. 1 7. xf7

13. !ae1 Here this sacrifice is not so effective. Play

ing for bri l liancy might merely have brought
The o n l y known g a m e at t h at t i m e ,
Carlsen unnecessary problems. The simple
Berkes - Stokke (Stockholm 2003), con
1 7. Axf6 Axf6 1 8. tbxc6 tbxc6 1 9. tbd5 Ae5
tinued 13. e4 h6 1 4 . Ad2 e5 1 5 . d xe5 d xe5
20 . .xc6 would have enabled White to win
16. tbd5 cxd5?! (an i nsufficiently justified
a pawn without any brilliancy.
queen sacrifice; Black should have decided
on 1 6 . . . xa2, when all Wh ite can do is 17 . . . !a xf7 1 8 . e5
force a d raw - 1 7. .a1 b2 1 8 . .b1 a2)
17. .xc4 Axc4 1 8 . ttlxe5! Axf1 1 9. xf1 (see next diagram)
CL!xe4? (1 9 . . . d xe4) 20. Ab4 .d8 2 1 . ttlxf7!
18 bd7?

CL!c6 22. ttlxd8 .xd8 23. Aa5 1 -0.

Black is rattled, and he does not even make
13. . .Wb4 1 4. e4 an attem pt to play on with an extra knight.
Apparently, after 18 . . .ttle8 he was afraid of
(see next diagram) the attack 1 9. ttle4! (1 9. AdS b4 20. e6

63
 Chapter 2
a b c d e f 9 h
8 .1. J!&\ 8
7. ... ... 7
6 j_
J!&\. 6
s'if 8 ii 5
4 4
a b c d e f 9 h
i s parried by 2 0 . . . .!::! f S 2 1 . tD e 4 Ac4
22. a3 '!Wb3) 19 . . . h6 (the pawn is taboo:
19 . . . AxeS? 20. tDcS .fS 2 1 . '!Wb3+) 20. e6!
.fS 2 1 . tDd6 hxgS 22. e7! and Black has
no defence. However, the threat of the e
pawn's advance could have been averted
by the bishop retreat 1 9 . . . Ac8, although
here too Wh ite's initiative is very danger
ous: 20. tDcS .!::!fS 2 1 . '!Wb3+ h8 22. g4!
.xgS 23. '!Wf7 /!i)c7 24 . .ed1 /!i)d7 2S. /!i)xd7
Axd7 26 . .xd7 '!WxeS 27. .xc7.
1 9. exf6
But here playing for bri l l iancy by 1 9. Axc6!
l:!c8 20. /!i)dS /!i)xdS 2 1 . e6 was q u ite ap-
propriate. However, the move i n the game
also ensures Wh ite an advantage.
1 9 . xf6 20. Axf6 Axf6
If 20 . . . l:! x f6 there wou l d h ave fol l owed
21 . /!i)dS! .f7 22. /!i)e7+ h8 23. tDxc6.
21 . e4 Ae7
It is i m portant to retai n this bishop - oth
erwise th ere i s not h i n g w i t h w h i c h t o d e
fend t h e dark squ ares. If 2 1 . . . AbS, then
22. /!i)xf6+ .xf6 23. l:!eS! with the th reats
Axc6 and Wb3+ is good e n o u g h to w i n ,
but there is also a spectac u l a r com puter
so l ution: 23. AdS+!? f8 24. Wd4! Wd 8
The Way to the Top
2S . .e6 ! . If 2 1 . . . '!W x a 2 ? t h e re also fo l
lows 2 2 . /!i)xf6+ .xf6 2 3 . '!Wd4! and B l ack
has no defence: 23 . . . .!::! f 7 24 . .xc6 '!WaS
2S . .ce6.
22. :xc6 Ab7
The black pieces coordinate more harmo
niously after 22 . . . l:!d8 23. '!Wa1 AbS 24. l:!e6
Ac4 2S . .eS '!Wb6.
23. :e6 :d8
If 23 . . .Wxa2? White wins by 24 . .xe7! .xe7
2S. /!i)f6+ f7 26 . .xe7+ xe7 27. /!i)dS+!
f8 28. '!Wf3+ .
24. 1l\'e2 Axe4
B l ac k stra i g htfo rw a rd l y (24 . . . WfS ? !
2 S . .xe7! .xe7 26. /!i)f6+) goes in for oppo
site-colour bishops. But in the presence of
other pieces, especially the queens, hopes
based on this factor are illusory: a bishop
which has no opponent is dangerous in an
attack on the king.
25. 'Wxe4 Ac5 26. :e2 :df8 27. Wd5!
A dangerous pin with an X-ray along the
a2-g8 diagonal .
27 . . 1l\'b4
Black does not sense the danger. It was
essential to 'unpin' the king - 27. . . g7 and
if 28. h2 to offer the exchange of queens
by 28 . . . Ab6. It should be mentioned that
28 . . . l:! xf2 ? is not poss i b l e because of
29 . .xf2 .xf2 30. l:!e7+! , when the mating
attack costs Black h i s q ueen: 30 . . . h 6
31 . g4 Wd2 3 2 . g S + ! '!W x g S 33 . .xh7+.
28 h2 g7
Not 28 . . . Axf2? 29 . .xf2 .xf2 30 . . xg6+.
64
 M. Carlsen K. Lie

a b c d e f 9 h 33. Ads Af6 34. h4 E:d7 35. Ae4 E:fd8

36. q;,h3 E:d6 37. E:xd6 E:xd6 38. E:c2
8 .! 8

7
.l . 7
a b c d e f 9 h
6 g 6
8 8
5
.i. 5
7
4
'iV 4
.!
6
.i..l 6
3 3 5 5
2 g 2
4
4
a b c d e f 9 h
3 3
2
g 2

29. Y!Ve5+?!
a b c d e f 9 h
There was a quick win by 29. f!.2e4! a3
30. eS+ g8 31 . f!.h4 d3 (if 31 . . . f!.g7
Endings with opposite-colour bishops are
there is 32. f!.xg6! hxg6 33. AdS+ with a
Carlsen's forte. On many subseq uent oc
mating attack) 32. xeS, with an extra
casions he was able to demonstrate the su
bishop for Wh ite.
periority of his bishop. In the g iven specific
29. . g8 30. a3
. position it is worth noth ing that White's ex
tra pawn is on the kingside, where Black's
This move allows B lack to exchange the
res o u rces are concentrated . Therefore
queens, which eases the problem of de
Carlsen keeps both wings in view.
fending his king. The th reat of 30 . . . d4
could have been parried by 30. f4! , having 38 . . . E:d7?!
in mind the variation 30 . . . d4? 31 . xd4
Pas s i ve d efence m a ke s t h i n g s eas i e r
Axd4 32. AdS! g7 33. f!.6e4, but by play
for Wh ite. 38 . . . f!.a6!? was more critical:
ing 30 . . . f!.fS Black would have forced the
39. f!.c7+ (after 39. f!.a2 f!.e6! 40. AdS f!.e3
queen to retreat to a less i m posing pos i
B l ac k attacks the g 3 - paw n , p ractically
tion: 31 . a1 A d 4 3 2 . d1 f!. d 8 etc. Also
forcing the exchange of rooks: 41 . a4 Ac3
possible is 30. f!.e4 f!.fS 31 . f!.xg6+ hxg6
42. f!.a3 Ad4) 39 . . . h6 40. h5 f!.xa3 41 . f!.c6
32. 'We6+, winning the q ueen , but how to
f!.e3 , pursuing the bishop. After the transi
convert the advantage i s not clear.
tion into a rook ending by 42. Axg6 hxg6
Playing to s u rrou n d the b i s h o p also
43. f!.xf6 xhS 44. f!.f7 h6 4S. f!.xa7 Black
does not ach i eve its a i m : 30. f!.c6 Ad4
has real chances of a d raw.
31. 'WdS Ab2! (from the long diagonal the
bishop g uarantees its king a shelter at g7) 39. q;,g4 Ae7 40. a4 E:d4 41 . Ac6 h5+?!
32. f!.cS Af6 33. f!.c7 g7.
T h i s i m pu l s ive m ove m e rely c reates a
30 ... Y!Vd4! 31 . 'Wxd4 Axd4 32. f4 g7 weakness on g6. 41 . . . f6 was more logi
cal .
The game has gone into an ending which
is not easy to win. 42. q;,h3 Ad6 43. q;,g2 Ab4 44. q;,h3

65
 Chapter 2 The Way to the Top
-------------------

Ad6 45. Aea l:id3 46. q;,g2 l:ie3 Game 1 3

47. l:ic6! B. Jobava - M. Carlsen
Skanderborg 2005
a b c d e f 9 h
Queen's Indian Defence [E12]
8
8

7 . 7 1 . c4 f6 2. d4 e6 3. f3 b6 4. Ag5
6
::t .i. 6
It was not hard for Mag n u s to prepare for
5 5
this game - at that time this variation was
4 Jobava's main weapon against the Queen's
3
K 3 I ndian Defence.
2 2
4 . . . h6 5. Ah4 Ae7 6. c3 c5

a b c d e f 9 h Carlsen follows the path of lvanchuk in his

47. . . l:ie6?
The rook ending after 47. . J!xe8 48. gxd6
ge7 49. c!>f3 or 47. . . Axf4 48. gxg6+ c!>f8
49. gxf4 gxe8 50. ga6 is clearly better for
Wh ite, who has an active rook. But moving
into a pin leads to the loss of the a7-pawn.
More was prom i sed by a symmetric at
tack on the g3-pawn - 47. . . Ab4! 48. Axg6
Ae1 49. Axh5 g xg3+ 50. c!>f1 (weaker is
50. c!> h2?! ga3 51 . ga6 A x h 4) 50 . . . ge3
51 . gg6+ c!>f8 52. gg4 Ad2 53. c!>f2 ge4
54. f5 gxg4 55. Axg4 with real d rawing
chances.
48. !:ia6 q;,f8 49. Ab5 iif6 50. Ad3!
It i s usefu l to gain a tempo. The rest does
not req u i re any commentary.
50 . . . q;,g7 51 . l:ixa7+ q;,f8 52. a5 Ab4
53. a6 l:id6 54. l:ic7 l:ixd3 55. l:ic8+
q;,g7 56. a7 l:id2+ 57. q;,f3 l:id3+
58. q;,e4 l:ia3 59. a8'W
Black resigned.
recent game with Jobava at the Capablan
ca Memorial (Havana 2005), in which Wh ite
suffered a crushing defeat. The classical
6 . . . Ab7 7. e3 tDe4 seemed rather ted ious
to h i m .
7. e4
The following contin uation is practically
unexplored: 7. d5 tDxd5 8 . tDxd5 (the ex
changing operation 8. cxd5 Axh4 9. d xe6
fxe6 1 0. Vt/c2 with the th reat of Vt/e4+ is
parried by the simple 10 . . . Af6 1 1 . Vt/g6+
c!>f8 with an acceptable game for Black,
who i s able to stabi l ise the position of his
king thanks to the threat of the exchange
1 2 . . . A x c 3 + 1 3 . bxc3 Vt/f6) 8 . . . A x h 4
9. tDxh4 Ab7 (weaker is 9 . . . exd5 1 0. Vt/xd5
tDc6 1 1 . tDf5) 1 0. g3 (after 1 0 . tDf3 exd5
1 1 . cxd5 0-0 12. e3 'Wf6 1 3. 'Wb3 tDa6 Black
successfu lly com pletes h i s development,
retaining some initiative) 10 . . . exd5 1 1 . Ag2
0-0 1 2. 0-0 'Wf6 1 3 . Axd5 (or 1 3 . b3 gea
1 4 . e3 Vt/e5 with the idea of 1 5. cxd5 g5!)
1 3 . . . A x d 5 1 4 . Vt/xd5 tDc6 1 5 . b3 gad8
1 6 . gad 1 gfe8 1 7. e3 ge5 1 8 . 'Wd6 ge6,
and Black maintai ns the balance.

7. . . cxd4 8. xd4 d6
 B. Jobava - M . Carlsen

a b c d e f g h to go i nto an endgame: 1 8 . ygxc5+ bxc5

8
I .i.'ii'fl I 8 1 9. b1 ab8 20. b3 Ac3+ 21 . f1 d2,
7 .i. 7 which, despite the pawn deficit, i s advan
tageous to B l ack. The game conti n ued
6
! r 6 1 8 . b4 yggs 19. d1 ygxg2 20. b5+ e8
5 5
[fj
r- 2 1 . xd8+ xd8 22. Af3 ygg5 23. bxa6
4 4
ygd2+ 24. f1 Ad4! 25. Ae2 Ac5 26. ygg3
3
--[fj 3 ygxa2 27. ygxg7? d1 +, and Black soon
2 2 won .
1 Jl Jl 1 1 0 . . . dxe5
a b c d e f g h
Real i s i n g that 1 0 . . . a x b 5 1 1 . exf6 g x f6
9. lDdb5! (1 1 . . . A x f6 ? ! 1 2 . A x f6 yg x f6 1 3 . ygxd6
i s dangero u s for B l ack) 1 2 . ttJxb5 Cbc6
With the idea of breaking up Black's king
1 3 . Ag3 leaves Wh ite with an enduri n g
side pawns by exchanging on f6. However,
i n itiative without a n y r i s k , Carlsen goes
as was shown by the Jobava - lvanchuk
i n for an exchange sacrifice. His central
game, it proves to be i l l u sory.
pawn mass prom i ses him sufficient com
9.. . a6 1 0. e5!? pensation.

This move, giving the play an almost forc 1 1 . Axf6

ing character, was undoubted ly prepared
Weaker is 1 1 . ygf3? e4! 1 2 . Cb xe4 a x b5
at home. Thereby Jobava i n d i rectly ad
1 3 . tDxf6+ gxf6 1 4 . Wxa8 Ab4+ 1 5 . e2
mits the effectiveness of the pawn sac
ygd2+ 1 6 . f3 e5 with a very dangerous
rifice made by lvanchuk: 1 0 . Axf6 Axf6
attack for B lack. For example: 1 7. ygds
1 1 . ttJxd6+ e7 1 2. ttJxc8+ ygxc8.
(Wh ite loses q u ickly after 1 7. ygxb8 ygf4+
!.'if !. 1 8. e2 0-0) 1 7. . . ygf4+ (as a reserve vari
. ation there is a draw: 1 7. . . Ab7 1 8 . ygxb7

I!J .i. ygf4+) 1 8 . e2 0-0! 1 9. Ag3 ygf5 20. ygd3

r- .-
(if 20. d1 there i s the decisive 20 . . . Ab7!
!:!:, !:!:, 21 . ygxb7 d8+) 20 . . . ygg5 (with the threat
of . . . ttJc6 and . . . d8) 2 1 . yge3 ygg6 22. c1
!:!:,!:!:, !:,!:,!:, bxc4 etc.
M M
Analysis diagram 1 1 ... Axf6

13. ttJa4?! (1 3. ygf3 followed by Ae2 and If B l ac k i nt e r p o s e s t h e exc h a n g e

0-0 was stronger) 13 . . . d8 14. ygb3 ttJd7
15. yga3+, and after 1 5 . . . ttJc5 1 6 . Ae2
(not 1 6. ttJxb6? ygc6 1 7. ttJxa8 because of
17. . . Axb2! 1 8. 'Wxb2 ygxe4+ 1 9. yge2 ttJd3+
20. d2 ttJf4+ and Black wins the queen)
16 . . _ygc6 1 7. ttJxc5 ygxc5 ! . lvanchuk offers
1 1 . . . ygxd1 + 1 2. xd1 axb5 1 3 . Axg7 (af
ter 1 3 . AxeS lDd7! 1 4 . A.xg7 g8 1 5 . lbxb5
xa2 1 6. Axh6 xb2 Black has quite good
cou nterplay) 13 . . . g8 1 4 . Axh6 Wh ite wins
a pawn without sufficient compensation
for the opponent.

67
 Chapter 2 The Way to the Top

1 2 . Wf3! strong initiative; 1 6. tLlxb4 Y#d4! 1 7. tLlxc6

Y#e4+ 1 8. 'i!?d1 Axc6 is also advantageous
Wh ite w i n s the exchange, but M ag n u s to Black) 15 . . .Ad7 1 6. 'Wd6 bxc3 1 7. bxc3
manages t o find defensive resources. e4 1 8. 0-0-0 Ae7 (1 8 . . . Axc3 1 9. Ae2 Y#c8
with the idea of .. . Aa4 is also possible)
12 . axb5 1 3. Wxa8 b4!
1 9. Wd4 Aa4 (1 9 . . . Af6 2 0 . Wxe4 We?
a b c d e f 9 h 2 1 . Y#a8+ Ac8 is also good) 20. Wxd8+
8 .i. 'IV. .I 8 Axd8 21 . gd4 Ac6 22. Ae2 Ac7 followed
by . . .'i!?e7. Carlsen assesses this position
7
7 as favou rable for Black.
6 .i. 6
5
5 14 .. J.d7 1 5. xf6+
4
B 4
Wh ite hurries to deform Black's kingside
3
ttJ 3 pawns, whereas he also had other pos
2 B B B B B 2 s i b i l ities. True, 1 5 . tLld6+ is u n convinc
1 11 l:, 1 i n g : 1 5 ...'i!?e7 1 6 . tLlb5 (or 1 6 . gd1 tLlc6
a b c d e f 9 h 1 7. 'Wxd8+ gxd8 1 8 . tLle4 tLld4) 16 .. .tLlc6
1 7. Wxd8+ gxd8 1 8. 0-0-0 tLld4 with suf-
The moment of truth for the assessment ficient com pensation for the exchange.
of the open ing idea devised by Jobava, But 1 5. gd1 , keeping the exchange on f6
who apparently underestimated this pos- in reserve, deserved serious consideration:
sibil ity, expecting only 1 3 . .. tLld7 1 4. tLlxb5 1 5 . .. tLlc6 (1 5 . . . Ae7 1 6. c5 0-0 1 7. Wa7 fa-
0-0 1 5 . 0-0-0 with advantage to Wh ite. vou rs White) 1 6. Y#xd8+ 'i!?xd8 1 7. c5 bxc5
H owever, Carlsen has planned a m o re 1 8 . tLlxc5 tLld4 1 9. tLlxd7 'i!?xd7 20. Ad3 ,
active role for the knight, by prepari ng a and Wh ite's position is preferable (Carlsen).
route for it to d4, from where it controls 1 7. tLlxf6 gxf6 1 8 . g3 'i!?c7 1 9. Ag2 tLld4
both wings. 20. f4 Ac6 21 . 0-0 also looks q u ite good,
diverting the knight to f5 for the defence
14. e4
of the f-pawns.
After 1 4. Y# x b 8 b x c 3 1 5. g d 1 A d ?
15 . . gxf6 1 6. c5
1 6. Y#xd8+ 'i!?xd8 1 7. bxc3 e 4 t h e pair of
strong bishops fu lly com pensates for the Wh ite is aiming for a clear pawn structure
sacrificed exchange. on the queenside, but 1 6. 0-0-0 was more
1 4 . \Wa4+ looks stro n g er: 1 4 ... tLlc6 logical , after which Carlsen was i ntending
(i n the event of 1 4 . . . Ad? 1 5 . Y#xb4 tLlc6 to play 16 . ..tLlc6 1 7. Y#a4 (of cou rse, the
1 6 . Y#d6 Ae7 1 7. Wd 1 tLld4 B l ac k h as queen exchange 1 7. \Wxd8+ 'i!?xd8 followed
some com pensation for the exchange, but by . . . tLld4 would satisfy Black) 1 7. . . 'i!?e7,
Wh ite's chances are nevertheless prefer and with a su bsequent . . . Wa8 Black acti
able) 1 5. Y#xc6+ (activity with 1 5 . tLld5 is vates his forces.
du bious: 1 5 . . . Ab7 1 6. Wb5 0-0 1 7. Y#xb6
16 .. bxc5 1 7. Wa7 c6 1 8. Wxc5 d4!
exd5 1 8 . Wxb7 tLld4, and the powerfu l
knight on d4 enables Black to develop a After carrying out this key defensive ma-

68
 B. Jobava - M . Carlsen
noeuvre, Black can consider that he has
solved his opening problems.
a b c d e f 9 h
8 .I 8
7 7
6 6
5 5
4 4
3 ation favou rs Black, since he can develop
a b c d e f 9 h
19. 0-0-0?
Jobava chooses a bad post for his king,
where it m ay came under attack by the
active black pieces. But it was not an easy
choice. For exam ple, if 1 9. Ad3 Black
can fight for the i n itiative: 19 . . . 'Wa8! (after
19 . . .'We? 20. 'Wb6 0-0 21 . 0-0 Ac6 22. l:!fc1
Black would have problems with the de
fence of his b4-pawn) 20. 'Wxb4 l:!g8! 21 . a3
(if 21 . 0-0-0, then 21 . . . 'Wxa2 is possible,
not fearing 22. 'Wb8+ <!;}e? 23. 'Wxg8, since
Black is g uaranteed a d raw by perpetual
check: 23 . . .tDb3+ 24. r!;}c2 tDd4+ 25. r!;}c1
tDb3+, as after 25. r!;}c3? 'Wb3+ 26. r!;}d2
'i'xb2+ 27. r!;}e1 e4! Wh ite is in danger of
losing) 21 . . . l:!g4 ! . An i mportant intermedi
ate move, t h e virtues o f which are revealed
in an assessment of the d i rect capt u re
21 . . . l:!xg2 22. 0-0-0 l:!xf2 23. l:!hg1 , with an
unclear game. Now, after the queen moves,
the rook i nvasion follows in a more favou r
able situation: 22. 'Wd6 l:!xg2 23. 0-0-0
xf2 24. l:!hg1 'Wc6+! , or 22. 'Wc3 l:!xg2
23 0-0-0 l:!xf2 24. l:!hg1 'Wc6! . Evidently
White should h ave restricted h i mself to
the modest 1 9. f3, when Black could again
offer the exchange of queens by 19 . . . 'We7.
69
------
ctJ
1 9 . . .'We7!
Magnus demonstrates a deep understand
ing of the position. The endgame with the
powerfu l k n i g ht on d4 is advantageous
to Black .
20. 'Wxe7+
After 20. 'Wb6 0-0 21 . Ad3 Ac6 the situ
an attack on the white king by playing his
rook to a8 and his q ueen to d5.
20 . . mxe7 21 . mb1 J.c6 22. f3
.
If 22. l:!g1 again 22 . . . l:!g8 is possible, with
the idea of . . .tDf3.
22 . . J g8 23. Eic1 f5 24. Eic4 Eib8!
Subtle conduct of the endgame - now it
is not easy for Wh ite to complete his de
velopment. I n a superior position it is i m
portant t o maintain t h e tension, b y forcing
the defending side to solve the most d if
ficult problems. The tem pting continuation
24 . . . AdS 25. l:!xb4 tbxf3! 26. a4 would have
g iven Wh ite some chances, associated
with his pair of connected passed pawns.
25. b3
2 5 . A d 3 is not poss i b l e because of
25 . . . Ab5. The attem pt to develop the h1 -
rook does not work, since if 25. h4 there
is 25 . . . f4 (25 . . . h5 26. l:!h3 f4 27. g3 l:!g8 is
also good) 26. h5 r!;}d6 27. l:!h4 tDfS, and
the rook has to return home.
2s . . . t4 2s. mb2 ts
The pawn mass d isplays its character! The
pair of central pawns is irresistible.
(see next diagram)
 Chapter 2 The Way to the Top
-- ---- ------------------

a b c d e f g h 31 . . J:tf8 32. gg4 gfs was also good, tying

8 .I 8 the rook to the h5-pawn , followed by . . .
7 7
e6-e5 and . . . ggs.

6
j_ 32. Elg4 e3 33. Elg6
5 5
The position also can not be held after
4 4
33. ge1 tDf5 34. Ae2 gd8 35. gg6 Ae4,
3 3
while if 33. Ad3 there is the decisive 33 . . .
e2! 34. ge1 (34. gg6 gc8!) 34 . . . gc8 35. gg6
gc3 .
a b c d e f g h
33 . . . Ae4 34. Elxh6 Elg8
27. h4
M agnus appears to be enjoying the posi-
tion, avoiding the win of a piece by 34 . . . e2.
If 2 7. Ad3 t h e re wou l d h ave fol lowed
27. . . AdS 28. gc7+ d6 29. gg7 e4 with 35. Elc5+ d6 36. Elc4 dS 37. c1
advantage to Black. e2 38. El xd4+ xd4 39 . .lxe2 Elxg2
40. Ad1
27. . . d6 28. h5 AdS 29. Elc1 e4
30. fxe4 After 40. d1 Black casts a m ating net
with 40 . . . e3 ! .
If 3 0 . Ac4, then 3 0 . . , gg8 31 . g h 2 e 3 was
possible. 40 . . . Elxa2

30 . . .fxe4 31 . Elh4 eS Wh ite resigned .

The World Cup in Khanty- M ansiysk - analogous to the world championship on the
knock-out system - was another reform of the world cham pionship, with a slight low
ering of its status. The top ten in the tournament gained the right to take part in the
Candidates tournament of the next ind ividual world cham pionshi p cycle in 2007. The
knock-out event of 1 28 grandmasters, held in a remote part of Russia, proved to be a
turning point in Carlsen's career on the eve of his 1 5th birthday. For M agnus this was
a second test in matches with the strongest players in the world. True, the first one -
in 2004 - had ended in the very first rou nd, but the maturing Carlsen , who had joi ned
the group of 'twenty-six hundreds' on the world rating list, now had the right to hope
for a more successfu l performance.
This was how his performance was assessed by the experienced grandmaster Viorel
Bologan: 'The revelation of the tournament was Carlsen . He is d istinguished by his
classical method of playing - he plays com petently and calculates wel l . For the mo
ment he still displays a tendency to play for l ittle traps, but i n principle this is m i n i mal.
Over the last two years he has been merci lessly beaten i n various tournaments, and
unexpected ly this path to success has proved the shortest. Now he himself will do the

70
 M. Carlsen - F. Amonatov

beating.' I n his preparations for the tou rnament Mag n u s had greatly improved his open
ing groundwork. ' It i s apparent that he i s doing such work on his open i n g repertoire',
Mikhail G u revich said in adm i ration, 'that experienced grandmasters don't know what
to play agai n st h i m ! ' The youngster had also greatly improved his endgame techniq ue.
I n contrast to Libya, i n Russia Carlsen played seven stages: fou r in the main tou rna
ment and three i n the add itional fight for qual ification to the Candidates tou rnament.
His res u lt could have been even better, since he missed a real chance to go through
to the quarter-fi nal of the main tou rnament, when in his match with Evgeny Bareev h e
lost a rapid game which was practically decided. B ut even without t h i s h i s l i st o f wins
was im pressive: among those defeated were Zu rab Azmai parashvili, Farrykh Amonatov,
Ivan Chepari nov, Joel Lautier, Vladimir Malakhov and Gata Kamsky.

Game 1 4 9. t'Lld5
M. Carlsen - F. Amonatov
World Cup, Khanty- Mansiysk 2005 9. Ag5 i s regarded as the main conti n u
Sicilian Defence [ 890] ation.
1. e4 c5 2. t'Llf3 d6 3. d4 cxd4 4. t'Llxd4 9 . .t'Llxd5 1 0. 'Wxd5 t'Llc6
.

f6 5. t'Llc3 a6 6 .te3 e5 7. t'Llf3

After 10 . . . Ab7 it i s not easy for Black to

The mai n continuation is 7. tDb3 Ae6 8. f3 com plete his development: 1 1 . b3 Axe4
with the development of an attack on the 1 2 . Ac4 b7 1 3 . 0-0-0 Axf3 (if 13 . . . Ag6
kingside. The move i n the game i s also
there fol l ows 1 4 . tD h 4 A h 5 1 5 . f3! tDc6
very popular and indicates that Carlsen is 1 6 . Axb6 o r 1 6 . xb6 with advantage
aiming for positional play.
to Wh ite) 1 4 . g xf3 tDc6 1 5 . <!>b1 ! (1 5. Ad5
7 . :Wc7 8. a4 b6
.
b5 1 6. f4 is also good) 15 . . . !:!b8 (1 5 . . .tDa5
1 6. Ad5) 1 6. Ad5 with an obvious positional
Black radically prevents a4-a5 , but this
advantage for Wh ite.
move has d rawbacks, which Carlsen i s
able t o disclose. 8 . . . Ae7 9. a 5 0-0 i s more
often played . For example, the game Jang
Jong -Sakaev, (China- Russia match 2004)
conti n u ed 1 0 . Ae2 tbc6 1 1 . Ab6 d7 6
1 2. tDd2 tDb4 1 3. tDc4 d5! 14. exd5 tDfxd5
5 5
1 5. tDxd5 tDxd5 1 6 . Af3 Ab4+ 1 7. <!>f1 !
tDxb6 1 8 . tD x b 6 xd 1 + 1 9 . !:! x d 1 !:! b 8
4
fj, 4

3 3
20. !:!d5! !:!e8? (20 . . . Ae6!? 21 . !:!xe5 !:!fd8
2
came i nto consideration, with s ufficient
compensation for the pawn) 2 1 . c3 Af8
22. tDxc8! !:!bxc8 23. Ae4! with advantage a b c d e f g h

to Wh ite in view of the weakness of the a6/

b7 pawn pair. 1 1 . t'Llg5

71
 Chapter 2 The Way to the Top

The less ambitious 1 1 . \Wd2 Ae7 1 2 . Ac4 14 . . . Eia8

0-0 allows Black to complete his develop
ment. The game Kir. Georgiev - l strates
cu (Oh rid 2001) conti nued 1 3 . Ad5 Ae6
1 4. 0-0 E!ac8 1 5 . E!ac1 tbb8! 1 6 . E!fd1 tbd7
with equal chances.
11 . . . Eib8
It is not possible for B l ack to complete
h i s q u eenside d eve l o p m ent normally -
1 1 . . .Ab7? 1 2. Axb6! \Wxb6 1 3 . \Wxf7+ d8
1 4. a5! and if 14 . . . tbxa5 1 5. tbe6+ White
g ives mate in three moves.
1 2 . Ac4
Carlsen's favou rite p roced u re from h i s
youth: i f i t is possi ble t o p u t t h e k i n g under
fire, he does this.
12 . . . d8 1 3. Wd3 h6
Black is i m patient to d rive away the an
noying knight. To be fai r, it should be men
tioned that the theory of this variation was
largely unexplored . 1 3 . . . Ae7 is stronger,
when the fol lowing line is not dangerous:
14. Axa6 Axa6 1 5 . "Wxa6 "Wxc2 1 6. Wb5+
Wc6 1 7. 0-0 "Wxb5 1 8 . axb5 Axg5 1 9. Axg5
tbe6 20. Ae3 tbc7 21 . E!fc1 tbxb5 22. Axb6
d7 with a reasonable endgame for Black
(Ku laots - Leyva, Olympiad, Turin 2006).
14. f3
1 4 . . . Ae7 is no longer possible, si nce after
1 5 . Axa6 Axa6 1 6. "Wxa6 Black's strategic
deficiencies are obvious. Let us conti nue:
16 . . .Wxc2 1 7. "Wb5+ "Wc6 (1 7. . . f8? 1 8. a5)
1 8 . e2! 0-0 1 9. "Wxc6 tbxc6 20. E!hc1 E!fc8
2 1 . E!c3 tbb4 22. Axb6, and Wh ite wins a
pawn. Therefore the a6- pawn has to be
defended , but the return of the rook to a8
is an ad m ission that B lack has problems
with his queenside development and it is
not easy for him to coordinate his pieces.
If 14 . . . "Wb7 there can follow 1 5. 0-0 Ae7
1 6. Ad5!? with enduring pressu re on the
q ueenside: 1 6 . . ."Wc7 1 7. tbd2 0-0 1 8 . Wc4
"Wd7 (not 1 8 . . . Wxc4? 1 9 . tbxc4 b5? be-
cause of 20. Aa7) 1 9. "Wb3 tbe6 20. tbc4.
1 5. d2 e6
After 15 . . . Ae7 1 6. Ad5 Ab7 1 7. 0-0 0-0
1 8 . tbc4 A x d 5 1 9 . Wxd5 E!b8 20. E!fd 1
White has an obvious spatial advantage.
1 6. AdS labS
1 6 . . . A b7 1 7. 0-0 Axd5 1 8 . exd5 tbc5
1 9. "Wc4 also favours White - he is th reat
ening 20. b4.
1 7. Wc4!

a b c d e f 9 h

8 K -*. -*- K8
7
'if .t .t 7

s .t .l .l l.&\ .t s
5 5 5
i;, .t 5

4 [j,
4 [j, if, [j, 4
VJII [j, 4

3
tt:J 3 3
if, 3

2 [j, [j, [j, fj, fj, 2 2 [j, [j, t;jj [j, [j, fj, 2
1 1 1 1
a b c d e f 9 h a b c d e f 9 h

72
 M. Carlsen - G. Kamsky

The move of a mature master! Si nce the 27. e4!

majority of the black pieces are undevel
oped , Carlsen exchanges t h e o n l y active Deep calculation! All Black's trickery leads
one - the queen, with the hope of convert merely to a red uction i n the material.
ing his spatial advantage in the endgame.
27. . . Ae6 2S. Axd6 Axe4 29. lieS! lieS
17. . . e5?!
After 29 . . . Axd6 30. xc6 the elegant fork
This attem pt to avoi d the exchange of
proves decisive.
queens leads by force to the loss of the ex
change. But in the endgame too Black can 30. l:ixe4 Axd6 31 . iia6 JabS
not hold out: 17 . . .'xc4 1 8. CL!xc4 CL!f4 (not
The alternative 31 . . .CL!e7 32. xc8+ CL!xc8
18 . . b5? 1 9. axb5 axb5 20. Aa7) 1 9. 0-0-0!
.

33. a8 i s equally hopeless.

Ad? (1 9 . . . CL!xd5 20. xd5 1eads to the loss
of a pawn) 20. Axf4 (20. CL!xb6 Ag4 21 . d2 32. g3 d4 33. l:ixd6 ii b1 + 34. 'i!?g2
tiJxg2 22. a5 CL!xe3 23. fxe3 i s also good) ii b2 35. e3 e2 36. 'i!?t1
20. . exf4 21 . e5! b5 (if 2 1 . . . d xe5 there fol
.

Black resig ned .

lows 22. CL!xe5 or 22. Axf7+) 22. CL!xd6+
Axd6 23. exd6, and Black cannot avoid
loss of material.
Game 1 5
1S. b4! b5 1 9. axb5 axb5 20. We3
M . Carlsen - G . Kamsky
ti:}e6 21 . Wxe7 xe7 22. Aa7 xd5
World Cup, Khanty- Mansiysk 2005
23. AxbS xb4 24. 0-0!
Sicilian Defence [ 843]
Including the second rook in the attack on
the queenside. 1 . e4 e5 2. f3 e6 3. d4 exd4 4. xd4
a6 5. e3 We7 6. Ad3 b5 7. 0-0 Ab7
24 . .e6 25. iifb1 Ae7
S. We2 e7 9. iie1
.

At last Black has managed to develop this

bishop. In the event of 25 . . . CL!xb8 26. aS a b c d e f 9 h

tiJc6 27. xc8+ 'i!?d7 28. aS its release 8

Kl.l\ e .t..
might lead to further losses. 7 .t.. 'if .l l.l\ .l
26. J:ixb5 0-0 6 .l .l 6
a b c d e f 9 h 5 5

8
.a. Ke 8 4
{jj /j, 4

7
.l .l 7 3
{jj 3

6 1.1\ .l 2 /j, /j, /j, 'i! 8 8 8 2

5
11 .l 5 1 : 11
4 8 4 a b c d e f 9 h

3 3
Th i s novelty, apparently saved up for such
2 8 {jj an i m portant game, best c haracterises
1 11 Carlsen's u n d e rstan d i n g of the game,
a b c d e
based on a knowledge of the classics.

73
 C hapter 2 The Way to the Top

Taking account of B l ack's retarded de-
velopment, he prepares the knight th rust
to d5. For this he needs to rei nforce the
pin on the e-fi le.
9 . . . bc6
The natural way of deve l o p i n g . If 9 . . . tLlg6
all the same there fo l l ows 1 0 . tLld 5 ! 'Wd 6
(1 0 . . . exd5? 1 1 . exd5+ tLle7 1 2 . tLlf5) 1 1 . c4
(th e ag g ressive 1 1 . 'W h 5 is j u stified af
ter 1 1 . . .exd5 1 2 . exd5+ d8 1 3 . Ag5+
c7 1 4 . c4 with an attack on the k i n g ,
b u t B l ack c a n cont i n u e h i s development
with 1 1 . . . tLl c 6 ! ?) 1 1 . . . bxc4 1 2 . A xc4 f6
( h e re too 1 2 . . . e x d 5 ? 1 3 . e x d 5 + d 8
1 4 . 'We8+ c7 1 5 . 'Wxf7 'Wf6 1 6 . tLle6+
b6 1 7. Ae3+ is dangero u s for B l ack; if
12 . . . tLlc6 there i s the u n p l easant 1 3 . tLlf5)
1 3 . Ad2 tLlc6 with the idea of 1 4 . tLlxc6
Axc6 1 5 . J;!ad 1 Ab5.
If 9 . . . g 6 B l ack has to rec kon with
1 0.tLldxb5!? axb5 1 1 . tLlxb5 'We5 1 2. Ae3
tLlbc6 1 3 . f4.
1 0. xc6 xc6
10 . . . dxc6 was more circumspect. To judge
by the game, Kamsky either underestimat
ed Wh ite's i n itiative after the knight th rust
to d5, or he altogether did not consider it.
a b c d e f 9 h
8 I. .1. I. 8
'iV
7 .1. .l .l .l .l 7
6 .\ .l 6 7
5 .l 5
4
8 4
3
tb 3
The World Cup form ula with a mini-match
of two classical games rai sed the val ue of
each of them . Therefore it i s hard to con-
demn Kamsky for taking a risk. Otherwise
he would have retreated with 1 1 . . .'Wd8 .
1 2 . exd5+ e7 1 3 . c4!
An i mportant move for the understanding
of this position. Since it is unreal for the
black king to move to its own wing, Wh ite
must prepare to meet it on the q ueenside.
13 . . . b4
Opening the position by 1 3 . . . bxc4 14. Axc4
allows the wh ite rook to j o i n the attack
along the c-fi le, and if 14 . . . 'Wd6 Black has
to reckon with 1 5 . g3 followed by Af4.
14 . .tg5 f6
The alternative 1 4 . . . 'Wc5 i s parried by
the seem ingly modest offer to exchange
the queens - 1 5 . 'We3 'Wxe3 1 6 . l;!xe3 f6
(1 6 . . . d6 1 7. Axe? Axe? 1 8 . J;!ae1 ) 1 7. d6
f7 (1 7. . .fxg5 1 8 . dxe7 f7 1 9. J;!ae1 J;!e8
20. exf8'W+ l;!hxf8 21 . l;!xe8 l;!xe8 22. l;!xe8
xe8 23. Axh7 1eads to the loss of a pawn)
1 8. dxe7 Axe? 1 9. Af4 with the better end
game for Wh ite.
1 5. 'Wh5+ g6 1 6. Wf3 fxg5 1 7. 'Wf6
0-0-0
a b c d e f 9 h
8 .l. 1. .1_ J. 8
'iV .l .\ 7
6 .l VJ!i .l 6
5 8 5
4
.l 8 4

2 888 VJ/i 8 8 8 2 3
3

1 Ir It: 2 88 888 2
a b d f h
1 Ir
9

c e
It:
1 1 . d5! exd5 a b c d e f 9 h

74
 M. Carlsen - G. Kamsky

18. 'Wxh8 25 . . . gxf5

After the tempting 1 8. d6 there is 18 . . .g8! 25 . . . Wxf5 loses to 26. l::! e 8+ a7 27. Wa3!
1 9. Wxh8 Axd6 and the i n itiative passes l::! c7 28. We3+.
to Black. From the material point of view,
the outcome of the open ing operation is 26. Iie6 f4 27. Iice1 Iid8 28. 'Wb6 Iic8
not wonderfu l for Wh ite. But as regards the 29. b4!
assessment of the position, it has been an Wh ite's main threat i nvolves the invasion
undoubted success: h i s pieces are domi of his rook at e8.
nant, whereas for the moment the black
bishops are shut i n . 29 .. fxg3

18 . . .'Wd6 If 29 . . . l::! xc4 there fol lows 30. b5 axb5

If 1 8 . . . h6 there follows 1 9. Wf6.
19. Ae4
With the th reat of c4-c5.
19 .. /b8 20. Iiac1 'Wf4 21 . g3 'Wf7
22. 'Wd4
22. We5+ followed by Wxg5 is also good
enough, but Carlsen is hoping for more.
22 . . . d6 23. 'Wb6 Iid7 24. 'Wxb4 t'Llf5
31 . l::! e 8+ l::! c 8 32. Wxb5 Wxd5 33. l::! x c8+
xc8 34. l::! e 8+ c7 3 5 . Wxd5 A x d 5
3 6 . l::! xf8 , o btai n i ng a tec h n ically won
ending.
3 0 . h x g 3 h5 31 . b5 'Wc7
Si nce Black cannot be satisfied with 31 . . .
axb5 32. Wxb5 when there is no defence
against 33. l::! e 8, he is forced to agree to
an endgame, where his bishops are sti l l
out of play.

a b c d e f 9 h 32. 'Wxc7+ xc7 33. a4 axb5 34. axb5

8
..t 8 b6 35. Iie8
7 ..t .1. 'iV 7
35. l::!f 6! is more forcefu l , when Black i s
6 . 6
practically i n zugzwang.
5 B 5
35 . . . g7 36. Ii8e6
4
VJ!I B 4

a b c d e f 9 h

8 .1. 8

7 ..t ..t 7
a b c d e f 9 h
6
: 6

25. Axf5 5 B B
An amazi ngly pragmatic approach to the
4
B 4

3 3
solution of the position. Wh ite exchanges
2 2
the active knight in order to invade with his
rook at e6, not paying any great importance
to the undoubling of the black pawns. a b c d e f 9 h

75
 Chapter 2 The Way to the To P __________

..;.

36 J ad8?
. 40. xc3 Axd5 41 . d1 Aa8 42. a3 d5
(after 42 . . . xb5 Wh ite wins by 43. da1
Black should have decided on 36 . . J xc4
Ac6 44. a5+, while if 42 . . . Ab7 there fol
37. xd6+ c5 38. g6 Ac3 39. d1 Axd5 lows 43. e3 and 44. e6) 43. da1 Ab7
40. xg5 Ad4 41 . xh5. It is not easy for 44. e3 d7 45. e5 g7 46. b1 , although
h i m to escape from the p i n , but the ad
here too White should be able to convert
vance of the f- and g-pawn pair requ i res
his exchange advantage.
time, which gives h i m some practical sav
ing chances. 39. xb7 f8 40. Jac7+ ckxd5 41 . d1 !

37. c5+! The final subtlety. 42. b6 is threatened.

An unexpected thrust, leading to the win
41 Ja b8 42. ag7 94 43. JagS+ ckc4
.

of one of the bishops.

44. Jaxh5 Ac5 45. JagS Jaxb5 46. Jaxg4+
37 .'itxc5 38. e7 Ad4
.
ckc3 47. ckg2 Aa3 48. Jagd4

38 . . . Ac3 was more resil ient: 39. c7+ b6 Black resigned .

Qual ifying for the Candidates series implies contacts with the world elite and the reach
ing of a new level of mastery. The chess world was expecting a great deal of Carlsen ,
but if you raise you r potential in slow steps, you should remember that time does not
work long for the young, and you can easily dissolve i nto the overall mass, as occurred
with a n u m ber of his talented contemporaries. But to gain in class, it is im portant to
make the next step. The step into chess high society. For the 1 5-year-old g rand master
a rating of 2570 on the results of 2005 gave every justification for this.
At that time the chess calendar offered two opportun ities to progress. The fi rst
was to qualify for the main Wijk aan Zee tournament. Here there was no problem, as
Magnus was a desired g uest at the most famous festival of the year. Another way of
breaking i nto elite society was offered by the wel l -known Dutch chess patron Joop
van Oosterom, who sponsored not only his trademark ' M elody Amber' tournament,
but also the ' Rising Stars' competition, where veterans tested talented young players,
the best of whom would qualify for the main tournament. Carlsen made use of both
opportun ities.
I n January 2006 he again , for the third year i n a row, played in Wij k aan lee. On this
occasion the organisers had invited a stronger field for the B tournament compared
with the previous year, raising it to FIDE category 1 5, where the brightest young talents
were guaranteed places. But such rivalry was to the advantage of Magnus, who had
rested and was eager for a fight. The battle for qual ification to the A tournament was
uncommonly d ramatic. It appeared that on this occasion Magnus would more than
hold his own : after 7 rounds he was confidently leading with 6 points, after gaining
a n u m ber of wins i n splendid attacking style. But he then slowed down with a series
of d raws, and in the 1 1 th round he was caught by Alexander Motylev, who won their

76
 M. Carlsen - A. Beliavsky

individual encounter. It was these two who were victorious in the tournament, after
overcoming the hurdle of the last round, where they had to defeat the lady players
Kateryna Lahno and H u m py Koneru . True, even with a share of 1 st-2nd places (9 points
out of 1 3) there were stil l g rounds for disappointment, since on the tie-break fi rst prize
went to Motylev. But the festival organisers reassu red Carlsen , saying that they had
decided to allow both players i nto the A tournament in 2007.

Game 1 6 7. . . Ac5?!
M . Carlsen - A. Beliavsky
Wijk aan Zee B 2006 Here this activity is i nappropriate, si nce it
Ruy Lopez {C91] al lows Wh ite to develop his knight at c3,
creating the threat of pinning the f6-knight
1 . e4 e5 2. f3 c6 3 . .tb5 a6 4. Aa4
(with the idea of tt::l c 3-d5), after which the
f6 s. o-o b5 6. Ab3 .tb7 7. d3
absence of the bishop from e7 is very per
a b c d e f g h ceptible. The main continuation is 7. . . Ae7,
8 i. 'ir' .t i. 8 as Bel iavsky had played in all his preceding

7 .t .l .l .l .l .\ 7 games. To the q uestion, why did he play

6 .\ 6 7. . . Ac5, Alexander merely sh rugged his

shoulders.
5 .l .l 5
8. c3
4 4

3 Jb ttJ 3 For the moment the i m mediate pin 8. Ag5

2 2
1 1:l ttJ Jb iV
:
a b c d e f g h
This modest pawn move is a kind of anti
Arkhangelsk variation, directed against the
active development of the bishop at c5,
which is Black's main response to one of
the basic contin uations 7. e1 . Another
such continuation is 7. c3, which was espe
cially popular at the end of the 20th century.
The greatest stir at that time was created
by the game N u n n - Bel i avsky (Am ster
dam 1 990), which continued 7. . .tt::l x e4 8. d4
tt::l a5 9. Ac2 exd4 1 0. b4 tt::l c 4 1 1 . Axe4
Axe4 1 2. e1 d5 1 3 . tt::l x d4 c5! 1 4 . bxc5
Axc5 1 5. f3 0-0 1 6. fxe4 dxe4 1 7. *h1 e8
18. Ae3 tt::l x e3 1 9. xe3 b4, and Beliavsky
won after a very sharp fight.
is premature, since after 8 . . . h6 9. Ah4 0-0
1 0. tt::l c 3 Ae7! Black retreats his bishop,
obtai ning an easy game. N ow, however,
it is not easy for h i m to choose a moment
to castle.
8 . . . d6
A natural and freq uently-employed move.
But 8 . . . 0-0!? also comes i nto considera
tion, with the idea after 9. Ag5 h6 1 0. Ah4
Ae7 of reaching a position from the varia
tion given in the previous note.
9. a4
A standard fl a n k attac k , w h i c h at the
same time solves the problem of retain
ing the light-sq uare bishop. After the mod
est 9. Ae3 0-0 1 0. a3 tt::l d 4 Black has no
problems.

77
 Chapter 2 The Way to the Top

9 . . . a5 a2 or effect a pin on g4. But in the g iven

A dubious move - the knight moves to the
edge of the board . The main conti nuation
is considered to be 9 . . . b4. Si nce the time
of the Kasparov - Anand world champion
ship match i n 1 995 it has been considered
that 1 0. !Dd5 1eads to the exchange of this
knight (1 0 . . .!Da5 1 1 . !Dxf6+ 'Wxf6), which
eases Black's defence: 1 2. Aa2 h6 13. c3
bxc3 14. bxc3 0-0 1 5. Ae3 ad8 1 6. b1
Ac8 (with the idea of . . . Ag4 or . . . Ae6)
with rou g h ly equal chances. More com pli
cated play resu lts from 1 0. !De2 0-0, and
if 1 1 . a5!? (preventing . . .!Da5), then 1 1 . . .
d5!? i s possible.
specific situation, for the moment this ma
noeuvre i s premature. Black should have
decided on 1 1 . . . 0-0!? 1 2. Ag5 h6 1 3 . Ah4,
when 13 . . . Ac8! i s now good .
1 2 . c3
If 1 2. Ae3 Black eq ualises by 1 2 . . . Axe3
1 3 . fxe3 Ae6 ! .
1 2 . . . bxc3 1 3 . bxc3 Ab6
A p rophylactic b i s h o p retreat , s i nce in
the event of 13 . . . Ae6 the 1 4. d4 advance
is made with g a i n of tem po, and after
1 4 . . . Axa2 1 5 . xa2 Ab6 1 6. Ag5 Wh ite
has an enduring advantage.

1 0 . Aa2 14. g3

No true 'Spanish' player would part with his Now if 14. Ag5 there can follow 14 . . . h6
trademark bishop for the sake of an extra, 1 5 . Ah4 g5!? (1 5 . . . 0-0?! 1 6. !Dg3!) 1 6. Ag3
dou bled pawn - 1 0 . axb5 !Dxb3 1 1 . cxb3 0-0.
a x b5 1 2 . x a8 A x a8 1 3 . !Dxb5 h6, for a b c d e f 9 h
which Black has s ufficient com pensation
in the form of the two bishops.
8 .!. .i.'' .!. 8
7 j. j. j. j. 7
a b c d e f 9 h
6 j. .i. j. 6
.!. 'if
.!. 8 5 5
j.
8

7 .i_ j. j. j. j. 7 4
[j, [j, 4
6 j. j. 6 3
fj, [j, t:D t:D 3
5 j. .i_ j. 5
[j, [j, /j, 2
J_

4
[j, [j, 4
.:r
3
f:D B ED 3
a b c d e f 9 h

2 /j, /j, /j, /j, /j, 2

1 1:r VJH 1:r 1 4 . . . Ae6?!
a b c d e f 9 h
Black decides against 14 . . . 0-0, feari ng the
10 . b4 1 1 . e2 Ac8?! pin 1 5 . Ag5 . But perhaps the devil is not
as terri ble as he's pai nted? For example,
On the whole this is a thematic bishop ma the game De Firmian - Lugo (San Diego
noeuvre, by which Black takes control of f5 2004) continued 15 . . . h6 1 6 . Ah4 Ag4 1 7. h3
(for where the knight i s aimi ng) and gains Axf3 1 8 . 'Wxf3 g5 1 9. tDf5 gxh4 20. !Dxh6+,
the opportunity to neutral ise the bishop on and here, by playing 20 . . . g7! 21 . !Df5+

78
 M . Carlsen - A. Beliavsky

<.!?h8, Black wou ld have held the position. a b c d e f 9 h

Instead of 1 9. f5 Vi ktor M i khalevsky rec- a .I 'YJV .. . 8
ommends 1 9. h5! xh5 20. '1Wxh5, which
is correct after 20 . . . gxh4? 21 . '1Wg6+ h8
7 I. 1. 1. 1. 7

22. '1Wxh6+ g8 23. h1 ! with dangerous

6 I. ..t I. 6

threats such as f2-f4. But 20 . . . '1Wf6! 21 . Ag3

5 I. 5

<.!?g7 22. l:!ab1 '1Wg6 is stronger, when Black 4

D. D. D. 4

successfu lly defends (Beliavsky). 3 D. tb tb 3

2
.: D. D. D. 2

15. d4! J.xa2?!

.:
a b c d e f 9 h
If was not yet too late to castle - 1 5 . . . 0-0
1 6. dxe5 dxe5 1 7. xe5 '1Wxd1 1 8 . l:!xd1 1 1. Ag5!
ad8 1 9 . l:!e1 l:!fe 8 , and B l ack's active
Wh ite has been able to implement the main
pieces g ive h i m q u ite good compensa-
conceptual idea in the 7. . . Ac5 variation.
tion for the pawn.
The pin on the knight decides the game.

16. !axa2 0-0?! 1 7. . . exd4

Now Wh ite's attack becomes i rresistible,
but it is already hard to offer Black good
advice. For example, he is u nable to return
the a5 -knight to the defence - 1 6 . . . c6
17. a5! x a5 (1 7 . . . Axa5? 1 8 . 'Wa4 '1Wd7
1 9. d5) 1 8. d xe5 d xe5 1 9. l:!d2 with numer
ous threats. The defence is also not eased
by the inclusion of 1 6 . . . exd4!? 1 7. cxd4 0-0
1 8. Ag5 h6 1 9. Ah4, when in view of the
threat of h5 Wh ite's attack is very dan
gerous. For exam ple: 19 . . . c5 20. h5 g5
21 . /:Dxg5 xh5 22. '1Wxh5 hxg5 23. Axg5 f6
24. '1Wg6+ h8 25. l:!a3! or 1 9 . . . l:!e8 20. '1Wc1
(with the th reat of h5) 20 . . .b3 2 1 . '1We3
xd4 22. xd4 Axd4 23. '1Wxd4 g5 24. f3!
gxh4 25. f5, and Black's position is de
stroyed (Mi khalevsky).
Things are not changed by 17 . . . h6 1 8 . Ah4.
Now 1 8 . . . exd4 1 9. cxd4 leads to a posi
tion considered i n the n ote to 16 . . . 0-0,
while if 18 . . .c6 there follows 1 9. h5! g5
20. xg5! xh5 2 1 . '1Wxh5 hxg5 22. Axg5
f6 23. Ah6, and after the capture of the
rook Wh ite's attack remains dangerous.
18. h5! dxc3 1 9. h4!
Wh ite's attack is so strong, that h e can
demonstrate the strength of his caval ry.
But also after the prosaic 1 9. Axf6 gxf6
20. '1Wc1 h8 21 . '1Wxc3 l:!g8 22. xf6 B lack
would not have held out for long.
19 . . . h8 20. f5
Black resigned.

Some commentators gave an enth usiastic assessment of this game. This wou l d
seem t o be a d e l u s i o n . Beliavsky's play was a l o n g way from h i s real potential , as
indicated by the series of medi ocre m oves , beg i n n i n g with his choice of opening
variation. Knowing the main idea of Wh ite's play against 7. . . Ac5, it was not hard

79
 Chapter 2 The Way to the Top

for M ag n u s to fi nd obvious m oves . But he must be g iven h i s d u e - he carried out

a sustained attack .
Carlsen encou ntered Beliavsky's real potential a few months l ater i n Amsterdam,
where he was also hoping to m ake use of the other opportun ity to move i nto ' h i g h
society'.
I n the summer of that year the Max Euwe Association, financed by Joop van Oost
erom, organised in Amsterdam a team tournament ' Rising Stars' v. ' Experience', held
on the Scheveni ngen system . Famous veterans tested promising young players. The
best of the young would receive an invitation to the Melody Amber tournament, held
i n Monaco with the partici pation of the world el ite. The com position of the junior team
was i m posing: the young g randmasters who had distingu ished themselves in Wijk
aan Zee - Sergey Karjaki n , who had shared 5th-6th places i n the A Tou rnament, and
Magnus Carlsen, who had shared 1 st-2nd places in the B Tou rnament, the home play
ers Daniel Stel lwagen and Jan Smeets, and the Chi nese player Wang Hao. As for the
five veterans, only the 53-year-old Alexander Beliavsky was sti l l active i n com petitions.
He became a strict examiner of the main favourites in the two concl uding rounds: he
defeated the leader Carlsen and did not allow Karjakin to catch h i m . As a result Carlsen
and Beliavsky fi nished fi rst in their respective teams with identical resu lts - 6% poi nts
from 1 0 games. For the second time Carlsen fulfi l led his main objective - to break
through to encounters with the el ite in a m i n i m u m of two tournaments in 2007. As for
the examination by the veteran , it was very instructive.

M. Carlsen - A. Beliavsky 32 . . J.xe3 33. Wxe3 9:f4

Amsterdam 2006
In a position which is difficult for the op
a b c d e f g h
ponent it is logical to increase the pressure .
8 .1 .1 Therefore Bel iavsky is not in a rush to win a
7 ... 'iV pawn - 33 . . . tDxe4 34 . .!:!f1 !? (34 . .!:!e2 tDd6)
6 ... ..t ... 6 34 . . . tDf6 35 . .!:!ae1 , which allows Wh ite to
activate his forces .
5 ... 5

4
D, 4
34. J.d1
3
j, D, 3
Magnus fol l ows the path of least resist
2 D, D, D, D, 2
ance. 34. Ac2 was stro n g er: 34 . . . .!:!ef8
1 .S, .s 1 35. c5 .!:!f2 36. xeS+ .!:!8f6 37. g3 xg3
a b c d e f g h
38. hxg3 .!:!xc2 39. e5 .!:!e6 40. exd6 .!:! xd6,
31 Wg4! 32. J.e3
.
and now not 41 . .!:!e8+?! ct1g7 42 . .!:!e7+ ct1h6
43 . .!:!xb7 .!:!dd2, but i m med iately 41 . .!:!e7!
White cannot take the knight by 32. xd6 .!:!xb2 42 . .!:!f1 , and the activity of the rooks
on account of an attractive mate: 32 . . . .!:!f2 secu res equal ity. Therefore B lack wou ld
33 . .!:!g1 xg2+! 34 . .!:!xg2 .!:!f1 + 35 . .!:!g1 .!:!xg1 #. apparently have had to reg roup his forces

80
 M . Carlsen - J. Nunn

by 35 . . .'We6 36. fiad1 Wf6 37. '.'!?g1 fif2, decisive. Two pawns down , White played
but after 38. Ad3 Wh ite holds the position . on for another ten moves before curtai ling
his hopeless resistance.
34 . . . Wg5 35. 'Wd3?

This allows a rook invasion, lead ing to the

loss of a pawn without any com pensa Game 1 7
tion. 35. Af3!? tilxe4 36. fiad1 was more M . Carlsen - J . N u n n
resilient, although even here after 36 . . . tDf6 Amsterdam 2006
followed by . . . e5-e4 the extra pawn should Sicilian Defence [ 890]
have the final word .
1 . e4 c5 2. f3 d6 3. d4 f6 4. c3
a b c d e f g h cxd4 5. xd4 a6 6. Ae3 e5 7. f3 Ae7
8 8. Ac4 0-0 9. 0-0 Ae6 1 0. Ab3 c6
7 1 1 . We2 aS 12. fd1 xb3 13. cxb3!?
6 .t 6
The main continuation 1 3. axb3 We? is con
5 5 sidered to be sound enough for B lack, and
4 4 so Carlsen turns onto a less well-explored
3 3 path, opening the c-file for his rook and pre
venting the development of the queen at c7.
a b c d e f g h

a b c d e f g h
'i 8
7 .i. .t .t .t 7
35 . . . fif2 36. Af3 d2 37. Wf1
.t .i. 6

Also after the q ueen exchange 37. We3 5

.t 5

Wxe3 38. fixe3 fixb2 the endgame is hope 4 4

less for White. 3 CD j, Ci:J 3

37.. J a xb2 38. ad1 d8 39. c4?! 2 Wif 2

The cunning 39. Ag4 (with the idea of fixd6)
1 : :
a b c d e f g h
is parried by the cool-headed 39 . . . '.'!?g7.

39 . . . g7 1 3 . . . 'We8

In time-trouble one wants to play as sol
idly as possi ble, but 39 . . . fixa2 was also
possible, i ntending to meet 40. c5 with
40 . . .tDbS 41 . Wc4 fiad2.
40. 'Wd3 f7 41 . 'Wa3 dd2 42. xd2
'i'xd2 43. d1 xa2 44. 'We7 Wc3
45. h4
If 45. Wxb7 or 45. fif1 , then 45 . . . fia1 is also
Two years later the game Carlsen - lvanchuk
(Morel ia/Linares 2008) continued 13 . . .tDd7
1 4. tDd2 Wc7 1 5. fiac1 fifeS 1 6. tDdS Wd8
1 7. fixeS fixeS 1 S . tDxe7+ Wxe7 1 9. tDc4
Axc4 20. bxc4 tileS 21 . b3 fic6 and Black
maintained the balance. With the exchange
of the dark-square bishop the d6-pawn is
weakened , of course, but it proves i mpos
sible to approach it.

81
 Chapter 2 The Way to the Top
----------------

14. e1 1 8 . . . a5
This weakening of the q ueenside squares
This move indicates that Carlsen is fami l
by Black is associated with an incorrect
i a r with t h e theory even o f t h i s fairly rare
idea. Possibly the experienced g randmas
variation. 1 4 . Ag5 , attem pting to exploit
ter was not attracted by the exchanging
the weakness of the d5-point, is parried
operation 1 S . . . Aa5 1 9. tDbd5 tDxd5 (or
by 14 . . . tDh5! (this possibility is based on
1 9 . . . Axd5 20. tDxd5 AdS 21 . llad1 tDe8
tactics: 1 5 . tDxe5 is not possible because
22. llc3 tDf6) 20. tDxd5 AdS (not 20 . . . b5
of 1 5 . . . f6), played in the game Movsesian
2 1 . Ab6) 2 1 . llad 1 , lead i n g to a passive
Babula (Czech Team Championsh ip 2005-
position, which, however, is not easy to
6). After the exchange of B l ack's ' bad '
breach.
bishop the chances are equal.
1 9. bd5 xd5 20. xd5 a4?!
14 . . . hS
I l logical - why undouble the white pawns?
Later, at the rap i d tournament in Cap 20 . . . lla6 21 . llad1 llac6 was in the spirit of
d 'Agde 2006, Karjakin played more ener the position, with the idea of 22. llc3 Axd5.
getically against Carlsen: 1 4 . . .tDg4 1 5. tDc2 20 . . . b5 and 20 . . .f5 were also possible.
f5 1 6. f3 tDxe3 1 7. tDxe3 fxe4 1 S. tDxe4 21 . bxa4 gxa4 22. b3 gas 23. gad1
lldS 1 9. tDd5, and a position with equal gas?
chances was reached . A strange move, to put it mildly, leading
1 5. c2 to the loss of a pawn. If Black was i ntend
ing to exchange on d5, 23 . . . Wb5 24. a4
The knight is transferred to the key point WeS 25. tDb4 llaaS was more logical, at
of the position - d5. least provoking a weakening of the b3-
paw n . 23 . . .f5 came i nto conside ration,
1 5 ... 'Wd7 1 S. b4 gfc8 1 7. f3 .idS
with the threat of 24 . . .fxe4 25. fxe4 Ag4,
1 8, gd3
while if 24. Wf2? White has to reckon with
By threatening the doubling of rooks on 24 . . . Axd5 25. llxd5 f4.
the d -file, Wh ite does not al low time for 24. bS .txbS 25 . .txbS gas 2S. gxdS
. . . b7-b5. We7
h a b c d e f 9 h

8 8 .I 8

7
'if ... ... 7 7 ... 'ir' .t. 7

.t. i. ... 6 6 .1. I!l i. ... 6

5 ... 5 5 ... 5

4 ttJ 4 4 4

3 ttJ I!l 3 3 3

2 vw 2 2 vw 2
I!l
a b c d e f 9 h a b c d e f 9 h

82
 M. Carlsen - J. N u n n

27. Wb2! 31 . . .Wxd2

Apparently w h e n he p l ayed 23 . . J a5
Black missed this possibil ity, hoping for
27. .d8+ .xd8 28. Axd8 C5+ 29. <!>h1
'Wa3, when he regains the pawn (30. f!d2?
Axb3 31 . axb3 c1 + 32 . .d1 f!a1 ). Now,
however, the queen is switched to its own
wing , where it supports the advance of
the pawns.
27. . .Wg5?!
Of course, 27. . . f! x b6? 28. f! x b 6 c5+
29. f2 does not work, but Black should
have reconciled himself to 27. . . f!aa8 28. a4
f6, si nce the attem pt to develop counter
play on the kingside is unrealistic.
28. a4 h5 29. a5 h4 30. b4 !3:aa8
White has m ade prog ress with his attack
on the q ueenside, whereas Black's play
has come to a standsti l l , si nce after 30 . . .
h3 31 . f!d8+ .xd8 32 . .xd8+ <!>h7 33. d2
'Wxd2 (otherwise b4-b5) 34 . .xd2 f!a8
35. g4 he loses his h3-pawn.
31. Wd2
a b c d e f 9 h
a .I. .i 8 less is 37. . . f!c3+ 38. f!d3 f!xd3+ 39. <!>xd3
7
7 f6 40. <!>c3, when White creates a passed
6 I:( .t 6
Black is forced to agree to the exchange
of q ueens, since otherwise after 31 . . . h5
32. h3 the queen remains out of play, pas
sively observing the breakthrough by the
a-pawn: 32 . . . g6 33. f!c1 .xc1 + 34. xc1
.c8 35. d2 (also good is 35. xc8+ AxeS
36 . .xg6 fxg6 37. b5 <!>f7 38. Ac7 <!>e6
39. <!>f2 , when the m arch of the king to
the q ueenside leads to the winning of the
e5- and h4-pawns) 35 . . . f!c4 36. Ae3 <!>h7
37. .b6, winning the b6-pawn .
32.!3:1 xd2 !3:c4 33. Ac5 !3:c8 34. !3:b6
!3:c7 35. f2 h7 36. Ad6 !3:d7
36 . . . f!c2 37. <!>e1 f!c1 + 38. f!d1 leads to
the exchange of a pair of rooks and the
loss of the e5-pawn.
37. e3 f6
Parting with the e5-pawn , but also hope-
pawn.

5 'ii' 5
38. Axe5 fxe5 39. !3: xe6 !3:xd2
4
4
40. xd2 !3:xb4 41 . !3: xe5 !3: b2+ 42. c3
3 3
!3:xg2 43. !3: h5+ g6 44. !3:xh4 !3:a2
2 2 45. b4
1:(
a b c d e f 9 h Black resigned.

After his notable performance i n the World Cup, Carlsen was also g uaranteed invita
tions to the m ajor all-play-al l tournaments. Nevertheless, he couldn't refrai n from a trip
to Reykjavik to take part i n the trad itional Open . Magnus was leading u p to the last
round, but he contrived to lose a better endgame to the Egyptian Ad ly, whose rating
was 1 50 points lower than his. The result was a share of 6th-9th places with a score of

83
 Chapter 2 The Way to the Top

6% poi nts out of 9. A more serious test awaited h i m at a F I D E category 1 7 tournament

in Sarajevo with a well-balanced young field of six partici pants. In the absence of any
obvious outsiders, the tournament concluded with a unique result: '+1 ' was sufficient
for a share of first place. Carlsen gai ned his only win with B lack against the Bosnian
Borki Predojevic, and for clear victory in the tournament he needed to do the same in
the last round with Wh ite, but he played on to a difficult endgame, which he saved only
by a miracle. The result was a share of 1 st-3rd places with N isipeanu and Malakhov
(5% points from 1 0 games).
Part of the 'compulsory' program for 2006 was the Olympiad in Turin. Although Norway
does not number among the 'chess' cou ntries and the team faced matches in the second
echelon, Magnus showed h imself to be a genuine leader. On board 1 he scored 6 points
in 8 games, which enabled the Norwegian team to finish in an honourable 31 st place.
Carlsen's star status d i d not affect h i s patriotic sense, and he thought that he
should also play in the Norweg ian Championsh ip. This stroll in an easy tournament
(FI D E category 9) proved in fact to be q u ite d ifficult - a share of 1 st-2nd places with
his fi rst mentor, Agdestein (7 poi nts out of 9). The match for the title of champion was
won 3-1 by the pupil.
The tournament i n Biel, held with the same format as in the previous year, attracted
an even stronger field (FI D E category 1 7), with the partici pation of the winners of the
preced ing tournaments, Alexander Morozevich (2003 and 2004) and Andrey Volokitin
(2005). It was the games with them that determ ined Carlsen's result. He achieved the
i m possible - he won twice against Morozevich , but he also lost twice to Volokitin. As
for Morozevich, on this occasion too he was irrepressible and he again won ' Biel' by
a marg i n of 1 % poi nts from Carlsen and Radjabov, who shared 2nd-3rd places with
6 points out of 1 0.
Standing apart in the series of strong tournaments was the rapid-play event in the
French town of Cap D'Agde (FI D E category 1 4), where in his sem i-fi nal match Carlsen
lost %-1 % to Karjaki n .

Game 1 8 averse to playing with a fianchettoed bish

M. Carlsen - M. Adams
Olympiad , Tu rin 2006
Nimzo-lndian Defence [ E20]
1 . f3 f6 2. c4 e6 3. c3 Ab4 4. g3
o-o 5. Ag2 d5 6. d4
Since the 1 4th world champion Vlad i m i r
Kram n i k i ntroduced t h e Catalan Open ing
into his repertoire, i nterest i n positions of
this type has greatly increased . Th is theme
was taken u p by Carlsen, who is also not
op. This same position can also arise in the
Romanishin Variation of the N i mzo- l ndian
Defence. Its virtues include piece pressure
on the centre, and its d rawbacks - the
vulnerabi l ity of the c4-pawn , which is not
the case in the Catalan Open ing.
6 ... dxc4 7. 0-0
Fi rst c larify i n g the bishop's position by
7. a3!? also comes into consideration.

84
 M. Carlsen - M. Adams

a b c d e f 9 h

K. 8

7
6
6

5 5

4 ..t. t:::. 4

3 !:::. liJ liJ !:::. 3

Analysis diagram
2 !:::. !:::, !:::, ii, :::. 2
a) 7. . . Axc3+ 8. bxc3 and after the attem pt 1 M ii, WH M
to defend the c4-pawn - 8 . . . c6 (if 8 . . . a b c d e f 9 h

b5, then 9. a4! is good, taking play along

simi lar l i n es) there can fol l ow 9. Cbe5!
tbd5 1 0. V!fc2 b5 1 1 . e4 Cbf6 1 2. a4 Ab7
13. Aa3 e8 1 4 . b1 ! with the th reat of
1 5. Cbxc4. 8 . . .Cbbd7 is more cautious,
after which Wh ite carries out a plan typi
cal of such positions: 9. 0-0 c6 1 0. a4
b5 1 1 . Aa3 e8 1 2 . Cbd2 Cbd5 1 3 . V!fc2
Ab7 1 4 . Cbe4;
b) 7. . . Ae7 8. V!fa4 a6 9. V!fxc4 b5 1 0 . V!id3
Ab7 1 1 . Ag5 Cbbd7 1 2 . b4 a5!? 1 3. 0-0
axb4 1 4 . a x b 4 A x b4 1 5 . Cb x b 5 c5
16. dxc5 xa1 1 7. xa1 Axc5 with an
equal game (Radjabov -Topalov, Sofia
2008) .
7. . . tDc6
Black is prepari ng an attack on the centre
by . . . e6-e5 . The alternative is the under
mining move 7. . . c5. Therefore it is more
accurate to clarify the bishop's position
on the 7th move.
8. a3
(see next diagram)
8 . Axc3
. .
If 8 . . . Ae7 White concentrates his forces
in the centre: 9. e4 Cba5 1 0. Ae3 a6 1 1 . V!fe2
b5 1 2. ad1 , threatening to launch an of
fensive. Noteworthy, for example, is the
game M atlakov - Em e l i n (St. Petersburg
2009): 1 2 . . . Ab7 1 3 . Cbg5 Cbd7 1 4 . h4 V!fe8
1 5. f4 Cbb3 1 6. e5 Axg2 1 7. xg2 h6 1 8. d5!
Cbdc5 (Wh ite also has a dangerous attack
after 1 8 . . . hxg5 1 9. hxg5 g6 20. V!fg4 g7
21 . Cbe4 h8 22. Cbf6) 1 9. V!fg4 d8 20. f5
exf5 21 . xf5 g6 22. Cbge4 h7 23. V!ff4
1 -0.
9. bxc3 Elb8 1 0. Ag5 b5
After the inclusion of 1 0 . . . h6 1 1 . Ah4 the
continuation 1 1 . . . b5 1 2 . e4 e5 leads to a
transposition of m oves. The dangers of
w i n n i n g the bishop - 1 1 . . . g 5 1 2 . Cb x g 5
h x g 5 1 3 . A x g 5 a r e wel l d e m o n strated
by the game lvan isevi c - Mas (Biel 2008):
1 3 . . . g7 1 4 . e4 Cbe7 1 5. f4 Cbh7 1 6 . V!fh5
Cbxg5 1 7. fxg5 e5 1 8. f6 Cbg8 1 9. af1 V!/e7
20. V!ih4 V!id8 21 . V!fh5 V!fe7 22. dxe5 Ae6
23. g6 Cbh6 24. g xf7 h8 25. f8Vff+ 1 -0.

Playing this position with Black, Carlsen
tried retreating his bishop - 8 . . . Aa5 9. e3
e5 10. tbxe5 Cbxe5 1 1 . dxe5 V!fxd1 1 2 . xd1
d7 1 3. f4 Cbc5 and he equalised i n Ped
ersen - Carlsen (Aal borg 2006).
1 1 . e4
Carlsen considers a more effective way
of exploiting the pin on the knight to be
1 1 . Cbd2!? Ab7 (after 1 1 . . .Cbe7 1 2 . Axf6 gxf6
1 3 . Cbe4 the weakness of the c5-sq uare is

85
 Chapter 2 The Way to the Top

d isclosed) 1 2. tDe4 tDa5 1 3 . tDxf6+ g xf6 16 . iie8?!

1 4 . Ah6.
Played on general g rounds: B l ac k pre
11 h6 1 2 . J.h4 e5
vents the transference of the bishop to

12 . . . g5?! is risky: 1 3. tDxg5 h xg5 1 4 . Axg5 the q ueenside. Annotating the game in
e5 1 5 . d5 with a d a n g erou s attac k for lnformator, Adams suggests a more prom
Wh ite. ising way to resist: 16 . . .tDxe5! 1 7. Ae7! e8
1 8. fxe5 xe7 1 9. d8+ i!?h7 20. Ah3 Axh3
a b c d e f 9 h
21 . xb8 xe5 22. b7! xe4. Despite the
8 .l .i. 'i . . 8
exchange advantage, the weakening of
7 7 White's peri phery allows Black to i n itiate
6 ,.. ,.. 6 counterplay, for example: 23. xb5 e2
5 5 24. c5 g2+ 25. it>h1 c2 26. xc4 Ag2+
4 i, 4 27. it>g1 c6 28. a4 Ad5 29. xa7 g2+
30. it>f1 xh2 31 . e1 it>g6 or 23. it>f2 e5
3 [fj 3
24. xc7 f5+ 25. it>e3 a6, and White has
2 i, 2
nothing real.
1 1:[ 1:[ 1
a b c d e f 9 h 1 7. e6!

A bri l l i ant positional pawn sacrifice for

13. xeS! the sake of including the g2-bishop in the
game.
For Magnus, who is noted for his concrete
thinking, there is no reason to avoi d an a b c d e f 9 h

endgame where he has a spatial advantage 8 .l .i. .I 8

and the two bishops. Nothing significant is 7 ,.. 7
promised by 1 3. d5 tDa5 - after 1 4 . a4 there 6 6
can fol low 1 4 . . . g5 1 5. tDxg5 hxg5 1 6. Axg5
5 5
'Wd6, and the pin does not succeed .
4 Jt 4
1 3 xe5 1 4. dxe5 Wxd1 1 5. fxd1
3 3
d7 2 i, 2
If 1 5 . . .tDg4 Carlsen was intending 1 6. AdS 1 1:[ 1:[
a b d e f 9 h
b7 1 7. f4 tDe3 1 8. d2 tDxg2 1 9. it>xg2, c

when White gains an enduring advantage

thanks to his control of the d -fi le. 1 7 fxe6

As though bewitched , Adams calculated

1 6. f4
B lack's position is unenviable: the roo k
on d 1 severs h i s defensive resources i n
two, a n d t h e bishop pair is ready t o beg i n
operating a t ful l power.
the consequences of capturing the pawn
and decided on the m ove i n the game.
The capture with the rook 17 . . . xe6 did
not appeal to him because of 1 8 . e5 g5
1 9. fxg5 tDxe5 20. d8+ i!?g7 21 . g xh6+

86
 M. Carlsen - M . Adams

gxh6 22. ge1 ! ge6 23. ge3, and now if
23 . . . d3? there fol l ows 24. gxe6 fxe6
when Black cannot d isentangle h i m self.
Rybka, which d oes not suffer from i l l u
sions, defends b y 2 3 . . .f6!? 24. A d 5 geb6
25. gg8+ h 6 , and altho u g h the k i n g 's
position looks suspicious, it cannot be cap
tured . However, 1 8 . Ad8! is more cunning
(to judge by the game, Carlsen had this
move in mind): 1 8 . . . c6 1 9. e5 with the threat
of 20. Ah3.
But it is not essential to take the pawn!
It is surprising, but i n their commentaries
on the game neither Adams nor Carlsen
even considered 17 . . .c5! 1 8 . exf7+ xf7
1 9. e5 Ab7, when Black safely com pletes
his development.
18. e5 xeS
The only possible way of offering at least
Carlsen's play in this endgame is ad m i
rable. Depth is combi ned w i t h concrete
decision-taking. It is wel l known that the
side who is the exchange ahead should
aim for the exchange of the opponent's
rook, to deprive him of cou nterplay. The
young Norweg ian is prepared to sacrifice
a pawn to achieve this objective.
22 . . . bxa4 23. S:db1 .tc6 24. S:b8 f7
Black cou ld have prevented the i nvasion
of the rook by playing 24 . . . gxb8 25. Axb8
a6 26. Ad6 f7 27. lab1 Ab5, but not for
long. After 28. g2 f6 29. f3 h5 30. e4
White advances his kingside pawns, and
after the open ing of one of the files the
rook penetrates i nto the rear.
25. S: xe8 xeS
a b c d e f 9 h

8 8
some resi stance. 1 8 . . . g 5 is d u b i o u s :
1 9. fxg5 xe5 2 0 . gxh6 Ab7 21 . Af6 d3 7 I. 7

22. g4, and the passed pawns are bound 6

.i. I. 6

to decide the game. 5 5

19. .td8! 4 I. I. ::, 4

An excel lent intermed iate move! Black, un 3

::, ::, 3

der the cross-fire of the bishops, is forced 2

::, 2
to give up the exchange. 1 :
19 . . d3 20. J.xc7 S:b7 21 . J.xb7 J.xb7
.
a b c d e f 9 h

a b c d e f 9 h
26 . .td6
8 .!. 8
The game enters the technical phase. First
7
the path of the a4- pawn must be blocked.
6 I. 6
26 . . . a6 27. E:b1 h5?!
5 5
I llogical. If possible, the invasion of the rook
4
::, 4
should be prevented - 27. . . Ab5. However,
3
::, ::, 3
after 28. g2 h5 29. f3 f7 30. e4 White
2 ::, 2
would have obtai ned the position consid
1 : : ered in the note to Black's 24th move. Now
a b c d e f 9 h Black loses a pawn.
22. a4! 28. S:b6 d7

87
 Chapter 2 The Way to the Top
--------------

If 28 . . . Ab5 there fol l ows 29 . . b7 Ad? 38 . .aS! was stronger: 38 . . . b6 39 . .c5

30 . .a?.
29. Af8 g6 30. xa6
Black has a pawn for the exchange, and
i n add ition it is passed. But this does not
change the evaluation of the ending: Wh ite
should win, although it will be a lengthy
process.
30 . . . e5 31 . fxe5 tilxe5 32. h4 tilg4
The c3-pawn is immune: 32 . . .tDf3+ 33. f2
tbd2 34. e3 tbe4? 35 . .xc6.
tbd6 40. e3 ebbS (in the event of 40 . . . <t;c?
41 . f4 b6 42 . .g5 the g6- pawn is lost)
41 . .xc4 a3 42 . Ac5+ c7 43 . .a4 tbxc3
44 . .xa3 and Wh ite wins the pawn.
38 . . . b6 39. f7 tilh6 40. e7 Ad5
Adams defends resourcefu l ly, doing eve
rything possible to prevent the white king
from crossing the 3rd ran k. But Carlsen
fi nds a s u btle rook manoeuvre, which
destabilises the defence.

33. Ag7 c7 34. Ad4 41 . d7! tilg4+ 42. e2 Ae4 43. d4!
Ad3+ 44. f3 tilh2+ 45. f4 tilf1
The bishop has handed over control of the
46. d6+ b5 47. d5+ b6 48. a5
a4-pawn to the rook, but it is perform ing
Ac2 49. e5 Ad3 50. f3
an i m portant role by preventing the ma
noeuvre of the knight to e4. The knight 50. g4!? hxg4 51 . xg4 was stronger, not
must find another way of tying the king to allowing Black the chance which occu rred
the defence of the g3- pawn, since against in the game.
passive defence it will head towards the
50 . . . tild2+ 51 . f4 tilf1 ?
a4-pawn.
34 . . . tilh6 35. f2 tilf5 36. Ac5 d7 Adams misses the opportun ity to advance
37. Ab4 his pawn to a3 - 51 . . . tDb1 ! , with chances
of holding the new defensive l i ne.
'The Moor has done his d uty, let him go!'
The rook is freed to go i n pursuit of the
pawns.
37. . . c7

a b c d e f 9 h

8 8

7
7

G g .l_ 6 Analysis diagram

5 Here are the variations g iven by Adams:
4 ... Jt ... a) 52 . .aS a3 53. g5 Ac2 54. g4 (54. f!a8
b7 55. f!a5 b6, pursuing the rook)
3
8 3

2 2
54 . . . h xg4 55. xg4 Ab3 56. g5 Ac2,
and by red u c i n g the pawn m aterial ,
Black holds on;
a b c d e f 9 h
b) 52 . e3! a3 53 . .aS Ac2 54. d4 Ab3
38. a7+ fol lowed by 55 . . . a2.

