The King's Indian Defence

Mar del Plata Variation

Svetozar Gligoric

Translated by
Biljana and Zoran !lie

B.T. Batsford Ltd,

London

Firs t published in 2002

Sve to za r Gligo ric, Zo ran Ili c 2002
ISBN 0 7134 8767 4
Bri tish Lib ra ry Ca talogui ng-in-Publica tion Da ta.
A ca talogue reco rd fo r this book is
a vailable from the Bri tish Lib ra ry.
All righ ts rese rved. No pa rt of this book may be
rep roduced, by any means, wi thou t p rio r pe rmission
of the publisher.
Prin ted in Grea t Bri tain by
C rea ti ve Prin t and Design (Wales ), Ebbw Va le
for the publishers,
B. T. Ba tsfo rd Ltd,
64 Brewe ry Road,
London N7 9N T
A

member

of ChrysalifBooks pic

A BA TSFORD CHESS BOOK

Author's Foreword
This book has sp rung by chance from my w ri tten tex t fo r a lectu re, las ti ng
many hou rs, which I ga ve to the mos t talen ted young playe rs of my
coun try. This lec tu re, which took place a t the Chess School of the Yugosla v
Chess Fede ra tion in Belg rade 1 999 , made a deep imp ression upon my chess
colleagues a t home and, wi thout any furthe r help from me, the in te rna tional
mas te r Sini sa Jok si c though t i t was wo rth making a li ttle book ou t of i t and
published this in Se rbo -C roa t in 2000 , while a second edi tion appea red in
German in 200 1 .
Now has come the time to exp ress my g ra ti tude also to in te rna tional
mas te r Mr. Zo ran Ili c, who o n his own ini tia ti ve comple ted the job by
making sugges tions fo r an enla rged edi tion in English, and to his wife, Mrs.
Bil jana Ili c who helped in the transla tion of my o riginal manusc rip t.

Svetozar Gligoric
July, 2002

Contents
List of lllustrative Games

Historical Introduction

Part One: The Development of Ideas

that led to the Mar del Plata Variation

The

Birth of the Mar del Plata Variation

A Short Review of the Ideas in the Period 1 953-2002
Variation 9 tt'le l tt'ld7 1 0 tt'ld3
Variation 9 tt'le 1 tt'ld7 1 0 .ie3

11
22
24
35

Part Two: Playable Lines after 9 tt'lel tt'le8

54

Variation 9 tt'le 1 tt'le8 1 0 f4

Variation 9 tt'le 1 tt'le8 1 0 tt'ld3 f5 1 1
Variation 9 tt'le 1 tt'le8 1 0 tt'ld3 f5 1 1
Variation 9 tt'le 1 tt'le8 1 0 tt'ld3 f5 1 1
Variations after 9 tt'le 1 tt'le8 1 0 .ie3

55
58
62
67
91

f4
.id2 tt'lf6
.id2 fxe4 1 2 tt'lxe4
f5 1 1 f3

Part Three:

Variation
Variation
Variation
Variation
Variation

9 tt'le 1 tt'ld7 1 0 f3 f5 1 1 g4
9 tt'ld2
9 .id2
9 .ig5
9 b4-"Bayonet Attack"

Index of Variations

91
97
1 15
121
1 27
1 60

Illustrative Games
Page
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

Najdorf-Gligoric, Mar del Plata 1 953

Eliskases-Gligoric, Mar del Plata 1 953
Taimanov-Najdorf, Candidates, Zurich 1 953
Gligoric-Fischer, Bled 1 961
L.Schmid-Gligoric, Hamburg 1 965
Larsen-Tal, Candidates, Eersel 1 969
Piket-Kasparov, Tilburg 1 989
Korchnoi-Xie Jun, Marbella 1 999
Korchnoi-Kasparov, Amsterdam 1 991
Atalik-Gufeld, Los A ngeles 1 999
Piket-Kasparov, Linares 1 99 7
Piket-lvanchuk, Wijk aan Zee 1 999
Sosonko-Nijboer, Dutch championship 1 994
Atalik-Nijboer, Wijk aan Zee 1 99 7
Lalic-McShane, Redbus Knockout, Southend 2000
Akopian-Dimitrov, Linares 1 996
Shirov-Uhlmann, Germany 1 994
Gelfand-Judit Polgar, Dos Hermanas 1 996
Korchnoi-Kasparov, Debrecen 1 992
Gelfand-Topalov, Wijk aan Zee 1 996
Shirov-Nunn, Amsterdam 1 995
Pinter-Nunn, Olympiad, Thessaloniki 1 988
Lesiege-Fedorowicz, New York 1 993
Gelfand-Kasparov, Reggio Emilia 1 991/ 92
Gligoric-Littleton, Zonal, Praia da Rocha 1 969
Korchnoi-Gligoric, Havana 1 969
Larsen-Fischer, Candidates, Denver 1 971
Van den Sterren-Gelfand, Munich 1 994
Sakaev-Nevednichy, Olympiad, Elista 1 998
Gligoric-Quinteros, Linares 1 981
Yermolinsky-Topalov, Wijk aan Zee 1 999
Petrosian-Gligoric, Rovinj/Zagreb 1 970
Gligoric-Donner, Berlin 1 971
Gligoric-Simeonidis, Zonal, Panormo 1 998
Bacrot-Motylev, Linares 2001
Gligoric-Nataf, Cannes 1 998
Kramnik-Shirov, Tilburg 1 99 7
Kramnik-Kasparov, Novgorod 1 99 7

12
16
19
24
29
31
35
37
39
41
45
50
55
58
63
61
71
14
78
85
87
91
93
95
91
98
1 04
1 05
1 10
1 15
121
128
135
136
141
1 46
151
1 55

A Historical Introduction
In chess opening theory it is common practice to name variations after the
places where they have been played for the first time, instead of naming
them, more justly, after their creators. Thus, a principal opening variation in
the Semi-Slav Defence, played for the first time at the international
tournament in Meran in 1 924, is not called the Rubinstein Defence (after
the famous grandmaster who invented it), but the Meran Defence.
There have also been errors of a different nature when naming openings.
Alekhine did not invent the so-called Alekhine Defence, although he did
secure it wide recognition by being the first to play it in great competitions.
Najdorf did not find a new variation in the Sicilian Defence (it was found
by Czech masters) but, as a top-level player, he was the first to understand
the importance of this idea while playing at the international tournament in
Prague in 1 946. As a matter of fact, the greatest contribution to the Najdorf
variation was made by Robert Fischer. Likewise, a variation of the King' s
Indian Defence, played by Karpov a s White i n his match against Kasparov
in New York 1 990, in the official American bulletin of the match was given
my name, although I was not the one who invented this early excursion of
the queen's bishop (perhaps it was Reshevsky?); true, I played it
successfully-but only to avoid, after I d4 tiJf6 2 c4 g6 3 lLlc3 g7 4 e4 d6
5 tiJf3 0-0 6 ..te2 e5 7 ..te3 , having to play against my own variation,
because if 7 . . . lLlc6? ! White is ready for a favourable regrouping of the
pieces with 8 d5 lLle7 (now 9 . . . ltJd4 doesn 't work) 9 tiJd2, as in the game
Gligoric-Kraidman, Tel Aviv 1 966.
One of the things I really did invent, however, is the Mar del Plata
Variation-a whole plan of development for Black, which I employed for
the first time at the international tournament in Mar del Plata in 1 953
against Najdorf, and two rounds later also against Eliskases, winning both
games. Shortly afterwards, in Europe, Najdorf copied my method as Black
in a Candidates tournament game against Taimanov, scoring an efficient
and classic victory, after which the variation, as fast and furious as a forest
fire, occupied a significant place in worldwide tournament practice, living a
full life for nearly half a century and well into the present day. In bygone
times, however, communications being far from what they are today, games
from far-away Argentina remained unnoticed for a long time and so it took
many years for me to 'lose my patience ' and announce the name of the
variation's ' anonymous ' creator. In fact, perhaps too modestly, I was
hesitant about claiming my authorship for three decades . . .

Part One :
The Development of Ideas that led
to the Mar del Plata Variation
If we ignore the different strategic concept in the Siimisch variation (5
f3), the Fianchetto line, with its pronounced space control, used to be the
most popular method for White. Then, shortly after the Second World War,
Boleslavsky and Bronstein, and later Geller as well, found dangerous
tactical chances for Black by abandoning the central pawn stronghold on e5
in order to exert pressure along the e-file and the b6-g l and g7-a l
diagonals-and combining this with various associated tactical options.
Aware that maintaining the pawn tension in the centre is a strong weapon,
because in such an unresolved and tight situation Black cannot find suitable
squares for all his pieces, White players turned to the Classical variation
with its quick kingside development: 1 d4 liJf6 2 c4 g6 3 liJc3 .tg7 4 e4 d6
5 liJf3 0-0 6 i.e2 eS 7 0-0. Black ' s first reaction was the natural move
7 liJbd7 but, by manoeuvring his rooks to e l and b l , White retained both
the tension in the centre and his spatial advantage.
. . .

h
8

8
7

2
1
a

Then, around about 1 950, the Soviet grandmaster Aronin promoted the
new move 7 liJc6! in order to increase the pressure on the d4 square and
force White to commit himself to a definite pawn formation in the centre.
This idea was immediately adopted by the top Soviet players, who had
...

10

The Development ofIdeas

a lready contributed enor mous ly to the popu larity of the King 's Indian
Defence. White p layers tried 8 .i.e3 !? in order to hinder B lack 's desire to
b lock the centre (after which he has a c lear p lan for coun terp lay by prepar
ing the pawn break . . .t/-f5), but the Argentinian grand master Na jdorf soon
found the right answer in 8 . . . :e8! when , if 9 d5 , there now fo llows 9 . . .lDd4 !
1 0 lDxd4 exd4 1 1 .i.xd4 lDxe4 with an e qua l ga me in a si mp li fied position.
Nor did Larsen 's atte mpt to p lay for an endga me advantage by 9 dxe5 dxe5
10 'i1Fxd8 pro ve to be e ffective enough to p lay for a wi n.
In a short period after 1 950- 1 952 White p layers fma lly accepted that after

7 0-0 lDc6 8 dS lDe7 (here the knight is for the present poor ly p laced , but it

is on the way to the batt le zone where B lack is gathering as much force as
he can for the impending attack on the kingside) there is a direct con fronta
tion of two initiatives : White 's on the queenside , and B lack 's on the king
s ide. White 's first reaction was to continue 9lDe1 (to deny the b lack knight
on f6 the active h5 and g4 s quares and to prepare a stab le white pawn cha in
leading towards the centre) , after which B lack 's best response was 9...lDd7 !
(making way for the b lack f-pawn and o ffering better contro l than 9 . . .lDe8!?
of the c5 and e5 s quares -where White can atte mpt a breakthrough by
c4-c5 or f2-f4).
Rea li zing that a ll this was the pre lude to a dra matic race between the two
p layers , each p lanning to be first to penetrate their respective flanks , and
that e ach tem po added to the gu lf be tween victo ry and defeat , the lead ing
So viet grandmasters were in a hu rry to launch B lack 's attack as quickly as
possib le a long the 'g 'or 'h '- file by a manoeuvre of the king 's rook from f8
to f6-g6 or h6 , after the se quence 10 lDd3 (or 1 0 .i.e3) 10.. f5 11 f3 f4 12
b4 gS 13 cS l:[f6!? etc.
.

Birth of the Mar del Plata Variation

the early spring of 1 953 (in the Southern hemisphere it was autumn ! ) ,
traditional international tournament i n Mar del Plata, it happened that
111 one of the first rounds I was to play with the black pieces against the
principal favourite, "the chess king of Latin America", grandmaster
Najdorf, and I was considered one of his main rivals, having won in Mar
del Plata in 1 950 when Najdorf did not take part.
In

a t the

I had

become attached to the King' s Indian Defence in 1 952, after which

it constantly in my repertoire for the next 25 years (which I
describe, within my annotations to the game, Geller-Gligoric, in my book I
/'lay against Pieces). I remember that I did not like at all the hurried
manoeuvre of the leading Soviet players . . Jlf8-f6-g6 or h6, which seemed
artificial as it left the black rook and knights obstructing one another in their
search for harmonious redeployment to more active squares.
I included

it was not my habit to rely too much even on authorities greater

myself, nor to spend hours on tedious home study, my preparation in
the hotel room for the forthcoming derby lasted at most ten minutes. I set up
the position after 9 tLle1 tLld7 10 i.e3 (I was most concerned about this ac
tivation of the white queen' s bishop) 10 f5 1 1 f3 f4 12 i.f2 g5 and played
also 13 tLld3 lLlf6 ! 14 c5 tLlg6! (although I had never played this line before,
if my game plan was going to be any good then I had to put all my faith in
this natural formation so as to combat White 's growing initiative on the
queenside) and then carried on 15 cxd6 cxd6 16 tLlbS when the opponent
comes first with his attack on the weak a7 square, where he threatens to take
the pawn and then the main trump of my counterattack, the bishop at c8.
Since

than

1
a

12 Birth of the Mar del Plata Variation

a

2
a

Now I saw that I could try 1 6... g4! without further ado, because the pawn
at e4 was not sufficiently well protected ( 1 7 fxg4? lll xe4 ), while against the
'terrible' 1 7 lll x a7 I noticed the decisive blow 17 ... g3!. This convinced me
that my new idea was not only worth the risk but appeared also to be
logical . Even today, I wonder why no one had come up with the idea
before.
In our game, however, the experienced Najdorf played in slightly
different fashion and I discovered a further plan which enabled Black to
manoeuvre quickly and easily over the board. This broader-based set of
moves was a fresh and powerful weapon for Black in the newly born Mar
del Plata Variation, and here is the matrix game with annotations:
Game 1

Naj dorf - Gligoric

Mar del Plata 1 953

1 d4 t'bf6 2 c4 g6 3 t'bc3 i.. g7 4 e4 d6 5 t'bf3 0-0 6 i.e2 e5 7 0-0

t'bc6
a

2
a

Birth of the Mar del Plata Variation 13

At the time this move (in place of 7 . . . l'Llbd7) was brand new, having just
hcen discovered by Aronin and seen in the game Taimanov-Bronstein,
I JSSR Championship 1 952.
8 d5 !De7 9 !Del ti::Jd 7
I

regarded this move as more precise than 9 . . . l'Lle8, which also opens the
for the f-pawn, because now White will have to spend one more tempo
order to achieve the thematic queenside breakthrough c4-c5 .

way
in

2
a

10 ti::Jd3

As a top-class player with a subtle understanding of the position, Najdorf

gave up the more active 1 0 i.e3 , as played in the game Taimanov
Hronstein, so as not to help promote Black's pawn advance . . . f7-f5-f4 on
White ' s endangered opposite flank. He was convinced that White would
now win the battle on the queenside and I, too, was not sure whether this
would be the case. Later, when he was reflecting on the unexpected
outcome of the game, another participant, IM Luckis (who had come from
the Baltic countries to the Olympiad in 1 939) gave a succinct explanation:
"In chess, the most important thing is the king ! ".

lO f5 1 1 f3 f4 1 2 ..id2 ti::Jf6 ! ?
...

Although at that time the system o f attack with 1 2 . . J:H6 , 1 3 . . .g 5 and

1 4 .. J::t g6 or 14 .. J::t h 6 had caused confusion among White players, I did not
l ike the disharmony of the manoeuvre, and I suppose that this was exactly
what Najdorf was ready for. The text move commences my new plan of
development, but nevertheless it shows my inexperience, otherwise I would
have played first 12 . . . g5.

1 3 b4! ?
Optimistic about gaining space, Najdorf corrects my mistake in the move
order, and wastes a tempo instead of exploiting the opportunity presented to
continue with an immediate 1 3 c5 ! . Now everything proceeds normally as if
the correct sequence 1 2 . . . g5 1 3 b4 l'Llf6 had been played.

1 4 Birth of the Mar del Plata Variation

13 ... g5 14 c5 h5 15 tiJf2 liJg6 !

a

A consistent idea. What should stand here is not the rook but the suddenly
activated, previously passive knight.

1 6 llcl
Expecting the continuation 10 e3, my "preparations" went only as far as
the 1 5'h move. I realized that my critically sensitive point was c7, and, over
the board, without much thought, I found the following move.

1 6 ...l:lt7!
This excellent manoeuvre both defends and attacks.
1 7 cxd6 cxd6 1 8 a4
The game is a classic example of a situation where the opponents
consistently pursue their respective operations on opposite flanks. However,
the more important flank is the one where the kings are, because Black can
even allow himself to lose the battle on the queenside and still deliver
tactical strokes which would enable him to win the decisive battle on the
opposite wing.

1 8 ... iH8!
Black consistently carries out his plan. In good time, he protects his king
from being disturbed along the sh rank via the opposite flank, while
additionally covering the weak pawn on d6 and releasing the g7 square for
the rook.
19 aS l:.g7
At this moment, the Mar del Plata Variation was born !

Birth of the Mar del Plata Variation

15

2
1
a

20 h3 l0h8 !
Preparing the decisive breakthrough . . . g5-g4.

21 l0b5
White tries to weaken his opponent on the queenside, but Black can more
or less ignore it.
2 l . .. g4 22 fxg4 hxg4 23 hxg4 a6 24 l0a3 i.d7 25 l0c4 l:.c8 26
li'lb6 llxcl 27 .i.xc l .i.e8
a

2
a

White 's only chance is to get to c8 with his rook, but this is a long way
and the black king is well-protected from everything that is happening
on the opposite side.
off

28 i.a3 l0f7 29 'ic2 l0h6 30 g5

Since the pawn is doomed anyway, White returns it in a way that makes
Black somewhat more sensitive on the 7'" rank.
30 . l%xg5 3 1 licl
..

16 Birth of the Mar del Plata Variation

3 1 'ir'c8 would not be of much help to White either-because of the weak

e4 and g2 squares.

3 1 . ..l:lg3 32 i.b2 lbfg4 33 lbxg4 lbxg4 34 .ixg4 l:hg4 35 lifl

i.g6 36 l:tc4 'ii e7 37 .tc3 'ii h 7 38 'ife2 l:th4 39 'itt fl f3 !
b

2
b

Removing the remaining shelter of the white king.

40 'ii e3 l:f4 41 gxf3 'if'h2+ 42 'itt e 1 'ifh l+ 43 'itt e2 .th5 44 'itt d 2

l:xf3 ! 45 lig5+ .ig7 46 'itt c 2
46 .l:tc8+ h7 is hopeless.

46 1:1fl+ 47 .id2 'ii'd l + 48 'itt c3 'ii a l+ White resigned.

Mate is unavoidable after 49 'iti>d3 i.e2+ 50 'iti>e3 (or 50 'iti>c2 i.d l + 5 1
'iti>d3 'ii'b l +) 50 . .l:tf3+ 5 1 'iti>xe2 'ii' fl mate.
..

. .

Two rounds later, in an atmosphere of disbelief that Black's attack could

really break through, GM Eliskases also entered the discussion on a similar
topic, but without success. I persuaded my cautious chess friend
Dr.Trifunovic to play the same sharp line against Najdorf in the same
round. He drew that game. In this way the Mar del Plata Variation obtained
its first visa for the world stage.
Game 2

Eliskases - Gligoric
Mar del Plata 1 953

1 d4 lbf6 2 c4 g6 3 lbc3 .tg7 4 e4 d6 5 lbf3 0-0 6 .te2 e5 7 0-0

lbc6 8 d5 lbe7 9 tile1 lbd7 10 .ie3
This attempt at more energetic development, in place of 10 d3 , was,
after subsequent tournament practice, regarded as weaker because it adds
water to the mill of the black pawn phalanx on the kingside. Only in the last

Birth of the Mar del Plata Variation 1 7

\Tars

ttl'W

IItts

of the 20'11 century, thanks to the efforts of Korchnoi and others, were
options found for White. which will be dealt with in a separate part of
book.

IO f5 1 1 f3 f4
...

Black captures space and gains a tempo.

II i.f2 g5

2
a

13 tbd3 tbf6 !

In the earlier game Taimanov-Bronstein, USSR 1 952, Black was not suc
ll'ssli.d with 1 3 . . . .l:r.f6, on which followed 14 c5 .l:r.h6 1 5 cxd6 cxd6 16 ltJb5
t_;\f'X 1 7 .te l a6 1 8 ltJa3 b5 1 9 ltJc2 ltJd7 20 a4 ! etc.

14 c5 tbg6 1 5 .:tel l:.ti!

Black

consistently follows his newly discovered patent for this position.

llt 'ii b3

Although this was not part of his repertoire, Dr Trifunovic too,

l'lllouraged by me, reached this position against Najdorf in the same round.
l lowcver, here Najdorf played 1 6 cxd6 cxd6 1 7 ltJb5 g4 1 8 'ii'd2 and
1hough, due to insufficiently energetic play, Black missed his chance to
ohlain an attack, he did obtain half a point through solid defence.
I (J g4 ! 17 fxg4 tbxg4 18 i.xg4 i.xg4 19 'irxb7 f3

1-'111

the sacrificed pawn, Black has obtained the attack.

20 ..lZ.c3 tbf4! 21 i.xf4 exf4 22 gxf3 i.h3 23 h1

11'23

...

Now

.l:r.t2, then 23 ... .td4 .

i.xfl 24 .l:txfl i.d4

Iu; best

it is difficult for White to defend himself along the g-file so he does

using tactical means.

25 c5 dxe5 26 tbe2 l1b8 27 'ii c 6

18

Birth of the Mar del Plata Variation

If 27 -.xa7 c6.

27 ...'iff6 28 xd4 exd4 29 d6! cxd6 30 'Yi'd5! h8

30 . . . dxc5 3 1 lbe5 .l::r. d 8 32 .l::r. g 1 + etc. is weak.
31 cxd6 .:g7 32 d7 l:r.d8 33 e5
a

2
a

33 ....:dxd7 ! !
This move destroys all White ' s plans.
34 xd7 'it'g6
Although at the moment he is a piece up, White cannot defend himself
against the mating threats along the g-file without material losses.

35 'ii a 8+ l:tg8 36 'ii'x g8+ 'ii' x g8 37 b3 'ii g 5

Cutting off the line of retreat for the white knight.

38 .:et d3 39 4Je5 d2 ! 40 f7+ g7 4 1 .1Id l 'ii'h 4 42 g2 'ifel

White resigned.
In the 50s, chess information travelled slowly, and so few people in
Europe would or could know what had happened in March 1 953 in far
away Argentina. When the Candidates tournament for the world champion
ship began in August 1 95 3 in Switzerland, it was there, in Europe, that the
l ine seen in the game Taimanov-Najdorf first attracted world attention, and
Najdorf received the prize for the most beautiful game.
The true origin of the Mar del Plata Variation remained obscured and,
believing that refraining from self-praise is a sign of fine manners and good
taste, for the next 30 or many more years I was hesitant about claiming my
authorship. I mentioned the fact, although it remained rather unnoticed, for
the first time in the Russian edition of my chess autobiography, in 1 98 3 ,
and said i t again, this time ' loudly' , when I was asked t o give a major

Birth of the Mar del Plata Variation 19

lecture for young p layers -and this was in Be lgrade , not until the year

1999!

When , seve ral months after Mar de l P lata , Na jdorf, as Black agai nst
Taimanov , adopted the s ystem I had p layed against him , he pr oduced his
uwn model example of how to p lay the line in a game which has since
hccome a classic in chess history.

Game3

Taimanov- Najdorf

Candidates tournament, Zurich 1953

I d4 f6 2 c4 g_6 3 c3 .ig7 4 e4 d6 5 f3 0-0 6 .ie2 e5 7 0-0
li:\c6 8 d5 e7 9 lll el d7 10 .ie3 f5 1 1 f3 f4 12 .if2 g5 13 d3

l/)f6 14 c5

Bronstein , in his famous tou rnament book , wrote : "In the 20th USSR
championship 1 952 , Taimanov , as White , won two games with this same
method , i.e. a pawn break a long the c- file , progress on the quee nside and
deep penetration of pieces into enemy ter rito ry along the d- , e- , f- and
g-files -after which the press wer e left with the impression that Black had
lost both games because of the o pening. However , there wer e pla yers who
practised the "re futed variation " with success. For instance , in Mar del Pla ta
the victim of fashion was Na jdorf who lost to Gligori c and d rew with
Trifunovi c. But , at the sta rt of the Zurich tou rnament , these games were not
known to Taimanov . Thus both p layers began this game with great ho pes.
Na jdorf, self-taught by practical ex perience , had studied Yugoslav anal ysis
which went up to at least the 2 1'1 move , whi le Taimanov was encouraged b y
his previous two victories. "
The r eader will know by now that I have to co rrect Bronstein 's descri p
tion a li ttle b y stating that the idea of Black 's system was entirel y mine and
that there was no pre limina ry Yu gos lav ana lysis whatsoever.

20 Birth of the Mar del Plata Variation

14 .. .lbg6 15 .l:tc1l1ti 1 6 .l:c2 i.f8

Najdorf has reached a position which is extremely similar to the one
which he had against me-but this time he is playing from the other side of
the chessboard ! In his book Bronstein comments : "The pawn on d6 is
protected, the rook can be transferred to the g-file, Black's pair of knights
are hanging over the kingside. Nevertheless, Taimanov hopes to maintain
an impregnable position and so proceeds with a daring onslaught on the
queens ide."

1 7 cxd6 cxd6 1 8 'it'd2 g4 1 9 IUcl g3 !

a

The first time that this promising pawn sacrifice, introducing attacking
tactics, was applied here.

20 hxg3 fxg3 2 1 i.xg3 lbh5 22 i.h2

White admits to having problems defending himself along the three
newly opened files on the kingside, otherwise he would have preferred the
more active 22 i.f2.

22 ... .te7! 23 lbb 1

This knight will need time to get back into play, but the active 23 ltlb5
would fail on 23 . . . i.g5 when, if 24 l'hc8 i.xd2 25 %:r.xd8+ %:r.xd8, White is
left the exchange down. Now, 23 . . . i.g5 would mean material losses for
Black because White ' s queen is protected.
23 ... ..td7 24 'i'e1 ..tg5 25 lbd2 i.e3+ 26 <it>h1 'i'g5
Bronstein remarks: "Step by step, Black's pieces arrive at the battlefield
on the kingside. Now Taimanov had to get rid of the bishop on e3 even by
the sacrifice of the exchange, and try 27 ltJc4."

27 .to l:.af8 28 l:td1 b5

Taking away the beautiful c4 square from White ' s knight.
29 a4 a6 30 axb5 axb5 3 1 .:tc7 l:.g7 32 lbb3 lbh4 33 l:.c2 .ih3 !

Birth of the Mar del Plata Variation 21

p icturesque position ! Black's queenside is completely deserted, seven

attack White ' s king. Now, the pawn on g2 is under fire four times,
.111<! il cannot be defended further. After 34 gxh3 there is a forced mate
1 \.1 .gl+ 35 .i.xg l .!:.xg l + 36 h2 lLlxf3 mate), and besides, the threat is
11 Jlxf3 .
!\

1'1<'\TS

. \-t

'We2 ltJxg2 35 Jl.xg2 Jl.xg2+ 36 1Wxg2 'iWh4

Wh ite loses the queen, because if it tries to run away, Black's knight
i'IVl'S a winning check on g3 .
.\7 'Wxg7+ xg7 38 l:tg2+ h8 39 tLlel tiJf4 40 .:tg3 i.. f2 4 1 .l::tg4
Wtd 42 tiJd2 h5
line,

the game was adjourned (the good old days ! ) , and White sealed the
After 43 . . . l:.g8 44

111ovc 43 l:tg5, but had to resign without further play.

Thl'.X I xg8 there would be no defence against mate.