88
 M. Carlsen - M. Adams
But Wh ite's play can be improved . In vari
ation b) instead of 53. aS he can win the
c4-pawn - 53. <!>d4! Ac2 (53 . . . <!>b7 54. as
b6 55. a4! Ac2 56. Ac5+ <!>c6 57. a6+!
b7 58. aS) 54. Ac5+! (of course, not
54. <!>xc4?? CL!d2+ and the paw n can
not be stopped) 54 . . . <!>c6 55. <!>xc4. To
convert the passed c3-pawn Wh ite must
lift the blockade of the c4-sq uare, which
Black can set u p by 55 . . .CL!d2+ 56. <!>b4
a2 57. e6+ d5 58. a6 Ab1 . This is
achieved by the manoeuvre of the bishop
to b2 - 59. a3! CL!c4 60. aa tLld2 61 . Ae3
lilc4 62. Ag5 CL!eS 63. <!>b3 CL!c4 64. Ac1
cs (64 . . . CL!d6 65. aS+) 65. a4 <!>bs
66. l::!. b 4+ <!>aS 67. Ab2, and the c- pawn
beg ins advancing.
Of course, under the strict time control it
was impossible to calculate all these sub
tleties, but this was Adams' only chance
of saving the game.
52. g4
There is no other way of breaching Black's
defences.
52 . . . hxg4 53. xg4 d2 54. f4 b3?
And h ere the o n l y sav i n g chance was
54 . .CL!b1 ! , with the hope of exchang i n g
.
the a-pawn for t h e h4- pawn. However, as
Adams shows in his analysis, after Wh ite's
best contin uation 55. <!>e3! he would have
retained winning chances.
Analysis diagram
a) 55 . . . a3 56. <!>d4 Ac2 (the attem pt to at
tack the h4-pawn after 56 . . . CL!d2 leads
to the temporary removal of the knight
to the edge of the board , which allows
Wh ite to pick up the a3-pawn and launch
an attack on the k i n g : 57. l::!. e 6+ <!>bS
58. Axa3 CL!f3+ 59. <!>d5 CL!xh4 60. Ac5
CL!f3 61 . l::!. b 6+ <!>as 62. b4 or 57. . . <!>b7
58. Axa3 CL!f3+ 59. <!>cS CL!xh4 60. l::!. e7+)
57. Ac5+! <!>c6 58. <!>xc4 tLld2+ 59. <!>b4
a2 60. l::!. e 6+ <!>d5 61 . l::!. a 6 Ab1 62. Ae3
CL!c4. Altho u g h B l ack h as set up a
blockade on c4, it is not difficult to l ift
it: 63. Ad4 tLld6 64. a5+ <!>e4 65. Af6
<!>d3 66. a3 CL!c4 67. aa CL!e3 68. c4,
and the pawn beg ins advancing;
b) 55 . . . b7 5 6 . as a3 57. <!>d4 CL!d2
58. <!>d5 (58 . xa3 leads to the loss of
the h4-pawn: 58 . . .CL!f3+ 59. <!>dS CL!xh4
60. Ac5 CL!f5 61 . l::!. a7+ <!>c8) 58 . . . CL!b1
59. l::!. b 5+ <!>c8 60. <!>d4 CL!d2 61 . <!>e3
tLlf1 + 62. <!>f2 CL!d2 63. a5 tLlb1 64. <!>e3
<!>b7 65. <!>d4 tLld2 66. <!>dS tLlb1 . In this
position, although White has retained an
advantage, Black has more chances of
holding the position, having in mind an
attack on the h4- pawn.
55. lie6+
The knight on b3 is essentially out of play
and it is easier to attack the pawn on a4
than on a2, as was exam ined in the note
to Black's 51 st move.
55 . . . b5
The a4-pawn is also not saved by 55 . . . <!>b7
56. <!>es CL!d2 57. <!>d5 CL!f3 (or 57 . . . CL!b1
58. e7+ <!>b6 59. Ac5+ <!>bS 60. b7+
<!>a6 61 . a7+ <!>bS 62. Ab4) 58. l::!. e7+ <!>b6
59. Ac5+ <!>bS 60. b7+ <!>a6 61 . l::!. a7+ <!>bS
62. Ae7 tLld2 63. Ad8 and the inevitable
aS+ (Adams).
89
 :; pt::
.; h.:;:a:;:,:
....C ...:._ T he Way to t he To p
e r'.:.2:...
....,
________

56. e8 c1 It was because of this reply, fixing the b4-

square, that the move 9. a4 went out of
The position cannot be held by 56 . . . b6
use. This makes it all the more i nteresting
57. !!a8 b5 58. e5 Ac2 59. d5 Ad3
- what has Carlsen prepared?
60. !!b8+ a6 61 . c6 Ae4+ 62. c7, when
Black is in zugzwang (Adams).
8
57. b8+ c6 58. a8 e2+ 59. f3
7
g1 + 60. e3 e2 61 . xa4
6 6
The titanic work has nevertheless con
s j. 5
cluded with the win of the pawn. Now Wh ite
can permit h i mself to switch to an attack 4 8 4
3 3
on the g6-pawn .
2
6 1 d5 62. a5+ e6 63. a 2 g3
.

64. f4 f5 65. a6+ f7 66. g5

g3 67. a7+ g8 68. f6 h5+
69. e5 g3 70. Ac5 f5 71 . Af2!
After the knight's manoeuvrabi lity has been
restricted , the game q u ickly concludes.
71 . . . h6 72. Ad4 f5 73. f6
Black resigned .
Game 1 9
M . Carlsen - A. Morozevich
Biel 2006
King's Indian Defence [ 97]
1 . d4 f6 2. c4 g6 3. c3 Ag7 4. e4 d6
5. f3 0-0 6. Ae2 e5 7. 0-0 c6 8. d5
e7 9. a4
A rare continuation in the classical variation
of the King's Indian Defence, which used to
be practised in the late 1 980s. Nowadays
it is considered that White's main plan in
volves the preparation of c4-c5, which can
be begun with an immediate 9. b4 or after the
preparatory moves 9. tLle1 , 9. tLld2 or 9. Ad2.
9 . . . a5
a b c d e f g h
1 0. b3!
Here it is! To support the move b4 White
brings out his bishop to a3 . This is a fresh
idea in the given position - the knight
remains at f3 , in order in the event of . . .
f7-f5 to proceed via g 5 to e6. It should
be mentioned that the fi rst serious test of
this variation occu rred i n the World Cup
in the game Korc h n o i - Kasparov (Bar
celona 1 989), which conti nued 1 0 . tLle1
tLld7 1 1 . Ae3 f5 1 2 . f3 tLlc5! 1 3 . tLld3 b6
1 4 . b4! tLlxd3! 1 5. 'Wxd3 axb4 1 6. tLlb5 h8
1 7. 'Wb3 tLlg8 1 8. 'Wxb4, and here, as Kasp
arov shows, Black would have gained an
advantage by 1 8 . . . fxe4! 1 9. fxe4 !!xf1 +
20. !!xf1 Ah6! 21 . Af2 tLlf6 ! .
1 0 . . . d7 1 1 . Aa3 Ah6
Of course, 1 1 . . .f5?! 1 2 . tLlg5! tLlf6 1 3. c5!
h6 1 4 . cxd6 cxd6 1 5. tLle6 Axe6 1 6. d xe6
would have been too big a concession to
White. Apart from the spike-pawn on e6,
which cramps Black, his own pawn on d6
needs defending.

90
 M . Carlsen - A. Morozevich

12. b4 axb4 1 3. Axb4 f5 event of c4-c5) and vacates the g8-square

a b c d e f g h
for the knight, which at e7 is too passively
placed. From here the knight can be played
s .l ..t 'iV .1 a
to f6, supporting the attack on the e4-pawn.
7 7
6 ..t 6 1 5. a5 Iaf7
.l .l
5 B .l .l 5 For the m o ment there i s noth i n g new.
Wh ite continues seizing space, while Black
4
BB B 4
strengthens his queenside defences and
3 CD CD 3
is not in a hu rry to remove the support of
2 B B B 2 the f5-pawn.
1 :. VJ!I :. 1 6. ttlb5
a b c d e f g h
Carlsen is true to his agg ressive style. He
Here we can take stock of White's opening presses on the q ueenside, d isplaying his
idea. Compared with the 9. b4 variation, he read i ness to sacrifice the e4- pawn and
is as though a tempo down , since his pawn hoping to use his pieces on the adjacent
has gone to b4 in two moves, which, how a2-g8 and a1 -h8 d iagonals.
ever, is not of great importance. The main
1 6 . . . ttlf6
ideas of this position are wel l known: White
prepares c4-c5, and Black, while trying to Black accepts the challenge. The posi
prevent this, ties down the white pieces by tion demands concrete decisions, and
the reinforcement of the f5-e4 pawn ten Morozevich thinks that the manoeuvre of
sion, not hurrying with the exchange. As for his other knight to f6 via g8 is too slow,
the plan of advancing the kingside pawns in si nce he has to reckon with 1 7. exf5 g xf5
connection with . . .f5-f4, in the g iven situa 1 8 . c2 when he again faces a d i lemma
tion (with the bishop on h6) it is unfounded, - how to operate with his knights.
since White gets in first with his attack on 1 7. c5 dxc5
the queenside. In this respect it is useful to After 17 . . . Axd2 1 8. xd2 tt::l x e4 Wh ite had
be fami liar with van Wely's games with lvan prepared 1 9. a2! dxc5 20. d6, attacking
chuk (Wijk aan Zee 1 999) and Piket (Monaco rook and knight.
1997).
a b c d e f g h
14. ttld2

A typical m anoeuvre. S ince the knight's 7

route to g5 is blocked , it defends the e4- 6
pawn and is ready from b3 to support the 5
c4-c5 advance. 4 4

3 3

Also a typical manoeuvre in this type of 2 CD B B B 2

position . The king moves off the a2-g8 di 1 1:, 'i :.
a b c d e f g h
agonal in good time (which is useful in the

91
 Chapter 2
1 8 .tc3
. An anti-positional move - the doubled c
Outwardly a thematic move: by capturing
the eS-pawn, the bishop puts the black
king u nder an 'X- ray' attack . But i n the
given situation this is going too far. Carlsen
comm its a mistake, typical of youth: he
fai ls to reckon with the opponent's cou n
terplay. There was no point i n rejecti ng
1 8. AxeS, since after 18 . . . Axd2 1 9. 'Mixd2
llxe4 20. 'Mib2! Black has no way of op
posing Wh ite's attack on the long diago
nal: 20 . . .llxcS 2 1 . 'MixeS+ g8 22. llxc7.
Now in the event of 22 . . . lld7 23. Wc3 !!a7
24. !!fd1 llf6 2S. d6 llc6 26. Ac4 Wh ite
real ises the attacking conception beg un
with 1 6 . !lbS - his pieces dominate on the
a2-g8 and a1 -h8 d i agonals. There only
remains 22 . . . !!b8, but after 23. lle6! (this
is stronger than 23. d6 lle4) Wh ite forcibly
transposes i nto a favou rable endgame:
23 . . .WxdS 24. WxdS !lxdS 2S. eLl xeS !!c7
26. !!fc1 llf4 27. Af1 .
1 8 . . . c6!
This way, without a moment's respite! In
the event of 1 8 . . . Axd2 1 9 . Wxd2 tD xe4
20. AxeS+ g8 21 . Wb2 Wh ite wou ld have
achieved his aim.
1 9. dxc6?!
Carlsen beats a retreat, relying only on
t h e positional p l u ses assoc i ated with
the k n i g ht's post o n t h e c4 b l ockad i n g
square . And w ro n g l y s o ! By the p i ece
sacrifice 1 9. AxeS cxbS 20. AxbS !!xaS
21 . !!xaS WxaS 22. tDc4 Wh ite could have
conti nued the attack: 22 . . . Wd8 23. lld6
!!f8 24. Axf6+ !!xf6 2S. Wa1 Wf8 26. eS,
regai ning the sacrificed material and retain
ing the i n itiative (Sergey Shi pov).
19 . . . bxc6?!
The Way to the Top
pawns are no adornment to Black's po
sition . The natural 1 9 . . .tDxc6! suggested
itself: 20. llc4 !!d7! 21 . Wc2 tDxe4 22. tLlb6
!!b8 23. llxd7 Axd7, and after . . .tDd4 Black
has more than sufficient com pensation for
the sacrificed exchange.
20. a3
a b c d e f g h
8 !. .i. 'ii' 8
7
.. 7
6 .s.. 6
5 8 5
4
8 4
3 3
2 8 8 8 2
1 :. :.
a b c d e f g h
20 . . . fxe4
The attempt to win this pawn with the knight
after 20 . . . Axd2 would have been parried
by the i ntermediate move 2 1 . tDc4! (after
21 . Axd2?! tDxe4 22. Ac3 Wc7 23. Ab2
tLldS 24. llc4 !!e7 2S. f4 Aa6 26. eLlxeS g8
the black knights dominate in the centre
of the board) 21 . . . tDxe4 22. AxeS+ g8
23. f3, and Wh ite regains the piece, retain
ing all the pl uses of his position. As Vik
tor M i khalevsky has shown , Black woul d
have had t o g o in for a forcing variation:
23 . . . WdS! 24. fxe4 Ae3+ 2S. tDxe3 WxeS
26. tDg4! fxg4 27. Wd8+ g7 28. !! xf7+
xf7 29. !!f1 + tDfS 30. Ac4+ g7 31 . Wg8+
h6 32. exfS gxfS 33. h1 !?, where the
doubled extra pawns are of no im portance,
but White's initiative remains. After the in
clusion of 33. Wf8+ gS 34. h1 Black
can sacrifice the exchange - 34 . . . Aa6!
92
 M. Carlsen - A. Morozevich
3S. 'Wxa8 Axc4 36. !::!g 1 AdS with the threat
of forcing a d raw by perpetual check after
. . . Axg2+.
21. ac4! ed5 22 .&.xeS .&.g7
23. d6
An illog ical , reckless move, depriving the
position of its strateg ic clarity for which
Carlsen was aiming. He should have con
sidered 2 3 . 'Wa4 e3 (23 . . . 'Wd7 24. 'Wa3
'fie? 2S. Ad6 favou rs Wh ite) 24. fxe3 tDxe3
2S. !::!f2, and after the inevitable exchange
of a pair of knights Wh ite controls the most
important points on the board .
23 . . J ae7 24.2c4 .te6?
This routine central play allows Wh ite to
advance h i s passed paw n . 24 . . . A a6 ! ,
keeping the white pieces under pressure,
was more logical .
25. a6! b4 26. 'Wc1
26. 'Wb1 !? (with the threat of tDb7) came into
consideration: 26 . . . 'Wg8 (both 26 . . .tDxa6
27. tDb6 and 26 . . . !::! x a6 27. !::! xa6 tD x a6
28. 'Wa1 are u nfavou ra b l e for B l ac k)
27. tDb6 !::! aa7 and now 28. 'Wc1 gains in
strength . For exam ple, if 28 . . .tDd3 there is
29. Axd3 exd3 30. 'WxcS.
26 . . . d3?!
Understandably, Morozevich did not like
26 . . . !::! x a6 27. !::! xa6 tDxa6 28. Wa1 ! !::! a?
29. f3! e3 30. tDxe3, when Black is u n
able t o d isentangle h i mself without los
ing material. But he could have taken the
pawn with the knight: 26 . . .tDxa6 27. !::! d 1
(after 27. Wf4 tDd5 the knig hts become
coordinated) 27. . . AdS 28. 'Wf4 'Wf8 , and
the knight on a6 moves out of 'custody'.
27 .&.xd3 exd3
a b c d e f 9 h
a .I 'iV 8
7 .1. ..t. 7
ti:J ..t. '.&
dt r-+
6 6
5 5
4
t2J 4
3 3
2 2
1 S s 1
a b c d e f 9 h
28. 'Wc3?!
A poor move, allowing Black to provoke
favou rable s i m p l ification while reta i n i n g
his extra material . White could have fought
for an advantage with the natural 28. !::! d 1 .
28 . . . Axc4! 29. 'Wxc4?!
After this move the d2-pawn remains alive
and White has to fight for equality. More ac
curate was 29. tDxc4 tDe4 30. 'Wb2 g8!?
31 . Axg7 !::! xg7 32. tDb6 !::! aa7 33. !::!fd1 with
the idea after 33 . . . Wd4 34. Wxd4 cxd4
35. !::! xd3 c5 of forcing a d raw: 36. tDc8
!::! a 8 37. tDb6.
29 . . .'Wg8! 30. 'Wxc5 d2
An impu lsive decision, provoked by a de
sire in time-trouble to fix Wh ite's attention
on the far-advanced passed d-pawn . From
the standpoint of fighting for an advantage,
30 . . . 'WdS! 31 . Wc3 (31 . Wxd5 tDxd5 32. f4
tDb4) 31 . . .tDh5 32. Axg7+ !::! x g7 33. 'Wb4 c5
34. We4 Wxe4 35. tDxe4 c4 was stronger.
31 . Jaad1 Jaxa6?!
One can understand Black's desire finally to
capture the annoying passed pawn. Espe
cially since in severe time-trouble it was hard
to evaluate the consequences of the ex-
93
 Chapter 2 The Way to the Top

changing variation 31 . . . \Wd5! 32. \Wc3 1:!xe5! to create a counter-th reat - 36 . . . f!ga7! .
33. \Wxe5 \Wxe5 34. tDf7+ g8 35. tDxe5 Now the piece sacrifice 37. f!db2?! \Wxd6
tDe4, where for the moment the d2-pawn 38. 1:!b8+ g7 39.1:!1 b7+ proves ineffective
remains alive, but as Shipov showed, Black in view of 39 . . . tDc7 (39 . . .tDe7?? 40. f!g8+!
is in no danger: 36. tDc4 Ac3 37. f3 Ad4+ leads to mate in three moves), and after
38. h1 tDf2+ 39. !:!xf2 Axf2 40. !:!xd2 1:!xa6, 40. !:!xa7 f!xa7 41 . f!g8+ f6 Wh ite can
and the draw is not far off. count o n l y on a d raw. M ost p robably
White would have had to restrict h imself
32. xd2
to 37. '1Wd4+ h7 with an unclear position.
a b c d e f g h
37. xb6 'We1 + 38. h2
8
'i . 8

7 .!. j_ j. 7 a b c d e f g h

6 .!, j. tt:J J!& j. 6 8

8

5
WI j, 5 7
.!. 7

4 4 6 tt:J j. j. 6

3 3 5
I.& 5

2
2 4
4

3 3

d e f g h 2 rJ;; 2

With the win of the d2-pawn Wh ite has 'iii

a b c d e f g h
also gained a solid positional advantage.

32 . . d5 33 .txg7+ xg7?! 38 xb6?

The e-file should not have been conceded .
33 . . . xg7 34. \Wd4+ tDf6 was stronger,
since now 34. !:!e2! would have put Black
in a critical position: 34 . . . !:!e7 (Black would
not hold out much longer after 34 . . .tDc7
35. !:!fe1 f!a8 36. \Wxc6) 35. f!xe7 tDxe7
36. tDe8! \Wxe8 37. \Wd4+ g8 38. \Wc4+.
But Wh ite missed this chance, and every
thing turned fu l l circle.
34. h3?! 'We6!
It is i mportant not to concede the e-fi le to
the white rooks.
35. b1 h6 36. 'Wc4 b6?
Black overlooks a q ueen manoeuvre, win
ning a pawn, just when salvation was at
hand: 38 . . . \Wxd2 39. !:!xc6 \!f/f4+ 40. \!f/xf4
tDxf4 and, in view of all the pawns being
on one wing, he can hope for a draw.
39. 'Wf4! d5?
Allowing a mating attack. But also after
39 . . . \We7 40. Wxh6+ g8 41 . Wf4 tDd5
42. Wg3 there was no chance of saving
the game.
40 xdS! cxdS 41 'WfS+ h? 42. ea
Black resigned .

By impu lsively aiming to simplify the posi- This defeat merely aroused Morozevich,
tion, Morozevich misses an opportun ity who won fou r games i n a row. But then

94
 A. Morozevich - M . Carlsen
again 'awaiting' h i m was Carlsen . On this
occasion Magnus had to demonstrate his
defensive ski l l in a d ifficult position.
Game 20
A. Morozevich - M . Carlsen
Biel 2006
Sicilian Defence [ 830]
1. e4 cS 2. f3 c6 3 . .tbS f6
Carlsen is seeking fresh ideas in an old
variation . 3 . . . g6 occurs far more often .
4 .txc6 dxc6 S. d3 d 7 6 . .tf4!?
.
With the knight retreat Black prepared the
fixing of the centre with . . . e7-e5, and White,
while continuing his development, prevents
this. The play is more committing after 6. e5
eab6. The source game Bologan - Lautier
(Poikovsky 2003) contin ued 7. tLlbd2 Af5
8. 0-0 e6 9. b3 Ae7 1 0. Ab2 0-0 1 1 . l:!e1 a5
1 2 . a4 tLld5 1 3 . tZ:Ie4 VUe? with com plicated
play. But in this variation too Carlsen had
some experience: 7. 0-0 c4! 8. d4 Ag4
9. c3 e6 1 0. Ae3 Ae7 1 1 . tLlbd2 0-0 1 2. vt/e2
'Wd5 1 3 . b3 vt/a5 1 4 . bxc4 tZ:Ia4 1 5. tZ:Ib3
(1 5. tZ:Ie4!?) 15 . . .vt/a6 1 6. h3 Ah5 1 7. Ad2
eab6 1 8. g4 Ag6 1 9 . Ag5 Axg5 20. tZ:Ixg5
'Wxc4 21 . vt/xc4 tL:Ixc4, and Black achieved
the better endgame (G u l laksen - Carlsen ,
Tromso 2006).
6. . . g6 7. 'Wc1 .tg7?!
In his comments on the game Carlsen con
siders that it was more logical to prevent
the exchange of his bishop by 7. . . h6, aim
ing to obtain a position simi lar to the main
variation 3 . . . g6 4. Axc6 dxc6.
8 .th6 WaS+!
.
A move with deep impl ications. M agnus
95
has found a way to avoid the exchange of
bishops. If 8 . . . 0-0 there cou ld have fol
lowed 9. h4.
9. c3
If 9. Ad2, then 9 . . . vt/b6 is possible.
9 eS! 1 0. xeS .txeS 1 1 . d2
. .
a b c d e f 9 h
J. s
5
4 4
3 3
::t: 1
a b c d e f 9 h
1 1 . . .'Wa6?!
A poor m ove, and the i n itial cause of
Black's problems. The d3-pawn is eas
ily defended , whereas for a long time the
queen proves to be out of play. The more
naturai 1 1 . . . Ae6 1 2 . tZ:Ib3 suggests itself,
and now 1 2 . . . vt/b5, provoking a weaken
ing of the d4-square (1 2 . . . vt/b6 1 3 . Ae3
Axb3 14. axb3) 1 3 . c4 vt/b6 14. Ae3 Ad4
with roughly equal chances (variation by
Carlsen).
12. 'Wc2
Those fol lowi ng this game on the internet
wondered what Morozevich was thinking
about for so long, since 12. tZ:Ic4 is such
a natu ral move. However, after 12 . . . Ac7
1 3 . vt/e3 (manoeuvres such as 1 3 . Ag7
l:!g8 1 4 . vt/h6 Ae6 1 5 . vt/xh7 0-0-0 can
hardly be seriously considered) 13 . . . Ae6
1 4 . vt/xc5 0-0-0 Wh ite is forced to switch
 Chapter 2 The Way to the Top

to defence - 1 S. Ag7 ghe8 1 6 . Ad4 <!>b8,
when his pieces are too cram ped . 1 7. 0-0
can be answered by 17 . . .fS!?, setting Wh ite
a choice - 1 8. AeS gxd3 1 9. A xe?+ <!>xc7
20. 'WeS+ <!>d? or 1 8 . exfS AxfS , in both
cases with an acceptable game for Black.
1 2 . . . .te6 1 3. f4
White drives back the bishop to c7, in order
to obtain the possibility of playing c3-c4,
rel ieving his concerns over the d3- pawn
and sti l l further restricti ng the q ueen on
a6. I n the event of 13. f3 Ac7 1 4. gS
Black need not fear the doubling of his a
pawns - 14 . . . 0-0-0 (weaker is 14 . . . Ad?
1 S . 'W b3) 1 S . xe6 fxe6, s i n ce it i s n ot
easy to defend the backward d3- pawn:
after 1 6. c4 AaS+ 1 7. <!>e2 B lack doubles
rooks on the d -file.
1 3 . . . .tc7 1 4. 0-0 f6
Preparing the retreat of the bishop to f7 in
the event of f4-fS .
1 cM 0-0-0 1 6. a4!
The vice tig htens. Black's position is des
perate, si nce Wh ite's only weakness - the
d3-pawn - is easily defended , whereas it
is easy for h i m to plan his attack on the
Carlsen recognises the danger of his posi
tion, and, realising that 1 6 . . . AaS 1 7. b3
Ab4 1 8 . aS is hopeless (the q ueen is still
out of play), he decides on rad ical meas
ures. He decides to carry out a combi ned
attack on the aS- and d3-pawns, display
ing his read iness to part with his trapped
queen . Half-measures aimed at cutting off
the bishop on h6 apparently did not ap
peal to him - 1 6 . . . ghg8 1 7. aS gS 1 8. b3
b6 1 9. axb6 'Wxb6 20. fxgS fxgS 21 . 'Wc3,
when after Ag7 W h ite controls the po
sition's i m portant hig hway and retains a
stable advantage.
1 7. a5 E:hd8 1 8. E:a3
The attack on the d3-pawn has come to
a standstill, whereas 1 9. fS Af7 20. Ae3
is th reatened , forcing B lack to weaken his
king's defences - . . . b7-b6. But Carlsen
had al ready taken a decision.
18 . . . .txa5!
It is better to die standing, than to l ive on
your knees!
a b c d e f g h
8
. .. 8

q ueenside: a4-aS , d2- b 3 , f4-fS and 7 j. j. !. .I. ... 7

6 'jV
Ah6-e3 . ... .i. .t. .t. .!. 6
a b c d e f g h 5 .i_ ... 5
8
<9 .1 .I 8 4 B BB 4

7 .t. .t. .i. .t.J- ... 7 3 1:,

B 3

6 'ifV ... .i. .t. .t. .!. 6 2 [j, VJJ{ ttJ BB 2

5 ... 5 1
: 1

4 B B BB 4 a b c d e f g h

3
B 3
1 9. f5!
2
B tiJ B B 2
White is not in a h urry. The move b3 will
1 1:, : 1
not run away, and for the moment it is use
a b c d e f g h
ful to include the bishop in the attack on
16 .. E:d7 the cS-pawn .

96
 A. M orozevich - M. Carlsen
19 . . . Af7 20. fxg6
If 20. ti)b3 there wou l d h ave fol l owed
20. . . xd3 21 . ti)xc5 (in the event of 21 . xa5
V!Vxc4 22. 'Wxc4 Axc4 23. ti)xc5 d1 24. xd1
l::l xd1 + 25. f2 b6 the piece is regained)
21 . . . Ab6!. It was on this queen sacrifice that
Carlen's defensive idea was based. After
22. xa6 Axc5+ 23. h1 bxa6 a position
arises, similar to that which occurred in the
game.
20 . . . hxg6 21 . eS Ae6 22. exf6
It was poss i b l e to w i n the exchange -
22. ti)e4 Af5 23. e6 d4 24. Ae3 , but Mo
rozevich is hoping for more.
22 . . . exf6 23. b3 xd3 24. xeS .tb6
A piece is lost after both 24 . . .'Wb6 25. xa5
V!Vxa5 26. ti)xe6, and 24 . . . 'Wxc4 25. 'Wxc4
Axc4 26. ti) x d 3 Ab6+ 2 7. ti)f2 A x f 1
28 . xf1 .
2S. xa6
For Wh ite to restrict h imself to the win of
the exchange by 25. xd3 Axc5+ 26. h1
27. 'Wxd3 Axc4 28. 'Wh3+ f5 29. e1 would
be too great a concession to Black.
2S . . . AxeS+ 26. h1 bxa6
a b c d e f g h
8
. ... 8 which gives Black a serious attack.
7 I."r--->--- 7
6 I.
5 5
4 4
3 3
2 ::, ::, 2
a 1
a b c d e f g h
This is the position for which Carlsen was
aiming i n his attem pt to save the game.
Black has sufficient material compensation
for the queen, and his control of the d-fi le
allows h i m to hope for counterplay in view
of Wh ite's i n adeq u ately d efended back
ran k. Surprising ly, he ignores this factor.
27. Ag7?
Fate is obviously u n kind to Morozevich.
Now he loses the game. Of course, it is
unjust to end up with a zero after such high
q uality play. But objectively, it has to be
ad m itted : the situation has changed , and
the coordi nation of the black pieces can
be disrupted only by concrete play: 27. b4!
Ab6 (27. . . Axb4 28. 'Wa4) 28. 'We2 ! Af7
(28 . . . Af5 29. g4!). Now Carlsen suggests
the following variation 29. 'We4 b7 30. b5
axb5 31 . cxb5 Ad5 32. 'We7+ (or 32. bxc6+
Axc6 33. 'Wxg6 d1 ) 32 . . . a8 33. bxc6
Axc6 34. 'Wxf6 3d6 with counterplay for
Black. But White can also consider 29. Ag7
d2 (29 . . .f5?! 30. c5 Ac7 is not good be
cause of 31 . Ae5!) 30. 'Wf3 8d6 31 . Af8
e6 32. 'Wc3 de2 33. Ac5 b7 34. 'Wd3
Ac7 35. b5, when he nevertheless suc
ceeds i n d isru pting the coord i nation of
the black pieces.
27.. Ag4!
N ow W h ite cannot h o l d h i s back ran k,
28. b4
Here this move is now too late.
28 . . . Ae3! 29. h3 d1 30. xd1 xd1 +
31 . h2 Af4+ 32 . g3 d2+ 33. Wxd2
J.xd2 34. hxg4
(see next diagram)
97
 Chapter 2 The Way to the Top

a b c d e f g h There was a q uicker win by the thematic

8 8 diverti ng-cum-i nterference move 34 . . . c5!
7 ... 7
35. bxc5 Ag5 36. Af8 aS 37. c6 Ac1 , when
the a-pawn cannot be stopped .
6 ... 6

5 5 35. Axf6 aS 36. g2 d7 37. f3 Ad6

4
::, ::, ::, 4
These tricks are unnecessary. There was
3 ::, 3 an elementary w i n by 37 . . . a4 3 8 . e2
2
j_ 2 a3 39. d3 e6 40. Aa1 Ae7 41 . c2
Af6.
a b c d e f g h
38. e4 Axg3 39. d3 e6 40. Ad4
a6 41 . c2 a4 42. b1 Ae5 43. Af2
34 Axb4?! d6 44. a2 Ac3 45. a3 e5
46. xa4 f4 47. Ab6 xg4 48. Aa5
For the moment Carlsen's endgame tech
Axa5 49. xa5 f4 50. b6 aS
nique is still shaky. The exchange of pawns
allows Wh ite to p rolong the resistance. Wh ite resig ned .

The sternest test of Carlsen's potential came at the M i khail Tal Memorial Tou rna
ment, which took place i n N ovem ber 2006 i n Moscow. This was the fi rst time he had
played i n a F I D E category 20 tournament with such a un iformly strong field . I ndeed ,
half of them were in the top ten in the world rating l ist , and although Carlsen h i mself
was just 2 poi nts short of the 2700 mark, already i n the 1 st round the difference was
felt. M agnus was g iven a gen u i ne strategy lesson by Gelfand : 'J ust l i ke Chigorin',
Yu ry Razuvaev commented on the manoeuvres of his knig hts, which outplayed the
opponent's bishops. After being g iven a master class by one of the pupils of the
Soviet Chess Schoo l , su bseq uently the youngster held on l i ke a seasoned com peti
tor. For the moment it was hard to hope for anyth ing more, and had it not been for
an em barrassment in h i s game with Aronian (on this occasion he lost a dead -d rawn
rook endgame), M agnus wou l d have saved face right to the end of the tou rnament.
But as it was - a share of 8th-9th places with 3% poi nts out of 9. A serious lesson
on the eve of his 1 6th birthday.
Contrary to expectations, Carlsen made a poor start to the year 2007. In his fi rst
A Tou rnament i n Wij k aan Zee (FI D E category 1 9) he suffered a complete fai l u re.
Magnus lost fou r games, without a single win, and shared last place with S h i rov.
The game between them from the last round was the only one where Carlsen was
gen u i nely close to a w i n . It was something else that was i m portant. M agnus man
aged to sum mon u p his courage and demonstrate the potential of a fig hter, ready
for forthcoming battles.

98
 M. Carlsen - A. S h i rov

M. Carlsen - A. Shi rov lead to the goal. H ere too 29. f!.gf3 was
Wij k aan Zee 2007 more logical.

a b c d e f 9 h 29 . . . xe5 30 . .tb2 bc4 31 . xg6+

8
.i .i 8
Wh ite has no other way of conti n u i ng the
7
.i. .i. .t 7 attack. 31 . Ad4 allows Black a respite to
6 .\ I.& .\ 'iV .t 6 bring up his reserves and parry the attack:
5 .\ 5 31 . . . f!.f6 32. d5 ..xf1 + 33. Wxf1 ..f8, re
tai ning the extra pawn.
4
8 1.&\ tD 8 tD 4

3 8M 8 3 31 . . hxg6 32 . laxf8+ la xf8 33. Axg6

2
1t 8 2
After 33 . .. x g 6 x b2 34. Wh5+ g8
1t M 1 35. f!. xg7+ xg7 B lack's m aterial com
a b c d e f 9 h
pensation for the q ueen is too great.
a b c d e f 9 h
26. f5!

Carlsen apparently thought that winning

the q ueen by 26. fe6 ..xe6 27. ..xf6 ..xf6
was an insufficient reward for the parryi ng
of the attack.

It is dangerous to take the knight - 26 . . .

gxf5 27. ..g3! h8 28. h5 '1Wd4 29. xg7!
(this is more effective than the capt u re
with the rook) 29 . . . Wxd1 30 . ..xd1 ..xe4
33 . . . .tg4?
(30. . . f!.g8 31 . xf5) 31 . Axe4 fxe4 32. h5,
winning the exchange with a conti nuing
attack: 33 . ..xd6 is th reatened (variation
by Carlsen).
27. xg7 Wxg7 28. lag3
28. f!.bf3!? was stronger.
28 . . . laf8 29. e5!?
A pawn sacrifice on the altar of the at
tack! B ri l liancy lovers will undoubtedly give
Carlsen's enterprise its due. And yet, ob
jectively speaking, this is not the most justi
fied sol ution to the position, since against
accu rate defence Wh ite's plan does not
a b c d e f 9 h
Carlsen comments that both he and Shi
rov considered t h i s to be the o n ly pos
sible defence, si nce a conti nuation of the
forcing play by 33 . . . f!.f1 + 34. Wxf1 xb2
35. Ae4! Wh6 36. AdS would have led to
a loss for B l ack, in view of the th reat of
f!.g8. But this proved to be delusion. After
33 . . . Ae8! Wh ite would have had nothing
better than to force a d raw by perpetu
al check: 34. Wh5+ g8 35. Ah7+ h8
36. Ag6+.
34. laxg4 Wt6 35. Ad4 Wf1 + 36. Wxf1
laxf1 + 37. h2 xa3 38. d3 lad1
39. lag3 ac4 40. h4

I
99
 Chapter 2 The Way to the Top

There is no point i n com menti ng on the
practically forced play i n the time scram
ble, except to suggest moving one of the
bishops out of the pin - 40. Ac3!?.
40 d2
...
The smoke of battle has cleared . With
material eq ual, Wh ite has the two bishops,
an i m portant factor, given the open char
acter of the position. But subseq uently he
missed chances to convert his advantage,

and the game ended in a d raw 20 moves

Only a computer would be capable of find later.
ing a defence such as 40 . . .&Db6!.

It is unl ikely that anyone was expecting Carlsen to succeed i n the most prestigious
tournament - Morelia/Linares (FI DE category 20), so poor had been his fi rst contacts
with the el ite g randmasters i n M oscow and Wijk aan Zee. It was all the more surprising
to see him lead ing together with Anand after the Mexican half of the tournament. This
did not su rprise Magnus himself, as he was the most 'motivated ' fighter in the i nitial
part of the tournament. 3% points in the fi rst five games - perhaps only Kasparov had
managed this! After an exchange of blows with Morozevich and Anand (Carlsen's 'evil
gen i us'), he gai ned successive wins over lvanchuk and Topalov, who clearly underes
timated the youngster. In the Span ish half of the tournament they played against him
with more respect, and only i n the last round was he unable to anything to counter the
solid strategy of Leko, who was eager to win at least one game. Carlsen was caught
by Morozevich, who played bri l l iantly in the second half of the event. Nevertheless, by
scoring 'plus one', he clung on to 2nd place, a point beh ind the win ner, Anand . The
main outcome of the tournament was that Magnus had learned how to win against
el ite g randmasters!
He confi rmed this at the ' Melody Am ber' Tournament, staged by Joop van Oosterom
in Monte Carlo. This was a unique contest with twelve of the strongest players in the
world , personally selected by the patron, who played two games a day - one bli ndfold
and one rapid. Carlsen performed worth ily in the rapid event, and with a score of 6%
poi nts he shared 2nd-5th places, beh ind only the world champion Vlad i m i r Kram nik.
He was less successfu l at bli ndfold play, where he scored 4 points.
One would have thought t h at , having passed thro u g h the cru cible of several
p rest i g i o u s to u rn aments , Carlsen had earned h i m self a rest . B u t ahead l ay the
Can d i d ates tournament, crown i n g his fi rst cycle i n the battle fo r the world cham
pionsh i p . Befo re trave l l i n g to E l i sta it was i m portant to raise h i s fig hting spi rit in
h i s homelan d , and M ag n u s deci ded to play i n Gausd a l , w h i c h was c l ose to his
heart. In o rder to ensure the m otivation of thei r i d o l , the o rganisers of the ' G ausdal
C l assic' g randmaster tournament did everyt h i n g possible to raise its level to F I DE
categ ory 1 2. T he tourn ament was i n the nature of a warm - u p : Magnus easily took
the fi rst prize. Over the short d i stance of 9 rou n d s he scored 7 points, fi n i s hing 1 %
ahead of h i s nearest rival.

100
 M. Carlsen - A. Morozevich

Game 2.....
1 _ thf3-e5) there can fol low 9 . . .thh5 1 0 . Ae3
M. Carlsen -A. Morozevich thd4!.
Morelia/Li nares 2007
8 . . . tba5 9. tbd2 e5
King's Indian Defence [ E66}
b d f g h
1. d4 tDf6 2. c4 g6 3. g3 .tg7 4 . .tg2
a c e

0-0 5. tbc3 d6 6. tDf3 c5 a!. .i. I. e 8

7 j.j. j..i_j. 7
This m ove characterises the Yugosl av
Variation, which was so named thanks to
6 j. 6

the analyses of Yugoslav players i n the

si.&J 1.&1 ., 5

1 950s. By attacking the centre from the 4 4

wing, Black tries to expand the range of 3

his bishop on g7, not feari ng the exchange
7. dxc5 d xc5, si nce in the absence of any
weaknesses Wh ite can hardly hope for an
advantage in this sym metric position. I n
the event of 7. d 5 with 7. . . b5!? Black can
switch to a sharp variation of the Benko
Gam bit. Therefore after 6 . . . c5 Wh ite usu
ally does not hu rry to close the centre, but
waits for Black to bring out his kn ight with
7. . .thc6, after which he plays 8. d5. In mod
ern tou rnaments the development 6 . . . thc6
7. 0-0 a6 is preferred , when after 8. d 5
tDa5 9 . thd2 c 5 a similar pawn structure
arises. The attack on the wh ite centre with
the central pawn 6 . . .thbd7 7. 0-0 e5 takes
us back to the outstand i ng 'fathers' of the
King's Ind ian Defence, David Bronstein and
Isaak Boleslavsky.
7. 0-0 tbc6 8. d5
Wh ite's m a i n conti n u ation, lead i n g to a
closed position. B ut in the Yugoslav Vari
ation B lack also has to reckon with the
transition into the symmetric variation after
the exchange 8. d xc5 d xc5, where Wh ite
can try to exploit the advantage of the fi rst
move. But here too B l ack h as cou nter
arguments. For exa m ple, after 9 . Ae3
Ae6 10. a4 he can avoid the symmetry
-1 0 . . .thd4!, while if 9. Af4 (with the idea of
I
101
tiJ 3
2 tiJ i, 2
1
1 i, 1
a b c d e f g h
By closing the centre, Black sign ifies that
his priorities lie in the centre and on the
ki ngside. The d rawback to his position is
the knight on a5, which is not easy to bring
i nto play.
The other plan with a counterattack i n
t h e centre a n d on t h e q ueenside is asso
ciated with the underm i n i ng of the pawn
chain by . . . e7-e6 and . . . b7-b5 and play on
the b-file. It is usually carried out via this
move order: 9 . . . a6 1 0 . c2 b8 1 1 . b3 b5.
1 0. b3
More usual is 1 0 . a3 b6 (1 0 . . . c7? fails to
1 1 . b4! cxb4 1 2. axb4 thxc4 1 3 . thb5 b6
14. thxc4 xb5 1 5. thxd6 xb4 1 6. Aa3
with an obvious advantage for White) 1 1 . b4
thb7 1 2. Ab2 thg4 (if 1 2 . . .thea Wh ite erects
the same barricade) 1 3 . h3 thh6 1 4. e3! f5
1 5. f4 with complicated play.
As for the obvious move 1 0. e4, it has
contrad ictory assessments. T h e out
stand i ng King's Ind ian expert Yefim Gel
ler regu larly employed it, hoping to exploit
the undermi n i ng move . . .f7-f5 to White's
advantage, si nce on the decisive part of
 Chapter 2
the battlefield he has one piece more. For
example, after 1 0 . . .CL!g4 i n his game with
Vel i m i rovic (Havana 1 97 1 ) he responded
1 1 . b3 f5 1 2. exf5 and after 12 . . . e4? 1 3. f6!
CL!xf6 1 4.CL!d xe4! CL! xe4 1 5 . CL! xe4 Axa1
1 6. Ag5 Af6 1 7. CL!xf6+ xf6 1 8. 'Wa1 Black
was unable to escape from the vice. But
also after 1 2 . . . gxf5 1 3. h3 CL!h6 1 4 . Ab2
with the preparation of f2-f4 it is not easy
for Black to create counterplay on the king
side. Nevertheless, the modern view on the
problems of this variation recommends
that Wh ite should avoid al lowing Black his
main counterplay resource.
With 1 0 . b3 Wh ite fol lows a d ifferent
conception: he pays no attention to the
out-of-play kn ight at a5 and tries to exploit
its absence for the creation of a numerical
superiority on other parts of the board ,
avoid i ng giving Black something to latch
onto for the creation of counterplay.
1 0 . tbg4
. .
The retreat 1 0 . . . CL!e8, preparing . . .f7-f5 ,
does not satisfy the non-routi nely thinking
Morozevich . However, now Black has to
reckon with h i s second kn ight also be
ing d riven to the edge of the board . T h is
creates the constant p roblem of fi n d
ing a suitable refuge a t t h e rear for the
kn ights, whose paths i ntersect at d8. In
this respect it i s useful to see how this
position was handled by Kasparov, who
in his game with Beliavsky (Linares 1 994)
preceded the kn ight move with 1 0 . . . Ad7.
There fol lowed 1 1 . Ab2 CL!g4 1 2. h3 CL!h6
1 3. e3 (Beliavsky chooses the plan with the
blockade of the e5/f5 pawn pai r; 1 3. e4 f5
1 4. exf5 gxf5 1 5 . 'Wc2 fol lowed by f2-f4 is
more often played) 13 . . .f5 14. f4 a6 1 5. 'Wc2
b5 1 6. CL!d1 (Wh ite takes his kn ight via f2
to d3, aiming to provoke a crisis at e5; i n
The Way to the Top
the event of 1 6. ae1 Black advances his
queenside pawns with gain of tempo: 1 6 . . .
b 4 1 7. CL!d1 CL! b7 and then . . . a6-a5-a4)
1 6 . . . b8 1 7. Ac3! (it is im portant to remove
the bishop from the 'X-ray' of the rook)
17 . . . e8 1 8. b1 CL!f7 1 9. CL!f2 exf4 (Black
has to keep the knight at a5 in view; for ex
ample, after the inclusion of the exchange
on c4, 1 9 . . . bxc4 20. bxc4 xb1 21 . xb1
exf4 22. exf4, he cannot play 22 . . . e3 be
cause of 23. Axg7 r:!l xg7 24. tLlf1 followed
by Wc3+) 20. exf4 e3 21 . Axg7 r:!l xg7
22. fe1 ! xe1 + (if 22 . . . We7 there fol lows
23. CL!f3 with the th reat of 24. xe3 Wxe3
25 . e1 ) 23. xe1 Wf6! (it is i m portant to
occupy the long diagonal) 24. cxb5 axb5
25 . b4 cxb4 26. 'We? 'Wd8 27. 'Wxd8 xd8
28. CL!d3 c8 with an equal game.
1 1 . h3 tbh6 1 2 . tbde4!?
A com pletely new idea in this position.
It is obvious that if 12 . . .f5 there follows
1 3. CL!g5 with the th reat of an i nvasion at
e6. I ncidentally, about the effect of a nov
elty. It often operates l i ke a psychological
bom b, and the reaction to it is burdened
by this psychological factor: the opponent
knows something that I don't know. And
to go along with the opponent involves a
risk: at the board will it be possible to find
everyth i ng which he prepared at home?
Morozevich d i d not take the risk, but he
went, as they say, from the frying pan into
the fire.
12 f6?!
...
Black refrains from active measures, pre
paring the knight retreat to f7 and reserving
the move . . . f7-f5 for better ti mes. Even so,
there were no grounds for rejecting the
main l i ne of play.
\
102
 M. Carlsen - A. Morozevich
------

a b c d e f 9 h But a con u n d rum was posed by

8 .!. .i. 'iiV .!. 8 1 5 . . . tbf5!?, which would have set W h ite
7 .. -*- 7
d ifficult problems. For example: 1 6 . Aa3
!;!e8!? (after 1 6 . ..Wc7 1 7. tbe4 b6 1 8. d6 Wf7
6 ' .&6 1 9. tbxf6+ Wxf6 20 . Axa8 Black achieves
5
1.& 5
less) 1 7. tbe4!? (1 7. g4 tbh4 1 8. d6 tbxg2
4 ttJ 4 1 9. xg2 Af8 with the th reat of . . . Wxd6
3 tt:J and . . .b7-b6) 17 . . . b6 1 8. g4 tbh4 1 9. tbd6
2 2 tbxg2 20 . tbxe8 tbf4 21 . tbxg7 xg7 with
1.: unclear play.
b d f 9 h
a c e
16. d6 e4 1 7. d7

13. xd6!? After t h e i nclusion of t h e exc h ange

1 7. Axh6 Axh6 the advance 1 8. d7 !;!f7
This positional sacrifice of a piece for two 1 9. dxc8W Wxc8 would have allowed Black
pawns (the one on c5 cannot be held) counterplay, based on the advantageous
sharply changes the character of the play. opposite-colour bishops.
And although it is to a certain extent i ntui
tive, with his poorly coord inated pieces it is 17 . . .f7 18. b1 We7 1 9. dxc8W axeS
not easy for Black to restrain the d5-pawn . 20. a4 fd8 21 . We1 c6
Taking account of t h e surprise effect, he is
forced to solve com pletely new problems.
8
13 . . .Wxd6 14. e4 Wd8 7
After 1 4 . . . \Wb6 1 5 . Ae3 the pawn can not 6 1.&
5
be saved .

15. xc5 f5
4 ttJ
3
The psychological effect of the piece sac 2
rifice has worked . It is not in Morozevich's
II
style to try and 'dig in'. By return ing the ma a b c d e f 9 h
terial, he coord inates his pieces. However,
with his extra pawn Magnus also feels com This was the position for which Morozevich
fortable. But did Black have anything bet was aiming, when he went in for the forcing
ter? He would have been unable to solve play begi n n i ng with 1 5 . . . f5. For the sacri
his problems after 1 5 . . . Wc7 1 6 . Aa3 !;!d8 ficed pawn Black has compensation , in the
1 7. tbd3 f5 1 8. Wc2. For example, it is risky form of his more active pieces.
to play 18 . . . e4 1 9. tbf4 Axa1 ?! 20. l;!xa1 ,
22. c3 d4 23. Ab2
when he m ay be unable to survive the
'draught' along the long diagonal. He would If 23. tbd5 Black could have sacrificed the
have lost material after 1 5 . . . tbf7? 1 6. tbe6 exchange: 23 . . . l;!xd5!? 24. cxd5 !;!c2, ob
Axe6 1 7. dxe6 tbd6 1 8. e7 Wxe7 1 9. Wd5+. tai n i ng defi n ite com pensation . But after

103
 Chapter 2 The Way to the Top
25 . h1 tbxe2 26. Ad2 tild4 27. c1 xa2 27. WaS
28. c8+ Af8 29. Ab4 tild6 30. 'Mr'd1 tD4d5
31 . c2 xc2 32. 'Mfxc2 White's chances
are sti l l better.
23 . . . bS 24. tildS
24. e3 suggests itself, after wh ich the rna-
noeuvre 24 . . .tbc2 25. 'Mfe2 Axc3 26. Axc3
tDa3 27. bd1 looks dubious i n view of
the weakness of the long d ark-square
d i agonal, while 24 . . . tilf3+ 25 . Axf3 exf3
26. tilxb5 would have led to the win of a
second pawn , although also with an ele-
ment of risk - the wh ite ki ng's defences
are weakened .
24 . . .'Wd6
24 . . . 'Mr'c5!? was more vigorous, when i n
the event of 25. tDf4 'Mr' d 6 26. cxb5 c2
27. Axd4 Axd4 28. d1 e3 29. fxe3 Axe3+
30 . h2 'Mr'e5 Black's active pieces would
have given him defi n ite compensation for
the sacrificed material.
2S. Axd4 Axd4 26. !ad1 AeS?
Black m isses a chance opportunity to
exchange a pai r of pawns: 26 . . . bxc4!?
27. xd4 cxb3 28. axb3 c5 29. f3 xd5
30 . xd5 'Mr'xd5 31 . fxe4 fxe4 32. 'Wc3,
when Wh ite's achievements are less than
in the game.
a b c d e 3
8 8
6 6
5 5
4 4
a b c d e f 9 h
1 04
It should not be forgotten that this game
was played in the fi rst round of the most
prestigious tournament of the year, and
after his fai lure in Wijk aan Zee Carlsen
was sub-consciousl y a i m i n g for clarity.
Possibly, therefore, he prefers to transpose
into a technical endgame, rather than play a
midd legame with a couple of extra pawns:
27. cxb5 'Mr'c5 28. b4 'Mr'd6 29. a4, wh ich
involves some risk.
27. . . bxc4 28. tile3! Wc7 29. Wxc7 Axc7
30. tilxc4 tileS
As a result of the exchanging operation
Wh ite has obtai ned an endgame with an
extra pawn and a stable pawn structure,
where Black's practical saving chances are
associated with the opposite-colour bish
ops. With every exchange these chances
are improved .
31 . E:i:xd8+ !axd8
a b c d e f
8 .!. 8
7 ... .t..
6 6
5 5
4
tlJ 4
::,
2 ::,
a b c d e f 9 h
32. E:i:c1 !
The active kn ight has to be exchanged ,
after which Wh ite wi l l 'revive' his bishop,
regroup his forces, and at a conven ient
moment underm i ne Black's pawn chain
with g3-g4.
 M. Carlsen - A. Morozevich

32 !Llxc4 33. xc4 d1 + 34. Af1 Ad6

. 43 .
35. e3 a5!
It is useful to halt the pawns at the distant
approaches to the queening squares.
36. mg2
The prophylactic 36. f!c2 was more ac
curate, preventing the possible 36 . . . f!d2.
36 mf7 37. c2 me7 38. Ae2 d5
..
After 38 . . . f!a1 White would probably have
carried out the same plan as in the game.
39. Ac4 d1 40. g4
mf6
43 . . . f!xd3? fails to 44. f!xe4+ and 45. f!xf4.
44. Axe4 d2! 45. xd2 Axd2 46. mg3
If White were able to advance f2-f4 with his
pawn on h3, the position would be won for
h i m . But he is not able to do this.
46 . . . Ae1 47. mt3 Ab4 48. h4 h6
49. me2 Ad6
The ki ngside should have been blocked -
49 . . . g5 50. h5, when it would not be easy
for Wh ite to fi nd a winning plan.
so. md3
a b c d e f 9 h

a b c d e f 9 h 8 8

8 8 7 7

6
..t . ...... 6
5

4 [j,[j, 4

3 [j, 3

2 [j, [j, 2

a b c d e f 9 h

40 . . f4?
A com m itting decision . Any neutral move,
such as 40 . . . f6, would h ave retai ned
practical d rawing chances. Now, however,
White wins the e4-pawn . True, it is with the
help of a tactical trick, which was easy to
overlook when making the last move before
the time control.
41. exf4 Axf4 42. e2 d4 43. Ad3!
50 .. Ac5
Black m isses the last opportunity to play
50 . . . g5, when after 51 . h5 e5 52. f3 Ab4
it would not be easy for Wh ite to convert
his material advantage. H i s chances are
associated with the creation of an outside
passed pawn: 53. c4 Aa3 54. b5 Ab4
55. a4 Ad2 56. a3 d6 57. b5 Ae1
58. b4 axb4 59. axb4 Ad2 60. c4.

Now Black has to play on two pawns down.

The only consolation is the possi bil ity of
exchanging rooks, transposing into a pure
ending with opposite-colour bishops. Re
taining the rooks - 43 . . . f!xd3 44. f!xe4+
leads to a hopeless rook end i ng. Analysis diagram

1 05
 Chapter 2 The Way to the Top
---------------------

The critical position . White would l i ke to a b c d e f 9 h

d i vert the black pieces by the advance 8 8
of the b-pawn and then create a second 7
passed pawn on the h -file with the help
6
of the f2-f4 breakth rough , which works,
for example, after 60 . . . Ae1 ? - 61 . f4! gxf4
5 ..t.
62. g5 . I n the event of 60 . . . Ac1 Wh ite wins 4
using the typical idea of restricting the op 3 [j, 3

ponent's ki ng- 61 . AdS! , enabl i ng his own 2 [j, 2

king to go to e4 and then on to the h6-pawn
- 61 . . . Af4 (if 61 . . . Ab2 the pawn advances a b c d e f 9 h
- 62. b5, wh ile if 61 . . . Ae3 62. d3 the king
breaks th rough to e4) 62. d4 Ad2 63. b5
58. f5 gxf5 59. Axh5 e7 was stronger,
Aa5 64. e4 Ab6 65. Ac4 and e4-f5
retai n i ng chances of saving the game.
(M. Marin).
The o n l y move lead i n g to a d raw i s 54. 'it>b5 Axf4 55. 'it>xa5 Ag3 56. 'it>bS
60 . . . Ae3! . I t i s i m portant to keep t h e Axh4 57. a4 Axg5
b6-square u n d e r contro l , without l ifti n g
Black has restored material equal ity, but
t h e control o f f4 , a n d t h e ' k n ight o p
the pawn race is obviously won for Wh ite.
position' o f the k i ngs with the sup port
of the b i s h o p e n a b l e s the position to a b c d e f 9 h

be hel d . If 6 1 . AdS or 6 1 . Aa8 there fol 8 8

lows 6 1 . . .e5 , a n d not h i ng i s given by 7 7

6 1 . b5 or 6 1 . d3 because of 6 1 . . . Ab 6 . 6 6
The march o f the k i n g t o a 6 - 6 1 . b 5
i s restrained by 6 1 . . . A d 2 . I ncidentally,
5 ..t. 5

here too the opposition can be m a i n

4 /j, 4

tai ned - 6 1 . . . c7 6 2 . a 6 b8, s i nce

3 [j, 3

2 2
after the advance of the pawn to b6 the
f3-f4 b reakth rough does not ach i eve its
aim: the passed h - pawn i s held by the a b c d e f 9 h

bishop, and if the wh ite k i ng goes to its

aid , Black wins the b6-pawn , by d iverting
the wh ite bishop with his passed pawns .
51 . f4 h 5 5 2 . g5+ 'it>g7 5 3 . 'it>c4
(see next diagram)
53 . . . Ad6?
58. aS 'it>f6 59. a6 Ae3 60. 'it>c6 g5
61 . b4 'it>es 62. b5 'it>xe4 63. b6 g4
64. a7 g3 65. a8'W 'it>f3 66. b7 Af4
67. Wf8 'it>e4 68. 'We8+
Black resigned .

It is i l logical to give the opponent con

nected passed pawns. 5 3 . . . Af2 5 4 . b5
Ae1 55. a3 f7 56. Ad3 g7 57. Ae2 f8

no6
 M. Carlsen - V. lvanchuk
--------------------
tt:J

M. Carlsen - V. Topalov .'Da5 . Nowadays plans are also practised

Morelia/Linares 2007 where Black does not hurry to relieve the
a b c d e f 9 h pawn tension. lvanchuk follows the tradi
tional path, retai ning t h e option after the
8 8
bishop retreat of reverting to fam i liar l i nes
7 'if "' 7 by exchanging pawns, but the course of
6 I. I. VJfi 6
events in the game, where the kn ight at a5
51.1.8 I. QJ 5 played the role of a spectator, may subse
4 4 quently have led h i m to another, possibly
3 88 3 more p rom i s i ng conti nuation: 10 . . . Vf!c7
1 1 . l::! c1 l::! d 8. The game Chepari nov - lv
2 8 2
anchuk (Cap d'Agde rapid 2008), continued
a b c d e f 9 h 1 2. Af4 Ae5!? 1 3. Axe5 .'Dxe5 1 4 . Ab3
.'Dg4 1 5 . .'Dg3 Vf!f4 1 6. Vf!f3 Vf!xf3 1 7. gxf3
Here somet h i n g i nexp l icable occurred LDf6 1 8. l::!fd1 f8 1 9 . .'De2 b6 20 . e5 .'Dh5
with Topalov. He resigned the game, not 21 . d xc5 Aa6 22 . .'Dg3 "Df4 23. cxb6 axb6
finding any way to save h i s kn ight, and with excel lent compensation for the pawn ,
he merely sh rugged his shoulders i n per typical of G ri.infeld-type positions.
plexity when Magnus showed him how to 1 1 . !J.d3
do this: 64 . . . Vf!d5+ 65. f3 e5. Wh ite has a b c d e f 9 h
nothing more than perpetual check by 8 .!. .i.
66. Vf!h7+ f8 67. Vf!h8+ Vf!g8 68 . .'Dh7+
<;!;>f7 69 . .'Dg5+.
:--:.,-_-
7 1._,..:,1.
6

5"' 5
4 4
Game 22
3 3
M. Carlsen - V. lvanchuk
Morelia/Li nares 2007
28
GrOnfeld Defence [ 087] 1:
1. d4 CLlf6 2. c4 g6 3. CL!c3 d5 4. cxd5
a b c d e f 9 h
tDxd5 5. e4 CL!xc3 6. bxc3 IJ.g7 7. /J.c4 11 b6
..

c5 8. CL!e2 CL!c6 9. !J.e3 0-0 1 0. 0-0
This is the so-cal led main variation of the
Gri.infeld Defence. It was here that the first
attem pts were made to refute what has
turned out to be the 'evergreen' idea of
the Austrian grandmaster Ernst G ri.infeld
(3 . . . d5!), fi rst carried out in 1 922.
10 CL!a5
. ful l com pensation .
Over previous decades Black's main con-
This variation has become popular in recent
years. Of course, Black could also have re
turned to the main line by playing 1 1 . . . cxd4.
1 2 . lac1
If White captures the pawn - 1 2. d xc5?!
bxc5 1 3. Axc5, then after 1 3 . . . Vf!c7 1 4. Ad4
e5 1 5 . Ae3 Ae6 1 6 . Vf!c2 .'Dc4 Black has
12 cxd4
.

tinuation was 10 . . . cxd4 1 1 . cxd4 Ag4 1 2. f3 With the rook on c1 , Carlsen considers

1 07
 Chapter 2
the open ing of the c-fi le to favour Wh ite.
I ndeed , to neutralise the control of the c
fi le Black has to go in for the exchange
of rooks, which reduces the possibility of
creating piece play typical of the G ri.in
feld Defence. 1 2 . . .'\Wc? looks to be i n the
spi rit of the position, although it is not
easy to shake the pawn centre. Black's
problems are i l lustrated by the game Ja
kovenko-Areshchenko (World Cup, Khan
ty-Mansiysk 2009): 1 3. 'Wd2 Ab7 1 4 . Ah6
fiad8 1 S . Axg7 xg7 1 6. f4 fS?! 1 7. exfS
gxfS 1 8. 'We3 'Wd6 1 9. ficd 1 c4 20. Ac2
AdS 21 . 'WeS+ WxeS 22. fxeS e6 23. f4
f7 24. fide1 e7 2S. fie3 bS 26. fib1 a6
27. fih3 fit? 28. fih6 c6 29. Ad 1 ! 1 -0.
After 1 3. 'Wd2 Black can consider 1 3 . . .
eS, when i n the event of 1 4 . d xeS AxeS
1 S . f4 Ag7 1 6 . eS fid8 or the i m med iate
1 4 .. . fid8 (i ntending 1 S . AgS Aa6!) he ob
tai ns excellent play. However, 1 4. f4!? is
more critical .
1 3. cxd4 e6
If 1 3 . . . A b7 there would h ave fol l owed
1 4. dS! , restricting the bishop, and Black is
obliged to prevent this move. However, this
leads to a weakening of the dark squares
on the kingside, which creates the grounds
for an attack.
14. 'Wd2 .tb7
a b c d e
8 .!.
'if 8 Baramidze (Bundesl iga 2007) continued
7 .l .i.
6 .l 6 1 9. WgS 'Wd8 20. 'Wg3 fixc1 21 . fixc1 fic8
5 5
4 4
3 3
2 12
a b c d e f g h
The Way to the Top
1 5. h4
A strong move, introduced by the German
grandmaster Rainer Knaak i n a game with
Malich (Halle 1 976). Wh ite begi ns a flank
attack without risking the loss of a pawn,
si nce in the event of 1 S . . . Wxh4? 1 6. Ag5
'WhS (th i ngs are not changed by 1 6 . . .'\Wg4
1 7. f3 'WhS 1 8. g3 Axd4+ 1 9. fif2) 1 7. g 3
'Wg4 1 8. Ae2 the queen is lost.
1 5 . 'We7?!
. .
Surprisingly, lvanchuk fol lows Mal ich, who
already then had become convi nced about
the d rawbacks to this move and had rec
ommended 1S . . .'Wd7, which Black adopted
in later games with his variation . True, with
out particular success. For example, the
game Knaak- Kir. Georgiev (East Germany
1 987) continued 1 6. Ah6 fiac8 1 7. hS! fixc1
1 8 . .xc1 fic8 1 9. Axg7 .xc1 + 20. Wxc1
xg7 21 . h6+! with advantage to White.
The move in the game is obviously not
in the spi rit of the position, si nce it assists
White's plan of exploiting the weakness of
the dark squares. Carlsen considers the
best defence to be 1S . . .c6, after which he
was i ntending to play 1 6. eS. But a rad ical
solution in the spi rit of the Sici l ian Dragon
also comes into consideration: 1 S . . . hS!?,
nipping in the bud the threat of the h-pawn's
advance. For example, the game Tim man
1 6. AgS 'Wd7 1 7. Ah6 fifc8 1 8. Axg7 xg7
22. fie1 c6 23. dS b4 24. Ab1 exd5
2S. a3 c2 26. fid 1 , and here 26 . . . We8!
would have led to equal play.
16. h5!
This is stronger than 1 6 . AgS f6 1 7. Af4, as
played in the Knaak- Mal ich game, where
1 08
 M. Carlsen - V. lvanchuk

Wh ite did not achieve anything significant. tDe2-f4. For exa m pl e: 22 . . .f8 (22 . . . e8
23. h xg6 h xg6 24. tDf4 tDxd4 25. tDh5! or
16 . . . l:UcS 1 7. e5!
22 . . . Aa6 23. Axa6 xa6 24. tDf4) 23. hxg6
a b c d e f 9 h h xg6 (23 . . . fxg6 24 . tDf4) 24. tDf4 A h 6
a .I .I 8 2 5 . h 4 tDxd4 2 6 . tDxg6 fxg6 27. xd4
etc.
7
.t j_ 'W.t .i. .t 7
6 .t .t .t 6 1 9. !axeS+ AxeS 20. Ag5 Y!/c7
5 1.1\ 5 Black misses the last chance for a suc
4
_..___
1-----
4
cessful defence: 20 . . .d7, as suggested by
3 3 lvanchuk in his commentary on the game.
2 Wl tiJ 2 'This is more resil ient', agrees Carlsen, ' But
even so after 2 1 . Af6 in any case it is very
1 .: .: 1
a b c d e f 9 h hard to believe that Black is able to sur
vive.' After the best reply 21 . . . tDc6 Black's
The concrete solving of problems by earls position is sti l l dangerous.
en , who i m mediately 'takes the bul l by the
horns', is impressive. Wh ite prepares the
ground for the exchange of the bishop on
g7, regard less of the weakening of the long
light-square diagonal, which Black is not
able to exploit . Previously 1 7. Ag5 a3
18. e5 was preferred, and, to judge by sto
ries about Magnus's outstand ing memory,
this idea was wel l known to h i m . Analysis diagram
Wh ite can conti nue the attack as i n the
1 7. . . 1a xc1 ?
game: 2 2 . g5 tDxd4 2 3 . Axg7 <;!;>xg7!
This was clearly not lvanchuk's day. The (weaker is 23 . . . tD x e2+ 24. Axe2 <;!;>xg7
exchange of all the rooks deprives Black 2 5 . h 6+! <;!;>fa 2 6 . f6 with the t h reat af
of his last hopes of creating counterplay. ter Ae2-d 3xg6 of quee n i ng the h - pawn ,
As Carlsen comments, first and foremost wh i l e i n t h e eve nt of 2 6 . . . d2 2 7. Af1
the kn ight should have been brought back Ab7 28. h8+ <;!;>e7 29. xh7 Wh ite picks
into play - 1 7. . . tDc6. up an i m portant pawn and succeeds in
The p rophylactic 17 . . . d7, p reparing to retu r n i n g w i t h h i s queen - 29 . . . A xg2
besiege the d4-pawn after 18 . . . tDc6, was 30. g7 d 1 31 . f6+ <;!;>fa 32. <;!;>xg2 with
also in the spi rit of the position. an extra piece) 24. f6+ <;!;>f8 25. tDf4!
(2 5 . h 6 is just i f i e d o n l y i n t h e event
1S. laxc1 lacS
of 2 5 . . . tD x e2+ 26. A x e 2 , but by p l ay-
If 18 . . . d7 Wh ite attacks: 1 9 . Ag5 l:::!.c 8 ing 25 . . . tDf5! B l ack h o l d s the position:
20. l:::!. x c8+ xc8 2 1 . Af6 tDc6 22. g5! 2 6 . h8+ <;!;>e7 27. tDf4 tDxh6 28. xh7
(22. Axg7 <;!;>xg7 23. Ae4 tDe7 al lows Black g5) 25 . . . g x h 5 2 6 . A x h7 c7 27. tD x h 5 ,
to h o l d the position) with t h e t h reat of and the attack conti nues .

09
 Chapter 2 The Way to the Top

Another way of attacki n g is 22. 'Wf4
with the idea of 23. Axg7 xg7 24. 'Wf6+
f8 25. Axg6! hxg6 26. h6 and the pawn
is i rresistible. Black loses after 22 . . . Axf6
23. gxf6 'Wd8 (23 . . . We8 24. Ab5 Ad7
25. Axc6 Axc6 26. We? leads to the loss of
a pawn) 24. hxg6 fxg6 (24 . . . h xg6 25. Wg5
f8 26. Axg6) 25. f7+! g7 26. Ab5 Ad7
27. Axc6 Axc6 28. f8W+, winning the bish
op. After the bishop retreat 22 . . . Af8 again
23. Wg5 is possible. Let us consider the
main variations:
a) 23. .tDxd4? 24. tDxd4 Wxd4 25. hxg6
fxg6 26. Axg6 and Wh ite wins;
b) 23 . . . Ab7?! 24. hxg6 fxg6 25. d5! exd5
(oth erwise 2 6 . d 6) 2 6 . A xg6 h xg6
27. Wxg6+ Ag7 28. e6 and wins;
White has carried out his plan, and now it
only remains to get to the king.
21 . . . c6
A belated return of the kn ight. But Black
has no defence against the th reat of Wd2-
g5 with a ' b reak-in' at g6. For exam ple:
2 1 . . . Af8 2 2 . Wg5 tDc6 23. h xg6 h xg6
24. Axg6 fxg6 25. Wxg6+ Ag7 26. Axg7
Wxg7 27. We8+, or 21 . . . Axf6 22. exf6 Wd8
23. Wg5 tDc6 24. hxg6 fxg6 25. Axg6! .
22. Wg5
Wh ite is inexorable in the conduct of his
plan. He is th reaten i ng to capture on g7
followed by h6+ and Wf6 , and if 22 . . . Af8
the bishop sacrifice is decisive.
22 . . . h6

c) 23 . . . tDe7 24. h xg6 fxg6 25. tDf4 A b7 A forced weake n i ng, s i n c e W h ite was
26. Wg4 AdS 27. Axe? 'Wxe7 (27. . . Axe? threatening 23. Axg7 xg7 24. Wf6+ f8
28. tDxg6!) 28. tDxd5 exd5 29. Wc8 with 25. h6 when the pawn cannot be stopped .
advantage to Wh ite - he is th reatening
23. Wc1 !
30. Wc6;
d) 23 . . . tDb4 (the most resi lient) 24. hxg6 The p i n o n the kn ight e m p h asi ses the
fxg6 25. Axg6! hxg6 26. Wxg6+ Ag7 hopelessness of Black's position.
27. tDf4 Wf7 28. Wg5 h7 29. Wh4+ g8 a b c d e f g h
30. Axg7, and by coordinating his queen
and kn ight (as Rybka ind icates), Wh ite
8
.t 8

combi nes an attack on the king with

7 .l 'iV .t .t 7

the th reat of winning one of the pieces:

6
.l j, .l .l 6

d 1 ) 30 . . . xg7 31 . tDh5+ f8 32. Wd8+

5 B B 5

We8 33 . Wf6+ Wf7 34. Wh8+ e7 4

B 4

35. tDf6 Wf8 36. Wh4; 3 j, 3

d2) 30 . . . Wxg7 31 . tDh5 Wh6 32. tDf6+ 2 B CiJ BB 2

g7 3 3 . Wg4+ Wg6 3 4 . tD e8+ VJg
h7 3 5 . W h 3 + W h 6 3 6 . tDf6+ a b c d e f g h
g6 37. Wg4+! Wg5 38. We4+
Wf5 39. Wh4 g7 40. tDe8+ g6 23 .. g5
41 . Wh8.
If 23 . . . Wd7 Wh ite wins by 24. hxg6 fxg6
21 . Af6! 25. tDf4 tDxd4 26. tDxg6 . But now to the

110
 M. Carlsen - V. lvanchuk tD

vertical pin is added a diagonal one, which a b c d e f 9 h

leads to the loss of a piece. 8 8

24. Ab5 Ad7 25. d5! exd5 26. d4 7

Axf6 27. exf6 'Wd6 28. A xc6 'Wxf6 6

29. Axd7 'Wxd4 30. g3 'Wc5 31 . 'Wxc5 5
bxc5 32. Ac6 d4 33. Ab5 f8 34. f4 4 4
gxf4 35. gxf4
3 3

Black resigned . 2
a b c d e f 9 h

The Candidates tournament in the new cycle of the individual world cham pionship
was a notable event in Carlsen's chess career. The very appearance of a 1 6 -year-old
candidate for the world c rown was an extraord i n ary phenomen o n . The only player
to have p revi ously achieved this was Bobby Fischer, among whose opponents were
the stars of the post-war wave of Soviet gran dmasters. Playing in the Candidates
Tou rnament in 1 959 were such titans of the Soviet Chess School as Paul Keres,
Vasily Smysl ov, Mikhail Tal and Tigran Petrosian , who are now l egends of world
chess. The present-day Candidates have their stars , and although the list of real
contenders for the chess crown is shorter, this does not reduce the i ntensity of the
fight at the top l evel .
At t h e Candidates Tournament i n Elista, h e l d on t h e knock-out system , Carlsen
was pai red with Levon Aronian , the brightest young representative in the chess el ite.
Forecasts regard ing the outcome of the match were one-sided. The clear favourite
was the Armenian gran d m aster, with whose strength and experience o n e could
hard ly com pare the potential of the uncommonly talented but not yet battle-hardened
you ngster. Here it was forgotten that, with prod igies, time moves on a different scale,
and that which takes others years, they resolve within months. And the Carlsen who
went into battle in Elista was a ' renovated' version, one who had m arked ly gained
in strength during his contact with the elite. True, in his opening preparation there
were still gaps , and the first game of the match appeared to confirm the unfavour
able forecasts. Carlse n , who was not yet ready for o pening disputes, went in for
an un pretentious anti - M arshall variation i n the Ruy Lopez and contrived to lose an
absolutely equal position . In the 2nd game the self-confid ent Aronian was not eager
to build on his success, and in the t h i rd game, like a sphinx, M agnus rose from the
ashes and level led the scores.

1 11
 Chapter 2 The Way to the Top

Game 23 c6 1 3 . Ab2 gives preference t o Wh ite,

M. Carlsen - L. Aronian Black's defensive resources should also
Candidates Match, 3rd Game n ot be underest i m ated . The attem pt to
Elista 2007 prevent the creation of the pawn centre
English Opening [A30] fai ls in this objective.

1 . f3 f6 2. c4 b6 3. g3 c5 4 .ag2
1 1 . e5! .axg2 12. mxg2 0-0 13. e4
.ab7 5. 0-0 e6 6. c3 .ae1 7. lae1 d5 a b c d e f 9 h
The m ost popular m ove, lead i ng to the 8 !.'.& 'if .. . 8
creation of a mobile pawn centre for Wh ite. 7 .l .S. .l .l 7
Lovers of the 'hedgehog' structure prefer 6 .l .l 6
7. . . d6 8. e4 a6 9. d4 cxd4 1 0 . xd4 V!fc7. 5 .l ttJ 5
8. cxd5 xd5 9. d4 4 88 4
This move is usually made after the pre 3 8 8 3
parato ry exchange 9 . xd5 V!! x d5. The 28 8r!? 8 2
following also occurs: 9. e4 b4 (9 . . .xc3 1 l,! :!
1 0 . bxc3 0-0 1 1 . d4 transposes) 1 0 . d4 a b c d e f 9 h
cxd4 1 1 . xd4 8c6 1 2 . xc6 V!! x d 1
1 3 . l::! x d1 Axc6. I n both cases this leads 1 3 . . . 'Wc8
to an early exchange of queens. A n ew idea for solving B l ack's opening
a b c d e f 9 h problems, compared with 1 3 . . . Af6 14. g4
8 J. I,& 'if. J. 8 cxd4, as was played in the game Yermolin
7 .l .a .a .l .l .l 7 sky -Atalik (Berkeley 2005). There fol lowed
1 5. Aa3 Ae7 1 6. Axe? V!! xe7 1 7. cxd4 l::! d8!
6 .l .l 6 1 8. d5 exd5 1 9. exd5 V!!d 6, and here White
5 .l '.& 5 could have retained the better chances by
4 8 4 20. l::! c 1 ! a6 2 1 . V!!f3 c5 22. l::! c d1 . Black
3 ttJ tt:J8 3 can also consider 14 . . .c6!? 1 5. Ae3 cxd4
2 /j,/j, 8 8 8 2 1 6 . cxd4 V!fd7 with an equal game.
1 :! :! rJ;; 1 4. 'Wg4
a b c d e f 9 h The alternative is 1 4. d5!? Af6 1 5. Af4 with
the better chances for Wh ite.
9 . . . xc3
14 . . . .at6 15. f3
T h e p re p a ratory exc h a nge 9 . . . c x d 4
1 0. V!! x d4 0-0 comes i nto consideration, With the th reat of e4-e5 and Ah6.
while if 10. xd5 there is the good reply 1 5 . . . mha 16. h4!
10 . . . V!! x d5.
M agnus always h as suc h an attacking
1 0. bxc3 .Q.e4 m ove i n mind, especially since 16 . . . h6
Although the experience of the main con does noth ing to prevent 1 7. Ag5! .
tinuation 10 . . . 0-0 1 1 . e4 cxd4 12. cxd4 1 6 . . . c6 1 7. .ag5!

112
-------
M. Carlsen - L. Aronian
-----------------
ctJ

a b c d e f 9 h of the simple 22. l:Xed1 fol lowed by an at

a ,E 'iV .. 8 tack on the queenside by a2-a4 .
7 .l .l.l.l 7
6 .l.i. 6
5 .l !k, 5
4 8 8 V/!i 8 4 6 6
3 8 l::iJ 8 3 8 .l 5
2 8 8 41 2 4 8 8 4
1 1:, : 3 l::iJ 8 3
a b c d e f 9 h 2 8 8 2
1 1:, :
1 7. . . cxd4 a b c d e f 9 h
For Black, who is under fire, it is not easy 22. h5!
to hold the position. 1 7. . . Axg5 1 8. hxg5
will not do, since it allows the rook to be Not only clearing the knight's path to the
included in the attack along the h-fi le. His key f5-point, but also with the aim of 'seal
difficulties are only aggravated by 1 7. . . d8 ing i n ' the king, which will be unable to
18. l:Xad1 cxd4 1 9. cxd4, when he has to render any assistance to its forces.
reckon with t h e d4-d5 b reakth roug h .
22 . . . c4 23. h4 d6 24. h6 Jac3?!
Therefore Black agrees t o t h e deforma
tion of his ki ngside pawn structure, which 24 . . _gc4 suggested itself.
may tel l not only i n the middlegame, but
25. l3:ac1 ! l3:fc8
also in the endgame.
For a rook, an open file is a gul p of fresh
18. Axf6 gxf6 1 9. cxd4
air. Aron ian makes use of the last practi
The queen exchange could have been cal chance. I n rejecting the capture of the
avoided - 1 9. f4 d8 20. cxd4 with the pawn - 25 . . . tll x e4, he appreciated Wh ite's
better game, but Carlsen has perceived cer i ntention of attaching 'epaulettes' to the
tain advantages for White in the endgame king at h 6 and d 6 - 26. f5! gd8 (after
and he provokes the opponent's reply. 26 . . . l:Xfc8? 27. l:Xxc3 xc3 the pin 28. l:Xc1
is decisive) 27. gxc3 xc3 28. d6.
19...e5
..
There a ppears to be n oth i n g better. If 1--
.t. .t. .t.
19 . . . d8 there is the good reply 20. f4,
1----,
[j .t. [j
when 20 . . .tbxd4? fai l s to 2 1 . ged 1 e5 .t.tt:J
22. xeS. I

20. Wxc8l3:axc8 21 . d5 a5 1.1\ ::.,

[j t::,cJ;;
Black transfers his kn ight to the d6 block :S:
ading point. 21 . . .b4 is illogical on account Analysis diagram

113
 Chapter 2 The Way to the Top

Tableau! - this was how the French would
have assessed the resulting position.
The i nvasion of the rook o n the c-fi le is
threatened, but supporting the kn ight leads
to the loss of a pawn, without essentially
changing the position: 28 . . . b 5 29. gc1
b4 30. a3 a5 3 1 . axb4 axb4 32. gc2.
Wh ite i ntends to play gb2, for example,
32 . . . gd7 33. gb2 gb7 33. gxb4! , since after
33 . . _gxb4 34. d7 gba 35. tDd6! Black has
no defence against the threats of 36. tDxf7+
and 36. tDc8.
26. xc3 xc3
29. e4! f8 30. g4 lic7
I n the event of 30 . . _gd3 31 . gg7 gxd5
32. g x h7 c2fg8 33. gg7+ c2fh8 34. gxf7
gas 35. gxf6 Black cannot cope with the
armada of white pawns.
31 . g7 b5 32. xh7 g8 33. g7+
h8
Wh ite is as though playing with an 'extra'
ki ng. If 33 . . . c2ff8 there fol l ows a b reak
t h rough by the h - pawn - 34. d 6! gd7
35. gg4 and 36. h7.
34. d6

a b c d e f g h a b c d e f g h
8 .8 8 .8
... ... 7 7 ... .I ... 7
6 ... 6 6 ... 6
5 ... 5 5 ... 5
4 ttJ 4 4 4
3 .I 3 3 3
2 2 2 2

a b c d e f g h a b c d e f g h
27. <tlf5! 34 . . . d7

34 . . . gb7 is more resil ient, with the idea of

This is where the depth of the h4-h5-h6
creating a passed pawn, but then a raid
manoeuvre is felt. White forcibly transposes
by the king proves decisive: 35. c2ff3 b4
into a rook ending where the black king is
36. c2fe4 a5 37. c2fd5 a4 38. c2fc6 b3 39. axb3
tightly 'sealed in' by the f5- and h6-pawns.
axb3 40. c2fxb7 b2 41 . d7 b1 'W+ 42. rl::lc7
27 <tlxf5 28. exf5 g8
Wc1 + 43. c2fd8 Wc5 44. g xf7, and after
ge7 the king h ides from the checks at g6.
Black cannot prevent the rook's i nvasion
o n the k i n gs i d e by 28 . . . gc4 in view of 35. f3 b4 36. e4 xd6
29. d 6 c2fg8 30. gd1 gca 31 . g4! when he After 36 . . . a5 37. c2fd5 a4 Wh ite wins by the
is i n zugzwang. It is not possible to avert same method as in the preced ing note.
the breakthrough by the g-pawn: 31 . . . c2ff8
37. xf7 a6 38. g4 g8 39. h7+ h8
32. d7 gd8 33. g5 c2fe7 (33 . . .fxg5 34. f6)
40. g5 fxg5 41 . f6
34. g6 and the h-pawn queens (variations
by Carlsen). Black resigned .

1 14
 M. Carlsen - L. Aronian

After the 4th game everything appeared to have come ful l circle. N ow on his guard ,
Aronian played with great composure and did not let Carlsen 'off the ropes'. For the
Norwegian, behind 1 1/2 -21/2 i n the match, only one chance remained - to make use of
his last Wh ite i n the 5th game. And he did this brilliantly.

Game 24 1 3 . . . c4
M. Carlsen - L. Aronian
Candidates M atch, 5th Game An unexpected decision in the context of
Elista 2007 the match .
Queen's Indian Defence [E12] I n making t h i s strategi cally risky move,
Aronian surely should have had a 'bomb' up
1. d4 f6 2. c4 e6 3. f3 b6 4. a3 Ab7
his sleeve. After all, by relieving the pawn
5. c3 d5 6. cxd5 xd5 7. c2 Ae7
tension (of course, 1 4. Axc4 Axe4 is un
8. e4 xc3 9. bxc3 0-0 1 0. Ad3 c5
favourable for White), he granted Wh ite a
1 1 . 0-0 c7
mobile pawn centre with prospects of an
1 1 . . . Wc8 came i nto consideration, with the attack on the kingside, which meant that he
idea of exchanging the bishop after . . . Aa6. had to aim to strike fi rst on the queenside.
But it would appear that he was merely
12. e2 d7 1 3 . Ab2
counting on Carlsen's psychological de
a b c d e f g h p ression after the defeated suffered the
a .! .1 a day before.
7 .i.. 'if,... .i.. 7 Usually in this variation Black mai ntains
6 6 the pawn tension by developing his rooks at
5 5 c8 and d8. H ere are two typical examples
on this theme:
4 4 Kas parov - K ra m n ik ( L i n ares 2004):
3 ttJ 3 13 . . .laac8 14. tDd2 lafd8 1 5 . lafd1 tDf6 1 6. a4
2 V/JJ 2 cxd4 1 7. cxd4 Ab4 18. tDf3 We? with equal
1.: .: chances;
a b c d e f g h Ponomariov - Leko (Tal Memorial , Mos
cow 2006): 1 3 . . .lafd8 1 4. d5 exd5 1 5 . exd5
Carlsen had experience of playing the Pet Af6 1 6. c4 b5! with good counterplay for
rosian Variation, although with the black Black.
pieces. At the tournament in Biel (2005)
Gelfand attacked h i m with 1 3 . e5, and 14. Ac2 b5 1 5. Ac1
this is what happened: 13 .. J:!.fd8 1 4 . h4
After Black has 'locked ' the long diagonal,
f8 1 5. Ae3 Wc6 1 6. lafd1 laac8 1 7. h 5 h 6
the bishop is doing nothing at b2 , and so
18. e1 tDh7 1 9. Wg4 g5 2 0 . h2 h8
it joins the attack on the ki ngside, ai ming
21 . a4 tDe4 22. ladc1 Ag5 23. f4 f5 24. We2
for g5 .
Ah4 25. tDf3 Ae7 26. Ab5 We? 27. Ac4
d7 28. Ab5 We? 29. Ac4 Wd7 1/2 -1/2 . 1 5 . . . a5 16. Etb1

11 5
 Chapter 2 The Way to the Top

a b c d e f g h This obvious move decides the outcome

s_! .. . 8 of the game, since the attack against the
7 .i. 'iV 1.& ll. I. I. 7
u n d efe n d e d k i n g s i d e l ooks extre m ely
6 I. 6 d angerous . But also i nterest i n g is the
p ragm at i c view of the young p rofes
5 s i o n a l , who c o n s i d e rs the best m ove
4 4 to be 1 9. x b4! with the win of a pawn .
3 [:::, 3 The rook cannot be taken because of a
2 stan dard attack: 1 9 . . . A x b4 20. A x h7+
'i!>xh7 (or 20 . . . 'i!>h8 2 1 . g5 g6 22. 'Wf3!
a b c d e f g h 'i!>g? 23. 'Wf4 with the th reats of xe6+
and 'Wh4) 2 1 . g5+ 'i!>g8 (2 1 . . . 'i!>g6 l eads
1 6 . . . Aa6?! to mate after 22. 'We4+ f5 2 3 . exf6+ 'i!>h5
24. 'W h7+ 'i!>g4 25. h3#) 22. 'W h 5 tea
When there are attacks on opposite wi ngs,
23. 'Wxf?+ 'i!>h8 24. e1 ! , and B l ack has
every tempo is i m portant. Therefore 1 6 . . .
n o defence against the switc h i ng of the
b4! 1 7. axb4 axb4 1 8. cxb4 a2 was more
rook to h3.
logical, with sufficient compensation for
the sacrificed pawn (Carlsen). 1 9 . . . ctlb6

1 7. e5 b4? Naturally, 19 . . . Axg5? 20. xg5 h6 21 . 'We4

f5 22. xe6 'Wc8 23. Wt4 would have been
But now this move is too late. It was time
suicidal.
for Black to concern himself with his ki ng:
1 7 . . . fe8 1 8. g5 f8 1 9 . 'W h 5 A xg5 20. 'We4 g6 21 . 'Wh4 a:a7
20. A xg5 with a somewhat i nferior, but
Not 21 . . . Axg5 22. xg5 h5 23. xb4 d5
sound position.
24. e4! , when 24 . . .xb4 fails to 25. f6+
If 1 7. . . fb8 Black has to reckon with the
'i!>g? 26. xh5+ with inevitable mate.
breakth rough 1 8. d5!? exd5 1 9. e6.
22. At6!?
1 8. axb4 axb4
a b c d e f g h
Of course, the extra pawn after 22. Axe?
a_! .. . 8 Wxe7 23. 'Wxe7 xe7 24. x b4 d 5 is
too small a gai n i n such a powerful po
7 'iV I.& .i. l.l.l. 7 siti o n . Carlsen was a i m i n g o n l y for an
6 .i_ I. 6 attack, but, h av i n g c o o l e d down after
5 [:::, 5 the battle, h i s p ragm at i c v i ew i n c l i ned
4 1.1.:::, 4 i n favour of 22. xb4, since here too the
3 [:::, 4J 3 win of the exchange - 22 . . . A x b4 23. Af6
2 'fl[:::,[:::,[:::, 2 b8 24. cxb4 - l eads to a very strong at
tack on the king (25 . g5 is th reatened).
I;! I;!
a b c d e f g h
After 2 2 . . . d5 23. a4 W h ite rem a i n s
a pawn up (23 . . . xc3? 2 4 . Axe? 'Wxe7
1 9. Ag5 25. 'Wxe7 xe7 26. xa6 leads to the l oss

116
 M. Carlsen - L. Aronian

of a piece), retaining th reats to the king. a b c d e f g h

For exam p le, in the event of 23 . .. .Ab5 8 I. 8
24 . .xa7 'Wxa7 25 . .Ae4 Black's position
is unenviable.
7 r
.t. .t. 7
6 ... ... 6
22 . . . Axf6 23. exf6 d5 5 .t. 5
a b c d e f g h 4 4
8 .. . 8 3 I. ttJ 3
7 I. 'iV .t. .t. 7 2 2
6 .*_ 6
5 '.& 5 a b c d e f g h

4 ... ... VJJJ 4 31 . h4!

3 ttJ 3 Signal ling the start of an attack on the
2 ii, 2 kingside. Although Black's defences ap
a b c d e f g h a decisive storm .
24. Ae4
Amazing pragmatism ! By th reatening the
exchange on d5 Carlsen provokes the ex
change of q ueens, preferring to go into an
endgame with an extra pawn, rather than
an attack with 24. tt::le 5, which looks very at
tractive. For example: 24 . . . b3 25. Ae4 'Wd6
26. f4 .e8 27. .Axd5 exd5 28. f5, building up
the threats. This is correct, but what can be
suggested after 24 . . . bxc3 25. Ae4 'Wd6?
24 . . .'fif4 25. Axd5 Wxh4 26. xh4
exd5 27. x b4
White's main tru m p in this ending is the
'nail' on f6. By sealing the kingside, it ef
fectively guarantees White an extra piece
- his king, the role of which in the endgame
is hard to overestimate.
27. . . Ac8 28. b6
Prophylaxis against the th reat of . . . .!:!a6.
28 . . . a3 29. c1 Ae6 30. f3 fa8
pear solid , Magnus sees a way to launch
31 . . . h6 32. e5 a1 33. xa1!
The only way of fighting for a win . If
33. .bb1 Black can reply 33 ....1 a3!, re
stricting White's offensive actions.
33 . . . xa1 + 34. h2 a3 35. b8+ h7
36. f4!
An essential link in White's plan is the diver
sion of the bishop ( f4-f5 is th reatened). To
convert his advantage it is important for
him to win the f7- pawn without exchang
ing the minor pieces, since in the event
of 36. .f8 .xc3 37. ti:lxf7 Axf7 38. .xf7+
<i!?g8 39 . .g7+ <i!?f8 40 . .xg6 .d3 the rook
ending is not worse for Black (variations
by Carlsen).
36 . . . xc3
(see next diagram)
37. h5!
A brilliant solution! The pawn offensive with
a small army creates a fine aesthetic im
pression. Black could not have prevented
this move by . . . h6-h5, since after ti:lf3-g5
he would have been mated.

117
 Chapter 2 The Way to the Top

a b c d e f g h Forced , since if 37. . . gS? there follows 38. fS

al :s Ia AxfS 39. xf7 with unavoidable mate.

7[ .l 1 7 38. laf8 laa3 39. f5! J.xf5 40. la xf7+

sl j_ .l.lls mg8 41 . lag7+ mts 42. la b7! laa8

51 .lti:J 15
43. mg3!

41 14 The king decides the outcome. If 43 . . . c3

3[ .! 13 there follows 44. mf4 c2 (44 . . . Ae6 4S. l:!e7)
21 1 2
4S. xfS! with mate in fou r moves.

1I 11 43 ladS 44. mt4 J.e4 45. g3! c3

a b c d e f g h 46. llt7+ mgs 47. llg7+ mts 48. tbd7+

laxd7 49. la xd7

37.. gxh5 Black resigned .

I n the 6th game Carlsen again deserved the highest praise - this time for the defence
of a difficult position.

Game 2S Capablanca- Vid mar (Carlsbad 1 929), the

L. Aronian M. Carlsen
- main continuation was considered to be
Candidates Match, 6th Game
Elista 2007
Slav Defence [017}
1 . d4 tbf6 2. tbf3 d5 3. c4 c6 4. tbc3
dxc4 5. a4 J.t5 6. tbe5 tbbd7 7. tbxc4
tbb6
ti:J
7. . .V!fc7 8. g3 eS 9 . d xeS xeS 1 0 . Af4
fd7 1 1 . Ag2 f6 , where in recent times
Black has begun to experience difficu lties
after 1 2 . xeS . For exam ple, the game
van Wely-1. Sokolov (Dutch Cham pion
ship 1 998) continued 12 . . . xeS 1 3. aS a6
1 4. 0-0 Ad6 1 S. e4 Ae7 1 6. V!ib3 with a
small but enduring positional advantage for
White. Therefore, not without reason , Ivan
Sokolov made a detailed study of 7. . . b6
8. eS aS and took it up. It was this that
6 Carlsen employed in this im portant game.
5
4 8. tbe5 a5 9. g3 tbfd7

3 t2J 3 This continuation is employed far more

2 2 rarely than 9... e6 1 0. Ag2 Ab4 1 1 . 0-0 0-0,
1 :s :s 1 and since the advance of the e-pawn is not
a b c d e f g h threatened , for the moment it is not essen
tial for Black to perform such pirouettes.
For many years, beginning with the game But also Carlsen's decision to exchange

118
 L. Aronian - M. Carlsen
the active knight rather earlier is probably
no accident.
10. ftlxd7
Nothing significant is p romised by the
knight retreat 1 0 . tDf3. For example, the
game Svid ler - Wang Yue (China 2008)
continued 1 0 . . . e6 1 1 . e4 Ag4 1 2 . Ag2 e5
1 3. 0-0 exd4 1 4 . xd4 f6 1 5 . Ae3 1/2-1/2.
If 1 0 . tiJd3 there can fol low 1 0 . . . e5!?
1 1 . d xe5 tDc5 with active counterplay for
the pawn.
1 0 . . .'Wxd7
In the event of 1 0 . . . tDxd7 1 1 . Ag2 e6
1 2 . 0-0 Ab4 1 3. e4 the centre has to be
conceded without sufficient reason .
1 1 . e4 .th3
In the game P. H. Nielsen -Timman (Malmo
2002) Black d ecided first to weaken
White's kingside: 1 1 . . . Ag4? ! 1 2 . f3 Ah3
13. A x h3 Wxh3 1 4 . W b3 .a6 1 5 . Ae3
Wg2? 1 6 . 0-0-0 Wxf3 17 . .he1 , and his
delay in castling led to defeat . It should
be mentioned that after the better 15 . . .
e6 1 6 . f2 White also has the advantage.
12. J.xh3 'Wxh3 1 3. 'Wb3 !aa6
a b c d e f g h en countering difficulties , Carlsen sac
The first impression is that Black has no
119
reason to be dissatisfied with his posi
tion. He has managed to remove White's
'sting' - the g2-bishop, one of the main
performers in such a position , supporting
the d4-d5 breakth rough, and at the same
time it is not easy for White to exploit the
awkward position of the rook on a6. How
ever, White's reply changes the picture.
14 . .tf4!
From here the bishop can attack the knight
on b6. In the game Gyimesi - Bu Xiangzhi
(Germany 2007) White developed his bish
op more passively - 1 4 . Ae3 and after 14 . . .
e6 1 5 . d 5 Ab4 1 6. d xe6 Wxe6! 1 7. Wxe6+
fxe6 1 8. Axb6 l:!xb6 1 9. 0-0-0 Ac5 20. f4
e7 he had an inferior endgame.
14 . . . e6 1 5. J.e5!
A well-known technique: before placing the
bishop on its designated square - c7, in
passing White tries to extract some benefit
by provoking a weakening of the kingside.
1 5 . . . .tb4
The d rawbacks to 1 5 . . .f6 a re revealed
by 1 6 . Ac7 Ab4 1 7. 0-0-0 0-0 1 8. tDe2 !
(after 1 8. tDa2 c5 1 9 . tD x b 4 c4!? Black
has serious counterplay: 20. Wa3 axb4
21 . Wxb4 tDxa4 22. Wxb7 l:!fa8) . Now, in
view of the th reat of tDf4 or even tDg1 , it is
not easy for Black to defend his e6-pawn .
He c a n hard ly contemplate either 1 8 . . . g5
1 9 . tDg1 Wg2 20. Wxe6+ h 8 2 1 . tDh3
with a com promised kingside, or 1 8 . . . Wg2
1 9 . Wxe6+ h8 20. tDf4! xf2 2 1 . l:!hf1
We3+ 22. b1 with a trapped queen. The
main d rawback of Black's position is ob
vious - his pieces are bad ly placed . On
rifices a paw n , hoping to gain time to
coordinate his forces.
 Chapter 2
16 . .txg7 lagS 1 7. Ae5 'Wg2
Black regains the paw n , a central one
moreover, which gives him the d5-square
- an excel lent base for his pieces. True,
for this he has to exchange his bishop,
weakening his dark-sq uare periphery. But
in White's position too the light squares are
weak. The question is whose strategy is
the more discerning.
1 8. 0-0-0 Axc3 1 9. 'Wxc3
White is hoping to extract more from the
position than the endgame after 1 9. bxc3
xe4 20. c4 e2 2 1 . c5 (not 2 1 . d3?
tbxa4! with the th reat of . . . b6) 21 . . . c4+
22. xc4 tbxc4.
19 . . .'Wxe4 20. la he1 'Wd5
Normally it is im portant to blockade the
weak point in front of an isolated pawn.
I n the given situation , where the king has
not castled , this is doubly im portant: the
opening of the position after a pawn break
through cannot be al lowed . This cou ld oc
cur after 20 . . .f5 2 1 . d5! tbxd5 (after 21 . . .
cxd5 22. c7 the b7-pawn is lost) 22. xd5
exd5 23. c5!, when Biack has no defence.
21 . b3 'Wd7 22. 'Wc2
a b c d e f 9 h
8 I. 8
7 .l 'iV .l .l 7
6 !. '.&\ .l .l 6
5 .l 5 24. 1ae3
4 [3::, [3::, 4 One gains the impression that both play
3 [3::, [3::, 3 ers are thinking in the same direction and
2 [3::, 2
[3::,
Ir ir
a b c d e f 9 h
Despite all his trickery, as yet Black has
not managed to consolidate his forces,
The Way to the Top
whereas White has his sights set on the
h7-pawn.
22 . . . 'We7
Carlsen dislikes passive play and he de
cides to sacrifice a pawn to activate his
pieces as m u c h as possible. There is
logic in this - there are plenty of weak
nesses in White's position, and Black can
greatly hinder the advance of the pawns.
In such a situation it is customary to rely
on intuition, but the conseq uences of pas
sive defence also have to be weighed up:
22 . . . g6!? 23. <!>b2 tDd5 24. h4 h5 25. e2
g4 26. f3 g6 27. g4 e7, and Black is
still holding the position . Nothing is given
by either 28. gxh5 h6, or 28. g5 b4,
while if 29. f4 , then 29 . . .f6 looks a satis
factory reply.
23. 'Wxh7 lag6
After the preparatory 23 . .. a3+ 24. <!>d2
g6 Black h as to reckon with 25. e3!
tbd5 26. h8+ <!>d7 27. f3, reminding
Black that he also has a king. Therefore
Carlsen wou ld have to seek the exchange
of q ueen s by 24 . . . a2+ 2 5 . c2 , but,
knowing Aronian's tech niq ue in the con
version of a minimal advantage in sim ple
positions, he does not want to al low White
a comfortable life and he prefers to go
fishing in troubled waters.
are not in a h u rry to exchange q ueens.
Carlsen is afraid to go into an endgame,
while Aronian is also not in a rush , hoping
to finding chances in the middlegame, in
which he is hardly taking any risk: it is dif
ficult for Black to create any cou nterplay,
1 20
 L. Aronian - M. Carlsen

since he has to worry about his own king. 35. 'We2

After 24. h8+ Black wou ld have been
unable to avoid the queen exchange, since
with his king in the centre he would have
to reckon with the opening of the position.
For example, if 24 . . . @ d7, then 25. d5! fol
lowed by 26. b8 (the immediate 25. b8
does not work because of 25 . . . a3+) and
the queen , coordinating with the bishop,
creates dangerous th reats.
24 . . . ttld5 25. f3 'Wf8 26. h4 h6
Carlsen again decides against going into
the endgame after 26 . . . h6+ 27. xh6
xh6. For the same reason after the retreat
of the rook he does not play 27. . . a3+.
Carlsen has nevertheless won the psycho
logical battle. White's manoeuvres have not
led to anything, and Aronian himself offers
the exchange of q ueens. I n the event of
35. xf5 xf5 36. e2 @ c8 Black finally
resolves the problem of his king.
35 . . . g8!?
And this is al ready the chal lenge of a self
confident fighter. Black avoids the natu ral
exchange 35 . . . xe2+ 36. xe2, not fearing
the dou bling of his pawns by 36. xg4
xg4 37. l:!xf5 exf5 , since White can not
prevent . . . f5-f4.

27. 'Wc2 a8 28. 'Wd2 h5 29. b2 36. 'We1

'Wg8 30. e1 'Wg6 31 . Ad6 d8
Aronian appreciates his opponent's audac
The young Norwegian manoeuvres with ity and returns to the previous manoeuvring
his pieces like a high ly-experienced grand strategy.
master, observing the basic ru les of actions
36 . . . g6 37. e2
by the defending side. He avoids pawn
advances and any activity, keeping a care Here the exchange 37. xf5 exf5 is u nfa
ful eye on the opponent's manoeuvres. vou rable on account of the possible ma
31 . . . @ d7, not removing the defence of the noeuvre . . . l:!g6- e6-e4.
aS-pawn , can also be recom mended . But
37. . . f3?!
Magnus prefers to defend this pawn with
counterplay. a b c d e f g h
32. Aa3 'Wg4 33. d3 f5 34. e5 d7 8 8
a b c d e f g h 7 7
8 8 6 .I 6
7 7 5 "' 5
6 6 4 [j, [j, 'iV [j, 4
5 5 3 /j, a .l [j, 3
4 [j, 2 a [j, 2
3 /j, 3
2 2 a b c d e f g h
Up till now Carlsen has acted irreproach
a b c d e f g h ably, but in time-trouble his compos u re

1 21
 Chapter 2 The Way to the Top

deserts him . He forces events, changing
the character of the position, although it
was possible to continue the strategy of
indirectly defending the aS-pawn, by at
tacking the f2- and d4-pawns: 37. . . .g8
38 . .c2 'l;\'g6 39 . .d1 (39. '1;\'xaS? ebf4)
39 . . . 'Mlf6 etc.
38. : xf3 'Wxf3 39. :c2 Wf6 40. Wxa5
Wxd4+ 41 . a2 :f6 42. f4?!
After Black's lapse on the 37th move
White has managed to break through to
the queenside, at the same time ridding
himself of his isolated d4-pawn . However,
by removing his f-pawn from attack, he cre
ates a backward pawn on g3, which signifi
cantly eases Black's defence. The correct
continuation was pointed out by Rogozen
ko: 42. Ac5! We4 43 . .d2 .f3 44. Wa7
rile? 45. Ae3! with the idea of 46. 'Mld4 ,
exploiting t h e fact that 45 . . . ebxe3?? i s not
possible on account of the mating attack
46. WaS+ b6 47. 'Mla7+.
42 . . . :g6 43. Was
This leads to sim plification , which could
have been avoided by 43 . .d2 with the
probable continuation 43 . . . 'l;\'e3! 44. Wb4
rile? 4 5 . Wd 6+ ril b 6 4 6 . Wd8+ rila6
47. Wa8+ rilb6 (47. . . 'I;\'a? 48. Wxa7+ rilxa7
49 . .d3 leads to a better endgame for
White) 48. aS+ rile? 49. Ad6!? rild7 (after
49 . . . rilxd6? 50. 'Mld8+ rileS 51 . .xd5+! and
52. 'Mlb6 the queen is lost), and Black is still
holding the position .
4 3 . . . b4+ 44. Axb4 Wxb4 45. :g2
The last attempt to retain an advantage -
45 . .c4 'Mld2+ 46. rila3 rile?! 47. Wf8 .xg3
48. 'Mlxf7+ Wd7 49. Wxd7+ rilxd7 would
have resulted in a d rawn rook endgame.
a b c d e f 9 h
8
7 7
6 6
5 5
8 8 4
8 3
:t 2
a b c d e f 9 h
45 . . . :xg3!
A worthy conclusion to this titanic struggle.
46. : xg3 Wd2+ 47. a3 Wc1 +
Draw.
Such a d raw is worth any win . To hold out
in a very difficult battle at a critical moment
of the match signified a moral victory for
Carlsen.

And so, ' normal time' ended in a 3-3 d raw, and the match continued with a tie-break
of fou r rapid games. And everything repeated itself again. Aronian won his first 'White'
game and for the third (!) time took the lead in the match . After two d raws Carlsen had
one last chance to make use of the white pieces in what was already the 1 0th game
of the match . And he performed a miracle, fighting 'to the last bullet', and win ning a
theoretical ly d rawn endgame.

1 22
 M. Carlsen - L. Aronian

M. Carlsen L. Aronian
- (making it taboo for Black to exchange
Candidates Match, 1 0th Game the queens), and the maxim u m activation
Elista 2007 of the king.
a b c d e f g h
69. h3 'Wb1 70. Wf6+ h7 71 . Wf3
8 8 g7 72. h4

The techniq ue of coordinating the king

and q ueen while keeping the pawn on
its initial square is impressive. In a rapid
play finish it is not easy to defend - it is
im portant not to run into an exchange of
q ueens. Black exaggeratedly plays with
out checks , on ly moving his king, pl acing
a b c d e f g h it on safe squares .

72 . . . h7 73. Wf7+ h8 74. Wh5+ g7

In pre- reform times, when games were
played with the classical time control and
there were adjou rn ments, one cannot im
agine such an ending being resumed . The
pawn cannot be converted. This is obvious,
but not for the present-day 'condensed '
times, when the competitive element pre
vails over logical play. Nowadays, fol lowing
the FIDE reforms, games are played to a
finish, and players are ensured against los
ing on time by the addition of extra seconds
after each move. And so Carlsen plays on .
What is striking is the professionalism with
which he plays this ending, even under the
rapid time regime: here there is the rule 'do
not hurry' (it is more difficult to defend than
to attack), restraint in advancing the pawn
75. Wg4+ h7 76. Wh5+ g7 77. We5+
h7 78. Wd5 g7 79. Wd4+ h7
80. Wd7+ h8 81 . g4
The pawn comes into play at the most ap
propriate moment, provoking Aronian into
giving checks with his q ueen .
81 . . .Wh1 + 82 . g5 Wc1 +?
This next im p u l sive check leads to de
feat. The position cou ld have been held by
82 . . . 'We4 or even 82 . . . 'Wa8!?, with a tacti
cal point: 83. 'Wf5 'Wg8+ 84. h 6 'Wg7+
85 . h5 'Wf7+ 86. Wxf7 - stalemate!
83. g6 Wb1 + 84. Wf5
Black resigned .

And so, the tie-break also ended in a d raw, 2-2 . The winner was d ue to be decided in
a blitz match of two games, and if necessary an 'Armageddon' game would be played .
But things did not come to this. The ability to play five-minute chess has its specific
features, and here the clear favou rite was Aronian , who won both games, and with
them the match .
Carlsen gave a critical appraisal of his defeat: 'Aronian i s very strong', he replied to a
correspondent of the magazine 64 Shakhmatnoe obozrenie. 'The match with him was
a good lesson ; I realised that there is still much to learn about chess, much to study.

1 23
 Chapter 2 The Way to the Top

Leven caused me a mass of unpleasant surprises in the middlegame. I n addition , I

basically cou ldn't get out of the opening with the black pieces.' But, despite the defeat,
it was this match that brought Carlsen up to a new level. The 1 6-year-old grand master
presented himself to the chess com m unity as a real contender for the world crown in
the near future. He presented himself in particular as an inflexible fighter! Three times
in a week he eq ualised , twice in decisive games - only those possessing a cham pion's
character are capable of this.

T h ree weeks for ' repairs' before the start of the third ' pi l l ar' of the world chess
calendar - the Sparkassen G M Tou rnament in Dort m u n d (FI D E category 20) was
not a long time. It is wel l known that this was a kind of personal fiefdom of Kram nik,
who by that time had won there on eight occasions. The other participants were
also no pushover. Carlsen was in a cautious mood : ' W hen there are so many solid
opponents: Kram nik , Anan d , Gelfa n d , Leko - playing is not easy. They exert pres
s u re with their every move, not losing concentration for a seco n d . In this sense
Dortm u n d was in general the most diffic ult su per-tourn ament of the year for me:
Morelia/Linares or Wijk aan Zee was fu l l of fighters such as l vanchuk or Topalov,
who themselves play and also al low their opponents to play. Here there were no
such pl ayers.' Therefore he was not too disappointed with the resu lt: a loss to Kram
nik (where he was thoroughly outplayed by the cham pion in his favou rite Catalan
Opening) and fighting d raws with the rest was no reason to be u pset . Especially
since t h ree weeks l ater there was the tou rnament in Bie l , where the opposition was
not q uite so strong.
On this occasion the organisers of the festival in Biel had changed the format
of their m ain tou rnament. They increased the n u m ber of participants to ten , and,
despite the absence of Morozevich , the main victor of the last few years, they were
able to raise the level to F I D E category 1 8. Carlsen was happy: the opponents were
of those who themselves play and al low others to play. It seemed the right time to
play fighting chess, but he was contin uing to work on his own game, and in Biel we
saw a self-confident 'tech nician'. Magnus happily exchanged the q ueens, relying on
his greatly improved endgame tech nique, as he demonstrated in the very first round
in this game with Bu Xiangzhi. He was not concerned by the protracted nature of the
battle (the conversion of a positional advantage with opposite-colour bishops and
rooks took 50 moves). The tou rnament went well, and th ree rounds before the end
Carlsen was leading by a margin of one point. It seemed that victory was assured .
But at that moment he committed a childish error, inad missible for a 1 7-year-old
grand master.

1 24
 M. Carlsen - A. Onischuk

M. Carlsen Y. Pelletier
- the first p rize. They fought to the l ast:
Biel 2007 with a score of 2-2 things came down
a b c d e 9 h to an 'Arm aged don' (d raw in favo u r of
8 .! 1.1\ .! 8 Black, but he has 4 min utes on the clock) .
7 ..t And Carlsen again displayed his char

6 6 acter. He won the game with Black, and

moreover he won 'on c l ass'. Passions
5 died down - M agn u s had won his first
4 8 4 su per-tou rnament.
3 8 [jj li, 8 3
2 8 888 2
.: 1:'( 1 Game 26
a b c d e f 9 h M. Carlsen A Onischuk
- .

Biel 2007
After the natu ral 1 6 . f3 White is slightly Queen's Gambit [043]
better, but that is all. But Magn us played
1 . d4 f6 2. c4 e6 3. f3 d5 4 .tg5
1 6. Axh7?? (one can't help remem bering

the first game of the 1 972 Spassky- Fischer
match), and after the natural reply 16 f5 . .
17. e4 g7 1 8. exf5 xh7 1 9. fxe6 fxe6 he
was a piece down for obviously inadeq uate
compensation.
What was the cause of such inexcusably
cavalier play? Perhaps he was seeking
revenge for the miniature in his first tou rna
ment in Biel 2005, which he lost to Pel letier
in the opening (1 . d4 f6 2. c4 e6 3. c3
Ab4 4. c2 d5 5. cxd5 c5 6. d xc5 xd5
7. Ad2 AxeS 8. xd5 xd5 9. e4 d4
10. 0-0-0 d7 1 1 . Ab5 0-0?? 1 2 . Ac3,
and Black lost a piece). On the principle - it
depends who you are playing!
This chil dish ness cost him dearly. As
it is said , misfortunes do not come sin
gly. T h rough inertia Carlsen lost his next
game to van Wely and allowed himself
to be passed by Radj a bov, whom he
had to p l ay in the l ast rou n d . All t h at
remained was to display his character
and win ' to order', and although he did
this , M agn u s was caught by O nisc h u k .
They h ad t o p l ay a b litz tie - b reak for
By a non-traditional way the game has
entered the c u rrently popu l a r M oscow
Variation , the classical route to which lies
via the Queen's Gam bit (1 . d4 d5 2. c4 e6
3. f3 f6 4. Ag5). The variation received
this name after the famous Moscow I nter
national Tou rnament of 1 925, where it was
emp loyed by Eman uel Lasker and Efim
Bogolj u bow.
4 . . . h6 5 . .txf6
A popular continuatio n . Those who like
sharp play p refer to test their strength
(and home preparation !) in the so-cal led
Anti- Moscow Variation 5 . Ah4 c6 6 . c3
dxc4. As for Carlse n , he prefers clearer
play.
5 . . .'Wxf6 6. c3 c6 7. 'Wb3
Along with 7. c2, one of the main continu
ations in the Moscow Variation .
7. . . dxc4
Another branch 7. . .d7 8. e4 dxe4 9. xe4
f4 1 0. Ad3 e5 1 1 . 0-0 has been known

1 25
 Chapter 2 The Way to the Top

since the time of the 1 938 AVRO Tourna a b c d e f g h

ment.
.. . 8
1.& .. 7
6 'if 6
5 5
4 'o/i/ 8 4
3 t2J 8 t2J 3
2 /j, /j, 8 8 2
1 M M
Analysis diagram a b c d e f g h

The Euwe - Fine game continued 1 1 . . . Ae7 1 2 . Jafd1 e5

1 2 . gfe1 exd4 1 3. &Dxd4 0-0 1 4. Ac2 tDf6
(nowad ays the more forceful 1 4 . . . tDc5
1 5 . '1Wc3 g d 8 i s p referred , with equal
play) 1 5 . gad 1 g6 1 6 . tDxf6+? ! (1 6 . '1Wc3!?)
1 6 . . . A x f 6 1 7. ge4 We? 1 8. '1We3 *g7
1 9 . h4 h 5 20. gf4 ge8! 2 1 . '1Wxe8 'Wxf4
2 2 . tDf3 g b a 23. '1We4 '1Wxe4 24. Axe4
Ag4 , and Bl ack won the endgame in ex
cellent style.
The modern sharp continuation 1 1 . . .
exd4 underwent a severe testing in the
game Mamedyarov -Topalov (Sofia 2007),
where Shakhriyar made use of a recom
mendation by Jan Tim man, that had not
come to the attention of the ex-world
champion - 1 2 . &Dxd4! , after which cas-
tling is not possible - 12 . . . Ae7 1 3. tDf5 ! .
T h e game continued 1 2 . . .tD c 5 1 3. &Dxc5
Axc5 1 4 . gae1 + *f8 1 5. ge4 '1Wf6 1 6. '1Wc3
a5?! (1 6 . . . Ad7!? 1 7. b4 Axd4 1 8. gxd4 gea
- Stohl) 1 7. gfe1 Ad7 1 8.g1 e3 Ab4 1 9. Wc2
'1Wd6 20. gf3 h5 2 1 . c5! Axc5 22. Ac4 f6
23. tDe6+ Axe6 24. gxe6, and White soon
won .
8. Wxc4 d7 9. e 3 g6 1 0. Ae2 Ag7
1 1 . 0-0 0-0
(see next diagram)
Carlsen t hinks t h at t h e p reparatory
12 . . . '\We?!? is more accurate.
1 3. e4
Black is seriously behind in the develop
ment of his queenside, and White's play
involves exploiting the weakness of the
d6-square for the invasion of his knight.
1 3 . . .'We7?!
Here this move does not oppose White's
plan , since it al lows a favourable break
through in the centre. As Carlsen shows in
his comments on the game, the problem
would have been solved by the paradoxical
move 1 3 . . . Wf5 ! . This seems strange - the
queen invites the knight to go to d6 with
gain of tem po, but the subtle point is that
Black forces the exchange on e5 and neu
tralises the danger: 1 4 . tDd6 Wf6 1 5 . d xe5
tDxe5 1 6. tDxe5 Wxe5 1 7. 'Wb4 Ae6 , and
1 8 . Wxb7?! is risky because of 18 . . . Ad5.
14. d5 cxd5 1 5. 'Wxd5 b6
The alternative 1 5 . . . tDf6 1 6. Wd 6 Wxd6
1 7. &Dxd6 would also have left White with
a favourable endgame.

"1 26
 M. Carlsen - A. Onischuk

a b c d e f g h 1 9. xb7
8 .i ..t. .i e 8 White m u st accept the pawn sacrifice,
7 .f. .t. 'ii .t. ..t. 7 since after 1 9 . ac1 e4 20. h xg4 exf3
6 '.& ... ... 6 2 1 . Axf3 A x b2 22. b1 xeS 23. x b2
5 ... 5 c7 he can not hope for anything.
4 tiJ 4 1 9 . . . Ae6
3 8 tb 3 The pair of long-range bishops supported
2 88 JL 8 8 8 2 by the rook put the queenside under con
1 1:, 1::. certed fire, forcing White onto the defen
a b c d e f g h sive.

16. Wc5!
20. b3

White can not get by without this move,

One can not help but be amazed at the
20. d2 is too passive: 20 . . . c7 21 . &baS
pragmatism of the you ng grand master. If
(2 1 . tbd6 d8) 21 . . . tea 22. ad 1 c2,
the position demands it, Carlsen is pre
while after a move by the knight - 20. tDd6
pared to try and squeeze out a minimal
the potential of the black pieces is fu lly
advantage, even in an obviously protracted
revealed : 20 . . . c2 21 . d2 xd2 22. tbxd2
endgame.
d8 23.tD2c4 e4 24. b1 fS , and again
16 . Wxc5 1 7. xeS Ag4
. . 2S. b3 cannot be avoided .

White's control of the d-file combined with 20 . . . 9:c2 21 . Af1

the pressu re of his knight on the b7-pawn
creates problems for Black over the devel
a b c d e f g h

opment of his light-square bishop. Onis 8 . . 8

chuk's decision to sacrifice the pawn for 7 ... tiJ '-1-
.t. ..t. 7
the sake of disclosing the potential of his 6 '.& ..t. ... ... 6
bishop pair seems a fu lly objective one. 5 ... 5
If he defends the pawn by 17 . . . b8, after 4 4
18. ac1 the problems of developing the
bishop are not relieved : if 18 . . . AfS there is
1 9. tbh4, while if 18 . . . Ag4 - 1 9. h3.

18. h3 a b c d e f g h

For the moment 1 8. tbxb7 does not work 21 . . . d5

because of 18 . . . e4.
Bl ack is at the c ross-roads: should he
18 . . . 9:ac8
continue his active cou rse, which Onis
Black has gained an im portant tem po to chuk settles for, by playing his knight to
secu re an excellent post for his bishop c3, or limit himself to waiting tactics? For
at e6. example, 21 . . . AdS 22. tDd6 c6 23. ebbS

1 27i
 Chapter 2 The Way to the Top

a6 is possible, d riving the knight to the 26. tileS Axc4+?

edge of the board (24. tDa3). But what to
do next? 24 . . . e4 is only a half-measure:
25. tDd4 1k5 26. f!ac1 and White neverthe
less transfers his knight to c4 - 26 . . . f!fc8
(or 26 . . . f!a5 27. tDc4 Axc4 28. bxc4 f!xa2
29. c5 tDc8 30 . f!a1 , win ning the a6- pawn)
27. f!xc5 f!xc5 28. tDc4! , and Black is u n
able to regain t h e pawn. If 2 4 . . . a 5 White
rep lies 2 5 . tD d 2 e4 (25 . . . a4 2 6 . Ab5)
26. f!ab1 f!a8 27. tDac4, stabilising the
a3-knight.
It wou ld appear that Black no longer had
any choice, other than to win the exchange:
26 . . . tDa4 27. tbxe6 Axa1 28. bxa4 fxe6
29. Axe6+ g7 30. tDxe4, where White
wou ld be faced with converting his com
pact pawn mass in a protracted endgame.
But, at the least, this was a p ractical
c h ance, whereas with the exch ange of
the light-square bishops the potential of
the extra pawn increases sharply.

22. Ac4 tilc3 23. l:id2 l:ixd2 24. tilxd2 27. t'il xc4 l:ib5
e4 25. f1
Ap parently O nisc h u k was pin ning his
a b c d e f 9 h
hopes on this move, thinking that after the

8 .I. e 8 knight moves to a6 it wil l be out of play. Af

ter 27 . . . tDa4 White can interpose 28. tDa6 .
7 .t. CD '--'r .t. .i. 7
6 .i. .t. .t. 6 28. t'ila6!
5 ] 5 This knight is destined to discredit the
4 .t. 1 4 rule of Siegbert Tarrasc h , that a knight
3 I.& 3 on the edge of the board always stands
2 CD 2 bad ly. In the given specific position the
white knights are excellently coordinated ,
::
a b c d e f 9 h
restricting the rook, which looks longingly
at the c6-square.
25 . . . 1:ib8 28 . . . Af8

Alexey Korotylev considers the best plan a b c d e

to be the switching of the light-sq u are 8 8
bishop to the long diagonal, from where
7 .t. 7
it defends the e4-pawn and keeps an eye
on the queen side pawns: 25 . . . Ad7!? 26. a4
6 CD
Ac6 27. tDd6 Ae5 28. f!c1 Axd6 29. f!xc3 5 .I 5
Ab4 30 . f!c2 f!d8, and the a4/b3 pawn 4 CD .t. 4
pair is sec u rely blockaded . It shou ld be 3 I.& 3
added that the attempt to advance the b 2 2
pawn - 26. tDc5 Ac6 27. b4 - is parried by 1 ::
27. . . f!d8 28. e1 a5! 29. bxa5 tDa4. Now, a b c d e f 9 h
however, Carlsen confidently finds the key
to Black's defences. 29. b4!

1 28
 M. Carlsen - T. Radjabov

And here is the key to the position. By sac Game 27

rificing a pawn, White completely ties down M. Carlsen - T. Radjabov
the black pieces. Biel 2007
Irregular Opening [ 807]
29 . . . .txb4 30. a4 lib7 31 . lic1 ltid5
32. lib1 f5 1 . e4 d6

Against the attempt to unravel with a 're
verse' pin , White has a choice between
two winning endgames: 32 . . J !e7 33. xb4
.b7 34. xd5 .xb1 + 35 . e2 and 34. d6
.xb4 35 . .xb4 xb4 36 . xe4.
33. a5!
a b c d e f g h specter of reputations, and he is one of
8 8
Despite the fact that in the last round Carls
en was more high ly motivated (to overtake
Radjabov, he had to win the game), his
opponent was by no means in a peace
able mood . This is also indicated by the
choice of opening. It should be mentioned
that Radjabov (as also Carlsen) is no re
the few who plays for a win with both col

7 .1 7 ours. It is worth remem bering Teimour's

win , at the age of 1 6, over the great Garry
s t[j Kasparov in Linares 2003. With the black
5 1.& 5 pieces, moreover! It is sufficient to say
4 .i. tZJ 4 that Teimour's main weapon against 1 . d4
3 is the King's Indian Defence, in which he
2 2 has gained numerous wins.

a b c d e f g h
The domination of the white knights over
the black rook is impressive! By taking
away the b6-square, White puts Black in
zugzwang. Carlsen confirms this with the
fol lowing variations: 33 . . . .e7 34. xb4
.b7 35 . b6 ! , or 33 . . . .b5 34. g3! with
the th reat of 35. d6 (after the immediate
34. d6 .xa5 35. xb4 xb4 36 . .xb4
.a1 + White has to give up his f2-pawn).
The performance is finished . The curtain
can be lowered .
33 . . . g7 34. ltie5 f4 35. exf4 e3
36. fxe3 lti xe3+ 37. g1 ltid5 38. h1
lib5 39. ltic6
2. d4 ltif6 3. ltic3 e5
A curious psychological moment. It turns
out that Radjabov was only 'scaring' Carls
en with the Pirc- Ufimtsev Defence; he does
not intend to fianchetto his bishop, and
he invites a transposition into the Philidor
Defence. He is not afraid of the exchange of
queens, realising that in view of the tourna
ment situation Carlsen will not go in for it .
4. ltige2
Magnus prefers something different, which
has no name in opening classification . The
transposition into the Philidor Defence
could have occurred after 4. f3 bd7.
4 . . . ltibd7

Black lost on time. Carlsen thinks that the development of

the bishops was more in the spirit of the

129
 Chapter 2 The Way to the Top

position - 4 . . . Ae7 S. g3 Ag4 6. Ag2 exd4 Haifa 1 989) Beliavsky preferred to restrict
7. xd4 tilc6. Black on the queenside in 'Spanish' style
5. g3 c6 6. Ag2 - 1 1 . b4. The game continued 1 1 . . . Ab7
1 2 . f!b1 axb4 1 3. axb4 tileS 1 4 . tilg3 g6
It was possible to forestall the pawn initia
1 S. Ah6 tilg7 1 6. tilce2 tilb6 1 7. f4 exf4
tive on the queenside by 6. a4.
1 S. tilxf4 AgS?! (1 S . . . tilc4! was more logi
6 . . . b5 7. a3 Ae7 8. 0-0 0-0 9. h3 aS cal) 1 9. AxgS WxgS 20. eS d xeS 21 . tile4
In the rapid game G risc h u k - Frid man e7 22. d xeS xeS 23. d6 ! and White
(Mainz 2006) Black included his pawn in developed a dangerous initiative.
the attack only after developing his bishop: 1 1 . . . b4?!
9 . . . Ab7 1 0 . g4 aS, but after 1 1 . tilg3 exd4 One can only try and guess why Radjabov
1 2 . xd4 cS 1 3. d1 b4 1 4 . tilbS!? Aa6 did not go in for the favourable opening
1 S. a4 it transpired that the d6-pawn re of the position : 1 1 . . . exd4 1 2 . xd4 b4
quired d efe nding. The game continued 1 3. tilce2 bxa3 1 4 . .xa3 (1 4. bxa3? tilxg4!)
1S . . . g6?! (1 S . . . tileS) 1 6 . Ah6 f!eS 1 7. gS 14 . . . dS, where the black pieces would have
tilhS 1 S. tilxhS gxhS 1 9. xhS 'i'b6 20. f4 become very active. But subsequently too
c4+ 21 . h1 f!acS 22. fS tileS 23. f6 AdS his play raises questions. Apparently it was
24. AfS 1 -0. one of those days for Radjabov.
1 0. g4 1 2 . ce2 bxa3
It is natural to try and take the lead in the By beginning with 12 . . . dS 1 3. exdS tilxdS,
flank attacks , although in his com ments Black would not have allowed the activa
on the game Carlsen considered this to be tion of the queen's rook and after 1 4 . f!e1
not the optimal decision, without, however, he would have achieved an equal game:
suggesting anything instead . 14 . . . bxa3 1 S. bxa3 Af6.
1 3. r:lxa3 d5
1 0 . . . Aa6
a b c d e f g h
a b c d e f g h 8 .1 8
8 .1 8 'i .1 9
'iV .. . 7 l.& .l. .l .l .l 7
7 l.& .l. .l .l .l 7 6 .1. .l 6
6 .1, .l .l 1.& 6 - 1.&
5 .l .l 5
5 .l 5 4 4
4 4
3 t2J 3 2
2 tD .il. 2
1 J::l .il. 'lW g 1
a b c d e f g h
a b c d e f g h
14. r:le3!
1 1 . g3 This shows a deep understanding of the
A provocative move, since it allows the position! The rook manoeuvre to the centre,
opening of the position. In his game with involving the doubling of rooks on the e-file,
Mokry (European Team Cham pionship, forces Black to release the pawn tension.

130
 M. Carlsen - T. Radjabov

This leads to the opening of the position , CDf8 22. Ab2 CDg6 23. CDxg7! CL'Idf4 24. CL'Idf5
making it easier for White to attack o n the CL'Ixh5 25. CDxh5 with a decisive attack.
kingside. 1 5 . . . h5, a move considered by Carlsen ,
1 4 . dxe4
. . also looks irrational , since it contradicts a
Black is forced to concede the centre. classical rule: don't move pawns on the
He vacates the d5-square for his knight, wing where you are weaker. After 1 6. gxh5!
hoping to d rive away the rook, but dis (1 6. g5 CL'Id5 17. ..xe4 Axg5 1 8. d xe5 CL'Ic5
illusionment awaits him . The alternative would suit Black) 1 6 . . . CL'Id5 17 . .. xe4 f5
1 4 . . . exd4 1 5 . CL'I x d 4 Ac5 (1 5 . . . A x f 1 ? 18 . ..xe5 CDxe5 1 9. d xe5 White has more
16. CDxc6 \WeB 1 7. exd5 1eads to the loss of than sufficient compensation for the ex
a piece) 1 6 . ..fe1 '!Wb6 (weaker is 1 6 . . . Axd4 change (Axd5 and CL'If4 is th reatened).
17. \Wxd4 c5 1 8. \Wa4 d4 19 . ..a3 with an Possibly the best solution to the situa
attack on the weak pawns) 1 7. c3 had its tion could have been 1 5 . . J e8.
drawbacks, but at least Black would have 16. f5 .td8?
activated his bishops. The bishop retreat conclusively disru pts
1 5. e1! the already poor coordination of the black
The rook moves out of the pin and the il pieces , dividing them into two camps.
lusions associated with 1 5 . . . CL'Id5 prove 16 . . . ..fe8 answers better the demands of
unfounded in view of 1 6 . ..xe4. Unexpect the defence.
edly it transpires that the black pieces are 1 7. g5 d5 1 8 . xe4 f6 1 9. eg3
mobilised far worse for the forthcoming play. a b c d e f g h
a b c d e f g h a !.
..t .. . a
a !. 'if .. . a 7
'il l.&\ 1. 1. 7
7
1.1\ ..tl.l.l. 7 6 j_ I. I. 6
6 j_ Jl. 1.1\ 6 5 I. 1.1\ 1. /:b 5
5 I. J I. 5 4 : 4
4 4
3 : l:b 3
2 l:b 2
1
iV : 1 a b c d e f g h

a b c d e f g h 1 9 . . . g6
It is al ready hard to offer Black any good
1 5 . . . Wc7
advice, since the white pieces are arriving
Not an easy choice. The inclusion of the
on the kingside as though by schedule. To
exchange 15 . . . exd4 1 6. CDxd4 V#c7 sharply
illustrate this we give two possible varia
activates the opponent's pieces: 1 7. g5
tions: 1 9 . . . ..f7 20. \Wg4 CDe7 2 1 . b3 CDg6
CDd5 18 . ..xe4 Ac5 1 9. CL'Igf5 '!Wb6 (if 1 9 . . .
22. Aa3 or 19 . . .CDe7 20. gxf6 CDxf5 21 . CDxf5
a4 there follows 20. \Wh5 ..fd8 21 . c4! Axc4
Axf6 22. dxe5 tDxe5 23. Af4. In both cases
22. CL'Ixg7!) 20. \W h 5 ..fd8 (or 20 . . . Axd4
White's advantage is obvious.
21 . CDxd4 g6 22. \Wh6 ..fd8 23. b3! with the
decisive threats of Ab2 and ..h4) 2 1 . b3 20. h6+ g7

1 31
 Chapter 2 The Way to the Top

If 20 . . .'> h8 there could h ave fol lowed 22. e6!

2 1 . gxf6 (2 1 . d xe5 fxe5 22. tbg4 is also
The decisive breakth rough . White clears
good) 21 . . . Axf6 22. tbg4 Ag? 23. tbxe5
the approaches to the black king.
tbxe5 24. d xe5 with an extra pawn.
22 . . . xh6
21 . dxe5 fxg5
If 2 2 . . .tD7f6 t here a l so fol lows 23. e7!
a b c d e f g h Axe? 24. l::! xe7+ tbxe7 25. l::! xe7+ VJ/xe7
a .I
..t .1. 8 26. tbhf5+.
7
'iV,.. .t 7 23. e7! 'Wb6
.t .t 6
,.. .t 5 The capture of the pawn leads to a mating
4 l:r 4 attack: 23 . . .tbxe7 24. l::! xe7 Axe? 25. l::! xe7
!!adS 26. 'Wd4 l::!f 6 27. 'Wh4#.
3 3
2 2 24. exf8'W+ tilxf8 25. c4 tilf4 26. 'Wd6
g7 27. xf4 gxf4 28. S:e7+
l:r
a b c d e f g h Black resigned .

An im portant l a n d m ark in Carlsen's self-assertion was his participation in the

second M ikhail Tal Memorial Tou rnament (FI D E category 20) . On this occasion
Carlsen arrived in M oscow not as a timid p u pi l , like a year earlier, but as a fighter,
tested in battle with elite gran d m asters. However, he did not manage to take part
in the battle for the first p rize. Kram nik was untouc hable. After the 1 4th world
cham pion had passed on his ' ruling aut horit y ' to Anan d , h e p l ayed freel y and
over the short distance of 9 rounds he scored +4, finishing 1% points ahead of
the seco n d - p l aced Shirov. But Carlsen's sh are of 3rd-5th p l aces with Peter Leko
and D mitry J akovenko also showed that he had effectivel y established himself in
the upper echelo n .
Playing in Moscow against a farge group o f successors o f t h e famous Soviet
Chess Schoof is always a serious test . Especially when you begin the tournament
with a game against Kram nik, and the fact that in Moscow the 1 4th world cham pion
always pl ays with particular com mitment and is always aiming to win . The intrigue
was increased by the fact that, in their previous meeting in the sum mer of that year
in Dortmun d , Carlsen had suffered a disaster in the Catalan Opening. The opening
was predictable.

13
 V. Kramnik - M. Carlsen
-------

Game 28 ready in a positional trap: 17. . . c6 1 8. tDd3 or

V. Kramnik - M . Carlsen 17 . . . axb4 1 8. tDc6 V!/d7 1 9. axb4 is equally
Moscow 2007 unpleasant) 1 7. tDb3! . The culmination of
Catalan Opening [ EOS} White's strategy - he achieves a complete
blockade of the queenside. After 17 . . . axb4
1. f3 f6 2. c4
1 8. tDa5 Aa8 1 9. tDac6 Axc6 20. tbxc6
Kramnik's favourite way of reaching the V!/d7 2 1 . Axd5! exd5 22. axb4 Black was
Catalan . com pletely paralysed and the b5- pawn
was doomed .
2. . . e6 3. g3 d5 4. d4 e7 5. g2 0-0
6. 0-0 d xc4 7. 'Wc2 a6 8. 'Wxc4 b5 1 1 . g5 bd7 1 2 . bd2 Jac8 1 3. b3
9. 'Wc2 J.b7 1 0. d2 c5

White is intending to place his bishop on Thus Black has carried out his plan, but
a5, to prevent the freeing advance . . . c7-c5. has associated it with a new idea.
a b c d e f g h 14. J.xf6 xf6!?
s ,i 'iV .i fl a
This is Carlsen's genuine novelty. He sacri
7 .i. .i. 7 fices a pawn. In the only previous game on
s 6
this theme, Hofman - Browne (Bad Wies
5 5 see 1 998), Black played 14 . . . V!/xf6 1 5. tDe5
4 4 cxd4 1 6. tDxd7, after which White did not
3
[jj 3 tempt fate and he offered a d raw.
2 V/H 2
1 l: D l:
a b c d e f g h

10. Ad6

Carlsen's home p reparatio n . Black va

cates the e7-square for his queen , when
the Aa5 thrust becomes irrelevant. Up til l
here t h e two players h a d move for move
repeated their game in Dortmund, which
continued 1 0 . . .tDc6 1 1 . e3 tDb4 1 2. Axb4
Axb4 1 3. a3 Ae7 14. tDbd2 lk8 1 5. b4 a5
1 6. tDe5! (apparently this move came as a
surprise to Carlsen, who was being guided
by the game Bruzon - Capo Vidal, Morelia
2007, which went 1 6. V!/b2 axb4 1 7. axb4
'Wd6 1 8. flfb1 flaB with good play for Black)
1 6 . . .tDd5? (Black should have exchanged
bishops - 1 6 . . . Axg2 1 7. xg2, but he is al-
Analysis diagram
What was it that Carlsen did not like here?
Sergey Shipov suggests the fol lowing vari
ation: 1 6 . . . V!/d8 1 7. tDbc5 Axg2 1 8. tDxf8
Axf1 1 9 . 'Wx h7+ x f8 20. tD b7 V!/d7
2 1 . tDxd6 flc6 22. tbxf7 xf7 23. flxf1 ,
and White win s a pawn. But 2 1 . . . V!/xd6
22. 'Wh8+ e7 23. 'Wxc8 Axe2 must also
be considered. White is the exchange up,
but the d4-pawn, supported by the bishop,
is a real force. In addition the white king
is a concern . Let us consider a possible

1 33
 Chapter 2 The Way to the Top

continuation: 24. l:;!c1 f6! (out of harm's
way!) 25. l:;!c7 (25 . Wfc7 Wle5!?) 25 . . . Wib4
26. g2 Wle1 27. l:;!c1 Wid2 28. l:;!c5 Ad3
29. Wlb7 g6 and Black is alright. But these
and other variations remained off-screen ,
since the two players continued playing at
high speed , demonstrating confidence in
their home preparation.
As Pushkin said: 'One dawn to replace an
other hastens'. Now White has to watch out
for the possible manoeuvre . ..e4- g5-f3.
22. Wxe5 laxc5 23. 'Wd4
If 23. Wif4 possible is 23 . . . g5 24. h1
h6 fol lowed by the invasion of the rook on
the 2nd rank.

1 5. xc5 a b c d e
If 1 5 . dxc5, then 1 5 . ..e4! is good .

15 . . J.e4 1 6. Wc3 e5!

For the sacrificed pawn Black has developed

great activity. For example, bad for White
is 1 7. xeS Axg2 18. xg2 Axe5 1 9. dxe5
e4 when he has to give up his queen.

1 7. e3 J.xc5 1 8. dxc5 Wd5

a b c d e f g h
It was only now that Kram nik stopped to
think. And he had plenty to ponder. Black 23 .. 1ac4!
has successfully solved his opening prob
lems. A multi-purpose move. The rook d rives
the queen off its central position and frees
a b c d e f g h
the knight for the manoeuvre to f3. I n the
8 I 8 event of 23...l:;!c2 24. f4 (after 24. e1
7
I .\ .\ .\ 7 f!e2 2 5 . f3 h 6 2 6 . a3 f!c8 the b l ack
6 .\ 6 pieces occupy their optimally active po
5 .\ 8 'if .\ -- 5 sitions) 24 . . .g5 25. 'l!l'ld5 f3+ (weaker
4 .t 4 is 25 ...Wixd5 26. f!xd5 f3+ 27. g2 d2
28. l:;!e1 with the threat of the f!e2 pin , while
3 8 CiJ 8 3 after 28 . . . f! x b2 29. d3! f! b1 30 . f!xb1
2 88 88 2 xb1 31 . b4 White again obtains an extra
1 1:,
.: w 1
pawn) 26. g2 Black has either to withdraw
a b c d e f g h
his knight or work out the consequences
of the mutual pin after 26 . . . d2 27. Wxa8
1 9. laad1 WaS f!xa8 28. f!fe1 l:;!d8.

1 9 . . . '1Wb7 was more active, but Carlsen 24. 'Wd3

only took into account the nuances known
to him with the defence of the back rank.
20. h4 J.xg2 21 . xg2 e4
After 24. Wle5 , apart from the t rivial
24 . . . f!c5, there are also other possibilities.
Black undoubted ly has full compensation

1 34
 D. Jakovenko - M. Carlsen

for the pawn, although it is sufficient only way to save the game is by a cou nter
to maintain the balance, not more. sacrifice of a rook.

24 .. .tig5 25. b3 c5 26. h4 a b c d e

8 8
To avoid standing for ever u nder the threat
of the check on f3.
7
'iV
6
26 . . . f3+ 27. h1 5 :t: .l 5
a b c d e 4 fj, 'iJJ fj, 4
8 'ii' 8
3 fj, 1.1\ B 3
7 2 fj, fj, 2
6 .l 6 :t: W 1
5 .l l. 5 a b c d e f g h

4 32. 'We6+ h8 33. 'Wxf5!

3 fj,
2 fj, Exploiting the absence of an escape square
for the king, White exchanges the q ueens.
a b c d e f g h 33 . . .'Wxf5 34. xf5 xf5 35. d1 g5

27. . .'Wc8 Black cannot get away with half-measu res

How can one refrain from threatening mate
on h3? But Black's position is so good , that
he does not need to h u rry. For example,
he cou ld first have developed his rook -
27. . . gea , opened an escape sq uare for
his king with . . . h7-h6, and then taken a
decision.
28. f4 'Wg4!
Otherwise with <;!;>g2 White will d rive away
the knight.
- 35 . . . g6 36. <;!;>g2 .
36. hxg5 xg5 37. fxg5 xf2 38. a4
bxa4 39. bxa4 a2 40. d4
Draw.
The tem po play in the final stage of the
game is i m p ressive. The two p l ayers
proved worthy of each other. Even so, there
was a small 'pl us' for Carlsen, even if only
a moral one - in his opening preparation
he su rpassed none other than Kram nik.

29. 'Wd7 f5!

Magnus's m otto is 'only forward ' ! The
exchange sacrifice on f4 is now on the
agenda. Kram nik has to find a way to save
himself.
30. d5 xf4 31 . exf4 f5
Again mate on h3 is th reatened . The only
Game 29
D. Jakovenko - M. Carlsen
Moscow 2007
Nimzo-lndian Defence [ E20]
1 . d4 f6 2. c4 e6 3. c3 Ab4 4. f3
White goes in for the Romanishin Variation
with the fianchetto of the bishop. The main

1 35
 Chapter 2 The Way to the Top

continuations are 4. e3 and 4. 'Wc2. 18. Af4

4 c5 5. g3 cxd4 6. xd4 0-0 7. J.g2

a b c d e
.1
.

8 8
d5 8. cxd5 xd5 9. 'Wb3
7 .l
a b c d e f g h
6 l.& .l 6
. . 8
5 'if 5
.l .l .l ? 4 .i..
6 .l 6
3
5 1.& 5 2
4 .i.. tiJ 4
1
3 ti} 3 a b c d e f g h
2
1 i, This battle took place at the peak of the
a b c d e f g h variation's development, where much was
decided by nuances, since this position was
A universal move in this type of position. studied at the highest level. And although it
White cou nters the pin on the knight and is considered favourable for White, who has
attacks the d5-point. the two bishops, Black's resources are also
quite considerable: his light-square bishop
9 . . . Wa5
can be included in the fight for the d-file and
9 . . .'Wb6 is weaker in view of 1 0 . Axd5 exd5 in the play on the kingside. True, there are
1 1 . Ae3. 9 . . . tDc6 1 0. tDxc6 bxc6 1 1 . 0-0 also drawbacks: Black's queenside pawns
Was 1 2. Ad2 leads to a transposition of are in need of defence.
moves. White began developing his bishop at
f4 after a detailed investigation had been
1 0. Ad2 c6 1 1 . xc6 bxc6 1 2 . 0-0
made of 1 8 . Ae3, putting the queenside
Axc3 1 3. bxc3
under attack. After Black carried out his
After the exchange of the bishop 1 3. Axc3 main idea of exploiting the weakness of the
tDxc3 1 4. bxc3 Aa6 Black easily equalises. light sq uares: 1 8 . . . Ae2 1 9 . e1 (1 9. d2
For example, the game Moiseenko - Zugic ab8, Kasparov - Anand, Wij k aan Zee
(Toronto 2003) continued 1 5. 'Wb4 Wxb4 2000) 1 9 . . . tDc4 20. Wc5 fd8 2 1 . e5?!
1 6. cxb4 ab8 1 7. a3 Axe2 1 8. fc1 fc8 (21 . h3!?) 21 . . .tDb2! 22. Wxc6 tDd3, where
1 9. Axc6 f8 20. f4 e7 21 . f2 Aa6 %-%. he achieved the better game (Bacrot
G rischuk, Cap d 'Agde 2003), interest in
13 Aa6 14. Elfd1 'Wc5
the move 1 8. Ae3 fel l away.

Preventing c3-c4, which is possible after

18 c5
1 4 . . . Axe2.

Black is at the parting of the ways: should

15. e4 Ac4 1 6. 'Wa4 b6 1 7. Wb4 Wh5
he d rive back the queen or the bishop?
Nothing is given by 17 . . . 'Wxb4 1 8 . c x b4 Carlsen begins by driving back the queen,
fd8. but in so doing he takes into account the

1 36
 D. Jakovenko - M. Carlsen
experience of the game G rischuk - Gel
fand (World Championsh ip, Mexico 2007),
where B l ack included 1 8 . . . Ae2 1 9 . l::! e 1
and wanted both the one and the other
- 1 9 . . . cS 20. Wb3 eS, expecting 2 1 . Ae3
tDc4. But there fol l owed 2 1 . Ac1 ! , when
Black faced an unpleasant white initiative
on the q ueenside: 21 . . . l::! a c8 22. a4 Ac4
23. Wb2 Wg6 24. aS tDd7 2S. l::! d 1 Wc6
26. Ah3 Ae6 27. Axe6 Wxe6 28. Ae3 c7
29. Wa2 ! (the exchange of queens guar
antees W h ite control of both the open
fi les, d and b) 29 . . . Wxa2 30. xa2 tDf6
31 . f3 , and Wh ite obtained clearly the bet
ter endgame.
The alternative 1 8 . . . eS a l l ows W h ite
to attack the c6- pawn. For example, the
game Najer- J akovenko (Moscow 2006)
conti n u ed 1 9 . Ae3 Ae2 20. l::! e 1 tDc4
21 . Wcs l::! fd8 22. l::! a b1 h6 23. h3! tDxe3
24 . Wxe3 Ac4 2S. a4 aS 26. WcS Ae6
27. h4 l::! d 2 ! , where for the pawn B l ack
gained good counterplay. But it is q u ite
poss i b l e t h at White's p l ay can be i m
proved .
1 9. Wb2
A month l ater at the 2007 World Cup i n
Khanty- Mansiysk, Cheparinov in his game
with Carlsen made an attem pt to improve
White's play with 1 9. Was. The game con
tinued 19 . . . eS 20. Ae3 Ae2 21 . l::! e1 tDc4
22. Wa6 tDxe3 23. xe2 tDxg2 24. xg2
and, despite the sym metry of the pawn
weaknesses on the queenside, the black
pawns proved more vulnerable.
19 . . . ad8
A more natural continuation than 19 . . . eS?!,
which makes the defence of the cS-pawn
more d iffi c u lt . For exa m p l e , the g a m e
Bacrot - G rischuk (Poikovsky 200S) con-
1 37
tinued 20. Ae3 l::!fc8 2 1 . f3 Wg6 22. a4 aS
23. Ah3 Ae6? (after 23 . . . !:!cbS 24. Wa3
Ae2! Black wou l d have had counterplay)
24. Wxb6 Axh3 2S. Wxg6 fxg6 (after 2S . . .
hxg6 26. g4 fS 27. dS fxg4 28. fxg4 Axg4
29. AxeS all the black pawns are weak)
26. f2 , and White achieved clearly the
better endgame.
20. e1
A d ubious decision to concede the d -fi le,
not supported subseq uently by any idea.
20. f3 came into consideration , but block
ing the diagonal of his bishop apparently
did not come into White's plans.
20 . . . d7
Carlsen's ability to place his pieces o n
their best squares characterises h i m as
an ad herent of the classical style of play.
Black prepares the doubling of his rooks,
at the same time defending the a7-pawn,
and leaves open the possibility of includ
ing the second rook in the defence of the
cS-pawn in the event of Wa3 .
21 . h3
White has t o reckon with t h e threat o f a
rook invasion of the second rank, and in
the event of this he prepares g3-g4.
21 . . . h6
Carlsen also makes a useful waiting move,
leaving open the possibility of . . . l::! c 8 in the
event of Wa3.
22. a4
(see next diagram)
A pointless advance in the given situation ,
since an attack on t h e queenside i s not a
reality.
 Chapter 2 The Way to the Top

a b c d e was intending to answer with 25 . . . g5! . Now

8 8 if 26. Ac1 ? there follows a rook invasion
- 26 . . J !d 1 27. Aa3 xe1 + 28. xe1 d1
7 ... 7
29. xd1 'Wxd 1 + 30. h 2 'Wa4 and the
6 queenside comes u nder attack by Black,
5 ... and after 31 . Af1 Ab7! 32. Axc4 'Wd 1 the
4 .i. kingside also collapses. White has to throw
3 caution to the winds: 26. g4 'Wg6 27. Ae4
2 Wi 'Wg7 28. Ag3 tbd2, but the initiative remains
1 : with Black.
a b c d e f g h
25 . . . e5 26. g4

22 . . . .ta6! W h ite loses after 26. Ac1 ? d1 27. g4

xe1 ! 28. gx h 5 d d 1 , also pick i n g u p
A s u btle move. The bishop vacates an
the bishop.
i mportant square for the knight, without
blocking the d -file, which cou ld have oc 26 . . . Wg6 27. Axc4 exf4!
cu rred after 22 . . . Ad3.
Transposing into a heavy piece ending by
23. Wa2 27. . . Axc4 28. Wxc4 exf4 29. Wxc5 d3
would have left White with cou nterplay -
This prophylactic queen move in the event
30 'Wf5 'Wd6 31 e5 'Wd5 32. e6.
of . . .tbc4 i n d icates that J akovenko has
reconciled himself to waiting tactics. It was 28. AdS f3!
more logical to return the rook to the d -file
Fixing the kingside in antici pation of . . .
by playing 23. g4 'Wh4 24. ed 1 .
h6-h5.
23 . . . fd8 24. a5 c4
a b c d e f g h

a b c d e 8 K 8

8 8 7 ... K ... ... 7

7 ... 7 6 .i_ . 6

6 .i_ 5 .l i, 5

5 4 4

4 3 ... 3
3 2 'iY 2
2 Wff 1 : .=.
1 : a b c d e f g h

a b c d e f g h
29. c4?
25. At1
It is na"lve to hope to shelter behind the
White could have parried the threat of 25. . . bishop, since Carlsen has already found
e5 with the rad ical 25. e5 , which Carlsen the key to the enemy king. Wh ite should

138
 D. Jakovenko - M. Carlsen

have given up the exchange: 29. l::!. e3 Ae2! h2, White has to accept the loss of the
(29 . . . Wf6 30. c4 h5 31 . e5 Wh4 32. l::!. x f3 exchange - 33. Wxc4 Wxa1 .
leaves White with some hopes) 30. l::!. x e2
32. Iiab1?
fxe2 31 . Wxe2 , when his active bishop
gives compensation. Jakovenko sensed the danger and moved his
29 . h5 30. h2 Wf6! rook away, but to the wrong square. 32. ad1 ,
defending the bishop, was essential.
Black moves his q ueen closer to the king,
keeping 'up his sleeve' a delayed-action 32 . . . Axc4!
mine. The g4-pawn wil l not run away, es
That's it! The fortress on d5 col lapses, and
pecially as after 30 . . . hxg4 White wou l d
the exposed king cannot withstand the
have defended with 31 . l::!.g 1 Wf6 32 . l::!. ad 1 !
attack by the heavy pieces.
gxh3 33. Wb3.
33. Wxc4
31 . Iig1?!
If 33. Axc4 Black decides m atters with
The rook on a1 should have been moved
33 . . . l::!. d 2, when 34. Axf7+ fails to 34 . . . f8.
out of range of the queen - 31 . l::!. a d1 , and
after 31 . . . \Wf4+ 32. h1 hxg4 White can
33 . . .Wf4+ 34. Iig3 Iixd5!
defend by 33. l::!.g 1 Wh6 34. l::!.g3 (variation
by Shipov). The e4-pawn is pinned , and all that remains
31 . . hxg4 for White is to sacrifice his queen for two
rooks, since after 35. hxg4 l::!. d 1 there is
a b c d e f g h no defence against . . . Wh6+ with a deci
B .i B sive invasion of the queen . White does not
7 .l .i .t .t 7 have sufficient compensation - this is not
6 .i_ 'iV 6 one of those cases where the rooks have
s f::, operational scope .
.t i.. 5

4 ::, ::, .t 4 35. Wxd5 Iixd5 36. exd5 c4

3 .t ::, 3
While the rook is pinned , the pawn swiftly
2 '{i ::, 2
advances towards the queening square .
1 .: .:
a b c d e f g h 37. Iid1 c3 38. d6 c2 39. Iid3 Wc4!

The axe is raised ! If 32 . xg4 Carlsen had One senses that Carlsen is savouring the
prepared 32 . . . Axc4! , and since 33. Axc4 position .
i s not possible on account o f 3 3 . . . d2
40. gr93 Wc6 41 . Iid3 Wc5!
34. Wa4 l::!. x f2+ 35. g1 fd2! with the ir-
resistible threats of . . . Wd4+ and . . . We5- White resigned .

139
..;;;.
c_
hapter 2 The Way to the To p
,_
___________ ---
---------

After the Tal Memorial, how cou ld he not play in the world blitz championship with the
participation of the world cham pion Anand , which was held two days later there in
Moscow, in the prestigious G U M building on Red Sq uare? Carlsen battled selflessly
in the double-round tournament with 20 participants, and finished in an honourable
9th place, with a score of 20% points. Of course, 38 games with the strongest blitz
players in the world was a severe strain , especially as two days later the World Cup
was starting in Khanty- Mansiysk.
Carlsen cou ld rightly consider himself a veteran of the knock-out system, he liked
the fierce struggle in mini- m atches, and he was improving tou rnament by tourna
ment. We should remind you : in Tripoli 2004 Magnus went out in the first round, in
Khanty- Mansiysk 2005 he reached the fourth , but in Khanty- Mansiysk 2007 he played
six stages. 'The knock-out suits the young more than experienced players,' Carlsen
was to say after the end of the tou rnament. ' I n particular because it is easier for us to
recover, and because nothing special is expected of us. Before the World Cup I didn't
have many expectations. I wasn't afraid of anyone, but also I did not consider myself
the favourite. I simply came along and played . It turned out wel l . I hope this wil l also
continue in the future.'
The start was difficult, as fatigue told after the continuous marathon. The respite
in the 'entry' match with the Australian Zhao Zong Yuan (2-0) was purely sym bolic for
Magnus, since his subsequent opponents did not al low him to relax. Much strength
was used on a protracted endgame in the first game with Naiditsch , whom he managed
to overcome only in the tie-break (3-1 ). Things took a difficult cou rse in the match with
Dominguez, who in the first game so zealously tried to spoil M ag nus's birthday, that
on this occasion too he was thinking about the tie-break. But the second game went
wel l (1 %-%), and subsequently his play improved . By the same score 1 %-% he won
his mini-matches against Adams and Cheparinov.

Game 30 The modern experience of the Sicilian De

M. Carlsen - L. Dominguez
World Cup, Khanty- Mansiysk 2007
Sicilian Defence [ 892]
1 . e4 c5 2. f3 d6 3. d4 cxd4 4. xd4
f6 5. c3 a6 6. J.e2 e5 7. b3 J.e7
8 . .tgs
The pin on the knight is a natural reaction
to 6 . . . e5. By the threat of the exchange
on f6 White discloses the weakness of the
d5-point and the d6-pawn .
fence (and earlier - of the King's Indian De
fence in the variation with . . . tDb8-d7 and . . .
e5xd4) has shown that the weakness of the
backward d6-pawn is fu lly compensated
by Black's active piece play in the centre.
Black is not concerned about his d6-
pawn being inadeq uately defended - he
can support it by retreating his bishop to
e7 or cover it with his knight from d4 after
the exchange of the knight on b3.
1 0. 'Wd3

8 . . J.e6 9 . .lxf6 .lxf6

. (see next diagram)

140
 M. Carlsen - L. Dominguez

a b c d e f g h Black makes a seemi ngly useful waiting

'if .!. a move; after 1 4 . . .0-0 he was wrongly afraid

7 al al al 7 of 1 S . h4, for the reason that Wh ite has

al .i. .i. 6
5 al 5
4 8
3 QJ QJ 3 counter has its d rawbacks: 1 7. exfS gxfS
2 8 8 8 8 8 8 2
1 : : 1
a b c d e f g h
1 0 . . . Ag5
An original idea - Black prevents Wh ite's
queenside castling. The classical continua
tion is considered to be 10 . . .c6 1 1 . 0-0-0
Ae7 1 2 . b1 0-0 1 3. dS AgS. A char
acteristic game, for example, is Kramnik
Anand (Wijk aan Zee 2004), which contin
ued 1 4 . h4 Axh4 1 S. g3 Af6 1 6. 'Wf3 AgS
17. 'WhS h6 1 8. f4 Af6 1 9. d2 d4 20. Ac4
c8! 2 1 . c3 bS 22. fS AxdS 23. AxdS
xc3! 24. 'Wg6! (or 24. bxc3 xc3+ 2S. c2
CL!xdS! 26. exdS 'WaS 27. 'Wf3 c8+ 28. b1
c3 with equal play) 24 . . . 'Wb6 2S. xh6!
insufficient resources to conduct an attack
on the kingside. The variation 1 S. 0-0 g6
1 6. Ag4 looks more real , with the idea of
4 advancing the pawn to fS, since the 1 6 . . . fS
1 8. Ah3!? (1 8. Af3 'Wd7 1 9. AdS+ h8
20. CL!e2 is also good) 18 . . . 'Wd7 1 9. f4, and
Black may have problems with the defence
of his fS-pawn.
1 5. o-o o-o 1 6. Ag4!
It is useful to retain this bishop, 'urgi ng' the
rook to go to c6 where, strangely enough ,
it proves to be out of play.
1 6 . . . Eic6 1 7. g3 g6 1 8. f4 Ah6 1 9. Elf2
Ag7
a b c d e f g h
8 'i .. . 8
7 J---.al._ .i. al 7
6 al .!. all I al 1 6

and the players agreed a d raw in view of 5 al 5

perpetual check - 2S . . .a3+ 26. a1 c2+ 4 8 8 8 4
27. b1 a3+.
3 8 Qj 8 3
The plan with control of the dS-point
is also possible - 1 0 . . . 0-0 1 1 . 0-0-0 Ae7
2 8 : 8 2
1 2 . dS AxdS 1 3. 'WxdS 'Wb6 1 4 . hf1
1 : 1
a b c d e f g h
d7 1 S. b1 f6 .

1 1 . El d 1 c6 1 2 . a 3 Somehow imperceptibly it has transpired

that in this unpretentious manner Carlsen
T h e variation w i t h the w i n o f a pawn -
has outplayed his opponent. Despite the
1 2 . 'Wxd6 'Wxd6 1 3 . f! x d 6 e7 1 4 . f!d1
apparently active placing of the knight on
b4 1 S . Ad3 ac8 1 6 . 0-0 f!hd8, where
d4 and the rook on the c-fi le, they are ef
the activity of the black pieces guarantees
fectively out of play. At the same time the
full com pensation - would undoubted ly be
wh ite pieces are ready for an offensive on
a bonus for the opponent.
the ki ngside, and it is not easy for Black to
12 . . . Axb3 13. cxb3 d4 14. b4 c8 create counterplay.