A Short Review of the Ideas

in the Period 1953-2002

After nearly 50 years of the existence of the Mar del Plata Variation,
today over 8,000 games can be found in computer databases, but there are
certainly many more games because even such exhaustive databases, if my
memory serves me well, fail to include some important games from the
past, not to mention possibly unrecorded, less significant encounters.
The basic position from which one can take various paths, arises after
eight moves:

Knowing Black' s intentions, in many years of tournament practice White

players have come up with no less than sixteen different moves in this
position. Some of them can be assessed as sporadic attempts, the further
research of which-due to the l imited space in this book-can be ignored.
Such continuations are 9 lLlh4, 9 l:tbl, 9 l:tel, 9 ii'c2, the preventive 9 <it>hl,
9 h3, then the perhaps premature 9 a4 (so as also to provoke some weak
squares on Black' s queenside as he attempts to blockade) or 9 b3.
Interesting could be 'neutral ' developing moves such as 9 i.d2 or 9 i.gS,
if we leave out Ljubojevic ' s attempt 'a Ia Reshevsky'-9 i.e3
.

A Short Review of the Ideas in the Period 1 953-2000 23

Nevertheless, the most frequent and first played continuation was 9 ll:\e1
which is worthy of particular attention.

.'. c17,

Now White has a choice of three completely

' dop his queenside initiative -10 ll:\d3, or 10
:\ n important later idea
lie e v en more efficient

111 I

In the

l.avonct

different tactical ways to de

i.e3, or 10 f3 and 11 g4.

is 9 ll:ld2, with the intention of moving the knight

c4 square.

last few years, special attention has been given to the so-called
attack ' , in its improved version, with 9 b4,
a

1
a

tries to gain space on the queenside as soon as possible,

he should not waste time on improving the position of his
prove that they are already well-placed where they are.

ll'lllll" White

"mg th a t

,, .. , , ... hut

Variation 9 tt:Jel t2Jd7 10 tiJd3

From the two matrix games given above it can be seen that the break
though of the black pawn to g4 is dangerous for White. I think that it was
Reshevsky who came up with the idea of radically eliminating this danger
after 10 f5 by reducing the number of advancing black pawns and
responding with the simple exchange 1 1 exf5 ! , when if l l . gxf5 then 12 f4 !
and White stops the opponent' s further advance and has a slight advantage
as well as a less sensitive pawn formation.
...

..

Fischer realized that a consistent reaction for Black is l l . lLlxf5 ! ,

activating the passive knight o n e 7 without losing any time, and finding
compensation for White's stronghold on e4 in the fact that he keeps both his
knights on the board !
. .

From Bobby Fischer's classic My 60 Memorable Games we quote his

complete annotations to one of his games entitled . . .
A

lyrical performance

"This draw has the charm of perfection. Each move is interesting and, to
this day, appears flawless. With 17 . . . c5 Fischer launches an intricate
double-Pawn sacrifice which involves exact timing. Gligoric rises to the
occasion, returning material in an attempt to wrest the advantage. The
economy and ingenuity displayed by both players produces a harmonius
flow of movement, remarkable in its esthetic appeal. The effect is of a pas
de deux in which each partner contributes equally to the total symmetry."
Game 4

Gligoric - Fischer
Bled 1 96 1

1 d 4 f6 2 c 4 g 6 3 c3 .ig7 4 e 4 d 6 5 f3 0-0 6 .ie2 e 5 7 0-0

c6 8 d5 e7 9 tll el d7 10 liJd3
The older 1 0 f3 f5 1 1 .te3 f4 1 2 .tf2 g5 has been abandoned. Black' s K
side attack has practically been worked out to a forced mate ! (This was
written in 1969. In recent years a pawn sacrifice on c5 has given new
chances for White in this 'desperate line', but Black has alternative
counterplay, but only here, in the otherwise dubious manoeuvre
... :,{8f 6-h6).

Variation 9 tiJe1 tiJd7 1 0 tiJd3 25

IO ... f5

I I cxf5

l'l't rosian-Tal, in this same tournament, continued (with Black 's tiJ on
q I I t"t exf4 1 2 i.xf4 fxe4 1 3 tiJxe4 tiJf5 14 i.g5 tiJf6 1 5 g4 tiJd4 1 6
: . \dl'2 "fie7 = (I chose this simple pawn exchange in our game for 'safety
I o'ol.\'0/1.1'1, having been impressed by the effectiveneSS of the pawn avalanche
:' .t:5-g4 in my own previous wins as Black-here I wanted to avoid such
ol /oil<' ()/1 the other side of the board!).

ll ...liJxf5
lu this

line White gets a grip on e4, Black on d4. 1 l . . .gxf5 is more

(But, after 12 f4!-a method devised by Reshevsky-White's
1 ,,,ff,m seems to be rather more favourable. The game Reshevsky-GligoriC,
liu,nu.\ Aires 1 960, continued 12 ... tiJg6 13 i.h5 'Wie7 1 4 l:.el e4 1 5 tiJj2 c5
'' ' 'i.t.d .Lc3 1 7 bxc3 "ikg7 1 8 l:.cl tlJf6 1 9 i.e2 i.d7 20 c:;h1 tiJh4 2 1 g3
: '' i .'.' Lj3 exj3 23 "ikxj3 1:.ae8 and a draw was agreed).

'"IJ'.dic.

I!. n 4Jf6 1 3 liJf2 liJd4 1 4 liJfe4

a

26 Variation 9 &Del li:!.d7 1 0 0. d3

14 ... lbh5
White has the c4-c5 lever; Black has the dynamic break with . . . g5-g4.
Chances are roughly even.

(I think that going to h5 with the knight is a phenomenal move and

deserves two exclamation marks-!!. Bobby produced the move quickly,
showing that he was, as usual, well prepared and that he understood this
specific kind of position deeper than anyone else and before it was ever
seen in practice. Its simplicity reveals the logic of a genius: White cannot
put both knights on e4, so he allows the other black knight to stay alive and
to be used to embarrass the opponent on the kingside.)

1 5 J.g5 'it'd7
Keeping an eye on the d6-pawn so that . . . c5 becomes possible.
16 g3
(I could not let the knight come to f4, although I was not too happy at
having to weaken the light squares around my king).
1 6 ... h6
In a later round Gligoric (as Black) played against Tal 1 6 . . . c5?. But after
1 7 0.b5 ! 0.xb5 1 8 cxb5 White obtained a bind.
17 .te3
a

2
a

1 7 ... c5 !
I was informed that Gligoric thought I had blundered a Pawn (correct-!
admit!), but it is a deliberate sac. On 1 7 . 0.xe2+ 1 8 'i!kxe2 g5 1 9 c5 White
has it all his own way.
..

18 .txd4
Not 1 8 0.b5 0.f5 1 9 ..td2 a6, etc.
1 8 ... exd4 1 9 lbb5 a6

Variation 9 tl:Jel ti:Jd7 1 0 ti:Jd3 27

Not 1 9 . . . i.e5? 20 f4 .

20 tDbxd6
Apparently Black has lost a Pawn without any visible compensation. His
Jlll:o:es, which are now so awkwardly placed, soon spring to life, however.
20 d3 ! 21 'ifxd3
A double-edged game would result from 2 1 i.xd3 i.d4+ 22 h l tl:Jxg3+
.'I tl:Jxg3 'i!Vxd6 24 'i!Vc2 i.h3 .
..

2 l ... i.d4+
The combination requires intricate footwork. A mistake would be
' l i.xb2 22 tl:Jxc8 i.xal 23 ti:Jb6 and it' s all over (23 . . . i.d4+ 24 'ii'xd4).
. . .

22 g2
After 22 'it>h l tl:Jxg3+ 23 tl:Jxg3 'i!Vxd6 White is weak on all the squares
and his K-side looks like Swiss cheese. Chances would be even.
22 lDxg3 !
.

This is the resource it was necessary to visualize as far back as move 1 7 .

a

2
1
a

28

Variation 9 !Del ti:Jd7 1 0 ti:Jd3

23 lbxc8 !
Best. Not 24 hxg3 (or 'it>xg3) 'ii' h 3 mate. On 24 !Dxg3 'ii'xd6 again is
good.

23 ... lbxn 24 lbb6 ! 'Yii'c 7!

Blow for blow ! The threat of mate on h2 keeps the exchange.

25 l:.xfl 'ii' x b6 26 b4 !
The saving clause.

26 .. .'iVxb4
I saw the draw coming but felt the position was too precarious to play for
a win. On 26 . . . cxb4 27 c5 ! xc5 28 !Dxc5 'ii'x c5 29 'it'xg6+ 'it>h8 3 0
'ii'x h6+ 'it>g8 3 1 'it> h 1 wins. The only other try is 2 6 . JH7 2 7 bxc5 xc5 28
l:.b 1 followed by d5-d6 with tons of play.
.

27 .l:.bl 1i'a5 28 lbxc5

On 28 l:.xb7 l:.f7.
28 .. 'ti"
. xc5 29 'ifxg6+ i.g7 30 l:Ixb7 'ifd4
The only move. Gligoric was so sure I ' d "find" it that he wrote it down
on his scoresheet while I was taking a minute to look for something better.
3 1 .id3 .l:.f4

3
2

32 'ife6+ 'it>h8 33 1i'g6 Drawn.

On the same topic, let me add one of my games (this time as Black) :

Variation 9 0,e] ti)d7 10 ti)d3 29

Game 5

L.S chmid - Gligoric

European Team Championship, Hamburg 1 965

I d4 lbf6 2 c4 g6 3 lbc3 .tg7 4 e4 d6 5 lbf3 0-0 6 .te2 e5 7 0-0

8 d5lbe7 9 ttJe1 lbd7 10 lbd3 f5 1 1 exf5 lDxf5

: W(,

7
6

3
2
1
a

12lbe4 ! ?
White i s in too much o f a hurry to occupy his stronghold. A more natural
\\,IY would be to bring another knight to this square after 1 2 f3 and 1 3 0,2 .
12 . .. lbf6 13 f3 lbd4 1 4 i.e3 c6!
Since there is no white knight on c3, Black emphasizes the sensitivity of
1lw d5 square.
15 l2Jxf6+ 'ifxf6 1 6 lDf2 cxd5 1 7 cxd5 Vllif7 1 8 i.c4 b5
lllack has obtained satisfactory counterplay.
19 .txd4 exd4 20 .txb5 'ii'x d5 21 'ii' a4 'ii'h 5 22 'ifb3+ h8 23

.:._,.(, i.e5

"'

I , ,,

'"' ,, I)

lik I li'Jd7

I()

li:Jd3

24 f4?
White is too greedy to win material and rushes into a hidden trap. He
should have continued 24 g3 .1La6, with chances for both sides.

24 ... .txf4 25 g3 .te3 26 .1Lxa8 .1Lh3

Black has the superior position, although he is a rook down.

27 'ifd5
White had calculated that this manoeuvre would destroy Black' s attack.
Not 27 .1Lg2 .1Lxg2 28 <li>xg2 'iff3+ 29 Wh3 .l:tf5 .

27 ...'ii'e 2 ! 28 'it'xd6
h
8

1
a

28 ...g7 ! !
The German grandmaster could not get over the shock when, after this
quiet move, he realized that he had no adequate response and that, even a
rook up, he was lost. 29 i.g2 .l:txf2 etc. is also no use.
29 'ii'x f8+ xf8 30 i.g2 i.xg2 31 xg2 'i'xb2 32 a4 'i'c2 33 h3
g7 34 l2Jg4 'i'c8 35 h4 g5+ White resigned.
After 9 li:Je 1 li:Jd7 1 0 li:Jd3 f5 White players later found a new finesse to
save an important tempo for the breakthrough c4-c5. This was by the inter
mediate development 1 1 i.d2 (instead of 1 1 f3) when, after l l . . .li:Jf6 1 2 f3
f4 1 3 c5, White will open the c-file a move earlier without spending a
tempo on the superfluous b2-b4 (as in Najdorf-Gligoric).

Variation 9llJeJllJd7 1 0 llJd3 3 1

Game 6

Larsen -Tal
Game Five, Candidates match,
Eerse/ 1 969

I tl'lf3 ctJf6 2 c4 g6 3 tl'lc3 i.g7 4 e4 d6 5 d4 0-0 6 i.e2 e5 7 0-0

:.;\c6 8 d5 tl'le7 9 tl'le1 tl'ld7 10 ctJd3 f5 1 1 i.d2 tl'lf6 12 f3 f4 13 c5
J!5 14 J:c1 tl'lg6
a

2
a

1 5 tl'lb5
1 5 cxd6 cxd6 1 6 llJb5 is simplest because it avoids the unclear complica
lions of 1 5 . . . a6 1 6 cxd6 ! ? .
1 5 :n
...

8
15 a6 (see diagram) 1 6 cxd6 ! ? is sharp 8
(passive is 1 6 llJa3 g4 1 7 cxd6 cxd6 1 8 llJc4 7
7
g3 1 9 h3 .i.xh3 20 gxh3 b5 2 1 l:te 1 bxc4 22
6
6
.!:hc4 llJh4 23 i.. fl h5 24 i.. g2 llJh7 25 l:tfl 5
li1g5 with the attack, Hoeksma-Riemersma, 4
4
llolland 1 987) 1 6 ... axb5 1 7 dxc7 'ii'd 7? !
3
3
( II uzman recommends as better 1 7 . . 'iVe8 ! ? 1 8
2
2
'iib 3 g4 1 9 i.. b4 l:tt7 20 l:tfd l gxf3 2 1 i.. x f3
li.Jh4 with counter-chances) 1 8 'ifb3 llJ e8 (if 1
a, b
c
d c f 8 h
1 8 . . . g4 1 9 i..b4 l:tf7 20 d6 ! g3 ? ! 2 1 l:tfd l
gxh2+ 22 <ifi>xh2 b6 23 i.. a 3 , the outcome favours White, Sariyadzanov-Klimov, St. Petersburg 1 997) 1 9 llJc5! 'ii'd 6 20
li.Je6! 'ii'b 6+ 21 <ifi>h1 .i.xe6 22 dxe6 llJxc7 23 e7+ l:tti 2 4 i..c4 llJh8 2 5
J:Ifd 1 lieS and in the game B iryukov-Solovyov, St.Petersburg 1 999, White
accepted a draw without noticing the road to victory after 26 i.. x f7+ llJxt7
27 'ii'c 2llJa6 28 'ii' c 8llJd6 29 'ife6+ <ifi>h8 30 .te l etc.
..

1 6 cxd6 cxd6

32 Variation 9 ti::J e l ti::J dl 1 0 ti::Jd3

3
2
a

1 7 'Wc2!
The threat ti::Jb 5-c7-e6 would annihilate Black's chances of an attack, so
now he has to lose two tempi to prevent it. Less precise is I 7 ti::J 2 i.f8 I 8
'ifc2 1i.d7 ! ? I 9 'ifc7 ti::J e 8 2 0 'ii'x d8 l:txd8 2 I ti::Jx a7 l:ta8 22 ti::Jb 5 l:txa2 23
1i.c3 l:ta8 24 l:ta I l:tb8 25 i.b4 i.e7 and B lack managed to hold the draw in
a passive endgame, Korchnoi-Gligoric, Leningrad I 973 .

1 7 ...g4
Later on, Black players turned to the more cautious continuation 17 ti::J e8
(see diagram) 18 a4 h5 1 9 ti::J fl 1i.f8 20 h3 (if 20 ti::Jx a7 l:tc7 2 I 1i.a5 l:txc2
22 1i.xd8 l:txe2 23 ti::Jx c8 l:txa4 24 ti::Jd 3 g4 25
.l:t2
l:te3 26 ti::Jei l:ta8 27 l:tfc2 l:tb3 28 'it2
8
8 l:ta2 29 l:tb I 'itf7 30 'ite2 1i.e7 3 I ti::J x e7 li::J x e7
7
7
32 ti::Jd 3 l:ta8 33 1i.xe7 rl;xe7 with equal play,
6
6
Novikov-Glek, Lvov I 98 5 ; or 20 'ifb3 l:tg7 2 I
5
5
h 3 ti::Jh4 22 l:tc2 a6 2 3 ti::Ja 3 1i.d7 24 l:tfc 1 l:tb8
4
4 25 li::J c 4 b6 26 a5 g4 ! 27 fxg4 ti::J f6 28 ti::J x b6
3
3 hxg4 29 hxg4 ti::J x g2 ! and instead of 30 'ii'h 3 ? !
2
2 ti::J e 3 3 1 i.xe3 l:th7 ! which favours Black,
lvanchuk-Timman, match, Hilversum 1 99 1 ;
-....a 1
Timman recommends as better 3 0 'itxg2 li::J xg4
3 1 i.xg4 i.xg4 32 'it>fl i.d7 ! with chances for
both
sides) 20 ... l:t g7 (see diagram) 2 1 a5 (if 2 I
8
8
ti::J x a7 l:tc7 [less clear is 2 1 . . .1i.d7 22 ti::Jb 5 ti::J h4
7
7
23 'ifb3 'ith8 24 a5 g4] 22 1i.a5 l:txc2 23 .txd8
6
6
l:txe2 24 ti::J x c8 l:txa4 25 ti::Jd 3 g4 26 .l:t2 l:te3
5
27 ti::J e 1 g3 28 l:tfc2 l:tb3 29 'itfl l:ta2 30 l:tb 1
4
4
'itf7 3 1 <iti>e2 l:ta8 32 ti::Jd 3 1i.e7 33 1i.xe7 ti::Jx e7
3
3 34 ti::J x e7 'itxe7 with equal play, Averkin2 Kasparov, USSR 1 979) 2I. .. Ji.d7 (or 2 I . . .ti::J h 8
22 Vb3 ti::J f7 23 l:tc2 ti::J h6 24 l:tfci a6 I,h- Ih,
1
iS!b...a
Drasko-Gligoric, Sarajevo 1 983) 22 'ii'b 3 ti::J h 4
..

Variation 9 li:Jeili:Jd7 1 0 li:J d3 33

!\ .!tel i.. e 7 with chances for both sides, Polugaevsky-Tal, match, Alma
1 980.

:\Ia

18 tiJc7 gxf3 1 9 gxf3

I9

J.xf3 is worth considering.

19 .th3 20 tDxa8
.

Or

20 li:Je6 'i'b6+ 2 1 lt2 J.xe6 22 dxe6 lte7 23 ,.a4 li:Jf8 24 i.. a 5 'We3
i.. fl li:Jxe6 26 J.d2 'it'b6 27 i.. a 5 ,.e3 28 i.. d 2 Ih-Ih, Podgaets
M.Gurevich, Sverdlovsk 1 984.
.

' 'l

20 ... tiJxe4 2 1 fxe4 'ifg5+ 22 f2

a

2
a

22 .. .'iVg2+? !
The correct continuation i s 2 2 . . ...h4+ with a draw, Averkin-Tal, USSR
1 969.

23 <it>e1 tiJh4 24 .te3 ?

The right defence was 24 li:J2 ! li:Jf3+ 25 d l li:Jd4 26 'ii'c 3 and White
should win.
24 ...'iVxe4 25 .tf2 f3 26 .txh4 'ifxh4+ 27 tiJf2 fxe2 28 'ifxe2 e4
29 l:[g1
Better was 29 ltc7 l:r.f5 30 ltc8+.

29 ... e3 30 'ifxe3 l1e7 31 I:txg7+ 'iitx g7 32 l:c7 .td7 33 l:txd7 llxd7

34 "iix a7?
A new error. Larsen recommends 34 'iic 3+ 'iif6 3 5 'ii'g 3+ f8 36 li:Jd3 .

34 ... 1:te7+ 35 d 1 'ifc4 ! 36 'ii b 6 'iffl + 37 d2 l:te2+ 38 c3

'ifc l+ 39 d4 'ii"e 3+ 40 c4 l:ic2+ White resigned.

34 Variation 9 ti:)e] ti:)d7 10 ti:)d3

However, after

1 d4 ltlf6 2 c4 g6 3 ltlc3 i.g7 4 e4 d6 5 ltlf3 0-0 6 i.e2 e5 7 0-0

ltlc6 8 d5ltle7 9ltle 1ltld7 10 ltld3 f5 1 1 i.d2
Black can exploit the opportunity to interrupt the white pawn chain:

l l .... fxe4 1 2ltlxe4ltlf6

Another continuation is 1 2 . . . ti:)f5 1 3 .tc3 ti:)f6 1 4 .tf3 ti:)h4 1 5 ti:)xf6+
'ii'x f6 1 6 .te4 .tf5 1 7 'ii' e2 M.Gurevich-Kupreichik, Minsk 1 987.
a

13 i.f3
If 1 3 f3 t;:)f5 1 4 ti:)d2 ti:)xe4 1 5 lt:Jxe4 c5 1 6 .td3 ti:)d4 1 7 l:tb l a5 IJS-th,
Glig_oric-Szabo, Dallas 1 957, while also playable is 14 . . . . c6 1 5 ti:)xf6+ .txf6
1 6 tDe4 cxd5 1 7 cxd5 i.. g 7 1 8 g4 ti:)d4 1 9 'lti>g2 'ii'b 6 20 b4 i.. d 7 2 1 a4
ti:)xe2 22 'ii'x e2 l:tac8 with chances for both sides, Robatsch-Giigoric, Lone
Pine 1 97 5 .
O r 1 3 ti:)x f6+ .txf6 1 4 .tc3 t;:) f5 1 5 i.. f3 i.. g7 1 6 i.. e4 ! 'ii'h4 1 7 l:t e 1 b 6
1 8 g 3 'ii' g 5 1 9 Vel 'ii'x c l 2 0 l:taxc l a5 IJS-IJS, Polugaevsky-Tal, match,
Alma Ata 1 980.

13 ...ltlf5 14ltlxf6+ Wxf6 1 5 i.e4 i.d7 1 6 l:Ie11:tae8 1 7 ltcl 'i'ih4

1 8 i.c3 ltld4 1 9 g3 Wg5 20 i.xd4 exd4 2 1 f4 'ii f6 22 l:c2 l:Ie7 23
'ii'o 1J2-I.h,
Sosonko-Giigoric, Ljublj ana!Portoroz 1 977.

Variation 9 tt:Jel ctJd7 1 0 .lte3

This variation, in which the white queen 's bishop takes the most active
diagonal to the black queenside, has for a long time been underestimated
hccause of the quick advance of the black pawn phalanx on the kingside.
l'hcn, in the late eighties, Korchnoi-and later Atalik and other
randmasters-tried to prove that White' s initiative on the queenside could
parry Black's attack on the white king. To illustrate various ideas, here are
:;cveral games.

Game 7

Piket - Kasparov
Tilburg 1 989
I d4 lDf6 2 lbf3 g6 3 c4 i.. g 7 4 lbc3 0-0 5 e4 d6 6 i.. e 2 eS 7 0-0
ll'\c6 8 dS lbe7 9 lDe1 lDd7 10 i.e3 fS 1 1 f3 f4 12 i.f2 gS
a

1
a

13 b4 ! ? liJf6 14 cS l2Jg6 1 5 cxd6

Korchnoi 's improvement 1 5 a4 ! is analyzed in the next game.

lS cxd6 16 lic l .Uf7 17 a4 i.. f8! 18 aS i.d7! 19 liJbS?! g4! 20

l/lc7? !
...

I n the case o f 2 0 lDxa7 g 3 2 1 .tb6 'fle7 2 2 lDb5 lDh5 2 3 hl gxh2 24

f2 i.xb5 25 .i.xb5 lDg3+ 26 .i.xg3 fxg3 Black would have had the
superior position. The best is 20 fxg4 lDxe4 2 1 lDc7 with chances for both
s ides.

3 6 Variation 9 tD e l CDd7 1 0 1l.e3

20 ... g3 !
8

2
a

2 1 li)xa8?
If 2 1 hxg3 fxg3 ! 22 i.xg3 i.h6 ! 23 CDxa8 CDh5 24 i.2 CDgf4 25 CDd3 !
l:tg7 26 CDxf4 i.xf4 27 g4 ! i.xc l 28 'ifxc l tDf4 29 'i1Ve3 h5 ! with the
initiative.

2 1 . .. li)h5 ! 22 <i1th l
The only move. If 22 i.xa7 'i1Vh4 23 h3 i.xh3 24 gxh3 'ifxh3 25 l:t2
gx2+ 26 'ifo>x2 CDh4 27 i.fl 'ii' h 2+ 28 CDg2 l:tg7 and Black wins.

22 ... gxfl 23 l:.xfl li)g3+! 24 gl 'ifxa8 25 i.c4 a6! 26 'ti'd3 ? !

I f 2 6 hxg3 fxg3 2 7 l:tb2 'ii' d 8 28 'ifo>fl i.h6 2 9 'ifo>e2 'ir'g5 wins, o r 2 6 CDd3
'iVa? 27 CDc5 i.b5 ! 28 i.xb5 axb5 29 hxg3 fxg3 30 l:tfc2 dxc5 3 1 bxc5 tDf4
with a winning position.

26 'i!Va7 27 b5 axb5 28 i.xb5

..

h
8

8
7
6

28 tDh l ! White resigned.

...

Variation 9 tiJeJ tiJd7 10 i.. e3 3 7

Game 8

Korchnoi - Xie Jun

Veterans-Ladies match, Marbella 1 999

I d4 f6 2 c4 g6 3 c3 i.g7 4 e4 d6 5 .ie2 0-0 6 f3 e5 7 0-0

!,; \c6 8 d5 e7 9 el d7 10 i.e3 f5 l l f3 f4 1 2 .ifl g5 13 b4 f6

1 4 c5 g6 1 5 a4 !

This is more ambitious than 1 5 cxd6 cxd6 as was played in the above
g i ven game, Piket-Kasparov. It is still possible for White to exert maximum
pressure on the queenside with c5-c6.

15 ... h5 16 c6 ! ?

2
a

White blocks the queenside where he is active, but the blockade is only
partial because on 16 . . . b6 he has the strong response 17 a5 .

38 Variation 9 /?Jei I?Jd7 10 i. e3

1 6 ... i.. h 6 1 7 bS b6 ! ? 1 8 aS g4 1 9 axb6 cxb6 20 fxg4

For the great strategic success he has achieved on the queenside, White
makes a positional concession on the other flank where he is inferior. It
would be too dangerous to allow . . . g4-g3 .

20 ... hxg4 2 1 g3 ! :n
After 2 l . . . f3 22 i.d3 , with the kingside blocked, it would be easier for
White to defend himself.

22 i.d3 h7 23 h1 gS 24 l:.a2
White makes useful preventive moves for defending the king.

24 ... 'iic 7 2S i.. e2 h3 26 gxf4

Not 26 .i.xg4 fxg3 27 hxg3 i.e3 28 'ii' h 5 'it?g7 ! with too many threats.

26 ...l1xf4 27 i.. g3 .:txfl + 28 i.. xfl gf4 29 d3 'iNti 30 f2 'ii g 6

31 i.. e2 h8 32 'ii fl 'ii'g S 33 i.d1 hS 34 xh3 xg3+ 3S hxg3
gxh3 36 'iif3 !
I n time to take the important d l -h5 diagonal.

36 ... a6 37 h2 g8 38 i.. e 2 aS 39 i.. d 1 l:ta7 40 .l:if2 l:tg7 4 1 llJa4

White ' s strategy-safety first and then an easy victory on the queenside
-is the way to triumph.
a

3
2

2
1

4 1 ...'ii'g6
Or 4 l . . .'ii'e 3 42 'ii'x e3 .i.xe3 43 .l:.f6 .l:.f7 44 .l:.xf7 'it?xf7 45 .i.e2 i.d4 46
i.fl .i.g4 47 i.xh3 i.d l 48 c7 winning.

42 :n
Covering the gl square, and parrying the threat 42 . . . .i.f4 .

Variation 9 tiJeJ tiJd7 10 i.. e3 3 9

42 ... i.. f4 43 gxf4 exf4 44 i.. e 2

Defending the rook on fl , so that if 44 .. .'ii' g 3+ 45 'ii'x g3 fxg3+ 46 <it>g l
47 g2 i.. h 3+ 48 <it>xh3 g2 he could respond with 49 <t>xh2.

h2+

44 ... 1i'g2+ 45 li'xg2 lbg2+ 46 hl .l:txe2 47 lbxb6 i.. g4 48 l:gl

11 49 l:txg4+ h7 50 c7 f2 51 l:.f4 Black resigned.