141
 Chapter 2 The Way to the Top
------
20. f5! 'Wg5 21 . Ah3 Ah6
I ntending to exploit the weakening of the
dark squares by 22 . . . 'We3.
22. e1 b5 23. mg2
a b c d e f 9 h
The temptation to d rive in the pawn wedge
23. f6, constructing a kind of trap arou nd
the enemy queen , was very great. But then
the position becomes closed and there is a
protracted battle in prospect. At the same
time, as Carlsen comments, he sensed that
the opponent was experiencing some dis
comfort and that it was possible to await a
more favourable outcome. Taking account
of this, a usefu l waiting move was made.
23 . . . a8?
Magnus's intu ition did not let him down.
Black is unable to find a sensible plan of
defence, and he commits a mistake, lead
ing to disaster. The idea of switching the
rook via a8-a7-c7, in order to double rooks,
looks rather abstract, since it requires too
much time. In anticipation of tDd5 he should
have considered prophylaxis - 23 . . . c;!>h8,
but after 24. tDd5 Wd8 again loom ing up
is the idea of the wedge f6 followed by the
advance of the h-pawn . 23 . . . lk7 24. tDd5
!!a? 25. !!ef1 is also too passive for Black.
142
It would appear that it is already hard to
offer him any good advice.
24. fxg6 hxg6
I n the event of 24 . . .fxg6 the unfortunate
position of the rook at a8 is d isclosed -
25. Ae6+! tDxe6 26. 'Wd5 !!ac8 27. Wxe6+
c;!>h8 28. tDd5 and the concerted actions
of the q ueen and kn ight shou l d bring a
win (variation by Carlsen). For example, if
28 . . . !!f8, then 29. tDf6 is strong, with the
threat of winning the queen by h2-h4.
25. Ad7!
The poor rook! Black has reason to regret
his superficial move 1 4 . . . !!c8. Incidentally,
the proud knight at d4 is also sti l l playing
the role of observer.
25 . . . c4 26. b3 c7 27. d5!
Coord i n ating the actions of kn ight and
bishop. Black can not avoid loss of ma
terial: u n der th reat is not only his rook,
but also his queen in view of the threat of
tDd5-f6+ and h2-h4.
a b c d e f 9 h
a .I 8
7 . 7
6 j_ 6
5 tt:J 'iV 5
4 4
3 3
2 l!t 2
l!t
a b c d e f 9 h
27. . . c1 ?
N ow Black loses his q ueen . H e should
have accepted the loss of the exchange
by retreating 27. . . 'Wd8, although it is clear
 M. Carlsen - M. Adams
that the win for White wou ld be merely a
question of time.
28. h4! xe1
If 28 . . . \Wd8, then 29. Z!xc1 Axc1 30. '1Wf1
is decisive.
29. hxg5 .txg5 30 .tg4 a1 31 . a4
Rook for queen is clearly insufficient com
pensation. Dominguez decided to play on
to the time control.
31 . . .g;,g7 32. c3 a3 33. axb5 axb5
34. xb5 xb5 35. 'Wxb5 .te3 36. f1
a2+ 37. .te2
Not only a retreat, but also the start of an
attack on the f7-pawn .
37. . . a7 38. g;,h3 a8 39 .tc4 h8+
40. g;,g4 h5 41 . xf7+ g;,h6 42. 'Wd7
Black resigned .
Game 31
M. Carlsen - M. Adams
World Cup, Khanty- Mansiysk 2007
Nimzo-/ndian Defence [ E36]
1 . d4 f6 2. c4 e6 3. c3 .tb4 4. Wc2
d5 5. a3 .txc3+ 6. 'Wxc3 dxc4
English solidarity. Fol lowing the prescrip
tion of Nigel Short, Michael Adams con
cedes the centre, in order to develop his
light-square bishop with gain of tem po.
The main contin uations are considered to
be 6 . . . c5 and 6 . . . tile4.
7. Wxc4 b6 8. Af4
Carlsen is ready to engage in a theoretical
discussion, by choosing the most critical
variation with the acceptance of the pawn
sacrifice. I n the 2nd rou n d of the World
Cup, Mikhail G u revich played 8. tilf3 0-0
9. Af4 against Adams, and after 9 . . . Aa6 he
retreated 1 0. 'Wc2 (1 0. '1Wxc7 Wxc7 1 1 . Axe?
Z!c8 1 2 . Af4 tilbd7 wou ld have transposed
into the present game) 1 0 . . .til bd7 1 1 . Z!d1
WeB 1 2 . g3 c5 1 3. Ag2 Ab7 1 4. dxc5 and
offered a d raw, which, natu ral ly, was ac
cepted . How can one reject u p holding the
black pieces without a fight in a mini-match
of two games? In the second game Adams
made successfu l use of the white pieces.
8 .. Aa6
The same idea can be carried out, without
sacrificing a pawn - 8 . . . tild5 9. Ag3 Aa6,
as Onischuk played against Carlsen in a
blitz match for the first prize at the tou rna
ment in Biel. There followed 10. Wa4+ \Wd?
1 1 . Wc2 '1Wb5 12. tilf3 tilc6? and here 1 3. b4!
with the threat of 1 4 . e4 wou ld have won .
Therefore 9 . . . 0-0 1 0 . tilf3 Aa6 is more
accu rate, as in the game Beliavsky- Rozen
talis (Baden 2007). There fol lowed 1 1 . \Wa4
Wd7 1 2 . 'Wc2 c5 1 3. tilg5 tilf6 1 4. Ae5 cxd4
1 5 . Axf6 d3 1 6 . exd3 gxf6 1 7. tile4 Wd4
1 8. Wc3 %-%.
a b c d e f 9 h
'iV . .1. 8
1. 1. 1. 7
6
5 5
4 VIHB 4
3 B 3
2 B BBBB 2
1 .: ttJ .: 1
a b c d e f 9 h
9. 'Wxc7 Wxc7 1 0. Axc7 0-0
Strangely enough, this natural move proved
to be a novelty. Black does not hu rry to
143
 Chapter 2
determine the position of his knight, keep
ing open the option of both . . .CL!d7 and
. . .CL!c6. Before this the main continuation
was considered to be 10 . . . CL!c6 1 1 . tDf3.
For example, the game Wang Hao - Jako
venko (Poikovsky 2008) continued 1 1 . . . );tc8
1 2 . Af4!? 0-0 1 3. );tb1 );tfd8 1 4 . e3 Axf1
1 5. );txf1 CL!e4 1 6. c;t>e2 tL!aS 1 7. );tfd1 CL!b3
1 8. CL!e1 g5 1 9. Ag3, and here after 1 9 . . .
hS 20. f3 CL!xg3+ 2 1 . hxg3 aS Black could
have held the position. The game went 19 . . .
f5 20. Ae5 c;t>f7 2 1 . f3 and White retained
an advantage.
1 1 . <tlf3 Jac8 1 2 . .tf4 <tl bd7
For the sacrificed pawn Black has a lead in
development and control of the c-fi le. The
question is whether Wh ite can neutralise
these tem porary factors, while retain i ng
his extra pawn.
1 3. <tld2!?
The start of an impressive manoeuvre.
13 Jac2
The best reply was 13 . . .tDh5!, according
to Adams, leaving the evidence in paren
theses.
14. Jab1 Jaac8
a b c d e f 9 h
8 .I 8 3
7 . 7 ii,
6 -*. . 6 ii, :J:l 1
5 5
4 4
3 3
2 .I CD 2 h i s kingside is u ndeveloped , but Black
:J:l ii, :J:l 1
a b c d e f 9 h
The Way to the Top
------
1 5. <tlb3!
Now the depth of Carlsen's idea becomes
apparent. He takes his knight to a1 , from
where it d rives away the rook.
15 .tc4
An original decision , to say the least. Black
goes in for a forcing l i ne, since his play
has come to a standsti l l . I n the event of
1 5 . . .tL!dS 1 6. Ad2 tD7f6 1 7. f3 the move
e2-e4 can not be avoided , while if he at
tem pts to prevent it with 1 6 . . .f5, then after
1 7. CL! a1 );t2c7 1 8 . e3 Axf1 1 9. c;t> xf1 eS
20. dxe5 tL!xeS 21 . c;t>e2 White consolidates
his forces, retaining a sound extra pawn.
1 6. <tla1 ! .ta2 1 7. <tlxc2 .lxb1 1 8. <tla1
<tld5
The attempt to cut off the bishop from the
c1 -square - 1 8 . . . );tc4 1 9. e3 );tc8 20. d1
CL!e4 21 . Ag3 does not prevent White from
completing his development after Af1 -a6.
1 9 .td2 e5
a b c d e
8
6
5
4
2
a b c d e f 9 h
A unique situation: White's passive pieces
block the enemy rook's access to his rear,
has no way of strengthening his position.
Therefore he makes a last attempt to sup-
1 44
 M . Carlsen - M. Adams

port his fading initiative, as otherwise after
f2-f3 and e3-e4 his opening idea will have
been a fail u re.
20. e3!?
Also possible is 20. d xe5 tbxe5 21 . e3 (in
tending Aa6 and e2) 21 . . . Ad3 (weaker
is 2 1 . . . tbd3+ 22. Axd3 Axd3 23. f3 f5
24. f2 f7 25. d1 ) 22. Ae2, when the
black pieces occupy highly active positions
(variation by Adams). But not for long . Let
us continue: 22 . . .f5 23. f4 Axe2 24. xe2
t'i:lc4 25. c1 c7 26. c2 b5 27. b3! t'i:lxa3
28. xc7 tbxc7, and with his knight cut off
at a3 it is unlikely that Black can hold the
Black succeeds in exchanging the lig ht
square bishop: 27. Ae3 t'i:lf4+ 28. f2 tbd3+
29. Axd3 Axd3, obtaining opposite-colour
bishops and setting up new defensive lines.
For exam ple: 30. d1 Aa6! 31 . d5 Ab7
32. d6 tbd5 33. tbc2 f6 and . . . d8.
24. f2 .tf5 25. b3
White has managed to retain his extra
pawn, but Black's control of the d 5 block
ading point leaves him with defensive re
sources.
f
a b c d e 9 h
8 .!. 8

position . There can follow 29. d3 (with

7 ... ... ... ... 7

the threat of Ab4) or 29. Ab4 t'i:lb1 30. d3. 6 1. 1.& 6

20 . . . exd4 21 . exd4 b8
5 1.& .i. 5

4
i, 4
The manoeuvre 21 . . . tbf8 22. Aa6 e8+ 3 [jj 3
23. d1 tbe6 looks more logical, threat
ening after . . . tbdc7 to regain the paw n .
2 i, 2
: 1
But here too White retains t h e advantage:
a b c d e f 9 h
24. e1 d8 25. tbb3.

22. f3 c6 23. c4 25 . . . .te6 26. flc1 f6 27. a4 aS?!

A careless move, allowing Black to rein-
force the blockade at d5. 23. f2 with the
threat of Aa6 was more accurate. Then
23 . . . Aa2 is dubious in view of 24. b3 t'i:lxd4
25. Ac4, when Black has problems in re-
leasing his stray bishop.
Therefore it is sounder to support the
knight on d5 - 23 . . . Af5 24. Ac4 Ae6, at
the same time setting a small trap: 25. d1
tt::l e3! 26. xe3 Axc4.
This natural reaction to the th reat of a4-
aS allows Wh ite to reg rou p his pieces.
27. . .f7 (with the idea of exchanging the
light-square bishops after . . .tbde7) 28. e1
tbce7 followed by . . .tbc7 was stronger.
28_ .tc3
White frees his knight from the defence of
the pawn, in order to transfer it to a more
active post.

23 . . . fld8 28 . . . At7

23 . . .tt::l c b4 came into consideration: 24. b3 Here also it was not yet too late for 28 . . . f7
t'i:ld3+ 25. f1 tbc7 (25 . . . d8? 26. Axd5 29. e1 &bee? 30. tbd2 tt::l c7.
xd5 27. e2) 2 6 . e2 e8+, and , by
giving up the blockade of the d4-pawn, 29. d2 de7

1 45
 Chapter 2
a b c 36. e3 Ab3 37. xd5 Axd5 suggests
8 8
7
6 6
5 .
4 ::, Jl. ::,
3
Jl. ::,
2 ::, ttJ ::, ::,
a b c d e f 9 h
30 . .lt1 !
' I could have exchanged the bishops, but
with almost no time left I suddenly found
a very good plan', Carlsen comments. He
sacrifices the d4-pawn, advantageously
transforming the position. I ndeed , on this
occasion too Magnus's intuition did not let
him down . After 30. &td1 xd4 31 . Axf7+
xf7 32. c4 the exchange of the d4-pawn
for the b6-pawn is indeed favourable for
White, but Black could also play 30 . . . Axc4
31 . xc4 d5, continuing his blockading
strategy.
30 . . . xd4 31 . e1 dc6 32. c4 d5
33. b1 !
'The point. Black cannot take o n c3 due
to the b6-weakness, and as a result White
will have time for a desirable regrouping.'
(Carlsen)
33 . . . f8?!
It is not surprising that in time-trouble Ad
ams does not have the time to see through
the opponent's idea, and he makes what
is in general a useful move, centralising
the king. But why not attack the a4-pawn?
33 . . . Ag6 34. &td1 (34. &tc1 &tea with the
idea of . . .cb4) 34 . . . Ac2 35. &txd5 &txd5
The Way to the Top
itself, retaining practical chances of a draw
(variation by Carlsen). 33 . . . Aea is also pos
sible, and if White tries to carry out his
intended plan - 34. Ae1 , then 34 . . .ce7
5 35. b3 Ag6, switc h i n g to an attack on
4
the b3-pawn. If 36. &tc1 there can fol low
3 36 . . . &tea 37. g1 e3 3a. Af2 7d5.
2
34. Ae1 e7?!
Conti nuing o n the same course. It was
n ot yet too l ate to 'jo i n u p on' the a4-
pawn - 34 . . . cb4 35. &td1 (or 35. g 1
&tea 36. Af2 &te6 37. &td1 .tea 3a. b3 Ag6,
forcing 39. a3) 35 . . . c2 36. Ad2 .tea
37. b3 Ag6 etc. 34 . . . Ag6 35. &td1 Ac2
36. &td2 Axa4 37. xb6 Ab3 3a. xd5
&txd5, simplifyi ng the position, was also
possible.
a b c
8
a b c d e f 9 h
35. g1 !
This was Carlsen's idea: the king moves
away from the centre, but vacates the f2-
square for the bishop, from where it attacks
the b6-pawn , after which, relying on the
strength of his bishop pair, Wh ite gains
good winning chances.
One never tires of admiring the depth
and precision of Carlsen's ideas! On the
1 3th move he embarked on a manoeuvre
1 46
 M . Carlsen - M. Adams

with his knight into the corner of the board , a b c d e

which enabled him to keep his extra pawn, 8 K 8
and on the 30th he devised a retreat by
his pieces, to obtain a strategically more
7

favourable position with equal m aterial.
6
ttJ 6

The ability to arrange the pieces on the 5

necessary squares, the depth of ideas and

the purity of their implementation i n some 3

way remind one of Jose Rau l Capablanca. 2

35 . . b8
.

Black takes his knight to d7, to support
the b6-pawn. More active play came i nto
consideration - 35 . . .tDeS!? 36. tDxe5 fxe5
37. Ah4+ tDf6 38. e1 d6 39. Ag3 tDd7
(39 . . . e8?! 40. Ab5 e7 41 . f4!) 40. Ab5
AdS 41 . d1 (41 . Axd7 xd7 42. AxeS+
c5 leads to the c reation of opposite
colour bishops) 41 . . .e6, and White has
to look for new ways.
36 .lf2
a b c d e f g h
40 . . . e5?!
A mistake on the last move before the time
control. The exchange of the bishop cre
ates a situation where the superiority of the
two bishops over the two knights, which
have no strong points, is very perceptible.
40 . . .tDc5 was also unsuitable: 41 . tDxf7
xf7 42. AxeS bxc5 43. Ac4 .!::l d 7 44. d1
e6 45. f4 g5 46. g3 and Black cannot

Carlsen comments that he used h i s re disentangle himself (Adams). But 40 . . . Ag8

maining time studying the king walk after
36. tDxb6! tDxb6 37. Axa5 tD8d7 (37. . . d6?
38. Ab4) 38. Ab4+ e6 39. e1 + f5
(39 . . . d5? 40. Ab5! is bad for Black) and
he was unable to find a win. But had there
been a little more time, he would surely have
found 40. Ad3+ g5 41 . e7 tDf8 42. xf7
.!::l x d3 43. Axf8 tDxa4 44 . .!::l xg7+ with new
material gains. It should be added that
39 . . . tDe5 would also not have saved Black
after 40. Aa5 .!::l b8 41 . f4 tDxa4 42. fxe5 d7
(42 . . .tDxb2? 43. exf6+ xf6 44. Ac3+ g6
45. b1 ; 42 . . .tDxb2? 43. Ac4+; 42 . . .fxe5?!
43. Ac7) 43. exf6 gxf6 44. b4 with a tech
nically won endgame. In short, this could
have been a worthy conclusion to the game.
36 d7 37. S:e1 + f8 38. S:d1 We7
. . .
39. E:e1 + Wf8 40. d6
was more resil ient.
41 . xf7 Wxf7 42. E:d1
The centra l i sed black k n i g hts are hob
b l ed by the w h ite bishops stan d i n g at
the rear.
42 . . . We7 43. f4 g4
B l ac k wou l d h ave lost m aterial after
43 . . .tDxf4 44. xd8 xd8 45. Axb6+.
44. E:e1 + Wf8
Consideration should have been given to
44 . . . d6!? 45. Ad4 tDxf4 46. e4!? g5
47. Axb6 b8 48. d4+ c6 49. Axa5
x b2 50. Ad2 (if 50. b4 there follows
50 . . . a2 with t h e t h reat of 51 . . . a1 )
50 . . .tbe6, and the black pieces are more
active than i n the game.

1 47
 Chapter 2 The Way to the Top
------------------

45. J.d4 !ad6 tDxd6 50. Axb6 the knights lose their last
strong point and are u nable to resist the
45 . . .tDxf4 was dubious in view of 46. Axb6 long-range bishops, as occurred in the
d5 47. e4 g5 48. g3 f5 49. d4. game, Even so, B l ac k cannot avoid the
exchange of rooks.
46. h3
49. g4 h6
a b c d e
After 49 . . .tDfe3? the pin 50. e1 is decisive.
8 8

7 50. f5 f7 51 . J.g2 f4 52. !axd6 xd6

53. J.xb6 c4 54. J.c5+ md7
6 .I 6

,. 5
a b c d e f g h
i, ,. 4
8 8
3 3
7
... ... 7
2 2
6 ... 6
J: Ji, 5 ... i, 5
a b c d e f g h
4 ,. ,. 4

3 3

Adams thinks that it was possible to capture
the pawn - 46 . . .tDxf4 47. e4 tbe6 48. xe6
xe6 49. hxg4 e4 50. Axb6 xa4, but
Black would hard ly be able to save such
an endgame. For example, 51 . Ab5 a1 +
(in the event of 51 . . . xg4 52. Axa5 r!:le7
53. Ac3 the outside passed pawn, sup
ported by the bishops, is bound to win)
52. r!lf2 b1 53. Ac5+ r!lf7 54. Ac4+ r!:le8
55. b3. The king approaches the centre,
after which the b6-pawn is won . Possibly
this was the only practical chance, since in
the subsequent play Carlsen did not offer
another one, unhurriedly bringing the game
to a victorious conclusion.
2 i, 2
a b c d e f g h
55 . .tf1
The position is a purely technical one, and
through i nertia Carlsen retains the bish
op pair. But, as it transpires, the knights
are able to stand up for themselves, and
avoiding the exchange merely prolongs the
game. 55. b3! tDxg2 (55 . . . tDd2 56. Ae3)
56. *xg2 tDd2 57. b4, creating an out
side passed pawn, was stronger. 57. . .tDe4
58. Af8 r!le8 59. bxa5! is an amusing vari
ation.

47. !ad1 ! 55 . . . xb2 56. J.b5+ mda

Dispelling Black's last i l l usions (47. . .tDxf4?
48. Ac5).
47. . .f5 48. J.f2 me7
After the rook exchange 48 . . .tDxf4 49. xd6
56 . . . r!lc7 was more resi l ient: 57. Af8 g6
(or 57. . . tDxh3+ 58. r!lh2 tDf2 59. g 5 fol
lowed b y A xg7) 58. fxg6 h x g 6 59. h4,
although here too Black is unable to save
the game.

1 48
 M . Carlsen - I. Cheparinov

57. Ab6+ e7 58. h2 The critical reply to the development of

the k n i g ht at d 2 is 4 . . . d5, restricting its
Capturing the pawn would have won more influence in the centre, but the move 4 . . .
quickly - 58. Axa5 tbxh3+ 59. g2 tbf4+ d 6 also has its vi rtues. However, from a
60. f3 tbd5 61 . e4. romantic such as C hepari nov one would
not expect anything else.
58 . . . d5 59 . .txa5 d6 60. Ad2 c5
5. e4 0-0 6. c3 c5
Now, when the king has come to the aid of
the knights, Wh ite cannot get by without a b c d e f g h
the exchange of one of his bishops. .i .i. 'if
8 .! 8

61. g3 c7 62. Ae3+ b4 63. Ad2+ 7 .\ .\ .l .l .i. .\ 7

mcs 64. Ac1 c4 65. Axc4 xc4 6 .\ 6
66. Ad2 a6 67. aS 5 i!. 5

Of course, i n this endgame White also has 4 4

an easy win. 3 {jj 3

67... b5 68. f3 c5 69. Ac3 h6 2 {jj 2

70. e3 c4 71 . Ad4 a6 72. e4 i!. M 1
a b c d e f g h
b4 73. h4 b5 74. Ac3 a6 75. d5
c5 76. Ad4 d3 77. e6
7. dxc5
Black resigned.
The game has transposed i nto a q u i et
variation of the P i rc - Ufi mtsev D efence
Game 32 where, as is well known, the attem pt to
M . Carlsen - I . Cheparinov maintain the pawn tension by 7. Ad3 leads
World Cup, Khanty- Mansiysk 2007 to the exchange of White's i mportant dark
square bishop: 7. . . cxd4 8. cxd4 h6 9. Ah4
Torre Attack [A48]
tbh5! with good prospects for B lack . I n
1. d4 f6 2. f3 g6 3 . .tg5 this respect the game T i m man -Topalov
(Belgrade 1 995) is instructive: 1 0 . 0-0 g5
Carlsen chooses the Torre Attack, avoiding
1 1 . Ag3 g4 12. tbh4 tbxg3 13. hxg3 Axd4
the main l ines of the Grunfeld and King's
1 4. CL!f5 Axf5 1 5. exf5 h 5 1 6 . f6 CL!d7 (1 6 . . .
Indian Defences, which are possible in re
e6 1 7. CL!e4 Axf6 1 8. d2 g7 or the imme
ply to 3. c4. A quite logical decision in a
diate 1 6 . . . Axf6 is stronger) 1 7. fxe7 xe7
mini-match in the World Cup, where with
1 8. Af5 CL!c5 1 9 . CL!c4 with sharp play, where
the 'classical ' time l i m it the white pieces
White has sufficient com pensation for the
were of primary i m portance and there was
pawn.
a risk of running i nto a prepared line by the
productive Bulgarian Topalov-Cheparinov 7. . . dxc5 8. Ac4
pairing.
After 8. Ae2 CL!c6 9. 0-0 c7 1 0. c2 Ae6
3 . . .tg7 4. bd2 d6
. Black has an easier game.

1 49
 Chapter 2 The Way to the Top

8. . . c6 9. 0-0 Yl/c7 1 4. Ag3 xg3 1 5. hxg3 b6 1 6. h2

9 . . . tila5!? is m o re active: 10 . Ae2 Ae6
1 1 . e1 a6 1 2. Wc2 (or 1 2. Af4 h6 1 3 . Wc2
b5 1 4 . h3 'Wb6 with equal chances) 1 2 . . . b5
1 3 . b3 xb3 14. axb3 We? with good play
for Black, Malaniuk- Marin (Romania 1 992).
1 0. Yl/e2
.thS
A provocative move: Chepari nov invites
the enemy fire. 1 6 . . . Ad7 was more cir
cumspect.
1 7. g4 .tg6 1 8. g3 Dad8 1 9. f4

The main continuation is considered to be a b c d e f 9 h

1 0. e1 , after which Black can carry out 8 .1 .i.e 8

the same l i n e with the exchange of the 7 .l 'iV .l .l .i. 7
dark-sq uare bishop: 10 . . . h6 1 1 . Ah4 h5 6 .ll.&\ ..t .l 6
1 2 . 'Wc2 Ag4 1 3 . Ag3 xg3 1 4 . hxg3 h5 5 .l .l 5
1 5. f1 Ah6, obtaining comfortable play, 4
Burmakin - Kurnosov (Ulan Ude 2009).
BBB 4
3 B VJ!J B 3
10 . . . h6 1 1 . Ah4 h5! 2 DB t2J 4:J2
The crucial manoeuvre i n this variation, 1 l:[ II
a b c d e f 9 h
leading to the exchange of the dark-square
bishop.
1 9 . . . d4!?
1 2 . 1Ue1
Nevertheless Ivan im poses his play! The
If 12. we3 there can follow 12 . . .a5 13. Ad3
advanced knight is immune on account of
c4 1 4 . Ac2 g5 1 5. Ag3 xg3 1 6. hxg3 d8
the loss of the queen.
with good play for Black.
20. flac1 bS! 21 . Af1
1 2 . . . Ag4 1 3. We3!

A useful move, moving the queen out of
the pin. Now the natural 1 3 . . . b6 is met by
1 4 . e5! , shutting the g7- bishop out of the
game.
1 3 . . . g5
After the inclusion of 1 3 . . . tila5 1 4 . Af1
and now 14 . . . g5 Black has to reckon with
1 5. xg5!? hxg5 1 6. 'Wxg5 Wf4! 1 7. h3 Ae6
1 8 . Wxf4 tilxf4 1 9. Axe? fc8 20. Ag5,
when White has three pawns for a knight,
which may be advantageous i n the end
game.
After the active 21 . AdS with the threat of
twice capturing on d4, White has to reckon
with 21 . . . 'Wb6!?, when the attem pt to shut
in the bishop - 22. f5 Ah7 23. df3 xf3+
24. tilxf3 e6 25. Ab3 c4 leads to the ex
change of queens, which does not come
into Carlsen's plans.
21 . . . gxf4 22. gxf4 e6 23. eS
The fork 23. fS is parried by 23 . . . Ae5
24. hf3 Af4 25. We2 g5, and if 26. fxg6
Black wins by 26 . . . Axd2 27. xd2 xd2
28. 'Wxd2 f3+.

15
 M. Carlsen -I. Chepari nov
------

a b c d e f g h 28. J::lx e5
8 8 K Ke
A prag m atic approac h . Without delving
'iV ..t 7 into detai ls, Magnus eliminates the pow
6 ..t 6 erful bishop.
5 5
28. . .'Wxe5 29. df3
4 4
3 Wi 3 Here Carlsen wrongly avoids going i nto an
2 2 endgame, since the pressure of the heavy
pieces on the central files is so great that
l:r l:rlt
a b c d e f g h there are no prospects of activating the
minor pieces. After 29. Ag2 'We2 30. tDdf3
23 . . . f5! gd 1 + 31 . g x d 1 'Wxd 1 + 3 2 . 'Wf1 'Wxf1 +
33. tDxf1 tDxf3+ 34. Axf3 gxg6 35. f2
Cheparinov is eager to sacrifice a piece. He he would have obtained a more favourable
cannot remain for ever under the impend version of the endgame than that which oc
ing threat of the pawn phalanx! curred in the game. As for Sergey Shipov's
recommendation 33. Axf1 tDxf3+ 34. tDxf3
24. exf6J::lx f6 25. f5 g5 26. fxg6J::le6
a6, this endgame is safer for Black, since
27. 'Wf2 .le5!
it is not possible to retain the g6-pawn -
a b c d e f g h 35. tDh4 ge4 36. Ag2 gxg4 37. tDf5 gg5
8 K 8 38. tDxe7+ f8 39. tDd5 gxg6, and with
his outside passed h-pawn Black is not in
7 'if 7
danger of losing. Also insufficient is 35. f2
6 K 6 gxg6 36. g3 gd6 37. Ae2 e5! 38. tDxe5
5 .i. 5
gd2, when the rook reaches the second
4 4 rank.

29 ... xf3+ 30. xf3 'Wf4 31 . J::le1 !

Carlsen finally realised that he needed to

a b c d e f g h exchange the queens.

For the sacrificed piece Black has a strong
attack . For exa m ple, Wh ite loses after
28. tDhf3? Ag3 29. 'Wg2 Axe1 30. tDxe1
'Wf4! or 30. gxe1 gxd2 ! . Somewhat better
is 28. tDdf3 Ag3 29. 'Wg2, when after ex
changes on e1 - 29 . . . Axe1 30. gxe1 gxe1
31 . tDxe1 - Black retains the initiative: 31 . . .
c4! 32. tDef3 'Wb6+ 33. h1 'Wxg6, since
White's minor pieces are sti l l passive (vari
ation by Shi pov).
31 . . .J::lx e1
It is not clear - who i n fact is playing for
a win? Black could have captured the g4-
pawn - 31 . . .'Wxg4+ 32. Ag2 gxe1 +, forcing
a d raw i n the variations 33. tDxe1 'Wxg6
34. 'Wxc5 gd1 35. 'Wxe7 'Wg3 36. f1 'Wf4+
37. g1 'Wg3+ or 33. 'Wxe1 'Wxg6 34. 'Wxe7
gd1 + 35. f2 'Wc2+ 36. g3 'Wg6+.
32. xe1 'Wxf2+ 33. xf2

151
 Chapter 2 The Way to the Top
--------------------

a b c d e f g h The subseq uent play in this ending resem

bles a three-act d rama. Act One - lost
8 8

7
ill usions. Carlsen ignores the sharp move
37. b4, aiming by the time control on move
6 6
40 to obtain a position with a clearer pawn
5 5
structu re.
4 4
37. a3
3 3

2 2 Good winning chances would have been

L:iJJt
a b c d e f g h
If this position is com pared with the one
exa m i ned i n the n otes to W h ite's 29th
move, it i s obvious that the situation is
worse for him.
33 . J;td2+ 34. e3 d1
.
Black makes things difficult for hi mself. He
should have reduced the number of pawns:
34 . . J:!xb2! 3S. tbd3 .c2 36. tDxcS .xc3+
37. <;!.>d4 .g3 38. Ae2 <;!,>g7, and after the
win of the g- or the a- pawn he is close to
a draw. This is confirmed by the variations
39. Ad3 .xg4+ 40. <;!.>es .g2 41 . tDe6+
g iven by the energetic 37. b4! , blocking
the queenside pawns, which would have
enabled the minor pieces to be activated .
Let us consider the main variations:
a) 37. . . . xa2+ 38. <;!.>e3 <;!,>g7 39. Ae4 e5
(otherwise <;!.>d4) 40. tDc2 .b2 41 . AfS! a6
42. <;!.>t3! .b3 (or 42 . . . .a2 43. tDe3 .a3
44. tDdS) 43. <;!,>e4 .xc3 44. <;!,>xeS, and
Wh ite develops an attack on the king;
b) 37. . . cxb3 38. axb3 a5 (in the event of
38 . . . g7 Wh ite attacks with 39. Ae4
.a3 40. tDg2! .xb3 41 . tDf4) 39. Ac6! a4
(39 . . . b4 40. c4) 40. b4! , and the a- pawn
will be stopped by tDc2.
37 b1 38. Ae4! xb2+ 39. tilc2 g7
..

<;!.>g8 42. tbc7 hS or 39. gS hxgS 40. Ad3

40. e3
a6 41 . tDxa6 eS+ 42. <;!,>e4 <;!,>xg6 43. tDcS
g4 44. <;!,>xeS+ <;!.>hS 4S. AxbS .a3, where a b c d e f g h
the outside passed pawns g ive Black suf 8 8
ficient cou nterplay.
7

35. e2 a1 36. Ag2 c4 6 6

a b c d e I g h 5 5
8 8 4
Jt 4
7
3
6 6 2 2
5 5

4 4 a b c d e f g h

3 3
This was the position Carlsen had aimed
2
Jt for. The black pawns have been stopped ,
L:2J and the wh ite pieces are excellently co
a b c d e f g h ordinated . It would appear that the win is

152
 M. Carlsen -I. Cheparinov
a matter of technique. I ndeed , after the
attem pt to i m pede the king by 40 . . . e5
White creates an attack against its oppo
site number: 41 . &Db4! b3 42. tDd5 xa3
43. Af5! b4 44. &De?! (with the th reat of
&De8+) 44 . . . f8 (44 . . . xc3+ 4 5 . e4)
45. e4 bxc3 46. xeS e7 (otherwise
f6) 47. &Dd5+ f8 48. &Df4! a6 49. d4,
winning the pawns (variations by Shi pov).
But all is not so clear.
40.. b3!
Act Two - disappointment. This accu rate
rook move, tying the king to the c3-pawn ,
is a kind of cold shower for White. It tran
spi res that after the i ntended 41 . tDd4
there fol l ows 41 . . . aS 42. c5 (42 . Af5
f6 and . . . e7-e6) 42 . . . xc3 43. x b5
g3 44. Af5 f6 45. &Dd4 xa3 46. xc4
e3 and Wh ite is stymied. If 47. &De6 there
follows 47. . . xe6, with the theme of the
dark-square corner in mind: 48. g7 c6+
(or 48 . . . xg7 49. Axe6 h5! 50. g5 g6)
49. b5 xg7 50. xc6 h5! 51 . g5 a4. He
is forced to retreat his king.
41 . d2 f6?
a b c d e f 9 h
8
8
7 52. xg5 hxg5 53. J.xc4 Wg7
6 6
5 5
4 4
3 6
2 2
a b c d e f 9 h
An unexpected conclusion to the Second
Act. After the pin 41 . . . b2 there is noth ing
more sensible for White than 42. e3, when
Black has to repeat moves 42 . . . b3 with a
probable draw. Now, however, White gains
an opportunity to support the g6-pawn and
he should be able to win.
42. d4!
In view of the threatened knight manoeu
vre &Dd4-e2-g3-h5, the idea of breaking
through with the a-pawn after 42 . . . b2+
43. e3 is too late, and so Black is forced
to exchange his b5-pawn for the a3-pawn,
which radically changes his pawn structure
for the worse.
42 . . . xa3 43. xb5 a5 44. c7 Wg7
Black also loses after 44 . . . g5 45. Af5 h5
46. &De8+ e5 47. gxh5! xf5 48. g7, when
he has to give up his rook.
45. e6+ h8 46. e3 a1 47. d4 aS
48. J.c6!
The bishop is hurryi ng to f7.
48 . . . a4 49 .tea g1 50. g5!
The g5-pawn joins the group supporting
the g6-pawn - it has to be covered .
50. . . a3! 51 . .tf7 xg5!
Here is a surprise! Just try in time-trouble
coping with the advancing pawns.
a b c d e f 9 h
8 8
7 7
6
5 5
4 4
3
2 2
a b c d e f 9 h
153
 Chapter 2
Act Three - swings. It has nothing in com
mon with the preced ing acts. It was per
formed in severe time-trouble by two actors
worn out by the very difficult struggle in the
midd legame. The strength of the fighters
was exhausted , and in such a condition,
with time running out, calculating work is
practically impossible. As fortunes swung
backwards and forwards, Caissa smi led
on her favourite.
54. d3?
This m ove th rows away the w i n , which
could have been achieved by the study
like 54. e5! xg6 55. Aa2. Now the race
to queen a pawn - 55 . . . g4 56. f4 h5
57. c4 h4 58. c5 g3 59. f3! h3 60. c6
ends with a winning check. It is also pos
sible to win more prosaically: 57. g3 g5
58. c4, and Black cannot hold his g4-pawn.
54 . . . xg6 55. c2 g4 56. b3 f5
In the pawn race, time is i m portant. After
56 . . . e5 57. AdS f5 58. c4 e4 59. c5 e3
Black would also queen one of his pawns.
57. xa3 g3?
57. . . e5 or 57. . . e4 wou l d have led to a
draw. There is no point i n commenting in
depth on instant play i n a time scram ble,
but we will merely touch on the errors.
58 . .tf1 ?
Why not the immediate 58. AdS, occupying
the long diagonal?
58 . . . e5?
The Way to the Top
-----------------
58 . . . e4 would have saved the game.
59. b3?
The blockading 59. Ag2 would have won.
59 . . . f4?
Here also it was not too late for 59 . . . e4.
60. J.g2
It is no longer possible to go wrong, as
. . . f3 was threatened .
60 . . . e3 61 . c4 f2 62 . .t.e4
Black resigned.
This and the preced ing games used up
much of Carlsen's strength . Everyone has
their l i m it, and in the sem i -fi nal he lost
without a particu lar fig ht to G ata Kam
sky (%-1%), who in fact won the World
Cup. Questioned whether he was disap
poi nted to have gone out of the World Cup
at the semi-final stage, Magnus replied: ' I
hadn't expected t o go s o far. A n d I was
very tired, so basically I was ready to go
out any point.' This was not surprising,
since for practically the entire year he had
not had any respite, playing tournament
after tournament. At a press conference
Carlsen looked extremely tired . I ndeed to
reach the sem i-final of the World Cup after
such a lengthy marathon of difficult battles
was an ind ication of his fighting spirit. An
important stage in his career had come to
an end - at the age of 1 7 Carlsen had made
his way to the top and was establ ished
among the world elite.
154
 Chapter 3

Life at the Top

In 2008 the organisers of the chess festival i n Wij k aan Zee were able to assemble
the strongest field for the mai n tournament i n its entire h i story (FI D E category 20).
They could hardly have imag ined that i n such prestigious company Carlsen wou l d
ach ieve the greatest success i n his short career. H e shared 1 st-2nd places with
Aronian, ahead of all the main nomi nees i n the battle for the world crow n . The last
year had not been wasted . From a complete fiasco in h i s fi rst A tournament in 2007,
without a single w i n , to a triumph with five won games - this was the res u lt of an
enormous amou nt of work. But as before, M ag n u s was modest: ' Before the tour
nament I thought that if I could score 50%, it wou l d not be a bad result,' h e said i n
a n i nterview for t h e magazi ne 6 4 Shakhmatnoe obozrenie. ' I wasn't expecting to
wi n , of course, and the share of fi rst place was a pleasant su rprise for me.' But, as
they say, appetite improves d u ring the meal. M ag n u s h imself realised how he was
improving from one tournament to the next . Along with the perfecti n g of h i s posi
tional understand i n g , the tec h n i q u e of playing simple positions and i n particu lar the
endgame, he was also accumu lating experience i n h i s meetings with the strongest
players in the world . His self-esteem was also g rowing: ' I can no longer consider
50% a good result. Of course, this is a big success, but i n the world there are sti l l
many players who are stronger t h a n m e . To become world champ i o n , I will have to
develop and improve in many respects.'

Game33 White's system of development associated

M. Carlsen P. Eljanov
-
with h i s l ast two moves has the aim of
Wijk aan Zee 2008 forcing Black to concede the centre. The
GrCmfeld Defence [ 091} roots of Wh ite's idea lie i n the game Ale
khine- Bogolj u bow (Vienna 1922), where
Wh ite first developed his bishop - 4. Ag5,
It is rare to win a simple position agai nst
after which there followed 4 . . .etJe4. Later
a stro n g g randmaster without any ap
things were refined: after the inclusion of
parent effort, based on factors that are
4. tDf3 Ag7 the move . . . etJe4 causes Wh ite
deeply hidden. This game, although with
less trouble.
out outward effects, is impressive for its
completeness.
(see next diagram)
1. d4 f6 2 . c4 g6 3. c3 d5 4. f3
.tg7 5 . .lg5

5
 Chapter 3 Life at the Top

a b c d e f 9 h hoping to exploit the opening of files for

8 i. ..i.'i. i. 8 his rooks, supported by the long-range
bishop on g7.
7 .\ .l .l .l..i. .\ 7
14. dxc5
6 .l 6

5 5 1 4 . lahd1 tbc6 1 S. dxcS is also played , but

ii,
4 4 here Black has already completed his de
velopment and he can begin counterplay
3 3
on the q ueenside, exploiting the position
of the white king in the centre. For exam
ple, the game N everov - Pavlovic (Hast
a b c d e f 9 h
ings 200S/6) continued 1 S . . . 'Wc8 1 6. laab1
bxcS 1 7. 'Wc4 tbaS 1 8. 'WbS tDc6 1 9. 'Wc4
5 tbe4 6 .lh4 tbxc3 7. bxc3 dxc4

(1 9. 'WxcS 'Wa6+) 1 9 . . . lab8 20. laxb8 'Wxb8

8. 'Wa4+ 'Wd7 9. 'Wxc4 b6 1 0. e3 .la6
1 1 . 'Wb3 .lxf1 1 2 . mxf1 0-0
Carlsen's reputation in questions of home
preparation had grown so much, that El
janov avoided 1 2 . . .tDc6 1 3 . e2 0-0, which
he had successfu lly employed in a game
with Kiri l Georg i ev (Wij k aan Zee 2006),
where after 14. lahd1 tDaS 1 S . 'Wb4 lafe8
1 6. flac1 laac8 1 7. c4 'Wg4 1 8 . h3 'We4
1 9 . 'Wa4 eS he gained q u ite good coun
2 1 . 'WxcS lac8 2 2 . A x e? (22 . f1 was
more cautious) 22 . . . 'Wb7 23. Ad6 'Wa6+
24. lad3 tDaS (24 . . .tDeS!? 2S. 'Wxc8+ 'Wxc8
26. AxeS AxeS 27. tbxeS 'We6 also came
i nto consideration, obtaining an outside
passed pawn after . . . 'Wxa2) 2S. 'WdS tbc4
26. tbgS tDb2 27. 'Wxf7+ h8 with sharp
play, favourable for Black.
14 tba6!?
.

terplay. Black intends to play his knight to the ex

cellent square cS with gain of tempo in
1 3. me2
the event of 1 S. cxb6 tDcS 1 6. 'Wc4 axb6
a b c d e f 9 h 1 7. lahd1 'Wb7, obtaining excellent compen
8 i. i.e 8 sation for the pawn ( . . . f!a4 is threatened).
7 .\ .l 'iV .l .l..i. .l 7 1 5. Elhd1 'Wb7
6 .\ L J .\ 6 The alternative is 1 S . . . 'Wc7. I n the game
5
- .-I 5 Lautier- lvanchuk (Terrasa 1 991) White de
cided to accept the pawn sacrifice, and this
4 [::, ii, 4
is what happened: 1 6 . cxb6 axb6 1 7. a4
3 V/i{[::, [::,CiJ 3
tDcS 1 8 . 'Wb4 laaS 1 9. Ag3 eS 20. tDd2 (or
2 [::, [::,[::,[::, 2
20. f1 lafa8 21 . 'Wc4 bS) 20 . . . lafa8 21 . CL!c4
1 1:( 1:( 1 tbxa4! 22. f1 Af8! 23. tDd6 tbxc3 24. laac1
a b c d e f 9 h
lacS 2S. lad3 tbe4! 0-1 .
13 . c5 But Wh ite is by no means obl iged to
I n provoking the capture on cS, which is accept the pawn sacrifice, and he can play
positionally advantageous for White, Black 1 6 . c6!?, as in the game.
associates this move with a pawn sacrifice, 1 6. c6

156
 M. Carlsen
Of course, 1 6. cxb6?! axb6 1 7. !;iab1 tDcS
18. 'ti'b4 tbe4 did not come into Carlsen's
plans. In his preparations for the g ame
he was aimi ng for a favourable endgame.
16 ... Wxc6 1 7. Axe7 Eife8 1 8. Aa3 Wxc3
Black regains the pawn, since in the event
of 18 . . . !;iad8, as played in the game Gri
vas - Leko (Dortmund 1 992), the compen
sation for it is insufficient: 1 9. !;iac1 tbc7
20. c4 tbe6 21 . f1 'We4 22. !;idS.
19. Wxc3 Axc3 20. Eiac1 Ab4 21 . JJ.b2!
White deploys his pieces in their most ac
tive positions.
21 . Af8 22. tbd4 tbc5
..
a b c d e 4
a
.I 8 3
7 2 ::,
6 6
5 5
4 4
3 3
a b c d e f 9 h
23. g4!
A deep move, aimed not only at seizi ng
space, but also at fighting for the weak
ened f6-point.
23 .. Eie4
Here there is nothing for the rook to do - the
knight at d4 stands like a rock. According to
the theory of positional play, formulated by
the fi rst world champion Wilhelm Steinitz,
'small advantages' should be accumulated .
Among these is a q u eenside pawn ma
jority, which Black has here, as in many
Grunfeld-type positions. But, in contrast to
other variations with a strong pawn centre
for Wh ite, where the black pawns are very
157
- P. Eljanov
mobile, in the given specific position they
are immobil ised . Therefore Black should
have thought about preparing to advance
them - 23 . . . a6 24. Aa3 !;iac8 etc.
24. f3 Eiae8 25. h3 f6
Log ical. Black has gone totally onto the
defensive and h e brings his king to the
centre.
26. Aa3 f7 27. Eic2
8 8
7
6 6
5 5
1.&
tt:J.I ::, 4
::, ::, 3
n ::, 2
n
a b c d e f 9 h
Carlsen 's technique in this endgame re
sembles t h e ' i ntricate' style of Anatoly
Karpov in his best years, when he would
lull his opponents with unhurried manoeu
vres, maintai ning the pressure and forcing
them sooner or later to weaken themselves.
Wh ite prepares to double his rooks on one
of the files, intending the knight sortie to bS.
27 tba6?
..
This is wrong. The knight at cS was ac
tively placed , and the decision to exchange
bishops is a strategic mistake. Now Black's
rear, especially the d6-square, is extremely
weak, and the wh ite rooks are able to in
vade. He should have stuck to waiting tac
tics, for example 27. . . !;i4e7, offering White a
choice: to play an endgame with an isolated
cS- pawn after 28. AxeS bxcS 29. ttlb3
(29 . !;ixcS !;ixe3+) 29 . . . !;ic7 30. !;ic4, or
continue manoeuvring - 28. tDbS !;ib7, al
though finding the key to Black's position
 Chapter 3
is not easy. Thus, here the exchange of the
f8-bishop is not so dangerous: 29. tbd6+
Axd6 30. xd6 c7, and Black sets up
new defensive li nes, since going into the
roo k e n d i n g by 31. Axc5 bxc5 has its
drawbacks, if only because of their draw
ing tendencies. Apparently, the spirit of
the preceding strategy could have been
mai ntai ned by a pawn offensive on the
kingside - 29. h4 with the idea of break
ing through, depending on circumstances,
with g4-g5 or h4-h5. Here are some pos
sible variations:
a) 29 . . . h6 30. c4 e6 31 . tDd4 (or 31 . d8
Ae7 32. h8 g7 33. a8) 31 . . . e8
32. tbe2 c8 33. h5;
b) 29 . . . e5 30. Ab2 e6 31. cd2 Ae7
32. d5 a6 33. tbd6+ Axd6 34. xd6
c7 35. d8.
28. .xf8 mxf8 29. grc6
The start of the decisive i nvasion.
29 . . .mg7 30. b5 gr4e7 31 . grdd6 c5
32. c7
The capture 32. xf6? tbd7 leads to the
loss of the exchange.
32 . . . grf8 33. h4 grff7 34. d5
a b c d e f 9 h
8 8
6 6
5 5
4 4
3 3
2 2
a b c d e f 9 h
Complete domi nation by White. Black is
forced to exchange a pair of rooks (oth-
Life at the Top
erwise the f6-pawn cannot be defended),
after which the endgame becomes of a
technical nature. Sooner or later the e- and
f-pawns will advance, creating a passed
pawn.
34 ... grd7 35. grxd7 xd7
35 . . . xd7 3 6 . tD x f6 f7 37. g 5 tbd7
38. e4 tDxf6+ 39. g xf6+ xf6 40. xf6
<.!?xf6 41 . d5 leads to a won pawn end
game for Wh ite.
36. mg3 c5 37. f3 h6
Activity with 37. . .f5 i s i n a p p ropriate -
38. gxf5 g xf5 39. h5. 38. h5 tbd3 39. g5
gxh5 40. h6 is also good for Wh ite.
38. f4 g5
38 . . . h7 39. h5 g5 40. tbd5 g7 41. e4
was little better for Black.
39. h5+ mg6 40. f4 gxf4+ 41 . exf4
mh7 42. t5 m98 43. mt3
The decisive word belongs to the king!
43 ... d7 44. me4 mf8 45. grC8+ me7
46. md5 b5 47. grh8 b6+ 48. mc6
c4 49. gra8 e5+ 50. mc5 d7+
51 . mxb5 md6 52. grx a7 grf8 53. mb4
c5 54. mc4
Black resigned .
Game34
V. Kramnik M. Carlsen
-
Wij k aan lee 2008
English Opening [A30]
I n this game Magnus demonstrated the
significance of a correct opening choice.
With the aid of the cunning ' hedgehog' he
was able to outplay the world champion.
158
 V. Kram n ik- M. Carlsen

1. f3 f6 2. c4 e6 3. c3 c5 4. g3 1 3 . b3 0-0 1 4. tbge4 d7 1 5 . Ab2 e8

b6 5 . .tg2 .lb7 6. 0-0 Ae7 7. d4 cxd4 1 6. ac1 d5 with excellent play for Black.
W
8. xd4 d6 9. d1 a6 But also i n the event of 1 3 . &Dce4 d7
a b c d e I g h 1 4 . &Dxf6+ Axf6 Wh ite cannot make use
8 I. '.&\ 'if I. 8
of tactics: 1 5 . &D x h7 x h7 1 6 . 'We4 h5

..t .t....... 7
1 7. 'Wxc6 c5 1 8 . 'We4 'Wc8 , regai n i ng the
7 ..t _j.... pawn with an equal game (Adorjan - Bis
6 ...I...I l.ll .t.'.&\ 6 choff, New York 1 986).
I 5
5
1 3. ce4 e8!
4 ::, fli 4

3 ttJ ::,
ttJ 3 It is important to defend the pawn without
2 /j, [j, !Jl, 2 simpl ifying the position, since the wh ite
1 !il, knig hts are ' hang i n g ' somewhat. In the
a b c d e I g h
event of 13 . . .&Dxe4?! 14. &Dxe4 a7 1 5. b3
d7 1 6. Ab2 the ' hedgehog 's' spines are
Black has arranged h i s pawns i n a so not so effective: 1 6 . . . d5 1 7. cxd5 x d 5
called ' hedgehog' set-up, the spines of 1 8 . &Dc3 x d 1 1 9 . x d 1 'W c 8 20. tb e 4 ,
wh ich are aimed at the central squares and Wh ite c reated th reats to the k i n g
and b5. (Csom - Kinderman n , Biel 1 986).

10. g5 1 4. b3 lia7 1 5 . .tb2

This continuation was i ntrod uced more a b c d e I g h

than 30 years earlier. White plays his knight
8 8
to e4 in order to attack the d6-pawn . 'ii' '.&\1..
7 J. ..t .t..t..t. 7
10 . . . Axg2 1 1 . xg2 c6 6 .f..t. '.&\ .t..t. 6
The developing 1 1 . . .&Dbd7 does not work 5 ttJ 5

because of 1 2 . tbge4, w i n n i n g the d 6 - 4 ::, tbfli 4

pawn. 3 ::, 3

12. 'Wf4 2 ::, !il, ::, ::, ::, 2

By playing his queen to this active posi
tion , White maintains the pressure on the
d6-pawn , intending to include his bishop
in the attack on it from a3 (after b2-b3).
12 . . . 0-0
The alternative is 1 2 . . . a7, i m mediately
switc h i n g the rook to d7. If Wh ite does
not react to this manoeuvre, the ' hedge
hog ' may demonstrate its spines, as in
the game Webb - Hart (Hastings 1 977/78):
1 1
a b c d e I g h
1 5 . . . d7
I n his commentary Carlsen remarks that
he had not expected such an old-fash
ioned variation as 1 0. tbg5 and he had to
beg i n thinking for h i m self. I ntu itively he
fol l owed the correct cou rse, and since
the young 'classic' is accustomed to plac
ing his pieces on the best squares, this

159
 Chapter 3
move came naturally to h im. I n the game
Wojtkiewicz - Ftac n i k ( B u da pest 1 993)
Black carried out the thematic ' hedge
hog' advance 15 . . . b5!? (with the idea after
1 6 . cxb5 axb5 of playi ng . . . d 6-d5). The
game conti n u ed 1 6 . f3 Was 17 . .ac1
h6 1 8 . ed2 (if 1 8 . c3 there can fol
low 1 8 . . . e5 w i t h the th reat o f . . . Ag5)
1 8 . . . .d 7 1 9 . <!> g 1 Vb7 2 0 . Ye4 bxc4
2 1 . Yxc4 .c7 22. 'id3 d5 with roughly
equal chances.
1 6. E[ac1
White completes his d evelopment, indi
rectly preventing . . . d6-d5.
1 6. . . c7
A new idea i n the ' hedgehog' structure,
fou n d at the board . B l a c k prepares a
pawn offensive with . . .f7-f5 and . . . g7-g5,
driving back the wh ite pieces with gain
of tempo, and before h a n d h e d efends
his e6-pawn . The standard plan i nvolves
the preparation of . . . b6-b5 - 1 6 . . . Ya8
1 7. f3 b5.
1 7. f3 f5 1 8. c3
After t h e retreat 1 8 . ed2 t h e k n i g ht
blocks the d -fi le for the rook, and Black
can switch to play i n the centre - 1 8 . . .
e5!? (here 1 8 . . . g 5 1 9. Ye3 g4 20. e1 is
less effective, since the b6- pawn needs
defending) 1 9. e3 e6 with the idea of
. . .as and . . . cd4.
18 . . . g5 1 9. Wd2 g4 20. e1
Everybody back! The active 20. d4?! is
worse because of 20 . . . Ag5 2 1 . e3 e5,
when the long a8-h1 diagonal passes into
Black's und ivided possession . There is an
especially yawning weakness at f3, which
makes it hard for White to create play on
Life at the Top
the q ueenside: after b3-b4 and c4-c5 the
d-file is opened, creating conditions for the
exchange sacrifice on d4.
20 . . . J.g5!
a b c d e I 9 h
a b c d e I 9 h
The hand of a great master! Black provokes
a weakening of the f3-sq uare, a desirable
one for a knight. Compared with the posi
tion after Black's 1 6th move, the picture
has changed radically. He is in complete
possession of the initiative.
21 . e3 E[ff7
A skilful combination of offensive with
prophylaxis: Black prepares in good time
for Wh ite's cou nterplay with b3-b4 and
c4-c5, by defending his rook on d7.
22. <!>g1 e8
This passive knight is switched to a more
active post.
23. e2 f6 24. f4 We8
' I would have liked to play . . . e4 immedi
ately, but I rejected it because of the queen
sac 24 . . . e4 25. xe6? WeB 26. xg5
xd2 27. xf7. I don't really understand
why though, since 27. . .e4 28. h6+ <!>f8,
with the poi nt of 29. xf5 <!>f7, wins for
Black.' (Carlsen).
160
 V. Kramnik - M. Carlsen

25. Wc3 !lg7 26. b4 e4 27. Wb3 !lge7 a b c d e f g h

The q ueen m u st be freed for more i m 8

'iV 8

portant matters than t h e defence o f the 7I I .l 7

e6-pawn. 6 .l .l 6

a b c d e f g h
5 '.& .l .S.. 5

8
'iV 8
4 '.&QJ.l 4

7 IL__Jr
II .l 7 3 3

6 .l '.& .l .l 6
2 2
5 .l .S.. 5 l:l: . <;t>
a b c d e f g h
4 '.&QJ .l 4

3 3
2 2
1 l:l:QJ <;t>
a b c d e f g h
28. Wa4?!
This attem pt to d ivert the q ueen to the
defence of the q u eenside weaknesses
does not achieve its aim. Kramn i k must
have taken the reply into account, but un
derestimated the dynamic potential of the
centralised knights. At a press conference
after the game Carlsen adm itted that he
judged the position to be equal or slightly
better for White. His i ntuition did not let
him down. As N igel Short pointed out, the
best solution for White was 28. h4! , and
since 28 . . . gxh3 29. Ctixh3 is unfavourable
for Black, it has to be ad mitted that the
knights on e4 and f4 are of equal value.
28 ... e5 29. Wxa6?
The strength of inertia is staggeri ng. White
should h ave stopped in good time and
returned his queen to b3. Other moves are
unsou nd. If 29. c5?! Black has 29 . . .Ctic4! ,
while after the exchange 29. Axe5? dxe5
White loses a piece (30. gxd7 exf4 31 . gcd1
et:lc3!).
29. . . !1a7
30. Wb5
Wh ite is forced to retreat. When he made
his queen sortie on the 28th move Kram
1
nik overlooked that after 30. Wxb6? geb7
31 . Wd4 Af6! his queen would be trapped
(32 . . .Ctif3+ is threatened). Now, however,
Black transposes into a clearly better end
game.
30 . . .Wxb5 31 . cxb5 xa2 32. c8+ f7
33. fd3 J.f6
Black also has a big advantage after
33 . . .Ctixd3 34. Ctixd3 Af6, but Carlsen was
absorbed in constructing a ' box' for the
wh ite king.
34. xe5+
I n the event of 34. Axe5 dxe5 35. gc6 ttig5
White has no defence against the threat of
. . .Ctih3+ and . . . e5-e4.
34 . . . dxe5
' 3 4 . . . A x e 5 was p ro b a b l y bette r, but
i n t i me-trouble I was rel u ctant to allow
35. Axe5 dxe5 36. Ctid3 , which however is
wel l met by 36 . . . f6 37. gfB+ g7 38. gda
(38. gb8 gd7) 38 . . .Ctic3 39. gf1 gd2! and
wins.' (Carlsen)
35. !lc2 !lea7 36. g2
36. f1 is stronger, since White cannot get
by without this move.
36. . . g5 37. d6?

161
 Chapter 3 Life at the Top

Belated activity. W h ite should have re-
stricted himself to 37. dd2, but all the
same the ' box' is slamming shut and the
game cannot be saved.
37.. e4! 38. Axf6 Wxf6 39. Wf1
If 39. xb6, then 39 . . . a1 40. e2 f3 is
decisive.
Loss of material cannot be avoided .
40. c.t?e2
If 40. cd2 Black wins by 40 . . . b1 41 . d1
xb4, when 42. xb6 is not possible on
account of 42 . . . b2 with the irresistible
th reat of . . . h3.
40 J:ib1 41 . :i:id1 :i:ixb4 42. g2 :xb5
.

39 .. a1 43. f4 :i:ic5 44. :i:ib2 b5 45. Wf1 :i:ic7

46. :bb1
a b c d e f 9 h
If 46. g2 Black exchanges a pair of rooks
8 8
by 46 . . . c1 47. xc1 xc1 48. xb5 f3
7 .! J t 7 49. e2 e1 50. b2 g5 and puts White
6 .s: 6 in zugzwang , after which he wins with the
5 5 pawn offensive . . . e6-e5 and . . . h7-h5-h4.
4 l 4 46 :i:ib7 47. :i:ib4 :l:ic4 48. :i:ib2 b4

3 3 49. :i:idb1 t3 50. c.t?g2 :i:id 7 !

2 .s: 2 Now White can not avoid t h e exchange of
1 ,! 1 rooks followed by zugzwang.
a b c d e f 9 h
51 . h3 e5 52. e2 :d2 53. hxg4 fxg4
Now all the wh ite pieces are tied by the 54. :xd2 xd2 55. :b2 f3 56. Wf1
threat of doubling rooks on the 1 st ran k. b3 57. g2 :i:ic2
On the other hand , . . . f3 is th reatened . Wh ite resigned .

Carlsen arrived at the main super-tournament of the year, Morel ia/Linares (FI DE cat
egory 2 1 ) as the victor from Wij k aan Zee. But he realised that to achieve such a suc
cess here would be more d ifficult: the evenly strong field did not permit any relaxation
- there were no outsiders. In Wij k aan Zee there were, after all, a few players against
whom you could hope to win or at least have a respite. But here Magnus did not relax
and throughout the tournament he sat on the 'wheel ' of the leader Anand.
Curious in this respect is the pen-portrait of one of the grandees of Russian chess
journalism, Yuri Vasil iev: ' Mag nus, this l ittle mongoose, rising sharply and swiftly over
the board , when he needs to grasp the nape of another cobra, is a new su per-hero!
To win alone, without a trainer, the tournament in Wijk aan Zee, and then to take
second place in "double-headed" Linares?! And all this at the boyish age of 1 7! ' In
Linares Carlsen's play was dynamic and uncompromising, and, as in Wij k aan Zee, he
scored five wins. Moreover, his contri bution to the eight decisive games considerably
influenced the statistics of the tournament, in which the proportion of draws slightly
exceeded 50%. The one to suffer most at the hands of the 'mongoose' was Topalov,
who lost both games to him.

162
 V. Topalov - M. Carlsen
Game35
V. Topalov M. Carlsen
- 1 0. l:!e1 0-0 1 1 . Wg3 tbxe5 1 2. dxe5 tbxc3
Morelia/Li nares 2008
Alekhine Defence [ 804]
1. e4 f6
Carlsen was so confident in his powers,
that he decided to surprise his opponent
with the Alekhine Defence, an opening from
his child hood. N owadays he employs it
only in bl itz games, and he was success
ful with it at the 2007 World Champion
ship in Moscow. Nevertheless, in su per
tournaments it is not customary to play
such open ings. How great is the degree
of risk? I remember a conversation with
Lev Polugayevsky, one of the strongest
Soviet grandmasters. Commenting on the
fascination for this opening of Vladimir Ba
girov, who wrote a book about it, he was
categorical : 'An incorrect opening! I don't
have the time, otherwise I would refute it'.
Knowing Polugayevsky's character and
his analytical powers, one can have no
dou bts about this.
2. e5 d5 3. d4 d6 4. f3
For a long time Wh ite tried to refute the
Alekhine Defence with 4. c4, but nowadays
he does not hurry with this move, preferring
to complete his development.
4 ... dxe5 5. xeS c6
Preparing the exchange of the active knight,
since if 5 . . . tbd7 Black has to reckon with
6. tbxf7 c;!;>xf7 7. Wh5+ c;!;>e6, when White
can choose between 8. c4 and 8. g3 with
an attack on the king.
At a tender age Magnus preferred 5 . . .
g6, as, for example, against lan Nepom
niachtchi in the World Youth Champion
ship (Heraklion 2002). The game continued
6. Ac4 c6 7. tbc3 Ae6 8. 0-0 tbd7 9. Wf3 Ag7
1 3. Wxc3 Axc4 1 4 . Wxc4 Wd5 1 5. We2
!!adS, and Black maintained the balance.
6. .ld3
At the World Bl itz Championsh i p (Mos
cow 2007) other continuations were tried
against Carlsen:
S h i rov - Carlsen: 6. Ac4 tbd7 7. tbf3
tb7f6 8. h3 Af5 9. 0-0 e6 1 0. Ag5 Ae7
1 1 . tbbd2 h6 1 2 . Axf6 tbxf6 1 3 . c3 0-0
14. Ab3 c5 1 5. Ac2 Axc2 1 6. Wxc2 l:!c8
1 7. Wb3 We? with an equal game;
Grischuk- Carlsen: 6. Ae2 Af5 7. 0-0
tbd7 8. tbg4 e6 9 . c4 tb5f6 1 0. tbe3 Ag6
1 1 . tbc3 Ad6 12. b3 Wc7 13. h3 0-0 14. Ab2
!!adS with complicated play.
6 . . . d7
a b c d e f 9 h
8 i. .i. 'tW .i. i. 8
7 7
6 6
5 tt:J 5
4 8 4
3 3
2 888 888 2
1 :s: tt:J :s: 1
a b c d e f 9 h
7. xd7
Five years earl ier, p l ay i n g against the
1 2-year-old Carlsen (Rethymnon 2003),
Emil Sutovsky did not bother to argue and
he agreed to an endgame with 7. 0-0 tbxe5
8. dxe5 tbb4 9. Ae4 Wxd1 1 0. l:!xd1 f5 1 1 . a3
tba6 1 2 . Af3 g6 1 3 . tbd2 Ae6 14. Ae2 tbc7
1 5. tbf3 Ad5 1 6. tbd4 Ag7 1 7. c4 Af7 1 8 . f4,
in which White had the better chances.
 C hapter 3
---------------------
At the M oscow B l itz tournament M i
chael Adams retreated h i s knight - 7. !Df3,
but also without success: 7. . .!D7f6 8. h3
!Db4! 9. Ac4 AfS 1 0. !Da3 (1 0 . Ab3 aS!=)
1 0 . . . e6 1 1 . c3 !DbdS 1 2. !Dc2 Ae7 1 3 . !De3
Ag6 1 4 . !DeS 0-0 1 S. 0-0 !Dxe3 1 6. Axe3
!DdS 1 7. Ad2 cS. The g a m e is equal.
7 .lxd7 8. 0-0 g6 9. d2
9. lae1 Ag7 1 0. c3 0-0 1 1 . AgSiooks pref
erable.
9 . . .tg7 1 0. f3 0-0 1 1 . Iie1
.
A sound move, made on general grounds,
but for g ai n i n g an advantage gen eral
grounds alone are insufficient, as Carlsen
e m p h asises with h i s s u bseq uent p l ay.
M eanwhile, the presence of the bishop
on g7 demands that attention be paid
to the d4- pawn, which could have been
rei nforced by 1 1 . c3 or at least the pin
on the knight prevented - 1 1 . h3. I n the
game Kulaots -Heim (Gausdal 2003) White
solved his development problems in origi
nal fashion: 1 1 . Wd2!? aS 1 2. a3 a4 1 3 . WgS
lac8 1 4. lad1 bS 1 S. Wh4 e6 1 6. AgS. It is
probable that Black's play here can be
improved , but the main thing is the innova
tive thinking.
11 . . . .lg4 1 2 . c3 c5!
a b c d e f 9 h
'if .. . 8 (Black is better after 1 6. c4?! WaS 1 7. Ag5
.t..t..i..t.7
6 .t. 6 WxdS 1 8. WxdS exdS when, compared with
5 5
4 4
3 3
a b c d e f 9 h
164
Life at the Top
--- ------------------
I n contrast to Topalov, Carlsen's play is
specific and pu rposefu l . With this blow
at the centre he casts doubts on the op
ponent 's opening conception.
1 3 . .le4?!
After the capture of the pawn 1 3. dxcS there
is the unpleasant 1 3 . . .!Dxc3! 14. bxc3 Axc3
1 S . Ah6 Axe1 1 6. Wxe1 Wxd3 1 7. !DeS
Wd4!? 1 8. c6 (the primitive 1 8. Axf8 xf8
1 9. !Dxg4 Wxg4 1eaves White a pawn down)
1 8 . . . bxc6 1 9. !Dxc6, and here there is a
choice: 1 9 . . . Wd6 20. !Dxe7+ h8 21 . Axf8
laxf8 22. Wc3+ f6 23. lae1 Ae6 24. We3
Af7 when the pawn capture 2S. Wxa7
is risky because of the p i n 2S . . . lae8, or
1 9 . . . Wf6 20. !D x e7+ h 8 2 1 . !DdS Wd4
22. Axf8 laxf8. In both cases Black has
a good game. But the decision to go in
for a position with an isolated d4-pawn is
also il logical: with the black pieces being
very active, this pawn comes under attack.
1 3 . Ae2 came into consideration, pro
voking 13 . . . Axf3 (1 3 . . . cxd4 14. !Dxd4 Axe2
1 S. !Dxe2 1eads to simpl ification favourable
to White) 1 4 . Axf3 cxd4 and at the least
Wh ite can try to ' m uddy the water' - 1 S. c4
!Db6 1 6. Axb7 lab8 1 7. Aa6 Wc7 (1 7. . . e5
1 8. b4) 1 8. AgS eS 1 9. lac1 .
1 3 . . . cxd4 14. cxd4
More cunning was 14. h3 Axf3 1 S. Wxf3 e6
!De3 or 1 7. 1ad1 fS) 1 6. cxd4 Axd4 1 7. Axd5
the position in the game, White has opened
an escape square for his king.
14 . . . e6 1 5. Wb3?!
1 S. h3 Axf3 1 6. Wxf3 was better, transpos
ing i nto the variation considered earlier.
But it was also possible to play 1 6. Axf3
 V. Topalov
'i'b6 (1 6 . . . Wd6!?) 1 7. Axd5 exd5 1 8. Ae3!?
'i'xb2 19. Wd3 Wb6 20. gab1 Was 21 . Wb5
'i'xb5 22. gxb5, regaining the pawn.
15 ... Axf3! 1 6. xf3 xd4 1 7. xd5
White takes play into an endgame, since
if he keeps the queens on with 1 7. Wxb7
the game clearly goes i n Black's favour:
1 7. . . Wa5 1 8 . gd1 gabS 1 9 . Wc6 gfc8
20. Wd6 gda 2 1. Wg3 A x b2 . 17. . . gba
18. Wa6 Wf6 1 9. gb1 gfc8 is also good .
17 .'i'xd5 1 8. 'i'xd5 exd5 1 9. :iad1 J.g7
..
a b c d e 21 . J.g5?!
8
6 6
5 5
4 4
3 3
2 ::, ::,
1 .: ib .S:
a b c d e f 9 h
Wh ite is a pawn dow n , and he must re
gret that h e did not g o in for the varia
tion g iven i n the note to his 9th move .
His d ark- s q u are b i s h o p i s n ot d evel
oped, a n d i n v i ew o f t h e weakness of
his bac k ran k h e is u nable to reg a i n the
pawn: 2 0 . g x d 5 gfd 8 2 1 . gxd8+ g x d 8
2 2 . Ae3 (22 . <!> f 1 gd 1 + 23. <!>e2 gh1 ) 22 . . .
b6 23. gb1 Ax b 2 .
20. <!>f1
20. Af4 gfd8 2 1. gac1 gd7 2 2 . b3 d4
23. gd3 gea 24. *f1 h 6 25. ge1 gxe1+
26. <!>xe1 f5 27. f3 <!>f7 28. h4 <!>e6 29. <!>e2
d5 30. gd1 , blocking the pawn, was more
resilient, although it is hard to believe that
this position can be saved .
165
- M. Carlsen
20 .. Jfd8
a b c d e f 9 h
8
.i. 7
6 6
5 5
4 4
3
2 /j,/j,
1 .: jb]d
a b c d e
Since Black's plans include the . . . d5-d4
advance, Wh ite must base his play on the
blockade of the d-pawn . The move i n the
game not only does not solve this problem,
but also allows Black to carry out the ad
vance with gain of tempo. I n this respect
straightforward development is insufficient:
21 . Af4 gd7 22. gd2 gca 23. gad1 d4, and
Black advances his d-pawn : 24. <!>e2 Af6
25. <!>d3 g5! 26. Ag3 gc3+! 27. <!>e2 (in the
event of 27. bxc3 d xc3+ 28. <!>c2 gxd2+
29. gxd2 cxd2 30. <!>xd2 <!>f8 31 . <!>d3 <!>e7
32. <!>e4 <!>e6 Black has winning chances
in the bishop endi ng) 27. . ,ge7+ 28. <!>f1
gc6, switching the rook to the queenside
(variations by Shi pov).
By all the rules of endgame play, it is
useful to centralise the king - 21 . <!>e2, aim
ing not only for the blockade of the pawn,
but also the exchange of bishops, si nce
in a rook endgame White's chances of a
draw are improved. For example: 21 . . ,gac8
22. <!>d3 gd6 23. Ae3 Axb2 24. gab1 d4
25. Ad2 gb6 26. Aa5 gbs 27. gd2 gxa5
28.gbxb2, transposing i nto a rook ending.
Stronger is 21 . . ,gd7!? 22. <!>d3 gea 23. gb1
gde7 24. Ae3 b6 25. gd2 ge4 26. gc1 d4!
 C hapter 3
Life at the Top

27. Ag5 h6 2S. f3 4e6 29. Af4 g5 30. Ac7
e3+, when Black should win.
21 ... l::ld 7 22. d2 h6 23. J.e3 d4
24. d3
The attem pt to exploit the c-file by 24. c1
proves ineffective: 24 . . . adS 25. Af4 g5
26. Ag3 f5 27. f3 d3!. The subseq uent win
ning plan i s as fol l ows: . . .f5-f4, . . . Ae5,
. . . c7, and after the exchange of a pair of
rooks the king comes i nto play.
The black pieces occupy ideal positions,
controlling all the most im portant l ines on
the battlefield. The technique for convert
ing the advantage is highly instructive.
27. a4 f5 28. b3 ec7 29. J.e1 f7
30. d2 c1 31 . xc1 xc1 32. e2
l::lb 1 33. l::ld 3 e6 34. h4 d5 35. Ad2
e4 36. l::lg3 f4! 37. l::ld 3
37. xg6 d3 mate would have been a pretty
fi nish!

37...J.e5 38. f3+ d5 39. Ae1 .ld6

24 ... c8 25. Ad2?!
40. Ad2 g5 41 . hxg5 hxg5 42. Ae1 g4
25. c1 was m o re res i l ient: 25 . . . ddS! 43. fxg4 e4 44. g5?
(in the event of 25 . . . xc1 + 26. A xc1 f5 And Wh ite resigned, without waiting for
27. *e2 the king blockades the pawn on the 44 . . . xe1 +.
d3-square) 26. cd1 c2 27. 3d2 deS!
(the exchange of rooks 27. . . xd2 is justi
fied after 2S. xd2 f5 29. f3 *f7 30. Af2 Game36
d3! 31 . *e1 b6, when after . . . d5 Black M. Carlsen - L. Aronian
transfers his king to the q ueenside, but Morelia/Li nares 200S
2S. Axd2! with the idea of c1 is stronger, Ruy Lopez [ CSB]
and it is not apparent how to breach White's
defences) 2S. Axd4 Axd4 29. xd4 xb2 1 . e4 e5 2. f3 c6 3. Ab5 a6 4. Aa4
30. 4d2 xd2 31 . xd2 b5 32. *e2 c3 f6 5. 0-0 Ae7 6. l::le1 b5 7. Ab3 0-0
33. d3 c2+ 34. d2 c4! , after which 8. a3
Black takes his king to the centre with real
winning chances.

25... l::lc2 26. l::lb 1 l::le7 !

a b c d e

5 5

4 4
a b c d e f g h
3

Carslen's favou rite way of avo i d i n g the

Marshall Attack.
a b c d e f g h
8...Ac5

166
 M. Carlsen - L. Aronian

For a long t i m e the m a i n conti n u ation 15 . .tg3

was the classical 8 . . . d6 (as Shirov played
agai nst Carlsen in Wij k aan Zee 2007).
But the move in the game is also logical:
Black first brings out his bishop to an ac
tive position.
9. c3 d6
In the event of 9 . . . d5?! 1 0. exd5 tbxd5
1 1 . d4 exd4 1 2. cxd4 Ab6 13. tbc3 Black
has a poor pawn structure on the queen
side.
Now in the event of 1 5. Axg5 hxg5 1 6. tbxg5
the move 1 6 . . . d5! gains in strength .
1 5. . . g4
The e4-pawn is taboo: 1 5 . . . l::! xe4? 1 6 . l::! xe4
tbxe4 1 7. AdS Y!/e8 1 8. Y!/c2 tbxg3 1 9. Y!/g6+
(1 9. Axc6? Af5) 1 9 . . . Cf8 20. Y!/xh6+ g8
(20 . . .'i!>e7 21 . tbc3) 21. Y!/xg5+ f8 22. Y!/xg3
tbxd4 23. tbc3 ! and, despite the material
equality, Black's position looks broken.

10. d4 .tb6 1 1 . h3 1 6. hxg4 .xg4

To mai ntain the pawn tension in the centre
a b c d e f g h
it is important to prevent the pin on the
knight at f3 . a J. 'if .! 8
11 . . . lire8!?
7 7
6 .t 6
Aronian's refi nement compared with the 5 5
game Carlsen - Svidler (Wijk aan Zee 2007).
4 j_ 4
Black does not consider it necessary to
spend time on . . . h7-h6. That game de 3 tb 3
veloped as follows: 1 1 . Ae3 h6 1 2. h3 l::!e 8 2 2
1 3 . tbbd2 Ab7 1 4 . d xe5 d xe5 1 5. Axb6 1 .: tLJ .: <it>
cxb6 1 6 . tbh2 Y!/e7 1 7. tbg4 l::! a d8 with a b c d e f g h

roughly equal chances.

12 . .tg5
With the bishop cut off from its wing by the
d6-pawn , this pin may cause Black prob
lems. He is forced to concede the centre.
12 . . . h6 13 . .th4 exd4
13 . . . g5 is risky in view of 1 4. Axg5 hxg5
1 5 . tbxg5 l::!f 8 (1 5 . . . d5?! 1 6 . Y!/f3 exd4
1 7. Axd5 tbe5 1 8 . Y!/g 3) 1 6 . l::! e 3 Y!/e7
1 7. l::!g 3.
14. cxd4 g5
After q u iet d evelopment with 1 4 . . . Ab7
1 5 . tbc3 Wh ite has a small but enduring
spatial advantage.
1 7. .th4!
White has lost the opening battle, and with
his knight undeveloped he is unable to hold
the d4-pawn. Therefore Carlsen renews the
pin, not hesitating to sacrifice the pawn.
1 7. . . txd4 1 8. c3 Axf3
It is hard to condemn this move, made after
a long think. Aronian opens the g-fi le for an
attack on the king, but carries out the idea
incorrectly. He should have taken control
of the d5-sq uare - 1 8 . . . c6! , and if Wh ite
nevertheless insists on pinning the knight
- 1 9. e5?! dxe5 20. tbe4, then his king is as
sailed by the powerfu l group of black piec-

167:
 C hapter 3 Life at the Top
--------

es concentrated i n the centre: 20 ...xe4! 20. d5 g8+ 21 . f1

21 . Axd8 l;!axd8 22. l;!xe4 (the queen has
to be retu rned , si nce neither 22. A xf7+ Of course, 2 1. h1? wou l d h ave been
xf7 23. g5+ xg5 24. Wxg4 gf3+! nor crazy - the lone king simply cannot hold
22. Vi'b1 Axf3 will do) 22 . . .xf3+ 23. gxf3 out agai n st the black army: 21 . . . g4!
l;!xd1+ 24. Axd1 Af5 25. J;!e2 Ad4, reach 22. Vi'xd4+ Axd4 23. Axd8 xf2+ 24. h2
ing an endgame with real winning chances l;!axd8 with the terri ble threat of . . . l;!g5.
for Black. However, White is not obliged
21 . . . g4 22. Wxd4+!
to throw caution to the winds and he can
switch to defence - 19. Aa2 g7 20. Vi'd3. Here is the d ifference compared with
19 . . . h7!? - the exchange of queens takes
1 9. gxf3
place with check, and the assessment of
a b c d e f 9 h the position changes rad ically.
a,! 'i.1 8 22 . . . .lxd4 23. Axd8 h2+ 24. e2
7 I. I. 7 axd8
6 I. .i. I. I. 6
5 I. 5 a b c d e

8
4 it 4
3 itti:J 3 7 7
2 2 I. 6
1 1::, 5 5
'iVl:.
a b c d e f 9 h 4 4
3
1 9 . . . h8? 2
Here the variation 19 . . . c6 20. e5 d xe5
a b c d e f 9 h
21 . e4 can no longer satisfy Black, while
the attempt to i nvade with the q ueen by
25. J;!ad1 !
20 . . . Yi'c8 21. A x f6 Vi' h 3 22. J;!e3 x b 3
23. J;!e4 does not promise anything real ,
The knight on h2 is doomed , but the bish
apart from perpetual check after 23. . Vi'g3+
op's position must be clarified , since if
24. h1 Yi'h3+. 25. J;!h1 there follows 25 . . . l;!g2.
However, the idea carried out i n the
game would also have been practicable, 25 . . . Axb2
but with appropriate refinement: 19 . . . h7!?
20. d5 l;!g8+ 21. f1 g4! 22. Vi'xd4 Now Wh ite wins a piece, practically with
(2 2 . A x d 8 ? ? h2 mate) 2 2 . . . Vi' x h 4 out compensation, but also after 25 . . . c5
2 3 . xb6 cxb6 24. fxg4 (or 24. e2 e5 26. l;!h1 l;!g2 27. l;!xd4! cxd4 28. f4 J;!g7
25. J;!h1 Vi'f6 26. J;!h3) 24 . . . Vi'h3+ 25. e2 29. l;!xh2 Black is doomed .
Vi'xb3 26. f3 J;!ac8 27. J;!ac1 J;!c2+ 28. l;!xc2
Wxc2+ 29. e3 and the storm d ies down. 26. h1 c6

168
------
M. Carlsen - V. Topalov
-----
tiJ

26 . . . Ae5 27. f4 or 26 . . . f!g2 27. !De3 was Game37

no better. M. Carlsen V. Topalov
-

27. f4 Ae5 28. d3 xf3 29. xf3
Ag7 30. !ah5!
Playing to restrict the opponent's mobility.
Black has three pawns for a knight, but the
white pieces are so active that he has no
hopes of saving the game.
30 . . . d5
If 30 . . . f!d7 there follows 31. !DeS, while
after 30 . . . c5 White captures the pawn -
31. Axf7 f!gf8 32. f!f5 f!d7 33. Ae6.
31 . exd5 Jad6 32. Jaf5 cxd5
Morelia/Linares 2008
English Opening [A28]
1 . c4
A rare move for Carlsen, with the psycho
logical idea of forcing the opponent to play
without home preparation. But Topalov is
a tough customer!
1 . . . e5 2. c3 f6 3. f3 c6 4. d3!?
Provoking the opponent i nto a Sicil ian De
fence with colours reversed. More usually
White fianchettoes his bishop - 4. g3, keep
ing the advance of the d-pawn in reserve.

a b c d e f 9 h 4 . d5
8 .. 8 Topalov accepts the challenge.
7 .i. 7
5. cxd5 xd5 6. e4 b6
6
5 The alternative is 6 . . .!Df6 7. Ae2. The game
4 4 Morozevich - G rischuk (World Champion
s h i p , M exico 2007) conti n u ed 7 . . . Ag4
3
8. 0-0 A x f3 9 . A x f3 !Dd4 10 . g 3 Ac5
2
1 1 . Ag2 0-0 12. !Da4 Ab6 1 3 . Ae3 Wd6
14. f4 !Dd7 1 5. c;t.h1 f!ad8 16. f!c1 with com
a b c d e f 9 h
plicated play.

33. Jac1 ! 7. Ae2 Ae7

Excellent technique for converting the ad a b c d e f 9 h

vantage - the pawns will not run away, but
8 I. .i. 'ii' e
it is always useful to activate a piece.
7 .i. .
33 Jaf6 34. Ja xf6 .txf6 35. Jac6 g7
..
6 1.11.1 6
36. f4 Ag5 37. h5+ h8 38. Jaxa6 5 5
d4 39. e4 Jag6 40. Jaa7
4 8 4
Black resigned . 3 ti:J 8 ti:J 3
2 88 888 2
i, VJH M 1
a b c d e f 9 h

69
 C hapter 3 Life at the Top

A position, typical of the Boleslavsky Vari
ation with colours reversed , has arisen. It
should be remembered that in Sicilian-type
positions the value of a tempo is qu ite high.
8. 0-0 0-0 9. a4
I n fu l l accordance with the i deas of the
afore-mentioned variation - to d rive the
knight from b6.
For example, M i hail Marin considers the
following variation: 12. b4!? et:lxb4 (1 2 . . .et:lb3
1 3 . b1 ttlbd4 1 4 . bS tt:lxf3+ 1 S . Axf3 ttld4
1 6. Ag4 b6 1 7. Axe6 fxe6 is also possible)
1 3. et:lxeS f6 1 4 . b1 fxeS 1 S. xb4 b6, in
both cases with a sound position for Black.
11 exd4 12. tlbxd4 tlbxd4 1 3. 'Wxd4 c6
a b c

9 Ae6!? a .!. 8
9 . . . a5 suggests itself, but Topalov ap
6 6
parently did not want to concede the bS
point to the knight, which would assist the
5 5
d3-d4 advance. For example, the game 4 ::.. fli ::.. 4
Agdestein - lvanchuk (Ti l burg 1 993) con 3 t2J 3
tin ued 1 0. Ae3 fS 1 1 . c1 f4 1 2 . Ad2 Ab4 2 ::.. ::.. ::.. ::.. 2
1 3 . et:lbS Axd2 14. et:lxd2 <;!i>h8 1S. h3 Ae6 1 : :
1 6. tt:lf3 Ag8, and here 1 7. d4! is good . a b c d e f g h

1 0. Ae3
The threat is stronger than its execution!
For the moment a4-aS is kept in reserve,
espec ially since the d i rect 1 0 . aS et:ld7
1 1 . d4 exd4 1 2 . et:lxd4 et:lxd4 1 3 . Wxd4
et:lcS, em phasising the weakness of the
b3-sq uare, leads to a position favourable
for Black.
Black has safely got through the open ing.
Wh ite's slight su periority in the centre is
of a temporary nature, whereas Black's
pawn majority on the q ueenside is a real
positional factor. If he should succeed in
playi ng . . . Wa5 and . . . d8, Black will be
able to seize the initiative.
14. a5 tlbc5

This offer to go into an endgame is based

The knight is withd rawn not only for the on positional factors: apart from his pawn
sake of pro p hylaxis against a4-aS, but majority on the queenside, Black can in
also with the aim of transferring it to b3. vade at the weak b3-point.

1 1 . d4 1 5. 'We5

This is the sign ificance of an extra tem po
in a Sicil ian structure. At one time, when
the Sicil ian was just beginning to become
established , the battle for equality was as
sociated with the preparation of the freeing
advance . . . d6-dS. There was no real point
in insisting on 1 1 . aS in view of 1 1 . . . et:lcS.
Magnus is happy to play an endgame, but
in the given specific situation he prefers to
retain his q ueen for an attack on the king.
15 tlbb3 1 6. Jaa4
A not altogether su itable post for the rook,
but otherwise the aS-pawn cannot be de-

170
 M. Carlsen - V. Topalov
-------

fended . After the i nterposition 1 6 . gfd 1 Axf6 24. exf6 and mate next move, and
Black forces the exchange of q ueens by Black also fails to save the game with 20 . . .
1 6 . . . 'llr b 8, transposing i nto a favou rable f6 21 . exf6 Axf6 22. tDxf6+ gxf6 23. AgS.
endgame. 19 .lgS?!

16 . . . .ld6 1 7. YlrhS This inconsistency is not typical of Carlsen.

True, after 1 9. h4 Ag7 20. Wf4 the offensive
a b c d e f g h would have come to a standstill, since he
a .! "if .l e 8 would have had to reconcile himself to the
7 .l .l .l .l .l 7 exchange of q ueens: 20 . . . 'llr b 8 21 . Yltxb8
6 .l .i. .i. 6 gfx b8 22. gd1 gda with good play for
Black. But, apparently, there was no longer
anything better - the aS-pawn is too weak.
For example, if 1 9. gd1 there can follow
1 9 . . . 'llrc7 20. f4 Ag7 21 . 'llrg S h6 22. 'lltg 3
tDxaS, wi nning the pawn.
1 9 Wc7 20 .le3

a b c d e f g h
The abil ity to ad mit your m i stakes is an
important qual ity for a professional. Carls
17 g6
. . .

en returns his bishop to the same sq uare

A cool-headed defence. The threat of e4- and , although he loses a pawn, as if noth
eS with the switching of the rook to h4 is in ing has happened he continues looking
the air, but Black has time to transfer his for chances on the kingside. M eanwhile,
bishop to g7, cementing together his cas 20. Wh4!? tD x aS 2 1 . f4 Ag7 22. fS also
tled position. The other way of defending came into consideration .
involves 17 . . . 'llrc7. However, it is not easy 20 xaS 21 . f4 .lg7 22. Wh4 .lb3
to work out the consequences of the at
22 . . . bS! was more vigorous, since now
tack after 1 8. eS!? AxeS 1 9. f4 (the frontal
23. gd4 would not have been such a pro
attack 1 9. gh4 h6 20. Axh6 gxh6 21 . 'llt x h6
d u ctive idea: 23 . . . ttJc4 24. Axc4 A xc4
is parried by 21 . . . gfe8 22. ghs Ag7 23. ggs
2S. gfd1 Axd4 26. Axd4 fS.
WeS!). For example: 1 9 . . . Af6 20. tDe4 We?
21 . fS AdS 22. tDxf6+ 'llt x f6 23. gh4 h 6 a b c d e t g h

24. gg4 etc. 1 9 . . . Axc3 20. bxc3 gfe8 21 . fS! . . 8

WeS 22. fxe6 Yltxe3+ 23. h1 fS 24. gh4 .l .i. .l 7

h6 2S. Ac4 is also unclear. 6 .l 6

18. 'llr h 6 .les

White's threats are vividly illustrated by the

variation 18 . . .tDxaS? 1 9. eS! Ae7 20. ttJe4.
Now 20 . . . tDb3 is m et by a combination
on the theme of dive rting the q ueen -
21 . Ab6! axb6 22. gxa8 Yltxa8 23. ttJf6+ a b c d e f g h

171
 C hapter 3
------
23. Jad4!
An excellent m ove in the psychological
sense. One can understand Topalov's feel
ings. It is not easy for an attack-mi nded
player to decide on the win of the exchange
by 23 . . . Axd4 24. Axd4, opening the flood
gates in the d i rection of his king.
23 . . . Jaad8?!
A human is afraid of things that a com
puter works through without hesitation:
23 . . . Axd4 24. A xd 4 fS 2 S . e x fS d 8 !
2 6 . Ylff2 flxfS 27. Ag4 flf7 a n d Rybka does
not see any d i rect contin uation of the at-
tack. It parries the obvious 28. tile4 with
28 . . . fl xf4! 29. ttlf6+ xf6 ! .
24. e5!
Here it is, the triumph of spirit over matter!
Wh ite's attack acquires a fresh impetus.
The way is opened for the knight to f6.
24 . . . xd4 25. Axd4 cS
There was also a strategic solution to the
position: 2S . . . fld8 26. f2 b6 27. tile4
cS with the switching of the knight to d4:
28. Ac3 tilc6.
26. Ae3 f6
The desire to activate the g7-bishop is un
derstandable, but 26 . . . b6!? 27. tile4 Ac4
28. Ylff2 fld8 was also not bad , with the
idea of playing the knight to d4.
27. bS 'Wd8 28. fS!?
The pawn makes a breakthrough, although
no light is apparent at the end of the tunnel.
Since if 28. AxeS Black had prepared 28 . . .
fxeS ! , White could have prepared it by first
playing 28. e6!? Axe6 (if 28 . . . e7 White
attacks by 29. fS gxfS 30. J;if3 Axe6 31 . flh3
172
Life at the Top
h6 32. flg3 h7 33. hS with the threat
of 34. Axh6 Axh6 3S. flg6), and now after
29. AxeS fS 30. f2 flf7 31 . Axa7 White
regains the pawn, since 31 . . . Axb2?! is dan
gerous because of 32. Ab6 d2 33. fld1
b4 34. fld8+ g7 3S. Acs a4 36. flaB.
28 . . . fxe5
The capture 28 . . . gxfS 29. flxfS would sig
nify a lack of ambition.
29. AgS b6 30. f6 C4+ 31 . h1
An auto m at i c m ove i n t i m e -t ro u ble.
31 . J;if2!? Axf6 32. Axf6 'tlfxbS 33. AhS! was
stronger, when the bishops gain scope:
33 . . .'tlfe8 34. Af3 'tlfe6 3S. Ae4.
31 . . .'Wxb5 32. fxg7 Jaxf1 + 33. J.xf1
a b c d e f g h
8 8
:::. 1. 7
6 I. 6
5
it
4 4
3 3
2 !:::. !:::. 2
it <i!t 1
a b c d e f g h
33 . . . xg7?!
In time-trouble Topalov decides against
leaving his king in front of the ' m i ne' on
g7. Even so, the pawn could have waited .
Black should have consolidated his forces,
since the aS-knight is clearly out of play.
True, he has to return one of the pawns
and the position is not completely clear.
Here are some variations:
a) 33 . . . dS 34. Ah6 tilc6 3S. Ae2 f7
36. g4 fS (36 . . .tild8 37. 'tlfc8 'tlfe7
 M. Carlsen - V. Topalov
------

38. Af3) 37. A xc4+ Axc4 38. 'W xc4+
'Wf7 39. 'Wd3 ! (39 . 'Wc1 tbd8) 39 . . . e4
40. 'Wd1 and then h2-h3;
b) 3 3 . . . 'Wd7 34. A h 6 tbc6 3 5 . Ae2 ! ?
(35. Axc4+ Axc4 36. 'Wxc4+ Wf7 37. Wc1
tbd8) 35 . . .'Wf5 36. h3 e4 37. Ag4 'Wf1 +
38. h2 'Wf7 39. Ad?! 'Wxd7 40. 'Wf6,
and the threat of mate on f8 ties Black
down (4o . . . 'Wc?+ 41 . h1 'Wf? 42. 'Wd6
We? 43. 'Wf4 'Wf7).
34. AdS! c6??
A blunder in time-trouble. Now Black gets
mated. He could h ave held the position
by 34 . . . 'Wd5 35. 'We?+ g8 or 34 . . . g8
35. 'We? Wd5 , when Wh ite has noth ing
more than perpetual check.
35. Wf6+ mga 36. We6+ mta 37. .lg5
Black resigned .

After his successes in Wijk aan Zee and Morelia/Linares, the appearance of Carlsen's
name among the prize-win ners of the strongest tournaments was no longer regarded
as a sensation. He also enhanced his reputation at the el ite ' M elody Amber' Tourna
ment, which had changed its constant venue from Monte Carlo to N ice. Magnus shared
2nd-5th places, performing equally sol idly in both disci plines (blindfold and rapid).
At the first tournament of the Grand Prix series in Baku (FI D E category 1 9) Carlsen
started as the rating favourite. Therefore it was somewhat unusual to see him keeping
in the background and playing second fidd le. But M agnus came to the fore in the last
two rounds - he won against Adams and Bacrot and shared 1 st-3rd places. True, few
realised that the turning point in his play had occurred back in the 7th round, when in his
game with Radjabov Carlsen took the risk with Black of going into the Dragon Variation
of the Sicilian Defence. This was not a one-off action, but a change of conception. The
hitherto typically White-biased player began also playing for a win with Black. Magnus,
as usual , was modest: 'I simply looked at the main li nes in the Dragon Variation and I
couldn't understand where Black had problems. The positions appealed to me, and I
decided to include this opening in my repertoire.'
The success of the Aeroflot Open festival in Moscow drew the attention of U krainian
patrons. Starting in 2006 they began organising prestigious all-play-all tournaments
in Foros in the Crimea. The resort area, which in not so distant times used to greet
the leaders of the USSR, and then by inheritance those of the U krai ne, was selected
by the Aerosvit Company, presenting an opportun ity for U krai nian professional chess
players to earn some money. In 2008 the organisers were also able to interest Carlsen,
by assembl ing a FIDE category 19 field. The outcome of the tournament was effectively
decided in the very first round, when Magnus confidently outplayed lvanchuk and took
the lead , not allowing his rivals the slig htest chance. Clear first place with a score of +5
in 1 1 rounds cannot be called anything but a triumph. On the cu rrent rating list Magnus
moved up to second place and closed right u p on Anand.
 C hapter 3
Game38------
T. Radjabov - M. Carlsen
Grand Prix, Baku 2008
Sicilian Defence [ 878]
1 . e4 c5 2. f3 d6 3. d4 cxd4 4. xd4
f6 5. c3 g6 6 .le3 .lg7
So, the Dragon Variation, a never-end ing
dispute in which players of all times and na
tions have engaged ! However many times
it has been buried , it has always climbed
back onto its feet. Not so long ago (after the
millennium), one of the authors of this book,
in collaboration with the u nforgettable Ed
uard Gufeld, whose tremendous love for
the g7-bishop was legendary, had to raise
the entire Dragon in the process of writing
a book about this open ing. This was by no
means a labour of Sisyphus, and rumours
about the tam ing of the Dragon proved
exaggerated . And the fact that the bright
est representatives of the chess young are
prepared to demonstrate this (Radjabov
also happily plays this variation with Black)
is the best confirmation.
7. f3 c6 8. 'Wd2 0-0 9. J.c4
The main continuation in the Rauzer At
tack, where the most recent research has
been made. Wh ite develops his bishop in
an active position, preventing . . . d6-dS,
which is possible after 9. 0-0-0.
9 . . . .ld7
A flexible move, leaving Black a wide range
of options. In particular, he can carry out
one of the main cou nterplay resources -
the . . . tilc6-eS-c4 manoeuvre - without
first developing his queen , thereby saving
a tempo.
10 .lb3
1 74
Life at the Top
This prophylactic bishop retreat is part of
the 'compu lsory prog ram' of this varia
tion . After 1 0. 0-0-0 with 1 0 . . . ftb8!? Black
can transpose into the so-called 'Chinese
Dragon', which Radjabov employed in his
game with Carlsen from the Grand Slam
tournament i n B i l bao 2008. 1 0 . . . flc8 or
10 . . .tileS leads to the main line of the vari
ation.
10 . Etc8
a b c d e t g h
8 .l. 'i .. . 8
7 .\ .t .i. .t .t .i. .\ 7
6 1.&\ .t 1.&\ .t 6
5 5
1 1 . h4
An important branch. Wh ite does not hurry
to castle, but first tests Black's reaction . In
recent times the more restrained develop
ment 1 1 . 0-0-0 tileS 1 2. b1 has been in
fash ion, where Wh ite is not i n a rush to
disclose his attacking plans on the king
side, retaining the option of playing both
h2-h4 and g2-g4. For example: 1 2 . . . tilc4
1 3 . Axc4 flxc4 14. g4 b5 1 5. b3! . 1t is im por
tant to disrupt the coordination of the rooks
- now 1 5 . . . flcS? 1 6. tile6! leads to the loss
of the exchange, and 1 5 . . . flc8 1 6 .tild xb5
'WaS 1 7. a4 is also in Wh ite's favour. But it
turns out that the rook does not have to
move. The game Socko - Stocek (Bu ndes
liga 2006) continued 15 . . . b4! 1 6. bxc4 bxc3
1 7. V!Vxc3 'Wc7 1 8 . gS tilhS 1 9. c1 flc8
20. 'Wd3 AeS with sufficient counterplay
for the sacrificed exchange.
 T. Radjabov - M. Carlsen

Apart from the standard plan , also in and now 1 5. h5 or 1 5. f4 1eads to very sharp
teresting is 1 2 . . . .!:!e8!? (with the idea at an play.
appropriate moment of playing . . . e7-e5), 14 . . . :ia e8!
which Carlsen tried two rounds later in his
A prophylactic move, after which the sac
game with Karjaki n: 13. h4 h5 14. Ah6 &Dc4
rificial attack, possible after 14 . . . b5 1 5. g4
15. Axc4 .!:!xc4 1 6. Axg7 xg7 1 7. &Dd5
hxg4 1 6 . h5 tDxh5 1 7. &Dd5 when Black is
e5!? 1 8. tDxf6 Wxf6 1 9. tDe2 flc6 20. tDc3
forced to play 1 7. . . .!:!e8, is now consid
Ae6 21 . &Dd5 Axd5 22. Wxd5 Wf4! 23. d2
ered insufficient, since in the variation with
Wxd2 24 . .xd2 f5, and Black maintai ned
14 . . . .!:!e8 the rook is already in the right
the balance.
place and Black saves an important tempo.
11 . . . h5! However, even in the first case he has ad
equate defensive resources: 1 8. flxh5 gxh5
For Dragon devotees this is the standard
19. Wh2 flxd5 20. Axd5 b6 21 . tDf5! Axf5
reaction to h2-h4, si nce if 1 1 . . . &De5 Black
22. exf5 gxf3 23. Wxh5 and now instead of
has to reckon with the pawn sacrifice 12. h5
23 . . . flc8 , which has occu rred in practice,
tDxh5 1 3 . g4 tDf6 1 4 . Ah6 Axh6 1 5 . xh6
consideration should be given to 23 . . . Wf2!?
with a dangerous in itiative for Wh ite.
24. flh1 Wg3 with sharp play.
12. 0-0-0 tUe5 1 3. Ag5
a b c d e f g h
8 8
The m ost usual cont i n uatio n . Wh ite at 'i .1.
tacks the knight which is preventing g2-g4, 7 -*- ..t 7
and he restricts Black's possibilities. The 6 6
direct attack 1 3 . Ah6 &Dc4 1 4 . Axc4 .!:!xc4
15. Axg7 xg7 1 6. g4 hxg4 1 7. h5 does
s .!. i, . s
not achieve its aim. For example, the game 4 CD !:::. :::. 4
Sax- Kir. Georgiev (Burgas 1 992) continued 3
i, {jj !:::. 3

1 7. . . flh8 1 8. hxg6 fxg6 1 9. f4 e5 20. fxe5 (or
20. &Dde2 .xh1 21 . .xh1 &Dxe4 22. &Dxe4
.xe4 23. fxe5 .xe2! 24. Wxe2 Wg5+ with
roughly equal chances - Stohl) 20 . . . dxe5
21 . &Dde2 flh3! with double-edged play.
13 . . .:iac5
This move has the aim of prepari ng the
pawn attack . . . b7-b5. I n addition , the rook
takes control of a n u m ber of i m portant
points along the 5th ran k. If 13 . . .&Dc4, then
14. We2 &Da5 1 5 . b1 is good , followed by
an attack on the kingside.
14. b1
A usefu l prophylactic move. 14. g4 hxg4
2 l:::, l:::, l:::, VI !:::. 2
1;l l;l 1
a b c d e f g h
1 5. g4?!
After the game Radjabov ad m itted that
he had m i xed u p two li nes. N ow Wh ite
is forced to th row caution to the winds.
15 . .!:!he1 is regarded as the main continua
tion. But what Wh ite had prepared in reply
to 15 . . .Wa5 remained a mystery, si nce ex
tensive practice has not demonstrated any
thing significant. For example, the game
Beliavsky - Ki r. Georg i ev (Wij k aan Zee
1 985) went 1 6. f4 &Dc4 (after 16 . . . &Deg4?
1 7. e5! dxe5 1 8. Axf7+ xf7 1 9. &Db3 Black
 C hapter 3 Life at the Top

loses material) 1 7. Axc4 xc4 1 8. Axf6 (the
1 8 . eS breakthrough does not achieve its
aim: 1 8 . . .tDdS! 1 9. tDxdS 'WxdS or 1 8. tDb3
'Wa6 1 9. eS tDg4 20. exd6 Axc3 21 . bxc3
e6!) 18 . . . Axf6 1 9. tDdS 'Wxd2 20. tDxf6+
exf6 21 . xd2 fS! with an equal endgame.
1 S. Ah6!? tDc4 1 6. Axc4 xc4 1 7. Axg7
xg7 1 8. tDdS leads to a position which
occurred in the afore-mentioned Karjak
in - Carlsen game.
15 . . . hxg4 1 6. h5
The alternative is to create a breakthrough
in the centre: 1 6. f4 tDc4 1 7. Axc4 xc4
1 8. Axf6 Axf6 1 9. eS Ag7 (weaker is 19 . . .
dxeS?! 20. fxeS AxeS 21 . tDde2 and then
h4-hS) 20. hS dxeS 21 . fxeS Ac6 with un
clear play.
1 6. . . ti)xh5 1 7. xh5
If 1 7. f4 Black has several possible con
tinuations: 17 . . . tDc4 1 8. 'Wd3 bS, 17 . . .tDc6
1 8 . eS 'Wc8 or 17 . . .tDf3 1 8 . tD xf3 Axc3
1 9. bxc3 g xf3.
After 1 7. tDdS Mikhail Golubev considers
the best to be 1 7. . . xdS! 1 8. AxdS 'Wb6
with dynamic play.
1 7 . . . gxh5 1 8. Wh2
a b c d e f g h
8 'iV X 8
note to 1 4 . . . e8, the white knight is at c3,
and Black has a wide choice of defensive
measures.
1 8 . . . ti)g6!?
Now if 1 9. tDdS there is 1 9 . . . h4. In Radja
bov's opinion 1 8 . . . Wb6!? 1 9. WxhS e6 is
also good .
1 9. Wxh5 Was
Black is obliged to reckon with the threat
ened capture on g6 (for the moment this is
unfavourable - 20. 'Wxg6? e6), and Carlsen
decides to return the exchange, but with a
more active position for his queen . Before
this 1 9 . . . xgS 20. WxgS e6 21 . Wxg4 Wf6
had been played .
The immediate 1 9 . . . e6!? is also interest
ing: 20. tDfS! 'WaS! 21 . h1 (21 . xd6 Ac6)
21 . . . xfS 22. exfS WxfS 23. fxg4 'WeS with
a sound position for Black (Stoh l).
20. f4!
If 20. tDdS there can follow 20 . . . e6 21 . h1
e x d S 2 2 . Wh7+ f8 2 3 . A h 6 A x h 6
24. 'Wxh6+ e7 2S. 'WgS+ f6 26. 'Wxg6
dxe4, and Wh ite does not have sufficient
compensation for the exchange.
20 . . . xg5
After 20 . . . xc3 21 . bxc3 'Wxc3 there is the
cou nter-sacrifice 22. Axf7+ xf7 23. fS.
21 . fxg5 e6
a b c d e f g h

... .I .t. .t. .i. 5

7 7
8 8
6 6
7
5 6 6
4
ED 4
5 'if
3 ED 3
4 4
2 WI 2
3 3
1 :a: 1 2
a b c d e f g h

I n contrast to the position exami ned in the a b c d e t g h

176
 T. Radjabov - M. Carlsen
------
22. t'Uf5?
Youth strives only to go forward , but this
leads merely to a red uction in White's at
tacking potential. He should have restricted
himself to 22. 'Wxg4, although, objectively
speaking, the spearhead of his attack has
been considerably blunted .
22 . . . exf5 23. xg6 Ae6 24. h5
The rook endgame arising after the ex-
changes 24. Axe6 xe6 25. xf5 Axc3!
26. bxc3 xf5 27. exf5 e4 is in Black's
favour (2B . xd6? is not possible because
of 2B . . . g3).
24. . . fxe4 25. f1
After 25. et:ld5 Carlsen g ives this varia
tion: 25 . . . e3! 26. g 6 fB 27. g xf7 Axf7
(or 27. . . e7 2B. g5 Axd5 29. f1 ! dB
30. x d 5 e2 3 1 . e1 ) 2 B . f1 'We1 + !
29. xe1 Axh5 3 0 . f1 + Af7 31 . <tlxe3 xe3
32. xf7+ eB with an obvious advantage
to Black, who has an extra pawn which is
also passed .
25 . . .We5!
Black pins his hopes on the passed g -
pawn - 25 . . . Axb3 26. axb3 e7 27. xg4
prom ises less.
26. xf7 Axb3
26 . . . fB!?, simplifying the position, is even
stronger.
27. axb3 g3 28. a2
With a rather transparent trap: 2B . . . g2?
29. f2! g 1 30. f7+ h7 31 . h5+ with
a draw.
28 . . . Etf8 29. xf8+ xf8 30. Wg4 e3
31 . g6
a b c d e f 9 h
8 8
7
.a
6 6
5 5
4 4
3 l:iJ 3
2 2
a b c d e t 9 h
31 . . . e2?
With this i m p u lsive move Carlsen could
have ru i n ed the resu lts of h i s work. H e
should fi rst have secured t h e position of
his king: 31 . . . e7 32. f3 e2! 33. x b7+
(33. xe2 xe2 34. et:lxe2 g2 35. c3 Ah6)
33 . . . f6 34. et:lxe2 xe2 35. f7+ g5
36. xg7 a6+ 37. b1 g2 and the pawn
can not be sto p ped . H ow, h owever, by
playing 32. d7! , Wh ite would have suc
cessfu lly ' terrorised ' the king: 32 . . . gB
(32 . . . e7 33. cB+ WeB 34. f5+ etc.)
3 3 . WeB+ AfB 34. x b7 Ae7 3 5 . 'Wh1
fB 36. h7 eB 37. et:lxe2 xe2
3 B . g7.
32. f3+?
White misses his chance, and the king es
capes from the checks.
32 . . . .l...
-=-e8 33. f7+
White loses after 33. xe2 xe2 34. <tlxe2
g2 followed by the manoeuvre of the bishop
to d4 or e3.
33 . . . d8 34. Wg8+ d7 35. Wf7+ We7
36. Wf5+ d8 37. Wa5+ b6 38. Wd5
e1W 39. Wa8+ d 7 40. Wb7+ e8
White resigned .
177
 C hapter 3 Life at the Top

Game39 sound arguments for 1 0 . . .ttif4. Therefore,

M. Carlsen - V. lvanchu k before launching the attack with c4-c5,
Foros 2008 Wh ite neutralises this possibility: the rook
King's Indian Defence [E97} vacates the f1 -square for the bishop. Mark
Taimanov, the founder of the Classical Vari
This game, which was annotated by earls ation, used to parry the threat of . . .ttih5-f4
en, enables us to u nderstand to a certain with the radical 1 0. g3, but ever since the
extent the logic of the decisions he takes, effectiveness of the counterplay with . . .
based not so much on calculating abil f5-f4 was demonstrated, White has pre
ity (which some experts consider to be ferred to defer this move u ntil better times.
the main reason for his achievements), as
10 . . . f5 1 1 . g5
on his depth of understanding of chess.
This, in our view, is Carlsen's main strength , The knight is aiming for e6, with White dis
placing him on a par with the great chess playing his read iness to sacrifice a pawn
players in the world. for the sake of increasing the activity of
his pieces.
1. d4 f6 2. c4 g6 3. c3 J.g7 4. e4 d6
5. f3 0-0 6. .le2 e5 7. 0-0 c6 8. d5 11 . . . f6
e7 9. b4
a b c d e f 9 h
a b c d e f 9 h
8 1. . . 8
.i. 'it'
8 I .i. 'i . . 8
7 j. j. .i. J. 7
7 j. j. J. .i. J. 7 6 j. 6
6 j. 6
5 j. j. t{) 5
5 5
4 4
4 4
3
I{) 3
3 3
I{) I{) 2 Ji 2
2 Ji 2
l:l
1
l:l
1 l:l Ji l:l a b c d e f 9 h
a b c d e f 9 h

In the Classical Variation this move is the
most popular. This is not surprising. The
characteristics of the pawn structure imply
the c4-c5 attack by White and . . .f7-f5 by
Black. The move 9. b4 enables Wh ite to
get going without beating about the bush.
9 . . . h5 1 0. !!e1
The modern continuation . At one time they
preferred to 'take the bull by the horns' with
1 0. c5, but it was realised that there were
1 2 . f3
'Th is has become the main line after a lot
of van Wely- Radjabov games' (Carlsen).
In this sentence M agnus expressed the
sum total of knowledge, accumulated in
ChessBase - for him the opinion of mod
ern experts is i mportant. But this is also
the fru it of the decades of labour by such
mighty King's Indian players as Yefi m Gel
ler, Eduard Gufeld, Garry Kasparov and
Boris Gelfand - the list could be extended .

178
 M. Carlsen
When conducting plans with an attack
on pawn chains, you r own chain should
be rei nforced . If 1 2. Af3 Black can reply
by undermining the spearhead of the pawn
chain - 1 2 . . . c6!?.
12 . . h8
.
A u sefu l p rophylactic king m ove, i ntro
duced into modern practice by Radjabov.
Black prepares for the possible open ing of
the a2-g8 diagonal in the event of 1 3 . e6
Axe6 1 4 . dxe6, assuming that it is danger
ous for his king to remain on it.
13. b5?!
'For some reason I decided against playing
the main line - with 1 3 . e6 - and came
up with this move instead , which clamps
down on the black queenside. It looked
interesting to me during the game, but now
I don't like it any more: it's probably just
too slow' (Carlsen). A typical admission.
Magnus is not inclined to think dogmati
cally, although he undoubtedly knew the
prospects for White in the event of 13. e6
Axe6 1 4 . dxe6.
Analysis diagram
The dispute i n this position between van
Wely and Radjabov began in 2005 at the
World C u p in Khanty- M ansiysk, when
Teimour tried to pick u p the pawn with his
e7-knight , which is passively placed , by
transferring it to c6 - 1 4 . . . fxe4 1 5. fxe4 c6.
179
- V. lvanchuk
It only remai ned to play . . . Cbd4, but Leek
prevented this with 1 6. Cbd5! (1 6 . . . d4?
1 7. e7) and after t h e forc i n g cont i n ua
tion 16 . . .xe4 1 7. Af3 f6 18. b5 Cbxd5
1 9. bxc6! White retained the initiative. There
followed 1 9 . . .b6? 20. cxb7 b8 21 . c5! e4
22. xe4! dxc5 23. Wxd8 fxd8 24. Ag5
e8 25. d1 Ad4+ 26.exd4! cxd4 27. e7
and the e-pawn decided the game. But
also after the better 19 . . .c3! 20. cxb7 b8
21 . Wc2 We8 22. Ad2 a4 (with the idea
of . . .c5) 23. Ae3 White, with his pair of
passed pawns, has an obvious advantage.
The duel conti n ued in W ij k aan Zee
2007, when Radjabov went for the e6-
paw n with his other k n i g ht : 14 . . . Cb h 5
1 5 . g3 Af6 (with t h e idea o f . . . g7) 1 6 . c 5
f4 1 7. g 4 g7 1 8 . Ac4 c6 1 9. cxd6 cxd6
20. e2 c8 2 1 . AdS xb4 22. b1 c2,
after which Black held the i n itiative and
won the game. At Monaco 2007 i n their
b l i n d fo l d g a m e van Wely began w i t h
1 5 . c5 a n d after 1 5 . . . f4 1 6 . A c 4 fxe4
1 7. xe4 f5 1 8 . Axf4?! exf4 1 9 . Cbd5
e8 Black safely got through the open
ing. And fi nal ly, i n B i e l 20 07, van Wely
corrected Wh ite's p l ay : 1 8 . g 3 ! ? Cbh3+
1 9 . <!>g2 g5 20. g4 xf3 2 1 . <!> xf3 e4+
22. xe4 A x a1 23. Ag5 tbe7+ 24. <!>g2
Ae5 25. f4 Axf4 , when after the fau lty
26. gxf4? with 26 . . . d5 B l ack seized the
i n itiative and went on to win the game.
M eanwhile, 26. Axf4! wou l d h ave g uar
anteed a d raw: 2 6 . . . d 5 2 7. Wa1 + <!> g 8
28. Ah6 Cbf5 29. e 7 Wxe7 30. Axd5+ f7
31 . f6+ <!>h8 32. h5+.
1 3 . . . e8
With the same idea of attacking the e6-
pawn in the future from g7 (after . . . Af6).
Of course, it is also possible to achieve the
same aim after 1 3 . . .h5.
 C hapter 3 Life at the Top
-- --------------

14. A.e3 with opposite-colour bishops is clearly to

I n the event of 1 4 . c5 dxc5 1 5. Ae3 Black
is not obl iged to defend the pawn - 1 5 . . .
b6, since here 1 6. tbe6 Axe6 1 7. d xe6
takes place in a situation favou rable for
Wh ite. Stronger i s 1 5 . . . A h 6 ! 1 6 . tDf7+
xf7 1 7. Axh6, exchanging the ' bad' dark
square bishop (variations by Carlsen).
14 . . . .tf6 1 5. e6 .txe6 1 6. dxe6 g7
1 7. Ah 6
The possession of the two bishops in the
closed type of position after 1 7. tDd5 tbxe6
1 8 . tbxf6 xf6 is not an advantage.
1 7 . . . xe6
Otherwise the knight manoeuvre to g7 does
not make sense. In the event of 1 7. . . tbg8
1 8 . Axg7+ Axg7 1 9. c5! Wh ite succeeds
in defending his e6-pawn .
1 8. Axf8 Wxf8
White's advantage.
1 9. . .1txc5
'After this the k n i g ht will be somewhat
sidelined. 1 9 . . . dxc5 was p robably bet
ter. I intended to conti nue 20. Wd7 Wc8
21 . Wxc8+ xc8 22. Ac4 with an about
equal ending' (Carlsen). A confirmation of
the previous comment. M agnus is objec
tive not only with respect to his pieces. Let
us continue the variation with 22 . . .tbd4, in
order to see the difference in the placing of
the knight: at d4 it is 'eternal', whereas at
c5 it is merely trad itional and has no future.
But let us be honest regard ing the natural
reaction to 1 9. c5 - to whom would it occur
to voluntari ly spoil his pawn structure in a
simi lar situation, if it did not even occur to
the non-trivially thinking lvanchuk?
20 .tc4 .tg5

The only way of activating the ' bad ' bishop.

a b c d e f g h
21 . We2 'Wh6 22. liad1
8 .1 'iV 8
Wh ite has managed to restrain the dark
7
_.t. 7 square bishop, by not allowing it to go to e3
6 .t. .i. .\ 6 or d2. But otherwise White's achievements
5 fj, .t..t..L- 5 are slight: an attack on the queenside for
4 fj, fj, 4 h i m is j u st as u n real istic as one on the
3 ttJ fj, 3
kingside for Black. The position is one of
dynamic balance.
2 fj, fj, fj, 2
22 . . . 1if8
1 : : 1
a b c d e f g h
a b c d e f g h
8 .i 8
1 9. c5! 7 .t..t.f.t.l .t. 7
' I think this is basically the only move, as 6 .t. .t. 'ii' 6
the highest priority is to activate the bish 5 fj, .t..t. .t 5
op' (Carlsen). This com ment shows the
4
fj, 4
essence of M ag nus's concrete thinking.
The main thing for him is the ideal plac
3 ttJ fj, 3

ing of his pieces, for the sake of which he 2 fj, fj, fj, 2

does not begrudge giving u p a pawn. With ld ld

the activation of the bishop, the situation a b c d e f g h

180
 M. Carlsen
23. a4
'I was not quite sure if including a4 and . . .
b6 was a good idea, as the a4-pawn will
require constant attention now, but it has
to be said that . . . b6 is also qu ite a conces
sion here, as it gives up any dream of ever
advancing the pawn centre' (Carlsen). One
cannot help agreeing with the evaluation
of 23 . . . b6, but as regards 23. a4 Magus
is understating someth ing, and he most
probably played t h i s intuitively. This is
something we will be able to ascertai n .
23 . . . b6?!
It was safer to stick to a waiting course,
as in the game - 23 . . . Wh4 24. g3 'Wh3 etc.
24. g3 'Wh3
An unpleasant move, preventing the con
solidating g2.
25. Wg2 Wh6 26. We2 Wh3 27. h1
The avoidance of the repetition has the
aim of dulling the veteran's vigilance. How
otherwise can his reply be explained?
27. . . d7?!
Why not sound out the opponent's inten
tions with 27. . . Wh6 ?
181
- V. l vanchuk
28. !l:a1 !
' I believe that the rook is a little better off
on a2 than on d1 , as it can now support a5,
and also go to c2 in some lines' (Carlsen).
Here it is, that amazing intuition! Magnus
could not have known Eduard G ufeld, and
had probably not read his book o n the
King's I ndian publ ished i n 2002, but he
exactly reproduced an idea of Tigran Pet
rosian in a similar situation with a bishop on
the h6-c1 diagonal, which he carried out in
their game from the USSR Championship
Final in 1 960. Here is the ad m i ring com
mentary by the loser: 'While I was marking
time, Petrosian gave me an object-lesson
on the theme that time in chess, as in life, is
irreversible. His gen ius was revealed in the
fact that he avoided the ' m ined' square c1 (
a2-a4, b2-b3, ga1 -a2-c2, gf1 -b1 -b2) and
succeeded in setting up a decisive bind
on the queenside. Such a plan was hard
to find and highly instructive. " I ron Tigran"
sim ply reduced the board to seven ran ks,
after which it transpired that my bishop
was indeed firing into empty space.' Here
you have the young Norwegian's lack of
'schooling ' ! N ot without reason do they
say that he absorbs ideas like a sponge.
28 . . . Wh6 29. !l:a2 f6?!
 Chapter 3 Life at the Top

a b c d e I 9 h xd5
8 .I 8
a b c d e I 9
7 .t.l l.t. ... 7
8
6 ... ... .t. 'i 6
5 .t..t. .i.. 5
6 6
4 4
5 5
3 fJ 3
4
2 2
3 3
2 2
a b c d e I 9 h

31 . d5! xd5 32. xd5 f4 33. Wf2 a b c d e I 9 h

Was the knight's run worth such effort, if

it can be parried in one move?
33 . . . fxe4 34 . .lxe4 Wg5 35. Jac2 d5?
Black's play has come to a standstill, but
why voluntari ly give up material? 35 . . . f7
was more resil ient, although after 36. Ab7!
with the idea of Ac8-g4 Wh ite neverthe
less will force him to declare his intentions.
39. Jace2!
S i m ple and busi ness- l i ke! The e5-pawn
cannot be held: if 39 . . . f5 there is 40. g4,
while if 39 . . . e8 there follows 40. f4.
39 . . .Wf6 40. Jaxe5 Wxf3 41 . Wxf3
Ja xf3+ 42. 'it?e2 Jaf5 43. Jaxf5 gxf5
44. 'it?d3 c5 45. Jae5 b4+ 46. 'it?d2

36. xd5 xg3 37. hxg3 f4+ 38. 'it?f1 Black resigned .

The uninterru pted series of successes created another wave of i nterest in the rapidly
prog ressing 1 7-year-old N orwegian. Carlsen's name was regularly among the first
prize-winners at such prestigious tournaments as Wijk aan Zee, Linares, Nice, Baku
and Foros - only successes, not a single failure. It appeared that to conquer the sum
mit - a rating of 2800 - only one last effort was needed . Magnus's confidence in his
own potential was impressive: he believed that he was capable of anyth ing, and that the
limit of his possibilities lay far ahead . However, among the chorus of eulog ies, doubts
were also expressed about the qual ity of his play, and it was said that his talent was
only of a calculating nature. Numerous cock-and-bull stories were whipped up, such
as his lack of 'schooling' with a su bstitute knowledge of computer dogmatism, and
the surprise at how Carlsen had contrived , without reading a single classical chess
book, to learn everything.
What can we say? One can merely feel sym pathy for the ill-wishers. As for his love
of chess, this question is indeed one that appeals only to basic instincts. Questions
about the qual ity of his play also sound rhetorical , since the appreciation of Magnus's
masterpieces depends on your own level of chess, and stories about his lack of school-

182
 M. Carlsen - V. lvanchuk

ing had long ago been refuted by his father. His schooling was at the level of a high
quality expert, and prizes at super-tournaments testify to the marks of the examiners.
And besides, can one talk about a lack of school ing, when you have such a mentor as
Simen Agdestein? The famous Russian trainer Yu ri Razuvaev liked to remember the
introd uctory sentence by M i khail Botvinnik at a session of his school: 'I m ust i m med i
ately warn you that you cannot b e taught t o play chess, you c a n o n l y learn ! ' Magnus
had learned . I ndependently. There were countless examples of this.
It is clear that the times of the old chess stars are long si nce gone. Time has
moved forward , and tech nical possibilities have also changed . And whereas i n the last
century new ideas and conceptions were tracked down l iterally piecemeal in chess
period icals, in the com puter age it would be silly not to use databases with m i l l ions
of games, processed accord i ng to certain criteria. ' It is sufficient s i m ply to click on
a few keys,' complains Alexander Khal ifman , whose main successes coincided with
the start of mass com puter usage. ' Based on this, it is completely d ifferent q ual ities
that are called for: calculation, memory and energy. Moreover, the key word here is
memory! What becomes i m portant is not the abi l ity to look at many positions from
various openings, but that all this should be i n your head .' One can argue with this.
From the existing work with juniors it transpires that many talented players possess
such qual ities as a good memory and excellent calculation, but only a few make it
to the top. It is i m portant not only to be able to calculate, but also to know what to
calculate. The main thing here is g rasping the depth of chess, understanding the
essence of a position, enabl ing a n u m ber of variations to be i m mediately discarded
and not considered . It is not possible to calculate everyth i n g . A player's class is
determ ined by his abi l ity to chop off branches of the calculation ' tree' and find the
correct continuation . This comes easily to only a few, among them Carlsen. 'I prob
ably have qu ite a good understanding of chess,' he let slip i n one of his i nterviews .
Th is the where the root of the q uestion l ies.
As regards memory, it is worth dwelling on this factor i n more detai l . H ere Carlsen
is undoubted ly superior to many of his contemporaries. During his apprenticesh ip,
Agdestei n wrote an introductory letter to Garry Kasparov, extracts from which were
published i n the magazi ne 64 Shakhmatnoe obozrenie. This was how he assessed
the potential of his pupil: ' F i rst of all, M agnus has an incred ible, photographic type
of memory, when a page is not read , but i m mediately i m pri nted on the brain and
instantly assi milated . He remem bers not only the game itself, but also where, when
and who played it. He is capable of read ing a book on a sharp open ing d u ring the
cou rse of one evening (usually without using a board and pieces) and of playi ng it
the following day even against a strong g randmaster. This explains why he has such
a broad open ing repertoi re. It often seems to me that he can play everything and
that i n the open ing he knows everything. It is wel l known that i n the Sicilian Defence
there are several incred i bly sharp l i nes, where Wh ite unexpected ly sacrifices a cou
ple of pieces, and it all ends with a better endgame - but suddenly Magnus tel ls
me that all these compl ications are theory, and he knows about them! This is not

183
 C hapter 3 Life at the Top

surprising - he l iterally swallows all chess material.' We should remind the reader
that Alexander Alekhine and M i khail Tal also had such a memory. And if to Carlsen's
phenomenal memory you add this staggering intuition, plus his absol ute lack of fear
of the strongest in the world - two i m portant qual ities remarked on by the famous
trainer Mark Dvoretsky - this says a great deal . M ag n us's potential allows him to
hope for a worthy place at the chess summit.
Carlsen was self-sufficient; he acknowledged the strength of the leaders of the
chess world, but he realised that he could beat any of them . The only one that Magnus
had not yet got his 'teeth' into was the world champion Anand - he was the only one
whom he had not managed to beat (not counting blitz games). Therefore their clash in
Mainz, even in a rapid event, was of particular interest. There was no real battle. After
suffering a fiasco in his favourite 'Dragon' in the first game, Magnus also conducted
the second game weakly. The 1 -3 defeat was undoubted ly a blow to his self-esteem.
However, there was nothing surprising about this, as rapid-play comes naturally to
the quick-acting Anand; here he is the acknowledged leader. As for Carlsen's fai lure,
there is a limit to everything, and a player can only do so much. The continuous bat
tles at the top level, and with the bar raised high, demanded far more expenditure of
energy than earl ier.
Magnus arrived in Germany from Switzerland, where he had just played an intense
tournament, and he looked very tired . It appeared that the battle in Biel did not herald
any sensation. What for Carlsen , a regular partici pant in super-tournaments, was a
FIDE category 1 8 tournament with a far from stellar field of six grandmasters! He set the
pace: he won the fi rst half, and consolidated his success by also winning his second
game agai nst the undoubted outsider Pelletier. But when it appeared that the fate of
the fi rst prize was decided, Magnus was let down by his maximalism: he decided to
win with Wh ite against Alekseev. He failed to finish off a slightly better position, and
the protracted battle ended with the blunder of a pawn on the 75th (!) move and a loss.
This game was the deciding point of Carlsen's performance in Biel, since a desperate
attem pt to halt the leader Dominguez with the help of the risky ' D ragon' almost led to
a second defeat. The outcome was 3rd place with 6 points out of 9. For the time being
the storm ing of the 2800 summit had to be deferred .

Game40 curs quite often. By turning onto a sideline,

V. Pelletier - M. Carlsen Carlsen was hoping to outplay the Swiss
Biel 2008 grandmaster, who was performing badly
Queen's Indian Defence [E15] in the tournament.

1. d4 tbf6 2. tbf3 e6 3. c4 b6 4. g3 J.a6

8. cxd5 exd5 9. tbc3 0-0 1 0. 0-0 Ab7
5. b3 J.b4+ 6. J.d2 Ae7 7 .tg2 d5
1 1 . c1

7. . . c6 is considered the main continuation,

but the move of the central pawn also oc- (see next diagram)

184
 Y. Pelletier - M. Carlsen

a b c d e f g h a b c d e f g h

a K ,.. 'iV K tfr 8 ax 8

'i' .1
7 .t .i. .l J[ .i. .t .t .t
7 7 .t .i. ,.. .i. .t .t .t
7
6 .t ,..6 6 .tj .t 6 ,..
5 .t 5 5 .t 5
4 4 4 ib 4
3 [jj [jj 3 3 [jj [jj 3
2 ib ib 2 2 ib, 2
1 1;1 1;1 1 1;1 1;1
a b c d e f g h a b c d e f g h

1 1 . . . bd7 14. l3cd1?!

In this position the more trad itional way of
developing is considered to be 1 1 . . . tba6
followed by playing the knight to c7 after the
c-pawn has moved . Of course, Wh ite can
counter this. Here the game Aronian - lvan
chuk (Morelia/Linares 2006) is of interest:
1 1 . . .tDa6 1 2 . Af4 l::! e8 1 3 . &DeS h6 14. tDbS!?
cS 1 S . tDc4 (1 S. dxcS bxcS 1 6. tDc4 Ac6
17. tDbd6 is possible) 1S . . . Ac6 1 6. tDcd6
l::!f8 1 7. dxcS bxcS, and after 1 8. d2!? (the
game went 1 8. tDd4 cxd4 1 9. l::! x c6 tDb4
20. tDb7 e8 21 . l::! c7 tDa6=) the hanging
pawns are in need of defence.
But after the classical knight manoeu
vre . . .tDb8-d7-f8-e6 (in the style of the
Carlsbad Variation of the Queen's Gambit)
the knight also assists the undermining of
the centre.
1 2 . J.f4 c6
Pelletier follows in the footsteps of Aronian,
taki ng into account the experience of the
game van Wely- Leko (Dortmund 2008),
where the attem pt to stop the c- pawn
did not work: 14. b2 tDf8 1 S . b4 tDe6
1 6. Ae3 Ad6 1 7. a4 e7 1 8. l::! b1 a6 1 9. Ad2
l::! a b8 20. e3 cS 21 . bxcS bxcS 22. a2 g6
23. dxcS AxeS 24. tbe2 tbe4 2S. l::!fd1 Aa8
26. l::! x b8 l::! x b8 with equal play. However,
what he was hoping for is not clear. After
Black's reply it transpires that White has
simply wasted a tempo - an obvious hole
in his home preparations. Carlsen does
not forg ive such things.
14 . . . Ab4!
With the u nequ ivocal threat of 1 S . . . Axc3
1 6 . xc3 l::! x e2. The problems are not re
lieved by 1 S. &DeS?! tDxeS 1 6. AxeS tDd7 ! .

After t h e development o f t h e knight at d 7 1 5. J. d 2 c8 1 6. 'W b 2 'We7 1 7. fe1 'Wf8

Black is forced to su pport his dS-pawn .
(see next diagram)
1 3. 'Wc2 l3e8

(see next diagram) As Carlsen hi mself thinks, in principle this

queen retreat is not essential, since the
The same position was reached by trans- threat of e4 with the idea . . . Axc3 exdS is
position in the game Aronian - H uzman not real istic. But he sensed that the op
(Warsaw 200S). ponent was agon ising over his choice of

185
 Chapter 3 Life at the Top

a b c d e f g h a b c d e f g h
8 I .l. 'ii' fr 8
7 .l j_ 1.1\ .l .l .l 7
6 .l .l 1.1\ 6
5 l r .l 5
4 .t 4
3 Ci:J Ci:J 3
2 2
l:l l:l
a b c d e f g h a b c d e f g h
22 . . . .xc3!?
p I an , an d so for t h e moment he improves
the placing of his pieces. When pawn tension exists it is never easy
to make a choice as regards the further
18 . .lh3 J:Icd8 1 9. 'Wc2 h6!? transformation of the pawn structure.
Cont i n u i n g the same waiting tactics. Ac Should Black go in for an ' isolani ' - 22 . . .
tivity with 1 9 . . . 'De4 leads to m ass ex cxd4 23. tDxd4 'DeS, which stabilises the
changes: 20. tDxe4 d xe4 2 1 . Axd7 xd7 position with the prospect of protracted
22. A x b4 'Wxb4 23. 'DeS xd4 24 . tDxc6 play, or decide on hanging pawns? Carlsen
Axc6 2S. Wxc6 ed8 26. xd4! with the prefers to complicate matters.
simpl ification of the position and a prob 23. J.xc3 e4 24. dxc5 xc3
able d raw, which did not suit Carlse n , 25. 'Wxc3 bxc5 26. 'WaS f6
who was a i m i n g for victory i n the tour
nament.
a b c d e f g h
8 .l i. 'ir'e 8
20. h4 7 .t .l .l 7
Magnus's tactics have worked . How can 6 .l 1.1\ .l 6
White refuse the invitation to go to fS? But 5 Ci:J 5
with this creeping out of the ' lair' the play 4 4
becomes much more lively, which in view of 3
the difference i n motivation is to Carlsen's 2
advantage.

20 . . . cs 21 . f5 a6
The bS-square has to be covered against
the possible invasion of the knight.
22. 'Wb2
a b c d e f g h
A typical position with hang i n g pawns.
Black's pieces are ready to support their
activity with . . . dS-d4 and . . .'Df6-e4, which
demands energetic cou nter-action from
Wh ite such as 27. c1 !? or 27. Ag2.

(see next diagram) 27. 'Wb6?

1 86
 L. Dominguez - M. Carlsen

A strange move for a grandmaster. White 36. Sle3?!

provokes a pin, which is advantageous for
There was a quick win by 36 . . . c8! 37. Wa3
his opponent.
xe2+! 3 8 . tD x e2 Wh3+ 39. f2 Wf3+
21. . . .tc8! 28. g2 40. g1 We3+ 41 . g2 Ah3+ with mate
Only now did Pelletier see that the intended in three moves (variation by Carlsen).
28. tDxh6+ gxh6 29. AxeS does not work
37. Wd4?
in view of the interposition 29 . . . d6!, when
White loses a piece. If 28. f3 there follows A time-trouble mistake, leading to a rapid
28 . . . Ae6 with the threat of 29 . . . g6. d efeat. 37. g1 was more resi lient, al
28 e4!
..
though even here after 37. . . c8 38. Wa3
ce8 Wh ite would also have been u nable
With the th reat of . . . tD g 5 . N ow Wh ite's to resist the pressure.
game goes downhill.
37 Wh3+ 38. g1 Slxg3+ 39. hxg3
29. f3 g5
..

Wxg3+ 40. h1 f5
29 . . . tDc3 30. d3 (no better is 30. d2 d4
with the threat of . . .tDd5) 30 . . .tDxe2 31 . e3 Wh ite resigned .
Axf5 32. 3xe2 xe2+ 33. xe2 Ad3 leads
to the loss of a pawn.
30. g4 h5! 31 . Elc1 g6 Game41
L. Dominguez - M . Carlsen
31 . . . tDxh3 32. xh3 hxg4+ 33. fxg4 g6
34. tDg3 Wh6+ 35. g2 Axg4 was more Biel 2008
accurate, in the end leading to a position Sicilian Defence [ 878]
which occu rred in the game. 1. e4 c5 2. f3 g6 3. d4 Ag7 4. c3
32. g3 xh3 33. xh3 hxg4+ cxd4 5. xd4 c6 6 . .te3 f6 7. Ac4
34. fxg4?! 0-0 8 . .tb3 d6 9. f3 Ad7 1 0. Vd2 Erc8
1 1 . 0-0-0 e5 1 2 . b1
34. g2 was more resilient.
34. . .Wh6+ 35. g2 xg4 36. Vxc5? A useful prophylactic move, without which
Wh ite cannot get by in most branches of
a b c the Dragon Variation. Wh ite takes his king
8 8 i nto safety, inviting Black to declare his
7 7 i ntentions. For example, i n the event of
6 . 1 2 . . . tDc4 1 3 . Axc4 xc4 Wh ite can begin
an attack with 1 4 . g4!?.
5 5
4 ..t 4 12 . a6
ttJ 3 This m ove i n combination with 9 . . . Ad7
r:J;; !::, 2 was Carlsen's main weapon in 2008, one
which he employed in five games. If 1 3 . g4
a b c d e f 9 h Black can play 1 3 . . . b5.

18
 C hapter 3
Life at the Top

1 3. h4 h5 1 6. Ah6

a b c d e f g h In this variation Dom i n g uez was a pio

8 .!. 'if 8
.. . neer, and it tu rned out that the exchange
7 .i. .l .l .i. 7 of the dark-squ are bishops was prema
.l 1.1\ .l 6 ture . S u bseq uently the more dangerous
5 1.1\ 5 c o nt i n u at i o n 1 6 . d g 1 was e m p l oyed
4 [jj 4 against Carlsen , and his fi rst attem pt was
a fai l u re. I n his rapid game against Anand
3 it [jj it 3
(Mainz 2008) M ag n u s p l ayed 1 6 . . . 'Wa5
2 Wi 2 and after 1 7. Ah6 xc3? 1 8 . Axg7 c!l xg7
1
1 9 . x h 5 ! it was time for h i m to res i g n .
a b c d e f g h Tru e , aft e r 1 9 . . . x b 3 A n a n d s at i s
14. g4 fied h i m self with t h e w i n o f t h e q u een
- 2 0 . 'W x a 5 , but he c o u l d a l s o h ave
This flank pawn breakthrough is nowadays played more forcefu lly: 20. 'Wh6+! c!if6
a standard plan of attack. Black's defence 2 1 . 'Wg 5+ c!lg7 2 2 . tD x b3 'We? 23. 'Wh6+
is eased by the early exchange of the dark c!if6 24. xe5.
square bishops: 1 4 . Ah6 tbc4 1 5 . Axc4 The experi ment with 16 . . . c5, as played
xc4 1 6. Axg7 c!ixg7, and here 1 7. g4 is ag ainst To palov in the G rand Slam tour
not so dangerous - 1 7. . . hxg4 1 8 . h5 h8. nament (Bil bao 2008), also ended i n fai l
I n the rapid game Anand - Carlsen (Mainz u re. T h e g a m e cont i n u ed 1 7. Ah6 c!ih7
2008) Wh ite preferred 1 7. he1 , and after 1 8 . A x g7 c!i xg7 1 9 . f4 tb c 4 2 0 . A x c4
1 7. . . e8!? 1 8 . tbde2 (1 8. tbd5 e5! 1 9. tbb3 xc4 21 . f5 e5 22. tbde2 h8 23. b3 c6?
tDxd5 20. 'Wxd5 Ab5=) 1 8 . . .'WaS 1 9. tDf4 24 . xg4 tDf6 25. gg1 x h 1 26. xh1
ec8 2 0 . tbcd5 'Wxd2 2 1 . xd2 tD x d 5 tbg4 27. tDg3 with an obvious advantage
22. tbxd5 e 8 he satisfied h i m self with to Wh ite. And although an i m p rovement
an equal endgame. 14. Ag5 is examined in was found - 23 . . . c8! 24. xg4 (24. 'Wxd6
the game Radjabov - Carlsen (Baku 2008). Axf5!) 24 . . . tbf6 25. xh8 (25 . gg1 'We?!)
14 . . . hxg4 1 5. h5 t"Llxh5 25 . . . 'W x h 8 2 6 . g 1 c 6 , w h e n B l ack
holds the position (A . Kuz m i n), Magnus
a b c d e f g h d i d not tem pt fate any further. But the
8 .l. 'i .I. e 8 D ragon had once ag a i n demonstrated
7 .i. .l .l .i. 7 its viab i l ity!
.l .l 6 The main events occurred in the vari
5 1.1\ 1.1\ 5 ation 1 6 . . . 'Wa5 1 7. A h 6 (where Carlsen
suffered a disaster after 17 . . . xc3?). At
4 [jj .l 4
the Grand Prix tournament in Sochi 2008
3 it [jj it 3 Radjabov shocked Karjakin (and all con
2 WI 2 noisseurs of the Dragon) with a fantastic
1 double exchange sacrifice.
a b c d e f g h (see next analysis diagram)

188
 L. Doming uez - M. Carlsen

impun ity, which would not be the case with

his king on c1 on account of the exchange
of queens after 1 8 . . . Axh6 1 9. 'Wxh6 f4+.

1 8 . . . .lxh6

Not 18 . . .et:lc4?! 1 9. Axc4 xd4 20. Ad3

Axh6 21 . xh6 'Wg7 22. e3 ttlf6 23. g5
etlh5 (after 23 . . .tileS 24. gg4 with the threat
Analysis diagram of ggh4 Black is forced to make a further
weakening of his king's defences) 24. Ae2
followed by Axh5.
17 . . . Af6! ! 1 8. fxg4 Axg4 1 9. Axf8 xf8
1 9. 'Wxh6 g7 20. 'We3 f6 21 . g5 h5
20. e3 gxc3! 21 . 'Wxc3 xc3 22. bxc3
22. de2
e6 and Black gained excellent com pensa
tion . Perhaps not sufficient for a win, but a b c d e f 9 h
the psychological effect was so powerfu l 8 8
.i ft
that Karjakin was unable to save the game.
7 .i. .l 'it' 7
16 . . . e6?! .l .l .l 6
This allows Wh ite to develop a dangerous 5 5
initiative. Black should have clarified the sit 4 4
uation by exchanging the dark-square bish 3 Ji tt:J WI 3
ops: 1 6 . . . h7! 1 7. Axg7 xg7 1 8 . f4 et:lc4
2 t2J 2
19. Axc4 gxc4 20. f5 e5, when he would
have achieved a good game (A. Kuzmin). <;!;> :: 1
a b c d e f 9 h
17. i:idg1 'Wf6
White's plan of attack is 23. gxh5 g xh5
24. etlf4 with the th reat of etlf4xh5-f6+.
Black's counterplay is associated with the
advance of his queenside pawns and the
exchange sacrifice on c3.

22 . . . a5

T h i s m ove is m o re l o g i ca l in re p l y to
22. et:lce2 , whereas here 22 . . . b5 i s more
appropriate. But i n that case too Wh ite
carries out the same plan of attack as
a b c d e f 9 h i n t h e g a m e : 2 3 . g x h 5 g x h 5 24. ttlf4
h 8 2 5 . g h 1 w i t h t h e t h reat of ttlf4x
18. fxg4!
h5-f6+, whereas B l ack's cou nterp lay is
This is where the val ue of 1 2. b1 shows too late: 25 . . . h4 2 6 . 'Wf2 g xc3 27. bxc3
itself - Wh ite can capture the pawn with et:lc4 28. e5! , blocking the com m u n i ca-

189
 C hapter 3 Life at the Top

tions to the k i n g . N ow a l l capt u res o n a b c d e

eS are bad : 28 . . . til xeS 2 9 . g x h 4 , 28 . . . 8 8
dxeS 29. g x h4 , o r 28 . . .'xeS 2 9 . 'Wxh4.
7 7
H owever, there is sti l l the possi b i l ity of
6 6
28 . . . gca, attempting to evacuate the king
to the queenside.
5 5
4 tD ,.. 4
23. xh5!? gxh5 24. f4 g4 3 3

The alternative 24 . . .'Wh8 2S. gh1 h4 26. 'Wf2 2 tL1 2
gfd8 27. gxh4 'Wg7 28. gh3 gcs (with the 1
idea of 29. &DhS &Dg4 30. 'Wf4 'WxgS) looks a b c d e f 9 h
more attractive.
30 . . . Ac6!
a b c d e
Carlsen does not miss such opportunities.
8 8 Now he is able to cover his king.
7 7 31 . 'Wh4 .te4! 32. c3
6 6 Of course, not 32. 'Wxg4?? Axc2+ 33. c1
5 5 (33. a1 gas mate) 33 . . . AfS+.
4 tD ,.. 4 32 . . . .tf5 33. h5 f6!
3 In amazing fashion Black has activated his
2 position. With the inclusion of the rooks in
the defence after 34. &Dxf6+ &Dxf6 3S. gxf6
a b c d e f 9 h f7 Wh ite's attack peters out.
34. g3 c7 35. xf5 exf5 36. gxf6
25. 'Wh3 'Wd4? xf6 37. 'Wg5+ c;t?t7!
A rare instance of Carlsen playing impul- The following was a false trai l: 37. . . gg7?!
sively. 2S . . ,gxc3 26. bxc3 &DeS 27. &DxhS 38. 'WxfS gg1 + 39. a2 ga8+ 40. &Da4 f7
'Wg6, intending . . . gda, was essential. 41 . 'WbS.
38. 'Wxf5
26. ce2! 'Wxe4 27. Wxh5?!
a b c d e f 9 h
It is not clear why Dom inguez rejected the
8 .! 8
thematic 27. tL!xhS! . After all, even 'by eye' it
is obvious that the variation 27. . . fS 28. gxf6
7 7
f7 29. tL!hg3 'Wg6 30. tL!f4 is u nable to 6 ,.. 6
save Black. He preferred to win the queen, 5 lWJ 5
thinking that the king would not run away, 4 4
but he underestimated Black's defensive 3 tD 3
resources. 2 2
27. . . a4 28. h1 'Wxh1 + 29. Wxh1 axb3
30. axb3 a b c d e f 9 h

190
 M. Carlsen - T. Radjabov

38 . . . i:le7 rtle7 45. Vd3 rtlt7 46. b5 rtle7

Black begins erecting a 'fortress'. 47. d4 rtlf7 48. 'Wf3 h5 49. f5 rtle6
39. b4 50. d4+

Not 39. d5?? because of m ate in two Having fai led to find the key to the fortress,
moves. Dominguez forces a draw by repetition.

39 . . . i:le5 40. 'Wf2 i:ld8 41 . b3 i:ld7 50 . . .rtlf7 51 . f5 rtle6 52. d4+

42. rtlb2 i:lde7 43. 'Wb6 i:ld7 44. 'Wd4 Draw.

A new tournament - the Grand Slam in Bilbao (FI D E category 2 1 ) occupied a special
place in the chess calendar. It was organ ised with a record prize fund for tournaments
without the status of a world championsh ip - 400,000 Euros. The idea of the organis
ers was that those playing in it should be the winners of the main super-tournaments
of the year. Those taking part in the first tournament were the winners in Wijk aan Zee,
Linares and Sofia - Aronian, Anand and lvanchuk, together with the personally invited
Carlsen, Topalov and Radjabov. A special feature of the tournament was the 'football '
system of scoring (3 points for a w i n , 1 point for a draw).
The battle of the ' big six' proved highly uncompromising. This was a benefit perfor
mance by Topalov, who was striving to demonstrate the rig htfulness of his partici pation
in the current world championsh ip cycle (he was facing a match with Kamsky). Veselin
won four games, two of them against Carlsen , gaining revenge for the 'double' in Linares.
The blow to Magnus's self-esteem was softened by second prize in the tournament:
on the tie-break he finished ahead of Aronian. In last place, to general su rprise, was
the world champion Anand.

Game42 a b c d e f g h

M. Carlsen T. Radjabov
-
8 K 'iV K. 8

Grand Slam, Bilbao 2008 7 .. .i. .i. . 7

Sicilian Defence [ 878] 6
1.& 1.& 6

1. e4 c5 2. f3 d6 3. d4 cxd4 4. xd4 5 5

f6 5. c3 g6 6. J.e3 g7 7. f3 0-0 4
ik ti:J 4

8. 'Wd2 c6 9 . .lc4 J.d7 1 0. 0-0-0 i:lb8 3

i:J ik 3

(see next diagram) 2 WJ 2

w .: .: 1

This is the co-called ' C h i nese D ragon'. a b c d e f g h

The author of this brand name is the Bel-

gian master Luc Henris, who under this point of this move is to exploit the b-file
title published an analysis of 10 . . . .b8 in after the exchange of the knight on c4.
2002 in the New in Chess Yearbook. The At the same time, the main tru m p of the

191
 Chapter 3
'Chi nese' is to i m mediately react to the
wing attack 1 1 . h4 or 1 1 . g4 with the pawn
sacrifice 1 1 . . . b5! . It is dangerous to accept
it. For exam ple, the game Stanford -Anka
(McMi nnville 2004) continued 1 1 . h4 b5!
1 2 .tDdxb5 tbe5 1 3 . Ae2 \Wa5 1 4 . tbd4 1:!fc8
1 5 . tbb3 1:!xb3! 1 6 . cxb3 (1 6 . axb3? loses
to 1 6 . . . tbxe4 1 7. fxe4 \Wa1 + 1 8. tDb1 tbc4!)
1 6 . . . \Wxa2 1 7. \Wc2 \WaS with an attack for
Black. However, 1 2.tbcxb5!? tba5 1 3 . tba3
is not so clear.
11 . .tb3 tba5 1 2 . c.t>b1
Th is prophylactic move is best made i m
mediately. After t h e d i rect attack 1 2 . h 4
b5 1 3 . h 5 ? ! (1 3 . b 1 tbc4 1eads t o a posi
tion from the game) 13 . . . tbc4 14. Axc4
bxc4 1 5 . Ah6 'lti!f b 6 W h ite i s forced to
weaken his k i n g 's d efe n ce s . A g a m e
Echevari a - d a Costa (correspondence
2003) cont i n u ed 1 6 . b3 Axh6! 1 7. \Wxh6
'!WaS 1 8 . b2 1:!fc8 1 9 . tbde2 Ae6 with
the i n itiative for Black.
12 . . . b5 1 3. h4
The effect of the move . . .tba5-c4 can be
weakened by 1 3 . Ah6 A x h 6 1 4. \Wxh6,
but, as often happens i n the Dragon, the
early exchange of bishops does not give
anything. For example, the game Berg - Ev
domikov (Moscow 2007) contin ued 1 4 . . .
e5!? 1 5 . tbde2 tbxb3 1 6. cxb3 b4 1 7. tba4
(or 1 7. tbd5 tbxd5 1 8 . 1:!xd5 Ae6 1 9. 1:!d2
1:!b6=) 17 . . .'!We7 1 8. tbg3 Ae6 1 9. h4 1:!fc8
20. hS 1:!c7 21 . '!Wg5 tbe8 22. '!Wd2 1:!c6 with
equal play.
1 3 . . . tbc4 14 . .txc4 bxc4 1 5. ma1 h5!
Essential prophylaxis against the h4-h5
attack.
1 6. !i b1
Life at the Top
a b c d e f g h
8 8
.i
7 7
...
6 6
5
4
3 3
2
a b c d e f g h
The position reached is typi cal of the
'Chi nese Dragon' structure. The last two
moves are necessary for the defence of the
king. Despite the enti re power of Black's
heavy artillery, its defences are not easily
breached.
1 6 . . .Wa5
Black must act energetically to cou nterbal
ance Wh ite's possible attack. Instructive,
for exam ple, is the game Efi menko - Char
bon neau (Montreal 200S), which contin
ued 1 6 . . . !:!e8 1 7. Ah6 Ah8 1 8 . g4 hxg4
1 9 . h5 '!Wa5 20. hxg6 fxg6 21 . 1:!h4 gxf3
22. tbxf3 tbxe4? 23. tbxe4, and all Black's
su bsequent efforts proved in vain: 23 . . . c3
24. tbxc3 !:!xb2 25. xb2 Axc3+ 26. '!Wxc3
!:!b8+ 27. '!Wb3+ !:!xb3+ 28. axb3 e5 29. 1:!d1
Ac6 30. tbg5 d5 31 . !:!f1 1 -0.
But there are q uestions. Apart from
1 9 . . . '!Wa5 , the captu re 1 9 . . . tbxh5 should
also be exam ined . Now 20. !:!xh5 is parried
by 20 . . . Axd4! (20 . . . gxh5? 21 . '!Wg5+ h7
22. '!Wxh5 with a mating attack) 21 . Wxd4 e5.
Also 20. fxg4 tDf6 21 . Ag5 tbxg4 22. tDd5
is unclear. And final ly, in the game White
would hardly have felt so carefree if instead
of 22 . . . tbxe4? Black had played 22 . . .tbg4!
23. tbd4 Af6.
1 7. .th6
192
----------------------
M Ca
. a ja ov
- T. R_ __
r l se n d b
_____________________ _

The specific natu re of the position is such
that the typical pawn attack 1 7. g4 hxg4
1 8. h5 ti:)xh5 does not have the strength to
break through, both after 1 9. fxg4 Axg4,
and in the event of 1 9 . Ah6 gxf3 ! . I n the
game Kulaots - Papp (H ungary 2008) White
stuck to central strategy: 1 7. ti:)de2 l::!.fc8
1 8. l::!. h d1 l::!. b 4 1 9. a3 l::!. c bB 20. Ad4 1::1. 4 b7
and did not achieve anything sign ificant.
We should add that also nothing is given by
the typical tactic 1 7. ti:)d5 \Wxd2 1 8. ti:)xe7+
h7 1 9. Axd2 ti:)xe4, when Black regains
his pawn with the better position .
17 . . . Axh6! 1 8. 'Wxh6
a b c d e f g h
After the game Magnus ad m itted that he
had forgotten about the typical idea 20. h5
g5!, cutting off the queen. Its release by
2 1 . ti:)f5 Axf5 22. \Wxg5+ <!>h7 2 3 . fxg4
ti:)xe4! 24. \Wxf5+ (24 . ti:) xe4?? Wxa2+!)
24 . . . Wxf5 25. g xf5 ti:)g3 leads to an end
game favourable for Black.
20 ... lafb8 21 . h5 g5
21 . . . ti:)xh5 22. fxg4 (22. ti:)f5 g xf5 23. laxh5
f4! is not good for White) 22 . . . Axg4 is
stronger, when the knight on c3 stil l cannot
move, since after 23. ti:)d5 l::!. a 6 24. a3 c3!
25. \Wxc3 (25. ti:)xc3? \Wxa3+!) 25 . . . Wxc3
26. ti:)xc3 White finds himself a pawn down.
a b c d e f g h
8
.! 8

7 . -*- 7

6
.! 6

s if 5

4
tLl 4

3
tLl Viii 3

2 2

1 1:1 1:1 1
a b c d e f g h a b c d e f g h

18 . . . lab6 22. fxg4 xg4 23. Wd2 16 24. f3 laa6

Note this manoeuvre. Black intends to dou
ble rooks and at the same time he restricts
the mobility of the knight on c3 - if Wh ite
removes the defence from the a2-pawn he
gets mated ( . . . \Wxa2+! and . . . l::!. a6).
1 9. g4
Usually this thrust is followed up with h4-
h5, but with the queen on h6 this does not
work. But what else can be suggested for
White? After all, 1 9. Wg5 l::!. a6 is too passive.
Radjabov is looking for a way to approach
the king, but he fails to find one. He should
have centralised his q ueen - 24 . . . Wc5 ! ,
w h e n he is i n co ntrol o f the situati o n .
For example, in t h e event o f 2 5 . ti:) d 5 c3
26. xc3 \We3 B l ac k transposes i nto a
favourable endgame (Notkin).
25. lahg1 la b4 26. a3
Parrying the th reatened double sacrifice
on a2 .

19 . . . hxg4 20. We3 26 . . . Ae6

193
 Chapter 3 Life at the Top

After the game Radjabov suggested the fai ls to achieve his aim with 30. e2 tLlh6
pro m i s i n g q u een sacrifice 2 6 . . . !!a4 ! ? 31 . g2, after which, apart from 31 . . . Ag4
27. tbxa4 xd2 2 8 . tbxd2 !!xa4. 32. d2 f5, Black has an interesting com
b i n ation: 31 . . . Axe4! 32. xe4 xa3+!
a b c d e f 9 h
33. bxa3 !!xa3+ 34. tba2 !!xa2+! 35. xa2
8 8
!!a4+ 36. b2 c3+ 37. xc3 !!xe4, trans
7 .l 7 posing i nto an unclear endgame.
6 J. 6
30. 'We2 h6
s 'ii
4 .l .l 4 The trappy 30 . . . !!ab6?! (with the idea of
3 /j,t2J 3 31 . !!xg4? !! x b2!) is d u bious i n view of
2 2
31 . f3 Ae6 (31 . . .tbh6 32. tLld5) 32. tbg5+!
[j, [j, W!J fxg5 33. tbe4, when Wh ite breaks through
1 g g to t h e k i n g . T h e ex p i atory sacrifice
a b c d e f 9 h
33 . . . !!b3 also does not help: 34. cxb3 !!xb3
35. tbxg5+ h6 36. f8+ g5 37. xe7+.
27. e5!
31 . g3
Sensing the somewhat ponderous nature
of Black's manoeuvres, Carlsen resolutely
a b c d e f 9 h
opens up the position, forcing the oppo
8 8
nent to solve concrete problems.
7 .l 7
27. . . dxe5 6 J. 6
The c o m b i n ation 27 . . . xa3+ 28. bxa3 s 'ii ::, 5
!!xa3+ 29. tba2 c3 is refuted by 30. xg5+! 4 .l .l 4
fxg5 31 . !!xb4. 3 ::, t2J g 3
28. xg5! .tf5 2 ::, ::, Wi 2

If 28 . . .fxg5 there wou l d h ave fol l owed
29. xg5+ f8 30. h6 and a piece has to
be given up for this pawn.
29. ge4 mh7
On general g rou nds it wou l d be usefu l
to remove the king from the danger zone
- 29 . . . f7!?. The attem pt to d rive back
the knight with 30. g3 wou ld be refuted
by su pporti ng it with the rook: 30 . . . !!b8!
31 . e2 tbh6 32. e3 tbg4 33. g1 tbh6,
and Wh ite has to curb his ambitions - since
34. g6? Axe4 moves the 'pointer' towards
his own king: 35. tbxe4 xa3+. He also
1 Wg
a b c d e f 9 h
31 . . . .te6?
It was essential to help the king - 31 . . . b8!
32. g2 h8 (33 . !!g1 b6), since now
after 32. g2 tLlf5 33. g7+! Black would
have had no defence. But is there any need
to comment on ti me-trouble play? Perhaps
only for the sake of establishing the truth .
32 lagS? fS?
At this speed of play it is no wonder that
a combination exam ined i n the notes to

194
 L. Aronian - M. Carlsen

Black's 29th move was missed: 32 . . . Af5! a b c d e f g h

33. g2 Axe4! 34. xe4 xa3+! etc. , al 8 I. .*- 'ii' I. 8
though here the d ifference i n the placing
of the kings tips the evaluation in favou r
7 .l .l .l .i. .l .l .l 7
6
of Wh ite.