Game 9

Korchnoi - Kasparov
A ms terdam 1 991

l lDf3 lDf6 2 c4 g6 3 lDc3 i.. g 7 4 e4 d6 5 d4 0-0 6 i.. e 2 e5 7 0-0

lilc6 8 d5 I?Je7 9 I?Jel lbd7 10 i.. e 3 f5 1 1 f3 f4 12 i.. fl g5 13 a4
a

By increasing the pressure on the f2-a7 diagonal Korchnoi defeated Nunn

Amsterdam in 1 990. However, in the game Korchnoi-Lanka, Linz 1 997,
White chose 13 tiJd3 tiJf6 1 4 c5 t!Jg6 15 a4 <it>h8 16 a5 llg8 17 cxd6 cxd6
1 8 tiJb5 g4 1 9 fxg4 t!Jxe4 20 i.. x a7 i.. d 7 2 1 i.b6 'ii' e 7 22 t!Jc7 llaf8 23 lla3
lLlg5 24 b4 e4 25 t!Je 1 , but Black missed the chance of the better 25 . . . t!Je5 !
threatening to launch an attack by f4- f3 .
in

13 lDg6
..

Later, at the Olympiad in Erevan in 1 996, in the game Yusupov

Kasparov, Black opted for the safer l 3 . . . a5 14 tiJd3 b6 1 5 b4 axb4 16 t!Jxb4
tiJf6 etc.

14 lbd3

40 Variation 9 l'iJeJ l'iJd7 10 i.. e3

More energetic is 14 aS! (see diagram)

14 ... .l:lti (weaker is 1 4 . . .<ifr>h8 1 5 l'iJd3 l:g8 ? !
7
7
1 6 c5 i.. f8 1 7 c6 ! bxc6 1 8 l'iJb4 ! l'iJe7 1 9
6
l'iJxc6 l'iJxc6 2 0 dxc6 l'iJf6 2 1 l'iJd5 !
6
5
Korchnoi-Shirov, Horgen 1 994) 1 5 cS! ( a
novelty from the g_ame Kozui-P.Popovic,
4
4
Novi Sad 1 992) 1 5 ... tll x c5 16 i.. x cS dxcS 17
3
3
i.. c 4 h8 18 a6 bxa6 (or 1 8 . . . .l:lf6 1 9 axb7
2
2
i.xb7 20 l'iJd3 i.. f8 D.Gurevich-Sherzer,
USA ch 1 992) 19 l'iJ d3 i.. f8 20 l'iJa4 .l:lb8
a
b c
d e f g h
with a positional initiative for White and
now, instead of 2 1 b3, more precise is 2 1
.l:lfl ! i.. d 6 22 l'iJaxcS 'fie7 2 3 b 4 ! with the advantage, Korchnoi-P.Popovic,
Bmo 1 992.
8

14 ... ltJf6 15 c5 h5 16 h3 l::t f7 17 c6

Better is 17 cxd6 cxd6 18 .l:lc 1 .
a

17 ... a5! ! 1 8 cxb7

If 1 8 b4 b6 ! 19 bxa5 bxa5 ! and after . . . i.. h6 and . . . .l:lg7 Black would have
an ideal set up of his pieces to organize a successful attack.

18 ... .i.xb7 19 b4 .tc8 20 bxa5 .th6! 2l ltJb4?

It would be somewhat better to slow down Black 's attack by 21 a6 !
i.. x a6 ! 22 l'iJb4 i.. c 8 with chances for both sides.

2 l ... g4 22 ltJc6 ? ! 'iff8 23 fxg4 hxg4 24 hxg4

If 24 i.xg4 l'iJxg4 25 hxg4, then 25 . . . f3 ! is decisive.

24 ... .i.g5 25 .i.f3 'ifh6 26 J:e t ltJh4 ! 27 .i.xh4 .i.xh4

Variation 9 &i)e] &i)d7 1 0 i.. e3 41

28 g5? !

I t would b e o f n o use to play 28 fl &i)xg4 29 i.xg4 i.xe 1 30 i.. x c8

h i + 3 1 e2 ,.xg2+ 32 xe 1 f3 , nor 29 e2 &i)e3 30 Wd2 i.. x e 1 3 1
J:he I &i)c4 and Black wins.

28 ...'ii'x g5 29 lte2 ltJg4 30 .Ub 1 .1g3 31 'ii'd 3 'ii'h 4 White

csigned.
Game 1 0

Atalik

Gufeld

Los A ngeles 1 999

I d4 liJf6 2 c4 g6 3 ltJc3 .1g7 4 e4 0-0 5 liJf3 d6 6 i.e2 e5 7 0-0

0\c6 8 d5 ltJe7 9 ltJe1 liJd7 10 .1e3 f5 1 1 f3

3
2

1 1 ...liJf6

In Havana in 1 997, the same opponents (who also played in Beij ing in
1 996, when Black won) played this line and after 1 l . . .f4 12 i.. 2 g5 13 Ac l

42 Variation 9 ltJ eJ tiJd7 1 0 i.e3

ltJg6 14 c5 ltJxc5 1 5 b4 ltJa6 1 6 ltJb5 i.d7 1 7 ltJxa7 -.., 8 1 8 tiJb5 l:tf7 1 9

a4 h 5 2 0 tiJd3 i.h6 2 1 l:c4 l:g7 2 2 h 1 'i'd8 Black obtained counterplay.
For more details about this line see the next game, Piket-Kasparov, Linares
1 997.

12 c5 f4 13 ..1f2 g5 14 a4 li:)g6 1 5 a5 h5 1 6 cxd6 cxd6 1 7 li:)b5 g4

2
b

1 8 ..1xa7
Up to this moment, White played all his moves very quickly. If 1 8 ltJxa7,
a
b
c
(see diagram) then Black, in an unclear
d e r g 11
8
position,
would obtain an initiative for the
8
pawn.
7
7
6

6
5

Nevertheless 18 ... i.d7 1 9 i.b6 'i'e7 20

i.b5 (or 20 l:c 1 g3 2 1 i.b5 gxh2+ 22 h 1
ltJe8 23 i.xd7 'i'xd7 24 tiJd3 i.f6 25 xh2
4
4
l:f7
26 l:h 1 l:g7 27 g1 ltJh4 28 l:.h2 l:.g3
3
3
29 h 1 <ili>h8 30 b4 ltJg7 3 1 l:c7 'i'e8 3 2
l:xb7 'ifg6 3 3 ltJe 1 l:g8 34 'ife2 ltJ e 8 3 5 ltJc6
ltJxg2
36 ltJxg2 h4 37 'iffl h3 38 ltJxf4 'ifh6
a
b
c
d e f g h
39 ltJxh3 l:g2 40 l:xg2 Black resigned,
Grabliauskas-Sutovsky, Pula 1 997) 20 ... g3 (if
20 . . . h4 2 1 i.xd7 'ii'x d7 22 l:c 1 l:f7 23 l:c7 'ii'e 8 24 l:xf7 with a clear ad
vantage for White, Kozul-Belotti, Reggio Emilia 1 993) 21 i.xd7 ltJxd7 22
h3 ltJxb6 23 axb6 'ii'd 8 24 'ii'b3 and White's game is superior,
Nikitin-Frolov, St. Petersburg 1 997.
Or 18 ... g3 1 9 i.b6 gxh2+ 20 xh2 'ii'e 7 21 ltJxc8 l:axc8 22 l:h l ! tiJd7
23 i.f2 i.f6 ! 24 g 1 i.h4 25 i.xh4 ltJxh4 26 i.b5 with advantage to
White, Kiryakov-Ulko, Moscow 1 999.

1 8 ... g3 1 9 ..1b6 ile7 20 li:)c7

Variation 9 li::J e J li::Jd 7 1 0 i.. e3 43

2
a

20 tbd7 !

...

lo

A bolt (a prepared one though ! ) from the blue. Black gains a vital tempo
carry out a devastating attack.

21 h3
The only sensible answer. If 2 1 lt:Jxa8 lt:Jxb6 22 lt:Jxb6 'ifh4 23 h3 i.xh3
24 gxh3 'ii' x h3 25 .l:tf2 gxf2+ 26 'it>xf2 White exposes himself to an unpleas
a nt attack, where Black has the choice between repeating moves and search
ing for a mating attack.

2l tbxb6 !
..

Stronger than 2 l . . .'ifh4 22 lt:Je6 lt:Jxb6 23 axb6 i.xe6 24 dxe6 .l:txa 1 25

._.xa 1 .l:tc8? (25 . . .'ifd8) 26 'ifa7 'ife7 27 i.a6 and Black lost the b-pawn and
l h e game, Koutsin-Zpevak, Slovenia 1 999.

22 axb6 l::r.x al 23 'ifxal

a

23 tbh8

..

Black chooses the most efficient itinerary . . . lt:Jh8-f7-g5 .

44 Variation 9 tiJeJ tiJd7 1 0 i.. e3

24 i.c4
If 24 tiJe6, then Black proceeds with 24 . . . i.. x e6 25 dxe6 'i'xe6 with equal
chances, while 24 . . . tiJf7 ! ? is unclear.

24 ... l2Jt7 25 lbe6 lbg5 ! 26 'ifa2

After 26 tiJxf8 i.. x f8 27 tiJd3 h4 28 i.. b 5 i.. xh3 29 gxh3 tiJxh3+ 30 g2
tiJg5 Black has an attack for the sacrificed material. White didn 't like this
position and so he returns the pawn and transposes to the endgame.

26 ... lbxe6 27 dxe6 l:te8 28 lbd3 i.xe6 29 i.xe6+

The sharper 29 tiJb4 was also possible.

29 ...'ifxe6 30 'ii x e6+ l:he6 3 l lbb4 l:le8 32 I:tcl

If 32 l:ta l then 32 . . . l:tc8 ! .

32 . . .lia8 33 fl i.f6 ! 34 lt:ld5 i.d8

35 e2

If 3 5 l:tc7 l:ta l + ! 36 e2 l:tg_l 3 7 l:txb7 ! .ixb6 ! 3 8 Axb6 Axg2+ 3 9 d3

Af2 40 l:tb7 g2 41 Ad7 g l ='il 42 tiJf6+ f8 43 tiJh7+ with a draw by
repetition of moves.

35 ... t7 36 d3 l:ta6 37 c8 i.xb6 38 l:tb8 i.gl 39 lhb7+ g6

40 l:d7 g5 41 h4+ g6
Not 4 l . . .xh4 42 Ag7 and Black is helpless against the mate tiJd5-e7-f5 .

42 c2 i.d4 43 b l i.c5 4 4 .l:.b7

If 44 b4? i..xb4 45 tiJxb4 l:tb6.

44 ... .Ua4 45 l:.b3 t7 46 .Uc3 <t>e8 47 l:d3 l:ta7 48 c2 d7 49

b4 ! I:r.a2+! 50 b3 :xg2 5 1 bxc5 l'if2 52 lbc3 c6 ! 53 :xd6+
If 53 cxd6 d7 ! 54 tiJa4 (54 Ad l Axf3 55 c4 g2 56 Ag l Axc3 57 xc3
f3 and Black wins) 54 . . . g2 55 CiJc5+ c6 56 d7 g l ='it' 57 d8='i' 'i'b l + with
a mating attack.

Variation 9 ti:Jel ti:Jd7 1 0 i.. e 3 45

53 xc5 54 l:ld5+ c6 55 l:.xe5 l:.xf3 56 ltg5 l1f2

...

5
4

1
a

57 lbd5 g2 58 lbb4+ d6 59 e5+ d7 60 l:[g7+ e8 61 lbd3 l:tf3

(,2 l:I.xg2 .l:Ixd3+ 63 c4 .l:.g3 64 Iia2 f3 65 d5 l:g2 66 a3 f2 67
t!f3 e7 68 l:tf4 l:tg4 69 11xf2 .:.xh4 70 }!ta2 lib4 71 lia7+ e8 72
.i?c6 h4 73 lih7 l:td4 74 c5 ltd 1 75 11xh4 e7 76 .Uh6 !td2 77
J1d6 l:r.xd6 78 exd6+ d7 79 d5
and on the 89th move a draw was agreed.
Game 1 1

Piket - Kasparov
Linares 1 99 7

1 d 4 lbf6 2 c4 g 6 3 lbc3 i.g7 4 e 4 d 6 5 lb f3 0-0 6 i.e2 e 5 7 0-0

lbc6 8 d5 lbe7 9 lbe1 lbd7 10 i.e3 f5 1 1 f3 f4 12 i.f2 g5
a

f
8

8
7
6

3
2
1

1
a

46 Variation 9 l:i:Jel l:i:Jd7 1 0 i.e3

13 l:tcl
This simple developing move (instead of 1 3 a4) is the introduction to a
modem plan with a pawn sacrifice and a gain of time to exert pressure
along the c-file, which ties the black forces to the defence of the c7 square,
instead of allowing them to attack. In case of 13 a4 l:i:Jg6 14 a5 'it>h8 1 5 l:i:Jb5
l:i:Jf6 16 l:i:Jxa7 i.d7 17 c5 g4 ! 1 8 c6 g3 1 9 hxg3 bxc6 20 l:i:Jxc6 i.xc6 2 1
dxc6 l:i:Jh5 ! 2 2 gxf4 exf4 2 3 i.c4 l:i:Jg3 Black has chances for a strong
attack, Ziegler-Shulman, Goteborg 1 999 (24 i.e6 'iif6 ! 25 i.xg3 fxg3 26
i.h3 l:i:Jf4 27 'ifd2 'iih4 28 l:i:Jc2 .:tfb 8 etc.).
a

13 li.Jg6

If 13 a5 ! ? 14 l:i:Jd3 b6 1 5 a3 l:i:Jg6 1 6 b4 l:i:Jf6 17 c5 bxc5 ! 1 8 bxc5 .:t7

1 9 cxd6 ! cxd6 with chances for both sides, Atalik-Salinnikov, Alushta
1 999.
...

It is only now, after the black kingside pawn

phalanx has advanced with maximum speed,
7 that Black can successfully try the older, less
6 recognized method of conducting the attack
6
5 on the white king with 13 .:tf6 ! ? (see dia5
4
gram) 14 b4 .:th6 1 5 l:i:Jd3 a6 1 6 c5 'ii'e 8 1 7
4
3 'it>h 1 b5 1 8 cxb6 cxb6 1 9 g4 fxg3_20 i.xg3 b5
3
21 l:i:Jf2 l:i:Jf6 22 l:i:Jg4 .:tg6 23 a4 i.d7 24 axb5
2
axb5 25 'iid 3 .:tb8 26 l:i:Jxf6 .:txf6 27 'iid2
l:i:Jg6 28 'iix g5 'ii'f8 29 'iie 3 i.h6 30 'ilia7 i.h3
a
b
c
d e f g h
3 1 .:ta l .txfl 32 .:txfl i.f4 33 'it>g2 .:tc8 34
l:i:Jxb5 'ifh6 3 5 'iia2 l:i:Jh4+ White resigned, Atalik-Kasimdzhanov, Gron
ingen 1 999. This is an encouraging example from Black's point of view but
in this line Black has faced bad experiences as well. For example 14 b4 a6
15 c5 .:t h6 16 l:i:Ja4 (Focusing on the weakened b6-square. Both 16 cxd6?!
cxd6 17 g4? ! fxg3 18 hxg3 l:iJg6 19 l:i:Jg2 l:i:Jf4 ! ! Speelman-Uhlmann, Lenin
grad 1 984, or 1 6 l:i:Jd3 ? ! 'ii'e 8 1 7 i.e 1 'ii'h 5 1 8 h3 l:i:Jxc5 19 bxc5 i.xh3 20
8
7

..

Variation 9 lD e i lDd7 1 0 i. e3 4 7

Ilf2 i.d7 2 1 Ilfl i.f6 2 2 i.d2 'ii'h 2+ 23 'ifr> f2 'it'g3+ 2 4 <iitg 1 l:.h2 25 Ilf2
'fi'h4 0- 1 Ramos-Bologan, Las Palmas 1 993, are weaker) 1 6 'ii'e8 17 'ith1
'ii'h 5 1 8 i.g1 'ii' h4 1 9 lDd3 ! b5 20 cxb6 cxb6 21 'ife l 'ir'h5 22 lDf2! with
advantage for White, Korchnoi-Nataf, Cannes 1 998
Also worth mentioning is White ' s thematic idea of sacrificing a pawn (as
in the main game) by playing 14 c5, trying to obtain pressure on the c-file
as quickly as possible. In the game Vera-Nataf, Havana 2002, White suc
ceeded in that after 14 ... lDxc5 15 b4 lDa6 16 lDd3 Ilh6 17 a4 'it'e8 18 'ith1
'iVh5 (More consistent than 1 8 ... c6 19 b5 lDc7 20 dxc6 bxc6 21 lDb4 c5 22
li.Jbd5 lDexd5 23 lDxdS lDe6 24 b6 with an initiative for the pawn sacrifice,
Lopez Martinez-Volokitin, Leon 200 1 ) 19 i.gl i.d7 20 lDa2 Ilc8 21 lDf2
'tieS 22 lDg4 l:lg6 23 b5 lDc5 24 i.xc5 dxc5 25 l:lxc5 b6 26 Ilc3 h5 27 lf::, f2
'ot?h8 28 lDb4 ltlg8 29 lDa6 'it'd8 30 h3 i.f8 31 'ii'c 2 i.d6 32 Ilcl with
strong pressure on the c-file, Vera-Nataf, Havana 2002 .
.

14 c5 !

Kozul was the first to come up with this idea, which was later analyzed
and tested in practice by Piket and Korchnoi. This pawn sacrifice enables
White to speed up his action on the queenside.

14 c!Llxc5 15 b4 c!Lla6 1 6 c!Lld3 !

Piket's improvement. There is no need to

to recapture the sacrificed pawn by 16
l,i)b5 (see diagram) 1 6 l:lti (or 1 6 . . . i.d7 1 7
4'1xa7 ! ? h5 1 8 a3 Korchnoi-M.Ivanov,
Enghien 1 997, when Black should have
wntinued with 1 8 . . . l:lf7 ! 1 9 'ii' c2 g4 20 i.xa6
hxa6 2 1 'it'c7 'it'g5 or 2 l . . .'it'f8 with chances
for both sides) 1 7 lDxa7 i.d7 1 8 a4 h5 19
b5 i.h6 20 l:lc4 Ilg7 2 1 h1 lDf8 22 g3
fxg3 23 i.xg3 lDg6 24 lDd3 f4 25 i.xf4 gxf4
26 l:lg1 i.g5 27 i.fl ! i.h4 28 Ilxg7+ xg7
h u rry

..

6
5

6
5

4
3

1
3
2

2
1

48 Variation 9 li:Jel li:Jd7

10

i.. e3

with a double-edged position, Atalik-Timoshchenko, Romania 1 998, or

1 7 . . .lt:Jxb4 Atalik-Osterman, Bled 1 999, when White recommends as the
best solution 1 8 a4 ! lt:Ja6 1 9 lt:Jxc8 'ii'x c8 etc.

16 hS 17 lbbS i.d7 18 a4 i.h6 1 9 .:tc3 b6! ?

.

In the _game Korchnoi-Cvitan, Pula 1 997, was played 1 9 . . .l:tf7 20 i[)xa7

llg7 2 1 llJb5 lt:Jf8 22 h3 lt:Jh7 23 i.e 1 i[)f6 24 lt:Jt2 lt:Jxb4 25 llxc7 lt:Ja6 26
llxb7 lt:Jc5 27 llc7 g4 28 hxg4 hxg4 29 fxg4 lt:Jfxe4 30 i[)xe4 lt:Jxe4 3 1 'ifb3
lt:Jg3 32 i.xg3 fxg3 33 'ii'xg3 i.xb5 34 Axg7+ rtixg7 35 g5 i.xg5 36 i.xb5
'ifb6+ 37 llfl 'ii'e3 38 'ii'f3 with equal chances.
.

20 i.e1 !
20 'ii'd2 g4 2 1 llfc 1 h7 22 lt:Jxc7 lt:Jxc7 23 llxc7 g3 is double-edged.

20 Jr.ti 21 lbf2 lbh4 22 lbxd6 cxd6 23 i.xa6 'fie8! 24 'fie2 ! ? g4!

.

Not 24 . . . i.xa4? 25 b5 .

25 fxg4
a

2 S Ilg7?

Black missed the chance for 25 . . . i[)xg2 ! ! 26 g2 hxg4 27 'ii'c 2 (if 27

h 1 ? f3 28 'ii'c2 g3 29 hxg3 i.e3 ! threatening a mating attack after . . . :h7
and . . .'ii'h5 ) 27 . . . llh7 ! (27 . . . f3+ 28 g 1 would be bad) 28 i[)h 1 'ii'h 5 29
i[)g3 fxg3 30 i.xg3 with unclear chances.

26 h3 'ifg6 27 i.bS! i.xbS 28 axbS .:us 29 lbd1

White has consolidated and his position is superior.

29 hxg4?!

I t was better t o defend the piece b y moving the queen t o f6 o r g5 .

30 i.xh4 f3 31 'fic2 gxh3

If 3 1 . . .'ii'h 5 32 g3 gxh3 33 h l .

Variation 9 CiJeJ CiJd7 1 0 i.. e 3 49

32 g3 ? ?
White could have won after 32 l:lfxf3 'ii'xg2+ ( i f 32 . . . l:lxf3 3 3 l:lxf3 'ii'g4
34 i.g3) 3 3 'iVxg2 l:lxg2+ 34 'it>h l l:lxf3 3 5 l:lxf3 .

32 ...l1f4! 33 tbe3 ?
Again correct was 33 l:lfxf3 l:lxh4 34 CiJ2 h2+ 35 'it>h l l:thS 36 l:tc8+
Wh7 37 l:lf8.

33 .lbe4 !
.

Black takes his chance. After a couple of inaccurate moves by White he

now has the advantage. Not 3 3 . . . .l:txh4 because of 34 CiJfS .

3 4 tt:Jf5 ? ! 'ifxf5 35 .l:.cxf3 'tig4 36 h l

36 .l:tf8+ h7 37 .l:te l 'ii'xh4 3 8 'ii'xe4+ 'iVxe4 39 .l:txe4 l:lxg3+ 40 h2
didn't work, because after 40 . . . .l:tb3 ! followed by . . . i.h6-f4 Black wins.

36 ... ltf4! 37 l:L3f2 l::tx f2 38 J:xf2 e4 ! 39 :r6 e3 40 l:te6

a

50 Variation 9 l'i::le l l'i::l d7 1 0 i.. e 3

40 .. .'.-fJ+ 41 gl l:tt7 42 Wg6+ i.g7 43 lle8+ l:U8 44 l:xf8+

xf8
Or 44 . . . "ikxf8 45 1i'e6+ "ikfl 46 'ii'x e3 'ifxd5 47 'ilfe2 ! and Black has no
advantage.

45 i.e7+! g8
If 45 . . . xe7 46 1i'e6+ f8 47 'i!fc8+ etc.

46 'ii' e6+ 'ift7 47 'ii'c 8+ i.f8 48 'ii'g4+ i.g7 49 'ii'c 8+ h7 50

'ii'x h3+ i.h6 51 i.g5 'iff2+ th-th.
Game 1 2

Piket - lvanchuk
Wijk aan Zee 1 999
1 d4 f6 2 c4 g6 3 c3 i.g7 4 e4 d6 5 f3 0-0 6 i.e2 e5 7 0-0
c6 8 d5 e7 9 el d7 10 i.e3 f5 1 1 f3 f4 12 i.f2 g5 13 l:c l
g6 14 c5 xeS 1 5 b4 a6 1 6 d3
a

3
2

As was mentioned in the 12revious game, this positional approach is more

solid than the sharper 16 lL!b5 as has often been played by Atalik and
Yermolinsky.

1 6 . l:t7
.

Ivanchuk opts for a quite different continuation from 16 . . . h5 1 7 l'i::!b 5 i.. d 7

1 8 a4 i.. h 6 19 l:tc3 with unclear chances, as was demonstrated in the above
game, Piket-Kasparov, Linares 1 997. In Korchnoi-Cvitan, Pula 1 997, Black
made some improvements, but here Ivanchuk thinks that Piket has a
prepared response to those as well.

1 7 b5

Variation 9 l?:JeJ &iJd7 1 0 Jl.e3 51

After 17 a4 White was wary of the possible transfer of the passive knight
from a6 via b8 and d7 in order to strengthen the delayed attack on the king
s ide, so, by threatening to capture directly on a7, he forces Black to occupy
t he d7 square with his bishop.

1 7 ... .i.d7 1 8 a4
A slower alternative was 1 8 l?:Ja3 , intending l?:Ja3-c4-a5 .

2
a

1 8 ...'ife8 !
The most efficient way for Black to discourage White from taking on c7.
The alternatives are
1 8 . . . i.xb5 1 9 axb5 l?:Jb8 20 'ic2 h5 2 1 'ifilh1 i.h6 22 l:.g 1 l:.g7 23 l:.a 1
l?:Jd7 24 l:.a3 l?:Jf6 25 l:.ga 1 g4 26 l:.xa7 l:.xa7 27 l:.xa7 g3 28 .i.e 1 b6 29
'ii'c 6 1h- 1h, Krivoshey-Dobrowolski, Koszalin 1 999, and
1 8 . . . h5 1 9 l?:Jxa7 i.f8 (or 1 9 . . . i.h6 20 l:.c3 l:.g7 2 1 l?:Jb5 l?:Jf8 22 h3 l?:Jh7
23 i.e 1 l?:Jf6 24 l?:Jt2 'iti>h8 Te.Johansen-Djurhuus, Norway 1 998) 20 l?:Jb5
l:.g7 2 1 l:.c4 l?:Jh8 22 g4 ! hxg4 (22 . . . fxg3 23 hxg3 h4 24 g4 is in White' s
favour) 23 fxg4 l:.h7 2 4 'iti>g2 'if6 25 h 3 c 6 2 6 l?:Jc3 l?:Jc7 2 7 i.f3 and White
stands better, Piket-Nijboer, Rotterdam 1 999.

1 9 b2
Here we have perhaps a critical position. After 19 l:.c3 h5 ! 20 'iib 3 g4
Black' s attack on the kingside might be faster than White ' s pressure on the
queens ide.

52 Variation 9 li:,e] li:,d7 1 0 i.e3

1 9 b8 ! ?
..

Or 1 9 . . . h5 20 li:,c4 i.f8 2 1 li:,a5 'ifb8 22 li:,c3 with an unclear game.

20 'ilc2 ! ?
White threatens to take o n c7 . I f 2 0 li:,c4 li:,xb4 or 2 0 'i!Vd2 h5 2 1 .l:tc4 g4
Black's attack would advance quickly .
a

2
1
a

20 ... i.e8?
A more promising concession would be the one with 20 . . . i.xb5 2 1 axb5
li:,xb4 22 'ifb3 (or 22 'i!Va4 a5 23 .l:tc4 b6 24 .l:ta l .l:tf8 25 i.e l with unclear
chances) 22 . . . a5 23 b6 i.f6 24 li:,a4 i.d8 25 .l:tc3 li:,e7 26 .l:tfc l li:,c8 27
i.b5 ! with the initiative in return for the material investment, or the more
enterprising 20 . . . c6 2 1 dxc6 bxc6 22 li:,xd6 etc .

2 1 'ii'd 2 h5 22 J::[ c4 .ih6 23 l:tfcl g7

Variation 9 liJeJ liJd7 1 0 i.. e3 53

23 .. .'.t>h7 would be similar with the possible continuation 24 'ii'c 3 c6 25

dxc6 bxc6 26 l:txc6 i.xc6 27 'iVxc6 liJxb4 28 1i'c4 l:tb7 29 1i'xb4 a6 etc.
White always has compensation for the sacrificed pawn in his queenside
pressure and in neutralising Black 's attempt to organize counterplay on the
t 1pposite flank.

24 ctJd3 !
Not so dangerous here for Black is 24 1i'c3 c6 25 dxc6 bxc6 26 l:txc6
.Jil.xc6 27 'iVxc6 liJxb4 28 'ifxd6.

24 .. .'ii d 8 25 'ii'c 3 g4 26 ltJxc7

This is a safer continuation than the more complicated 26 l:txc7 .

26 ltJxc7 27 l:.xc7 g3 28 .tel

..