33. 'Wg4!
5
r :
4 [ 4

Now Carlsen is in charge! 3 3

33 . . . h4 34. g5+! fxg5
After 34 . . . r;t>h8 Wh ite gives a li near mate -
35. tDf7+ Axf7 36. l:!h6.
35. 'Wxg5 xg6 36. 'Wxg6+ Wh8
37. g1
Black resigned .
Game43
L. Aronian - M . Carlsen
G rand Slam, Bilbao 2008
English Opening [A32]
1. c4 c5 2. f3 f6 3. d4 cxd4 4. xd4
e6 5. g3 b4+ 6. d2
Before this Aro n ian had played 6 . Ad2 ,
including against Carlsen i n their famous
2007 Candidates match i n Elista. True, this
occu rred in the 1 1 th , bl itz game: 6 . . . Ac5
7. tDb3 Ae7 8. Ag2 tDc6 9. tDc3 b6 1 0. Af4
Ab7 1 1 . 0-0 0-0 1 2 . Ad6, where Magnus
played the superficial 1 2 . . . tDa5?! (1 2 . . .tDe8
is sounder) 1 3. tDxa5 Axg2 14. r;t>xg2 bxa5
1 5 . e4 l:!c8 1 6 . b3 and ended up in an i n
ferior position . But s i n c e that t i m e more
than a year had passed - a gigantic d is
tance for the swiftly- i m p roving Carlse n .
B y defending against t h e check with h i s
knight, Leven i nvites Magnus t o go i n for
a ' hedgehog' structure.
6. . . c6 7. c2 e7 8. g2 0-0 9. 0-0
2
[jj [jj 2
1 J: :s w
a b c d e f g h
After complet i n g h i s kingside d evelop
ment B l ack has to d e c i d e o n h i s s u b
s eq u e nt p l a n , taki n g i nto acc o u nt t h e
features o f the positio n . It is clear that ,
w i t h his b i s h o p o n g2, Wh ite w i l l aim to
coord i n ate it with his k n i g hts, for which
the natural posts are c3 and d4, but for
the transference of t h e k n i g ht from d2
time is req u i red. I n the spi rit of the posi
tion is 9 . . . d5 1 0 . cxd5 exd5 1 1 . tDb3 Af5
1 2 . tDcd4 Ae4, when with his isolated d
pawn Black's pi eces are actively placed
for the forthcom ing battle. Another plan
i nvolves the setting u p of a ' hedgehog ' :
9 . . . b6 1 0 . b3 A b7 1 1 . A b 2 c7 1 2 . e 4
d6 1 3 . tD e 3 , where Wh ite h a s a spatial
advantage .
9 . . . b8!?
A new idea. Black is intending to undermine
the c4-pawn.
1 0. e4
Wh ite is intending to seize control of the
weak d6-point. Another route for the knight
comes into consideration - 1 0. tDb3 with
the idea of 1 0 . . . b5 1 1 . c5. It is true that this
allows Black to develop pawn activity on
the queenside, but, without the support of
the pieces, this does not come to anyth ing:

195
 Chapter 3 Life at the Top

1 1 . . . a5 1 2. tDcd4 Yfic7 (1 2 . . .tDxd4 1 3 . Yfixd4) n atural, with the prospect of converting

1 3 . Ad2!? (1 3. a4 bxa4 expands the scope
of the rook on b8) 13 . . . a4 1 4 . tDxc6 dxc6
1 5 . tDa5 (nothing is given by 1 5 . Aa5 Yfie5
1 6. Ac3 Yfic7) 15 . . . Ad? 1 6. Yfic2, and after
1 7. b4 White blocks the queenside, freeing
his hands for play i n the centre.
10 . . . b5 1 1 . cxb5 xb5 1 2 . ttld6 Axd6
After the roo k retreat 1 2 . . J b 8 W h ite
consolidates his hold on d6 - 1 3 . b3 Aa6
1 4. Aa3 Yfia5 1 5 . l:!e1 l:!fd8 1 6. e4 with the
idea of 1 6 . . . tDe8 1 7. e5, and 17 . . . tDxe5?
is not possible in view of 1 8 . tDc4! tDxc4
1 9. Axe?, winning the exchange.
1 3. Wxd6 Ab7 14. ttla3
14. b3 tDe7 1 5. Axb7 l:!xb7 1 6. Ab2 came
into consideration, retaining pressure on
the dark squares.
14 . . . b6
a b c d e f g h
8 'if .. . 8
7 I. ..t I. 1. 1. 1. 7
the q ueenside pawn majority.
15 . . . xb2 1 6. Ac5 e8 1 7. ab1
W h ite c o u l d h ave won the exc hange:
1 7. tDc4 l:!xe2 (not 17 . . . l:!b5?! 1 8 . a4 l:!b3
1 9. Yfid1 l:!c3 20. tDd6, winning material)
1 8 . Y!id3 Aa6 1 9. Yfixe2, but after 1 9 . . . d5
fol l owed by . . . Axc4 B l ac k wou l d have
gained adeq uate material compensation.
1 7. . . xb1 18. xb1 Aa6
1 8 . . . Aa8 was weaker: 1 9 . tDb5 Y!ib8
20. l:! b 3 ! ? with the idea of 2 0 . . .tDa5
21 . Axa7, regai ning the pawn.
1 9. ttlb5 Axb5 20. xb5 Wc8 21 . a4 h6
It is useful to open an escape square for the
king, prepari ng . . . Yfia6, which for the mo
ment does not work because of 21 . . . Yfia6
22. Axc6 dxc6? 23. l:!b8.
a b c d e f g h
8 'iV I. 8
7 I. I. 1. 1. 7

6 6
6
iV I. I. 6
.. . 5
:t. iL 5
5 5
4 /j, 4

a b c d e f g h
a b c d e f g h

22. Aa3
1 5. Ae3?!
The logical com pletion of Wh ite's strategy
The ambitions of the two players have con
was 22. e4!? Y!ia6 23. e5, when the advan
verged on the d6-point. For the sake of
tage of the two bishops is more apparent.
this Aronian is ready to sacrifice a pawn,
Now the black q ueen comes out into the
although fu ll compensation is not apparent.
open.
1 5 . tDc4 l:!a6 1 6. Ad2 (preventing . . .tDa5)
1 6 . . . tDe7 1 7. Y!id3 Axg2 1 8 . c.!>xg2 is more 22 . . .Wa6! 23. Ab2 Wxa4 24. Axc6

196
 L. Aronian - M . Carlsen

dxc6 25. b4 Chess history knows many examples of

Si nce 25. 'Wxc6 a6 26. c5 'Wxc6 27. xc6
a8 28. Axf6 gxf6 leads to a rook end
game a pawn dow n , Aron ian prefers to
keep the queens o n , with prospects of
an attack agai nst the somewhat exposed
black k i n g.
25 . . . 'Wa5 2 6 . .1xf6 gxf6 27. g4+ h7
28. Wxc6
The suggestion of M i khail Golubev was
more subtle: 28. 'Wd7 f8 29. f4!? and if
29 . . . h5 (cutting off the rook from g4), only
now 30. 'Wxc6.
a b c d e f 9 h
8 K
the conversion of an outside passed pawn
with the same number of heavy pieces - re
member the famous 34th game of the Ale
kh ine- Capablanca match (Buenos Ai res
1 927). But there the king was covered by
a continuous pawn chai n .
2 9 . . . f5 30. a4?
Too late, as now the i n itiative passes con
cl usively to B l ac k . 30. g2 was better,
with the faint hope of 30 . . . We1 ?! 31 . 'We?,
but Black plays 30 . . . d2! 31 . a4 'Wd5+
32. e4 fxe4 33. 'Wxe4+ 'Wxe4+ 34. xe4
a2, transposing into a rook endgame with
an outside passed pawn and real winning
chances.
30 .. 'We1 + 31 . g2 d1 32. Wc7
8

.7 32. x a7? 'Wh1 + 33. h 3 g 1 leads to

.l s
'-...!-'
m ate .
5 32 . . . g6!
M 4
Played i n the style of Steinitz - the king
8 3 can defend itself.
88 8 2
33. f3!
a b c d e f 9 h
Aronian is n ot lacking in i n g e n u ity, and
within a couple of moves Carlsen's ambi
29. 'Wc2+?
tions will be curbed som ewhat.
A critical moment. Wh ite q u ite logically
33 . . Wh1 + 34. e3 a1 !
ai m s to provoke the weakening . . .f6-f5 ,
but he carries it out inaccu rately, allowi ng The kings have succeeded i n defending
an i m portant transposition of m oves . It themselves. M ag n u s offers to take play
was correct to beg i n with 29. a4!. Since i nto a queen endgame.
29 . . . 'We1 + 30. g2 d1 is i neffective on
accou nt of 31 . We? ! , the queen wou l d 35. Wc2
have t o move off the im portant diagonal
W h ite c a n n ot refu se: 3 5 . c4? a3+
- 29 . . . 'Wb6, and now 30. 'Wc2+ f5 31 . e4!
36. d2 'Wd5+ 37. e1 a2.
fxe4 32. xe4 exposes the king, forcing
Black to concern himself with its defence. 35 . . . xa4 36. Wxa4

197
 Chapter 3
Life at the Top

a b c d e f g h Game44
8 8 M . Carlsen - L. Aron ian
7
Grand Slam, Bil bao 2008
Sem i-Slav Defence [ 04 7]
6

5 5 1 . d4 d5 2. c4 c6 3. f3 f6 4. c3 e6
4 4 5. e3 bd7 6. J.d3 dxc4 7. J.xc4 b5
8. J.d3 J.b7
3
8 3

2 88 8 2 a b c d e f g h

a .! iV ..t. i. 8
a b c d e f g h 7 I. .i.
J!l\ 1. 1. 1. 7

36 . . . 'Wc1 +
6 I. I. J!a\ 6

5
[ 5

The transition i nto a pawn endgame by 4

3
8 1 4

36 . . . 'M'e4+ 37. 'M'xe4 fxe4 38. <;!;lxe4 is for
lovers of pai nstaking home analysis, but,
speaking on general grounds, it is u n real
to hope to save this ending where the op
ponent has an outside passed pawn.
37. f3 Wc3+ 38. g2 a5 39. g4
39. e4 was objectively somewhat strong
er.
39 . . . We5 40. gxf5+ xf5 41 . We8 g6
42 . 'Wf8 a4 43. e3 'We4+ 44. g3 Wd3
45. h4
It is poi ntless to p l ay 4 5 . 'M'g8+ <;!;lt6
46. Wh8+ <;!;le7, when the king escapes
from the checks.
45 ... a3 46. h2 'Wf5
Black also wins by 46 . . . a2 47. h5+ <;!;lt6
48. 'M'xh6+ <;!;le7 49. Wg5+ <;!;Ita 50. Wf6
'M'd1 .
47. Wxa3 Wxf2+ 48. h3 'Wf3+ 49. h2
h5 50. 'Wf8 'Wf2+ 51 . h1 g4
Wh ite resigned.
3
CiJ i, 8 CD
2 88 888 2
1 1:, 1:, 1
i,
a b c d e f g h
Cu rrently the main continuation, which has
supplanted the classical 8 . . . a6 9. e4 c5
1 0 . e5 cxd4 1 1 . &Dxb5 axb5 1 2 . exf6 gxf6
1 3 . 0-0 'M'b6 14. 'M'e2 . Of the recent battles
on this theme, mention should be made
of the Kram n i k - Anand duel in the match
for the world cham pionsh i p (Bonn 2008),
where Black was able to demonstrate an in
tegrated plan of attack on the king side. The
5th game of the match continued 1 4 . . . Ab7!
1 5 . Axb5 l::! g 8! (the most accu rate move
order) 1 6. Af4 Ad6 1 7. Ag3 f5 1 8 . l::!fc1 ! f4
1 9. Ah4 Ae7 20. a4 Axh4 21 . &Dxh4 <;!;le?!
with complicated play.
9. a3! b4 1 0. e4 xe4 1 1 . J. xe4 bxa3
1 2 . 0-0 f6
Black's play is based on n uances. By de
laying the development of his dark-square
bishop, he hopes to save a tempo in com
parison , for exam ple, with the game Aro
nian - Grischuk (G rand Prix, Sochi 2008),

1 98
 M . Carlsen - L. Aro n i a n
------
tD

which cont i n ued 1 2 . . . Ad6 1 3 . b3 f6 1 6. Ab5+ d7 1 7. e5 with a mortal pin)

14. d2 Y!ic7 1 5 . Af3 ! Axh2+ 1 6. <!>h1 Ad6
1 7. c4 Ae7 1 8. Axa3 0-0 1 9. Ac5! );!.fd8
20. b4 with a spatial advantage for Wh ite.
If 1 2 . . . Ae7 there can follow 1 3. b3 0-0
14. Axa3 Axa3 1 5 . l;!.xa3 Y!ie7 1 6 . b4 with
pressure for Wh ite on the q ueenside.
13 . .td3 axb2 14 . .txb2 a5
1 6. d4! Ae7 1 7. f5, and now in the event
of castling - 1 7 . . . 0-0? Black comes under
a mating attack: 1 8 . xg7! <!>xg7 1 9. Y!lg4+
<!>h8 20. A xf6+ Axf6 21 . Y!if5 .
1 6. tbe5 tbfS 1 7 Y!Va4 .tb4 18. tbxc6
.txcS 1 9. Y!VxcS+ me7 20. fd1 c8
21 . 'Wf3 'Wb6

Black is ready to develop his bishop on b4, Vacating the square for the rook.
neutralising Wh ite's pressure on the open
22 . .td4 'WbS 23 . .tas cdS 24 . .tb7!?
files on the queenside.
By blocking the q u e e n , Wh ite prepares
a b c d e f g h

'if .i.
an exchange sacrifice on b4, after which
.I 8
the black king w i l l come u n d er the dan
7 .\ 7 gerous c ross-fire of the bishops. I n the
6 6
. "' event of the d i rect 24. l;!.ab1 with the idea
5 of 25. Ac5+ B l ack can either sacrifice
4 his q u e e n : 24 . . . l;!. h e 8 25. Ac5+ A x c 5
3 2 6 . l;!. x b8 l;!. xd 1 + 2 7. Y!i x d 1 . x b8 w i t h
2 a d efe n s i b l e positi o n , o r m e c h a n i c a l l y
forestall Wh ite's threat: 2 4 . . . l;!.d 6 25. Ab7
1 _:
. h d 8 , prepari n g the evac uati o n of the
a b c d e f g h
king.
1 5 . d5! a b c d e f g h

A pawn sacrifice, w h i c h c h aracterises

8 'if .I .I 8

Carl sen's concrete t h i n ki n g style. It re

7 7
sembles his pawn sacrifice agai n st lv 6
l . "' 6

anc h u k (Foros 2008) . W h ite opens the 5 5

diagonal for his dark-sq uare bishop, and 4 .i. 4

now t h e king will be u na b l e to cast l e . 3 [j 3

Before this the g a m e Gelfand - Kram n i k 2 [j [j [j 2
(World Cham pionship, Mexico 2007) went
1 5 . Y!la4 Ab4 1 6 . Aa3 d5 1 7. e4 b6
: :
a b c d e f g h
1 8 . Y!ib3 Y!ie7 1 9 . l;!.ab1 Axa3 20. Y!ixb6
Ab4 2 1 . e1 0-0, and Black mai ntai ned 24 . . . h 5?.
the balance.
Black has succeeded i n coordinating his
15 . . . tbxd5
forces, but this move gives a new impe
The main idea of the pawn sacrifice is tus to Wh ite's i n itiative. Correct was 24 . . .
revealed after 15 . . . exd5 (not 15 . . . cxd5? e5! 2 5 . Ab6 .d6 26 . . x d 6 Y!ixd6 (not

199
 C hapter 3 Life at the Top

26 . . . Axd6 27. l::! x a5 with the threat of !::!a?) It is hard to comment on this type of care-
27. Ax aS Ax aS 28. l::! x a5 l::! d 8, and Black lessness without taking the preceding play
completes his development. into accou nt. It seems incred ible that such
a powerful g randmaster as Aronian was
25. h3
unable to assess the conseq uences of
If 25. l::! ab1 there fol lows 25 . . . tbg4 26. g3 f6, Wh ite's attack after the obvious 28. Ac5+.
preparing an escape square for the king. But simi lar 'black-outs' often occur with
Carlsen's opponents, who are u nable to
25 . . . h4 withstand the prolonged tensi o n , when
Here too 25 . . . e5 should have been played . they have to fi nd a series of only moves.
Now the black king ends up in a mating net
26. S:ab1 ! e5 and the game concludes quickly. 27. . . exd4
28. l::! b xd4 'lflc7 was essential , parrying the
Too late!
d i rect threats, although the ill-considered
27. S:xb4! advance of the h-pawn has seriously com
promised Black's position. For exam ple:
a b c d e f h
9 29. AdS tbxd5 30. l::! x d5 l::! x d5 31 . 'lflxd5,
8 if .I .I 8 and in view of the th reat of '/flg5+ he can
7 j, .,.,. 7 hard l y avoi d a further weakening of his
6 "' 6 position - 31 . . .f6.
5 I. I. 5 28. Ac5+ e6 29. S:a1 !
4 j, I. 4 The game is decided.
3 VIii 3
29 . . . S:d6 30. Axd6 xd6 31 . 'Wc6+
2 2 e7 32. S:a8! 'Wd6 33. 'Wxd6+ xd6
34. S:xh8 b3 35. Aa6 ttid7 36. S:xh4
a b c d e f g h ttic5

27 . . axb4? Black resigned.

At the Olympiad i n Dresden Carlsen performed as a genuine leader. He confidently led

the Norwegian team in all eleven matches, in which he contributed 71/2 points. The entire
team also played wel l and finished i n 1 9th place, its most successfu l in recent h istory.
Among the young player's games, special note should be made of the one with the
veteran Alexander Beliavsky, in which Magnus was given another master class by one
of the outstanding grandmasters from the 'golden age' of chess.

200
 M. Carlsen - A. Bel i avsky

Game 45 He undoubted ly knew that after the bishop

M. Carlsen A B eliavsky
-
. retreat 1 2. Ac2 with 12 . . .tDe6 1 3 . tDg3 Axf3
Olympiad , Dresden 2008 14. xf3 Ag5 Black can exchange his 'bad '
Ruy Lope z [ C84] dark-square bishop and obtain the bet
ter game. I ndeed , in the i m mediate future
1 . e4 e5 2. f3 c6 3 . .tb5 a6 4 . .ta4
Wh ite can not advance d3-d4, whereas
f6 5. 0-0 .te7 6. d3 d6 7. c3 0-0
Black has a clear plan - . . . g7-g6, . . . fg7,
8. It e1
. . . g5 and the preparation of . . .f7-f5 .
Alexander B e l i avsky i s known for be
1 2 . . . bxc6 1 3 . g4
ing faithfu l to the open ings he employs.
Therefore i n his preparations for the game The main line is considered to be 1 3 . tDg3
Carlsen was probably expecting the vari A xf3 1 4 . xf3 tDe6 1 5 . Ae3 c5, where
ation with 8 . . . tDd7, which Beliavsky had Black has a comfortable game.
employed against h i m i n the ' Rising Stars'
13 . . . .tg6 14. g3 e6 15. mg2
tournament in Amsterdam 2006 in reply to
8. tDbd2, and also in this Olympiad against Wh ite delays with the logical 1 5. d4, and
Bachman, who i n fact played 8. l::! e1 . But after Black's reply it is no longer a reality.
the experienced g randmaster changes
1 5 . . . c5 1 6. It h1 f6 1 7. h4 d5
course. Besides, the game with Bac h
man, w h i c h d eveloped i n very u n usual a b c d e f g h
fashion - 8 . . . tDd7 9 . d4 exd4 1 0 . cxd4
8 .!. 'iV .!. 8
tDb6 1 1 . Axc6 bxc6 1 2. tDc3 aS 13. Ae3
f5 1 4 . e5 f4 1 5. Ac1 d5, left some questions
7 .l .i. .l .l 7
unanswered . l.il .l .i. 6
5 .l.l 5
8 . . . .tg4 9. h3 .th5 1 0. bd2 d7
4 ::, ::,{:::. 4
1 1 . f1 c5
3 t::, t::, CbtD 3
a b c d e f g h 2 ::, ::, ::, 2
8 .!. 'iV .!. 8 1 1:: iL 1:: 1
7 .l.l .i..l.l.l 7 a b c d e f g h
6 .l l.il.l 6
1 8 . c4
5 "' }. .i. 5
4 jL ::, 4 Carlsen recognises in good time that he has
3 t::, t::, CiJ t::, 3 not achieved anything from the opening and
2 ::, ::, ::, ::, 2 has even ended up in an inferior position.
Therefore he fixes the pawn structure in
1 1:: iL l::Cb the centre, realising that otherwise Black
a b c d e t g h
himself will play . . . c5-c4 and gain control
1 2 . .txc6 of the i mportant d3-point. But now the d4-
point becomes weak, and it may prove an
It is hard to surprise Carlsen in the open ing. excellent outpost for the black pieces.

201
 Ch apter 3 Life at the Top

18 .. dxe4 The bishop makes way for the knight, ex

ploiting the fact that the cS- pawn is ta
There i s n o point in B l ack avo i d i n g the
boo: 30. AxeS E!xd3 31 . E!xd3 AxeS; or
exchange of queens (1 8 . . . c6), si nce after
30. tDxcS tDb4 31 . E!xd8 E!xd8 32. E!c1 E!d2
1 9. tDfS Wh ite has active play on the king
33. tDxa4 E!xf2+ 34. xf2 tDd3+ 3S. e3
side. I n the endgame Black gains control
tDxc1 , and White has to make considerable
of the d-fi le, and the attractiveness of the
efforts to regain the knight.
d4-point is not red uced .
30. et:\e3 et:\d4 31 . !ac1 et:\e6 32. !acd1
1 9. dxe4 'Wxd1 20. !axd1 !afd8

This rook, since the f8-sq uare has to be a b c d e f g h

.J..e 7
freed in the event of tDfS . 8 I. I. 8
7
.J.. 6
21 . .ie3 et:\d4 22. !ad2 .tf7 23. !ac1
IaabS 24. et:\e1 6
5 5
b ad cf g h
e
B
B BB 4
8 I. I. 8 3 tiJtiJB 3
7 7 2BB :: 2
6 . I .J...J..
t 6 ::
5 5 b d f g h
BB
a c e

4 BB 4
3 tLJ 3 32 .. !ad4!
2BB :: Bil 2 A m ove i n t h e sty l e of t h e n i nth world
1 I :: tLJ 1 c h a m p i o n Tigran Petrosian . B l ac k dis
b ad cf g h
e
p l ays h i s read i ness to sacrifice the ex
change: in the absence of his dark-square
24 .. a5! bishop, Wh ite will h ave nothing with which
Bel i avsky gives the young g randmaster to defend the dark squ ares . 32 . . . E!xd3?!
a good lesson in strategy. Having estab wou l d have been a false tra i l : 33. E!xd3
l ished control of the b-fi le with . . J !ab8 and tDf4 + 3 4 . g3 tD x d 3 3 S . E! x d 3 E!xb2
. . . a7-aS, he at the same time weakens the 36. E!d 8 g6 37. tDd 1 !, and i n view of the
dark squares in Wh ite's position. th reat of AxeS Black is forced to g ive up
the exchange: 37. . . E!xf2 38. xf2 Axc4
25. et:\d3 .te6! 26. f3 a4 27. et:\f5 .tf8 39. tDc3 f7 40. E!a8 , when he has a dif
28 . .tf2 ficu lt endgame.
I n the event of the exchange on d4 -
33. h2
28. tDxd4 cxd4 29. Af2 g6 30. E!cc2 cS
Black strengthens h i s centre, and after Removing the threat of . . . E!xd3.
. . . Ad6 he prepares . . .f6-fS .
33 !a bd8 34. et:\d5 !axc4 35. b3 axb3
.

28 . . et:\c6 29. !acc2 .tf7! 36. axb3 !ad4!

202
 M . Carlsen- A. Bel i avsky
__._________ .::::;;,
ttJ

N evertheless forcing the acceptance of 4 8 . . . h5 49. f!:b8

the exchange sacrifice, as otherwise af
ter . . . c7-c6 the pin on the d-fi le goes i nto a b c d e f 9 h
operation. 8 .! .: .i. . 8
37. Axd4 xd4 38. g2 xb3 39. f!:b2
7 .i. .t. 7
c6 40. e3 c4 41 . f2 d4 42. f!:b7 6 I. 6
5 .t. .t. 5
a b c d e f 9 h 4 fj, fj,fj, 4
8 .! .i.. 8 3 I:J!j, 3
7 .i. .t. .t. 7 2 .: 2
6 6
5 5 a b c d e f 9 h
4
3 3 49 . . .f!:a3!

2 2
In an endgame the exc h a n g e down, it
is i m portant to retai n the rook to create
a b c d e f 9 h
cou nterplay.
42 . . . c3? 50. f!:c8 hxg4 51 . fxg4 h7 52. f!:b7
Ag6 53. f!:xc4 c5
The 37-year handicap nevertheless means
something! After safely reaching the time f 9 h
a b c d e
control, the tired Bel iavsky conti nues play
8 .i. 8
ing 'automatically' and blunders a pawn.
After 42 . . J:Ia8! he would have had a stra
7 .: ....7
teg ically won position. 6 .t. .i. 6
5 .t. 5
43. f!:d3 c2 44. xc2 c5
4 .: fj, fj,fj, 4
A material advantage, although a mini mal I:J 3
one, is now on White's side. Beliavsky faces 2 2
a difficult struggle for a d raw, in which he
successfu lly demonstrates his defensive
I:J
a b c d e f 9 h
mastery.

45. f!:d2 f!:a8 46. e3 c4 47. fd1 e6 54. f!:bb4

48. f!:db2
48. d5!? was more energetic, with the
idea after 48 . . .d4 of playing 49. fid b2 ,
intend ing the exchange of rooks, and in the
event of 49 . . . Axd5 50. exd5 Black has to
reckon with the passed d-pawn .
54. fib8! Ad6 55. fic8 was stronger. For
example, 55 . . . Axe4+?! does not work i n
view o f 5 6 . g3 with t h e threat o f fixc5.
55 . . . fia2+ 56. f3 (56. g1 fia5 57. f2
e6 58. f5 is also i nteresting) 56 . . . b3
is safer, but here White can return the ex-

03
 Chapter 3 Life at the Top

change by playing 57. c3 (57. c!>g3 d2
58. fk2 !!xc2 59. !!xc2 xe4+ leads to a
reduction in the material) 57. . .d2+ 58. c!>e2
!!b2 59. c!>d3 with dangerous threats.
54 . . . iia2+
pawns, in which by some m i racle the d3-
pawn survives. 57. . . !!d3!? 58. !!e2 g61ooks
more prosaic, but perfectly sound.
58. iixd4 exd4 59. f5 d3 60. iib2
xe4+ 61 . mt4 c5 62. f2 g6

54 . . . Af7 55. !!c3 !!a1 was stronger.
55. mg3
B lack's defence wou l d have been more
d ifficult after 55. c!>f3!? e6 56. c3 !!h2
57. !!b8 (57. h5? g5+) 57. . . !!h3+ 58. c!>e2
Ac5 59. f5 .
55 . . . iid2 56. h5 .tf7 57. iic2
a b c d e f 9 h
8 .i. 8
7 .t..t. 7 e4+ 76. xe4+ Axe4 leads to the loss
6 6
5 [j_ 5
4 [j_ 4
3 3
2 2
a b c d e f 9 h
57. . . iid4!
An accurately-calculated exchanging op
eration with a red uction in the n u m ber of
The n u m ber of pawns is i n exorably re
d uced .
63. hxg6+ mxg6 64. me3 J.e6
s5. h4+ mg5 ss. t3+ mgs s1. iibs
mt7
After 67. . . Ah6+ 68. c!>d4 d2 the passed
d - pawn conti n ues to s u rvive: 69. !!c6
(69 . f! b1 d7 70. c!>d3 C5+ 71 . c!>d4
d7) 69 . . . b3+ 70 . c!>c3 Ad5 71 . h4+
c!>g5 72 . !!d6 Ab7! 73. f5 (73 . !!b6 c5)
73 . . . Af8 74. !!d7 c5, and now 75. !!xd2
of the last pawn.
68. iic6 .td7 69. iic7 mea 70. d4
Ah6+ 71 . mt3 e6 12. xe6 Axe6
73. xd3 f5 74. g xf5 Axf5
Al ready here peace cou ld have been con
cluded.
75. e5 Af8 76. iia7 Ad6 77. c4 Ah2
78. e3 .td7 79. me4 Ag1 so. iia8+
mt7 81 . iia7
Draw.

The year 2009 began i n an u n usual way for Carlsen . The N o rweg ian's ad m i rers
were su rprised by h i s peaceab l e n ess at the su per-to u rnament i n W ij k aan Zee
(FI D E category 1 9) . It seemed that t h i s was the very t i m e to b u i l d on the success
of the previous year, since fo r vari ous reasons the main favou rites for the chess
crown h ad decli ned to p l ay in the tournament: Anan d , Kram n i k and To palov. But
alongside Carlsen's name in the tournament table a depress i n g series of ' half
poi nts' appeared . H owever, it was clear to the experts that the reasons were of a
perso nal nature . It is wel l known that Carlsen n ever plays d e l i berately for a d raw,
as i n d i cated at the l east by the length of h i s games. One of the reasons for such

04
 M. Carlsen - L. Dom i nguez

an occu rrence was h i s g reatly increased prestige: opponents began p l aying more
cautiously ag ainst h i m . But there were also some notable set- bac ks. Playi ng ' by
tec h n i q ue' i n the endgame d i d not always work, as, for exam ple, i n heavy piece
end i n g s with Radj abov and Karjaki n , which by their protracted n atu re resem b l ed
individual games. N evertheless, by w i n n i n g i n rou n d s 1 0 and 1 2 ag ainst Dom i n g uez
and Smeets , Carlsen j o i ned the battle for the fi rst prize. But his strength was on
the wane, and i n the l ast rou n d he was outplayed by Wang Yue. Com pared with the
fai ry-tale previous year, h i s fi nal score of +1 with ten d raws and a share of 5th-6th
places looked very modest i n d eed .

Game 46 Wxd3 1 5 . l::! x d3 &Dc6 1 6 . 0-0 &DeS gives

M. Carlsen - L. Dominguez Black an easy game.
Wij k aan Zee 2009 12 . . . c6 13. f3
G runfeld Defence [ 081] For the moment it is too early to separate
1 . d4 f6 2. c4 g6 3. c3 d5 4. 'Wb3 the d-pawn from its base. I n the event of
dxc4 5. 'Wxc4 .tg7 6. e4 0-0 7. .te2 1 3 . d6 Ae6 14. &Df3 Wf6 Black is threatening
fd7 8 . .te3 b6 9. 'Wd3 to su rround it: 1 5 . Wc2 l::! d 8 1 6. 0-0 &DeB.
13 ... cxd5 1 4. xd5 xd5 1 5. 'Wb3
a b c d e f g h b c d
... 8
a e

8 --..t'if
..t7
6 6
5 5
4 4
3 ti:J il, 3 3
2 il, 2
1 <;t> tt:J 1
a b c d f
e g h
9 . . . f5
Dom inguez tries to carry out this idea more
accu rately than i n his game with Kram n i k
from t h e Olympiad in Dresden, where h e
played 9 . . .&D c 6 1 0. l::! d 1 f5.
1 0. l::!d 1 f4 1 1 . .tc1 e5 1 2 . d5
The critical continuation. This pawn has the
prospect of becoming passed. The feature
less 1 2 . &Df3 exd4 1 3 . &Dxd4 c5 1 4 . &Ddb5
8
6 6
5 5
4 4
3
a b c d e f g h
This was undoubted ly prepared at home.
The exchange of q ueens, carried out in
the game Jakovenko - Mamedyarov (G rand
Prix, Elista 2008), did not g ive Wh ite any
t h i n g : 1 5 . Wxd5+ Wxd5 1 6 . e x d 5 Af5
1 7. 0-0 &Dd7. Carlsen is not against the
queen exchange, but in a better version.
For exam ple: 1 5 . . . W b 6 1 6 . Ac4 Wxb3
1 7. Axb3 &Dc6 1 8 . Axd5+ h8 1 9. h3.
15 . . . h8

2 05
 Chapter 3
Life at the Top

The pawn sacrifice 1 5 . . . Ae6 1 6. exd5 Af7 1 9 . . . b5!

1 7. 'l!flxb7 tDd7 looks promising, but after Now Black solves his open ing problems.
Carlsen's meteoric rise the psychological 20. Axb5
approach of his opponents was al ready Also after 20. Ad2 bxc4 (weaker is 20 . . . a5
d ifferent. For exam ple, after 1 8 . d6 Axa2 21 . Axb4 axb4 22. Ad5) 2 1 . 'l!flxb4 '!ftc?!?
1 9. 0-0 Ae6 20. f!fe1 the wh ite pieces co (21 . . . Wxb4 22. Axb4 f!e8 23. 0-0) 22. Ac3
ord i nate very harmoniously, and the d6- Ab7 23. tDg5 f!ae8 Black has noth ing to
pawn may cause Black certain problems. fear.
Doming uez sensibly prefers not to throw 20 . . . xa2 21 . Ad2 b8 22. Y!la4 Y!lb7
caution to the winds and to retain a stable 23. 0-0 Y!lxb5 24. Y!lxa2 Y!lxb2 25. Y!lxa7
position. Y!l b7 26. Y!lc5
1 6. Ac4 c6 1 7. Axd5 Y!le7 1 8 . h3 b4 The position has stabil ised . The weakness
The alternative was 1 8 . . . tDd4 1 9. tDxd4 of the e5- pawn is compensated by the
exd4 2 0 . f! d 3 (20 . 0 - 0 ? f3) 2 0 . . . Af5 weakness of the e4-pawn.
21 . 'l!flxb7 'l!fte5! with good compensation 26 . . . Ae6
for the pawn. For example: 22. 0-0 (22. f!f3 26 . . .'l!ff xe4 is also possible, but Black has to
g5 looks u n natu ral) 22 . . . f!ab8 23. 'l!ff xa7 withstand some initiative by Wh ite: 27. Ac3
Axe4 24. Axe4 'l!flxe4 25. f!a3 f!a8 etc. Ab7 28. Axe5 Axe5 29. f!fe1 f!bc8 30. Wa3
1 9. Ac4 f!a8 31 . Wb3 Wa4. From the practical point
Carlsen avoids the exchange of his ac of view, with ti me-trouble approaching, the
tive bishop, overloo k i n g B l ack's rep ly. move in the game is safer.
Meanwhile, after the simple 1 9. 0-0 tDxd5 27. Y!ld6 Ab3 28. b1 Y!lxe4 29. Ac3
20. exd5 'l!ff f 7 21 . f!fe1 (otherwise there fol a b c d e
lows the blockad ing tDf3-g5-e4) 2 1 . . . Wf5
8 I.
22. d6 this pawn causes Blac k consid
erable problems. For exam ple: 22 . . . f!b8
7
23. 'l!ftc3 (23 . Ad2 e4 24. Ac3 ! exf3 25. f!e7 6 6
is tempting, but Black can play 23 . . . Ae6) 5 .t. 5
23 . . . Ae6 24. f!xe5! Axe5 25. tDxe5 <;!;lg8 4 'if .t. 4
26. d7 Wf6 27. 'l!ff c7. 3 .I. it t2J 8 3
a b c d e f 9 h 2 88 2
8 I. 1:. 1:.
'iV .i. .t. 7 a b c d e f 9 h

6 .t. 6 29 . . . bd8?!
5 5 It is from such trifles that the precond itions
for an attack are created . If one looks at the
4 it 8.t. 4
d ifference in the defences of the kings, it
3 t2J 8 3 is obvious that after exchanges on e5 the
288 88 2 black king w i l l to subject to a 'draught'
it 1:. 1:. 1 along the adjacent a1 -h8 and a2-g8 di
a b c d e f 9 h ag o n a l s , w h i c h i s d a n g e ro u s with the

2 06
 M. Carlsen - L. Do m i nguez
---- ---

queens on the board . M eanwh i l e , after Over the last three moves the situation has
the simple 29 . . . dS the th reat of . . . eS-e4 changed sharply.
wou ld have forced Wh ite to exchange the
32 . . . .td3?
queens - 30. AxeS AxeS 31 . xeS+ xeS
32. tDxeS AdS with an i m m i nent d raw. The bishop s h o u l d h ave been u rg e ntly
30. 'Wa3 Ac2? switched to the a2-g8 diagonal - 32 . . . Ae4
Taking into accou nt what was said earl ier, 33. AxeS a? 34. tDgS AdS 3S . .!;d1 h6
the bishop should have been kept on the 36 . .!;bxdS ,l;xdS 37. .!;xdS hxgS. Black has
a2-g8 d i agonal, for example 30 . . . AdS. a broken pawn structure, but the win is sti l l
Carlsen i m med iately exploits Black's m is a l o n g way off. One g a i n s the i m pression
take. that Dom i n g uez fai ls point-blank to see
31 . gb5! 'Wa4? the vital d iagonal.
Another inaccu racy. The q ueen coord i
33. gb7! Wc2
nates better with t h e other pieces after
31 . . . c6 , and alt h o u g h W h ite can sti l l There is no longer any way of saving the
torment h i s opponent, there i s n o d i rect game. T h i n g s are hopeless after both
win: 32 . .!;xeS!? (or 32. b2 c8 33. AxeS 33 . . . Axf1 34. AxeS .!;d7 3S. Axg7+ and
AxeS 34. tDxeS c3 3S. a2 .!;c7) 32 . . . a8 33 . . . .!;d7 34. CD xeS AxeS 3S. AxeS+ g8
33. b2 Ae4! (after 33 . . . .!;fc8 34. Ad4 36 . .!;xd7 xd7 37. b3+ ,l;f7 3 8 . .!;c1 ,
Wh ite is domi nant on the long d iagonal : when the th reats of .!;c7 or Ab2 and c3 ,
34 . . . .!;ab8 3S . .!;e7! ,l;xb2 36 . .!;xg7 with with an attack on the long diagonal, are
the win of a pawn - 37. .!;xg6) 34. tDd4 irresistible.
b7 3S . .!;bS f7 etc.
34. 'Wb4 gfe8 35. ge1
32. Wb2
a b c d e 3S. AxeS! AxeS 36. tDxeS .!;xeS 37. xf4
8 .I was more energetic .
7 35 . . Ae2
6 6
3S . . . Aa6 36 . .!;a? .!;a8 37. AxeS wou l d
5 :: 5
merely h ave delayed B l ack's i n evitable
4 'iV .l 4 defeat.
3 J, 3
2 .i. 2 36. xe5 Axe5 37. Axe5+ gxe5
38. Wxf4 Wf5 39. Wh6
:
a b c d e f 9 h Black resigned .

I n Linares (FI DE category 2 1 ) Carlsen a s though by inertia contin ued his ' Wij k' drawing
series and began the tournament with five draws. But the essence is not i n the n u mber
of draws - in such stel lar com pany this is natural - but i n the qual ity of the play. And
here open ing preparation comes to the fore. G iven the modern state of theory, it is only

2 07
 Chapter 3
Life at the Top

with the wh ite pieces that it is realistic to hope to win against a player of equal class.
But this demands pai nstaki ng work and searches in the labyri nths of theory, which is
not to Magn us's taste. Typical, for example, is this ad m ission in one of his interviews:
'I have never l i ked studying opening theory. For me a normal position is sufficient. If
I have an advantage in an end i n g , I can play for a long time, trying to force the oppo
nent to make a mistake.' But in su per-tournaments it is na"lve to rely on players making
mistakes. And the deficiencies of Carlsen's opening preparation were seen already in
the second round, when with Wh ite he was obliged to force a d raw in his game with
Doming uez after m issing a trick.

1. c4 c5 2. tbf3 tbc6 3. d4 cxd4 two successive losses, to Aro n i an and

4. tbxd4 tbt6 5. tbc3 e6 6. g3 'Wb6 Wan g Vue, in w h i c h , p l ay i n g Wh ite, he
7. tbdb5 tbe5 8 . .tf4 tbfg4 9. 'Wa4 (9. e3 not only fai l ed to extract anything from
is more often played) 9 . . g5! 10 .txe5.
. the open ing, but even ended up in i nferior
positions. And only an i m portant win in
a b c d e f g h
the 1 Oth round over the l eader G risch uk
8 J. .i. .a .1. 8 enabled h i m to fi nish the tournament wor
7 7 thily in 3rd place.
6 if
5 ttJ
J
iL
6
5 ,!
4 [J 1.& 4 Game 47
3 ttJ [J 3 M . Carlsen - V. Anand

2 /J /J [J[J [J 2 Linares 2009

1 1:, iL 1:, 1 Semi -Sla v Defence [ 045]

a b c d e f g h
1 . d4 d5 2. c4 c6 3. tbc3 tbt6 4. e3 e6
5. tbf3 tb bd7 6. 'Wc2 .td6 7. g4 tbxg4
1 0 'Wxf2+! (in the game Aron ian - Leko,
. .

8. lag1 'Wf6 9. laxg4 'Wxf3 1 0. laxg7 tbf6

Morel ia/Linares 2008, 10 . . .tL!xe5 was
1 1 . h3
played) 1 1 . d1 tbxe5 1 2 . tbc7+ d8
1 3. tbxa8 'Wd4+ 14. c2 tbxc4, and
a b c d e f g h
Carlsen went i n for a repetition of moves:
1 5. b3 tbd2+ 1 6. c2 tbc4 1 7. b3
8 .1. .i. .1. 8
tbd2+ 1 8 . c2 112-112.
7 :7
6 .*-'.& 6
However, i n the 6th round M agnus nev 5 5
ertheless managed to ach i eve what he
4 /J/J 4
wanted - he gai ned a m i n i mal endgame
advantage and i n a d rawn-out game he
3 ttJ [J 'iV [J 3
defeated Anand for the fi rst time. But this 2 [J [J 'Jfi [J 2
proved to be a ' pyrrh ic' victory. Magnus 1 : iL iL
began the second half of the event with a b c d e f g h

2 08
 M. Carlsen - V. Anand

Beg i n n i n g a chase of the q u e e n - l:!gS a b c d e f g h

and Ag2 i s th reaten e d . I f Wh ite begins
with 1 1 . l:!gS, t h e n 1 1 . . .l e 4 1 2. tD x e 4
(1 2 . Ag2 'Wf6) 1 2 . . . d x e 4 1 3 . Ag2 'Wf6
14. l:!g4 eS is possible, with roughly equal
chances.
11 . . . Wf5
It is hard to explain Anand 's d ecision to go
into an obviously i nferior endgame (possi
bly the reasons were of a personal natu re),
where the presence of the pawn islands
dS, fS , f7 and h7 condemns B l ac k to a
cheerless d efence. Apparently he doubted
whether it was possible to break open such
a fortress. But it is wel l known that Carlsen
is not afraid of hard work, and h e is ready
to play an endgame with a stable positional
advantage.
The play was more l ively in the game
Aro n i a n - M o rozev i c h ( M o re l i a/Li n ares
2007): 1 1 . . . h 6 (preventing l:!gS with the
th reat of tra p p i n g t h e q u ee n by Ag2)
12. Ad2 eS 1 3 . g3 'WhS 14. cxdS exd4
1S. tDe4 tDxe4 1 6. 'Wxe4+ 'WeS 1 7. 'MixeS+
AxeS 1 8. f3 cxdS 1 9. AbS+ e7 20. exd4
Ad6 (20 . . . Axd4? 21 . Ab4+ e6 22. 0-0-0
AeS 23. l:!e3 leads to the loss of the bishop)
21 . f1 Ae6 with equal chances.
1 1 . . .eS is risky: 1 2 . dxeS AxeS 1 3 . l:!gS
tDd7 1 4 . cxdS cxdS 1 S. l:!fS with advantage
to White, Tisdaii- Sidselrud (Norway 2002).
12. Wxf5
Of course, if 1 2 . Ad3 Black w i l l happily
capture the pawn - 12 . . .'Wxh3.
1 2 . . . exf5 1 3. cxd5 cxd5
The exchange of knights came into consid
eration: 13 . . .tDxdS 1 4 . tDxdS cxdS 1 S . Ad3
e7 followed by . . . Ae6 and . . . l:!ag8.
14. t'Ub5 Ab4+ 1 5. Ad2 Axd2+
1 6. xd2 e7 1 7. Ad3 Ae6
.!,a
.t.:t.t. 7
6 -*.'.& 6
5 tZJ ... ... 5
4 4
3 3
2 2
1:1
a b c d e f g h
1 S. t'Uc7!?
An indication of a deep unde rstand i ng of
the endgame . Few would have conceived
the idea of exchang ing the knight for the
' bad ' bishop. But Carlsen observed that
the bishop was fulfilling the i m portant task
of d efending the fS- and dS-pawns. For
the solving of a strateg ic obj ective Wh ite
uses a tactical feature of the position: the
presence of the rook on g7 forces Black
to capture the knight with his king , which
will be tied to the d efence of the fS- and
dS-pawns.
1S . . . laagS 1 9. t'Uxe6 xe6 20. laxgS
t'UxgS
Black does not want to concede the c-fi l e
after 20 . . . 1:!xg8 21 . l:!c1 .
21 . e2
If 21 . l:!c1 there follows 21 . . . tiJe7 22. l:!c7
l:!b8 and then . . . d6.
21 ... t'Ue7 22. f3 lacS 23. a4 lac7
24. a5 h6 25. h4 f6 26. h5 t'UcS
27. f4
While Black has been forced to adopt a
waiting position, Wh ite has been seizing
space.
27. . . t'Ud6 2S. lag1 lacS 29. f3 e6
It was hard to decide on the further weak
ening 29 . . . a6, but this was possibly the

2 09
 Chapter 3 Life at the Top

only way of not allowing the roo k i nto 36. g1 !

the rear. Wh ite wou l d h ave had to look
Wh ite g ives his opponent the move (com
for other ways to strengthen his position:
pare this position with the one after Black's
30. .e1 (both 30. Ab1 e6 and 30. fta1
33rd move). Si nce Black cannot concede
.g8 31 . .a3 .g5 are insufficient) 30 . . . ftg8
the c-fi le, he is forced to move his knight.
31 . Ac2 t2Jc4 (weaker is 31 . . . ftg2 32. Ab3
. x b2 33. A x d 5 . b 5 34. e4) 32. A x f5 36 . . . e8
(32 . b4 .g5 33 . .h1 e6 34. Ad3 tDd6)
In the event of 36 . . .t2Jc4 Black has to reck
32 . . . t2Jxa5 33 . .c1 tD c 6 3 4. .c3 tDb4
on with the possibility of White transposing
(34 . . . ftg2 35. Ac8) 35. Ag4 etc.
into a favou rable rook endgame: 37. Axc4
30. g7
. xc4 (or 37. . . d xc4 38 . .g7 .h8 39. e4)
With the threat of .h7. 38 . .g7 .a4 39. b4 . x b4 40 . . h7 .a4
30 . . . h8 41 . .xh6+ e7 42 . .h8 .xa5 43 . .b8. He
a b c d e f g h can also consider 37. e4!? fxe4 38. fxe4,
8 .Ia putting the d5-pawn under pressure, since
38 . . . d xe4? fai ls to 39 . .c1 , transposing
:S 7
i nto a won pawn e n d g a m e : 39 . . . d5
6 ,.. . 6 40. Axc4+ .xc4 41 . .xc4 xc4 42. xe4.
!:, 5 At the same time the manoeuvre Ab3-d1-
4 !:, 4 g4+ is th reatened .
3 8 !:, 3
37. e4!
2 !:, 2
White changes the pawn structure, gaining
a b c d e f g h access to the h-pawn .

31 . Ac2! 37. . . fxe4 38. fxe4 f6 39. e5 e4

A subtle manoeuvre of the bishop to b3, In the event of 39 . . .ttJxh5+? 40. e3 Black
creating the threat of e3-e4. could have lost his knight.
31 . . . c8 32. Ab3 h8 33. g1 c8
40. e3
34. g7 h8 35. g2 c8
a b c d e f g h a b c d e f g h

8 .I 8 8
7 7
6 ,.. . 6 6
5 !:, !:, 5 5 8 8
4 !:, 4 4 8,..
3 !:,!:, 3 3
2 !:, :s 2 2 !:,
:s
a b c d e f g h a b c d e f g h

210
 M . Carlsen - V. Anand

40 . . . b6 e6 53. h6 was stronger, when the h-pawn

must decide the game.
If 40 . . . ttlg5 Black has to reckon with the
th reat of sacrifi c i n g the exc h a n g e fo r 50 . . . Elh2+ 51 . d1 c4 52. Elxf6+ e7
two pawns: 41 . .f1 ttle4 42 . .xf7!? xf7 53. Ag6 Eld2+
43 . Axd5+ e7 44. Axe4! (44 . A x b7 .b8 a b c d e f g h
45. Axe4 .xb2 is i nsufficient) 44 . . . .c1
8 8
(44 . . . .c4 45. d5 .a4 46. f4) 45. d5 and
the pawns, 'taken by the hand', advance
7 7
toward s the q u e e n i n g s q uare: 45 . . . .f1 6 :s 6
(or 45 . . . b6 46. axb6 axb6 47. f4 .f1 + 5 .l8 8 5
48. Af3 d 7 4 9 . d 6 e6 5 0 . b4! b5 4 8 4
51 . e4 and the king goes to c5) 46. d6+ 3 3
e6 47. Af3 .e1 + 48. d4 .g 1 49. Ad5+ 2 8 .!. 2
d7 50. Axb7 etc.

41 . axb6 axb6 a b c d e f g h
Another positional achievement by White: 54. c1
his rook obtains new manoeuvri ng pos 54. e1 was more accu rate: 54 . . . f!xd4
sibilities. 55. h6 .h4 56. Af5 d4 57. .e6+ f7 58. b3!
42 . d3 f2+ 43. e2 e4 44. e3 ttle3 59. f!f6+ e7 60. Ac8! f!h2 61 . f!d6
e8 (in the event of 61 . . . f8 62. h7 White
Again giving the opponent the move by exchanges the rooks and w i n s the d4-
the 'triangulation' m ethod. Black has to pawn) 62. Aa6 ttlf5 63. Ab5+ e7 64. f!xb6
decide whether to allow the rook into his ttlxh6. Wh ite wins by coord i nating the ac
rear or block the g-fi le. tions of his pieces and the e-pawn: 65. Ac4
ttlf5 66. f!b7+ e8 67. Ab5+ f8 68. e6
44 . . . f6
ttle7 69. Ac4.
44 . . . ttlg5 was more resil ient: 45. .a1 (here
54 . . . Elxd4 55. b3 xeS 56. Elxb6 Elh4
45. .f1 ttle4 46. .xf7 xf7 47. Axd5+ e7
57. Ats
48. Axe4 e6 is now insufficient) 45 . . . f6!?
(weaker is 45 . . .ttle4 46. .a6 .b8 47. Ad1 a b c d e f g h

with the threat of Ag4+) 46. exf6 ttle4 47. f7 8 8

.f8 48. .a6 .xf7 49. .xb6+ ttld6, setting
up a new defensive line.
45. Elg6 Elc1 46. Elxh6 Elh1 47. Ac2
Elh3+
47. . . .e1 +! 48. d3 .h1 was stronger.
48. f4 Elh4+ 49. f3 d2+ 50. e2
7 7
6 6
5 8 5
4 .!. 4
3 3
2 2

50. g3! .xd4 51 . .xf6+ xe5 52. .f5+ a b c d e f g h

211
 Chapter 3
Life at the Top

57 tbf3?
. .
a b c d e f 9 h

Th is was obviously not one of Anand's best 8 I. ... . 8

days. By 57. . . .lah1 + he could have hindered 7 .t 7
.t .i..t.t.S..t
6 6
- . =-

the advance of the h-pawn . 58. b2 tDc4+!

59. bxc4 .!axh5 60. Ag6 .lah2+ 61 . c3 dxc4 .-.t .til,
5 5
leads to the theoretically drawn ending rook
4 CiJ B 4
and bishop against rook. After 58. c2
tDf3 59. Ag4 tDg5 Black somehow holds
3 il,CiJ B 3
on, and without the passed b-pawn Wh ite 2 BB BW/ BB 2
cannot get by. 1 M M1
a b c d e f 9 h
58. h6 tbd4 59. h7!
A wel l -known tech nique. Now the h-pawn away the knight by . . . e7-e5. I n this case
will cost the exchange. the d6-pawn is not so vul nerable, since
59 . . . tbxf5 60. b8 tbd4 61 . mb2 md6 after . . . b7-b5 Black has the multi-purpose
62. h8'W xh8 63. xh8 manoeuvre . . . .!ab6, when the rook defends
The game enters the technical conversion the pawn while also taki ng part in an attack
phase. on the king ( . . . .!aa6, . . .'<WaS). True, all the
same the shadow of tDd4-f5 hangs over
63 mc5 64. h5 tbc6 65. h4 tbb4
the king's position.
. .

66. ma3 d4 67. h5+ tbd5 68. mb2

mc6 69. ma3 mc5 70. h4 tb b4 12 . . .txh6
71 . h8 tbc6 72 . h5+ md6 73. b4 d3
It is useful to divert the q ueen from the
74. h3 tbe5 75. mb3 d2 76. mc2 tbc6
centre - in the event of the wing attack
n. ah4 md5
h2-h4-h5 there is the reply . . . g6-g5.
Black resigned .
1 3. 'Wxh6 b5 14. g4

It is an eternal di lemma in the Dragon- with

which pawn to beg i n the assau lt? White
Game 48
beg ins with the g-pawn , not al lowing 14 . . .
L. Dominguez - M . Carlsen
e5 i n view of 1 5. tDf5 ! , but now Black suc
Linares 2009
ceeds in i n itiating cou nterplay.
Sicilian Defence [ 8 78]
The alternative is 1 4 . h4 e5!? 1 5 . tDde2
1 . e4 c5 2. tbf3 d6 3. d4 cxd4 4. tbxd4 b4 1 6 . tDd5 tDx b3+ 1 7. axb3 (there is no
tbt6 5. tbc3 g6 6 .te3 .tg7 7. f3 tbc6
particular poi nt in 1 7. cxb3 tDxd5 1 8. .xd5
8. 'Wd2 0-0 9 .tc4 .td7 1 0. 0-0-0 la bS
Ae6 19 . .!ad2 .!ab6, when apart from the
1 1 . .tb3 tba5 1 2 . .th6 attack with . . . .laa6 and . . .'<Was Black has
the d istant prospect of a favourable end
(see next diagram)
game, L:Am i - Tiviakov (Leeuwarden 2004)
The exchange of bishops deprives . . . tbc4 1 7. . .tDxd5 1 8 . .!axd5 Ae6 (1 8 . . . .!ab6 is also
of m u c h of its strength . But it also has good , with the idea of 1 9. h5 g5!) 1 9. .d3
drawbacks: in certai n l i nes Black can drive (1 9. h5 is not dangerous in view of 19 . . . We?!

212
 L. Dom i n guez - M. Carlsen

20. hxg6 fxg6 21 . .xd6 .fc8 22. c4 bxc3 18. !a he1

23. eDxc3 . b 6 with a good game for
Black, Lasti n - B . Savchenko, Sochi 2007) Abandoning the flank strategy - 1 8 . h4 e5
19 . . . .b6 20. f4 Ag4 21 . h5 V!/c7 22 . .g3 1 9. dxe6 fxe6 20. h5 g5 .
.c8 with chances for both sides, Bolo
1 8 . . . e5 1 9. d xe6 fxe6
gan - B . Savchenko (Sochi 2007).
a b c d e f 9 h
14 . . . xb3+
8
It is pointless playing 14 . . . eDc4 1 5 . h4 e5?
7
1 6. eDdxb5.
6 .!.
15. xb3 5 5
a b c d e t 9 h
4 ::, 4
8 .!. 3 tZJ ::, 3
7 2::,::,::, ::, 2
6 1::r1::r
5 5 a b c d e f 9 h

4 4 Objectively speaki ng, Black's position is

3 already the more prom ising. Wh ite's f3 -
::, 2
1::r 1
a b c d e f 9 h
With this move Wh ite effectively abandons
the potential th reat of eDd4-f5, for the sake
of which g2-g4 was played . 1 5 . axb3!?
b4 1 6 . eDce2 came i nto c o n s i d e rati o n ,
and now i f 1 6 . . . e5?! there is t h e good re
ply 1 7. tDf5 ! g xf5 1 8 . g xf5 h8 19 . .hg1
.g8 20 . .xg8+ eDxg8 2 1 . Wxd6 and then
22. V!/xe5, obtai n i n g three pawns for the
knight. Apparently the best for Black is
16 . . . .b6, defending the pawn beforehand
in the event of 1 7. h4 e5 1 8 . eDf5 g xf5
1 9 . gxf5 h8 20 . .hg1 .g8.
1 5 . . . b4 1 6. d5 xd5 1 7. exd5 !a b6!
The best square for the 'Chinese' rook.
From here it is ready for an attack on the
king, and in the given specific case it also
su pports the . . . e7-e5 advance.
pawn is weak, and at the same time the
d 6/e6 'small centre' is ready to take up
more active positions. I n addition, an at
tack on the king looks more real for Black
than for Wh ite.
20. !ae3 !af7
Al exey Kuzm i n reco m m e n d s 2 0 . . . e5!?
21 . h4 V!/c7 22. h5 Ae8 with good play for
Black .
21 . d2 d5 22. b3
The e6/d5 pawn pair - which is usually a
headache for Black i n the classical varia
tion of the French Defence - does not con
stitute a problem here, since it can not be
blockaded by the knight. On the contrary,
in a number of variations it can become
mobile. For exam ple, White can not fix this
pair with 22. f4. By playing 22 . . . V!/f6 23 . .f3
Ab5 (with the th reat of . . . Ae2) 24 . .f2 e5
Black breaks through i n the centre.

213
 C hapter 3 Life at the Top

22 . . . Wc7 23. b1 la bS 24. lade1 26 . . . d4!

It is obvious that Wh ite's play on the king Black clears the approaches to the king.
side h as come to a standst i l l , whereas
27. lae5 d3 28. cxd3 laxf3 29. d4?
Black's threats on the opposite wing are
real. Therefore it was logical to bring back Carlsen's e n ergy has had its effect.
the q u een to support the other pieces: Doming uez, as though hypnotised , watch
24. h4 .!::! c 8 25. f2, going onto the de es the opponent's actions. The q u een
fensive. should have been u rgently brought back
to the defence - 29. d2 .
24 . . . lac8
2 9 . . . A b 5 30. la2e3 Ad3+ 31 . a1
W h i l e W h ite has been s ee k i n g t h e re-
q u i red regrou p i n g , B l ack has taken firm
a b c d e f g h
possession of the c-fi l e and is ready for
8 8
action.

25. la 1 e2 Wb6

f g h 5 5
a b c d e

8 8 4 I.
.i
3 tLJ 3
7 I. j_ I. 7 .i
6 'iV 2 fj, fj, 2
I. I. Vjf 6
5 5 1
I. a b c d e f g h
4 I. fj, 4
3 tLJ fj, 3 31 .. Wxd4!?
2 /j,/j,/j, fj, 2
The conseq uence of the back ran k be
ing u ndefended . It is hard to resist such
a b c d e f g h
a move, but the ruthless Rybka poi nts out
26. h4 that 31 . . . .!::!f2! , with the same th reat, is more
effective. 32 . .e1 allows a mating attack:
A pointless move. The advance of the d 32 . . . .!::! x b2! 33. xb2 .c2+ 34. a1 c6!
pawn should have been prevented , but this 35 . .c5 . x a2 36. x a2 a4+ 37. b2
is not easily done. For example, after 26 . .!::!d 2 a3#.
.!::!c4 27. .!::!e d3 .!::!cf4 Black switches to a siege
32. laxe6 laf1 + 33. lae1 Wxg4
of the f3-pawn - 28. d4 c7 and prepares
. . . e6-e5 (the direct 28 . . . e5 29. e2 leads With the th reat of 34 . . . d1 + .
to simplification of the position), retaining a
34. laxf1 Wxe6 35. c5 We2 36. lac1
serious initiative.
At5
Possibly the best for White was 26. c1 ,
setting up a distant defensive line: 26 . . . .!::!cf8 White has somehow pl ugged the holes in
(26 . . . d4) 27. h4. his position. But Black has an extra pawn

2 14
 M. Carlse n - A. Grischuk

and prospects of a pawn storm against 49. 'We3 E[e8 50. 'Wc3 E[e2
the king's residence. 50 . . .Wd4! was more forcefu l .
37. Wf4 aS 38. h 5 51 . b 3 E[xc2 52. Wxc2 We5+

With the faint hope of 38 . . . YNxh5? 39. YNc4+ The trap 52 . . . YNxb3?? 53. YNg 6+ is only
c;!>f8 40. Wd4. for beg i nners.
53. b1 g7 54. Wd2 xb3
38 . . We7 39. Wc4+?
Wh ite resigned .
In time-trouble Dominguez fails to exploit
his last chance. The preparatory opening
of the h-fi le by the exchange 39. hxg6 hxg6
would not have done Wh ite any harm . After Game 49
40. Wc4+ the reply 40 . . . Ae6 is now insuf M. Carlsen -A. Grischuk
ficient in view of 41 . 'Wd4 with the threat Linares 2009
of h1 . Black would have had to restrict Sicilian Defence [ 885]
himself to 40 . . . g7 41 . Wd4+ Wf6, red uc
ing the tempo of the attack. However, this 1 . e4 c5 2. f3 d6 3. d4 cxd4 4. xd4
would not have saved the game for Wh ite. f6 5. c3 a6 6 .te2 e6 7. 0-0 .te7

8. a4 c6 9 . .te3 0-0 1 0. f4 Wc7

39. .te6 40. Wc2 Wg5! 1 1 . h1 E[e8 1 2 . At3 At8 1 3. Wd2 E[b8
a b c d e f 9 h 14. Wf2
8 .I 8
This variation was wel l known to G rischu k
7 7
from h i s Candidates match with Rublevsky
6 6 (Eiista 2007), where he upheld the wh ite
5 . ttJ 'i' 5 side. In the 8th game of the match Grischuk
4 4 played 1 4 . ad1 and after 1 4 . . . e5 1 5 . tDde2
3 3 b5 1 6. axb5 axb5 1 7. f5 b4 1 8 . tDd5 tDxd5
--.-

1 9. Wxd5 Aa6 20. 'Wd2 tDd4 21 . Axd4 exd4

2
22. fe1 Axe2 23. Wxe2 Ae7 24. xd4 Af6
25. c4 WaS White had the better chances.
a b c d e f 9 h
a b c d e f 9 h
Now Wh ite is tied , as they say, hand and 8 .Ij.J .111.. 8
foot. Dom i n g uez could h ave cal m ly re
signed, but he was as though in a lethargic
7
.'i' 7

sleep. The rest is all Black.

6 . .& .& 6

5 5
41 . hxg6 hxg6 42. a3 bxa3 43. Wc3
4 tt:J 4
A piece is lost after 43. bxa3 'We5+ 44. 'Wb2 3 ttJ JtJt 3
xc5. 2 VIIi 2
43 axb2+ 44. xb2 Wd5 45. E[c2 a4
.
1.: .: 1
46. a1 a3 47. We3 .tf7 48. Wc3 g5 a b c d e f 9 h

15
 Ch apter 3 Life at the Top

14 . . . e5 f8, and leaving Wh ite with a bad bishop on

After 1 4 . . . Ad7 1 5 . g4! e5 White gains an
important tempo for the attack on the king
side. The game Carlsen - Anand (Wij k aan
Zee 2008) conti nued 1 6. tDf5 exf4 1 7. Axf4!
Ae6 18 . .ad1 tDe5 1 9. AxeS d xe5 20. g5
tDd7 21 . tDd5 Wc6 22. Ag2 Wc5 , and here
Magnus launched an attack on the king:
23. Wh4 Wxc2 24 . .c1 Wxa4 25. b3!?
K K..te
A
tjj tjj::,
'iV !::,
!::,
!Jl,::,
: :
Analysis diagram
25 . . . WaS? 26 . .c3 g6 27. .h3 h5, and here
instead of 28. Af3 Axd5! 29. exd5 Ag7, after
which Black was able to defend, according
to analysis by Maxim Notkin Wh ite could
have concluded the game in spectacular
fashion: 28. tDg3! ! (with the th reat of tDxh5)
28 . . . Ag4 29. tDxh5! g x h 5 (or 29 . . . Axh5
30. Wf2 f5 31 . g xf6 .e6 32 . . x h 5 gxh5
33. Ah3) 30. Wf2! f5 (30 . . . Ae6 31 . f!xh5+)
31 . g xf6 and wins.
It remains to add that, in the diagram
position, instead of 25 . . . WaS? Black could
f3 . However, the move has some poi nts as
well: the bad bishop might turn out to actu
ally be very strong after a su bsequent tDd5
tDxd5 exd5, and the possibility of creating a
passed pawn and a dangerous queenside
majority is generally dangerous for Black.'
1 5 . . . dxe5
The capt u re with the k n i g ht 1 5 . . . tDxe5
l eads to loss of m aterial after 1 6 . Ag5!
Ae7 1 7. Axf6 Axf6 1 8. tDd5 Wd8 1 9. Ah5.
1 6. b3
This is where a subtle point of the pawn
exchange on move 1 5 is seen. If instead of
1 5. fxe5 Wh ite had retreated 1 5. tDb3, then
after 1 5 . . . b5 1 6. axb5 axb5 1 7. tLld5 tDxd5
1 8. exd5 exf4 Black would have obtai ned
the e5-square. But now 1 6 . . . b5 wou ld be a
blan k shot - 1 7. axb5 axb5 1 8. tDd5 tDxd5
1 9. exd5, and the knight is driven to a less
good post.
1 6 . . . b4 1 7. j,a7
A wel l-known technique: before placing his
bishop on b6 Wh ite d rives the rook to its
initial square, removing a defender of the
b7- pawn, which may be of i m portance in
some variations.
1 7 .. J;!:a8 18. j,b6 'We7 1 9. Elad1
a b c d e f 9 h

have seized the in itiative with a q ueen sac 8 K j_ K.t.e 8

rifice: 25 . . . Wxb3! 26 . .c3 Axd5! 27. .xb3 7 .l 'if .l .l .l 7
A x b3 28 . .f3 .bc8! 29 . . x b3?! .c1 + 6 .l 1.1\ 6
30. Af1 .xf1 + 31 . g2 .f4! (Notkin). 5 .l .,L---J 5
1 5. fxe5 4 Bl.l\ B 4
Carlsen, as usual, is concrete in his deci
3 ti:Jti:J 3
sions: 'At fi rst sight it looks a bit i l logical, 2 BB VJ!I B B 2
letting Black get rid of his backward pawn I;! I;!
on d6 and opening up for his bishop on a b c d e f 9 h

2 16
 M. Carlsen - A. Gri schuk

19 . . Ae6?! the placing of his pieces and prepare a

A routine move, which leads to Wh ite cre breakthrough on the q ueenside.
ating a dangerous passed pawn. 'Safer
26 . . tUxb6 27. axb6 flab8?
was 1 9 . . . Ag4! , which looks paradoxical ,
as Black offers to exchange his "good" Grischuk returns the favou r. H e should not
bishop for my "bad" bishop, but actually it have clung on to the pawn - the position
makes a lot of sense, as the key move now could have been held by the activation of
will be less dangerous. Wh ite sti l l keeps a the rook: 27. . . l;!.ec8 28. b4 l;!.c6 29. d7 Axc5
slight initiative, but Black has a perfectly 30. bxc5 (30 . d 8+ is also i n s ufficient:
decent position.' (Carlsen) 30 . . . l;!.xd8 31 . l;!.xd8+ Af8 32. Axa6 l;!.xb6)
20. tUd5 Axd5 21 . exd5 e4 22. d6 30 . . . l;!.d8 31 . Axa6 l;!.xc5 (31 . . . bxa6 32. b7
'We6? leads to the loss of the exchange) 32. Axb7
It is wel l known that the q ueen is a poor l;!.b5, picking u p the pawns.
blockader, and 22 . . . d7 23. tbc5 f5 a b c d e f 9 h
wou l d have led to a position from the J...te 8
game, but why present the opponent with
.l.l.l7
a tem po? 22 . . . e5 23. d7 tbxd7 24. l;!.xd7 6
exf3 25. xf3 f6 26. tbd4 e4 was the best

way out of the situation.
5
4
23. tUc5 'Wf5 24. Ae2 'Wxf2 25. flxf2
tUbd5
3
a b c d e f 9 h

8J. J...t 8 a b c d e f 9 h
7 .l .l.l.l7
6 t::o 6 28. flxf6!
5 tb 5 Clearing the barrier in front of the passed
4 t::o .l 4 paw n . Carlsen confidently converts h i s
3 3 advantage.

28 gxf6 29. tUd7 15 30. c4 a5 31 . c5

. .

a b c d e f 9 h
Ag7 32. tUxb8 flxb8 33. Aa6! Af6
34. Axb7 flxb7 35. c6 flxb6 36. flc1 !
26. a5?
Throwing away the fruits of his previous The final subtlety - after 36. c7? l;!.c6 Wh ite
work, whereas after 26. l;!.xf6 tb xf6 the would have lost.
move 27. a5 would have practically placed 36 . Axb2 37. d7
Black in zugzwang . It is hard for him to find
a move, whereas White can calmly im prove Black resigned .

2 17
 Chapter 3
Life at the Top

Carlen's fi rst appearance at the Mtel Masters super-tou rnament in Sofia (FI DE category
2 1 ) was reassuring. In the very fi rst round he defeated the main tournament favourite
Topalov. In the second half Magnus gained add itional wins over Dom inguez and Wang
Yue and was wearing the yellow jersey going into the last rou nd, half a point ahead of
Shirov and Topalov. The fate of the fi rst prize was decided in the Shirov - Carlsen game.
It was sufficient for the Norweg ian to make a draw, and it appeared that the game was
fol lowing a perfected scenario: he played his favourite Chelyabinsk Variation, where he
had the reputation of being invincible. But it was here that a serious hole in Magnus's
open ing preparation was revealed - he fel l into a prepared line by Shirov and lost. As
a result, the fi rst prize was won by Shirov, and Carlsen had to be satisfied with a share
of 2nd-3rd places with Topalov.

Game 50 venting the squeezing of Black's position,

M. Ca r l s e n - V. Topalov occurred in the rapid game Carlsen - Kar
Sofia 2009 jaki n (Nice 2009), which contin ued 1 2 . d5
Semi -Sla v Defence [ 043] b6 1 3 . Ab3 Ag4 1 4 . f!c1 ! (in the event of
1 4. h3 Axf3 1 5 . V!Jxf3 V!t!xf3 1 6 . gxf3 e7
1 . d4 d5 2. c4 c6 3. f3 f6 4. c3
Black central ises his ki ng) 1 4 . . . 0-0 1 5 . h3
e6 5 . .tg5 h6 6 . .txf6 Wxf6 7. e3 d7
Axf3 1 6. V!Jxf3 V!Jxf3 1 7. gxf3 gfd8 1 8. gfd1
8 .td3 dxc4 9 . .txc4 g6
.

Af6 1 9. d xc6 bxc6 20. gxd8+ f!xd8.

a b c d e f 9 h
.!
8 .!. ..t ..t .!. 8 .l .l
7 1.& 7 .i..l
.l
6 'i 6 fj,
5 5 fj, fj,
Jblb
4 JL 4 fj, fj, fj,
3 ttJ ttJ 3
2 2 Analysis diagram
1 .s: .s: 1
a b c d e f 9 h I t u s e d to b e t h o u g ht t h at t h e e n d
g a m e ari s i n g i n t h i s variat i o n was n ot
The o l d variation 9 . . . Ad6 1 0. 0-0 V!Je? dangero u s for B l ac k . H owever, it is by
1 1 . e4 Ac7 1 2 . f!c1 , where Black has a no means easy to m ake a d raw. Wh ite
cram ped position, has hardly been played has a s m a l l advantag e , n ot o n l y thanks
in recent times. to h i s s u p e r i o r pawn structu re (wea k
pawns o n a ? and c 6 ) , but also the fa
1 0. 0-0 .tg7 1 1 . e4 0-0
vourable opposite - c o l o u r b i s h o ps : the
Castling usually leads to more complicated l i g ht-squ are bishop ties t h e b l ack k i n g
play, which is in accordance with Topalov's to t h e defence o f t h e f7-paw n , where
style. Another critical branch , 1 1 . . . e5, pre- as t h e dark- s q u are bishop i s passive.

2 18
 M . Carlsen - V. To palov
Carlsen h a p p i l y p l ays t h i s type of e n d
i n g : 2 1 . ti":ld 1 .d 6 2 2 . . c 5 f 8 2 3 . f1
h5 24 . ti":le3 e7 ? ! (th e b i s h o p s h o u l d
have b e e n freed fro m h av i n g t o d efend
the e 5 - pawn - 24 . . . Ah4! 25. e2 f6)
25 . e2 Ag7 26. ti":lc2 , and B l ack was
unable to d efe n d h i s weak paw n s .
12. e 5 Y!/e7 1 3. Y!/e2
A typical structure in the Moscow Varia
tion. Black's cou nterplay i nvolves devel
oping his bishop at b7 and prepari ng . . .
c6-c5. Wh ite prepares the manoeuvre of
his bishop to e4, in order to meet . . . c6-
c5 with the exchange of the lig ht-square
bishops and the possible central break
through d4-d5 .
a b c d e f g h
13 .. b5
1 3 . . . b6 is more often preferred: 1 4 . .fe1
.d8 (after 1 4 . . . Ab7 1 5 . Aa6 W h ite ad
vantageously exchanges the light-square
bishops) 1 5 . .ad1 . Now Black has to de
cide whether to forestall the exchange of
bishops by 1 5 . . . a5 or continue manoeu
vri ng - 1 5 . . . .b8.
a) 1 5 . . . a5 (in combination with . . . c6-c5
t h i s m ove somehow d oes n ot look
right, si nce it leads to a weakening of
the q ueenside) 1 6 . Ad3 Ab7 1 7. Ae4 b5
1 8. h4 ttlb6 1 9. Ab1 c5 20. tt::l x b5 Aa6,
and here instead of 21 . h5 g5 with dou
ble-edged play (Kram nik-Svidler, World
Championsh ip, Mexico 2007), 2 1 . Ad3!
wou ld have retained the better chances.
b) 1 5 . . . .b8 1 6 . Ad3 A b7 17. Ae4 c5
1 8 . Axb7 .xb7 1 9. Y!/e4 .bb8 20. ti":lb5
(20. d5 exd5 21 . ti::l x d5 is also possible)
20 . . . cxd4 (20 . . . a6 21 . ttla7) 2 1 . ti":lbxd4,
and the weakness of B l ack's l i g ht
square periphery guarantees Wh ite an
enduring advantage.
1 4 .td3 .tb7
In the game G raf - Kraemer (Bundesliga
2007) B l ack p re m atu rely forced the . . .
c6-c5 advance: 1 4 . . . b4? ! 1 5. tDa4 c5
1 6 . .ac1 cxd4 1 7. Y!/e4 .b8 1 8 . .c7 and
Wh ite gained the advantage.
15 . .te4 E:fd8
Black carries out the standard rook deploy
ment in this variation, prepari ng the freeing
. . . c6-c5. The i m mediate 15 . . . .ab8 is also
possible, although here too after 1 6. .ac1
a6 1 7. .fd1 it is not easy to make this ad
vance, since 17 . . . c5? is parried by 1 8. Axb7
.xb7 1 9. tt::l e 4 with the occupation of d6.
Its preparation requires time. For example:
17. . . .fc8 1 8 . h4 h5 1 9. tt::l g 5! Aa8 20. We3
c5 2 1 . f4 cxd4 22. Wxd4 with advantage
to Wh ite in view of the chronic weakness
of the d6-point.
1 6. E:ac1
If 1 6 . ti::l x b 5 there wou l d h ave fol l owed
1 6 . . .tt::l x e5.
16 . E:ab8 1 7. E:fd1 a6
(see next diagram)
219
 Chapter 3 Life at the Top

a b c d e f 9 h 21 . a4?!
8 .I .I 8 21 . 'We3 c5 22. Axa8 .!:!xa8 23. d5 came
7 .i. l.tJ 'iV .l .i. 7 i nto consideration.
6 .l .l .l .l .l 6 21 . . c5?!
5 .l 8 5 'After t h e s i m p l e 21 . . . b4 2 2 . b1 c5
4 8 4 23. Axa8 .!:!xa8 24. d xc5 .!:! xc5 (24 . . . tL!xc5
3 tt:J tt:J 3 25. Wc4!) 25. bd2 .!:!xc2 26 . .!:!xc2 .!:!c8
28 8 V/i/ 8 8 8 2 Black even has a sl ight i n itiative, thanks
to the weaknesses on e5 and a4. It's hard
1: 1:
a b c d e f 9 h to explain why both players had missed
this.' (Carlsen)
1 8. h4
22. axb5 cxd4 23. xd4
The manoeuvre 1 8. b1 ! was more i n the
spi rit of the position, exploiting the res a b c d e f 9 h
pite (1 8 . . . c5? is not possible on account of 8 .i_ .l .l 'if . 8
1 9. Axb7 .!:!xb7 20. d xc5) to play the knight 7 ,.. .l .i. 7
to b3. A possible continuation is 1 8 . . . .!:!dc8 6 .l
1 9. bd2 Aa8 (now if 1 9 . . . c5? there fol
.l .l .l 6
lows 20. Axb7 .!:!xb7 2 1 . e4 with an inva
5 8 8 5
sion at d6) '20. '1Mfe3! (a great preparatory
4 tt:J 84
move, shown by the engi nes. 20. b3 c5 3 tt:J 3
2 1 . Axa8 .!:!xa8 with equal ity is what dis- 2 8 1: V/i/ 8 8 2
couraged me from playing 1 8 . b1 ) 20 . . . 1 1: I 1
c5 21 . Axa8 .!:!xa8 22. e4, and since 22 . . . a b c d e f 9 h
cxd4 can now be met by 23. 'Wxd4, Wh ite
has a clear i n itiative.' (Carlsen) 23 . . . Axe4?
1 8 . . . Aa8 A strateg i c m i stake, allowing complete
H e re 1 8 . . . c 5 is n ow poss i b l e , s i n c e domi nation by the knig hts in the centre of
19. Axb7 .!:!xb7 2 0 . e4 i s less effective be the board . After 23 . . . axb5 24. Axa8 .!:!xa8
cause of 20 . . . cxd4 21 . .!:!xd4 (21 . d6 .!:!bb8 the weakness of the b5-pawn is balanced
22 . .!:!xd4 xe5) 21 . . .xe5 22 . .!:!xd8+ Wxd8 by the weakness of the e5- pawn , which
23. xe5 Axe5 24. c5 '1Mlc7 25. d3 Ah2+ gives Black adeq uate cou nterplay. If 25. f4
26. h1 '1Mlb8. there is the good reply 25 . . . 'Wd8 (with the
th reat of . . .tL!xe5) 26. 'Wf2 and Black has a
1 9. Etc2
choice between 26 . . . g5 and 26 . . . b4.
1 9. h5 g5 20. h2 was more energetic,
switching the knight to g4. 24. xe4 Etxc2 25. Et xc2 axb5

19 . . Etdc8 20. Etdc1 'Wf8 25 . . . A x e 5 is d a n g e ro u s because of

Since 20 . . . c5?! 2 1 . Axa8 .!:!xa8 22. tLle4 26 . .!:!d2 ! , when the black pieces are hang
leads to a pin on the c-fi le, Black defends ing (tLlc6 is threatened).
his rook on c8. 26. c6! Etb6 27. f4

220
 M. Carlsen - Wang Yue

Playing for a complete bind; it only remains 33. 'Wd3! 'Wxe7

to 'seal ' Black up with tDd6. 3 3 . . . \W x c7 a l l ow s a m at i n g attac k :
34. tDg5+! hxg5 35. \Wxg6+ h8 36. \Wh5+
a b c d e f g h
Ah6 37. \Wxh6#.
8 'if . 8
34. Etxd7 'Wh4+ 35. f3 'Wh5+
7 1.& .l .l. 7
6 .i 4J .l .l .l 6 If 35 . . . h8, then 36. !! xg7 xg7 37. tDf6
is decisive.
5 .l B 5
4 4 36. g3
4J B B
3 3 Black resigned .
2 B l:. B 2
* Game 51
a b c d e f g h
M. Carlsen - Wang Vue
27 'Wa8
. .
Sofia 2009
The queen is more active after 27. . . tDb8 Sla v Defence {015}
28. tDa7! b4, but it cannot do anything on
its own: 29. h5 !!b7 30. tDc8 !!d7 31 . tDcd6 There are few players who do not have
etc. awkward opponents. Carlsen is no excep
28. tile7+ h7 29. h5 Eta6 30. hxg6+ tion. He was one of the Lviv player And rey
fxg6 31 . Etc7! Eta1 + 32. f2? Volokiti n's 'clients' (0-5!). But that was a
'I thought that the most natu ral 32. h2 long time ago (by Magn us's standards) -
would give Black unnecessary counterplay in 2006, and with the existing rating table
after 32 . . .tDxe5, but this fai ls to the simple their paths are not likely to cross in the near
33. tDf6+ Axf6 34. tDd5+ Ag7 35. \Wxe5 future. But another 'offender', Wang Yue,
f8 36. !!xg7+ \Wxg7 37. tDf6+, when Black is among his present-day rivals. The Chi
is mated or loses the queen.' (Carlsen) nese grandmaster is one of Carlen's most
32 'Wd8 awkward opponents: in their six games
32 . . . !!a4 33. tDc6! tDf8 34. tDf6+ also does before the present one with the classical
not help. time control he not only did not suffer a
a b c d e f g h
single loss, but he also twice forced the
Norweg ian to lay down his arms. Moreover,
8 'if 8
this was in the main tou rnaments of 2009,
7 l:. '.& 4J .. . 7 in Wijk aan Zee and Linares, where Carlsen
6 .l .l .l 6 was a real contender for fi rst prize.
5 B 5 1 . c4 c6 2. d4 d5 3. tilf3 tilf6 4. tilc3 a6
4 4J B 4
The Chebanenko Variation, named after the
3 3 well-known Moldovan trai ner, who made a
2 B * B 2 detai led analysis of this system of defence.
1 .1 Wang Yue is one of the most fervent sup
a b c d e f g h porters of this variation.

221
 Ch apter 3
5. c5
The most critical reply: Wh ite tries to gain
space o n the q u eenside. In a p revious
game with Wang Vue (Linares 2009), the
. . . a7-a6 theme was performed i n a differ
ent scenario: 1 . d4 d5 2. c4 c6 3. c3 f6
4. e3 a6 and after 5. f3 b5 the play took
on a q u ite d ifferent aspect.
5 .tf5
. .
Wang Vue's favou rite reply in this variation.
The main continuation is considered to be
5 . . .tbbd7, with which Black prepares the
undermining of the pawn chain by . . . e6-e5
and . . . b7-b6.
6. t'ilh4
Magnus takes the decision to exchange
the bishop, giving the position clear stra
teg ic contou rs. In the main line - 6. Af4
tbbd7 7. e3 e6 S . Ae2 there is the possible
S . . .tbe4, when Black's light-square bishop
plays an i m portant role.
6 . . . .tg6
I n the event of 6 . . . Ag4 7. h3 Ah5 S. g4 Ag6
all the same the bishop has to retreat to g6.
The inclusion of h2-h3 and g2-g4 not only
expands White's space on the kingside, but
also gives h i m a tempo for developing his
bishop with 9. Ag2, from where it supports
the e2-e4 advance.
7. .tf4 t'ilbd7
a b c d e f g h
Life at the Top
8. tilxg6!?
The concrete nature of Carlsen's thi nking
is one of the i m portant com ponents of his
playing style. Wh ite n i ps in the bud the
possi bility of . . . e6-e5 .
At the Amber Tou rnament (N ice 2009)
in his rapid game with Wang Vue, Anand
made this exchange after S. e3 e6 9. h3
Ae7 1 0. tbxg6 hxg6 1 1 . Ad3, and also re
solved his main strateg ic objectives: along
with . . .tbe4 he also excl uded . . . e6-e5. The
game cont i n ued 1 1 . . .'cS 1 2 . 0-0 0-0
1 3 . b4 !XeS 1 4. Ah2! AdS 1 5 . f4! b6 1 6. a3
a5 1 7. c2 b7 1 S . g4 Ac7 1 9. g2 axb4
20. axb4 !!xa1 2 1 . !!xa1 !!aS 22. !!xaS+
xaS 23. a4 xa4 24. tbxa4, and Wh ite
achieved the better endgame. Apart from
Black's compromised pawn structure on
the kingside, his c6-pawn is also weak,
and if 24 . . . b5 he has to reckon with the
sacrifice of a piece for a pair of connected
passed pawns.
However, after S. e3 Black could also
play S . . . e5. Now noth ing is promised by
9. A xe5 xe5 1 0. d xe5 tDd7, while in the
event of 9. d xe5 tDh5 the move Ag5 is no
longer possible.
8 . . . hxg6 9. h3!
I m med iately prepari ng the retreat of the
bishop. If 9. e3 there would have followed
9 . . . e5! .
9 . . . b6
Here if 9 . . . e5 there is 1 0. dxe5 tDh5 1 1 . Ah2
Axc5, and if White does not l i ke the com
plications after 1 2. g4 b6 1 3 . e3 xb2
1 4 . !!c1 d4, h e can restrict h i mself to
1 2 . d2 e7 1 3 . g4. Now after 1 3 . . . d4 it
is possible to transpose i nto a favourable
endgame: 1 4 . tbe4 Ab4 1 5 . tbd6+ xd6
222
 M. Carlsen - Wang Yue
16. exd6 Axd2+ 1 7. xd2 lDhf6 1 8 . .d1 .
Instead of 1 0 . . .lDh5, 1 0 . . .'c7 1 1 . e3 tDxe5
12. b4 Ae7 1 3 . Ae2 0-0 1 4 . 0-0 a5 1 5 . a3
lDfd7 looks more sol i d , with an accept
able game. For exam ple, if 1 6 . e4, then
16 . . . axb4 1 7. a x b4 . x a1 1 8 . xa1 d4!
19. lDa4 lDf3+ 20. A xf3 xf4 2 1 . xd4
tDe5 22. Ad1 .d8 23. c3 Af6 is possible,
with excellent com pensation for the pawn.
16. d2!? is more circumspect.
10. cxb6
If 1 0. b4?! there follows 1 0 . . . bxc5 1 1 . bxc5
e5! .
10 . . .'Wxb6 1 1 . :b1 e 6 1 2 . e 3 c5
a b c d e f g h
8 .I -*. .I 8
7 7
s . 'tW 6 xb3 19 . .xb3 leads to the two bishops
5 5
4 jb 4 1 5. Axd6 Wxd6 1 6. xc5 xc5
3 Qj 3 1 7. dxc5 Wxc5 18. Wa4+ e7 1 9. Ad3
2 2
1:. 'iV jb 1:. 1
a b c d e f g h
Although the opening battle has concluded
safely for Black, certain problems sti l l re
mai n . Wh ite has the advantage of the two
bishops, and the exchange of the dark
square pair w i l l enhance the rol e of the
light-square bishop i n su pporting a pawn
offensive on the q ueenside.
13. a3
Wh ite's plan i s to advance h i s paw n s
a n d create a passed paw n . If 1 3 . . . Ae7,
for exam ple, there follows 1 4. d xc5 tDxc5
1 5 . b4 tDce4 1 6. lDa4 d8 1 7. Ad3, and
the q ueenside pawns, supported by the
bishops, are ready to advance.
13 . . . Ad6
After the other version of the bishop ex
change - 1 3 . . . cxd4 1 4 . exd4 Ad6 - Wh ite
gains a serious positional plus: a pawn
majority on the queenside. This could have
been avoided by the rad ical 1 3 . . . c4, but
then Black sti l l has the problem of defend
ing his q ueenside i n the event of b2-b3 .
For example: 1 4. Ae2 (after the i mmediate
1 4 . b3 there is the good reply 1 4 . . . a5!)
1 4 . . . Ad6 1 5 . A x d 6 xd6 1 6 . 0-0 0-0
1 7. a4 .fb8 1 8. b3 lDb6 19. a5 cxb3
20 . .xb3.
14. a4 Wc7
The provocative 1 4 . . . a5+? 1 5. b4 cxb4
1 6 . Axd6 b3+ 1 7. Ab4 xa4 1 8 . xb3
being a real advantage.
a5
The position has clarified . To play for a win
White must not only create a passed pawn,
but also keep the q ueens on, as otherwise
this pawn will be easily blockaded.
20. c.!;le2!
A highly concrete move, which as regards
the opponent's reply also has a psychologi
cal implication: White displays his readiness
to give up his two rooks for the queen. I n
t h e event o f 2 0 . 0-0 .hc8 21 . .fc1 xc1 +
22 . .xc1 .xc1 + 23. h2 .b8 Black suc
ceeds in activating his rooks. If the rooks are
retained, it is not easy for White to create a
passed pawn: 21 . .fd1 (21 . d1 a4) 21 . . . .a?
22 . .d2 b6 23 . .c2 .xc2 24. xc2 .c7
(24 . . . a4 25 . .c1 ) 25. a4 .a? etc.
23
 Chapter 3
Life at the Top

a b c d e t g h a b c d e t g h

8 J. .I 8 8 J. 8

7
.t .t 7 7 7

6
.l.l 6 6 6

5 .l
iV.l 5 5

4 4
3 B B B 3
2 B D B 2

M M 1
a b c d e t g h a b c d e t g h

20 . . J;thc8 23 . . . q;,f8?!