1
a

28 ... gxh2+?
With this move Black deprives himself of any counterplay. Better
resistance would have been offered by playing 28 . . . i.xa4 29 l:txf7+ xf7
30 'ii'c 7+ liJe7 3 1 1i'xb7 with a slight advantage for White.

29 'lt>h l .i.g5 30 litxti+ .i.xti

If 30 . . . xf7 3 1 'iVc7+ 'iVe7 32 b5 etc.

3 1 'ii c 7 'ii b 8 32 b5 .i.d8 33 'ii x b8 lixb8 34 i.b4 .i.e7 35 :c7 f8

36 ctJb2 r,t>e8 37 ctJc4 'lt>d8 38 .i.xd6 Black resigned.
After 38 . . . i.xd6 39 l:txf7 i.c5 40 l:tf6 the position collapses.

Part Two : Lines after 9 ltJel l2Je8

I n the early 1 990s several o f the world 's top players tried t o reach ' Mar
del Plata' positions in a different way by substituting the standard 9 . . . /t)d7,
considered by the creator of this system as the correct reply, with 9 . . . /t)e8 .
a

An advantage of this less frequently played move is the preliminary

protection of the weak square c7 (and d6), if possible to avoid the continu
ation 9 /t) e1 /t)d7 10 /t)d3 f5 1 1 i.d2 /t)f6 12 f3 f4 13 c5 g5 14 .l:.cl /t)g6
15 /t)b5 .l:.ti 16 cxd6 cxd6 17 'ii'c 2 /t)e8 (for 1 7 . . . g4 see the comments to
Larsen-Tal, Eersel 1 969, Game 6) and White gains one tempo (but not two
tempi thanks to Black' s involuntary manoeuvre . . . /t)d7-f6-e8-f6, since
White ' s 1 7'h move is an intermediate move and the queen will need to spend
one more move leaving the c2 square to make room for a doubling of white
rooks along the c-file) in the race for the initiative on opposite wings.
The disadvantage of 9 . . . /t)e8 ! ? is the knight's reduced control over c5
and e5, squares of the eventual pawn break in White ' s strategy, so that
White obtains an additional option for the break 1 0 f4, or 1 0 l2Jd3 f5 1 1 f4
(the games which follow illustrate both possibilities) on the kingside, with
unclear consequences for both sides, but with faint hopes for Black of
obtaining a strong attack on the kingside, since White has sufficient space
available on that wing too.
If White is not pleased with the outcome of the pawn break on the king
side, he may go for standard methods, like 10 /t)d3 f5 1 1 i.d2, 1 2 f3, etc . ,
aiming for the usual break c4-c5 o n the queenside, o r 9 /t)e 1 /t)e8 1 0 i.e3
(similar to games in the preceding chapter) .

Variation 9 ttJel ttJe8 1 0 f4

Game 1 3

Sosonko - Nij boer

Dutch champion ship 1 994
I d4 ti:lf6 2 c4 g6 3 ti:lc3 i.g7 4 e4 d6 5 i.e2 0-0 6 ti:lf3 e5 7 0-0
t;i)c6 8 d5 ti:le7 9 ti:lel

An interesting idea is 9 ..tih 1 ltle8 1 0 ltle 1 f5 1 1 exf5 gxf5 1 2 f4 ltlg6 1 3

rx c5 dxe5 1 4 .ie3 f4 ! ? 1 5 .ic5 ltld6 1 6 .id3 .if5 1 7 .ixf5 l:txf5 1 8 .ixd6
rxd6 1 9 ltle4 .l:r.c8 20 .l:r.c 1 'ife7 2 1 ltlt3 .if6 22 c5 dxc5 23 d6 'ii'f8 24
'ir'd5+ ..tih8 25 d7 .l:r.b8 26 .l:r.xc5 with a strategically won game, .Kramnik
M iklos, Mainz 200 1 .

9 ti:le8 1 0 f4
...

2
1
a

l O exf4

..

Better than 10 . . . f5 1 1 fxeS dxeS (more cautious is 1 l . . . .ixe5 1 2 .ih6

ltlg7 13 'ifd2 fxe4 14 .l:r.xf8+ 'iix f8 1 5 ltlxe4 ltlef5 with a tough position,
Teschler-Erwes, Hessenliga 1 994) 12 ltlf3 (playable is 12 ltld3 c6 1 3 .ig5
h6 14 .ih4 g5 1 5 .ig3 cxd5 16 cxd5 lDg6 1 7 exf5 .ixf5 1 8 .ig4 with a
clear advantage, Petran-Kiss, Budapest 1 99 5 ; or 1 2 .ig5 h6 1 3 .ih4 .if6 1 4
.txf6 lDxf6 1 5 ltld3 fxe4 1 6 ltlc5 ltl f5 1 7 'ifd2 'iti>h7 1 8 ltl5xe4 with the
superior game, Koper-Kolar, Prague 1 995) 12 lDd6 13 lDgS Hillarp
Persson-J.Hansen, Nordic Grand Prix 1 997.
. .

Unclear is 1 0 c6! ? 1 1 dxc6 bxc6 12 fxe5 'iib 6+ 13 <i>h 1 dxe5 14 b3

'ifc7 1 5 .ta3 .ie6 16 lDd3 .l:r.d8 1 7 'ife 1 f5 1 8 :d 1 with slight white
pressure, Neverov-Iskusnyh, Azov 1 99 5 .
. .

56 Variation 9 ti:Jel ti:Je8 1 0 f4

1 1 xf4 h6
a

A solid method to fight for control of the e5

square. Inferior is l l . . f5 1 2 'iVd2 ti:Jf6 (worse
7 is I 2 . . . fxe4 I 3 ti:Jxe4 and Black has weak g5
7
6 and e6 squares) 13 exf5 ti:Jxf5 (see diagram)
6
I 4 g4 (or I 4 ti:Jc2 h5 I 5 i.g5 l:lf7 I 6 i.d3 'Wf8
5
I
7 h3 c5 I 8 l:lf3 ti:Jd7 I 9 i.xf5 l:lxf5 20 l:lxf5
4
gxf5 2 I l:lfl ti:Je5 22 ti:Je3 'iVf7 23 ti:Je2 1Wg6
3

24 h4 tl:Jg4 25 ti:Jf4 'iVf7 26 ti:Jxg4 hxg4 27

2
2
l:le i i.e5 Hillarp Persson-Hjartarson, Gausdal
ab.='l 1
I 996, where-instead of the risky 28 l:lxe5? !
a
b
r
d c f g h
-the simple continuation 28 ti:Jd3 i.d4+ 29
h i i.f6 30 i.xf6 1Wxf6 3 I 'iVg5+ 'iVxg5 32
hxg5 f7 3 3 ti:Jf4 offered a solid initiative) I 4 . . . ti:Je7 I5 h3 h5 I6 i.h6
hxg4 I 7 i.xg7 xg7 I 8 hxg4 c5 I 9 ti:Jg2 i.d7 20 l:lae i with the superior
game, Weglarz-S.Farago, Hungary I 999.
8

Playable is 1 1 . . . c6 (see diagram) 1 2 'iWd2

cxd5 13 cxd5 (or I 3 exd5 ti:Jf6 I 4 h i i.f5
I 5 ti:Jf3 a6 I 6 l:lae I l:le8 I 7 ti:Jd4 tl:Je4 I 8
6
6
ti:Jxe4 i.xe4 I 9 i.g4 ti:Jf5 20 ti:Je6 fxe6 2 I
5
l:lxe4 e 5 2 2 i.g5 i.f6 2 3 i.e3 tl:Jxe3 24 'iVxe3
4
i.g5 25 'iVh3 l;H..S , Wells-Kupreichik,
3
: Regensburg I 997) 13 . . . f5 14 ti:Jd3 fxe4
2
2
( I 4 . . . ti:Jf6) 15 ti:Jf2 e3 ! (Gaining an important
1
tempo. The chances are equal.) 1 6 i.xe3 ti:Jf5
a
b
c
d e f g h
17 i.f4 h6 1 8 i.d3 g5 19 i.e3 ( I 9 i.xf5
..txf5 is okay for Black) 19 . . . ti:Jf6 20 l:lae1
ti:Jxe3 2 1 11i'xe3 i.d7 22 h3 'iWb6 (22 . . . 1Wa5 ! ?)
23 'ii' x b6 axb6 24 ..tg6 i.e8 25 i.xe8 l:lfxe8 26 l:le6 l:lxe6 27 dxe6 l:le8 28
l:le1 ti:Jh5 29 ti:Jd5 i.e5 30 e7 ti:Jf4 3 1 ti:Jxf4 gxf4 32 ti:Jd3 l:lxe7 33 ti:Jxf4
lfl-lh, Yusupov-Kupreichik, Eupen I 996.
7

1 2 ti:Jc2
8

8
7

6
5

....
r

3
2
1

White players have tried other continuations

as well: (see diagram)
12 h4 ! ? f5 1 3 'iVd2 'itt h 7 I4 ti:Jf3 fxe4 I 5
ti:Jxe4 ti:Jf5 I 6 g3 ti:Jf6 I 7 ti:Jxf6+ 1Wxf6 I 8 ..td3
i.d7 I 9 l:lae i b6 20 g2 l:lae8 Wells-Kocsis,
Zalakaros I 996.
1 2 g4 g5 I 3 i.g3 ti:Jg6 I 4 ti:Jd3 i.d4+ I 5
g2 f6 I 6 1Wd2 c 5 I 7 ti:Jd I KasperDeutschmann, St.Veit 2000, and instead of
I 7 . . . b5 ? ! I 8 cxb5 a6 I 9 bxa6 ti:Jc7? 20 ti:Jb4 !
etc., Black should have continued with the
solid I 7 . . . ti:Je5 .

Variation 9 t:iJeJ t:lJe8 1 0f4 5 7

1 2 tiJ f3 f5 (Better i s 1 2 . . . g5 ! 1 3 .itd2 f5 1 4 exf5 tiJxf5 with even chances)

3 e5 g5 14 exd6 cxd6 1 5 ..ie3 f4 16 ..if2 tiJf5 1 7 tiJd4 .ite5 I 8 tiJe6 ! ?
i.xe6 I 9 dxe6 tiJc7 2 0 ..i f3 h8 2 1 'i'd3? ! (2 1 .itxb7) 2 I . . . tiJxe6 2 2 .itxb7
tnc5 23 .itxc5 dxc5 24 'i'h3 .l:!.b8 25 i.e4 g4 ! 26 'i'xg4 .l:!.xb2 27 tiJd5 'i'g5
28 'i'h3 tiJd6 29 :tae i tlJxe4 30 Ihe4 l1g8 3 1 g4 fxg3 32 l1g4 i.d4+ and
White resigned, Wells-Kupreichik, Bad Worishofen I 997.
I

1 2 ..ie3 f5 I3 ..id4 g5 (there is no reason to delay this move; inferior is

1 3 . . . tiJf6 I4 e5 dxe5 I5 i.xe5 Wells-S iebrecht, Bad Worishofen I 997) 14
4'lf3 tiJf6 1 5 e5? ! ( I 5 i.d3) 1 5 . . . dxe5 I6 tlJxe5 f4 was in Black favour,
Lavrov-Galkin, Rostov-on-Don I 996.

12 . . . f5 1 3 exf5 g5
a

3
2
a

14 i.d2
Or I4 i.e3 tiJxf5 I 5 i.f2 ( I 5 'i'd3 ! ? 'i'e7 I6 i.d2 .ite5 1 7 l1f2 tlJeg7 1 8
llafl Gavrilov-Navrotescu, Eoforie I 996) I 5 . . . i.e5 1 6 ..td3 'i'e7 I 7 'ife2
"iig 7 I 8 h i tiJf6 I 9 i.g i tiJh4 20 ..id4 ..ig4 2 I 'ifd2 tiJd7 22 i.e2 1h-1h,
Sosonko-Van der Wiel, Dutch championship I 994.

14 . . . tLlxf5 15 i.d3 'fie7 1 6 :n tLld4 17 'ii e l 'ifxel+ 1 8 lbel i.f5

19 i.xf5 1lxf5 20 tLlxd4 i.xd4 21 i.e3 xfl 22 i.xfl i.xc3 23 bxc3
<l;f] 24 i.d4 c5 25 i.fl
Better was 25 dxc6 bxc6 26 c5 d5 27 :tb I ltc8 with an even game.

25 . . . tLlf6 ! 26 i.g3 l:le8 ! 27 l:Ib l tLle4

More enterprising was 27 . . . b6 28 i.xd6 tlJe4 29 i.g3 tlJxc3 .

28 llxb7+ l:te7 29 lbe7+ xe7 30 i.el h5 3 1 fl f6 32 e2

f5 33 <it>f3 a5 34 h3 h4 35 a3 a4 36 e3 e5 37 <t>f3 <t>f5 38 e3
tLlf6 39 f3 tLld7 40 g4+ <t>f6 41 e4 tLle5 42 i.d2 tLlxc4 43 .tct
g6 44 d3 tLle5+ 45 e3 tLld7 46 i.b2 c4 47 i.ct tLlc5 48 f3
tLld3 49 i.e3 tLlel + 50 <t>e2 tLlc2 5 1 .tel that 52 i.d2 tLlc2 1/z-1/z ,

Variation 9 ltJe1 ltJe8 1 0 tiJd3 f5

1 1 f4
Game l4

Atalik - Nij boer

Wijk aan Zee 1 99 7

1 d4 tbf6 2 c 4 g 6 3 tbc3 i.g7 4 e 4 d 6 5 i.e2 0-0 6 tb f3 e 5 7 0-0
tbc6 8 d5 tbe7 9 tbe1 tbe8 10 tbd3
The immediate 1 0 f4 was analyzed in the previous game.

1 0 . . . f5 1 1 f4
a

3
2

The idea of this system is to prevent the activity by Black on the kingside
and, by opening the position, to exploit the weakness of the e6 square.

l l . exf4
. .

Alternatives are 1 1 . . . c6 and 1 1 . . . fxe4. For example:

l l c6 12 fxe5 dxe5 13 i.g5 (more active than 1 3 lLlf2 cxd5 1 4 cxd5
lLld6 1 5 i.e3 i.d7 1 6 "ifb3 [playable was 16 .i.c5 lLlec8 1 7 a4 b6 1 8 i.a3
i.h6 with chances for both sides] 1 6 . . . b6 and Black had a promising
position, Gavrilov-Sakaev, St.Petersburg 1 994) 13 h6 14 i.h4 cxd5 1 5
cxd5 lLld6 (or 1 5 . . . g5 1 6 i. f2 b 6 1 7 exf5 .i.xfS 1 8 i.g3 lLlg6 1 9 i.g4 lLlf4
. . .

. . .

Variation 9 ti:Jel ti:Je8 1 0 ti:Jd3 f5 I I f4 59

20 i.xf5 Axf5 21 ti::J f2 ti::Jd6 22 ti::J fe4 Neverov-Burovic, Porca 1 993) 16

ti::J c 5 'ii'c 7 1 7 i.f2 b6 1 8 ti::J e 6 i.xe6 1 9 dxe6 ti::J x e4 20 ti::J x e4 fxe4 2 1 1:.cl
with a clear advantage for White, Hiibner-Herbrechtsmeier, Germany 1 982.
ll . fxe4 1 2 ti::J x e4 i.f5 (inferior is 12 . . . c6 13 fxe5 Axfl + 1 4 Wxfl dxe5
1 5 .i.g5 cxd5 16 cxd5 'ii'b6+ 1 7 ti::Jd c5 ti::J f5 1 8 <iif h 1 as in Petran-Orso,
Budapest 1 978) 13 .i f3 ! ( 1 3 ti::Jg 3 e4 14 ti::J f2 ti::J f6 1 5 'it'b3 c6 16 1i'xb7
cxd5 1 7 cxd5 Ab8 1 8 'ii'x a7 ti::J e xd5 1 9 .i.c4 <iif h 8 and Black's pieces were
much better coordinated, Johansson-Kavalek, Halle 1 963) 13 c6 14 1i'b3
cxd5 15 cxd5 .ixe4 16 .txe4 ti::J f6 17 Ae1 with potential pressure on the
light squares, Smiov-Jones, Canberra 2000.
. .

1 2 lbxf4
The possible continuation 1 2 ii.xf4 (see
diagram) is lel)S frequently played:

6
s

12 . . . .i.xc3 ? ! 1 3 bxc3 fxe4 1 4 ti::Jb 4 ti::J f5 1 5

'lli'd2 ti::J f6 1 6 ti::J c2 'ii'e 7 1 7 g4 with advantage,
Gavrilov-Galkin, St.Petersburg 1 994.

7
5

3
2

12 . . . fxe4 1 3 ti::J xe4 ti::J f5 ! 1 4 ii.g5 ( 1 4

'ii'd 2 ! ? ; 14 <iit h l ! ?) 1 4 . . . ti::J f6 1 5 g4 ti::Jd4 1 6
ti::Jd f2 'ife7 ! !Hh, Petrosian-Tal, Bled 1 96 1 .

1 2 . h6 1 3 exf5 ( 1 3 e5 ! ? g5 and now either

the pawn sacrifice with 14 .ig3 f4 1 5 .if2
dxe5 1 6 .ig4 or 1 4 cxd6 seems to be more dangerous for Black) 13 . . ti::J xf5
(stronger than 1 3 . . . g5 1 4 .id2 i.xf5 1 5 'ifb3 'ifc8 1 6 l::ta e 1 Fraguela
Hiibner, Las Palmas 1 976) 14 g4 ti::J d 4 15 g2 ti::J f6 16 h3 h5 17 xh5
ti::J x e2 18 Wxe2 ti::J x h5 19 .i.d2 .tf5 20 'ii'f3 'ifh4 21 ti::J f4 ti::J xf4+ 22 xf4
Aae8 White resigned, Beckmann-Hillarp Persson, Recklinghausen 1 999.
. .

1
a

12

. . .

fxe4

60 Variation 9 &Del 0.e8 1 0 0. d3 f5 1 1 f4

6
4

3
2
a

8
7

5
4

3
2

.'1

1:1

"

e .:..
f
g 1...1..a
,

8
7

6
5

1:1

...a

8
7

2
"

-....a

3
2
1

Other options are:

12 . . . i.xc3 ? ! (see diagram) 13 bxc3 fxe4 14
0.e6 ! i.xe6 1 5 !:txf8+ Wxf8 1 6 dxe6 0.g8 1 7
'ii' d 5 0.ef6 1 8 'i'xb7 'ii' c 8 1 9 l:t b l We7 20 c5
'ii' x b7 21 !:txb7 Wxe6 22 !:txc7 PengKotronias, Wijk aan Zee 1 995.

12 . . . 0.f6 (see diagram) 13 exfS 0.xf5 1 4

g4 ! 0.e7 (or 1 4 . . . 0.h4 1 5 0.e6 i.xe6 1 6 dxe6
ll:Je8 17 0.d5 c6 1 8 lhf8+ i.. x f8 19 fl 0.f5
20 gxf5 cxd5 2 1 fxg6 i.. g 7 22 'ii' f7+ W h8 23
e7 'ifb6+ 24 '1t>h 1 hxg6 25 'ii'x g6 'ii'c 6 26 i.f3
0. f6 27 ..ih6 Black resigned, Kozul-Grosar,
Pula 1 996) 15 0.bS ! c6 16 dxc6 bxc6 17
0.xd6 i.xg4 18 cS li'd7 19 !:tel 0.ed5 20
0.xd5 cxd5 21 ..igS ..ixe2 22 !:txe2 'ii' c 6 23
!:tel 0.g4 ? (23 . . . 0.e4) 24 l:te7 0.f6 25 b4 a6
26 'ii' e2 0.e4 27 0.xe4 dxe4 28 'ii' x e4 'ii' a 4 29
!:txg7+ Wxg7 30 'ii'b 7+ Wh8 3 1 i.h6 !:tg8 32
..id2 Black resigned, Kozul-Cvitan, Osijek
1 992.
12 . . . c6 (see diagram) 13 ..ie3 ( 1 3 Wh 1 0.f6
1 4 exf5 0.xf5 1 5 g4 0.e7 1 6 i.e3 is slightly
better for White) 13 . . . 0.f6 14 exf5 0.xf5 1 5
..if2 !:te8 1 6 i.. f3 ..i d 7 17 !:tel ..i h 6 18 0.e6
..ixe6 19 dxe6 ..if4 20 g3 ..ie5 21 c5 i.xc3 22
bxc3 dxcS 23 ..ixc5 'ifa5 24 .i.f2 'ii'x c3 25
!:tel 'ti'a3 26 'ii' e2 !:tad8 27 ..ig2 'ii' d 3 28 'ii'b 2
0.g4 29 ..ixa7 0.xg3 30 !:tc3 a6 3 1 !:ta3 !:td2
32 'i'xd2 'ii' x a3 33 i.b6 &Drs 34 'ii' d 7 ike7 35
'ii' x e7 !:txe7 36 i.c5 !:te8 37 i.fl 0.f6 38 !:tb l
0.e4 39 i.a7 0.fd6 White resigned, KozulBlees, Munich 1 993.
1 2 . . . ..ie5 (see diagram) 13 0.d3 i.d4+ 14
0.f2 (Better is 1 4 Wh 1 ) 14 . . . i.g7 15 g3 h6 1 6
i.. e3 0.f6 17 c 5 fxe4 1 8 0.fxe4 ll:Jxe4 19 0.xe4
!:txfl + 20 'ifxfl &Drs 21 i.. f2 i.xb2 with the
preferable position for Black, Yermolinsky
Van Wely, Wijk aan Zee 1 997.

Variation 9 l1JeJ l1Je8 1 0 l1Jd3 f5 1 1 f4 61

1 3 tbxe4 tbf5 14 '1th 1 'ii/ e 7

Or 14 . . . l1Jf6 1 5 l1Jg5 l1Jg4 1 6 i.xg4 'ii'x g5 1 7 l1Je6 l1Jg3+ 1 8 hxg3 l:lxfl +
1 9 'ii'xfl 'ii'xg4 20 l1Jxg7 xg7 2 1 i.d2 i.f5 22 i.c3+ <Ji;f7 and Black is not
worse, Chiong-Howell, Calcutta 1 996.

1 5 i.d3 i.e5 16 a4
PreP.aring the activation of White ' s queen ' s rook via a3 . If 1 6 g3 l1Jeg7 1 7
l1Jg2 'i.fe8 1 8 l:lb 1 i.d7 1 9 b3 c5 20 l:le 1 'ii'c 8 2 1 i.f4 i.xf4 22 l1Jxf4 'i!lc7
23 b4 b6 24 bxc5 dxc5 25 l1Jg5 l1Jd4 lh- lh, Yennolinsky-Rogozenko, Bad
Zwesten 1 997.

16 . . . tbf6 1 7 tbg5 tbg4 ! 1 8 'YWxg4 tbe3

a

19 tbxg6 !

Playable was 1 9 'ii'g 3 l:iJxfl 20 i.xfl 'ii'f6 2 1 'ii'h4 h5 22 g3 with the

better chances.

19 . . . .l:.xfl+ 20 i.xfl hxg6 21 'ii e4 ! 'ti'xg5

Not 2 1 . . . l:iJxfl ? 22 'ii' x g6+ i.g7 (if 22 . . . 'ii'g 7 23 'ii'e 8+ 'ii' f8 24 'illx f8+
xf8 25 l1Jf3 ! i.g7 26 i.g5 i.xb2 27 l:txfl +) 23 i.f4 'ili'e2 24 i.e5 ! ! 'i!lxe5
2 5 'ii'h 7+ <Ji;f8 26 l:txfl + winning.

22 i.xe3 'iVh5 23 i.f4 i.f5 24 'ii e3 .txb2 25 lle1 i.f6 26 c5!

i.d7 ! 27 'i!Vb3 l:.e8 28 :xeS+ i.xe8 29 cxd6 cxd6 30 'ii'x b7! Wf5 3 1
b8 !
The only reply.

31 . . .'ii' xf4 ! 32 'i'xe8+ <it>g7 33 'iiid7+ <it>h6 34 'ii'h 3+ <it>g7 35

'YWd7+
Not 35 'l'd3 i.e5 36 g3 'i!lxa4.
35 . . . <it>h6 36 'i!Vh3+ 1/2 -1/2,

Variations after 9 ttJe1 ttJe8

1 0 ttJd3 f5 1 1 d2 ttJf6
Game 1 5

LaUe - McShane
Redbus Knockout, Southend 2000
1 d4 tiJf6 2 c4 g6 3 ltJcJ i.g7 4 e4 d6 5 tiJf3 0-0 6 i.e2 e5 7 0-0
ltJc6 8 d5 ltJe7 9 ltJe1
Another order of moves has also been tried: 9 .td2 to provoke 9 . . . tbe8
10 tbe 1 f5 1 1 tbd3 and reach the same position.

. . .

ltJe8 10 tiJdJ f5 1 1 i.d2

1 1 f3 f4 (or ll . h5 12 c5 Wh7 13 cxd6 cxd6 14 tbt2 f4 15 .td2 g5 1 6

llc 1 tbg8 1 7 a4 tbh6 1 8 lbb5 llg8 1 9 h3 .tf8 20 a S .td7 2 1 llc3 :g7 22 b4
a6 with chances for both sides, B .Maksimovic-Kupreichik, Yugoslavia
1 993) 12 g4 h5 1 3 h3 .tf6 14 .td2 Wg7 1 5 .te l l:th8 16 Wg2 with an
unclear position, Gelfand-J.Polgar, Linares 1 994.
. .

ll

. . .

tiJf6

The continuation 1 1 . . . fxe4 is given in the next game.

Premature is 1 1 . . . f4?! because of 12 .ig4 with the exchange of light
squared bishops which is strategically favourable to White as, for instance,
in Kozul-Williams, Istanbul 2000.

Variations after 9 ll:le1 ll:le8 1 0 l:i:ld3 f5 1 1 i..d2 ll:lf6 63

Now, the same position has been reached as in the game Larsen-Tal,
Eersel 1 969 (compare the earlier comments on it in this book, Game 6).
Playable is the waiting method (imitating White's opening tactics) with
.. h8, (see diagram) but White can also
b
1
d e r 8 '
"
c
insist on improving his pieces with 12 Act
8
(less consistent is 1 2 f3 f4 1 3 c5 h5 1 4 Ac l g5 8
7
1 5 cxd6 cxd6 1 6 l:i:lt2 Af6 1 7 h3 l:tg6 1 8 a4 7
6
ll:lg8 1 9 l'l:lb5 l'l:lh6 20 Ac3 a6 2 1 ll:la3 ll:lf6 22 6
5
5
lC!c4 b5 23 l:i:laS g4 24 ll:lc6 1i'e8 25 fxg4 hxg4
4
26 hxg4 l'l:lhxg4 27 ll:lxg4 i.xg4 28 i..xg4 4
3
Axg4 29 Ah3+ g8 30 i. e l i.f8 3 1 i.h4 3
2
iVhS 32 iVe l 1i'h7 33 l:tff3 iVxe4 34 iVxe4 2
lC!xe4 and Black won in 58 moves, Piket- 1
J.Polgar, Monte Carlo 1 994) 12 . . . l:i:lg8 13 f4
a
b
c
d e r 8 h
fxe4 14 lC!xe4 i.. f5 1 5 i.. f3 lLlef6 16 Ae1 l:i:ld7
17 fxe5 lC!xe5 18 ll:lxe5 i.xe5 19 i.. c3 Ae8 20 1i'd2 i.. x e4 21 Axe4 i.xc3
22 1i'xc3+ 1i'f6 23 Af4 iVxc3 24 Axc3 g7 25 c5 Ae7 26 cxd6 cxd6 27
i.g4 Af8 28 Axf8 xf8 29 i.. e6 with the better endgame, Shirov-J.Polgar,
Monte Carlo (blindfold) 1 995.
11.

1 2 f3
a

2
1

1 2 f4

...

An unclear alternative is 1 2 . . . c5 (see

diagram) 13 g4 (if White wants to play the
idea g2-g4 it is better to do it immediately, i.e.
before Black plays . . . f4; if he delays g4 then
Black has the option . . . fxg3 , for example 1 3
Ab l f4 1 4 g4 fxg3 1 5 hxg3 Wd7 1 6 l'l:lt2 lLlh5
1 7 g2 h8 1 8 Ah l ll:lg8 1 9 Wc 1 a6 20
iLe3 ! ? iVf7 2 1 '1i'd2?! ll:lf4+ 22 gxf4 exf4 23
i.. x f4 iVxf4 24 iVxf4 Axf4 25 ll:ld3 Af7 26 f4
i.. d4 27 l:tbfl b5 28 cxb5 i.. x c3 29 bxc3 axb5
30 ll:lc l ll:lf6 and Black obtained a strong

6
5

4
3
2
1

64 Variations after 9 I1J e1 I1Je8 1 0 l:iJd3 f5 1 1 i.. d2 I1Jj6

initiative, Conrad-S.Ivanov, St.Ingbert 1 989) 1 3 f4 1 4 h4 h6 l S :n 'it>h8

1 6 :h2 i.. d 7 17 :bt aS 1 8 b3 b6 19 a3 'it>h7 20 i.. e l l:Ui 21 b4 axb4 22
axb4 11Jc8 23 i.. fl i.. e8 24 :at :ra7 2S :xa7 :xa7 26 'ikbl .td7 27 CiJbS
:as 28 :a2 :xa2 29 'ikxa2 11Je8 30 bxcS bxcS 31 'ifa8 with the initiative,
Kozui-Radj abov, Leon 200 1 .
. .

11

Possible is 1 2 . . . 'it>h8 (see diagram) 13 :ct

c6 (or 1 3 . . . c5 ! ? 1 4 dxc6 I1Jxc6 1 5 i.. e 3 i.. e 6
7
7
1 6 b3 11Jd4 1 7 11Jf4 i.. g 8 1 8 11Jfd5 i.. x d5 1 9
6
exd5 a6 20 i.. d 3 I1Jh5 2 1 'ifd2 :cs 22 :fd l
6
'ika5 2 3 i.. fl CiJf6 2 4 I1Je2 I1Jxe2+ 25 i.. x e2
'ika3 26 i.. b6 I1Jd7 27 i.. f2 a5 28 :c2 I1Jc5
4
4
with good play for Black, Akopian-Smirin,
3
3
Haifa 1 995) 14 i.. e3 f4 l S i.. f2 cS 1 6 b4 !
2
cxb4 1 7 11Jxb4 b6 18 a4 gS 1 9 aS :gs 20
:at bxaS 2 1 I1Jc6 'ifd7 22 cS I1Jg6 23 cxd6
"
b
c
d e f 8 h
.trs 24 :xaS g4 2S fxg4 11Jxg4 26 .txg4
'ifxg4 27 'ifxg4 i.. x g4 28 .tcs :es 29 :n lLlh4 30 liJbS .th3 3 1 :aa2
:gs 32 'it>hl i.. g4 33 h3 i.. d 7 34 :xa7 :g7 3S 11Jc7 :cs 36 l:iJxeS f3 37
11Jxf3 Black resigned, Ehlvest-Van Wely, New York 1 994.
8

1 3 c5
The reader has seen 13 :ct g5 14 c5 I1Jg6 1 5 cxd6 cxd6 16 11Jb5 :n 1 7
11J t2 i.. f8 1 8 'ifc2 i.. d7 1 9 'ifc7 (Korchnoi-Gligoric, Leningrad 1 973) in
comments on the line 9 i.. d2
A cautious plan is 13 g4 g5 1 4 .te l h5 1 5 h3 11Jg6 1 6 'it>g2 :n 1 7 :h i
i.. f8 1 8 .tf2 i.. d 7 1 9 c5 as in Hoeksma-Van Wely, Dutch championship
1 994, and White won after 1 9 . . . 11Jh4+ ! ? 20 .txh4 gxh4 2 1 'ifb3 etc. , in 5 8
moves.

13

. . .

g5 14 cxd6

More precise than 14 :c l , see the game Larsen-Tal, Eersel l 969.

14

. .

cxd6
a

1
a

Variations after 9 t'i'J e i t'i'Je8 1 0 t'i'Jd3 f5 1 1 i.d2 t'i'Jf6 65

1 5 lDfl
Playable is 1 5 l:tc l (the other rook would stand even better here)
1 6 t'i'Jf2 l:tf7 1 7 t'i'Jb5 i.f8 1 8 c2 t'i'Je8 19 a4 h5 20 b3 i.d7 2 1 h3
l:tg7 22 a5 t'i'Jh4 23 d l g4 24 fxg4 a6 25 t'i'Ja3 t'i'Jf6 26 .te l l:tc8 27 t'i'Jc4
hxg4 28 hxg4 e7 29 l:tc3 i.e8 30 l:th3 t'i'Jh7 3 1 i.b4 f6 32 t'i'Jh 1 t'i'Jg6 3 3
t'i'Jg3 t'i'Jg5 34 l:th 1 l:th7 3 5 .l:Ixh7 'it;xh7 3 6 t'i'Jf5 t'i'Jxe4 37 'ifd3 t'i'Jc5 3 8
'ir'h3+ '1i>g8 39 g 5 'ii'd 8 4 0 t'i'Jfxd6 i.xd6 4 1 t'i'Jxd6 Black resigned, Podgaets
Loginov, Tashkent 1 97 3 .
1 5 . . . tllg 6

1 5 . . . h5

b
c
d
e
r g h
Similar is 15 . . . t'i'Jg6 (see diagram) 1 6 'ii'c 2 !
8
8
%If7 1 7 l:tfc 1 a6 (or 1 7 . . . i.f8 1 8 t'i'Jb5 t'i'Je8 1 9
7
a4 h5 20 h3 t'i'Jh4 2 1 l:ta3 a6 22 l:tc3 i.d7 23 7
6
t'i'Ja3 l:tg7 24 'ifd l l:tb8 25 i.e l 'iff6 26 t'i'Jc4 6
b5 27 axb5 axb5 28 t'i'Ja5 'it'g6 29 t'i'Jc6 l:ta8 30 5
l:ta3 l:txa3 3 1 bxa3 t'i'Jf6 32 t'i'Jb8 i.e8 3 3 l:tc8 4
4
and White was quicker, Kozul-A. Srebmic, 3
3
N o va Gorica 1 999) 18 a4 h5 19 h3 g4 20 fxg4 2
2
hxg4 2 1 hxg4 i.f8 22 a5 b5 23 axb6 'it'xb6 24 1
t'i'Ja4 a7 25 i.a5 l:tb8 26 g5 t'i'Jg4 27 i.f3
"
b
c
d
e
f g h
t'i'Je3 28 e2 l:ttb7 29 l:tc6 l:tg7 30 l:tac1 i.d7
3 1 lba6 'it'd4 32 t'i'Jb6 t'i'Jh4 33 t'i'Jxd7 t'i'Jxf3+ 34 xf3 l:txd7 35 i.c3 'ii'c4
36 l:tc6 'ifb5 37 t'i'Jg4 t'i'Jxg4 38 xg4 l:tg7 39 'ii'e6+ h7 40 ..tr>f2 Black
resigned, Kozul-Shchekachev, Linares 1 996.

1 6 h3 l:.fi 17 'ii' c2 l2Jg6

Or 17 ... a6 (see diagram) 18 a4 i.f8 19 a5
(better is 19 l:tfc l and 20 d l ) 19 . . . l:tg7 20
ll:la4 g4 2 1 t'i'Jb6?? (White underestimated
Black's attacking potential, surprisingly he is
n o w lost by force; he should have played 2 1
fxg4 hxg4 22 hxg4 i.xg4 23 t'i'Jxg4 t'i'Jxg4 24
l:ta3 e8 with an unclear position) 2 1 . . . gxh3
22 t'i'Jxa8 l:txg2+ 23 h1 e8 ! (White is
defenceless against . . Ji'g6) 24 l:tg1 'iVg6 25
'ifxc8 t'i'Jxc8 26 t'i'Jxh3 l:tg3 2 7 'it;h2 i.h6
White resigned, Sosonko-Gadj ilu, Pula 1 997.

1 8 l:tfcl a6 1 9 a4 l2Jh4
More passive is 19 ... i.f8 (see diagram) 20
a5 g4 (20 . . . b5?! 2 1 axb6 xb6 22 tll a4 a7
23 i. a5) 2 1 fxg4 hxg4 22 hxg4 b5 23 axb6
'ifxb6 24 t'i'Ja4 "fia7 25 i.a5 l:tb8 26 g5 ! t'i'Jh7
( 2 6 . f3 ? ! 27 i.xf3 t'i'Jg4 28 d2 ! l:th7 29 i.b6
l::!:. x b6 30 llxc8 was superior for White,
Spassov-Danailov, Pamporovo 1 98 1 ) 27 t'i'Jb6
f3 28 i.xf3 l:r.xb6! 29 i.xb6 'iVxb6 30 xeS
. .

8
7

3
2

3
2

4
3

2
a

66 Variations after 9 llJe 1 llJe8 1 0 llJd3 f5 1 1 i.d2 llJf6

llJxg5 3 1 l:lxa6 'ife3 ? (Black' s only chance was 3 1 . . .'ii'xb2 32 'ii'c 2 ! _.d4 3 3
l:la3) 32 l:la3 'iib6 33 'ii'g4 i. h 6 34 l:lc2 <Jo>h7 35 g3 ! and White won,
Rogers-Sznapik, Olympiad, Thessaloniki 1 988.
a

2
a

20 'iid 1
An improvement o n 2 0 a S g4 2 1 fxg4 hxg4 2 2 hxg4 b 5 2 3 axb6 'ifxb6
with an unclear position, Kozul-Arapovic, Porec 1 998.

20 . . . g4 2 1 fxg4 hxg4 22 hxg4 lbh7 23 aS lbg5 24 .t.fl b5 25

axb6 'iix b6 26 lba4 'ii a 7 27 .tas l;:tb7 28 l:ic6 ltab8 29 i.b6 ltxb6
30 lbxb6 i.b7 31 'ii'b 3 lbe6 32 lbd7 .l:.a8 33 1tcxa6 i.xa6 34 ltxa6
'ii x d7 35 dxe6 'ifc8 36 e7+ h7 37 l:xa8 'ifxa8 38 'ii ti Black
resigned.

Variations after 9 ttJe1 ttJe8

1 0 ttJd3 f5 1 1 iLd2 fxe4 12 ttJxe4
Game 1 6

Akopian - Dimitrov
Linares 1 996
1 d4 l2Jf6 2 c4 g6 3 l2Jc3 i.. g7 4 e4 d6 5 l2Jf3 0-0 6 i.. e 2 e5 7 0-0
l2Jc6 8 d5 l2Je7 9 l2Je1 l2Je8 10 l2Jd3 f5 1 1 i.. d 2 fxe4 1 2 l2Jxe4
Black has refrained from the more ambitious retention of pawn tension in
the centre and is now looking for simplification and potential pressure on
White's less well-protected d5-pawn.
a

3
2

12

. . .

c6

Or 12 . . . lt:\f6 (see diagram) 13 .if3 (the

alternative is 1 3 f3 lLifS 14 lLidf2 <ilih8 1 5 g4
lt:\d4 1 6 .ie3 c6 1 7 lt:\xf6 'ii'xf6 1 8 lt:\e4 'fle7
1 9 <ilig2 cxd5 20 cxd5 b6 2 1 .i.d3 .i.b7 22 lt:\c3
ac8 23 .i.e4 c5 with a balanced position,
Sharavdorj-Irzhanov, Singapore 1 995) 13 . . . c6
(if 1 3 . . . lt:\fS 14 .ic3 lt:\h4 1 5 lt:\xf6+ 'flxf6 1 6
.ie4 .if5 1 7 f3 .id7 1 8 f4 lLifS 1 9 fxe5 with a
winning position, Koehler-Seibold, Germany
1 992) 14 dxc6 bxc6 1 5 .i. gS .irs 1 6 .ixf6
i.. xf6 1 7 lLib4 b8 18 'fla4 .id7 19 'fla3 cS

8
7

4
3

3
2

2
1

68 Variations after 9 liJe 1 liJe 8 10 liJd3 f5 1 1 Ji.d2 fxe4 1 2 ltJxe4

20 liJxf6+ .l:[xf6 2 1 liJdS liJxdS 22 Ji.xd5+ Ji.e6 23 Ji.xe6+ .l:[xe6 24 .l:[ad 1

'ill c 7 25 .l:[d2 .l:[f8 26 g3 e4 27 .l:[e1 'illf7 28 'ili'e3 .l:[f6 29 b3 1/z-l/z, Pribyl
F.Portisch, Zalaegerszeg 1 977.
8

6
5

-....u 1
f

Another logical possibility is 12 . . . liJfS (see

diagram) . Black has done quite well with this
continuation in tournament practice. For
example:
13 i.g4 c6 1 4 .l:[c l aS 1 S .l:[e1 cxdS 1 6 cxdS
h6 1 7 i.h3 bS 1 8 f3 i.b7 1 9 liJdf2 liJc7 20
i.xfS gxfS 21 liJg3 lLlxdS 22 liJhS 'illb 6 with
material advantage for Black, Hertneck
Kupreichik, Badenweiler 1 990.

13 i.e3 c6 1 4 .l:[c 1 h8 1 S a4? ! liJc7 1 6 dxc6

bxc6 17 liJb4 i.b7 1 8 aS .l:[b8 19 ..tg4 cS 20
liJdS i.xdS 2 1 cxdS liJd4 and White was again outplayed, Holmsgaard
Kupreichik, Norresundby 1 992.
13 i.c3 c6 14 .l:[e 1 h8 lS i.fl aS 1 6 a4 cxdS 17 cxdS 1i'b6 18 b4 axb4
1 9 i.xb4 liJf6 20 aS 'illd 8 2 1 liJb2 b6 22 liJc4 liJxe4 23 .l:[xe4 bxaS 24 i.xaS
'ir'f6 2S liJb6 'iVb8 26 'ili'f3 liJd4 27 'illx f6 i.xf6 28 liJxc8 .l:[fc8 29 i.e 1 .l:[cS
30 ];[aS .l:[c 1 3 1 f3 liJb3 32 .l:[a2 liJcS 3 3 .l:[ee2 .l:[bb 1 34 .l:[a8+ g7 3 S .l:[a7+
g8 36 .l:[a8+ f7 3 7 .l:[a7+ i.e? White resigned, Barlov-Kupreichik,
Cetinje 1 992.
13 f3 c6 1 4 'it>h 1 cxdS 1S cxdS i.d7 16 g liJd4 1 7 f4 liJf6 1 8 liJxf6+
.l:[xf6 1 9 i.c3 'iVb6 20 fxeS dxeS 2 1 .l:[xf6 'ii'x f6 22 liJc5 liJxe2 23 'illx e2
'i'd6 and Black won a pawn, T.Ivanov-Iskusnyh, Novosibirsk 1 99 S .
a

3
2
1
a

1 3 dxc6 !
Alternatives are less efficient:

Variations after 9 0:Je1 0:Je8 1 0 0:Jd3 f5 1 1 i.. d2 fxe4 1 2 0.xe4 69

1 3 i.g4 (see diagram) ( 1 3 i.g5 cxd5 1 4

, xd5 0Jf6 leads t o a n approximately even
) '.a me) 1 3 . . . i.f5 1 4 i.xf5 gxf5 1 5 0Jg5 0Jc7
I (J dxc6 0Jxc6 1 7 'ii' h 5 h6 1 8 0Jh3 (the knight
1s misplaced here) 1 8 . . . 'ii'e 8 1 9 'ii'x e8 .l:tfxe8
. 0 l:r.fd l l:r.ad8 2 1 0Jb4 0Jxb4 22 i.xb4 d5 23
r xd5 0Jxd5 24 i.e 1 e4 with a clear advantage
1 1 1 1 Black, Klarenbeek-Nijboer, Dutch champ ionship 1 99 5 .

.'1

1 3 'ii'b 3 (see diagram) 1 3 . . . <ifi> h 8 1 4 f4 b5 1 5

r 5 dxc5 1 6 0Jexc5 exf4 1 7 0Jxf4 0Jxd5 1 8

(i )xd5 .l:txfl + 1 9 .l:txfl 'ii'x d5 20 'ii'x d5 cxd5
I i.xb5 i.d4+ 22 <ifi>h 1 i.xc5 23 i.xe8 i.a6
.4 i.c6 i.xfl 25 i.xa8 d4 26 b4 i.d6 27 a4
J;g7 28 <it>g l i.d3 29 g3 <ifi>f6 3 0 <ifi>f2 <it>e6 3 1
J.f3 i.e5 32 b5 i.c4 33 i.c6 i.d5+ 34
.'11L x d5+ <it>xd5 3 5 a5 i.c7 36 .te l <it>c4 3 7
J.>e4 <it>xb5 3 8 i.f2 i.xa5 39 i.xd4 i.b6 4 0
.'11L e 5 <ifi>c4 White resigned, Sosonko-Nijboer,
l l olland 1 99 3 .

8
7

5
4

2
a

13 . . . bxc6 14 i.c3 l'bf5

In case of 1 4 . . . 0Jf6 1 5 0Jxf6+ i.xf6 1 6 c5 0Jd5 1 7 cxd6 0Jxc3 1 8 bxc3

'i!Vxd6 1 9 'i!i'a4 White retains some pressure.

l 5 l:tel !
Not 1 5 0Jxe5? i.xe5 1 6 i.xe5 'i!i'e7 and White would lose a piece.

15 . . . lbc7
White is better after 1 5 . . . 0Jd4 16 i.xd4 exd4 1 7 i.g4 .

l 6 'ii a4 !
8

2
a

70 Variations after 9 lt:lel lt:le8 J O I.Lld3 f5 1 1 d2fxe4 1 2 'Dxe4

1 6 c5
..

After 16 . . . d5 17 cxd5 cxd5 1 8 lt:lg3 White retains his unpleasant pressure

on Black' s central pawns.

1 7 l:ad 1 i.d7 18 'ii a 3 liJe6

a

1
a

1 9 liJxd6 ! liJxd6 20 ltJxe5 'iig 5?

The lesser evil was 20 . . . .txe5 21 xeS liJf5 22 .if3 and White has the
initiative.

2 1 liJf3 ! 'ii f4
If 2 1 . . .'iVe7 22 xg7 xg7 (or 22 . . . lLlxg7 23 'i'xc5) 23 'i'd3 ! lLlf4 24
'ii'xd6 lLlxe2+ 25 fl .

22 g3 'ii'f5 23 i.xg7 ltJxg7 24 l:txd6 liae8 25 'i'd3 i.c8 26 'ii xf5

liJxf5 27 l:.d3 i.b7 28 i.d1 liJd4 29 liJe5
In time pressure White did not go for 29 l:txe8 l:txe8 30 fl .ic8 3 1 Ae3
.ih3+ 32 e 1 with a winning game.

29 a5 30 f4 g5 31 fxg5 l:f5 32 l:de3 l:.xg5 33 liJd3 lbe3 34

lbe3 .l:f5 35 l:r.e8+
..

In a hurry White forgets the immediate 35 l:te5 .

35 . . . fi 36 .l:e5 l:xe5 37 ltJxe5+ e6 38 liJd3 d6 39 f2 i.a6

40 b3 a4 4 1 e3 a3 42 d2 Black resigned.

Variations after 9 ttJe1 ttJe8

1 0 j,e3 f5 1 1 f3
This is one of the crucial lines (and similar to the continuation 9 li)e 1
d7 1 0 .i.e3 f5 1 1 f3 shown in the previous chapter). White puts his dark
squared bishop on its most efficient diagonal and, while exerting strong
pressure on the queenside, simultaneously takes the risk of having to meet
the opponent' s equally rapid attack on the kingside.
Game 1 7
Shirov Uhlmann
Germany 1 994

l d4 tiJf6 2 c4 g6 3 tiJc3 i.g7 4 e4 d6 5 i.e2 0-0 6 tiJO e5 7 0-0

tiJc6 8 d5 tiJe7 9 tiJel tiJe8 10 .t e3 f5 l l 0
a

l l ... c5 ! ?
This i s a less frequently played move to try and block the queenside first.

12 dxc6
White may agree with Black' s conception and continue 12 li)d3 f4 1 3
.i.f2 h 5 1 4 b4 b 6 1 5 a4 a5 ! ? 1 6 bxc5 bxc5 1 7 J:lb 1 g 5 1 8 li)xc5 ! dxc5 1 9
.i.xc5 li)g6 2 0 .i.b6 ! (The passed d-pawn will even compensate for White ' s

72 Variations after 9 &iJe1 &iJe8 1 0 i.e3 f5 1 1 j3

soon being temporarily a rook down ! ) 20 . . . 'iff6 2 1 c5 g4 22 d6 gxf3 23

gxf3 i.h3 24 h 1 i.xfl 25 i.xfl 11r'e6 26 d7 &iJf6 27 d8=1ir' .l:r.fxd8 28
i.. x d8 &iJd7 29 i.. g 5 l:lc8 30 'ir'd5 &iJgf8 3 1 i.e7 &iJxc5 32 i.h3 'ii'x d5 3 3
&iJxd5 Black resigned, Korchnoi-Nijboer, Holland 1 993 .

1 2 . bxc6
.

Inferior is 12 . . . &iJxc6? ! 1 3 &iJc2 .

1 3 'ifd2 i.. e 6 14 l:td 1 W/c7

a

a
c

15 b4
8

1 5 f4 ! (see diagram) with the idea

&iJe 1 -f3 -g5 seems to be more dangerous for
Black. For example:

15 . . . exf4 (if 15 . . . g5 16 fxe5 f4 17 exd6

fxe3 1 8 .l:r.xf8+ i.. x f8 1 9 1lr'd3 White is better)
1 6 i.xf4 fxe4 1 7 &iJxe4 'ii'b6+ 1 8 i.e3 .l:r.xfl +
3
:1
1 9 Wxfl 'i'xb2 20 'i'xb2 i.. xb2 2 1 &iJxd6 a5
2
22 ..tc5 with a slightly better endgame which
1
=a.;.a......a
White won in 39 moves, Korchnoi-Nijboer,
"
b ,.
d c r 8 h
Wijk aan Zee 1 993 . 22 &iJf3 &iJf6 23 &iJg5 is
probably even more promising for White, as in the game Pogorelov
J.Rodriguez, La Coruna 1 996.
1

1 5 . . . fxe4 16 &iJxe4 &iJf5 17 fxe5 i.. x e5 18 &iJd3 i.g7 ( 1 8 . . . &iJxe3 1 9

.l:r.xf8+ Wxf8 2 0 'i'xe3 i.g7 2 1 &iJg5 i.. xc4 2 2 'i'f4+ i.. f6 2 3 &iJxh7+ 'ii'x h7
24 'it'xc4 'irb7 25 &iJf4 with a decisive attack, A.Nikitin-S.Farago, Budapest
1 995) 1 9 &iJf4 &iJxe3 20 'i'xe3 i.f5 2 1 g4 i.. xe4 22 'i'xe4 &iJf6 23 'iie 6+
Wh8 24 g5 &iJd7 25 &iJxg6+ hxg6 26 'i'h3+ Wg8 27 'ti'e6+ Wh8 28 .l:r.d3
.l:r.xfl + 29 i.xfl &iJf8 30 .l:r.h3+ &iJh7 3 1 'ir'xg6 i.d4+ 32 Wh 1 'ti'g7 33 'ii' e4
g8 34 'ife6+ 'ii' f7 3 5 'ii' x f7+ Wxf7 36 .l:r.xh7+ and White had a material
advantage in the endgame, Pogorelov-Kaminski, Zaragoza 1 996.

15

Ild8 16 b5

Variations after 9 &D el &DeB 1 0 i.. e3 f5 1 1 j3 73

White can _play this move after removing his queen from d-file. For
example, 1 6 'ifc2 I:Df6 1 7 b5 &DeS 1 8 bxc6 'i!Vxc6 1 9 I:Dd5 l:.d7 20 a4 a5 2 1
c 5 &Da7 2 2 cxd6 'ii'x c2 2 3 &Dxc2 fxe4 24 I:Dxf6+ .l:txf6 2 5 fxe4 l:.xfl + 26
i.. xfl with a clear advantage for White, Finegold-Aitounian, Las Vegas
1 994.
a

2
a

1 6 . . . d5? !
Better was 1 6 . . . fxe4 ! 1 7 I:Dxe4 (Black is fine after 1 7 fxe4 .l:txfl + 1 8
i.. x fl I:Df6) 1 7 . . . I:Df5 1 8 bxc6 &Dxe3 1 9 'ii'x e3 'ii'x c6 20 &Dc2 a5 ! with
chances for both sides.

17 bxc6 d4 1 8 J.gs dxc3 !

The best chance. Inferior is 1 8 . . . 1:Df6 1 9 I:Dd5 'ii'd6 20 c5 ! 'ii'x c5 2 1 I:Dd3
'ilfd6 22 &Dxe7+ 'ii'x e7 23 &Dxe5 .

19 'iVxd8 'ii x d8 20 l:xd8 lbxc6 2 1 l:.d 1 ! fxe4 ! 22 fxe4 lbd4 23

J.d3 lbd6?
Correct was 23 . . . i.. g4 ! 24 .l:tc 1 i.. e 2 ! 25 .l:txf8+ i.. x f8 26 .l:lxc3 (if 26 'it>f2
i.. x d3 27 &Dxd3 I:Dd6 ! 28 &Dxe5 i.. g 7 ! 29 i.. f4 g5 30 e3 &De6 ! 3 1 i.. g 3
i.. x e5 32 i.. x e5 &Dxc4+ 33 'iti>d3 &Dxe5+ 34 xc3 fl with even chances)
26 . . . i..b4 27 .l:tb3 &Dxb3 28 axb3 i.. c 5+ 29 'iti>h 1 i.. d l ! 30 i.. c 2 ! (if 30 &Dc2
rJ;; fl intending I:Df6) 30 . . . i.. x c2 3 1 &Dxc2 I:Dd6 32 i.. f6 &Dxe4 33 i.. x e5
ctJd2 ! 34 &Da 1 ! with a slight advantage (Shirov).

24 :c t ! l:.c8
White has the better game after 24 . . . i.. xc4 25 .l:txc3 i.. x d3 26 .l:txf8+
i.. x f8 27 .l:txd3 &Dxe4 28 i.. e 3 .

25 l:.f2 .i.f8 26 lbc3 lbxc4

26 . . . I:D6b5 27 .l:tc l i.. a 3 28 .l:tb l &Dc3 29 .l:tb7 i.. c 5 30 fl is in White 's
favour.

74 Variations after 9 li::l e 1 li::l e8 1 0 .i.e3 f5 1 1 j3

27 .txc4 J:.xc4
27 . . . i.xc4 28 li::ld 3 li::lb 5 29 .l:.cc2 li::l d4 30 .l:.cd2 is also better for White.
28 l:.xc4 .txc4 29 .:tb2 ! aS 30 f2 i.a3 3 1 l:tb8+ f7 32 ltb7+
e6
lf 32 . . .<itg8 33 .i.f6 White wins easily.

33 ltxh7 .txa2 34 l:.g7 i.b1 35 :xg6+ d7

If 35 <itf7 36 .l:.a6 .i.b4 37 li::l f3 wins .
. . .

36 lbo lbxo

Bl ck had no good choice: 36 . . . li::l c 6 37 li::ld2 .i.d3 (if 37 . . . .i.c2 38 li::lc4

.i.c5+ 39 .i.e3 .i.xe4 40 .l:.g7+ .i.e7 4 1 .i.c5) 3 8 h4! a4 39 h5 is winning, or
36 . . . .i.xe4 3 7 li::lx e5+ cl;c7 38 .l:.g7+ just the same.

37 gxO a4 38 lta6 i.. c 2 39 h4 i.c5+ 40 i.e3 Black resigned.

Game l 8

Gelfand - Judit Polgar

Dos Hermanas 1 996
1 d4 lbf6 2 liJO g6 3 c4 i.g7 4 lbc3 0-0 5 e4 d6 6 i.e2 e5 7 0-0
lbc6 8 d5 lbe7 9 lbe1 lbe8 10 i. e3 f5 1 1 0 h8
Another less frequently played alternative to the main line 1 1 . . . f4. Black
vacates the g8 square for the activation of his passive knight on e7, while
maintaining flexible pawn tension in the centre and keeping the f4 square
and the h6-c 1 diagonal free for an eventual incursion of his pieces.