Passive defence has no chance of success.

A critical moment. Focused on a passive
stance, Wang Yue misses an opportu nity
to exploit the position of the king on e2 by
20 . . . d4! . After 2 1 . \Wxd4 Wxd4 22. exd4
hd8 the pawn is regained: 23. e3 fbd5+
24. f3 lbf6 , and i n the endgame there
is nothing to fear, while if 21 . exd4 Black
obtains an excel lent post at d5 - 2 1 . . .Wh5+
22. g4 \Wd5 (G ieizerov).
21 . hc1
Usually it is not considered dangerous to
exchange two rooks for a queen, but i n
t h e given specific position t h e long-range
bishop, su pporting the passed pawn, may
be able to restrict the rooks' mobil ity.
21 . . . Wxc1
After 21 . . . Wb6 22. xc8 xc8 23. b4 axb4
24. xb4 We? 25. Wb3, thanks to his out
side passed pawn, Wh ite has a small but
enduring advantage.
Black shou ld have decided on 23 . . .1be4!?.
In the event of the exchange of minor piec
es - 24. Axe4 d xe4 25. b5! (after 25. bxa5
c7 26. Wxe4 ca7 Black picks up both
pawns) 25 . . . f5 (25 . . . b8!? 26. Wxa5 bc8
27. Wb4+ 8c5 looks tem pting, but after
28. f4! the king escapes from the danger)
26. f3 exf3+ 27. xf3 e5 28. Wb3 it is easier
for the rooks to fight against the passed
paw n , although even here the chances
of saving the game are i l l usory. However,
Wh ite is not obl iged to exchange his bish
op; he can also consider 24. f3!? fbd2+
25. f4, and although it looks provocative
to bring the king out, Black has no real
th reats: 25 . . . f6 26. bxa5 lbc4 27. \Wb4
xb5 28. '!Wb8 etc.
The other active possibility 23 . . . e5!? is
parried by 24. d2 cc8 25. bxa5.
24. bxa5 q;,ga 25. a6
Wh ite has a big advantage, but its conver
sion should take a considerable time.

22. xc1 xc1 23. b4 25 . . . c7 26. Wf4 ca7

26 . . . ac8 was m o re res i l ient. N ow i n

(see next diagram) t h e event o f 27. \Wb4 lbe4 2 8 . Wb6 c6

224
 M. Carlsen - Wang Vue

Black somehow holds out, but 27. f3 tDd7 34 . . .tDd7?! 35. Ac6) 35. h4 l:!c7 36. Wb6
28. 'Wd6 iDeS 29. Ab5 is stronger, when lka7 37. g4. By advancing h4-h5 at the
Wh ite should win. appropriate moment, Wh ite breaks u p the
27. 'Wd4! king's defences and gains new offensive
lines:
The q ueen ties down both rooks. It is no
a) 37. . .tDd7 38. Wd6 tDb8 39. Wd8+ <.!?h7
longer possible to return: 27. . . 1k7 28. Wb6
40. h5! tDxa6 41 . hxg6+ fxg6 42. Wb6
l:!ac8 29. a7.
tDb4 43. Wxe6;
27. . . e4 28. 'Wb6 b) 37. . . <.!? h7 38. h 5 g x h 5 39. Ad3+! g6
Preventi ng . . .tDd6 with the switching of 40. gxh5 tDxa6 41 . hxg6.
the knight to c4. 29. e1
a b c d e E l i m i n ating the th reat of a check on c3.
a .! 8 Th is same aim would have been achieved
7 .! 7 by 29. Wc6 followed by the d riving back
6 fj of the knight and an attack on the central
V/11 6
pawns: 29 . . . l:!b8 30. f3 tDf6 31 . Wd 6 l:!b2+
5 5
32. <.!?d1 or 29 . . . l:!d8 30. f3 tDf6 31 . Wb6
4 4
da8 32. Wc5.
3 fj
29 . . . g5
2 2
If 29 . . . tDf6 Wh ite attacks with 30. a4 tDd7
31 . Wd6.
a b c d e f g h
30. J.b5 e6 31 . a4 d4
28 . . . e5? Black loses after 31 . . .tDc7 32. Ac6 but,
Inappropriate activity, leading to a weaken in view of the th reat of 32. Wd6, he also
ing of Black's position and a rapid defeat. cannot get by without advancing his pawn.
He should have tied the queen to the de 32. a5 f!.c7?
fence of the a6-pawn - 28 . . .tDf6 29. Ab5
Overlooking the loss of the exchange. But
tDd7. However, this is a reciprocal linkage.
32 . . . dxe3 33. fxe3 tDf8 34. Wd6 would have
By the cyc l i c manoeuvre 30. Wd 6 tDb8
led to the loss of the e-pawn.
31 . Wd8+ <.!?h7 32. Wb6 tDd7 33. Wd4 tDb8
Wh ite gains a tempo for a pawn offensive: 33. J.c6
34. a4 <.!?g8 (not 34 . . . c7? 35. Wh4+, or Black resig ned .

For Carlsen the next super-tournament, the Sparkassen G M i n Dortmund (FI D E cat
egory 20), followed a simi lar pattern to the tournament i n Sofia. The final result was
also similar. I n the first round Magnus won against Jakovenko, and after a present in
the 5th round from Naid itsch, who blundered a pawn, he could only be stopped by
Kram nik, who by trad ition is especially form idable i n Dortmund (he al ready had nine
victories to his cred it). Carlsen played against the favourite as though sleep-wal king,
making mistakes in a fairly simple position. Defeat i n the decisive game, the third in

2 25
 Chapter 3 Life at the Top

that year (after Wij k aan Zee and Sofia), was a warn ing symptom in the form ing of his
champion's character. However, in such company it is not customary to call a share
of 2nd-4th places a fai l u re.

Game 52 The start i n g p o s i t i o n of t h e variation,

M. Carlsen - D. Jakovenko
Dortm und 2009
Ruy Lope z [ C67]
1 . e4 e5 2. ttif3 ttic6 3 . .tb5 ttif6
The Berlin Defence was employed when
the study of the Ruy Lopez was sti l l in its
infancy, including matches i nvolving the
fi rst world champion Wilhelm Steinitz. But
for its popu larity i n the 21 st century it is
indebted to Vlad i m i r Kram nik, after he re
vealed it as his main weapon against 1 . e4
in his match for the world championsh i p
with Garry Kasparov i n 2000 i n London.
4. 0-0 ttixe4 5. d4 ttid6 6. Axc6 dxc6
7. dxe5 ttif5 8. 'Wxd8+
Wh ite has no reason to avoid the queen ex
change, si nce 8 . Yf/e2 1{)d4 9. l{)xd4 Yf/xd4
10 . .d1 Ag4 leads to the same thing, while
the attem pt to be stubborn with 1 0. l{)c3
Ae6 1 1 . . d 1 Yf/c4 1 2 . .d3 Ac5 g ives
Black a comfortable game. For exam ple,
if 13. l{)e4?! there follows 13 . . . Af5.
a . . . mxd8
w h i c h s i n c e the t i m e of the Kas p
arov - Kram n i k m atch h a s b e e n called
the ' B erl i n Wal l '. Even the 1 3th wo rld
cham pion was unable to breach it. As
for D m itry Jakove n ko, l i ke no one else
he knows about its sol i d ity, s i nce he has
ofte n attacked it fro m t h e w h ite side.
H oweve r, w h i l e t h e c h o i c e of t h i s de
fence ag ainst Carlsen has a psychologi
cal i m p l icat i o n , it can hard ly be deemed
a successful one. Everyone kn ows with
what pleasu re M ag n u s plays a compli
cated endgame, al beit with a smal l but
enduring advantage. In the g iven position
Wh ite can hope to make use of his pawn
majority on the kingside.
9. ttic3 mea
Kram n i k played 9 . . . Ad? and moved his
king to c8, but later experience in defend
ing the Berl i n Wal l recommends keeping
the king i n the centre. Therefore Black
moves it away from possible checks on
g5 and d1 .

a b c d e f 9 h 1 0. h3 h5 1 1 . ttie2 Ae7 1 2 . Ag5 Ae6

.J. s
This path was laid by one of the best So
=--' --"'""'-,'-'=- ,..'
.l 7 viet defensive players , Ratmir Kholmov.
6 6
The plan with t h e d eve l o p ment of the
5 5 bishop at b7 has been tried by the Chi
4 4 nese g randmaster Wang Yue, who also
has the reputation of being an expert on
3
tiJ 3

:::. :::. :::. 2 the Berl i n Wal l .

;g: 1 3 . ttif4 .td5 1 4 . .txe7 mxe7 1 5. ttig5

a b c d e f 9 h ttid4

2 26
 M . Carlsen - D. J akoven ko

a b c d e f g h By the march of the pawn to a4 Black pre

s .!. .I s pares the development of his rook via a5 ,
and at the same time he th reatens to cap
7 .\ .l *"" 7
ture the a2-pawn, si nce the bishop cannot
6 .l 6
be cut off by b2-b3 on account of . . . a5-a4 .
5 .i. CD .l 5
4 1.& CD 4 1 9. a3 a4 20. l:Xfe1 g6 21 . f3 l:Xa5
3 22. c3 l:Xb5 23. l:Xe2 l:Xa8 24. l:Xd4 l:Xaa5
25. <!>f2

In the event of 25. !!b4 !!xb4 26. cxb4 !!b5

a b c d e f g h 27. <!>f2 Black plays 27. . . c5.

1 6. l:Xad1 25 . . . l:X xe5

Strictly speaking, a novelty. But you don't
have to be a Carlsen to make such a logical
move - i n the g iven situation the rooks are
best deployed in the centre at d1 and e1 ,
since all the same Black will not captu re
the c2-pawn (1 5 . . . tbxc2? 1 6 . tbxd5+ cxd5
1 7. lk1 ). Nevertheless, previously 1 6. f!fd1
was played . For example, the game H ra
cek - Kholmov (Pardu bice 1 999) conti nued
16 . . . tbe6 1 7. Cbxe6 Axe6 1 8 . h4 f!d8 1 9. f3
g6 20. <!>f2 Af5 2 1 . c3 f6 22. exf6 <!>xf6
23 . Cbe2 c5 with equal play.
As a rule, it is sound practice to exchange
a wing pawn for a central one. In ad d i
t i o n , i n t h e g iven specific position Black
e l i m i n ates Wh ite's main advantage i n the
Berl i n Wal l - his pawn m aj o rity o n the
ki ngside.
26. l:X xe5 l:Xxe5 27. l:Xxa4 l:Xb5
27. . . b6 28. !!a? <!>d6 is also possible, with
approximate equal ity.
a b c d e f g h

s s
16 . . .tile6 1 7.gxe6 Axe6 1 8. h4
It is i m portant not only to fix the h5-pawn , 7 .l .l 7

but also to secure the knight at f4 against 6 .l 6

. . . g7-g5. 5 .!.
18 . . . a5 4 : CD
a b c d e f g h
3 3
.I s 2 2
7
e .t .t 7

6 .i. 6 a b c d e f g h

.l 5 It is hard to imagine that Black can lose

4 CD 4 such a solid position, but one of Carlsen's
3 strengths is his abil ity to exploit the slight
est chances, the opponent only need ing
to weaken his concentration.
a b c d e f g h 28. b4 c5 29. l:Xa7

227
 Ch apter 3 Life at the Top

Wh ite could have won a pawn: 29. tbxe6
xe6 30. c4 .!ab6 31 . bxc5, but the tran
sition i nto a rook endgame wou l d have
dispelled all his i l lusions: 31 . . . .!ab3 32 . .!ab4
.!axa3 33 . .!axb7 d7 with a d raw.
29 . . . cxb4 30. cxb4 d7
This would appear to be an insign ificant
transposition of moves, but it allows Wh ite
to improve the position of his king. The ac
curate 30 . . . i.c4 31 . e3 d6 would have
restricted Wh ite's possibilities.
31 . ttie2! !a b6
The decisive m i stake. The exchange of
rooks only favou rs Wh ite. Black should
have retreated with 34 . . . i.e6.
35. !a xa6+ bxa6 36. g4!
After this move the game is decided. Black
faces a dilemma: whether to allow the crea
tion of an outside passed pawn, or a weak
ness on h5.
a b c d e f 9 h
8 8
7 7
6

Another slight inaccu racy. Black should 5

not have conceded the 5th ran k to Wh ite, 4
which he can use for an attack on the h5-
pawn after prepari ng g2-g4. 31 . . . c8 was
3 [3::, 3

2 2
simpler, not feari ng 32. tDd4 .!ae5.
a b c d e f 9 h a b c d e f 9 h
8 8
36 . . . hxg4
7
Black follows the fi rst course. The second
6
would also have lost. Play involving 36 . . .
5
c5 i s too late: 37. gxh5! gxh5 (37. . . cxd4+
4 [3::, 38. xd4) 38. tDf5 e5 39. tDg3. No bet
3
[3::, 3 ter is 36 . . . e5 37. gxh5 gxh5 38. f4+ <;!>f6
2 2 (or 38 . . . d5 39. tDf5 Ab3 40. tbg7 i.d1
41 . tDe8! etc.) 39. e4 Aa2 40. a4, and
a b c d e f 9 h
in view of the th reat of a4-a5 and b4-b5
Black has to go in for new concessions -
32. e3 .tc4 40 . . . e7 41 . tDf5+ f6 42 . tDg3.
The rook's position should have been im 37. fxg4 e5 38. ttic6+ f6 39. f4
proved - 32 . . . .!ad6, intending 33 . .!axb7 .!aa6. e6 40. h5! gxh5 41 . gxh5 Ad3
33. ttid4 d6? Also after other continuations the coor
Here also it was not too late for 33 . . . .!ad6. di nation of the h- pawn with the knight is
decisive. For example: 41 . . .f6 42. tbd8+
34. l:ta5!
e7 43. h 6 i.d3 44. tb b7 f7 45. tDc5
Now g2-g4 is th reatened , with the idea of Af1 46. h7 g7 47. tbe6+ xh7 48. <;!>f5,
creating an outside passed pawn on the and after capturing the f6- pawn Wh ite also
kingside. picks up the remaining pawns.
34 . . . !aa6? 42. e3!

2 28
 M. Carlsen - D. Jakoven ko

The king moves closer to the q ueenside. 44. e5!

42 . At1 43. h6 f6
. .

A study-l i ke theme! The knight is attacking

a b c d e f 9 h the f7-pawn.
8 8
44 . . . Abs 45. d4 Aa4 46. h7 g7
7 i i 7 47. xf7 x h7 48. g5+ g6 49. e6
6 i tiJ fj, 6
Black resigned .
5 5

4
fj, 4

3 /j, 3
2 2

.i.
a b c d e f 9 h

For any representative of the el ite (apart from those seeking the champion's regalia),
consistently fi nishing i n the top three in tournaments can in no way be regarded as a
fai lure. For any, but not for Carlsen . His swift rise impl ied other summits. He h i m self
real ised that, despite outward successes, i n recent times he had slowed down and too
often fi nished on the i m mediate approaches to the tournament pedestal.
The kind of stupor that had occurred with Carlsen did not go unnoticed . After Dort
mund there was a storm of criticism. The comment by Konstantin Landa, published i n
t h e magazine 64 Shakhmatnoe obo z renie, looks qu ite fair: 'What, i n m y view, is prevent
ing the Norweg ian from conquering the summits? Previously chess was his favou rite
game, he was cal led the second Fischer, a genius . . . but at some point for M agnus it
became d ifficult everyday work (which in fact is true; powerful preparation is needed).
Serious mistakes are occurri ng, and even one-move blunders are creeping in. I don't
know what he is doing and how he is prepari ng with his trainers, but, i n my view, the
fresh ness in his play has gone. Something needs to the changed ! '
This d i d not last long. After al l , Carlsen h imself realised perfectly wel l that i t was u nfai r
to put his problems down to fate. For everything you have to answer you rself, and in
particular you have to get to the bottom of what is happen ing. One of the causes h i n
deri ng his further progress was qu ite obvious. He needed t o review his own conception
of work on the opening, si nce a correctly organised game wou ld lead to comfortable
play. Apparently help was also needed i n overcoming a certain psychological barrier,
which was h i ndering his further i m provement.
Help for the strongest player i n the West came from the East. I n the late summer of
2009 sensational news flew round the chess world. Garry Kasparov had beg un work
ing with Magnus Carlsen! From two weeks in August they trai ned in Croatia, and then
they worked for a few days i n Septem ber. Apart from face-to-face meetings, chess

229
 Chapter 3 Life at the Top

problems were regularly discussed by email and Skype. The results were i m mediately
apparent! After his contact with the Maestro, Magnus began appreciating more deeply
the i mportance of open ing organisation. This was undoubted ly an i m portant result
of his contact with Kasparov! The results of the col laboration were al ready seen in
October at the super-tou rnament in Nanjing (FI DE category 21). The chess world was
staggered by the ease with which Carlsen defeated his opponents one after another,
allotting them the role of observers in the battle for fi rst prize.
In the h i story of tournaments of such standard it is u n l i kely that one can find
an analog ue of the fi rst half, when fou r of M ag n us's five opponents fi n i shed on a
m i n u s score, the excepti o n being Wang Yue, who gained a d raw. Their ' m i nuses'
c l u bbed together to make M ag n us's ' p l us'. G arry's shadow hovered over Magnus
in the very fi rst ro u n d , when in a Scotch Game he confidently d efeated Leko. There
then fol l owed w i n s over Topalov, Jakoven ko and Radjabov. The confidence and
strength issu i n g from M ag n u s were ad m i rable. ' It seems to m e that an enormous
rol e is played i n particular by the emotional aspect of this col laboration,' commented
Al exey Kuzm i n , who was p resent at the tournament. ' For Carlsen such contact,
apart fro m the u n d o u bted c hess benefit , also h as a g reat psychological effect.
Carlsen's respo n s i b i l ity has sharply increased: he is now answerable not only to
h i mself, hut also to G arry Ki m ov i c h . '
I n the second h a l f o f the tournament Carlsen eased off, b u t he ended with the
enormous (for such a u n iform ly strong field) lead over his nearest rivals of 2% points.
Such a thing was achieved only by Alexander Alekh ine in the years of his greatest
achievements!
From the very first rounds the play of his main rivals gave the i m pression of a kind of
doom . ' I f was as though M agnus did not do anything at all, except that, l i ke a python,
he simply appeared on the path and lay down,' the reviewer of the Chesspro website,
Sergey Zagrebelsky, picturesquely commented on his play. But if you dig deeper, in
the games with Leko and Topalov the strategy of playing for restriction can be traced .

Game 53
M . Carlsen - P. Leko
Nanj i ng 2009
Scotch Game [ C45]
Before this game Peter Leko was the only
representative of the elite who Carlsen had
not succeeded in beating in play with the
classical time control. He was the first to
experience the consequences of the col
l aboration of the young N orweg ian with
Garry Kasparov.
1 . e4 e5 2. f3 c6 3. d4
Here it is, the fi rst sign! Up till then the
Scotch Game did not appear in Carlsen's
repertoire, whereas in Kasparov's it used
to occupy an honourable place. Among
those to suffer was Leko.
3 . . . exd4 4. xd4 Ac5
For those who like history we will also give
Leko's only 'Scotch' experience with Kasp
arov (Tilburg 1 997). Peter went in for a re-

230
 M . Carlsen - P. Leko

serve variation: 4 . . . Ab4+ 5. c3 Ac5 6. tt::l xc6 1 1 . tt::l d 2 Vi!!t'g 6 1 2. tLlb5.

bxc6 7. Ad3 Vi!!t' h 4 8. Vi!!t'e2 tLlf6 9. h3 0-0,
10 . 'Wxe4 1 1 . Af2 Axd4 1 2 . cxd4 tLl5g6
where he decided to try and refute Garry's
pawn offensive 1 0. g3 Vi!!t' h 5 1 1 . g4 Vi!!t'e 5 a b c d e f g h
12. g5 with the piece sacrifice 12 . . .tt::l xe4?
13. Axe4 e8 14. Af3 Vi!!t'xe2+ 1 5. Axe2 Aa6 .i s
1 6. c4 Axc4 1 7. tLic3 and lost. .& 7
6 1.& 6
5. Ae3 5 5

One of the main opening trends, along with 4

5. tt::l xc6 bxc6 or 5 . . .Wf6 6. Vi!!t'd 2 dxc6, which 3

in recent times has occurred more often . 2
a b c d e f g h 1 : CiJ
8 .i .i. 'Vi' . .i 8 a b c d e f g h

7 .f. 7
1 3. g3!?
6 1.& 6

5 .i. 5 This move, which was i ntrod uced by Mo

4
Cjj 4 rozevich, takes the game onto positional
l i nes: after defending the f4-pawn, Wh ite
3
ib 3

2 2 i ntends d4-d5, when the two bishops are

1 :( Cjj ib : 1
a b c d e f g h
5 'Wf6 6. c3 tt:lge7 7. Ac4 tLle5 8. Ae2
..
'Wg6 9. 0-0 d6
The sharp conti n u ation 9 . . . d5 1 0. Ah5!
Wxe4 1 1 . tLld2 allows White to develop a
strong i n itiative. For example, the game
Ehlvest - Be l i avsky (World Cup, Reykja
vik 1 99 1 ) went 1 1 . . . Wd 3 1 2 .tt::l 4 f3 Ad6
(1 2 . . . Axe3 1 3 . tLixe5 'Wxd2 14. 'Wf3! favou rs
Wh ite) 1 3 . tt::l x e5 Axe5 1 4 . Ac5 g6 1 5. Ae2
'Mif5 1 6. tt::lf3 Af6 1 7. e1 0-0 1 8. Ad3 Wd7
1 9. tLie5 'Mld8 20. Wf3 g7 21 . h4! and Black
came under an attack.
1 0. f4
Sergey Rublevsky, one of the main Scotch
devotees, usually sacrifices the pawn with
a prophylactic king move - 1 0. h1 Vi!!t' x e4
a real advantage. The attem pt to l i n k this
idea with 13. tt::l c 3 'Mixf4 and now 1 4. d5!?
(1 4. tt::l b5 0-0 1 5 . tLixc7 b8 1 6 . d5 is usually
played, but after 16 . . . b6 followed by . . . Ab7
W h ite's i n itiative comes to a sta n d sti l l)
was tried i n the game M i khalchish i n - Be
liavsky (Lviv 2002): 14 . . . 0-0 1 5. tt::l b 5 Ad?
1 6. tLixc7 ac8 1 7. c1 Wg5 1 8 . Axa7 Ah3
1 9 . Af3 tLlh4 20. We2 tLieg6 21 . Ae3 V!fe7
22. tLlb5 tLixg2 23. xc8 tLixe3 24. xf8+
xf8 25. e1 tt::lf4 26. Wxe3 Wg5+ 27. f2
V!fh4+ 28. g1 Wg5+ %-%.
13 0-0?!
A careless move, allowing Wh ite to car
ry out t h e m a i n i d ea of 1 3 . g 3 . B l ack
p l ayed m o re e n e rg etically i n the game
M o rozevich - K ram n i k (Dortmund 2001):
1 3 . . . Ah3! 1 4 . Af3 Vi!!t'f5 1 5 . e1 d5 1 6. Vi!!t' b 3
0-0 (1 6 . . . 0-0-0 is also interesting) 1 7. tt::l c3
c6 1 8 . Vi!!t' x b7 fb8 1 9. Vi!!t'c7 Vi!!t'f6 %-%.

231
 Chapter 3 Life at the Top

14. c3 'Wf5 1 5 . d5 fends the d5-pawn , relieving the queen of

this d uty, but also prepares the centralisa
tion of his knight via c3-e2-d4, attacking
8 I 8
the q ueen with gain of tempo.
18 . . . ac8 1 9. Wb3 b5
6 6
19 . . . b6 was more circumspect, h i ndering
5 5
the attack on the pawn weaknesses.
4 4
20. e2 Wh3 21 . d4 Ag4?
3 3
A loss of time. 21 . . .g8! was correct. Now
the attempt to win a pawn by 22. 'Wa3 tDf6
23. Wxa6 runs i nto 23 . . . g4, when af
a b c d e f g h
ter 24. Axg4?! 'Wxg4 Wh ite's light-square
1 5 . . . a6 periphery is weakened. Also insufficient
is 22. c6 f6 23. Ag2 'Wf5 (23 . . . 'Wh5?
The positional pressure could have been 24. Ad4) with a defensible position. 22. Ag2
avoided only by returning the pawn: 15 . . . is more accurate, provoking 22 . . . 'Wh5, and
c5!? 1 6. dxc6 xc6 1 7. 'Wxd6 l::! d 8 1 8. 'Wa3 now 23. c6 l::!fe8 fol lowed by . . .g8-f6
Ae6 1 9. l::!fd1 ge7, when Black succeeds with a tenable position.
in coord i nating his pieces. For exam ple:
22. Ag2 Wh5
20. Af3 'Wg6 2 1 . l::! a c1 Ag4 22. l::! xd8+
l::! x d8 23. Ag2, and although Wh ite retains a b c d e f g h
pressure on the q ueenside, the position is
defensible. I n the opinion of Arshak Petro
8 I 8
I
sian , consideration should also have been
7 .l .l .l 7
.l
given to the return of the q ueen to the rear .l 6
- 1 5 . . . 'Wd7!? followed by the fianchetto of 5 B 'ii' 5
the bishop, tyi ng White to the defence of 4
t:2J B .i. 4
the d5-pawn . 3 B 3

1 6. e1 h8!? 2 BB ib ib B 2
An i nterest i n g i de a . B l ac k i ntends to
1!l 1!l
a b c d e f g h
transfer his knight to f6 . Carlsen t h i n ks
that B l ack's p roblems wou l d hardly be 23. h4!
solved by 1 6 . . . b5 1 7. a3 , since he cannot
'The key move. Now all the black pieces
play 1 7. . . Ab7? on accou nt of 1 8. Ag4 'Wf6
on the kingside look rather stu pid, while
1 9. Ad4, when the q ueen is trapped . But
the black q ueenside pawns are ripe for
here also it was not too late for 1 6 . . . 'Wd7,
harvesti ng.' (Carlsen)
when Black holds the position.
23 . . . g8
1 7. c1 Ad7 1 8. Af3!
Now this manoeuvre is too late, but Black's
A multi-purpose move: White not only de- position is so comprom ised , that it is hard

32
 M. Carlsen - V. To palov

to offer h i m any good advice. Apparently

there was no longer any way of saving the 8 .!.
game. If 23 . . . Ad7 there follows 24. J::! c 3
with the doubling of the rooks.
7 .l
6 J: .l
24. :E!:c6 f6 25. :E!:xa6 5
tZJ
The gathering of the harvest begins.
4
3
25 Ad7 2
..

25 . . . J::!fe8 26. J::! x e8+ J::! x e8 27. J::! a7! is also

hopeless for Black. a b c d e t g h

26. xbS :E!:b8 27. a4 g4 28. Af3 Wh6 32 WhS 33. Wf4 AxdS 34. xc7 Ab7

Black also cannot be satisfied with 28 . . . 'Wf5 3S. :E!:b6 f6 36. Ad4 Wt7 37. e6 :E!:g8
29. Ae4 'Wh5 30. J::! a7. 38. g;,f2 :E!: bc8 39. Ac3 AdS 40. aS :E!:c4
41 . d4 Aa8 42. Wxd6 WhS 43. Wf4
a b c d e f g h :E!:cc8 44. :E!: be6
8 .!. .!. 8

7
.l .t .l .l .l 7 Black resigned .
6 1: .l 'if 6

5
tZJ 5
Game 54
4 4
M. Carlsen - V. Topalov
3 0 3
Nanj i ng 2009
2 --;
2 King's Indian Defence [ E90]
1: w
a b c d e t g h In 2009 Vesel i n Topalov headed the FIDE
rating list, but surprising ly, it was against
29. Wc4 h i m that Carlsen played especially suc
cessfu l ly, compared with the other repre
' I spent q u ite a bit of time aro u n d here sentatives of the top ten . After the tourna
trying to allow as l ittle counterplay as pos ment in Nanj i ng he improved the score to
sible. With this move I prevent 29 . . . tilxf2 5-3 (with 4 draws).
30. <;!;>xf2 tilxh4.' (Carlsen)
1 . d4 f6 2. c4 g6 3. c3 Ag7 4. e4 d6
29 xh4 s. f3 0-0 6. h3
Desperation, but otherwise Black will suf Carlsen avoids the Classical Variation, be
focate. ginning with 6. Ae2 , and prepares the de
velopment of his dark-square bishop at e3.
30. Axg4 Axg4 31 . gxh4 Af3 32. fS
6 a6
Only accu racy is req u i red of Wh ite - the
check at g6 was threatened . The main continuation here is 6 . . . e5 7. d5

33
 Chapter 3
a5 or 7. . . tll a 6. Also, 6 . . . c5, which is more
in accordance with Topalov's style, is often
played .
7. J.e3 e5 8. d5
The relieving of the pawn tension 8. dxe5
dxe5 is harmless after the queen exchange
9. 'Wxd8 f!xd8, since the e5-pawn is indi
rectly defended (1 0 . .'Dxe5 .'Dxe4). But Black
has to reckon with 9. c5. For example, in the
game Radjabov -V. M ilov (Warsaw 2005)
after 9 . . ..'Db4 1 0. Wa4 a5 1 1 . f!d1 We7 1 2 . a3
.'Dc6 White could have gained an advantage
by 1 3 . Ab5!?. However, here too 9 . . . 'Wxd 1 +
1 0. f! x d 1 .'Db4 would have relieved Black's
problems: 1 1 . f!d2 Ae6 1 2. a3 tll a2 13 . .'Dxa2
Axa2 14. Ad3 f!fd8 with the idea of 1 5. 0-0
f!xd3 1 6 . f!xd4 Ac4.
8 . . . c6
a b c d e f g h
Kasparov preferred the classical 8 . . ..'Dh5
with the idea of playing the knight to f4 and
advancing . . .f7-f5 . In the game C . Hans
en - Kasparov (Sven d bo rg 1 99 0) Wh ite
forestalled this by 9 . .'Dh2 'We8 1 0. Ae2, but
nevertheless Kasparov carried out this idea
with a pawn sacrifice - 1 0 . . .f5! 1 1 . exf5 .'Df4.
Now, as in a number of simi lar variations,
the exchange of the f4-knight is advanta
geous to B l ack: 1 2 . Axf4 exf4 1 3 . fxg6
Q34
Life at the Top
'Wxg6 1 4 . Ag4 Axg4 1 5 . h xg4 f!ae8+
1 6 . <;t?f1 Ill e S (Kasparov). The game con
tinued 1 2. 0-0 Axf5 (1 2 . . ..'Dxe2+ 13 . .'Dxe2
gxf5 is also possible) 1 3 . f!e1 Wf7 1 4 . .'Df1 ,
and here, accord ing to Kasparov's analy
sis, Black could have mai ntai ned equal ity
by 14 . . . .'Db4! 1 5 . .'Dg3 .'Dc2 1 6. Axf4 .'Dxe1
1 7. .'Dxf5 gxf5 1 8. Ad2 .'Dxg2 1 9. <;t?xg2 <!>h8
with com plicated play.
9. g4!
Wh ite prevents Black's standard play on
the kingside.
9 . . . c5 1 0. d2 a5 1 1 . a3
Wh ite prepares to drive away the knight.
After the standard development 1 1 . Wc2
IDea 1 2 . 0-0-0 f5 B l ac k has sufficient
cou nterplay. With a Samisch structure
there is the interesting idea of switching
the dark-square bishop to the queenside:
1 1 . f3 .'Dfd7 1 2. h4 Af6 1 3. Af2 'We8 1 4 . Ae2
Ad8 1 5 . 'Wc2 Ab6 with com plicated play
(Kolev - Kam inski , Odessa 1 989). Topalov
tries to carry out this idea in the course of
the game.
1 1 . . .fd7
Black prepares not only . . .f7-f5, but also
a bind on the q ueenside with . . . a5-a4,
which does not work i m med iately - 1 1 . . .
a4 1 2 . Axc5 d xc5 1 3 . .'Dxa4.
1 2 . lag1
1 2 . b4 is prematu re: 12 . . . axb4 13. axb4
f!xa1 1 4. Wxa1 .'Da6 1 5. Wa3 c5 and the
knight establishes itself at b4. The move in
the game has the aim of preventing . . .f7-f5.
1 2 . . . a4 1 3. 'Wc2 b6 14. 0-0-0 .td7
1 5. b1 cxd5
Black determ ines the pawn structu re too
 M. Carlsen - V. To palov

soon, which is i m med iately exploited by
Carlsen . The prophylactic 1 5 . . . h8, pre
pari ng . . .f7-f5 , was more in the spi rit of
the position.
1 6. cxd5 l:!:c8
1 6 . . . l;!a5 came i nto consideration, when
although the rook occupies an unusual po
sition for the King's Ind ian Defence (usually
the queen is brought out to a5), it performs
the important task of su pporting the knight
on c5 and controlling the b5-square.
8 8
creep in about the prospects of the bishop
manoeuvre.
M eanwh ile, the position demanded a
m o re concrete decision - the prepara
tion of . . . b7-b5. This wou l d h ave been
answered by 19 . . . l;!c7!? 20. h4 l;!a8 2 1 . h5
tDc8 or the immediate 19 . . . l;!a8!?, preparing
. . .tbc8 and . . . b7-b5.
20. g5
I n contrast to his opponent, Magn us's play
is always concrete. He ' u rges' the bishop
towards its d u bious goal and prepares an
attack on the king.

7 20 . . . Ad8
6
Black should have ad m itted the faultiness
5 5 of his idea and reverted to the plan with the
4 . 4 preparation of . . . b7-b5 - 20 . . . Ag7 21 . h4
l;!c7 22. h5 l;!a8 23. l;!h1 tbc8.
2 2
21 . h4! a8?

If this is a pawn sacrifice, then for what

a b c d e f g h
benefits? But Black's pieces are so uncoor
1 7. Ab5!
d inated , that possibly this is an attem pt to
bring them together. An intervention such
Who said that Carlsen lacks 'schooling'? as 21 . . . h3 22. l;!h1 g4 23. h5 merely
The move i n the game as though copies assists Wh ite's i n itiative.
the style of Tigran Petros ian , who knew all
22. AxeS! l:!:xc5 23. Wxa4 Wc8 24. l:!:c1
about such exchanges. Wh ite exchang
b6
es his ' bad ' bishop (which is ru n n i ng up
against his own pawn chai ns) for the 'good'
bishop on d7.

1 7. . . Axb5 1 8. xb5 'Wd7 1 9. c3

Af6?!

A rather artificial manoeuvre. Black takes

his bishop to a5 , but, as the song goes
'the road is long'. And if it is taken i nto ac
count that a barrier - the knight on b6 - will
have to be removed , which will weaken the
defence of the a4-pawn, doubts beg i n to a b c d e f g h

Q35
 Chapter 3 Life at the Top

An ad m i ssion of t h e fai l u re of the idea 29. gxf6+ I::i xf6

24 . . . Aa5, after which there follows 25. l;!g3 b c d f
a e g h
with the threat of b2-b4.
8 .i. 8

25. Wd1 ! 7 at 'iV at 7

Carlsen , as we know, i s able to deploy 6 at .!. at 6

his pieces in their best positions. If White 5 .!. at 5

places his queen in an advanced post with 4 4

25. Wb4, the black q ueen sets off on a 3 Qj 3
journey into the opponent's rear - 25 . . . Wh3
26. t/Jb3 l;!c8 27. l;!h1 Wf3 , eyeing not only
2 a QJ 2

the f2-pawn, but also the q ueenside. For

1 I a 1
a b c d e f g h
exam ple, if 28. Wxd6?! Black can sacri
fice the exchange - 28 . . . .xc3 29. bxc3 30. h5!
Wxe4+, causing the king to become nerv When an attack can be launched , Magnus
ous: 30. a1 t!Jc4 31 . Wb4 Ae7! 32. Wxe7 is not afraid to sacrifice material .
.a8 . 30 . . I::i x f2 31 . hxg6 h6
25 . 'Wh3 26. Wf3 Wd7 If 31 . . . hxg6 there follows 32 . .xg6+ xg6
33. Wg3+ Ag5 34. Wxf2.
26 . . . Wxh4? 27. .!::! g 4 leads to the loss of
32 ti)d1 !
the queen , while after the q ueen exchange
26 . . . 'Wxf3 27. t/Jxf3 t/Jc4 Wh ite sets about Simpl ification is the most effective way of
converting his pawn majority: 28. b4! .c8 converting the material advantage.
29. t/Jb5 with the threat of doubling rooks on 32 I::i x c2 33. ti)xf2 I::i c 8 34. ti)g4 .tg5
. .

the c-fi le. For exam ple, if 29 . . . f6 there can 35. ti)f3 ti)c4 36. ti)xg5 hxg5
fol low 30 . .c2 t/Jb6 31 . l;!xc8 t/Jxc8 32 . .c1 b c d
a e f g h
Ae7 33 . .c7 with a decisive rook invasion,
while in the event of 29 . . .f5 30. exf5 .xf5
8 .!. 8

31 . .xc4! .xc4 32. t/Jxd6 .xb4+ 33. axb4

7 at 'if 7

.xf3 34. t/Jxb7 Ab6 35. d 6 Black has to 6 at 6

reckon with the passed pawns. 5 at at 5

27. Wd3
4 Qj 4

3 3
27. h5 suggests itself, but with an extra
pawn ' i n h an d' M ag n u s does n ot rush ,
2 2

preferri ng prophylaxis i n the event of pos

a
a b c d e f g h
sible cou nterplay for Black such as . . . Ae7,
. . .t/Jc4 and . . . .fc8. 37. ti)e3
O n ly here does Carlsen m ake an error.
27 g7 28. I::i c 2 f6
. .

37. Wh3! t/Jd2+ 38. a2 l;!h8 39. Wxh8+!

28 . . . .!::! h 8 followed by . . . h7-h6 would pos xh8 40 . .h1 + g7 41 . .h7+ was more
sibly have g iven Black better chances. forcefu l .

36
 M . Carlsen - T. Radjabov

37. . . xe3 38. 'Wxe3 Wa4 a b c d e f g h

By playing 38 . . . 'We7 39. xg5 h8 Black 8 .I .t 'if fl .I 8

could have retarded, but not prevented 7 .t .t .t l.&l .t .t .t 7
Wh ite's offensive - 40. f5 ! . 6
"' 6
39. Wxg5 Wxe4+ 40. a 1 J;le8 41 . J;lc1 5 ii. .t 5
Black resigned. 4

Game 55
M . Carlsen - T. Radjabov a b c d e f g h
Nanjing 2009
Sicilian Defence [ 830] 5 . . a6 6 . .ta4 b5

1 . e4 c5 2. f3 c6 3 . .tb5!
A pragmatic approach, typical of the 'new'
Carlsen. In the main branches of the Chely
abinsk Variation Radjabov has analysed
everything, and there is no poi nt in com
peting with him in knowledge. It is more
log ical to switch to positional l i nes, where
Mag n us is stronger. Especially since i n
Linares that year they had already played
a very interesting game in this variation.
3 . . . e6
The greatest expert on this variation, Evg
eny Sveshni kov, considers this move to
be the strongest. Earl ier Radjabov chose
other variations, but d i d n ot play them
in the best way. For exam ple, the game
Grisch u k - Radjabov (Moscow 2002) went
3 . . . d6 4. 0-0 Ad7 5. e1 tDf6 6. c3 a6
7. Af1 Ag4 8 . d3 Axf3 9. "1Wxf3 g6 1 0. "1Wd 1
Ag7 1 1 . tDd2 0-0 1 2 . tDf3 lDg4 1 3 . a3 c8
1 4. b1 tDge5 1 5 . tDxe5 tDxe5 1 6 . Ag5
Wd7 1 7. "1Wd2 me8 1 8 . bc1 b5 1 9. h3 tDc6
20. Ae3 with comfortable play for Wh ite.
4. 0-0 ge7 5. c3
Other solid conti n uations are 5. b3 and
5. e1 . Carlsen sticks to central strategy.
6 . . . c4 7. "1We2 b5 8 . Ac2 tD g 6 prevents
Wh ite from setting u p a powerfu l centre,
but he gains the opportu n ity to attack the
q ueenside - 9. b3! .
7. .tc2 Ab7
The central pawn cou nter 7. . . d5 8 . e5
d4 9. Ae4 lDd5 leads to a very com p l i
cated positi o n . For exam ple, the rap i d
g a m e S h i rov - l van c h u k (Ta l l i n n 2 0 0 6)
continued 1 0. d3 Ab7 1 1 . c4 lDb6 1 2 . Af4
Ae7 1 3 . lDbd2 g5 1 4 . Ag3 h5 1 5 . h3 'Wd7
1 6. e1 0-0-0 1 7. cxb5 axb5 1 8. a4?! tDxa4
1 9. tDc4 bxc4 20. xa4 cxd3 2 1 . 'Wxd3 ,
and t h e position was s o com plicated that
only the players themselves could under
stand it.
8. We2 d5
I n Linares 2009 Radjabov played 8 . . . tDg6
ag a i n st Carlse n . There fol l owed 9 . d4
cxd4 1 0. cxd4 c8 1 1 . a3 Ae7 1 2 . tDc3
0-0 1 3 . d1 tDa5 1 4 . g3 tDc4 1 5 . h4 tDxa3
1 6 . bxa3 (1 6. Ab3!?) 1 6 . . . xc3 1 7. h5 Vflc7
1 8 . Ad3 tDh8 1 9. Ab2 b3 20. ac1 Vfla5
21 . d5 Axa3 22. Axa3 xa3 23. "1We3 with
sharp play, where White has com pensation
for the sacrificed pawns.

2 37
 Chapter 3
Life at the Top

The inclusion of 8 . . . c4 9. b3 d5 would A standard device, with which White opens

have led after 1 0. exd5 ti:)xd5 1 1 . bxc4 ti:)f4 the fi le for his rook. Black cannot prevent
1 2. e3 Ad6 to a very complicated posi this: 12 . . . b4 weakens the c4-square, which
tion, where Black has real cou nterplay. w i l l be occu pied by the knight on b1 . If
8 . . . g6 is a log ical conti nuation, and it 1 2 . Ag5 Wh ite has to reckon with 12 . . . h6
may possibly be the theme of futu re en 1 3 . Ah4 g5 1 4 . Ag3 Ag7.
counters. 1 2 . . . ttid5 1 3. axb5 axb5 14. cxd4
9. e5 d4 1 0. Ae4 The exchang ing operation 1 4. Ag5 Ae7
1 5 . Axd5 flxd5 1 6 . Axe? ti:)xe7 1 7. ti:)a3
a b c d e f g h
fld7 1 8. c4 b4 1 9. ti:)bS 0-0 would not have
a i. given Wh ite anything real.
7 .i. 14 . . . cxd4
6 .l ,.. 6

5 5 The knight exchange 14 . . . ti:)xd4 1 5 . ti:)xd4

4 4
cxd4 would have opened the way for the
wh ite queen to the kingside: 1 6. Ag5 flc8
3 3
1 7. ti:)a3 Ae7 1 8 . Axd5 Axd5 1 9 . Axe?
<ll xe7 20. g4 and it is not easy for the king
to find a q u iet refuge (20 . . . flhd8 21 . h4+
a b c d e f g h
<ll e a 22. xh7 etc.).
1 0 . . . 'Wb6?! 1 5. ttibd2

A rather rare m ove. The m a i n m ove i s a b c d e f g h

1 0 . . .ti:) d 5 , a s Kasparov h i mself played i n 8 i. e .I.
a rapid game against G rischuk (M oscow
2002). The continuation was 1 1 . cxd4 ti:)xd4
7 .i. .l
1 2. ti:)xd4 cxd4 1 3 . a4 b4 14. d3 c7 1 5 . f4?!
6
'if ,.. .l 6

flc8 1 6. ti:)d2 ti:)e3 1 7. flf3 Axe4 1 8 . d xe4

5 .l ,.. 5

d3! 1 9. xd3 ti:)c2 20. flb1 ti:)e1 , and Black 4 .l i;, 4

won the exchange. 3 tiJ 3

1 0 . . . ti:)g6 looks i nsufficiently logical: 2 tiJ VJJJ 2

1 1 . cxd4 cxd4 1 2. d3 Ae7 1 3 . g3! 0-0 1 4. h4 1
.: i;, .:
with an unpleasant initiative for Wh ite on a b c d e f g h
the kingside.
1 5 . . . ttif4?
1 1 . d3
It was more logical for Black to complete
1 1 . a4 deserved consideration, but Carlsen ,
his development by 1 5 . . . Ae7 1 6 . ti:)b3 0-0
a devotee of the classical style of play,
1 7. Ag5 , when Wh ite merely retains the
wants to complete h i s d evelopment as
advantage of the first move. But Radjabov,
soon as possible.
with his active style, is aiming to seize the
1 1 . . . lad8 1 2. a4 initiative.

38
 M. Carlsen - T. Radjabov

a b c d e f 9 h
16. Wd1 b4?!
16 . . . tbg6 was safer. But Teimour did not
8 .I -*- .I 8

declare war, in order to retreat!

7 ... ... ... 7

6
'if I .t. 6
17. b3 Axe4 1 8. dxe4 fd3 5 ... ::, !k, 5

All in the same style - only forward ! But 4

J!&\ J!&\ttJ 3:, 4

18 . . . tbg6 1 9.tbfxd4 tbxe5 20. h5 tbc4 3

ttJ 3

21 . Ag5 would no longer have solved the 2

V/!f ::, ::, 3:, 2

main problem - the rapid development of

the pieces. a b c d e f 9 h

1 9. Ags
19 . . . Ae7? will not do, since after 20. Axe?
xe7 21 . d2 Black has problems with his
knights. For exam ple, 2 1 . . J !e8 22. tbc1
xc1 23. xb4+.
19 . . . ftc8 20.fxd4 xb2
If 20 . . . tbxe5 there fol l ows 2 1 . f!a5 ! , w i n
n i n g the b5- paw n . T h e development of
the kingside wo u l d also have led to the
loss of this paw n : 20 . . . h 6 2 1 . Ae3 b7
22. f3 tbxe5 23. e2 Ae7 24. f!a5 0-0
25. f! x b 5.
21 . We2 c 4 2 2 . ftfc1
22 . . . AcS?
22 . . . Ae7 was better, although here too after
23. tbxb5 xb5 24. Axe? xe7 25. f!a7+
f8 26. f!a5 b6 27. f! xc4 Wh ite has a
very dangerous i n itiative.
23. xbS! 0-0?
Overl ooking the loss of a p i ece. True,
23 . . . Axf2+ 24. xf2 xb5 25. e2 0-0
26. tbd4 c5 27. Ae3 would also have led
to an obvious advantage for Wh ite.
24. xeS xeS
Th ings would not have been changed by
24 . . . f!xc5 25. Ae7 xb5 26. Axc5.
2S. Ae7
Black resigned.

Partici pation in the Tal Memorial Tournament has particular implications for Carlsen. It
is not only a matter of conti nuing the strong rival ry with the best players in the world ,
but also a n opportun ity t o make contact with t h e rich chess a n d cultural trad itions of
Moscow. The fi rst tournament after the fai ry-tale victory i n Nanj i ng also i m posed cer
tain obligations. Something extraord inary was expected of Magnus, although certai n
reviewers warned that a repetition o f Nanjing was not possible i n Moscow. A n d i t was
not only a matter of the very high level of the partici pants (FI DE category 21), but above
all the powerfu l potential of the grandmasters themselves. After all, most of them, to
some degree or other, were prod ucts of the famous Soviet Chess School and were i n
the top ten o f t h e world rating l ist.
In the very fi rst round he had to withstand the pressure of Vlad i m i r Kram n i k, who
with Black l iterally right from the opening forced Carlsen onto the defensive. But
the fol lowing series of d raws had a q u ite mu ndane explanation: d u ring the fi rst days

2 39
 Ch apter 3 Life at the Top

M agnus was u nwe l l , and i n such a condition he was u nable to play for a win. The
turn i n g point came in the second half of the tournament, when after five rounds the
main favou rites Kram n i k and Anand were confidently l ead i n g . The game with the
cu rrent world champion was awaited with particular i nterest . This was a fierce bat
tle between two powerfu l fig hters , who proved worthy of one another. The dynamic
nature of the play i n d icated that M agnus had recu perated , and although i n the 7th
round the d rawing series conti n ued - he was forced to concl ude peace with the Rus
sian champion Peter Svidler, his courage was now restored . In the next two rounds
Carlsen won ag ainst Ponomariov and Leko, and fi nished among the tournament
win ners, sharing 2nd-3rd places with lvanchuk, only half a point beh ind Kram nik.
A worthy performance i n such stel lar com pany, confi rm ing his status as the No.1
player from the West .
But the c u l m i nation of Carlsen's performance i n Moscow was his uncond itional
victory in the World Blitz Championship. In a double-round tournament of 22 grand
masters M agus scored 31 points, fi nishing 3 poi nts ahead of Anand and 6 ahead of
the third-placed Karjaki n .

Game 56 a b c d e f 9 h
M. Carlsen V. Anand
-
8 .! 1.1\ .t 'Vi .! 8
M i khail Tal Memorial Tou rnament
7 ... 7
Moscow 2009
Queen 's Gambit {039}
6
. . 6

1 . d4 f6 2. c4 e6 3. f3 d5 4. c3
dxc4
The acceptance of the pawn sacrifice is
cu rrently topical not only i n the Catalan
Open i n g , but also i n the Queen's Gambit,
for example i n the Botvi n n i k Variation or
the Moscow Variat i o n . But for the mo
ment Black is on course for the Vienna
Variation.
5 . .t g 5 .t b 4 6. a 3
Carlsen avoids the Vienna Variat i o n , to
which 6. e4 would have led , demonstrat
ing his read iness to sacrifice a pawn.
6 . . . .txc3+ 7. bxc3
5 5
4 4
3 tD 3
2 2
1 .:g .:g 1
a b c d e f 9 h
7. . . h6
Black clarifies the opponent's i ntentions
and al lows the regai ning of the pawn . 7. . .
b 5 8 . a4 c 6 i s m o re critical . The game
G risch u k - van Wely (Wij k aan Zee 2005)
contin ued 9. g3 h6 1 0. Axf6 (1 0. Ac1 also
deserves consideration, retaining the bish
ops with the idea of exploiting the weak
dark squares on the a3-f8 diagonal) 10 . . .

240
 M. Carlsen - V. Anand
------
gxf6 (after 1 0 . . .'xf6 1 1 . Ag2 Ab7 1 2 . e5
it is n ot easy for Black to complete h i s
development) 1 1 . A g 2 Ab7 1 2 . W b 1 a 6
1 3 . d2 We? ( xc4 was t h reatened)
1 4 . 0-0 d7 15. Wb4 c5 1 6. d xc5 Axg2
1 7. xg2 Wc6+ 18. g1 e7 and Black
maintained the balance.
Instead of 1 5 . Wb4 Wh ite can consider
the preparatory exchange 1 5 . axb5 axb5
(or 1 5 . . . cxb5 1 6 . A x b7 Wxb7 1 7. xc4)
1 6. xa8+ Axa8, and now 1 7. Wb4 with
the better game i n view of the th reats of
e4 and a1 . For exam ple: 1 7. . .f5 1 8 . a1
Ab7 1 9. e4 c5 20. Wxb5 cxd4 2 1 . cxd4
(21 . xc4 0-0 22. a7 is also good) 21 . . .
c3 22. c1 ! .
8 . xf6 Vkxf6 9 . e3
Magnus did not sacrifice the pawn in order
to regain it: 9. Wa4+ Ad7 1 0 . Wxc4, si nce
after 1 0 . . . Ac6 1 1 . e3 d7 Black success
fu lly completes his development.
9 . . . b5 1 0. a4 c6 1 1 . e5 a6
a b c d e f g h
8 .!. .a. .!. 8
7 7
6 . . 'iV 6
5 ttJ 5
4 fj, . fj, 4 b) 1 4 . axb5 axb5 1 5 . xa8+ Axa8 1 6 . Wa1
3 fj, fj, 3 Wd8 1 7. xc6 Wc8, and all that remains
2 fj, fj, fj, 2
1 M Jt M 1
a b c d e f g h
If 1 1 . . . 0-0 there c o u l d h ave fo l l owed
1 2 . Ae2 Ab7 1 3 . W b1 (an 'X-ray' of the
queen on the b7-bishop, typical of this
type of pawn structure) 1 3 . . . a6 1 4. xc4,
regai ning the pawn i n a favourable situa-
tion . The fianchetto of the bishop is also
possible - 1 2. g3 Ab7 1 3 . Ag2, or, as in
the game, 12. g4!? Ab7 13. Ag2.
12. g4!?
A fresh idea, combi ning the fianchetto with
the idea of an offensive on the king side - an
indirect indication of Carlsen's recovery.
12 .. .tb7 1 3 . .tg2 Vke7
There are two methods of playing as Black
with this q ueenside pawn structu re: he
can do his utmost to hold on to the pawn,
displaying his read iness to withstand an
attack by Wh ite on the q ueenside (in the
Catalan Opening this course is success
fu lly ad hered to by Evgeny Svesh n i kov), or
at an appropriate moment he can return the
pawn for the sake of completing his devel
opment. Anand prefers the former course,
demonstrating his i ntention of fighting for
a win (by no means the last consideration
here was Carlsen's physical exhaustion),
whereas he could have demonstrated his
peacefu l intentions and returned the pawn
with 1 3 . . . d7!?, provoking simpl ification:
a) 14. Axc6 Axc6 1 5. xc6 0-0 1 6. 0-0,
and now after 16 . . .b6 or 16 . . . e5 the
g4-pawn looks out of place. It is hard
to believe i n the real ity of an attack with
such l i m ited reserves;
for Wh ite is to force a draw by repetition
- 1 8 . a7 Wb8 1 9. c6 Wc8 (variations
by Shi pov).
14. 0-0 0-0 1 5. f4
(see next diagram)
Wh ite has seized the i n itiative, and with
the ' hook' on h6 the attacking pai r g4/f4
241
 C hapter 3 Life at the Top

a b c d e f 9 h threats to his king, and now he relieves

8 !. 1.1\ !. 8 the pawn tension on the q ueenside, hop
ing to exploit the a-fi le. 19 . . . b4 also came
7 .i. 'ii l. l. 7
into consideration, not only getting rid of
6 I. I. 1. 6
the potentially weak b5- pawn (after the
5 I. ttJ 5
exchange 20. axb5 axb5 it is very vulner
4 ... 4 able), but also creating a passed c-pawn. If
3 3 20. 'fie2 there can fol low 20 . . . flea 21 . cxb4
2
JL 2 'Wxb4, not fearing 22. g5 hxg5 23. fxg5 c3,
1 l;l VJ!i : w when the c3-pawn diverts Wh ite from the
a b c d e f 9 h attack on the king.

20. axb5 axb5 21 . !!xa7 'Wxa7

looks far more active than B l ac k's rigid
pawn construction on the q ueenside. This d e
a b c
is where the idea of the move 1 2. g4!? is
8 8
revealed.
7 'iV 7
15 .. d7 6 1.1\ I. 6
Anand becomes aware of the real ity of 5 I. 5

Wh ite's attack and he returns the pawn, 4 . 4

exchanging the active knight. In the event of 3 3
the preparatory 1 5 . . . 'fie? he wou ld have to
reckon with 1 6. g5! with dangerous threats,
for example: 1 6 . . . hxg5 1 7. 'fih5 tLld7 (1 7. . .
a b c d e f 9 h
g4 1 a. f5 exf5 1 9. flxf5) 1 a . CL!xd7 'fixd7
1 9. fxg5 with the th reat of switching the 22. 'Wa1
rook onto the h -fi le.
Wh ite switches his attention to the queen
1 6. xc6 'Wd6
side.
1 6 . . . A x c 6 i s weaker: 1 7. A x c 6 flaba
22 '*'e7
1a. axb5 axb5 1 9. fla7 flfda 20. 'fif3 b4
(20 . . . 'fid6? 2 1 . g5) 2 1 . flc1 with the initia Logical . Only by keeping the queens on can
tive for Wh ite. Black hope to get at Wh ite's weaknesses.
After the exchange 22 . . . 'fixa1 23. flxa1
1 7. e7+
CL!a4 24. fla3! flea 25. f2 Wh ite retai ns a
U nderstandably, a p l ayer with classical small but enduring advantage in the end
convictions rejects movi ng i nto a pin by game. Carlsen plays such endings with
1 7. 'fif3 , with unclear conseq uences after great pleasure, and therefore such a turn
1 7. . . ha. of events could not satisfy Anand.

1 7 'Wxe7 1 8 . Axb7 Eta7 1 9. Ag2 b6

. . 23. e4

Black has succeeded i n neutral ising the Time will not wait. After the incl usion of his

242
 M. Carlsen - V. Anand
rook Black can set up a passed pawn on
the queenside, and Wh ite forestal ls this
with the th reat of an attack on the king.
23 .. J:ia8 24. 'Wb2 EtaS 25. e5!
Showing an excel l e nt fee l i n g for the dy
nam ics of the posit i o n ! Wh ite i n c l udes
h i s b i s h o p in the p l ay and i nte n d s to
cont i n u e the advance of his pawn pha
lanx by f4-f5 .
a b c d e f 9 h
8 8
7 'Vi' I. I. 7
6 I. I. 6
5 .!. 1. 5 a4.
4 . 4
3 3
2
VJH
a b c d e f 9 h
25 Wh4
.
By launching a cou nterattack, d isclosing
the d rawbacks of g2-g4, which has led
to a weake n i n g of the k i n g 's defences,
Anand d e m o n st rates h i s read i n ess to
play for a win. However, he is taki ng a risk.
At a press-co nfere n c e after the game
he mentioned i n passing that 25 . . . d5
wou l d have guaranteed an easy d raw. A
straightforward analysis enables us to
ascertai n that the world champion's as
sessment was correct: 26. f5 (26 . Axd5
exd5 27. f5 gS) 26 . . . e3 27. f3 xg2
28. *xg2 g S ! ? (28 . . . h4 2 9 . h 3 hS
30. f2 leads to more critical play) 29. h3
exf5 30. xfS e3 , and the queen pen
etrates i nto the opponent's rear, forcing
Wh ite to concern h i mself with h i s own
king.
26. We2
This carefu l move gives Black an i m por
tant tempo for i n c l u d i n g h i s rook i n the
counterattack, whereas the resolute 26. f5!
wou l d have forced h i m to switch to de
fence. If 26 . . .xg4 there is the unpleasant
27. h3 gS (27. . . g3 leads to simi lar play)
28. fxe6 fxe6 29. b4! a? 30. d6 a2
31 . xe6+ *h7 32. g4 Wxg4 33. h xg4
and the passed e-pawn is very dangerous.
26 . . Eta3! 27. f5!
It would be fatal to switch to defence. Af
ter 27. c1 or 27. b1 there follows 27 . . .
27. . J:ixc3 28. fxe6 fxe6 29. We4! 'Wd8!
30. Wg6 Wxd4+ 31 . h1 Ete3!
Attack and defence are at the highest level.
a b c d e f 9 h
8 8
7 I. 7
6 I. 6
5 5
4 4
3 3
2
2
:1 1
a b c d e f 9 h
32. Wxe6+
Carlsen forces a d raw in a situation where
there was a last possibil ity of testing the
solid ity of Black's position by 32. Ac6 ! ,
with t h e threat o f 3 3 . f8+ * xf8 3 4 . e8
mate, which Anand mentioned in the press
conference. Black has to find the only de
fence 32 . . . Wc5! . H e loses after 32 . . .d8
33. g5 hxgS 34. f7+ *h8 35. h5+ *g8
243
 C hapter 3 Life at the Top

36. Ae8, renewing the threat of 37. f!f8+.
But after Black 32 . . . c5! Black holds on
by a fine thread of only moves: 33. xe6+
*h7! (33 . . *h8? 34. e8+ *h7 35. g5 hxg5
loses to 36. f!d1 ! with the threat of 37. h5+
*g8 38. f!d8+) 34. g5 h xg5 35. Ae4+ f!xe4
36. f5+ *g8 37. xe4 d5! 38. xd5
tbxd5 39. f!d1 with a d raw.
32 . . . mh7 33. Wf5+ mhs 34. wta+ mh7
35. Wf5+ mha 36. wta+
Draw.
Game 57
M. Carlsen - R. Ponomariov
M i khail Tal Memorial Tou rnament
Moscow 2009
Sicilian Defence [ 890}
1 . e4 c5 2. f3 d6 3. d4 cxd4 4. xd4
f6 5. c3 a6 6 . .te3 e6 7. f3 b5
8. 'Wd2 bd7 9. g4 h6 1 0. 0-0-0
a b c d e f
8 .!. j_ 'iV j_ .!. 8
7 1.1\ 1. 1. 7
I. I. 1.1\ I. 6
a b c d e f
10 . . e5
b6 1 7. tbe2 d xe4 1 8 . Axe4 !adS with
excellent play for Black (Leko - lvanchuk,
M u kachevo 2007);
b) 1 1 . tbce2 c7 1 2 . h4 d5 1 3 . tDf4 e5
1 4 . tDfe6 fxe6 1 5. tbxe6 a5 1 6. exd5
xa2 1 7. d3 *f7 1 8. g5 tbxd5 1 9. Ah3
tb xe3 20. tDd8+ * e7 2 1 . tbc6+ f7
22. g6+ *g8 23. xe3 with very com
p l i cated p l ay ( M o rozev i c h - Vachier
Lagrave, Biel 2009).
1 1 . We1 !
A new idea i n this variation After . . . b5-
b4 Wh ite moves away his q ueen in good
time in the event of . . .tbe5-c4. Despite his
fai l u re in this game, Ponomariov did not
change his assessment of 1 0 . . . tbe5 at the
World Bl itz Championship, which was held
i m med iately after the end of the tourna
ment. His opponents played differently.
The modest 1 1 . a3 was tried in the game
Dominguez- Ponomariov, which contin ued
1 1 . . .tDfd7 1 2. f2 c7 1 3 . f!g1 g5 14. h4
f!g8 1 5 . *b1 f!b8 1 6 . tba2 Ae7 1 7. f!h1 b6
9 h
1 8 . tDb4 tDbc4 1 9. Axc4 bxc4 20. hxg5
h xg5 2 1 . Ad2 a5 22. tba2 c3! 23. Axc3
tbc4 with com plicated play.
But in his game with Ponomariov, Mo
rozevich made a reckless pawn sacrifice:
1 1 . f4!? tbexg4 1 2 . e5 b4 1 3 . exf6 bxc3
14. xc3 tbxf6 (1 4 . . . gxf6 suggests itself)
1 5. Ag2 Ad? (1 5 . . . d5 1 6 . f5 would have al
lowed White scope for an attack) 1 6. Axa8
xa8 1 7. f!hg1 tbe4 1 8. a5 g6 1 9. tbe2! Ag7
9 h
20. Ad4, and White gained the advantage.
1 1 . . . Wc7 1 2 . h4 b4?!

In a number of recent games the more reso
lute move 1 0 . . . b4 has occurred , with good
results for Black. Here are some examples:
a) 1 1 . tba4 tbe5 1 2 . b3 Ad? 1 3 . tbb2 d5
14. Af4 c7 1 5. tbd3 tbxd3+ 1 6. Axd3
Premature, apparently. The more natu ral
1 2 . . . Ab7 1 3 . Ae2 tbc4 1 4. A xc4 xc4
1 5 . g5 tbd7 1 6 . g6 tbe5 would have led
to an u n clear position with chances for
both sides.

244
 M. Carlsen - R. Ponom ariov

13. ti:\ce2 ti:\c4 cessfu l . It was better to transfer the knight

to c5 - 1 6 . . . CL!d7, not fearing 1 7.fxe6 fxe6
13 . . . d5!? comes into consideration.
1 8. Axe6, after which 1 8 . . .c5 (or imme
a b c d e f 9 h
diately 18 . . . Ab7) 1 9. AdS Ab7 is possi ble.
8 .i .i. .i. .i 8
1 7. Wb3!
7
'W 1. 1. 7

6 1. 1. 1. 1. 6 Wh ite h as focused his attention o n the

5 5 e6- pawn: the th reat of a piece sacrifice
becomes real.
17 .. d5

The opening of the position merely ag

g ravates B lack's problems: his king has
a b c d e f 9 h
been caught in the centre and presents an
easy target for White's mobil ised pieces.
14. ti:\t4!
The only way to hold the position was by
An amazi ngly bold move, si nce all the dou 1 7. . .'e5.
ble attacks associated with . . . e6-e5 had
1 8. exd5 Ad6
to be calcu lated .

14 . ti:\xe3 a b c d e f 9 h

In the event of 14 . . . e5 1 5 . d5 x d 5
8 ,1 .i. .i 8

1 6. exd5 exd4 1 7. Axd4+ e5 1 8 . f4 B lack 7 1. 1. 7

faces d ifficu lties. 6 I. ,------ .i. I. I. 6

15. Wxe3 Wb6?
5 'W B 5

The pri mary cause of Black's later prob

4
l. ttJ ti:J B B 4
lems. H e allows the development of the
3 B 3

bishop on c4, after which the concentrated 2 BBB 2

pressure on e6 may become un bearable. 1 w 1

a b c d e f 9 h
15 . . . e5?! wou l d have been a mistake in
view of 1 6. d5 xd5 1 7. exd5 Ab7 1 8 . f4!
1 9.ti:\fxe6
with a dangerous position for B lack. H e
shou ld have completed his development The fru it was ripe, but 1 9.dxe6! looked
with 1 5 . . . Ae7. stronger and more natural . Magnus occa
16. Ac4! sionally makes little 'slips' in the conversion
of an advantage.
In the style of Bobby Fischer, who so l i ked
to attack the e6-point! 19 .. fxe6 20. d xe6 Ae7 21 . Wd3!

16 .. Wc5?! The q ueen has performed its role on b3,

and now it moves closer to the opponent's
The . . .'c7-b6-c5 manoeuvre is u nsuc- king.

245
 Chapter 3 Life at the Top

21 . . . 0-0 23. g5 h7

a b c d e f g h Of c o u rse, 23 . . J x d 4 24. '1Wxd4 '1Wxd4

8 !. .i. I. e 8 25 . .xd4 tDe8 26. gxh6 would have ena
7 .i. .t 7 bled Black to avoid a mating attack, but
6 .t 8 1.1 .t 6 there was no serious hope of saving the
5 'if 5 game.
4 .t ii. tt:J 88 4 24. gxh6 h5 25. e4 xh6+ 26. mb1
3 8 3 laa7 27. f5 laxd1+ 28. laxd1 f6
2 888 2
a b c d e f g h
g g 8
a b c d e f g h
8 .i.
7 J. .i. .t l.i 7
22. J.b3 6 .t 8 'Yi' 6
5 tZJ 5
Wh ite wants to win in comfort: move in to
4 8 4
g6 with his queen followed by tDf5 . But,
strangely enoug h , even with such a con
3 8 3
centration of force a delay i n the attack 2
allows a respite. There was a quick win by
22. g5! tDe8 23. gxh6 Ab7 24. hxg7, com a b c d e f g h

pletely elimi nating Black's pawn defences

with a conti nuing attack. 29. lad7!

22 J d8?
.
Ponomariov misses the only chance to fight
on with 22 . . . Ab7, when Black somehow
sets up a defence: 23. \Wg6 tDe8 24. tDf5
f!f6 or 23. g5 tDh5 24. gxh6 tDf4. Now his
game goes downh i l l .
I n concl usion, a l ittle combination in the
style of Capablanca, which decides the
game. The 'crude' 29 . .d8+ would have
led only to the win of the bishop.
29 . . . .txd7 30. exd7+ mta 31 . d5
Black resig ned .

The l ast tournament of 2009 was the London Chess C l assic with the partici pation
of e i g ht g ran d m asters (FI D E category 1 8) . The fi eld was selected in accordance
with E n g l i s h trad itions with respect to their com patriots . The fou r strongest Engl ish
players were g ranted the op portu n ity to cross swords with the m a i n challengers
for the world crown , Vlad i m i r Kram n i k and M ag n u s Carlse n . The favou rites were
d rawn together i n the very fi rst rou n d , and Carlsen 's w i n over the ex-world cham
pion determ i ned the further course of the tou rnament. H e confi d ently converted
h i s l ead i nto victory in the tou rnament (5 poi nts out of 7), al lowi ng h i s rival to close
only wit h i n half a point.

246
 M. Carlsen - V. Kram n i k
------

I n h i s game with Kram n i k , M ag n u s was able to demonstrate those attrib utes

of h i s , in wh i c h h e is su perior to the strongest players in the world - an abi l ity to
con d u ct l engthy p l ay, balancing on the barely d iscern i b l e bound ary separat i n g a
m i n i mal advantage from equal ity. Here it is normal ly the opponent who makes t h e
fi rst m i stake.

Game 58 a b c d e I 9 h
M . Carlsen - V. Kramnik 8 !. .i. 'iV !. 8
London 2009
English Opening [A29]
? .l .l .l .i. .l .l .l ?
6 6
1. c4 5 5

For Carlse n , games with Kram n i k have 4 4

a special i m p l icat i o n . Despite t h e fact 3 8 t2J QJ 8 3

that in his i nterviews he does not dem 2 8 8 8 88 2
onstrate any i nterest i n the fight for the 1 .: WH .:t
chess crow n , M ag n u s regards the play a b c d e I 9 h
of the 1 4th world champion as a kind of
beacon , by which he should be guided. 8 . 0-0
. .

Since for the present he is i nferior to the

ex-champion as regards the scale of his
home laboratory, this attempt to avoid sur
prises with the help of the English Open
ing, which he rarely plays, is fu l ly justified .
It should be mentioned that Carlsen also
used a s i m i lar idea i n his game with To
palov i n Linares 2008.
1 . . .e5 2 . tbc3 tbf6 3. tbf3 tbc6 4. g3
I n the game with Topalov, Magnus played
4. d3, provoking his opponent i nto a re
versed Sicilian. He adopts the same tactics
in the present game.
4 . . d5 5. cxd5 tbxd5 6. Ag2 tbb6 7. 0-0
.
8 . . . Ae6 is also possible, since if 9. b4 White
has to reckon with 9 . . . tbd4 with the threat
of 1 0 . . . Ab3, and after 1 0 . .b1 f6 1 1 . e3
(nothing is given by 1 1 . tbxd4 exd4 1 2. tbe4
AdS 1 3 . Ab2 fS) 1 1 . . .tbxf3+ 1 2 . 'Wxf3 c6
1 3 . bS .c8 Black has good play.
White's flank attack can be prevented by
8 . . . aS 9. d3 0-0 1 0. Ae3 Ae6. This varia
tion occurred in the famous game Botvin
nik- Portisch (Monte Carlo 1 968), where
the 6th world champion carried out a com
bination in which he twice sacrificed the
exchange: 1 1 . tba4 tbxa4 1 2 . 'Wxa4 AdS
13 . .fc1 .e8 14 . .c2 Af8 1S . .ac1 tbb8?

(1 S . . .'Wd7 is correct; the hope of playing . . .

Ae7 8. a3
(see next diagram)
Wh ite prepares a pawn offensive on the
queenside. 8 . d3 0-0 9. a3 leads to an
insign ificant transposition of moves.
c7-c6 is illusory) 1 6 . .xc7! Ac6 1 7. .1 xc6!
bxc6 18 . .xf7! and White won . Nowadays
1 1 . tbd2 is preferred , but here too after
11 . . .tbdS 1 2. tbxdS AxdS 1 3 . AxdS 'WxdS
14 . .c1 fS Black has adequate counterplay
(Cvetkovic- Mikhalchishin, Yugoslavia 1 993).

247
 Chapter 3
9. b4 .te6 1 0. la b1
1 0 . b5 is pre m at u re , as after 1 0 . . . tDd4
Wh ite can not play 1 1 . tDxe5? because of
1 1 . . . Ab3.
10 .. f6 1 1 . d3
A well-known position in the English Open
ing: White has deployed his forces as in the
Dragon Variation of the Sicil ian Defence.
This type of 'reversed ' play has a certain
impl ication. I n dynam ic Dragon structu res
an extra tempo is i m portant, and Carlsen,
who has a fai r amount of experience in
such positions, is satisfied by such a turn
of events.
a b c d e f 9 h
a i 'it' I . a
7 1. 1. 1. .a 1. 1. 7
6 j_ I. 6
5 5
a b c d e f 9 h
1 1 . . . a5
Before placing his knig hts on the central
squares d4 and d5, Black aims to deter
mine the pawn structure on the q ueenside
and obtain an outpost for his pieces at
b3, but after White's obvious reply he has
problems with his other queenside pawns.
It looks more promising to pursue central
strategy with 1 1 . . . tDd4, which Carlsen en
countered in his game with Radjabov at the
2008 Grand Prix tournament in Baku. There
followed 1 2 . tDd2 c6 1 3. e3 (1 3. tDde4 tDd5)
1 3 . . .tDb5!? (with the idea after 1 4 . &Dxb5
248
Life at the Top
cxb5 1 5 . &Db3 tDa4 of attacking the weak
nesses on the q ueenside) 1 4 . c2 &Dxc3
1 5. xc3 d7 with a good game for Black
in view of his pressu re on the d3-pawn .
It remains to add that after 1 2 . Ae3
tDxf3+ 1 3 . A xf3 c6 fol l owed by . . . tDd5
Black also has roughly equal chances.
1 2 . b5 ttid4 1 3. ttid2
After the exchange of knights Wh ite's c3-
square would have been weak.
1 3 . . . 'Wc8
The main contin uation is considered to
be 1 3 . . . Ad5 (not 1 3 . . . tDd5? 1 4 . Axd5
Axd5 15. e3, when Black loses a piece)
1 4 . tDxd5 tDxd5. The central ised knights
ensure Black a good game.
14. e3 ttif5 15. 'Wc2
a b c d e f 9 h
4 4
a b c d e f 9 h
The price of driving away the knight is a
weakening of the d3-pawn , which has to
be defended . The alternative is 1 5. e2
with the idea of attacking Black's centre by
f2-f4, but in this case the problems with the
defence of the d3-pawn are not relieved .
For exam ple, the game Sarg issian - Yako
vich (Moscow 2009) contin ued 1 5 . . . tDd6
1 6 . tDde4 tDf7 1 7. f!d1 Ag4 1 8 . f3 Ae6
1 9. d4, and here instead of routine play
 M. Carlsen - V. Kram n i k

agai nst an ' isolan i ' - 19 . . . exd4 20. exd4 able post, not only removing the threat of
e8, which in general is not bad , 19 . . . d8!? an attack by the bishop, but also preparing
came i nto consideration, with the idea of f2-f4.
20. tDc5 Ac4 2 1 . e1 tDd5, controlling the
a b c d e
central squares.
In Bareev - Sh i rov (Poi kovsky 2 0 0 6)
White preferred 1 6. a4, but this weakens the
b4-sq uare, which expands the possibilities
of the black bishops. The game conti n ued
1 6 . . . Ag4 (1 6 . . . d8 is also good , i ntend ing
to answer 1 7. Aa3 with 17 . . . tD e4!?, gaining
the advantage of the two bishops) 1 7. f3
Ae6 1 8 . f4 tDf7 1 9 . tDb3 Ab4! 20. c2 (the
q ueen's journey to e2 has proved fru itless)
a b c d e f g h
20 . . . Ah3 with excellent play for Black.

1 5 J ad8 1 6. b2 a4
.
1 9 . . . f8

Black prepares the surrounding of the b5-
pawn and takes control of the b3-sq uare,
restricting the manoeuvrabil ity of the wh ite
knights. The drawback to this move is that
the a4-pawn may become a target, as in
fact happened in the game.
1 7. fc1
A logical move; i ntensifying the pressu re
on the c-fi l e is i n the spi rit of the posi
tion, and W h ite is i ntend i n g to transfer
his knight to c5. After the routine 1 7. fd 1
tD d 6 both wh ite knig hts are ' hobbled' by
the need to defend the b5- pawn and the
b3-sq uare.
17 . . . tDd6 1 8 . tDde4 tDe8
Black not only prevents tDe4-c5, but also
defends the c7-pawn, not ru ling out the
poss i b i l ity of retu rning to the attack on
the b5-pawn.
1 9. 'We2!
A concrete approach to the position. Mag
nus transfers his queen to a more comfort-
The resulting position can be assessed
as one of dynam ic eq u i l i bri u m , which is
not easy to u n balance. By making t h i s
wait i n g m ove, Kram n i k i nvites h i s o p
ponent t o do t h i s . M o re active p l ay for
Black cou ld have been associated with
the preparation of . . . f6-f5 - the pivotal
idea of the given variation. In this respect,
consid erat i o n s h o u l d h ave been g iven
to 1 9 . . . Ah3 20. Ah1 (if 20. Af3 , t h e n
2 0 . . . Ae6 is appropriate, emphasising that
Wh ite's hopes associ ated with f2-f4 are
i l l u sory) 20 . . .f5 2 1 . tDd2 Ag4 2 2 . f1 (if
2 2 . tDf3 , then 22 . . . e6 with t h e th reat
of . . . e5-e4 is good) 22 . . . e4 23. f3 exf3
24 . tDxf3 e6 , keeping Wh ite's so-called
' l ittle pawn centre' under contro l .
20. f4
There is no other active possibil ity. The at
tempt fi rst to regroup the rooks encounters
a refutation, for example: 20. d1 (with
the idea of d3-d4) 20 . . . Ag4 2 1 . f3 Ae6
22. d4 Ab3 or 20. f1 d7 21 . bd1 Ab3
22. d2 f5.

249
 Chapter 3
20 ... exf4
An essentially forced exchange, conced ing
ground in the centre: Black does not want
to al low the isolation of his e- pawn and
the open ing of the f-fi le for the wh ite rook.
21 . gxf4 'Wd7 22. d4
Persisting with the e4-c5 manoeuvre,
exchanging the dark-square bishop. The
passive 22. gd1 is met by 22... Ab3 23. gd2
f5! 24. g5 d6 with excellent play.
a b c d e f 9 h
22 ... c6
'Th is was more or l ess the o n ly m ove I
considered , as I thought that Black had
to do something to prevent b7 from col
lapsing completely,' Carlsen comments.
'22...Ac4!? 23. Wf2 Ad3 was actually quite
good , but Kramn i k may have decided that
there was no reason to start concrete play,
since 24. c5! looked strong d u ring the
game. Sti l l , Black is probably fine here af
ter 24 ...Axc5 25. dxc5 c4! (25 ...Axb1
26. gxb1 gives Wh ite excel l ent compen
sation for the exchange, as Black can not
play 26 ...c4, due to 27. Ad5+) 26. Axb7
Wg4+ 27. Wg2 Wxg2+ 28. Axg2 gabS.'
23. cS AxeS 24. dxcS c4
Life at the Top
By l ifti n g h i s cont ro l of the a4- pawn ,
Black embarks on a forceful tactical path,
which complicates the defence. The nat
u ral 24 ...d5 suggests itself: 25. xd5
(25. bxc6 bxc6 26. xd5 leads to a simi lar
position) 25 ...Axd5 26. e4 We7 27. bxc6
bxc6 28. Af3 Ab3, and by contro l l i ng the
d-fi le Black holds the position without any
particular d ifficulty.
2S. d1
a b c d e f 9 h
2S . . .'Wc7 26. Ac1 as
T h e c o n s eq u e n c e of/the i n cautious
24...c4. Si nce Black h as to reckon with
the exchanges on c6 and d8 with the loss
of a pawn , he is forced to return his knight,
defending the more valuable of the pawns.
With the loss of the a4-pawn the conver
sion of the passed pawn is too distant and
Black has time to activate his pieces. 26 .. .
cxb5? did not work i n view of 27. gxb5
a5 28. d5 Wf7 29. b6, when a pawn
is nevertheless lost, and the rook remains
out of play.
27. bxc6 bxc6 28. xa4
'I was very happy to d iscover that this
m ove does n ot lose any m aterial, even
though it looks a bit loose. 28. gb4 was
another move I considered , but I d i d not
250
 M. Carlsen V. Kram n i k

l i ke 28...tt:lb3 (28 ...Ab3 29. xd 8 xd8 32. e2

30. Af1 , a n d a4 fal l s without any real
a b c d e f g h
com pensation) 29. tDxa4 tDxc1 30. xc1
f5 , w h e n Black has some cou nterplay.
8
Il.i) 8
However, Wh ite is probably sti l l a bit bet 7 1.1. 7
ter, and I should probably have gone for 6 I. .i.l. 6

this.' (Carlsen) sl.j) B 'iV 5

28 J1xd1 +
4 B 4

The alternative 28 ... Ab3 29. xd8 xd8 3 B B 3

30. tDb6 (if 30. tDc3, then 30...'fie? 31. Axc6 2

tZJ B 2
Ac4! 32. Wg2 Ad3 and .. .'xeS is possi
I(
ble) 30 . .. d1 + 31 . <;!;>f2 We? 32. xb3 xc1 a b c d e f g h
33. d3 would have led to simpl ification,
but allowed Wh ite to activate his rook. 32 .. .tf5?