1 2 a4

Variations after 9 t'iJe1 t'iJe8 1 0 i. e3 f5 1 1 f3 75

a
c
d e f g h
b
Playable is 1 2 cS (see diagram) 1 2 ... t'iJg8 13
8
cxd6 (or 13 h 1 i.h6 1 4 i.g 1 t'iJgf6 15 t'iJd3 s
7
t'iJh5 1 6 g3 fxe4 1 7 t'iJxe4 t'iJef6 1 8 'ii'b 3 i.f5 7
1 9 t'iJdf2 dxc5 th- Ih , Ikonnikov-Hillarp 6
6
Persson, Berlin 1 997) 13 . . . cxd6 14 a4 (in the s
s
game Fang-S.Farago, Budapest 1 998, White 4
4
treated this line in a manner that deserves 3
3
attention: 1 4 t'iJd3 i.h6 1 5 i.xh6 t'iJxh6 1 6 f4 2
2
t'iJfl 1 7 d2 t'iJf6 1 8 i.f3 i.d7 1 9 l:tae 1 with
the preferable position) 14 . . . i.h6 1 S i.f2
t'iJgf6 1 6 aS t'iJhS 17 t'iJd3 a6? ! 1 8 t'iJa4 t'iJef6
1 9 t'iJb6 l:tb8 20 l:ta4 i.d7 2 1 l:tb4 (2 1 t'iJxd7 was good too; White loses his
grip on c-file, but his solid possession of the light squares guarantees his
king safety and he can put pressure on the b-file instead) 21 . . . i.b5 22 g3
e8 23 hl 1if7 24 l:tg1 l:tbd8 2S l:tg2 l:tfe8 with chances for both sides,
Ziegler-Hillarp Persson, Excelsior Cup 1 997.

1 2 . . . ltJgS 1 3 aS
a
b
c
d e
f g h
In the earlier encounter Gelfand-J.Polgar,
8
Dos Hermanas 1 995, the game continued 13 8
t'iJd3 (see diagram) 13 ...f4 14 i.f2 hS 1 S cS gS 7
7
1 6 cxd6 cxd6 ( 1 6 . . . t'iJxd6 1 7 t'iJc5 ! ) 1 7 t'iJbS 6
6
a6 1 8 t'iJa3 t'iJh6 19 t'iJc4 g4 20 h1 l:tb8 2 1 s
s
l:tcl ! ? l:tg8 (unclear is 2 l . . . g3 22 i.g 1 gxh2 4
4
23 i.a7 l:ta8 24 i.f2) 22 t'iJb6 i.f6 ? ! 3
3
(22 . . . i.d7 is solid, while 22 . . . g3 allows 23
2
l:t x c 8 ! l:txc8 24 t'iJxc8 'ifxc8 (24 . . . gxf2 25 1
'ii'c 2) 25 i.b6) 23 t'iJxc8 l:txc8 24 'ii'b3
a
b
c
d e f g h
(stronger was 24 i.b6 ! 'ii'd 7 25 l:txc8 'ifxc8 26
'irb3 and if 26 . . . h4? ! 27 t'iJ f2 ! with the better
game) 24 . . . l:tb8! 2S i.b6 d7 26 l:tc4 i.d8 27 l:tfc1 t'iJf7 28 t'iJf2 i.h4 29
'ii'd 1 t'iJh6 30 b3? (30 t'iJd3 i.d8 3 1 i.g l ! ?) 30 . . . i.d8! 3 1 i.xd8 l:txd8 32
t'iJd3 'ifh7 33 l:tc8 (time pressure) 33 ...l:td7 34 l:ta8 l:tdg7 3S l:tcc8 g6 36
lLlf2 t'iJf6 37 l:txg8+ t'iJfxg8 38 'ii'c 2 h7 ? ! (3 8 . . . 'ii'g 5 ! ?) 39 'ifc8 'ifgS 40
l:tb8 'ii' h 4 4 1 t'iJd3 'ife7? (4 l . . . gxf3 ! 42 i.xf3 t'iJg4) 42 b4 t'iJf6 43 'We6!
g6 and a draw was agreed, but White missed 44 'ii'x e7 ! l:txe7 45 l:tc8 !
with a clear advantage.

There is another interesting example from

Judit' s practice: 13 cS (see diagram) 13 ... i.h6
1 4 i.f2 t'iJgf6 1 S aS (an idea is to transfer the
rook to the c4 square as in the game Grabliauskas-Sandstrom, Copenhagen 1 99 8 ; after 1 5
t'iJd3 t'iJh5 1 6 a5 lDef6 1 7 l:ta4 e7 1 8 l:tc4
t'iJe8 1 9 b4 a6 20 cxd6 t'iJxd6 2 1 l:tc5 t'iJf6 22
t'iJxe5 ! fxe4 23 t'iJc4 e3 24 i.g3 i.g7 25 l:txc7
White won the tactical battle) l S . . . t'iJhS 1 6 g3
fxe4 1 7 t'iJxe4 t'iJef6 1 8 cxd6 cxd6 1 9 t'iJc3

6
5
4

3
2
1

8
7

76 Variations after 9 &iJe1 &iJe8 1 0 i.e3 f5 1 1 f3

i.h3 20 &iJg2 a6 21 ii.b6 'ile7 22 .l:!:a4 &iJg7 23 .l:!:el 'fkf7 24 g4 i.xg2 25

'itt x g2 .l:!:ac8 26 i.c4 with a complicated position, Korchnoi-Judit Polgar,
Prague 1 99 5 .

1 3 . . . .th6 1 4 .tn ttJgf6

a

2
1

E J.oi):S

12

Ttr

-- --i
- - - - - - - - ,r

, ... 6

.1.%

''jf
v;r


-

ft
/////h
W///h
p"-
- ---&
-g
ft :if{

'

- --
a

1 5 ttJd3

s
7

- -

The possible 1 5 g4 seems to be too provocative in this concrete position.

For example 1 5 . . . fxg4 16 fxg4 'ild7 (better was 1 6 . . . i.f4 ! 17 &iJg2 &iJd7 1 8
&iJxf4 exf4 1 9 i.d4+ &iJe5) 1 7 h3 i.f4 1 8 &iJd3 g5 1 9 'ittg2 &iJg8 20 &iJxf4
gxf4 2 1 i.f3 &iJe7 22 .l:!:h 1 &iJg6 23 'ile2 'ile7 24 h4 .l:!:g8 25 fl i.d7 26
'itt e 1 c5 27 h5 &iJf8 28 i.h4 .l:!:g5 29 h6 &iJg6 30 i.xg5 'ikxg5 3 1 .l:!:h5 'ikd8 32
g5 with a material advantage, Van der Sterren-M.Ivanov, Reykjavik 1 994.

1 5 . . . ttJh5 1 6 c5 ttJef6 1 7 b4
If 17 c6 ! ? a6 ! with counter-chances.

17 . . . a6!
17 . . . 'ile7 would allow 1 8 b5.

1 8 ttJb2 ttJf4 19 ttJc4 We7 20 <it>h1

Playable is 20 cxd6 cxd6 21 &iJb6 .l:!:b8 22 b5 ! ? axb5 23 i.xb5 fxe4 with
mutual chances.
8

6
5

20 . . . .td7 2 1 cxd6 cxd6 22 ttJb6 liad8

23 exf5
The alternative 23 b5 (see diagram) offers
chances for both sides. According to Gelfand
the possible lines are: (i) 23 . . . &iJxe2 24 'ii'x e2
axb5 25 &iJxd7 'iixd7 26 l:ab 1 ; (ii) 23 . . . fxe4 24
fxe4 axb5 25 &iJxd7 l:.xd7 26 i.xb5 .l:!:c7 27
l:.a3 ; (iii) 23 . . . axb5 24 &iJxd7 .l:!:xd7 25 i.xb5
.l:!:c7 26 l:ta3 l:tfc8 27 'ii'a 1 fxe4 28 fxe4 &iJg4 29
i.g 1 &iJh5 ! ? 30 &iJe2 .

Variations after 9 lt:Je1 lt:Je8 1 0 i.e3 f5 1 1 j3 77

23 . . . gxf5
Also playable is 23 . . . i.xf5 24 :e 1 g5.

24 b5 i.xb5 ? !
More promising i s either 2 4 . . . lt:Jxe2 25 'i!Vxe2 ii.xb5 2 6 lt:Jxb5 axb5 27
'Wxb5 :g8 or 24 . . . axb5 25 lt:Jxd7 (25 ..txb5 i.xb5 26 lt:Jxb5 :g8)
25 . . . :xd7 (25 . . . lt:Jxe2 26 'i!Vxe2 'iVxd7 27 :ab 1 with even chances) 26
i.xb5 :c7 with active play.

25 lLlxb5 axb5 26 i.xb5 llg8 27 l:tgl l:.g5

An interesting line is 27 . . . lt:J6h5 28 i.fl (28 :a2 :g6) 28 . . . lt:Jg3+ 29
xg3 :xg3 but not 29 hxg3 'Wg5 30 g4 i.f8 3 1 ii.g3 :g6 32 'iVe 1 :h6+ 3 3
ii.h2 xh2+ 34 'iti>xh2 'ilr'h6+ 3 5 'iti>g3 i.e7 36 'iti> t2 'i!Vh2 and Black wins.

28 i.fl li!dg8 29 g3 liJ4h5 30 lLlc4 ..if8 31 a6 bxa6 32 l:txa6

Black has concentrated all her forces on the kingside but White ' s king is
pretty safe there. White develops an initiative on the queenside.

32 . . . 'iid 7 33 i.e2 .l:5g6 34 l:.a7 'iic 8 35 lLlb6 'iid 8 36 'ii'c 2 lLlg7

37 l:a8 'ike7 38 1iga l lLlth5

a

39 :8a7
More consistent was 39 : 1 a7 'i!Vg5 40 lt:Jd7 h6 (or 40 . . . i.e7 4 1 xg8+
'lftxg8 42 'i!Vc8+ 'iti>fl 43 a8 lt:Jf6 44 lt:Jxe5+ dxe5 45 d6 h6 46 dxe7 and
White wins) 4 1 'iti>g 1 i.e7 42 xg8+ 'iti>xg8 43 a8+ 'iti>fl 44 'iVc8 'iVd2 45
ii.fl with a clear advantage.

39 . . . 'ikg5 40 l:gt lLle8 4 1 i.d3 liJhg7 42 g4

Another possibility was 42 'iVe2 followed by g4 .

42 . . .'tlff4 43 'ii' e 2 lith6 44 i.g3 'it'g5

If 44 . . . 'iVd4 45 :b 1 threatening 46 a4.

78 Variations after 9 tlJe1 tlJe8 1 0 i.. e 3 f5 1 1 j3

45 lbd7 lbf6 46 lbxf8 l:xf8 47 gxf5 'tih5 48 .txe5

48 l:txg7 ! rJ;;x g7 49 i.. x e5+ rJi;h8 50 i.. b 2 was attractive.
48

. .

dxe5 49 l:taxg7 tlJxdS?

Missing the chance for strong resistance by 49 . . . e4 ! 50 fxe4 'ii'x e2 5 1

i..x e2 tlJxe4 52 l:te7 tiJf2+ 5 3 rJi;g2 l:txf5 etc.

50 .tc4 lbf6 5 1 'ii' g 2 'ii' e 8 52 'ii g 5 Black resigned.

Game 1 9

Korchnoi - Kasparov
Debrecen 1 992
1 d4 lbf6 2 c4 g6 3 lbc3 .tg7 4 e4 d6 5 .te2 0-0 6 llJtJ e5 7 0-0
lbc6 8 d5 lbe7 9 lbel lbe8 10 i.e3 f5 1 1 t3 f4 1 2 i.f2
a

h
8

8
7

1 2 ... h5

If 1 2 . . . g5 first, White has the unclear option

1 3 g4 ! , capturing space around his king, but in
7 practice he usually continues straightforward
7
6 action on the queenside by playing 13 cS (see
6
5 diagram) . For example 1J . . . tlJg6 (interesting
5
is 1 3 . . . h5 1 4 a4 l:.f6 I S cxd6 cxd6 1 6 a5 tlJc7
4
1
1 7 tiJd3 l:tg6 1 8 g4 hxg4 1 9 fxg4 l:th6 20 rJi;g2
3
3
1h- th, P.H.Nielsen-Uhlmann, team tournament
2
2
2002 ; or 1 3 . . . l:tf7 1 4 cxd6 cxd6 1 5 a4 tlJg6 1 6
1
...
a5 tiJf6 1 7 rJi;h 1 .i.f8 1 8 tiJb5 g4 1 9 tl:Jxa7

b
c
d e r g h
i..d 7 20 i.. b6 'ii'e 8 2 1 'Wb3 g3 with an unclear
game, Teplitsky-Bacrot, Mennaid Beach Club
1 999) 14 a4 (playable is 14 cxd6 tlJxd6 1 5 tiJd3 l:tf6 1 6 l:tc l h5 17 tlJc5
tiJf8 1 8 tiJb5 lL:Ixb5 19 i.. x bS l:tg6 20 'ife l g4 2 1 i.. h 4 i.. f6 22 .i.xf6 'ifxf6
23 tiJd3 ltg7 24 "ii'a 5 with advantage for White, Ivanchuk-Van Wely,
8

Variations after 9 &[je] &[je8 1 0 j_ eJ f5 1 1 j3 79

Monte Carlo (rapidplay) 1 997; more natural is 14 . . . cxd6 1 5 a4 llf7 1 6 aS

&[jf6 1 7 h I .i.f8 1 8 lDbS .i.d7 1 9 &[jxa7 g4 20 i.b6 'ii'e 8 2 1 'ifh3 llg7 22
&[jbS llc8 with the initiative for the sacrificed pawn, Narciso Dublan
Illescas Cordoba, Palencia 1 999) 14 . . . llf7 15 &[jd3 j_f8 16 a5 llg7 17 a6!
bxa6 ? ! (tougher is 17 . . . b6 although White is better after 1 8 &[jb4 ! dxcS 1 9
&[jc6 'iff6 2 0 &[jbS) 1 8 &[jb4 t:{jf6 1 9 &[jc6 'ii'd 7 2 0 &[jxa7 dxc5 2 1 &[jxc8 ! ?
(2 1 .i.xa6 ! ) 2 1 . . .'ii' x c8 22 ll c l h5? ! 23 t:{ja4 g4 2 4 &[jxc5 g 3 25 .i. e l gxh2+
26 xh2 .i.xc5 27 llxc5 &[jeS 28 llh1 'ii'd 8 29 g1 'ii'g 5 30 llh2 l:tb8 3 1
b4 llb6 3 2 lla5 h 4 3 3 j_xa6 &[jd6 3 4 j_ f} with a decisive advantage for
White, Korchnoi-Van Wely, Antwerp 1 997.

1 3 c5 g5
a

1 4 a4 !

Korchnoi ' s improvement. White simply increases the pressure on the

queenside, avoiding the more direct 14 llc l and 1 4 cxd6. As far as these
older continuations are concerned the following examples are interesting.

b
c d e r g h
14 llcl (see diagram) (White intends 1 5

&[jbS a6 1 6 cxd6 cxd6 1 7 &[jc7 &[jxc7 1 8 .i.b6) 8

8
14 . . . dxc5 ? ! (After this Black gets a strong 7
7
blockading square for his knight. But in gen- 6
6
era! Black has to avoid the opening up of the 5
5
centre. He remains with permanent weak- 4
4
nesses along the c-file and also the eS-pawn. 3
3
A knight on d6 cannot fully compensate for 2
2
the threat of &[jd3-c5-e6 Therefore the more 1
1

-..
common continuations 1 4 . . . llf6 or 1 4 . . _ &[jg6
a
b c
d e f g h
are better options.) 1 5 .i.xc5 llf6 1 6 'ii'b 3
(safer is 1 6 lDd3 .l:r.g6 1 7 t:{jf2) 16 . . . h7 17 &[jd3 .l:r.g6 18 &[jb5 a6 19 t:{ja7
llxa7 20 .i.xa7 b6 21 .l:r.c2 .i.d7 22 llfc1 &[jc8 23 j_b8 &[jcd6 24 j_xc7
&[jxc7 25 llxc7 &[jb5 with counter-chances, Korchnoi-G.Hernandez, Merida
1 996.

80 Variations after 9 tLle 1 tLle8 1 0 i. e3 f5 1 1 j3

"

l>

14 cxd6 (see diagram) 14 ... cxd6 (or

14 . . . lL!xd6 1 5 lL!d3 lL!g6 16 a4 l:tf7 I 7 a5 i.f6
1
7
1 8 lL!c5 lL!f8 1 9 'iib 3 'ii' e 8 20 a6 b6 2 1 lL!b7
6
6 i.e7 22 lL!b5 lL!xb5 23 i.xb5 i.d7 24 i.xd7
5 'it'xd7 25 'it'a4 g4 26 'it'xd7 lL!xd7 27 l:tfc l lL!c5
4 28 lL!xc5 i.xc5 29 i.xc5 bxc5 30 l:txc5 and
White won the endgame, Korchnoi-Relange,
j
Cannes
1 996) 15 a4 (playable is 1 5 'itth l lL!g6
2
1 6 l:tc l i.d7 1 7 a4 i.h6 1 8 lL!d3 [ 1 8 i.b5 was
possible, trading white-squared bishops] with
a
b
c
d
e
f g h
chances for both sides, Korchnoi-J.Polgar,
Vienna Millenium 1 996) 15 . . . lL!g6 ( 1 5 . . . l:tf6 ! ?) 1 6 lL!b5 ( 1 6 l:tc l ) 16 . . . a6
1 7 lL!a3 lL!f6 18 lL!c4 g4 19 'ir'b3 g3 20 hxg3 (or 20 i.b6 'it'e7 2 1 'itt h I ! with
unclear chances, but not 2 1 'iib 4? lL!xd5 ! 22 exd5 'i*'h4 23 h3 .i.xh3 24
gxh3 'ifxh3 25 l:tf2 gxf2+ 26 ..txf2 l:tf5 27 i.fl l:tg5+ 28 i.g2 e4 and Black
wins) 20 . . . fxg3 21 i.xg3 h4 22 ..th2 lL!h5 23 'it'b6 'it'g5 24 lL!d3
(dangerous is 24 lL!xd6 t'l)g3 25 l:tf2 lL!f4 26 i.fl h3 with a strong attack)
24 . . . lL!g3 25 l:tf2 (wrong is 25 i.xg3 hxg3 26 l:tfc l 'ifh5 27 lL!xd6 'i*'h2+
28 'itt fl lL!h4 ! 29 'ii' g l lL!xf3 ! 30 'ifxh2 gxh2 3 1 'iftf2 lL!g l ! winning)
25 . . . lL!xe2+ 26 J:txe2 l:txf3 27 lL!xd6! l:txd3 28 lL!xc8 'itt h 7! 29 lL!d6! lLif4
30 i.xf4 (if 30 l:tc2 h3) 30 . . . exf4 3 1 lL!f5 l:tg8
a
l>
c
d e f g h
32 'it'xb7 f3 33 l:tf2 l:td2 34 'it'a7 (see diagram)
8 34 ... 'itt h 8? (according to Kasparov correct was
8
7 34 . . . l:txf2 35 'ir'xf2 fxg2 36 lL!xg7 l:txg7 37
7
6 'iff5+ 'ifxf5 38 exf5 h3 39 l:td l l:td7 40 l:td3
6
5 'itt g 7 4 1 b4 'iftf6 42 l:txh3 'iftxf5 ! 43 l:th6 l:tb7
4 44 l:txa6 l:txb4 45 'ittx g2 'itt e 5 46 d6 l:td4 and
4
3 Black makes a draw) 35 'ir'e3 ?? (White misses
3

35 l:txd2 ! 'ir'xd2 36 'i*'f2 'i*'xf2+ 37 'iftxf2 fxg2

2
3 8 lL!xg7 'ittx g7 ! [not 38 . . . l:txg7 39 l:td l h3 40
-.....a 1
'itt g l winning] 39 'ittxg2 'iftf6 40 'itth 3 'itt e 5 4 1
d
"
f & h
l:te l l:th8 42 a5 ! and Black cannot save the endgame) (see diagram) 35 . . . 'ifxe3 ? (now Black
had a straight win with 35 . . . i.d4 ! 36 'it'xd4+
d
e
a
b
c
f g h
l:txd4 37 lL!xd4 'ife3 38 lL!xf3 h3 39 <it>fl l:txg2
8
8
40 l:txg2 'ii'x f3+ 4 1 l:tf2 'ifh l +) 36 lL!xe3 i.d4
7 37 l:txd2 i.xe3+ 38 'itt h 2 i.xd2 39 gxf3 l:tg3
7
6
6 40 l:td l i.f4 41 d6 l:txf3+ 42 'iii> g 2 l:tg3+ 43
5 'it.?hl (if 43 'itt f2 i.e3+ 44 'itt e2 i.g5) 43 . . . l:tg8
4
4
44 d7 l:td8 45 'itt g2 'itt g7 46 l:td5 i.c7 47 b4
j 'itt t7 48 a5? ! (a better chance to put up resisj
2 tance was 48 'itt h 3 'itt e 6 49 'ittx h4 l:txd7 50 l:tc5)
48 . . . l:tg8+! 49 'itt h3 i.d8 50 b5 axb5 5 1 l:txb5
1

l:tg3+ 52 'itt h 2 l:ta3 53 l:th5 'itt e 6 54 a6 'itt x d7

,
e d
,.
L>
f g !..1...a

55 l:th6 i.g5 ! 56 l:th5 i.e7 57 l:th6 l:ta4 58 e5

l:ta5 59 l:tb6 l:txe5 60 a7 l:ta5 61 l:tb7+ 'itt e 6 62 'itt h3 l:ta4 White resigned,
Shirov-Kasparov, Olympiad, Manila 1 992.
8

Variations after 9 tiJe1 tiJe8 1 0 .i.. e3 f5 1 1 j3 81

14

tbg6

There are several important alternatives. 14 . . . dxc5 and 14 .l:.f6 are

handled separately in the games Gelfand-Topalov, Wijk aan Zee I 996, and
Shirov-Nunn, Amsterdam I 995, respectively. Also interesting are I 4 . . .'ith8
with the idea . . . tiJg8-h6 and I 4 . . . tiJf6. These moves have been played quite
often in high-level tournaments. There are several instructive examples:
. . .

14 . . .'it h8 (see diagram) 1 5 aS dxc5 ! ?

11
a
e
b
r
c
g
d
( 1 5 . tiJg8 ! ?) 1 6 i.. x c5 tiJd6 (or I 6 . . . .l:.f6 I 7
8
8
tt:lb5 a6 I 8 lll a 3 tiJd6 I 9 tiJd3 .l:.g6 20 i.. xd6
7
7
cxd6 2 I tiJc4 g4 22 tiJb6 .l:.b8 23 'ifa4 tiJg8 24
6
6
.!lac 1 gxf3 25 i.. x f3 i.g4 26 tiJe I i.. f6 27 .l:.c3
5
i.h4 with an unclear position, Korchnoi-Xie
4
Jun, San Francisco I 995) 1 7 tiJd3 (instead of 4
-
this, a more aggressive approach is better, 3
2
2
either by I 7 a6 or I 7 tiJb5 ! , for example: I 7
I
a6 bxa6 I 8 i.xa6 i.d7 I 9 tiJd3 g4 20 h i
tt:lg6 2 I i.g I ll!h4 22 tiJc5 1Wg5? 23 tiJxd7
a
b
c
d
e
f g h
gxf3 24 g3 ! fxg3 25 hxg3 1Wxg3 26 .l:.f2 .l:.f4
27 .l:.h2 'ifg4 28 i.. f2 and White successfully defended his king, Ikonnikov
Gallagher, Geneva I 994; I 7 tiJb5 ! tiJxb5 [ I 7 . . . tiJg8 I 8 tiJxa7 i.d7 I 9 a6]
1 8 i.xb5 .l:.f6 I 9 .l:.f2 tiJg8 20 a6 b6 2 1 i.. a 3 g4 22 .ll c2 i.. d 7 23 i.. x d7
xd7 24 .l:.ac i .l:.c8 25 .l:.c6 .l:.xc6 26 .l:.xc6 g3 27 h3 tiJe7 28 i.xe7 1Wxe7
29 tiJd3 with a strategically winning position, Spraggett-F.Gonzalez,
Manresa I 993.) 17 . . . tiJg8 18 a6 (More enterprising than I 8 b4? ! tiJh6 I 9
l,ijf2 .l:.g8 2 0 .l:.a3 i. 8 2 I h3 b6 ! 2 2 axb6 cxb6 2 3 i.. xd6 i.. x d6 24 tiJb5
i.xb4 25 .l:.xa7 .l:.xa7 26 lll x a7 i.. c 5 27 tiJc6 1i'f6 28 1i'a i tiJfl 29 .l:.c i g4 30
hxg4 hxg4 3 I fxg4 i.. xg4 3 2 i.xg4 i.xf2+ 33 xf2 .l:.xg4 34 .l:.h i + g7 3 5
a3? [White had the chance t o defend with 3 5 1i'a7] 3 5 . . . 1Wg6 36 1i'h3
!!g3 and White is defenceless, Illescas-Topalov, Madrid I 993) 18 . . . bxa6
( 1 8 b6 ! ) 1 9 lll b 4 ll!h6 20 i.xa6 i.xa6 21 .l:.xa6 g4 22 1i'e2 .l:.g8 23 i.. x d6
( 23 .ll fa i ; 23 'it>h 1 ) 2 3 . . . cxd6 2 4 tiJb5 i. f8 2 5 tiJc6 1i'c8 2 6 .l:.xa7 .l:.xa7 27
t/) bxa7 1i'd7 28 lLlb5 tiJfi 29 ll:lc3 tiJg5 30 hl gxf3 31 .l:.xf3 ll:lxf3 32
xf3 h4 33 b4 h3 34 gxh3 1i'h7 35 b5 1i'h4 36 1i'O f3 37 tiJd1 1Wxe4 38
. .

. . .

82 Variations after 9 li:'Je1 li:'Je8 1 0 i.. e3 f5 1 1 j3

li:'Jf2 'ii'e2 39 'ii'x e2 fxe2 40 li:'Jd3 e4 4 1 li:'Je1 i.. h 6 42 l'Je7 i.. d 2 White
resigned, Shirov-J.Polgar, Dos Hermanas 1 99S.
a

1 4 . . . li:'Jf6 (see diagram) (in practice Black has

had poor results with this continuation) 1 5 li:'Jb5
7
7 (or 1 S aS g4 1 6 cxd6 cxd6 1 7 li:'JbS g3 1 8 i.. x a7
6
6 l0g6 1 9 i.. b 6 'ii' e 7 20 l'Jc7 li:'Jh7 2 1 h 1 .l:tb8 22
5
5 .l:tc 1 li:'JgS [22 .. .'ii' h4 23 i.. g 1 ] 23 i.. a7 li:'Jfl 24
4
4 l'Je6 [24 i.. x b8? 'ii'h4 wins] 24 ... b6 2S .l:txc8
3
3 l:tfxc8 26 .i.xb8 .l:txb8 27 a6 with a decisive
2 advantage, Koutsin-Zulfugaryi, Nikolaev 1 99S,
1 and 1S cxd6 cxd6 16 aS l'Jg6 17 li:'JbS g4 1 8
=
l'Jxa7 g3 1 9 i..b 6 'iile 7 20 .l:tc 1 li:'Jh7 2 1 tll x c8
d e f
c
b
g l..oh...il
a
.l:tfxc8 22 .l:txc8+ l:txc8 23 i..b 5' l'Jgf8 24 li:'Jd3
'ii'h4 2S h3 with a strategically winning position for White, Sulava
Steinermayr, Aschach 1 996) 15 ... g4 16 cxd6 cxd6 17 aS g3 ( 1 7 . . . gxf3 1 8
i..x f3 l'Jg4 1 9 i.. x a7 l:txa7 2 0 l'Jxa7 i.. d7 2 1 'iilh 3 'fic7 2 2 li:'Jd3 Wb8 23
'ii'b 6 is very bad for Black, Summerscale-Jackson, Torquay 1 998) 1 8 i.. x a7
l'J g6 19 i.. b 6 We7 20 h1 ! (this prophylactic move is a typical way to neu
tralize the attack on the kingside; the impatient 20 l'Jc7 li:'Jd7 2 1 l'Je6 li:'Jxb6
22 axb6 .l:txa 1 23 'ii'x a 1 i.xe6 24 dxe6 'ifxe6 2S 'ii'a 7 'ii'b 3 26 li:'Jd3 i.. f6 27
Wxb7 i.. d 8 is good for Black, Crawley-Hassan, Tanta City 2000) 20 ... gxh2
21 l'Jc7 l:[b8 22 i.. c 4! (White forces the exchange of the i.. c 8 for the ti:Jc7
via the e6-square, after which Black ' s attack loses its edge) 22 ... h4 23 l'Je6
li:'Jh5 24 xh2 l'Jg3 25 l:tg1 .l:te8 26 li:'Jd3 ! li:'Jf8 27 li:'Jf2! l'Jxe6 28 dxe6
i.xe6 29 'ii'b 3 .i.xc4 30 'ii'x c4+ h7 31 .l:tac l and White penetrated via the
c-file, Reinderman-Van der Wiel, Rotterdam 2000.
8

15 aS
a

2
1

1 5 J.h6 ! ?

...

O r 1 S . . . dxcS? ! 1 6 i..x cS li:'Jd6 1 7 li:'Jd3 .l:t fl 1 8 a 6 b 6 1 9 i.xd6 Wxd6 2 0

li:'JbS 'ii'd 7 2 I li:'Jb4 li:'Jh4 22 .l:tc 1 li:'Jxg2? 23 xg2 'ii'h 3+ 24 lil;>h 1 g4 2S i.. c4

Variations after 9 tlJeJ tlJe8 1 0 i. e3 f5 1 1 f3 83

g3 26 :c2 'i!;lh7 27 d6 l:tf6 28 tlJxc7 and White won, Korchnoi-Nemet,

Switzerland 1 994.

16 b4
Stronger than 16 c6 bxc6 1 7 dxc6 a6 ! . Another possibility is 16 &iJb5 a6
1 7 tlJa3 'i!;lh8 1 8 tlJc4 :g8 19 :a3 &iJf6 20 cxd6 cxd6 2 1 &iJb6 (playable was
2 1 it'b3 ile7 22 it'b4 ! ..tf8 [or 22 . . . :d8 23 &iJb6 :b8 24 h3 with the better
game] 23 &iJb6 :b8 24 :c3 i.h6 25 &iJd3 g4 26 :rei) 2 1 . . .:b8 22 l:tc3 g4 !
23 fxg4 tlJxe4 ! ? 24 l:txc8 l:txc8 25 tlJxc8 fixeS 26 ..tb6 tlJe7 27 gxh5 &iJf6
with unclear chances, Korchnoi-Shirov, Buenos Aires 1 993 .

1 6 . . . h7 1 7 cxd6
Unclear is 1 7 a6 b6 1 8 i.b5 .

1 7 . . . cxd6 1 8 lDb5 g4
After 1 8 . . . a6 1 9 tlJa3 &iJf6 20 l:tc l g4 2 1 i.b6 'fke8 22 l:tc7+ l:tf7 23 tlJc4
White had a somewhat preferable position in Zuger-Gallagher, Switzerland
1 993 . In the present game Kasparov wants to accelerate his attack by a
pawn sacrifice.

19 fxg4 hxg4 20 .i.xg4 .i.xg4 21 'i'xg4 lDf6

a

2
1

22 'i'O

22 ilf5 was tested in the game Zuger-Cvitan, Berne 1 99 3 . There followed

22 . . . f3 ! (22 . . . a6 23 i.b6 ! ile7 24 tlJc7 or 22 . . . tlJxd5 23 'fke6 tlJxb4 24
&iJf3 are better for White) 23 tlJxf3 tlJxd5 24 ..th4 :xf5 25 ..txd8 l:lf4 26
tlJg5+ ..txg5 27 i.xg5 l:txe4 28 tlJxd6 l:tg4 29 i.d2 tlJh4 30 g3 e4 and by
active play Black succeeded in keeping the balance.

22 . . .'ifd7! 23 lDxa7!
Passive is 23 tlJc3 l:tac8 with pressure.

23 . . . lDg4 24 b5! lDxf2 25 'ii x f2 0 !

The right moment for the counterattack-while White i s busy with
consolidation of his queenside.

84 Variations after 9 t:Dei t:De8 10 i.. e 3 f5 1 1 f3

26 lDxf3 'iig4 27 b6
There is no time for 27 .l:.ae i t:Df4 with many threats.

27 . . .'Yi'xe4 28 l:tfel Vixd5 29 Vic2 ! l:.ae8 !

The only good reply.

30 l:.ad l
White rejects the possible 30 .l:.ed i 'ii'e 6 3 I t:Db5 e4 32 t:Dc7 'ii'g4 3 3
t:Dxe8 exf3 34 t:Dxd6 h8 ! with a n extremely sharp position. Both rivals
stopped counting material in their quest for the initiative.

30 . . . Wxa5 31 'i'c7+
Of course not 3 I .l:.xd6? 'ii'x e I + winning.

3 1 . . . .l:.e7 32 'ii' x d6 e4 33 lDh4 llg7

a

34 ltJxg6

If 34 .l:.xe4? 'ii' a 2 ! 35 'ii'd 3 .l:.d8 ! ! 36 "ii'x d8 t:Dxh4 37 .l:.g4 ! .l:.xg4 38 .l:.d7+

i.. g 7 (or 38 . . . .l:.g7 39 .l:.xg7+ xg7 40 'ii'xh4 'ii'h i + 4 I f2 'ii'x b6+ 42
e2 ! with even chances) 39 .l:.xg7+ .l:.xg7 40 'ii'xh4+ with an equal game.

34 . . Jbg6 35 'ii e7+ ..tg7 36 'ikh4+ g8 37 'ii'x e4 l:.xb6!

Weaker is 3 7 . . . 'ihb6+? 3 8 h i .l:.gf6 39 'ii'c4+ h8 40 t:Dc8 .

38 'iid S+ 'ii x d5 39 .:txd5 :b2 ! 40 lin

If 40 t:Db5 i.. h 6 ! .

40 . . . ..th6
Or 40 . . . .l:.a8 4 I t:Db5 .l:.aa2 42 .l:.g5 .l:.a5 43 .l:.ff5 lla i + 43 .l:.fl .l:.a5 with a
draw.

4 1 l:xf8+ xf8 42 l:tf5+ e7 43 lDb5 ..te3+ 44 fl 1h-1h.

The drawing line would be 44 . . . e6 45 .l:.f3 .l:.b i + 46 e2 ..tg i .

Variations after 9 ti:Je1 ti:Je8 1 0 i.. e 3 f5 1 1 j3 85

Game 20

Gelfand - Topalov
Wijk aan Zee 1 996
1 ctJfJ ctJf6 2 c4 g6 3 ctJc3 i.g7 4 e4 d6 5 d4 0-0 6 i.e2 e5 7 0-0
ctJc6 8 d5 ctJe7 9 ctJe1 ctJe8 10 .ie3 f5 1 1 f3 f4 12 i.f2 h5 13 c5 g5
14 a4 dxc5 ? !
More consistent i s 1 4 . . . l:r.f6-see the next game, Shirov-Nunn, Amster
dam 1 99 5 . The idea of the move in the game is to spend a tempo forcing the
opponent' s bishop to the c5 square, which ideally White' s knight would like
to occupy.

1 5 i.xc5 l:tf6 1 6 ctJd3 l:tg6 1 7 i.f2!

a

4
3

4
3

1
a

More ambitious than 1 7 ti:Jf2 ! ?. White vacates the best square c5 for his
knight, and prepares to meet an eventual incursion of the opponent' s queen
by occupying the h4-d8 diagonal first.

17

. . .

g4

If 1 7 . . . ti:Jd6 1 8 ti:Jc5 b6 1 9 ti:Je6 i.. x e6 2 0 dxe6 l:r.xe6 2 1 'i'b3 . Or

i.. f6 1 8 g4 ! hxg4 ( 1 8 . . . fxg3 1 9 .i.xg3) 1 9 fxg4 ti:Jd6 20 ti:Jc5 with
advantage to White.
17.

. .

18 i.h4 a6
The alternative was 1 8 . . . ti:Jd6 1 9 ti:Jc5 'i'f8 20 ti:Jb5 ! ? (or 20 ti:Je6 .i.xe6

2 1 dxe6 i.. f6 22 i..x f6 'iixf6 23 fxg4 with unclear chances) 20 . . . ti:Jxb5 2 1

axb5 .i.f6 2 2 .te l ! with the better game.

19 g3 ! 'i!Vd6

If 1 9 . . . fxg3 20 hxg3 ; or 1 9 . . . gxf3 20 .i.xf3 .i.g4 2 1 .i.xg4 with the

advantage.

20 fxg4 i.xg4 2 1 i.xg4 l:txg4 22 ctJf2 l:g6 23 'ii'x h5 ctJf6 24

'iVe2 ? !

86 Variations after 9 ti:Je1 ti:Je8 1 0 i.. e3 f5 1 1 j3

Simpler was 24 .ixf6 ! itxf6 25 ti:Jg4 'ii'b6+ 26 .l:tt2 .l:tf8 27 .l:tafl , or

24 . . . i..x f6 25 c;a;,h l etc.

24

..

<it>ti! 25 <it>hl .l:h8 26 l::tg l fxg3 27 xg3 ti:Jh5

a

2
1
a

28 ti:Jg4?
White misses the strongest line: 28 ti:Jd3 ! ! .l:txg3 29 'ii'f2 + 'ii'f6 3 0 .l:txg3 ,
or 28 . . . .l:tgh6 29 .l:tafl + e8 30 .l:tt2 with the better chances .

28

. .

<it>e8! 29 ..tf2 ti:Jf4 30 'iVdl 'it'd7 3 1 l:lg3 Ith3 ?

It was Black' s tum to make a mistake missing the win by 3 1 . . . .l:th4 ! 32

ti:Je3 .l:txh2+ 33 xh2 .l:txg3 .
,

32 'ii'f3 ! llxg3 33 hxg3 'i!Vxg4 34 'iixg4 l:xg4 35 gxf4 exf4 36 ltgl

:.xgl+ 37 xgl ti:Jc8 38 ..tc5 ! d7 39 <it>g2 ? !
39 ..t f2 was the precise move, leaving chances for both sides

39

..

ti:Jd6 40 i.xd6 xd6 41 lL'Ie2

41 c;a;,o i.. x c3 42 bxc3 <it>e5 would leave a winning p awn endgame for
Black.

41 . xb2 42 ti:Jxf4 b5 43 ti:Jd3 i.e5

.

If 43 . . . i.. d4 44 axb5 a5 45 <it>f.3 a4 46 <it>e2 a3 4 7 ti:Jc 1 .ib2 (or 4 7 . . . <it>c5

48 ti:Jb3+ <it>c4 49 ti:Jxd4 <it>xd4 50 b6) 48 b6 cxb6 49 ti:Ja2 <it>c5 5 0 d3 b5
5 1 d6! xd6 (or 5 l . . . b4 52 d7 .if6 53 <it>c2 <it>c4 54 ti:Jc l ) 52 <it>c2 with a
draw (Gelfand).

44 axb5 aS! 45 <it>f3 a4 46 e2 a3 47 b6! cxb6 48 ti:Jc l <it>c5 49

d3 ..tf4
If 49 . . . b5 50 d6 ! ! .

5 0 c2 <it>c4 5 1 bl b5 52 d6! i.xd6 53 ct>a2 ct>d4 54 ti:Jb3+

ct>xe4 55 ti:Jcl ct>d4 56 lbb3+ ct>c3 57 ti:Jcl <it>b4 58 lbb3 <it>a4 1/z- 1/z .

Variations after

t:l:Jel t:l:Je8 1 0 J.e3 f5

I 1 j3

87

Game 2 1

Shirov - Nunn
Amsterdam 1 995
1 d4 tLlf6 2 c4 g6 3 tLlc3 i.g7 4 e4 d6 5 i.e2 0-0 6 tLlf3 e5 7 0-0
tLlc6 8 d5 tLle7 9 tLle1 tLle8 10 i.e3 f5 1 1 f3 f4 12 i.f2 h5 13 c5 g5
14 a4 l;lf6
a

2
1
a

This manoeuvre-in what proves to be a 'premature ' direct attack

makes sense only in the double-edged lines with 1 0 J.e3 where Black' s
counterattack has gained in speed because o f the exposed position o f the
aggressive white dark-squared bishop which is subject to the onslaught of
Black 's kingside pawn mass. In practice this continuation has been
popularised by Judit Polgar and John Nunn.

15 a5
Playable is 1 5 cxd6 t:l:Jxd6 (or 1 5 . . . cxd6 16 aS t:l:Jc7 [Black prevents t:i:Jb5
but thereby also wastes precious time; nevertheless 1 6 . . J 1g6 17 ll'lb5 a6 I 8
..ib6 d7 I 9 ll'la7 is also preferable for White] I 7 h i l::th6 1 8 ll'lc2 ll'lg6
1 9 ll'la3 g4 20 fxg4 hxg4 2 1 J.xg4 'ikg5 22 .i xc8 l::tx c8 23 ll'lc4 ll'le8 24
t:i:Jd2 ll'lf6 25 ll'lf3 1i'h5 26 h3 and Black could not strengthen his attack,
Kozul-Fercec, Pula I 993) 16 ll'lb5 ( 1 6 ll'ld3 ll'lg6 1 7 ll'lc5 is another plan)
16 a6 1 7 ll'lxd6 cxd6 and now:
. . .

18 aS (see diagram) 1 8 ... l::tg 6 1 9 h3 ( I 9 b6

8
f8 20 h3 <ifi>h8 2 I lD d3 ll'lg8 22 ll'lf2 ll'lh6 23
7
l:!.c i f6 24 l::t c 3 g8 25 c2 g_4 26 fxg4 J.h4
27 l::tc i J.d7 28 h i hxg4 29 tllxg4 t:l:Jxg4 30 6
..ixg4 l::txg4 3 I hxg g3 32 g_1 1i'xg4 33 5
l:!.c7 l::tg 8 34 l::txd7 Wxd7 35 e2 'ifh7 36 l::tc 7 4
'i'i'h6 37 l:!.c3 l::tg 5 3 8 fl g7 39 l::t c 7+ 'ili>g6 3
40 ..ig i 1i'h 1 4 1 l::td7 J.h2 42 l::txd6+ c1i;g7 43 2
1i'f2 h-h, Nikitin-Motylev, Tula 1 999) 1
1 9 . .'iti>h8 20 l::t a3 ll'lg8 2 1 J.b6 'iff8 22 ll'ld3
.

8
7

6
1

3
2

88 Variations after 9 ti::l e 1 ti::l e8 1 0 Jl.e3 f5 1 1 j3

Jl.f6 23 ti::l f2 ti::l h 6 24 .l:r.c3 1i'g8 25 ii'c2 g4 26 fxg4 Jl.h4 27 .l:r.cl Jl.d7 28
.l:r.c7 .l:r.g7 29 Jl.f3 hxg4 30 ti::l x g4 ti::l x g4 3 1 i.xg4 i.xg4 32 hxg4 .l:r.xg4 33
'ife2 .l:r.f8 34 'iff3 i.g3 35 'iti>n .l:r.h4 3 6 'iti>e2 'ii'g 6 37 .l:r.e7 .l:r.h2 38 .l:r.e6 'ifg5
39 'iti>d3 i.h4 40 .l:r.c2 .l:r.g8 41 .l:r.xd6 .tel 42 .l:r.e6 .l:r.h6 43 .l:r.xe5 ii'g7 44 .l:r.h5
1-0 Grabliauskas-McShane, Roskilde 1 99 8 .
8

"

=.!0-.....a

18 ti::l d 3 (see diagram) 1 8 ... .l:r.g6 19 'ii'b3 g4

20 'ifb6 'iff8 21 .l:r.fc l 'iff6 22 i.el h4 (or
22 . . . 'iti>h7 23 .l:r.c7 gxf3 24 Jl.xf3 i.g4 25 'iti>fl
.l:r.e8 26 'ifxb7 i.xf3 27 gxf3 ii'g5 28 ti::l f2
Shaked-Babula, Las Vegas 1 999) 23 ti::l f2 gxf3
24 i.xf3 h3 25 .l:r.xc8+! (after this dangerous
attacking bishop is eliminated Black will have
problems organizing his kingside attack)
25 . . . .l:r.xc8 26 ti::l x h3 'iti>h8 27 tll f2 ti::l g 8 28
'ifxb7 .l:r.c2 29 b4 i.f8 30 ii'xa6 with a decisive
advantage, Nielsen-Kantsler, Ohrid 200 1 .

1 5 Jlg6 1 6 cxd6 lbxd6

. .

16 . . . cxd6 17 ti::l b 5 a6 ( 1 7 . . . g4? 1 8 ti::l xa7 g3 19 i.b6 ii'd7 20 i.b5 and

Black wins) 1 8 i.b6 'i!Vd7 1 9 'Wb3 (even stronger is 19 ti::l a 7 ! g4 20 .l:r.c 1 )
1 9 . . . g4 2 0 .l:r.c 1 i. f8 2 1 ti::l c 7 ti::l x c7 2 2 .l:r.xc7 with advantage for White,
Odendahl-Kupreichik, Germany 1 992.
a

3
2

17 lbd3
8

t!

4
3

Alternatives are:
1 7 h3 (see diagram) (too cautious) 1 7 . . . h8
18 ti::l d 3 ti::l g 8 1 9 ti::l c S ti::l h 6 ( 1 9 . . . 'ife7 led to
messy play after 20 .l:r.a3 ti::lh6 2 1 ti::lb 5 g4 22
fxg4 hxg4 23 ti::lxd6 cxd6 24 ti::l e6 i.xe6 25
dxe6 f3 in Korchnoi-J.Polgar, Monaco
blindfold 1 994) 20 ti::l e 6 (a standard move
blocking the i.c8 's diagonal and thereby reducing the power of . . . g5-g4; however now it

Variations after 9 ti::J e 1 ti::J e 8 1 0 .i. e3 f5 1 1 j3 89

will be difficult for White to keep Black occupied on the queenside)

20 Ji'e7 (20 . . . .i.xe6 2 1 dxe6 .l:f.xe6 was certainly an option but Black
prefers to keep the tension and aim for . . . g4) 21 'ifb3 .i.f6 22 .l:f.a3 .i.d7 (the
text prepares . . . .l:f.g8 and creates the possibility of . . . c6) 23 g4 (an interesting
attempt to change the nature of the position; Black was ready with . . . .l:f.a8-g8
and . . . g5-g4) 23 ... fxg3 24 .i.xg3 with an extremely sharp position,
Hj artarson- Dj urhuus, Reykjavik 1 996.
_
a
c
d
e
f
g
h
b
1 7 ti::Jb 5 ! ? (see diagram) 1 7 g4 18 ti::J x d6
( 18 "ifb3 ? g3 19 hxg3 fxg3 20 .i.c5 ti::J ef5 2 1 f4 8
8
7
ti::J x e4 22 .i.b4 ti::J ed6 23 .i.xd6 cxd6 24 .i.d3 7
e4 0- 1 Zhu Chunhui-Gabriel, Halle 1 99 5 ; 1 8 6
6
ti::J x a7 .i.d7 ! with compensation) 18 . . . cxd6 1 9 s
.ih4 'iff8 ( 1 9 . . . gxf3 20 .i.xf3 .i.g4 2 1 ti::Jd3 4
4
'ifd7 22 'ifa4 'ifxa4 23 .l:f.xa4 .i.f6 24 .i.xf6 3
3
.l:f.xf6 25 .l:f.b4 .l:f.b8 26 .l:f.c 1 with slightly better l
2
chances for White, Becker-Kupreichik, Eupen 1
1 993) 20 'ifa4 ! .i.f6 2 1 .i.xf6 .l:f.xf6 22 .l:f.c1
a
b
c
d
e
f g h
tLlg6 23 a6 .l:f.ti 24 .i.b5 'ifd8 ! 25 axb7 .i.xb7
26 .i.c6 'irb6+ 27 <Ji>h1 .i.a6 28 .l:f.g1 .l:f.c8 29 .l:f.c2 .l:f.g7 with an unclear
game, Ftacnik-Wojtkiewicz, Budapest 1 993 .
..

..

1 7 . . . g4 1 8 i.h4 'tif8
1 8 . . . g3 is premature because of 1 9 hxg3 fxg3 20 'ife 1 ti::Jc4 2 1 ti::J d 1 .i.h6

22 .i.xg3 and White won quickly in Gutman-Varitski, Rovno 2000.

19 g3 i.f6 20 fxg4 hxg4
a

h
8

8
7

2
1
a

2 1 liJxf4?
A wrong combination. Correct was 2 1 .i.xf6 'ifxf6 22 l:lf2 (Nunn' s
suggestion i s very interesting: 22 gxf4 g 3 23 Ji.h5 ! 'ii'h4 2 4 'i!Ve2 gxh2++ 2 5
'it>h 1 .l:f.g 1 + 2 6 .l:f.xg l hxg 1 ='if+ 27 <Ji>xg 1 .i.h3 28 ti::Jx e5 <Ji>h8 29 f5 ! 'ifg3+
30 <Ji>h 1 'ifxe5 3 1 l'fh2 l'fxh2+ 32 xh2 i.xf5 33 exf5 ti::Je xf5 with an equal

90 Variations a

9 Ci:Je 1 Ci:Je8 1 0 i.. e3 f5 1 1 j3

game) 22 . . f3 2
fl i.. d7 24 Ci:Jc5 .l:.d8 25 'ifb3 a6 26 Ci:Jd 1 i.. c 8 27 Ci:Je3
q;g7 28 .l:.c 1 .l:.f8 with complex play, Kiriakov-Sotnikov, Moscow 1 995.
.

2 1 . .. exf4 22 ltxf4 'ifh6 ! 23 i.xf6

If 23 .l:.xf6 .l:.xf6 24 e5? 'ife3+ wins.

23 . . . .U.xf6 24 l:xf6
If 24 e5 .l:.xf4 25 exd6 .l:.f2 26 xf2 'iVxh2+ 27 e 1 (or 27 fl Ci:Jf5)
27 . . . 'ilxg3+ 28 d2 'ikf4+ (equally good is 28 . . . 'iVxd6) 29 c2 Ci:Jf5 3 0
b 1 Ci:Je3 with a strong initiative.

24 . . . 'ifxf6 25 .txg4 tbg6 26 i.xc8 llxc8 27 'ilig4 l:te8

Not 27 . . . .l:.f8? 28 e5 with counterplay.

28 :n 'ilid4+ 29 g2 g7
a

8
7
6
5
4
3

30 l:tf4 ? !
Or 3 0 h4 lLlc4 3 1 .l:.e 1 .l:.f8 ! 32 h5 l:tf2+ 33 h3 (if 33 h 1 'ii'd2 34
'ii'xg6+ <itt f8 wins) 33 . . . Ci:Jce5 34 'ikd 1 (34 'iig 5 l:.h2+) 34 . . . .l:.d2 with a
mating attack.

30

. .

tbc4 3 1 'ilig5 'ilid2+ 32 h3

Or 32 <li>g 1 'ikc l + 32 'iti>g2 'ikxb2+ 33 g 1 'i'xc3 .

32 . . J:th8+ 33 g4 tbe5+ 34 f5 l:lf8+ 35 e6 tbd3 ! White

resigned.
After 36 .l:.t7+ .l:.xt7 37 'ikxd2 Ci:Jc5 is mate.

Part Three :
Variation 9 etJe 1 etJd7 1 0 f3 f5 1 1 g4
This, so to speak, prophylactic vanat10n has long enjoyed sporadic
popularity. Before Black tightens his grip around the white king by advanc
ing the g-pawn to g3 , White is quick to grab space in front of his king with
g2-g4, thereby reducing the risk of being exposed to a mating attack and
potentially giving him a free hand to increase his queenside initiative
undisturbed. The continuation has both advantages and drawbacks-the
latter being White' s problem of how to successfully maintain the kingside
blockade and not run into trouble from any opening of the h-file. As with
the other variations discussed here, due to the abundance of material we
shall select only a few games which characteristically represent the ideas of
the line.
Game 22

Pinter - Nunn
Olympiad, Thessaloniki 1 988
1 d4 tL!f6 2 c4 g6 3 tLlc3 .i.g7 4 e4 d6 5 i.e2 0-0 6 tL!fJ e5 7 0-0
tLlc6 8 d5 tL!e7 9 tLle1 tLld7 10 f3 f5 1 1 g4 <it>h8
a

1 2 tLlg2
This move shows that White's primary concern is a successful blockade
of the kingside. Two years earl ier, at the Olympiad in Dubai, the same

92 Variations after 9 t'Lle I t'Lld7 I 0 j3 f5

I I g4

opponents played 1 2 .ie3 t'Llg8 1 3 1!Vd2 (preventing the exchange of the

dark-sq ed bishops) -ll._ . a6 14 t'Llg2 f4 1 5 .if2 h5 1 6 gxh5 g5 1 7 h4 .if6
1 8 'if .l:tt7 1 9 h2 .l:th7 W l:th 1 ife8 21 hxg5 'ii'x h5+ 22 .ih4 .ixg5 23
g l'Llf8 24 f2 t'Llg6 25 .ixg5 1!Vxg5 26 .l:txh7+ xh7 27 'ii' h l + g7 28
'if t'Llf6 and Black had some initiative but the game was drawn.
.

lso played is 12 t'Lld3 c5 (or 1 2 . . . t'Llg8 1 3 h 1 aS 14 .l:tg 1 .l:tt7 1 5 _g5 f4

1 6 fl .if8 1 7 b3 .ie7 1 8 t'Llb5 b6 1 9 .ih3 t'Llc5 20 .ixc8 t'Llxd3 2 1 ifxd3
'fix 22 'ii'fl h6 23 gxh6 g5 24 a3 .l:th7 25 .l:ta2 .l:txh6 26 .l:tag2 'ii'd 7 with
an approximately equal game, Vaganian-Uhlmann, Niksic 1 978) 1 3 g2
t'Llg8 14 .ie3 .ih6 1 5 .if2 .ig5 16 'iVe 1 fxg4 1 7 fxg4 t'Lldf6 1 8 h3 h5 1 9
gxh5 t'Llxh5 2 0 .ixh5 gxh5 2 1 'ii' e 2 t'Llf6 2 2 .ie3 .ih4 2 3 h2 'ii'd7 24 .l:tf3
'ifh7 25 .l:tafl .id7 with chances for both sides, Agzamov-Lanka, USSR
1 979.

1 2 ... a5
Preparing to exert pressure on the e4-pawn with the knights on c5 and f6.

13 h4
Or 1 3 h3 t'Llc5 1 4 .ie3 b6 1 5 b3 .id7 ! with a good game, Miladinovic
lstratescu, Olympiad, Elista 1 998.

1 2 ...tbc5 14 i.e3 tbg8 1 5 .l:b 1 i.d7 1 6 b3 b6 1 7 a3 a4 ! 18 b4

tt:lb3
Aiming at the stronghold on d4 . Black has the advantage.

1 9 tLlbS tt:lf6
Exerting pressure on both e4 and g4 .

20 exfS gxfS 2 1 tbc3 e4 ! 22 gS tbhS 23 fxe4 f4 ! ?

I t was simpler to play 2 3 . . . t'Llg3 24 : n t'Llxe2+ 2 5 t'Llxe2 fxe4 with the
better game.