29. V!Yxd1 lidS 30. V!Yc2 V!Yf7 31 . c3 ' Perhaps his only real mistake in this game.
He could have retained dynam ic equal
By bri nging u p his knight to the centre of
ity with the amazing 32 ...<;!;>f8!! , with the
events, Wh ite not only prevents the rook
idea of 33 . . . Ab3 ,' com ments Carlsen. 'The
i nvasion at d 1 , but also coordinates his
point, which we probably both m issed , is
forces. I n time-trouble one could hard ly
that White cannot play 33. tDd4 because
expect h i m to decide on 31. tDb6. After
of 33 ...xd4 34. exd4 Af5, and Black is
31 . . . Ab3!? (if 31.. .tDb3, then 32. e4 tDd4
at l east n ot worse, as 35. Ae4 fai l s to
33. Wf2 fo l l owed by Ae3 is poss i b l e)
3 5...Ylffg 6 + 36. Ylftg2 Ylf!xg2+!. 32 ...Ag4
32. Ylf!c3 d 1 + 33. <;!;>f2 Ylff h 5 34. xb3
33. tDg3 d1+ 34. Ylftxd1 is probably why
tDxb3 35. Ylftxb3 <;!;>f8 Wh ite wou ld have
he decided to play 32 ...Af5.' However, it
gained a m aterial advantage. However,
should be mentioned that to find the saving
his forces are uncoordi nated , and it is only
computer move 32 ...<;!;>f8!! i n time-trouble
possible to defend the bishop by 36. Ylftc4
is practically i m possible. But the move in
or 36. Wc2 (36. Ab2? loses to 36 . ..d2+),
the game i s also hard to explai n , s i n ce
but then there fol l ows 36 ... Ylf!h4+, when
after the opponent's obvious reply Black's
Black picks u p the h2-pawn, not leaving
position is u ntenable.
the king i n peace.
33. e4 .tg4 34. g3 V!Yf7 35 .tf1 !

31 .. V!Yh5?!
An i m p u lsive m ove, w h i c h s i g n ificant
ly c o m p l i c ates t h e d efe n c e , w h e reas
the bishop manoeuvre 31 ...Ab3 (wh i c h
Black must undoubtedly have considered)
32. Ylffe 4 Ac4 would have enabled him to
hold the position. For example, in the event
of 33. b6 Ylftf8 Black regains the c5-pawn ,
since 34. Ylfff5 d3 35. Ab2 xe3 is not
good for Wh ite.
In fi nishing off the opponent, Mag us, as
always, is accurate. First a technical proce
dure, shutting the knight out of the game.
35 .te6 36. V!Yc3!
And now the rook is d riven i nto the corner.
36 . 1Ia8
Otherwise the knight cannot be saved . If
36...tDc4 (b7) there would have followed
37. f5.

251
 Chapter 3 Life at the Top

a b c d e f g h
37. b4 'Wd7 38. t5 At7 39. At4 'Wd1
40. f2 aJ. 8

(see next diagram)

7
.i.l.l. 7

6 I. I. 6

40... ttlb3? 5 8 8 5

Now Black loses his knight. He should have

4 : 8 4

brought his queen back. 3 8 VJH tiJ 3

2 8 2
41 . Ae2 'Wb1 42. Ac4 xa3 43. ttle2
1
'iV 1

Black resigned. a b c d e f g h

Carlsen began the year 2010 as the clear rating favourite, now establ ished at the 2810
mark, and after the positive end to the previous year, few doubted his ambitions at the
super-tou rnament in Wijk aan Zee (FI DE category 19). And that is what happened. For
the fi rst time Mag nus won the fi rst prize (if you don't count the share of 1st-2nd places
in 2008, where on the tie-break the prize was awarded to Aronian). And the fact that
he did this i n com petition with the strongest players i n the world, including the last
two champions Kram n i k and Anand, consolidated his status as a real contender for
the world championsh ip. But in qual ity of play it was as yet premature to talk about
Carlsen's total superiority. Especially painful was his loss to Kram nik, the only player
whom he had both beaten and lost to in 2009. However, his ability to play with full
concentration was something that not all his famous opponents could withstand.

Game 59 ever, in individual meetings in 'classical'

S. Karjakin M. Carlsen
- chess, neither had won - five d raws. In
Wijk aan Zee 2010
French Defence [C11}
For Carlsen, games with Karjakin have a
particular implication. Since their childhood
years the chess paths of the two prodigies
have taken the course of a secret rival ry,
and they fi rst crossed here i n Wijk aan
Zee, where Karjakin was the fi rst to break
through to the A tournament and the fi rst
to win the main prize. Th is had occurred
in the p revious year in the p resence of
Carlsen, who by that time had confidently
overtaken Karjakin i n the rating list. How-
the 1Oth round of the current 'Wijk' the two
players arrived with the same n u mber of
poi nts, but different moods: the previous
day Carlsen had lost with Wh ite to Kram
nik, whereas Karjaki n had won with Black
against Nakam u ra.
1 . e4 e6
Undoubted ly a su rprise for Karjakin - ap
parently for the fi rst time, Carlsen plays the
French Defence. But here is his motivation:
' I n my younger years I used to consider it
at best a second-rate openi ng. I believe
that 1...c5 and 1...e5! are better choices,
but since I desperately wanted to win this

252
 S. Karjaki n - M. Carlsen

game (I was trai ling the leaders Kram n i k
a n d Shirov b y a poi nt at t h i s stage) I de
cided to try something new.' It can be as
sumed that Carlsen was following the path
of Rodriguez Vila, who at the 2009 World
Cup i n Khanty- Mansiysk used the French
Defence to try and save his m i n i -m atch
against Karjaki n , and d rew appropriate
conclusions from this experience.
2. d4 d5 3. CC!c3 CC!f6 4. e5 CClfd7 5. f4
c5 6. CClf3 CC!c6 7 .te3 .te7
sufficed only for a d raw, the com petitive
objective of qualifying for the second round
of the World Cup was fulfilled.
The waiting move 1 1 . h1 retains the op
tion of tba4 if Black carries out the plan of
...b5-b4, ...a6-a5 and ...Ac8-a6, although
moving the knight to the edge of the board
has its drawbacks and demands additional
justification compared with the retreat to d1.
1 1 . ..'Wc7

I n recent times this move has often been Carlsen also sticks to waiting tactics, to
employed by one of the last Mohicans of clarify Wh ite's plans, n ot considering it
the French Defence, Mikhail Gurevich. The necessary to check 11 ...b4 1 2. tba4. Mean
classical continuation is 7...cxd4 followed while, the game N ijboer- M. Gurevich (Am
by ...b6. sterdam 2006) showed that Black has quite
8. 'Wd2 0-0 9. .te2 a6 1 0. 0-0 b5 good prospects. There followed 1 2 ...cxd4
1 1 . h1 13 . tbxd4 tbxd4 1 4. Axd4 Ab7 1 5. f5 exf5
1 6. xf5 Ac6 1 7. b3 Axa4 18. bxa4 tbc5
b c d e f g h
1 9. af1 tbe6 20. Aa1 a5, and it was rather
a

8 .1. .i.'ti'8 .. Wh ite who had problems.

7 .&.t. 7
6 . "' 6 1 2. a3 .tb7 1 3. laad1 laac8 1 4. 'We1

5 ...... 8 .
5 Strictly speak i n g , a new, although qu ite
4 8 8 4 logical m ove. Wh ite switches his queen
3 t2J ittb 3 to the kingside, but his formulation of the
idea is unsuccessful. Before this the routine
2 8 8 8 Wiit 8 8 2 1 4. dxc5 AxeS 15. tDd4 was played.
1: : <;i(1
a b c d e f g h 14 cxd4 1 5. CC!xd4 CC!xd4 16 .txd4

.tc5
Karjaki n decides not to conti nue with the
a b c d e f g h
standard 1 1 . a3 and he d iverges onto a
l ittle-studied path, although in the afore 8 .i .. 8
mentioned game with Rodriguez Vila he 7 .i. 'YJV"' 7
was able to b u i l d up a very dangerous 6. 6
attack o n t h e k i ng: 11 ... b 6 1 2. tbd1 5 .t. 8 5
a5 1 3. c3 a4 1 4. tDf2 tba5 1 5. ad1 Ab7
4 it 8 4
1 6. Ad3 tbc4 1 7. e2 tbxe3 1 8. xe3 b4
1 9. f5 Aa6 20. f6 gxf6 2 1 . h6 f5 22. dxc5
3 /j, t2J 3
xc5 23. axb4 b6 24. h1 f6 25. g4 2 8 8 it 8 8 2
Axd3 26. tbxd3 fxg4 27. g1 f5 28. h3 f7 .SiV.S <;i( 1
29. hxg4 f4!. And although Wh ite's attack a b c d e f g h

253
 Ch apter 3 Life at the Top

1 7. 'Wh4? remains 23. b4 d4 24. fxeS dxc3 2S. exf6

cxb2 26. xb2 xf6 with the better game
A surpri s i n g l y n a'ive m ove for such an
for Black.
erud ite grandmaster. Surely Wh ite wasn't
seriously hoping for an attack on the king? 20 . . xf6
.

After all, for this he would need to have at

a b c d e f g h
least a pair of m i nor pieces ai med at the
kingside. I n the spirit of the position was
8 .1. 8

1 7. f2, from where, in the event of ...f7-f6, 7 .t'ii .l 7

the queen retai ns control over the central 6 .I. 6

dark squares (in particular d4). True, the 5
JA1 5
loss of time on the king move to h1 makes 4 l:r 12 Wi{ 4
itself felt, and Wh ite has no g rounds for 3 12 tij 3
gai n i n g a real advantage. For example:
1 7. f2 f6 1 8. exf6 xf6!? (more interesting
2
12 12 1212 2
than 18 ...tDxf6 1 9. AxeS xeS 20. xeS
1 11 <i!f1
a b c d e f g h
xeS 2 1 . *g1 , after which Black has to
forget about any ambitions) and Wh ite has
The simplest, since it becomes clear that
to reckon with the threat to his f4-pawn. For
White's play has come to a standstill. The
example: 1 9. AxeS tDxcS 20. d4 tDe4!?
rook on d4 is bad ly placed , and not only is
2 1 . tDxe4 dxe4 22. c3 cf8 23. g3 gS! ,
...cf8 threatened , but he also has to reck
or 1 9. e3 cf8 20. *g1 Axd4 2 1 . xd4
on with the th reat of ...e6-eS. 20 .. .tDxf6
b6, i ntending a possible ...tDd7-b8-c6.
is also possible, eyeing the option of play
Apart from 17...f6 Black can also consider
ing the knight to e4. For example: 2 1 . d1
17...tDb6 1 8. Ad3 (i ntending f4-fS, which
(21 . e1 tDe4!?) 21 ...eS 22. fxeS xeS with
for the m o ment i s p re m at u re - 1 8 . fS?!
the idea of .. .tDe4.
Axd4 1 9. xd4 f6! 20. fxe6 fxeS) 1 8 . . .tDc4
1 9. Axc4 (not 1 9. b3? tDb2!) 1 9 ...Axd4! 21 . f5
20. xd4 xc4 with an equal game.
The opening of the position merely increas
1 7 ... Axd4 1 8. xd4 f6! es the activity of the black pieces. But the
attempt by Wh ite to consol idate his forces
Of cou rse. As is well known, in the French by 2 1 . g4 cf8 22. f3 does not succeed
Defence there are two standard u nderm in after 22...tDb8! with the th reat of . . .tDc6.
ing m oves: . . . c7-cS and ...f7-f6. It now For example: 23. Ag6 Ac8! 24. g3 tDc6
tra n s p i res t h at there i s noth i n g for the 2S. d2 tDe7 26. Ad3 xf4 and Black wins
queen to do on h4. a pawn.
1 9. Ad3 h6 20. exf6 21 ... cf8

Nothing is given by 20. g4 b6 2 1 . exf6 Real i s i n g t h at W h ite is c o n d e m ned to

tDxf6 22. d1 eS! , when Wh ite cannot play passive d efence, Black cal m l y i n c reas
23. fxeS xd4 24. Ah7+ tDxh7, since it turns es the p ress u re, assu m i n g that w i n n i ng
out that his back ran k is weak. There only the exchange by 21 ...eS 22. tDxdS AxdS

254
 S. Karjakin - M. Carlsen

23. !!xd5 tDb6 24. '!We4 tDxd5 25. '!Wxd5+ 27 ... Eif1 28. d4 Eixg1+ 29. xg1 Eie8
is an i nsufficient return for his advantage.
Forc i n g t h e exchange of queens, after
22. Eig1 c5! 23. fxe6 xe6 24. Eig4
which Black can set about converting his
Afte r 24. !!xd5 tDf4 25. !!d4 '!Wc5 o r exchange advantage without any particular
24. tDxd 5 Axd 5 2 5 . !!xd 5 tDf4 26. !!d4 problems.
Yf/c5 Wh ite has to g ive u p the exchange.
30. h4 We1 + 31. h2 Wxg3+ 32. xg3
24.. f4 25. Wg3
a b c d e
a b c d e f g h 8 8
8
.I tit 8 7 7
7 .t. 7
.I .t. 6 5 5
5 5 4 ::, 4
4
.&:r 4
3
3 ::, tb 3 2 ::, 2
2 ::,::, ::,::, 2

:r
a b c d e f g h
25 . We7!
An accu rate m ove. By taking control of
the e2-square with the th reat of ...d5-d4,
Black puts h i s opponent in zugzwang. A
characteristic variation is given by earls
en: 26. h3 d4 2 7. tDd 1 (27. tDe2 tDxe2
28. Axe2 Ac8 leads to the loss of the
exchange) 27...Ac8 28. !!e1 Vff/ f 7 29. !!h4
Ab7 30. !!g1 tDxg2! 3 1 . !!xg2 !!f3 32. Yf/g4
!!e8, and Wh ite cannot d efend aga i n st
...!!e1 .
26. Eixf4
White could not find anyth ing better than to
give u p the exchange. But now it is merely
a matter of tech nique.
1 a b c d e f g h
32 ..f7!
The tech n ique of playing simple positions
i s a favou rite 'hobby-horse' of Mag n u s
Carlsen. He i s n o t tem pted b y the w i n
o f a pawn after t h e i nvasion o f t h e rook:
32...!!e1 33. 'i!?f4 !!b1 34. b3 !!a1 35. 'i!?e5
!!xa3 36. 'i!?d6, and he prevents any activity
by the wh ite king with the su pport of the
blockad ing knight on d4 - 33. 'i!?f4 'i!?f6,
th reatening ...g7-g5.
33. f2 f6 34. g3 Ac8 35. c3 A g4
36. Ac2 g5 37. hxg5+ hxg5 38. Ab3
Against the attem pt to stand fi rm with
38. Ad3 Black decides matters by 38 ...!!h8
39. 'i!?g2 Ah5 with the switc h i n g of the
bishop to g6.

26... Eixf4 27. e2 38 e5 39. Ac2 Eif8+ 40. g2 Ad7

..

41. f3+ f6 42. Ab3 g4 43. d4 e5

27. Vff/ g 6 !!8f6 28. Vffi h7+ 'i!?f7 was hopeless
for White. (see next diagram)

5
 Chapter 3 Life at the Top

a b c d e f g h The black pieces occupy ideal positions.

8 I. 8 Wh ite is u nable to prevent the advance of
the b-pawn .
7
.i. 7

6 44. Ac2 a5 45. Ad1 c!>e4

5 5 Wh ite resigned.
4
tiJ 4

3 3

2 2

a b c d e f g h

After h i s vi ctory in Wijk aan Zee the l eader of the world rating l i st permitted h i mself
to miss the prestigious Linares tou rnament and to relax i n N ice at the reg u l ar Am ber
Tou rnament of Joop van Oostero m. It stands to reason that here too a complete
tri u m p h was expected of Carlsen. After al l , in 2009 he had been o n ly fou rth after
Aro n i a n , Kram n i k and Anand . Mag nus was so sure of success , that he perm itted
h i mself to beg i n the tournament with the move 1. a3 in his b l i ndfold game with lv
anch u k . I n this way h e so m otivated the outstanding U krai n ian g randmaster, that
the latter won both games and developed such a pace, that to catch h i m Mag nus
had to score seven (!) successive wins. He began with a bri l l iant attack, conducted
b l i ndfo l d .

L. Aronian - M. Carlsen 'Th is is what Aron ian had missed . Once

N ice 2010 the bishop joins the attack via e5 or d4,
a b c d e f g h Wh ite's days will be n u m bered ,' Carlsen
8J. 'iV J. .8 comments. The fu rther c o m ments also
belong to him, and were publ ished i n the
7
.i. "' .i. 7 Russian magazine 64 Shakh matnoe oboz
6 '-" 6 renie and also in New in Ch ess.
5 . 5
4 t{j 4 23. xe4 Ad4+ 24. Etf2
3 3

2 ii,fij 2 '24. g2 looks very dangerous, but might

have been Wh ite's best chance, together
l:iV .: with Ag5 on the next move: a) 24...h4
a b c d e f g h
was what I had actually i ntended, and it
1 8 ...fxe4! 1 9. fxe4 cxd5 20. exd5 does indeed work very wel l after the ob
exd5! 21 . cxd5 xd5 22. <!>g1 e4! vious 25. Ag5 (25. Af3 is the only move

256
 L. Aronian - M. Carlsen

that allows Wh ite to d efend) 25 ...ttie3+ 25. tbxf2?

26. Axe3 Axe4 27. Af3 xg4+; b) 24...e7!
25. Af3 !!xf3 26. xf3 ttlf6 27. !!ce1 Axe4 'This move, which my opponent made in-
2S. xe4 ttlxe4, and Black is a clear pawn stantly, loses trivially. Had he taken a bit more
up, but there is sti l l plenty of fight ahead.' time, he would surely have noticed the re
source 25. Ag5! , which, surprisingly enough,
24. .. l!xf2 keeps White in the game. I had intended to
play the simple 25 .. . Ae3. N ow, after the
'When I played this, I believed that I was
more or less forced line 26. Axe3 !!g2+
winning on the spot, but then I immediately
27. h1 ttlxe3 2S. d4+ gS 29. xe3 !!xe2
noticed his only defence.'
30. xe2 e7 Black will once again emerge
a b c d e f g h
with an extra pawn, but not without some
drawing chances for his opponent. 25...Af6
ali is probably best, but lines like 26. ttlxf6 ttle3!!
71 ..t. 27. d4 !!g2+ 2S. h1 !!xe2+ 29. ttle4+ gS
6 .t 30. AxdS lacS!! 31. xe3 !!xe3 32. !axeS
5 .t !!xe4! 33. Axa5 AxeS are not that easy to
4 CLJ.t ..t.tt:J calculate blindfolded.'
3
25 .. .Wh4 26. We1 l!fS 27 .tf3 l!xf3
21 2S. We4 Wxf2+

1I :
a b c d e f g h Wh ite resigned.

Before the last round Carlsen still had real chances of sole victory in the tournament, but
in his blindfold game with Grischuk he left his queen en prise. The result was a share of
1st-2nd places with lvanchuk and an overall score of 14% points from 22 games. The
winners also made the same scores in both categories: in the blindfold they scored 6%
points (the same as Kramnik), but in the rapid they were unrivalled with S points.
The double-round tournament of six grandmasters in the smal l Romanian town of
Bazna (FI DE category 20) received the prefix 'super' thanks to the partici pation of the
leader of the world rating l ist Magnus Carlsen and the winner of the World Cup Boris
Gelfand. And although the echo of Carlsen's victory by a marg i n of two moves (+5 in
10 games!) was i n keeping with his victory i n Nanji ng, i n the cou rse of the tournament
his superiority did not seem so indisputable. At a certain moment he could have been
challenged by Gelfand , i n the game with whom the Norweg ian had to defend by find
ing virtually the only moves. I n the fi rst three rounds Carlsen had to be satisfied with
d raws, but at the critical moment he as though acquired a second wind and won four
games in a row. What operated was the phenomenon of Mag nus's 'Fischer style' of
play, ready to fight until resources are completely exhausted, and winning apparently
drawn positions. According to Gelfand: 'Carlsen has an incred ibly developed "ki l ler

257
 Chapter 3
Life at the Top

instinct": to players who are not i n their best form , he gives no quarter - neither with
Wh ite, nor with Black.' The impression is such that, u nder the magic of his powerful
play, some players, including ones with high ratings, come along to a game as though
doomed beforehand.

Game 60 a b c d e f 9 h
L.-0. Nisi peanu - M. Carlsen s.!. ..t.'ii' ... 8
Bazna 201 0
Sicilian Defence [ 876]
7 .i. 7
6
6
1. e4 c5 2. f3 d6 3. d4 cxd4 4. xd4 5
,---}
5
f6 5. c3 g6 6. J.e3 J.g7 7. f3 c6
8. 'Wd2 0-0 9. 0-0-0
4
li:J!j, 4

T h e fi rst 'warn i n g b e l l ' , i n d i cat i n g t h e

3
o/ ib, fj, 3

m a g i c a l effect of Carl s en o n h i s o p
ponents. N is i peanu i s o p e n l y afrai d of
Mag n u s , and he avo ids the critical m ove
9. Ac4, in w h i c h the l atter had gained
several b r i l l i a n t wins, i n favour of the
old conti n u at i o n. After 9. 0-0-0 Wh ite
need not fear Black's cou nterplay after
...Ad?, ...l::!: c 8 and ...&Dc6-a5-c4, since
h i s bishop can captu re the k n i g ht on c4
not from b3 (to where it retreats after be
i n g deve l o ped at c4), but from f1 , sav i n g
a whole tem po. B u t he h a s t o reckon with
the attack i n the centre, as occu rred in
the game.
9...d5 10. b1
The second 'bel l ': N isi peanu avoids the
main continuation 1 0. exd5 &Dxd5 1 1 . &Dxc6
bxc6 , where White w i n s a pawn i n the
variations 1 2. &Dxd5 cxd5 1 3. 'Wxd5 and
1 2. Ad4 e5 1 3. Ac5 l::!:e 8 1 4. &Dxd5 cxd5
1 5. 'Wxd5, with active counterplay for Black.
The tactical justification of 1 0. b1 is that
1 0 ...dxe4? is not possible on accou nt of
1 1 . &Dxc6 Wxd2 1 2 . &Dxe7+. But Carlsen is
also prepared for this turn of events, as is
indicated by his 1 3th move.
2
fj, fj, fj,'fl fj,fj, 2
1 :g ib, :g 1
a b c d e f 9 h
10 .. Jab8
N i s i pe a n u was p r o b a b l y a1m1ng for
1 0 ...6Dxd4 1 1 . e5! (an i m portant i nterpo
sition) 1 1 ...6Df5 1 2. exf6 exf6 (1 2 ...Axf6
1 3. &Dxd5 'Wxd5! 1 4 . Wxd5 &Dxe3 1 5. Wd2
&Dxd1 1 6. Wxd1 Ae6 leads to very sharp
play, where Black has compensation for the
sacrificed queen) 1 3. Ac5 d4 14. Axf8 Wxf8,
being gu ided by the game Short - Carlsen
(London 2009), which contin ued 1 5. &Db5
&De3 1 6.l::!:c 1 Ah6 1 7. 'Wxd4 &Df5 1 8. Wc3
Axc1 1 9. xc1 Ad? 20. Ad3l::!:c 8 2 1 . 'Wd2
Axb5 22. Axb5 Wc5 with equal play, where
he had most probably prepared a surprise.
But in the Dragon it has already been
establ ished that 1 0 . . l::!:
. b 8 is a usefu l move
(an indication being the emergence of the
'Ch inese Dragon' in the variation 9. Ac4
Ad? 1 0. 0-0-0 l::!: b 8), after which it is un
favou rable for Wh ite to capture &Dxc6, and
at the same time he has to reckon with the
threat of ...e7-e5. The tactical justification
of the rook development at b8 is reinforced

58
 L.-D. Nisipeanu- M. Carlsen
by the fact that 1 1 . Af4? is u nfavourable,
as there fol lows 1 1 ...e5! 1 2. tDxc6 bxc6
1 3. Axe5 tDxe4 1 4. d4 c5! 1 5. xe4 dxe4
1 6. f!xd8 AxeS 1 7. f!xf8+ xf8 18. c1
exf3 1 9. gxf3, and Wh ite has a strateg ically
lost position.
1 1 . tildb5
The most popular move - Wh ite plays his
knight to a7 with the idea of exchang ing
the bishop on c8. After the retreat 1 1 . tDb3
Black can go into an endgam e - 1 1 ...dxe4
1 2. xd8 f!xd8 1 3. f!xd8+ CDxd8 1 4. Axa7
f!a8 15. Ab6 Af5= (L. Milov - Ovsejevitsch ,
N i kolaev 1 995).
11 .. a6 1 2. tila7 e6 1 3. g4
N ot h i n g i s g i ven by t h e s i m p l ifyi n g
1 3. exd 5 exd5 1 4. tDxc8 f!xc8 1 5. tDxdS
tDxdS 1 6. xdS xdS 1 7. f!xd 5 tDb4
1 8. f!cS (or 1 8. f!d2 f!fe8 with the idea
of 1 9. Af2 Ah6!) 18 ...f!ce8! 1 9. Ad2 f!d8
with equal chances (Sjugi rov - N arayanan ,
Kirishi 2006).
1 3 Jae8!
.
Black prepares for the opening of the posi
tion after the captu re on dS.
1 4. g5 tilh5
8 8
5 example, White cannot play 1 8. tDxc6?
4 4
3 3
B 2
g1
a b c d e f g h
----------------------
ctJ
N ot 1 4 ...tDd7? 1 5. exd5 exd5 1 6. tDxdS
tDxa7 1 7. Axa7 f!a8 1 8. Af2 with an over
whel ming advantage for Wh ite.
1 5 Af2?!
The i m pression is that Carlsen has outwit
ted his opponent, and it has to be admitted
that the knight raid to a7 has proved inef
fective. If 1 5. tDxc8 Wh ite has to reckon
with 15 ...d4! , while i n the event of 1 5. exd5
exd5 1 6. tDxdS Ae6 1 7. c4 tDeS the knight
cannot escape from a7, and he is forced to
return the pawn by 1 7. tDxc6 bxc6 1 8. tDf6+,
transposing i nto an i nferior endgame. But
after the move in the game Wh ite is un
able to exchange his knight, which 1 5. Ac5
wou l d perm it, preventing 1 5 ... Ad7? be
cause of 1 6. exd5 exd5 1 7. tDxdS! with the
threat of 1 8. tDxc6 bxc6 1 9. tDe7+. Better
here is 1 5 ...a5 1 6. e3 with the alterna
tive of sacrificing a pawn - 1 6 ...d4 1 7. tDxc6
xc5 1 8. tDxd4, obtaining a powerful dark
square bishop as com pensation, or going
into a roughly equal endgame - 1 6 ...Axc3
1 7. xc3 xc3 1 8. bxc3.
15 Ad7!
The loss of a pawn is of no sign ificance,
since the knight on a7 is not even a notori
ous 'Tarrasch knight', but rather a 'donkey'
in a trap.
1 6. exd5 exd5 1 7. 'Wxd5
After 1 7. tDxdS Ae6! Black has more than
sufficient compensation for the pawn. For
bxc6 1 9. tDf6+ tDxf6 20. gxf6 xf6, when
without l oss of m aterial be is u nable to
withstand the massed attack of the black
pieces. But possibly he al ready had no
choice other than to go 'fishing in trou bled
waters', by playing 1 8. c4 tDeS.
259
 Chapter 3
Life at the Top

1 7...tbe5 l:!bd8) 2 1 . 'Wc1 c'll c 4, with clear and easy

play for Black.
The trap has operated - the knight is out
1 8 ...'Wxg5 1 9. tbe4 'Wf4
of play f or ever.
a
The outcome is decided, since Black is
b
effectively playing with an extra piece.
8 X 8

7 QJ I. 20 . .te2 .te6 21 . Wa3 tbc4 22 . .txc4

J.xc4 23. tbd6 .tf8 24. fld4 We5
6 I. 6
25. f!xc4 .txd6 26. Wd3 .tf8 27. a4
5
I.& I:Ibd8 28. Wb3 'Wd5 29. I:Ic3 Wd2
4 4
30. Ae3 We2
3 Qj 3

2 Ji 2 a b c d e f g h

M Ji M1 8
a b c d e f g h 7 QJ I.
1 8. Wb3?
6 I.
5

W h ite g i ve s u p h i s g 5 - paw n , w h i c h
c o u l d h ave b e e n d efe n d ed: 1 8. Ae3 3
'We? 1 9. Ae2 (or 1 9. f4? Ae6 20. 'Wa5 2
c'll g 4 and Wh ite l oses after 2 1 . Ac5 b6
22. Axe? bxa5 , wh ile after 2 1 . Ab6 the
M1
a b c d e f g h
p l a c i n g of his pieces looks strange, to
say the least: Black can cont i n u e either 31 . .tb6?
2 1 ...Axc3 or 2 1 ...Af5 with the th reats of
...c'llxf4 and ...Axc3) 19 ... Ae6 20. 'Wd2 and Wh ite resigned, without waiting for
l:!bd8 (20 ...c'll c 4 21. Axc4 Axc4 22. c'll e 4 31 ...l:!d1 + 32. l:!xd1 'Wxd1 + 33. a2 l:!e1 .

I n the following game it was a top-class player who fel l under the magic of Carlsen's
i nfal libil ity. After obtai ning an almost winning position, at a critical moment Ruslan
Ponomariov took Carlsen 'at his word ', was frightened by some illusory threats, and
i n the end he lost.

260

Game 61
R. Ponomariov - M. Carlsen
Bazna 201 0
King's Indian Defence [ E81]
1 . d4 t'Of6 2. c4 g6 3. t'Oc3 b.g7
I n recent times, possibly under the i nfl u
ence of Garry Kasparov, the King's I n d ian
Defence has ap peared in Carl sen's rep
erto i re . But his choice of it at a decisive
stage of the tournament, in the game with
Ponomariov, apparently took acco unt of
the variation employed by Ruslan against
Radjabov at t h e start of t h e t o u r n a
m ent.
4. e4 d6 5. f3 0-0 6. t'Oge2
It was this move, employed by Ponomariov
against Radjabov, the leading King's Ind ian
player of modern ti mes (the classical con
tin uation is 6. Ae3), that drew Magn us's
attention in prepari ng a new line of play.
a b c d e f
8 .!. 1.1\ .i.. 'iiV .!.
7.\.l .l.l.i...\ 7
6
5 5
4 888 4
3
8 3
288 CD 88 2
1 .: .!,.!, .: 1
a b c d e f
6 t'Obd7
Radjabov played 6 .. . e5?!, and after 7 Ag5!
he experienced some difficulties. There
fore Carlsen develops his knight, reducing
the potential of 7. Ag5. But it should be
mentioned that, when Wh ite delays the
development of h i s bishop, Black m ost
R. Ponomari ov- M. Carlsen
often responds with an i m med iate flank
attack - 6 ...c5.
7. b.e3 c5
Here i s the su rprise. It tra n s p i res that
Carlsen is not even contemplating ...e7-e5
and he prefers to transpose i nto a Modern
Benon i structure, where the bishop at e3
is not on its best square.
8. d5 fOeS 9. t'Og3 h5
The play is based on nuances. Black hurries
to drive the knight to f1 , as Wh ite can not
prevent the advance of the pawn by 1 0. h4,
since in this case, because of the weakness
of the g4-square, he will never be able to
carry out the thematic f3-f4. After 9...e6
1 0. Ae2 exd5 1 1 . cxd5 a6 1 2. a4, a natural
move order for the Modern Benon i , 12 ...h5
proves ineffective, since after 1 3. 0-0 h4
1 4 . tDh1 the knight switches to the excel
lent post f2, from where it forestalls Black's
threats on the kingside.
9 h
1 0. b.e2 h4 1 1 . t'Of1 e6 1 2. t'Od2
8
1 2 . f4 suggests itself, but then Black is not
obl iged to retreat, since after 1 2 . . .tDeg4
6
1 3. Axg4 tDxg4 1 4. xg4 exd5 1 5. f5 d4
he regains the piece, obtai ning good play.
12 . exd5 1 3. cxd5
8
7
9 h
6 6
5 5
4 .l 4
3 3
a b c d e f 9 h
61
 Chapter 3
1 3 .. . Ad7
At first sight this is a transposition of moves
com pared with 1 3 ...a6 1 4. a4 Ad7 1 5. 0-0
b5, which has been played in n u merous
games. The best known of these, Vyzhma
navin - J. Polgar (Gron ingen 1 994), contin
ued 1 6. axb5 axb5 1 7.l::!x a8 xa8 1 8. h3?!
c4 1 9. e1 (1 9. f4 tbd3 20. Axd3 cxd3
21. b4 a3 is advantageous to Black) 19 ...
b4 20. tDd1 a2! 2 1 . xh4 Aa4 22. f4 tDd3
with good play for Black. But, of course,
1 8. Axb5 Axb5 1 9. tbxb5 a6 20. tbc3
l::! b 8 is more critical , with com pensation
for the pawn typical of such structu res
(Psakh is- Piket, Amsterdam 1 990), or i m
med iately 1 7. Axb5 Axb5 1 8. tbxb5 l::! b 8.
Peter Heine N ielsen recommends 1 3 . ..
h3!? with the idea after 1 4. g4?! of sacrific
ing the bishop by 14...Axg4 1 5. fxg4 tDfxg4
1 6. Af4 Ah6 1 7. AxeS tbxe5 with excel lent
com pensation in the form of complete con
trol of the dark squares. Therefore 1 3...h3
should be answered by 14 g3.
1 4. 0-0 b5
This is the idea of 1 3 ...Ad7, but after the
next few moves one gains the i m pression
that the prepared line is not too success
ful. But there are no sensible alternatives.
If 14 ...a6, with the idea after 1 5. a4 b5 of
transposing into the fami l iar position, Wh ite
can play 1 5. h3! with the threat of f3-f4. It is
too late for 1 4. . .h3 1 5. g4 a6 in view of the
unpleasant 1 6. g5! fol lowed by f3-f4, and
therefore strong measures are necessary.
1 5. xb5 Axb5
If 1 5 . . . b 6 W h ite rep l i es with the typi
cal manoeuvre 1 6. tba3 xb2 1 7. tbc4,
obtai ning a serious positional advantage.
1 5 . . . e7 came into consideration, defend-
Life at the Top
ing the d6-pawn , and then attacking the
knight with . ..l::!a b8.
1 6. Axb5 :Sb8 1 7. a4
a b c d e f g h
17 h5
. .
I n pri n c i ple, a d u bious sortie, removing
Black's control of the g4-square - after
all, Wh ite will not allow the knight to go
to f4. But Carlsen was undoubted ly look
ing further ahead. I n the event of 1 7... h3
1 8. g4 (1 8. gxh3 c8) 1 8 ...a6 1 9. Axa6
l::!x b2 20.l::! b1 White exchanges the active
rook and Black has no compensation for
the extra passed pawn. The compensa
tion for the sacrificed pawn is also insuf
ficient after 17...a5 1 8. e2 (or 1 8. c2 a6
1 9. Ae2l::! b 4 20. b3) 1 8 ...a6 1 9. tbc4 ttlxc4
20. Axc4 tbd7 2 1 . l::!f b1 . It would appear
that Carlsen recognises how d ifficult his
position is, and he openly bl uffs, cal mly
staking everything.
1 8. f4!
After this move White has an obvious ad
vantage, but all the m i racles lie ahead.
18 .. d7 1 9. 'Wg4 a6! 20. Axa6
After 20. Axd7 tDf6 2 1 . xh4 tbxd7
22. xd8 l::!fxd8 Black regains one of the
Q62
 R. Ponomariov- M. Carlsen
------ --------====

sacrificed pawns (23. c4 b6 24. a5 by no means ephemeral , for exam ple, i n

.e8), retaining qu ite good compensation the variation 24. xh4 xe4 2 5 . xd 8
for the other one. Ponomariov prefers to .xd8 26. Ae1 ? g3+! 27. hxg3 g7 etc.
have a passed pawn. But is it so simple, when short of time, to
fi nd all the n uances in the defence of the
20... laxb2 21 . laab1
king? Apart from 24. xh4 W h ite also
I
Aiming for the exchange of the active rook. has to calcu late 24. f3 g3+ 25. hxg3
But Carlsen carries out his plan of an at hxg3 and fi n d a series of exact m oves:
tack on the king, begun with his 1 7th move. 26. Ae1 !, with the cal c u l ation branches
26 ...g7!? 27. Axg3 .h8+ 28. Ah2 xe4
21 . laxd2!
29. .b8! and 26...xe4!? 27. Axg3 xg3+
Everything else is hopeless! 28. xg3 g7 29. .f2!. Such resou rces,
incidental ly, already on the 7th move of
22. Axd2 Ad4+
the calculation (beg i n n i ng with 23. h1 )
a b c d e f g h
are not easy to d iscover. But now the ad
vantage passes to Black .
8 'iV .!e 8

7
1.&\ 7 23 ... Axf2+ 24. xf2 df6 25. Yf3
6 Ji 6 We8!
5 1.&\ 5 The e4- and a4-pawns are u nder attack.
4 .i. 4 Which of them should be defended?
3 3 b f
a c d e g h
2
iL 2 8
'i.1. 8

g g 7 7
a b c d e f g h
6 Ji '-' 6

23. laf2?
5
1.&\ 5
4 4
T h e b l uff works! Po n o m ari ov, taki n g 3 3
Mag n u s a t h i s word , b l o c ked w i t h his
rook almost without t h i n k i n g . Of course,
commenting on the game on the I nternet,
a b c d e f g h
and especially in the qu ite of one's study,
one can consu lt with Rybka and i n a few 26. e5
m i n utes obtai n an answer regard i n g the
safety of 23. h1 df6 . It is another mat Ponomariov takes the more energetic deci
ter to take such a decision at the board. sion. In the event of 26. Ad3 xa4 27. .c1
Po n o m ariov o bv i o u s l y i m agined t h e re d4+ 28. f1 .e8 29. Ac3 xe4 30. Axd4
were t h reats i nvo l v i n g k n i g ht leaps to xd2+ 31 . f2 xf3 32. xf3 cxd4 Wh ite
g3, when his king wou l d come under a still has problems. 26. Ab5 ti:)xe4+ 27. e3!
cross-fi re not only on the a7-g1 d iagonal, xd2+ 28. xd2 is stronger, with a defen
but also the h -fi l e. These th reats were sible position.

263
 Ch apter 3 Life at the Top

26 ...'Wxa4 27. exf6? 29...'Wf5

This is the consequence of the psycho Now, despite the material equality, White's
logical shock, following the loss of a cer position is strategically hopeless.
tain advantage. After the straightforward
30. l:!f1 l:!e8 31 . Aa1 l:!a8 32. 'We3 h7
27. Ab5 &De4+ 28. e1 &Dxd2 29. Axa4
33. Ab2 Iib8 34. Ac1 Iib1
&Dxf3+ 30. gxf3 &Dxf4 31 . exd6 d8 32. d7
&Dxd5 33. c1 White could have hoped to By placing his rook on its optim um square,
save the endgame. Black has put White in zugzwang. The d5-
and f6-pawns cannot be simu ltaneously
27...'Wxa6 28. Ac3 'Wc8?
defended.
Carlsen too has to pay for his past experi
35. 'We8 Wxd5 36. f5 gxf5 37. 'We3
ence. He intends to switch his queen to
f5 , whereas the game wou l d have been
a b c d e f g h
quickly concluded by 28...c4 with an
8 8
attack not only on the d5- and f4-pawns,
but also indirectly on the f6-pawn (29. f5
7 I. 7

&Dxf6! 30. Axf6 C2+). 6 6

a b c d e f g h
5 I. '.& 5

4 I. 4
8
.. . 8
3 3
7 I. 7

6 6
2 2

5
'.& 5 .1 :g
a b c d e f g h
4 I. 4

3
'if 3 37 ...Wd4
2 2
37... f4 was more forceful: 38. f2 (38. h3
d4+ 39. h1 xf6) 38...g6 39. xh4
a b c d e f g h
d4+ 40. f2 xf2 41 . xf2 d 5. How
ever, the passed d-pawn also guarantees
29. g1?
a win. Carlsen's su bsequent technique is
Ponomariov misses a chance opportu nity impeccable.
to exchange the queens, making it harder
38. 'Wxd4 cxd4 39. A g5 l:!xf1 +
for Black to convert his advantage: 29. f5!
40. xf1 h3 41 . gxh3 g6 42. Ah4 <tlf4
xf5 30. xf5 gxf5 3 1 . b6! d8 32. f3.
43. A g3 g5 44. Af2 <tle6 45. e2 f4
And since if 32 ...h7 there is the strong
46. f3 d3 47. h4+ f5 48. Ab6 <tics
reply 33 Ae1 ! h3 34 gxh3 g6 35 Ag3!,
49. h5 d2 50. e2 <tle4
Black wou l d have had to fin d 32...a8!
with winning chances (variation by Nielsen). White resig ned.

264
 M. Carlsen - E. Bacrot

I n July 2 0 1 0 Mag n u s su rprised t h e c h ess world by beg i n n i n g an 'alternative ca

reer' - with h i s rugged good l o o ks h e became a fas h i o n m o d e l fo r t h e fi rm G - Star
Raw. A few weeks l ater h e took part in the Raw Wo r l d C h ess C h a l l e n g e in N ew
Yo rk, with over 4 0 , 0 0 0 o p p o n ents vot i n g o n l i n e for t h e best m ove out of t h ose
sel ected by an e l ite g ra n d m aster pan e l , and Garry Kas parov a m o n g the c o m m e n
t a r y team. Whether t h i s w a s a l l p rov i n g a d i stract i o n i s not c l ear, but i n Carlsen's
n ext event, the Arct i c Securities C h ess Stars , a rap i d - p l ay eve nt of fo u r p l ayers
h e l d in h i s h o m e cou ntry and o rgan ised by the fi rm w h i c h was n ow sponsoring
h i m , h e i n it i a l l y had to p l ay second fiddle to V i s hy Anand. The world c h a m p i o n
d o m i n ated the a l l - p l ay-a l l sect i o n b y scori ng 5/6 as o p posed to Carlsen's 31/2,
but the fo rmat provided fo r a two-game p l ay- off between the top two, and here
Carlsen p reva i l e d 1 %-1/2.
Carlsen o pted to l ead the Norweg ian team at the 39th O l y m p i ad , held i n the
remote S i berian town of Khanty- Mansiysk, although he m ust have known that he
wo u l d m a i n l y be fac i n g l ower- rated p l ayers and i n danger of losing rat i n g poi nts.
W h i c h i s what in fact happened, as he suffered as m any as t h ree d efeats with
Black, i n c l u d i n g one to Michael Ad ams after ventu ring the experi mental ' N orth Sea
Defence' 1 . e4 g 6 2. d4 tDf6 3 . e5 tDh5. Then i n the G rand Slam Masters i n B i l bao
(FI D E category 22!) , he fi n i shed t h i rd beh i n d Kram n i k and Anand in t h i s fo u r- p l ayer
d o u b l e - ro u n d event, his o n l y win c o m i n g at the expense of the botto m-p l aced
Al exey S h i rov.
Tal k of a crisis in Carlsen's play was dispelled by his fine victory in the third Pearl
Spring tou rnament in Nanji ng (FI D E category 2 1 ) . Facing five 2700+ g randmasters,
including world champion Vishy Anand and former FIDE champion Veselin Topalov,
Carlsen went through the ten rounds undefeated and with 7/1 0 fi n ished a point ahead
of the 2nd-placed Anand. An incisive attack against Etienne Bacrot made a particularly
pleasing impression.

Game 62 as i nnocuous as it doesn't pose Black any

M. Carlsen E. Bacrot
- particular problems.
Nanji ng 201 0
5 ... Ab6 6. c3
Scotch Game [ C45]
Carlsen played 6. Y!fe2 in four games at the
1. e4 e5 2. f3 c6 3. d4 exd4
World Bl itz Cham pionsh i p in N ovember,
4. xd4 Ac5 5. b3
but i n a 'real game' he prefers someth ing
Kas parov tended to favou r 5. Ae3 o r more mainstream ...
5. tbxc6, but his protege prefers this slightly
quieter treatment. Some would describe it 6...f6

265
 Chapter 3
Life at the Top

a b c d e I g h weakening than on the c-file (after 11. tbxd4

.I .I Axd4 and 1 2 ... Axc3) proves to be spot on .
a
.t'ii' a

7AAAA AAA 7 11 ttlxb3 1 2. cxb3 9:e8 13. 0-0-0

..

6
.tl.a\ 1.1\ 6

5 5 a b c d e I g h

4 ]= 4 a .I .t 'i .1 a

3
Qjt[j 3 7 AAA AA 7

2 2 6
.t 1.1\ A6
1 ,S ibib .s 1 5 A 5

a b c d e I g h 4 ib 4
3 QJW/ 3
7. Ye2!? 2 2
... or does he? This seems to have been .s ib .s 1
a b c d e I g h
fi rst played by the 1 1 -year-old Sergey Kar
jakin, and not very successfu lly at that. To W h ite s i g n a l l ed h i s i ntent i o n to castle
date it has not had a good track record in queenside as soon as he played 7. 'We2.
terms of results scored. Added to his dubi His pressure along the d-fi le and the h4-d8
ous open ing experiments at the Olympiad, d iagonal com pensate for the structural
further questions might have been asked weakness of his queenside.
of Carlsen's state of mind had this game
gone awry, but this proves to be one of his 13 .. d6 14. Yc2 Ad7?!
more inspired ideas. 14...c61ooks more natural, depriving White
7...0-0 of the use of b5 and d5 for his minor pieces.

The solid 7. . . d6 has scored wel l for Black, 1 5. Ac4 Ae6

with 8. Ag5 h6 9. Ah4 and now 9 .. . g5 be
Moving this piece twice along the same
ing a possibil ity, as Black has yet to castle
diagonal gives the impression that Bacrot
kingside.
is unsure of himself. Perhaps he intended
8 .tgs h6 9. Ah4
1 5...Ac6 but then feared an i m m i nent g
pawn advance by Wh ite. Whatever his
9. h4?! d6 1 0. f3? hxg5 1 1 . hxg5 tbg4! (any
t h i n ki ng, we can adapt an o l d p roverb
thing else and Wh ite will play f4 followed
and conclude that 'he who hesitates (when
by '11V h 5) 1 2. fxg4 'Wxg5 1 3. '11Vf3 Axg4 and
playing Black against Carlsen) is lost '.
Black al ready had a winning position in the
game Rublevsky -Anand, Bastia Rapidplay 1 6. 9:he1
2004.
1 6 . e5? wou l d cost Wh ite a pawn after
9 . a5 1 0. a4 ttld4 1 1 . Yd3! 1 6 ...Axc4 and a capture on e5, but now
the e-pawn advance becomes a real threat.
Carlsen's judgement, that the forthcoming
doubled pawns on the b-fi le will be less 1 6...Ye7

266
 M. Carlsen- E. Bacrot

Black's o ption was to be brave and try advantage. Black's queen is overworked
1 6 ...g5 to drive away the annoying bishop, trying to defend the vulnerable king and he
but the resulting weakness of his kingside now tries to get his poorly-placed bishop
would be a problem. back i nto the game.
a b c d e f g h 25 .. .tc5 26. e4 .te7 27. :h3! g7?
a !. I. a
Of course, not 27...Wxe4? when 28. Wxh6+
7 .t..\ 'iV.t..t. 7
soon leads to mate, but the text move also
6 .i. ... .. ... 6
has a tactical flaw. 27...l::!.d 8 28.l::!.x h6l::!.x d2
5 ... 5
29. l::!.xh7+ xh7 30. tDxd2 is better, but
4 [j, iL [j, iL 4 it leaves Wh ite a pawn up in a probably
3 [j,{jj 3 w i n n i n g endgame. 27. ..g 8 28. Wf4 f5
2 [j, [j,[j,[j, 2 29. l::!.x h6 Wg7 30. l::!. h 3 also costs Black
a pawn.
1 :;g:;g 1
a b c d e f g h
28. 'Wd7 ! f7

17. e5! dxe5 18. :xe5 Yf8 28 . . . Wxe4 l oses to 29. Wxe7+ g 6
30. l::!.g 3+ f5 3 1 . W h7+ e5 32. l::!. e 3 etc.
Th is allows a weakening of Black's king
a b c d e f g h
side, but it may be the best of a bad job.
8
18 ...c6 would prevent the deadly threat of
tDd5, but 1 9. tDe4 is sti l l very strong; 1 8 . . . 7

g5 is sti l l a possibil ity, b u t with t h e same 6

drawbacks as before. 5 ... 5

19. .txf6 gxf6 20. :e2 Wg7 21. J.xe6 4 [j, 4

a:xe6 22. :xe6 fxe6 23. :d3 h8 3 [j,

24. a:g3 Wh7 25. Wd2 2 [j,
a b c d e f g h

a b c d e f g h
a !. a

7 ... 'iV 7
29. g5+!
6 .i. ... 6
Computers poi nts out that 29. tDxf6! is
5 ... 5
simi larly lethal.
4 [j, 4

3 [j,{jj :;g 3 29...fxg5 30. :f3+ g8 31. Wxe6+ h8

2 [j, [j,[j,[j, 2 32. :f7 .td6

If the queen flees with 32 . .. W d 3 , then

a b c d e f g h
33. Wxh6+ and mate next move.

With a series of straig htforward m oves 33. :xh7+ xh7 34. Yf7+ h8 35. g3
Wh ite has established a clear positional a:a6

267
 Chapter 3 Life at the Top

35 ...f!f8 36. 'Wg6 .xf2 37. 'Wxh6+ g8 36. b1 .lb4 37. f4 gxf4 38. gxf4
38. 'Wxg5+ and the kingside pawns w i l l
decide. Black resigned.

The next el ite event - the Tal Memorial Tou rnament in Moscow - came too soon for the
Nanjing contestants to take part, but Carlsen travel led to the Russian capital for the
World Blitz Championshi p which followed i m mediately afterwards. After his stu n n i ng
victory the previous year, Carlsen was the clear favourite, especially as the players
from the Tal Memorial had enjoyed just one day's rest before the start of the 20-player
double-rou nd bl itz event. But Leven Aronian, one of the joint win ners of the Tal Memo
rial , overcame his fatigue to score a deserved win, and Carlsen , who lost as many as
ten games, had to be content with third place. However, Mag nus was delighted to be
presented with the 2009 Chess 'Oscar', the resu lt of a vote organ ised by the Russian
chess magazine 64.
Around this time Carlsen shocked and disappoi nted the chess world by announcing
that he was withdrawing from the world championshi p Cand idates matches, due to
be held in the spring of 201 1 . The reasons he gave were the changes made during the
cycle, and the fact that he considered the current system to be insufficiently 'modern
and fair'. There was speculation that advice from his former mentor Garry Kasparov
was beh ind the decision , but the ex-world champion flatly denied this.
I n the second London Chess Classic, held in December 201 0, i n itially it seemed
u n l i kely that Carlsen would repeat his success of the previous year: a su rprise loss
in the fi rst round to Luke McShane was followed by another in round 3 against Vishy
Anand. And yet, with the aid of his fighting spirit, a miraculous endgame save against
Vlad i m i r Kramnik i n the penulti mate round, and also the 'football' scoring system (3
poi nts for a win, 1 for a d raw) Carlsen nevertheless emerged the winner.
I n Jan uary 201 1 , as usual most of the world 's top players l i ned u p i n the Tata Steel
tournament in Wijk aan Zee. Carlsen made a hesitant start: two d raws were fol l owed
by a sensational loss to the 1 6 -year-old Anish G i ri - it must have been a novel expe
rience for the Norweg ian to lose to a player so much you nger than h i m self. Thi ngs
i m p roved , and a sacrificial attack in the 8th rou n d against the leader N akam u ra
brought Carlsen right back i nto contention. But this was fol l owed two rounds later
by a bad loss against lan Nepom n i achtc h i , and in the end Carlsen had to settle for
a share of t h i rd place, beh i n d N akam u ra, who achieved the best result of his career,
and the world champion Anan d .

268
 M. Carlsen- H. Naka m u ra

-=G a=m e 63 a b c d e f g h

M. Carlsen- H. Nakamura 8 .I ii ... 8

Wijk aan Zee 201 1 7 .i. 7
Sicilian Defence [892}
6 .I.&\ .i. 6
5 J!a\ 5
1. e4 c5 2. tilf3 d6 3. d4 cxd4 4. tilxd4
tilf6 5. tilc3 a6 6. e2 e5 7. t"ilb3 e7
4 Jt 4
8. e3 0-0 9. g4
3 tiJtiJ 3
2V/Jt 2
a b c d e t g h
.a. .a. 1
8 I. 1.&\ .i. iY I. 8 a b c d e f g h

7 .i. 7
1.&\ 6 16...Wc7 17. h5 Elfe8 18. t,!;>a1

5 5 A prophylactic move, ai med against a pos

4 4 sible knight check on a3 after the i m medi
3 tiJtiJ Jt 3 ate 1 8. tild4 tilbc4 1 9. e1 b6 20. Ac1
2 Jt 2 tila3+ 2 1 . a1 xd4 or 21 . ..tilc6.
1 .a. .a. 1 18...f8 1 9. tild4
a b c d e f g h

a b c d e f g h
An u n usual twist, with the aim of catching
the opponent off-balance. Normally Wh ite
8 I. l..i.9> 8
quietly castles kingside in this positio n , 7 .ii 7
whereas i f he wants t o castle queenside 6 .I.&\ .i. 6

he prepares it with f2-f3 (rather than Ae2). 5 1.&\ 5

9...e6 10. g5 tilfd7 11. h4 4 tiJJt 4
Th is position has occurred in very few top 3 tiJ 3
level games. An alternative for Black is to 2V/Jt 2
advance his a-pawn, with the aim of induc 1 .a. .a. 1
i n g the weakening a2-a4. For exam ple, a b c d e f g h

the game Khal ifman - Gelfand (Las Vegas

1 999) went 1 1 . d2 aS 1 2. a4 tilc6 1 3. f4
fS with double-edged play.
11 ...tilb6 12. Wd2 til8d7 13. f4 exf4
14. xf4 tile5 1 5. 0-0-0 lacS 16. t,!;>b1
The outcome is a fairly standard type of
position from the English Attack, where
Wh ite has advanced his pawn to f4 i n one
go, but his bishop may not be best placed
on e2.
19 ...Wc5?
Black is evidently aiming to bring his queen
to an attacking position at b4 or aS , but
this real ly is too slow. After the game Na
kam u ra sugg ested 1 9 ...g 6 , the log ical
consequence of his preced ing play, but
then Wh ite can continue 20. hxg6 hxg6
2 1 . tilxe6 l;!.xe6 22. Ae3 , the point being
that 2 2...tilbc4 23. Axc4 tilxc4 a l l ows

69
 Chapter 3 Life at the Top

24. Ad4! (or 24. '1Wh2 Ag7 25. Ad4). How a b c d e f g h

ever, the i m med iate 1 9 ...tbec4 1ooks quite
adequate for Black.
20. g6 ec4 21. Axc4 xc4 22. 'Wd3
Wh ite has to keep his knight on c3 pro
tected in the event of ...tbxb2.
.!.
8 8
7 .l 7
6
5 5
4 4
3 3

22 ... fxg6 23. hxg6 h6 24. Wg3 Wb6 2

25. Ac1 Wa5

After this Wh ite has to counter the threat

of Black capturing on b2 and then on c3, 29. Axh6! gxh6
followed by mate on a2. After some twenty
min utes' thought Carlsen fou n d a good If 29 . . . '1Wxd4 Wh ite gives mate with 30. Ae3
reply. fol lowed by 31 . h8+.

30. g7 Ae7
26. l::i d f1 e5
30 ... Axg7 loses immediately to 31 . tbf5
If now 26 .. . tbxb2 , then White's rook on d1
c? 32. tbxh6+ followed by 33. tbg4+ or
is no longer attacked , and after 27. tbxe6
33. <uf7+.
r....

xc3 28. Wf4 he is the one who forces

mate. 31. xh6 f7 32. Wg6 xh6 33. Wxh6
At6 34. Wh8+ rJ;Jt7 35. g8W+ i::i x g8
27. d5 Axd5 28. exd5 Wxd5 36. Wxf6+ rJ;Jea 37. l::i e1+

After this Wh ite's attack crashes through. Black resigned.

For the past two d ecades years th e Melody Amber to u rnament, with its gener
ous sponsorsh i p by Joo p van Oosterom and its u n ique fo rmat of b l i ndfo l d and
rap i d c h ess, has been one of the h i g h l i g hts of the e l ite c h ess calend ar. Sadly,
it was a n n o u nced that t h e 20th event, h e l d in Mo n aco in Marc h 2 0 1 1 , was to be
the l ast . As u s u a l , a truly ste l l ar l i n e - u p h ad been i nvited. Carlsen made the same
score as the p rev i o u s year when h e f i n i shed equal fi rst - 1 4% out of 22, i n c l u d
i n g a p h e n o m e n a l 9% out o f 1 1 i n the rap i d gam es! - but t h i s t i m e it was good
e n o u g h o n l y fo r seco n d p l ace beh i n d the i n s p i red Levon Aro n i a n , who fi n i shed
a fu l l p o i nt ah ead.
With no Li nares tournament i n 201 1 , and with Carlsen not taking part in the Can
didates matches, he had a lengthy break between Monaco and his next tournament
in Romania in June. Meanwhile the Candidates matches went ahead in Kazan and
produced a deserved but unexpected winner, Boris Gelfand. But there was also a
storm of criticism, as only three of the thirty main games produced a decisive result,

270
 M. Carlsen - L.- 0 . Nisipeanu

there was a spate of short d raws, and four of the matches were decided in rapid or
even bl itz games. The feeling was that Carlsen's principled stand against the current
worlp championsh ip format had been justified.
After a ten -week break Carlsen was g l ad to be back in action in Romania at the
5th Kings Tou rnament, a six- p l ayer d o u b l e - ro u n d event (FI D E category 2 1 ), held
this year i n Med i as. He began with an i n c i sive w i n over the Wijk aan Zee victor
N akam u ra , and after smoothly overc o m i n g N i si peanu after an i n n ocuous-looking
open i n g he led at the half-way stag e with 3%/5 . His o n ly real rival was Sergey
Karjaki n , who d rew l evel i n the p e n u l t i m ate ro u n d with a crush i n g 22-move win
over the u nderperfo rm i n g Vasi l y l vanc h u k. The stag e seemed set for a g randstand
fi n i s h , with the two players pai red together in the l ast round. But Mag n u s had one
of those days that he wou l d ' l i ke to forget ', and i n the end, despite play i n g Wh ite,
he was rel i eved to d raw and share fi rst p l ace with his o p ponent on 6% out of 1 0 ,
two poi nts ah ead of the rest of the field. Carlsen was p l aced fi rst o n the tie- break,
and his result was good enough to regain h i m the N o. 1 spot in the world rating
l i st , p u b l ished on 1 July 201 1 .

Game 64 he sacrificed a pawn for active play, and,

M. Carlsen- L.-0. Nisipeanu although heavi ly outgraded by the rest of
Med ias 201 1 the players, this gained him a very respect
Queen's Gambit Accepted {027] able resu lt.

1. d4 d5 2. c4 dxc4 3. e3 f6 4. J.xc4 7 . Axc5

. .

e6 5. f3 c5 6. 0-0 a6 7. dxc5
7...'t'/xd1 8. l:!xd1 AxeS is the more usual
A see m i n g ly i n nocuous l i n e , lead i n g to continuation.
an early exchange of queens, but Wh ite
retai ns a nagg i n g edge and Black has 8. 'Wxd8+ xd8
to defend accu rately. It was a favou rite
with Boris Spassky in the later years of a b c d e f g h

his career (in particular, he played it fou r J. a

times agai nst Bobby Fischer i n t h e i r ' re 7 7
turn match ' in 1 992), and it also su its the 6 . .&J 6
tec h n ical style of Vlad i m i r Kram n i k (for
5 5
exam ple, he played it against Garry Kasp
4 4
arov in their world championsh i p match
3
i n 2000).
Psyc h o l o g i ca l l y t h i s wou l d seem to
be a good c h o i c e aga i n st Livi u - D i eter
N isi pean u , the lead i ng Romanian player, a b c d e f g h

who is noted for his aggressive style. For

exam ple, in four of his games in Med ias 9. Ae2

71
 Ch apter 3 Life at the Top

9. bd2 is more popular, although this

may simply transpose.

9 m e7 10. ctlbd2 Ad7

..

An alternative set- u p i nvolves 1 0 ...b6,

...Ab7 and ...bd7.

11 . ctlb3 AdS

Now Black will be more or less obl iged to a b c d e t g h

concede the bishop pair, which he could

have avoided by 11 ...Aa7. 22 ...m ea

12. ctla5 a7 13. ctlc4 Ab5 14. b3 d8 This looks rather passive. Perhaps N i
15. Ab2 Axc4 16. Axc4 sipeanu should have aimed for counterplay
with 22 . . .cb4:
a b c d e f g h a) 23. a3 a2 24. l;txc8 l;txc8 25. Ad3 b5,
and Black's control of c3 promises him
8 8
reasonable chances, despite the exotic
. ... ... ... 7
position of his knight on a2;
..t.t. 6
b) 23. e4 l;txc1 24. l;txc1 xa2 25. exd5
5 5
xc1 26. Axc1 l;tc8 2 7. Axh6 l;tc3
4
il 4
28. Ae3 l;txb3 29. h5, with a sharp end
3
D, D, Cjj 3 ing where the wh ite h-pawn is more of
2 D, il D, D, D, 2 an i m med iate danger than Black's con
1 1:, nected passed pawns on the queenside.
1:. 1 '
a b c d e f g h
23. g5 hxg5

Wh ite's two bishops give h i m a slight ad- It would surely have been safer to keep the
vantage, but as yet there are no real weak- h-fi le closed by 23 ... h5.
nesses in Black's position. 24. hxg5 Ae7 25. m g2

16 ctlc6 17. fd1 aa8 18. h3 g6

.. A necessary preparation for the planned
19. g4 invasion on the h-fi le.

25 . ctlb6
Wh ite sets about g ai n i n g space on the
kingside. Again too passive. 25 ... cb4 26. a4 c6
followed by ...a5 would have given Black
19 h6 20. Af1 ac8 21. ac1 ctld5
.. some counterplay against Wh ite's queen
22. h4 side pawns.

72
 M. Carlsen - L.- 0. Nisipeanu

26. Ad3! a b c d e f 9 h

8
... . 8
A nice pawn sacrifice, which activates the
light-square bishop and clears the way for
7 ..t. . 7

a rook to go to h1.
6 . . 6

5 5
2 6 !0b4 27. Ae4 !Oxa2 28. xd8+
..
4
m xd8 29. h1

(see next diagram)

Suddenly Black is lost. All the white pieces a b c d e f 9 h

are very active and there is no satisfactory
defence against the invasion by the rook.
29...!Od5 30. !Oe5 f5 31. Axd5
B l a c k resig n e d in view of 31 ...e x d 5
32. h8+ c7 33. h7 d6 (or 33 ...d8
34. Ad4) 34. Aa3+ xe5 35. xe7 mate.
'I thin k I played a very decent game!'
(Carlsen)

The world championship match sched uled for 201 2 wil l be between two players wel l
into their forties. Mag nus Carlsen i s less than half their age - it i s abundantly clear that
time is still wel l on his side!

273
 Tournament and M atch Record

I n d ex

To u r n a m ent a n d M atch Record

Sven Magnus 0en Carlsen born 30th Novem ber 1 990 in T0nsberg , Norway. Gai ned
the I nternational Master title in 2003, and the I nternational Grandmaster title in 2004.
Became the No. 1 rated player in the world in January 201 0, a position which he regained
in July 201 1 . Awarded the Chess 'Oscar' in 2009 and 201 0.

MAJOR TOU R NAMENTS

2001 + =

Trol l Masters, Gausdal 40th-45th 3/9 2 2 5

Norweg ian Open Championsh ip, Oslo 9th-1 3th 5%/9 5 1 3
Gausdal Classics 8th 2Y2/9 0 5 4
Nordic Championships, Bergen ?1 st-80th 3%/9 2 3 4
Bavarian Open Championship, Bad Wiessee 5/9
Norweg ian Junior Championship, Porsgrunn 1 st 6/7

2002
Trol l Masters, Gausdal 24th 4%/9 3 3 3
Gausdal Classics 8th 2%/9 2 1 6
I nternational Shakkinet Tournament, Helsinki 1 0th 4/1 1 3 2 6
Open Tou rnament, Bergen 1 1 th-1 8th 5/9 4 2 3
Open Tournament, Kiel 1 5th 6/9
European under-1 2 Championship, Pen iscola 6th 6/9
World under-1 2 Championship, Heraklio 2nd 9/1 1

2003
Trol l Masters, Gausdal 3rd-6th 7/1 0 6 2 2
Valle d 'Aosta Open 73rd 5/9
Gausdal Classics 1 1 th-1 2th 4/1 1 2 4 5
Norweg ian Championship, Fredri kstad 3rd-7th 5%/9 3 5
Pol itiken Cup, Copenhagen 7th-1 6th 8/1 1 6 4
Schwarzacher Open 2nd-8th 6Y2/9 5 3
European Under-1 4 Championship, Budva 3rd 6%/9 6 2

274
 Tournament and M atch Record tb

+ =

World Under-1 4 championship, Hal kidiki 9th-1 3th 7%/1 1 7 3

Pecaut Memorial , Taormina 2nd-5th 5%/9 3 5 1
Fi rst Saturday, Budapest 4th-5th 8/1 3 5 6 2

2004
I nternational Tou rnament, Wijk aan Zee C 1 st 1 0%/1 3 9 3 1
Aeroflot Open , Moscow 40th 5%/9 5 1 3
Reykjavik Open 33rd-43rd 4%/9 4 1 4
Dubai Open 2nd-1 1 th 6%/9 4 5 0
Sigeman Tou rnament, Mal mo/Copenhagen 3rd 5%/9 4 3 2
N orwegian Championship, Molde 1 st-2nd 7/9 6 2 1
Pol itiken Cup, Copenhagen 4th-1 3th 7%/1 0 5 5 0
5th Young Masters, Lausanne 5th 6%/1 1 4 5 2
Gausdal Classic 5th 5/9 4 2 3
Essent Crown, Hoogeveen 4th 2/6 1 2 3

2005
Smartfish Masters, Drammen 8th-9th 3/9 1 4 4
I nternational Tournament, Wijk aan Zee B 7th 7/1 3 3 8 2
Gausdal Classic 6th-8th 4/9 3 2 4
I nternational Tournament, Leon (rapid) 3rd-4th 1 /4 0 2 2
European Championship, Warsaw 1 8th-53rd 8/1 3 5 6 2
Norweg ian Championship, Sandnes 1 st-2nd 7/9 6 2 1
Play-off v. Agdestein, Oslo 2nd 3%-4% 2 3 3
I nternational Tou rnament, Biel 6th 4/1 0 0 8 2
6th Young Masters, Lausanne 5th 5%/8 4 3 1
Bygger'n Masters, Gausdal 1 st 8/9 7 2 0
Scanda Cup, Skanderborg 8th 4/9 1 6 2

2006
I nternational Tou rnament, Wijk aan Zee B 1 st-2nd 9/1 3 6 6 1
Open Tournament, Reykjavik 6th-9th 6%/9 6 1 2
I nternational Tou rnament, Sarajevo 1 st-3rd 5%/1 0 1 9 0
Midnight Sun Tou rnament, Tromso 2nd 7/9 6 2 1
N orweg ian Championship, Moss 1 st-2nd 7/9 6 2 1
Play-off v. Agdestein, Oslo 1 st 3-1 2 2 0
I nternational Tournament, Biel 2nd-3rd 6/1 0 4 4 2
Rising Stars v. Experience, Amsterdam 1 st 6%/1 0 4 5 1
World Bl itz Championship, Rishon Le Zion 8th 7%/1 5 5 5 5
I nternational Tou rnament, Cap d 'Agde (rapid) sem i -final 7/1 1 4 6 1
Tal Memorial Tournament, Moscow 8th-9th 3%/9 0 7 2

75
 Tournament and M atch Record

2007 + =

I nternational Tou rnament, Wijk aan Zee 1 3th-1 4th 4%/1 3 0 9 4

I nternational Tou rnament, Morelia/Linares 2nd 7%/1 4 4 7 3
Melody Amber Tou rnament, Monaco 8th-9th 1 0%/22 3 15 4
I nternational Tou rnament, Gausdal 1 st 7/9 5 4 0
Sparkassen Tou rnament, Dortm und 6th 3/7 0 6 1
I nternational Tou rnament, Biel 1 st-2nd 5%/9 4 3 2
Bl itz play-off v. Onisch uk 1 st 3/5 1 4 0
Open Tou rnament, Tromso 2nd-4th 7/9 5 4 0
Bli ndfold World Cup, Bil bao 2nd-3rd 6/1 0 4 4 2
Tal Memorial Tournament, Moscow 3rd-5th 4%/9 1 7 1
World Bl itz Championship, Moscow 9th 20%/38 15 11 12

2008
I nternational Tou rnament, Wijk aan Zee 1 st-2nd 8/1 3 5 6 2
I nternational Tou rnament, Morel ia/Linares 2nd 8/1 4 5 6 3
Melody Amber Tou rnament, N ice 2nd-5th 1 2/22 7 10 5
Grand Prix Tou rnament, Baku 1 st-3rd 8/1 3 4 8 1
Aerosvit Tou rnament, Foros 1 st 8/1 1 5 6 0
I nternational Tou rnament, Biel 3rd 6/1 0 3 6
I nternational Tournament, Mainz (rapid) 2nd 4%/1 0 7 2
Tal Memorial Blitz Tou rnament, Moscow 3rd 2 1 /34 16 10 8
Grand Slam Final, Bil bao 2nd-3rd 1 3/30* 3 4 3
I nternational Tou rnament, Cap d 'Agde (rapid) sem i -fi nal 8%/1 3 5 7 1

2009
Aker I nternational Tou rnament, Gjovik (rapid) 2nd 4%/8 3 3 2
I nternational Tou rnament, Wijk aan Zee 5th-6th 7/1 3 2 10
I nternational Tou rnament, Linares 3rd 71/2/1 4 3 9 2
Melody Amber Tou rnament, N ice 4th 1 3/22 8 10 4
Mtel Masters Tou rnament, Sofia 2nd-3rd 6/1 0 3 6 1
I nternational Tou rnament, Leon (rapid) 1 st 7%/1 3 4 7 2
Sparkassen Tou rnament, Dortm und 2nd-4th 5%/1 0 2 7 1
Pearl Spring Tou rnament, Nanji ng 1 st 8/1 0 6 4 0
Tal Memorial Tou rnament, Moscow 2nd-3rd 51/2/9 2 7 0
World Bl itz Championsh ip, Moscow 1 st 31/42 28 6 8
Chess Classic Tournament, London 1 st 1 3/2 1 * 3 4 0

2010
I nternational Tou rnament, Wijk aan Zee 1 st 8%/1 3 5 7 1
Melody Amber Tou rnament, N ice 1 st-2nd 1 4%/22 13 3 6
King's Tournament, Bazna 1 st 7%/1 0 5 5 0

276
 Tournament and M atch Record ttJ

+ =

Arctic Stars Tou rnament, Kristiansund (rapid) 1 st 5/8 3 4 1

G rand Slam Final, Bil bao 3rd 2%/6 1 3 2
Pearl Spring Tou rnament, Nanji ng 1 st 7/1 0 4 6 0
World Blitz Tournament, Moscow 3rd 23%/38 19 9 10
Chess Classic Tou rnament, London 1 st 1 3/21 * 4 1 2

2011
I nternational Tournament, Wijk aan Zee 3rd-4th 8/1 3 5 6 2
Melody Amber Tou rnament, Monaco 2nd 1 4%/22 11 7 4
King's Tou rnament, Med ias 1 st-2nd 6%/1 0 3 7 0
I nternational Tou rnament, Biel 1 st 1 9/30* 5 4

*These tou rnaments employed the scoring system of 3 poi nts for a win and 1 point
for a draw.

I ND IVIDUAL MATCHES

FIDE World Championship

2004 v. Aron ian, Tri polis (1 st round) 1 %-2% 0 3
2007 v. Aronian, El ista (Candidates 1 /8 final) 5-7 3 4 5

FIDE World Cup

2005 Khanty- Mansiysk
v. Azmai parashvili (1 st round) 3-1 3 0
v. Amonatov (2nd round) 1 %-% 1 0 1
v. Chepari nov (3rd round) 3-1 3 0 1
v. Bareev (4th round) 2%-1 % 1 3 0
v. Lautier (places 9-1 6) 1 %-% 1 0
v. Malakhov (places 9-1 2) 3%-2% 2 3
v. Kamsky (places 9-1 0) 1 -3 1 0 3

2007 Khanty- Mansiysk

v. Zhao Zong Yuan (1 st round) 2-0 2 0 0
v. Naid itsch (2nd round) 3-1 2 2 0
v. Dominguez (3rd round) 1 %-% 1 1 0
v. Adams (1 /8 final) 1 %-% 1 1 0
v. Cheparinov (1/4 final) 1 %-% 1 1 0
v. Kamsky (1 /2 fi nal) %-1 1/2 0 1 1

77
 Tournament and M atch Record

Other Matches + =

2004 v. Kasparov (rapid) %-1 % 0 1 1

2006 v. Svid ler, Longyearbyen (rapid) %-1 % 0 1
v. van Wely, Schagen 5%-2% 4 3
2007 v. Radjabov, Porto Vecchio 2-3 1 2 2
2008 v. Leko, Miskolc (rapid) 5-3 2 6 0

TEAM EVENTS

World Chess Olympiad

2004 Calvia, board 1 3/5 2 2
2006 Tu rin, board 1 6/8 4 4 0
2008 Dresden, board 1 7%/1 1 5 5 1
201 0 Khanty- Mansiysk, board 1 4%/8 4 1 3

European Team Championshi p

2007 Crete, board 1 6%/9 5 3

European Club Cup

2001 Panormo, board 6 4%/7 3 3
2003 Rethym non, board 1 3%/7 3 1 3
2007 Kerner, board 3 4%/7 3 3 1
2008 Kal l ithea, board 1 3/6 2 2 2

278
 I ndex of Carlsen's Opponents

I n dex of Carlsen's O p p o n ents

A L
Adams, M . .. . . 84, 1 43
. . . . . . . . . . . . . . Leko, P . . . . . . . . . . . . . . . . . . . . . . . . . 230
Ad ly, A . . .
. . . . 19
. . . . . . . . . . . . . . . . . . . Lie, K . . . . . . . . . . . . . . . . . . . . . . . . . . .62
Amonatov, F . . . 71
. . . . . . . . . . . . . . . . . . .
M
Anand, V . . . . . . 208, 240
. . . . . . . . . . . . .
Maki U u ro, M . . . . . . . . . . . . . . . . . 14
. . . .

Aronian, L. 1 7, 20, 1 1 2, 1 1 5, 1 1 8 , 1 23 , 1 66, Morozevich, A . . . . . . . . . . . . . 90, 9 5 , 1 01

1 95, 1 98, 256
N
B Naid itsch, A . . . . . . . . . . . . . . . . . . 19. . .

Bacrot, E . . . . . . . . . . . . . . . . . . 265
. . . . .
Nakam ura, H . . . . . . . . . . . . . . . . .269. .

Bel iavsky, A . . . . . . . . . . . . . . 77, 80, 201

.
N ielsen , P. H . . . . . . . . . . . . . . . . . . 50. . .

Bindrich , F . . . . . . . . . . . . . . . . . 15
. . . . . .
N i kolic, P . . . . . . . . . . . . . . . . . . . . . 57. . .

Nisipean u , L.-D . . . . . . . . . . . . . . 258, 271

c Nunn, J . . . . . . . . . . . . . . . . . . . . . . 81. . .

Cheparinov, I. . . . . . . . . . . . . . . . . 55, 1 49 0

D Onischuk, A . . . . . . . . . . . . . . . . . . 1 7, 1 25
Dolmatov, S . . 34
. . . . . . . . . . . . . . . . . . . . P
Doming uez, L. . . . . . . 1 40, 1 87, 205, 2 1 2 Palo, D . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Pavasovic, D . 16
E
. . . . . . . . . . . . . . . . . . . .

Pedersen, C . 18
El.Janov, p. . . . . . . . . . . . . . . . . . . . . . . 1 55
. . . . . . . . . . . . . . . . . . . . .

Pelletier, Y . . . . . . . . . . . . . . . . . . 1 25 , 1 84
Ernst, s .. . . . . 31 . . . . . . . . . . . . . . . . . . .

Ponomariov, R 244, 261. . . . . . . . . . . . . . .

G
R
G risch uk, A. 215 '
R a dJa bov, T. . . . 1 29 , 1 74 , 1 91 , 237

. . . . .

H Rozentalis, E . . . . . . . . . . . . . . .16 . . . . . .

Harestad , H . . . . . . . . . . . . . . . . . . . . . . 23
s
Shaposh nikov, E . . . . . . . . . . . . . . . . . . 35
lvanchuk, V . . . . . . . . . . . . . . . . . 1 07, 1 78 Shi rov, A . . . . . . . . . . . . . . . . . . . . . 51 , 99
Short, N . . . . . . . 46
. . . . . . . . . . . . . . . . . .

J
Jakovenko, D . . . 1 35, 226
. . . . . . . . . . . . .
T
Jobava, B . . . . . . . . . . . . . . . . . . . . . . . 66 Topalov, V . . . . . . 1 07, 1 63, 1 69, 2 1 8 , 233

K v

Kacheishvi l i , G . . . . . . . . . . . . . . . . 14 . . . .
Vladimirov, E . . . . . . . . . . . . . . . . . . . . . 41
Kamsky, G . . . . . . . . . . . . . . . . . 1 8 , 73
. . . W
Karjaki n , S . . . . . . . . . . . . . . . . . . . . . . 252 Wang Yue . . . . . . . . . . . . . . . . . . . . . . 221
Kasparov, G . . . . . . . . . . . . . . . . . . 39 . . . .

Ked ik, D . . . . . . . . . . . . . . . . . . . . . . . . . 15 Z
Kram nik, V . .
. . . . . . . . . . . . 1 33, 1 58 , 247 Zhigalko, S . . . . . . . . . . . . . . . . . . . . . . . 26

279
 I ndex of Open ings

I n dex of Open i n g s

Alekhine Defence [ 804 ] 1 63 Queen's I ndian Defence [ E1 2 ] 66, 1 1 5,

[ E1 5 ] 1 84
Caro - Kann Defence [ 81 9 ] 31
Catalan Open ing [ E05 ] 1 33 Reti Open ing [ A04 ] 34

Engl ish Open ing [ A28 ] 1 69, [ A29 ] 247, Ruy Lopez [ C67 ] 226, [ C84 ] 201 ,
[ A30 ] 1 1 2 , 1 58 , [ A32 ] 1 95 [ C88 ] 1 66, [ C91 ] 77, [ C95 ] 51 ,
French Defence [ COS ] 57, [ C1 1 ] 252 [ C98 ] 23

G ri.infeld Defence [ D81 ] 205, [ D87 ] 1 07, Scotch Game [ C45 ] 230, 265
[ D91 ] 1 55
Semi-Slav Defence [ D43 ] 2 1 8 ,
I rreg ular Open ing [ 807 ] 1 29
[ D45 ] 208, [ D47 ] 1 98
King's I ndian Defence [ E62 ] 62,
[ E66 ] 1 01 , [ E81 ] 261 , [ E90 ] 233, Sicil ian Defence [ 830 ] 95, 237,
[ E97 ] 90, 1 78 [ 833 ] 26, 55, [ 843 ] 73 , [ 867 ] 41 ,
[ 876 ] 258, [ 878 ] 1 74, 1 87, 1 91 ,
N i mzo- l ndian Defence [ E20 ] 84, 1 35 ,
2 1 2 , [ 885 ] 2 1 5 , [ 890 ] 71 , 81 , 244,
[ E32 ] 4 3 , [ E36 ] 1 43, [ E38 ] 35
[ 892 ] 1 40, [ 892 ] 269
Queen's Gambit Accepted [ D27 ] 271
Slav Defence [ D1 5 ] 221 , [ D1 7 ] 1 1 8
Queen's Gam bit [ D37 ] 46, [ D39 ] 240,
[ D43 ] 1 25 Torre Attack [ A48 ] 1 49

280