Variations after 9 li:Je 1 li:Jd7 1 0 j3 f5 1 1 g4 93

24 i.d2 lbxd2 25 'iVxd2 'ife8 26 .to lbg3 27 l:tfel ..te5 28 lbe2

lbxe4 ! 29 i.xe4 f3 30 lbef4 fxg2 31 lbxg2 'ir'h5 32 'ifd3 i.g4 33
e3 'fit7 34 'ii'd 2 'it'g7 35 l::ld 3 t7 36 l:tel .:.af8 37 lbe3 l:tf4 38
tiJg2 .l:l4t7 39 lbe3 i.h5 40 .:n ?
The only chance of resistance lay in 40 ltJg2 although after 40 . Af2 4 1
'ii' x f2 Axf2 42 Wxf2 i.d4+ Black would have had a clear advantage.
..

40 ... l::t xfl + 4 1 lbxfl llf4 42 'ii' e l i.d4+ 43 g2 'ife5 44 lbg3

i.g4 ! 45 b5 i.f2 46 'iVxf2 i.h3+ White resigned.
If 47 Wg 1 then 47 . . . 'ika l + wins.
Game 23

Lesiege - Fedorowicz
New York 1 993

1 d4 lbf6 2 c4 g6 3 ctJc3 ..tg7 4 e4 d6 5 ctJf3 0-0 6 ..te2 e5 7 0-0

ctJc6 8 d5 lbe7 9 lbel lbd7 1 0 f3 f5 1 1 g4 h8 1 2 i.e3 lbg8 1 3

ctJg2 f4 ! ?

Black didn't like 1 3 . . . i.h6 1 4 i.xh6 ltJxh6 1 5 gxf5 gxf5 1 6 f4 with

slightly better chances for White.

14 ..tf2 h5! ?
More solid is 1 4 . . . g5 1 5 b4 (If 1 5 h4 h5 1 6 gxh5 ltJdf6 1 7 hxg5 ltJxh5
with a very good game) 1 5 . . . h5 16 h3 Af7 with equal chances.

1 5 g5 !
Stronger than 1 5 h3 hxg4 1 6 hxg4 i.f6 1 7 i.e 1 with an equal position.

15 .. .'i'xg5 16 i.h4 Vi'h6 17 lbb5 g5 18 i.f2 lLldf6 19 ctJxc7 i.h3 !

20 ctJxa8 .:txa8

94 Variations after 9 &De1 &Dd7 1 0 f3 f5 1 1 g4

2 1 hl ?
A wrong move order. The correct one was 2 1 c 5 g4 22 h l ! 'i'g5 2 3 ltg 1
dxc5 24 .1Lxc5 h4 25 b4 with a good game.

2 1 . .. h4 !
Now Black is better.

22 c5 lLlh5 23 gl lLlg3 24 cxd6 lLlf6 25 ..tel g4 26 'ifd3

If 26 ltf2 ltg8 ! 27 iLfl &Dxfl 28 ltxfl (28 xfl g3 also loses) 28 . . . .1Lf8
and Black wins.

26 lLlxfl 27 ..txfl gxf3 28 lLlxh4

..

If 28 'i'xf3 .i.g4 29 'ii'd 3 'ifg6 etc.

28 ... ..txfl 29 'it'xfl

If 29 x fl &Dg4 30 &Dxf3 'ifh3+ 3 1 g l &De3 3 2 lLlh4 .i.f6 33 'ife2 ltg8+
34 'ifilh 1 ltg4 and Black wins.

29 ... lLlxe4 30 'it'xf3 lLJxd6 3 1 hl .if6

Not 3 l . . .e4?? 32 'ii'x f4 ! .

3 2 .ib4
a

8
7
6
5
4
3
2

32 ... ..txh4! 33 .ixd6 'ii'x d6 34 'ifh5+ g7 35 l:gl + .ig3 36 hxg3

'it'xd5+ 37 h2 f3 ! 38 'ifg5+ f7 39 'ii'f5+ e7 40 'ii'h7+ d6 4 1
'ilfb6+ c5 4 2 b4+
If 42 'ife3+ b5 43 ltc 1 lth8+ 44 g l f2+ 45 fl lth l + 46 e2 'ifc4+ ! .

42 xb4 43 l:b l+ a4 44 llb2 e4 45 l:id2 'ii'e 5 White resigned.

..

The next, very interesting game has a different move order, but reaches a
position quite typical of this variation.

Variations after 9 l:i:!.eJ I:i:!.d7 1 0 j3 f5 1 1 g4 95

Game 24

Gelfand - Kasparov
Reggio Emilia 1 991/92
I d4 tLlf6 2 c4 g6 3 tLlc3 i.g7 4 e4 d6 5 i.e2 0-0 6 tLlf3 e5 7 0-0

01c6 8 d5 tLle7 9 tLle1 tLld7 1 0 tLld3 f5 1 1 J.d2 tLlf6 1 2 f3 f4

Black believes in his kingside initiative even if White carries out 1 3 c5

with gain of a tempo in comparison with the matrix game. However later on
White decides to block the kingside.

13 g4 g5 1 4 b4
Worth mentioning is 14 e 1 h5 1 5 h3 l:i:!.g6 16 c5 (or 16 g2 l:tf7 1 7
f2 f8 1 8 b4 l:th7 1 9 l:th 1 e7 20 'iib 3 g7 2 1 c5 .id7 22 l:tac 1 a6 !
with an unclear game, Lukacs-Grivas, Budapest 1 993) 1 6 . . . l:tf7 1 7 cxd6
r xd6 1 8 l:tc 1 .if8 1 9 .i.f2 l:th7 20 'iib 3 hxg4 2 1 hxg4 l:i:!.h4 22 ..t.xh4 l:txh4
23 l:i:!.t2 a6 ! 24 a4 l:[h7 25 g2 d7 with equal chances, Khalifinan-Spasov,
Manila 1 990.

1 4 h5 15 h3 fi
.

This is a quicker way to take over the h-file.

16 i.e1 l:th8 1 7 g2 tLlg6

Or 17 . . . l:th6 18 ..t.2 'i'h8 19 l:th 1 .id7 20 'iib 3 'i'h7 21 l:tag 1 l:th8 22
..,d 1 hxg4 23 hxg4 l:th2+ 24 l:txh2 'ii'x h2+ 25 fl with an unclear position.

18 c5
More solid was 1 8 .if2 a5 1 9 a3 hxg4 20 hxg4 l:i:!.h4+ 2 1 .ixh4 l:[xh4 22
l:th 1 'ii'h 8 with an equal game.

1 8 hxg4 1 9 hxg4
..

1
a

96 Variations after 9 l[je1 l[jd7 1 0 j3 f5 1 1 g4

1 9 ... ll'lh5 !
White would be slightly better after 1 9 . . . l[jh4+ 20 .ixh4 l:hh4 2 1 .l:th l .

20 l:h 1
lf 20 gxh5 l:txh5 (20 . . . .ih3+ 2 1 g l ! , but not 2 1 xh3 l:txh5+ 22 g2
'ifh8 23 .th4 l[jxh4+ 24 f2 l[jg2 ! ! 25 l:th1 l:th2) 2 1 l:th 1 (or 2 1 l[jf2
l[jh4+ ! 22 g 1 'ii' h 8 with the attack) 2 1 . . .l:txh 1 22 <iti>xh 1 'ii'h 8+ 23 <itl>g2
'ii'h 5 24 .tfl g4 with a strong attack.

20 ...ll'lg3 21 xg3 fxg3 22 'ii'd 2 l':th4 23 xg3 d7 24 l:.xh4

gxh4+ 25 Wh2 f6
The chances are approximately equal because Black has good control
the dark squares.

of

26 b5 g5 27 'ii'e l e3
More precise is 27 . . . <iti>g7 ! 28 c6 .ic8 29 a4 a6 etc.

28 c6 c8
Better than 28 . . . bxc6 29 bxc6 .ic8 30 l:tb l .

29 ll'ld1 d4 30 cxb7 i.xb7 3 1 llc l a6 32 'i!Vd2 axb5

Weaker is 32 . . . <iti>g7 33 l[je3 .ixe3 34 'flxe3 axb5 35 l[jb4 etc.

33 ifh6 ll'lf8 34 I!c2 b6 35 ll'le3 'iif6 36 'ii h 5+ 'i!Vg6 37 ll'lf5

ifxh5 38 gxh5
a

38 ... c8

In time pressure Black misses a better continuation: 38 . . . l[jd7 39 f4 exf4

40 l[jxf4 l[jf6.

39 ll'lxh4 ll'lh7 40 f4 exf4 4 1 ll'lxf4 ll'lg5 42 xb5 ll'lxe4 43 a4

1/z-1/z.

Black can play 43 . . .l[jf6 etc.

Variation 9 l2Jd2
This continuation was the result o f a search for a plan t o develop White ' s
queenside initiative a s actively as possible and i t has enj oyed great popular
ity in grandmaster practice. White is not happy with the d3 square for the
king ' s knight and instead wants to send it to its most efficient post-the c4
square.
One of Black' s first ideas was to take the opportunity to exchange the
dark-squared bishop with 9 . . . i.h6, but this makes his kingside vulnerable
and speeds up the development of White ' s queenside after an exchange on
cl.
Game 2 5

Gligoric - Littleton
Zonal, Praia da Rocha 1 969
1 d4 lbf6 2 c4 g6 3 lbc3 g7 4 e4 d6 5 l2Jf3 0-0 6 e2 e5 7 0-0
lbc6 8 d5 lbe7 9 lbd2 h6 10 b4 lbe8
a

1 1 lbb3
In the game Gligoric-Vukic, Saraj evo 1 969, I played the riskier 1 1 c5 f5
1 2 exf5 gxf5 1 3 ll':lc4 i.xc l 1 4 l:lxc l ll':lg6 1 5 i.h5 'ile7 1 6 i.xg6 hxg6 1 7
f4 exf4 1 8 l:lxf4 dxc5 1 9 bxc5 'fixeS+ 2 0 h i 'iie 7 2 1 'ii'd2 ll':ld6 22 ll:lxd6
cxd6 23 l:le l 'iff6 24 l:lf3 i.d7 and won after 25 ll':le4 fxe4 26 l:txf6 l:txf6 27
h3 l:le8 28 'ifd4 etc.

98 Variation 9 1:iJd2

l l ... i.xcl 1 2 'i!Vxc l ! f5 1 3 f4 ttlf6

Black is behind in develoment in a position that is about to be opened. He
accepts a weaker pawn formation but this is the beginning of his defeat.

14 fxe5 dxe5 15 'it'g5 '3;g7 1 6 exf5 ttlxf5 17 i.d3 h6 1 8 'ii'd l ttlg4

19 ..txf5 ..txf5
1 9 . . . gxf5 20 h3 1:iJf6 2 1 .l:.ae 1 e4 22 l:iJd4 etc. is not good either.

20 h3 ttlf6 2 1 l:ael 'ill e 7 22 ttlc5 b6

a

1
a

23 l%xf5 gxf5 24 ttle6+ h7 25 lixe5 'i!Vxb4 26 'ti'd3 ! Black

resigned.
Another alternative for Black is to copy the original method from the
above given variations-9 . . . 1:iJd7 .
During the following game Korchnoi and I looked with completely differ
ent eyes at what was going on. When he resigned, Korchnoi said "I could
not guess any of your moves". This could mean that, in fact, I played badly,
but Korchnoi added that "perhaps this was the reason for my defeat".
Game 26

Korchnoi - Gligoric
Havana 1 969
1 d4 ttlf6 2 c4 g6 3 ttlc3 ..tg7 4 e4 d6 5 ttlf3 0-0 6 .tel e5 7 0-0
ttlc6 8 d5 ttle7 9 ttldl !
Several months earlier, this move was played in the game Taimanov
Tatai, Venice 1 969 (9 l:iJd2 .i.h6 1 0 b4 a5 1 1 bxa5 .l:ba5 12 l:iJb3 i.xc 1 1 3

Variation 9 /:i)d2 99

'ii'x c l l:[a8 1 4 a4 c5 1 5 f4 ! etc.), so even at this moment it was a kind of a

surprise. Of course the information flow at that time was at a much lower
level than the present day.

9 ... d7 1 0 b4 f5

l l b3

White prepares c4-c5, but more energetic is the sacrifice 1 1 c5. Now
accepting the challenge by taking the pawn 1 l . . .dxc5 1 2 bxc5 /:i)xc5 1 3
i.a3 b6 1 4 i.. x c5 bxc5 1 5 /:i)b3 'iti>h8 1 6 'i!fd2 fxe4 1 7 /:i)xc5 /:i)f5 1 8 /:i)5xe4
i.. h6 1 9 'ii'd 3 is slightly better for White, Lputian-Akopian, Erevan 1 98 8 .
A more complex game arises after 1 1 . . ./:i)f6 1 2 f3 f4 (Also bad i s 1 2 . . . a 5
1 3 bxa5 dxc5 1 4 /:i)c4 Geller-Zaitsev, USSR
1 969) 13 /:i)c4 g5 (see diagram)
8

In this position the idea g4 is not efficient 7

7
for White: 14 g4 /:i)g6 (Weaker is 1 4 . . . fxg3 1 5 6
6
hxg3 /:i)h5 1 6 'iti>g2 /:i)g6 1 7 l:[h 1 /:i)h4+ 1 8 5
gxh4 gxh4 1 9 'ifi>2 h3 20 'iti>e3 Sherbakov- 4
4
Motylev, Russia 1 997) 1 5 i.. d2 l:[f7 1 6 i.e 1 3
3
i.f8 1 7 c6 b6 1 8 b5 h5 1 9 h3 i.. e 7 20 .l:[2 2
2
'ii' f8 with good play for Black, Ulibin-Hillarp
Persson, Sweden 1 99 8 . In practice White

b
c
d e f g h
players have to an equal extent increased the
pressure on both the queenside and the centre with the continuations 1 4
i.. a 3 and 1 4 a4 .

14 i.. a 3 /:i)g6 15 b5 /:i)e8 (Stronger than 1 5 . . . dxc5 although with this con
tinuation too Black players have been successful in practice. For example
1 6 .i.xc5 l:[f7 1 7 a4 h5 1 8 a5 g4 19 b6 [ 19 i.. 2 i.. f8 20 'ii'd2 g3 2 1 hxg3
l:[g7 ! with counterplay, Lputian-Gadj ily, Dubai 2000 (0- 1 , 37)] 1 9 . . . g3 20
'iti>h 1 /:i)h7 2 1 d6 'iih4 22 i.. g 1 i.h3 23 bxc7? [23 gxh3 ! 'ilfxh3 24 .l:[2 gx2
25 .i.x2] 23 . . . .i.xg2+ 24 'it.?xg2 'it'h3+ 25 xh3 lLlg5+ 26 g2 /:i)h4+ 0- 1
Ftacnik-Cvitan, Germany 1 997) 1 6 b6! cxb6 ! ? ( 1 6 . . . axb6 1 7 cxb6 cxb6 1 8
'ii'b 3 Shirov-Akopian, Daugavpils 1 989) 1 7 /:i)xd6? ! (Better is 1 7 cxd6 ! ?

1 00 Variation 9 0.d2

or 1 7 cxb6 ! ?) 17 ... bxc5 1 8 0.xe8 l:txe8 19 b5 d7 20 d6 b6! 2 1 'j!Vd5+

'iti>h8 22 l:tad 1 ? (22 l:tfd l ) 22 ... a6 and Black slowly pushed back the white
pieces and successfully carried out his plan, . . . 0.f8-e6-d4, Vaganian-Shirov,
Manila 1 990.
14 a4 (Probably the most efficient way for White) 14 ...0.g6 1 5 a3
(More promising than 1 5 cxd6 cxd6 1 6 0.b5 0.e8 1 7 d2 l:[f7 [ 1 7 . . . l:tf6 1 8
.te l 0.f8 1 9 .i. 2 l:th6 20 0.xa7 l:txa7 2 1 xa7 0.f6 22 0.b6 Gleizerov
Sulava, Cappelle Ia Grande 1 998] 1 8 e 1 ..id7 1 9 0.a5 8 20 ..i2 f6
2 1 3 ..id8 22 l:tfc 1 ..ixa5 23 bxa5 h5 24 'ii'd l 'ilfd8 25 a6 bxa6 26 0.xa7
g4 27 fxg4 f3 28 xf3 hxg4 29 xg4 l::t x 2 30 0.c6 'ili'g5 3 1 'it>x2 xg4
with a strong attack, Pelletier-Maiwald, Germany 2000) 15 . . . .l:r.fi 16 b5 0.e8
( 1 6 . . . ..if8 1 7 b6 cxb6 1 8 cxd6 [ 1 8 0.xd6?! l:tg7 ! + Epishin-Akopian, Minsk
1 990] 1 8 . . . 0.e8 1 9 0.b5 ..id7 20 'ili'b3 ..ixb5 2 1 axb5 0.xd6 22 l:tfc l is
superior for White, Ftacnik-Gross, Slovakia 1 998) 1 7 a5 ..if8 18 0.a4 h5
(or 1 8 . . . dxc5 1 9 0.xc5 ..ixc5+ 20 ..ixc5 0.d6 2 1 b6 axb6 22 axb6 l:txa l 23
'ii'x a l 0.xc4 24 xc4 cxb6 25 2 'it>g7 26 l:tb 1 !.Farago-Tepper,
Rimavska Sobota 1 974) 19 b6 d7 20 bxc7 'j!Vxc7 21 a6! bxa6 22 c6 and
the strong passed c6 pawn paralyzed Black's position, Beliavsky-Solak,
Saint Vincent 2000.

On 1 1 f3 Black has a good response in

l l ... h6 (see diagram) with an exchange of
8
8
bishops in circumstances more favourable
7 than in the above given games. Nevertheless
7
6
6 after 1 2 0.b3 ..ixc 1 1 3 l:txc 1 f4 1 4 c5 0.f6 1 5
5 0.b5 0.e8 1 6 c2 d7 1 7 0.a5 a6 1 8 0.xb7
5
4 8 1 9 c6 c8 20 0.a3 ..ixb7 2 1 cxb7 'ii'x b7
3 22 c3 g5 23 0.c4 0.g6 24 0.a5 'ili'b6+ 25
3
2 .l:r.2 0.f6 26 a3 g4 27 xc7 'ilfe3 28 l:tc3 l:tac8
29 'ilfxc8 l:txc8 30 l:txe3 fxe3 3 1 l:tfl l:tc2 32
1
..a
l:te 1 li:Jf4 33 'it>fl g3 34 hxg3 li:J6h5 Black had
to defend resourcefully to obtain a draw in
Portisch-Gligoric, Hastings 1 969/70.

l l . . . fxe4 !
The motive for this move lies in the fact that one of the white knights has
travelled a long way from the centre and the e4 square. The exclamation
mark relates more to the fact that the opponent 'did not understand ' the
move than to giving any advantage to Black. In any case, the exchange in
the centre decreases White 's pressure on the queenside, because the pawn
on d5 is now not so solid, and at the same time it frees the f5 square for the
activation of the passive knight on e7.
In fact, I applied a similar idea against Korchnoi at the Interzonal tourna
ment in Sousse 1 967, after 9 li:Je 1 tt:\d7 1 0 d2 f5 1 1 l:tc 1 (White expected
1 l . . .li:Jf6 1 2 f3 f4 when he could play 1 3 c5 gaining a tempo without the
preparatory b2-b4, as in Larsen-Najdorf, Santa Monica 1 966). I played
1 l . . . fxe4 ! 12 li:Jxe4 li:Jf6 1 3 .i.d3 li:Jf5 14 .i.g5 c5 ! 1 5 .l:r.c3 'ilfc7 16 'ilfd2
li:Jd4 and a draw was agreed here, but Black's position is really very good.

Variation 9 CDd2 1 01

1 2 tbxe4 tbf6
a

13 tbg3 ! ?

This was the first move i n the game on which White spent any amount of
time. It was more natural to play 1 3 i.f3, intending to maintain the strong
hold on e4. Perhaps Korchnoi remembered our game in Sousse where he
had not achieved much with such a strategy and therefore preferred to try
something else in order to surprise his opponent. But th ' surprise' is not
unpleasant, because the manoeuvre actually enhances Black's development.
White probably wanted to free some space and remove the tension in the
centre so as to gain a free hand on the queenside. Perhaps he also reckoned
that it was better to exchange the second black knight, which was threaten
ing to become active.

13 ... tbf5 14 tbxf5 i.xf5

This may be another move that White "could not guess" and whose
strength he underestimated. But Black estimates that it is better to get
developed without weak squares than to form a ' strong ' pawn mass with
1 4 . . gxf5 which White can immobilize with 1 5 f4 and make vulnerable to
the pressure of the white pieces.
.

15 i.e3 h5!
During the game this move too was assessed as weak by White, who sim
ply took it as a weakening of the black kingside. In fact, the move enables
the black bishop to take up a strong position on f5 , and, more importantly,
threatens 1 6 tl:Jg4, which provokes White to start moving one of his pawns
on the kingside.
. . .

16 f3
The move seems necessary, but it also makes the kingside (inaccessible
while all the pawns were on the second rank) liable to be opened up-and
Black's whole strategy is actually based on the sharpening of the game on
this part of the board and in the centre. The pawn on f3 is the target of the
. . . e5-e4 breakthrough.

1 02 Variation 9 ti:ld2

1 6 'iVd7 1 7 'ii d 2
.

White provides a better protection of his pieces along the e-file, because
Black can ignore the aggressive 1 7 Ci:la5 and reply with 1 7 . e4.
.

1 7 ... b6
On the direct 1 7 :ae8 Black has to reckon on 1 8 .i.xa 7 b6 1 9 c5. The
preventive move in the game restricts the activity of the white knight and
the dark-squared bishop.
...

The alternative was to attack the white centre with 17 c6, but this would
be more favourable for White since his pieces are well-placed for exerting
pressure along the queenside files.
...

18 a4 ltae8
While White plays on the queenside and attacks over there, Black is com
pletely ready to react with a counter-action in the centre which is the more
important part of the board.

1 9 aS
On 1 9 c5, threatening 20 i.. b 5, Black can easily respond with 1 9 .'Wt7.
..

1 9 ... e4
This is the main point of Black's strategy. The number of squares through
which the black pieces can penetrate to the kingside increases more than
enough to neutralize the pressure that White exerts on the queenside.

20 axb6
20 f4 Ci:lg4 2 1 i.. xg4 hxg4 ! also provides a good game for Black.

20 ... exf3
Black exploits his chances to open the position in the centre of the board
as quickly as possible.

21 .i.xf3 axb6 22 l:.a7

a

Variation 9 ltJd2 1 03

White constantly plays under the wrong impression that he is the one who
has the initiative. Instead of this move, he should have tried to maintain the
balance in the centre, which is more important for the outcome of the battle.
But after the better 22 ltJd4 ltJe4 23 .ixe4 xe4 or 23 'iic l .ig4 24 .ixg4
xfl + 25 'iix fl 'iixg4 White is the one who has to defend himself, which is
the result of his incorrect conception, starting with the 1 3th move, as well as
of his underestimation of Black's counter-chances in the centre.

22 ... ltJe4 ! 23 'iic l

After 23 .ixe4 .ixe4 2 4 xf8+ xf8 25 .ixb6 White would b e exposed
to a destructive attack by 25 . . . 'iig4 and 26 . . . .ixg2 ! .

23 i.g4!
..

Black doesn 't lose a single moment on the threat 24 .ixb6, because just
one passive move could reduce the impact of his attack on the white king.

24 i.xg4
Obviously 24 .ixe4 xfl + 25 'iixfl xe4 26 .ixb6 'iie 8 27 .i2 .ie2
cannot appeal to White, nor can 24 .ixb6 .ixf3 25 gxf3 xf3 ! .

24 ...'ifxg4 25 l!xc7
It would be too late now for White even if he had realized the impending
danger. 25 h3 'iig 3 26 .if4 was no good because the knight on b3 is
hanging.
a

1
a

This is the climax in the Rashomon story-where every side has its own,
different truth. White believes he is winning, when he is, actually, lost! He
took the pawn on c7 quickly enough, thinking that the direct threat 25 . . . ltJc3
was not dangerous because of 26 xg7+ and that Black's whole action was
merely a reflection of his despair.

25 ... i.e5!
Only now, when the white rook has become useless on the vacated
seventh rank, did White exactly appreciate his predicament. There is no

1 04 Variation 9 &i::ld2

adequate defence _g ainst the threat of 26 . . . .txh2+. For instance, 26 &i::ld4

i.xh2+ 27 'itxh2 Wh4+ 28 g 1 &i::lg 3 29 &i::l f3 &i::l e 2 mate, or 29 l:tf3 (29
l:txf8+ l:txf8 30 &i::l f3 leads to a similar end for White) 29 . . . 'ii'h l + 30 cJilt2
&i::le4+ 3 1 'ite2 'it'xg2+ and Black has an unstoppable attack.

26 i.d4
White plays the only move to prevent the sacrifice on h2. But Black has
other ways as well to continue his deadly attack.

26 ... .l:.xfl+ 27 'ii' x fl ilf8 28 'ifa1

Everything loses ! For instance: 28 'ii' c 1 i.xd4+ 29 &i::lxd4 .1:1f2 30 g3 'ii' h 3 ,
or 28 'ii'd 3 Wf4 and further as in the game.

28 .. .'ii f4 ! 29 g3
29 i.xe5 is no good because of 29 . . . 'ii' f2 + and mate on the back rank.

29 ... tDxg3 White resigned.

Looking for a sound way to first neutralize White' s pressure on the
queenside, Black players turned to the blockading move 9 . . . c5.
Game 27

Larsen - Fischer
Game Four, Candidates match, Denver 1 9 71
1 c4 g6 2 ttJO i.g7 3 d4 ttJf6 4 tDc3 0-0 5 e4 d6 6 i.e2 e5 7 0-0
ttJc6 8 d5 tDe7 9 ttJd2 c5
This move has additional justification because the knight on d2 has
obstructed the d-file, so Black doesn ' t have to worry about pressure on the
d5 and d6 squares.
a

10 .l:.b 1 !

2
a

White doesn't lose a tempo on the preparatory 1 0 a3 . Not dangerous is 1 0

dxc6 bxc6 1 1 b4 d5 ! 1 2 .ta3 a6 1 3 'ii'c 2 dxe4 1 4 &i::ldxe4 &i::l f5 1 5 &i::l xf6+

Variation 9 !Cd2 1 05

'iWxf6 1 6 .l:.ae 1 !Cd4 1 7 'ifh 1 !Cxe2+ 1 8 .l:.xe2 i.f5 Sofrevski-Reshevsky,

Skopje 1 976.

1 0 ... ttJe8
This is better than 10 . . . a5 ! ? (Black should not tamper too much with the
side where he is weaker! ) 1 1 a3 !Ce8 12 b4 axb4 1 3 axb4 b6 14 bxc5 bxc5
1 5 !Cb3 f5 16 f3 h8 1 7 i.g5 f4 1 8 .l:.a l .l:.b8 19 i.h4 ..tf6 20 i.e ! g5 2 1
!CbS .l:.g8 22 .l:.a7 .l:.b7 2 3 .l:.a8 'iWd7 24 !CaS and the two open files on the
queenside made the penetration of the white pieces easier and this led to
victory on the 45th move, Gligoric-Zuckerman, Los Angeles 1 974.

1 1 b4 b6 1 2 a4
No better is 12 'ir'a4 f5 13 bxc5 bxc5 14 !Cb3 fxe4 15 !Cxe4 !Cf6 16 f3
!Cxe4 1 7 fxe4 .l:.xfl + 1 8 ..txfl h8 1 9 'ii' a 5 'iWd7 20 ..te2 !Cg8 I_h - I.h ,
Gligoric-Ciocaltea, Hastings 1 97 1 172.
Or 12 bxc5 bxc5 13 !Cb3 'it>h8 ! 14 i.d2 f5 15 !CbS ! ? fxe4 ! 16 ..ta5 'ifd7
1 7 !Cd2 'ii' f5 ! 1 8 !Cc3 !Cf6 with equal play, Shirov-Lanka, Torey 1 990.

1 2 ... f5 1 3 aS ttJf6 1 4 'ifa4 i.d7 15 'ii' a 3 i.h6 ! 1 6 i.. d 3 'flc7

Black is holding the position tightly and now White loses patience.

1 7 bxc5 bxc5 18 exf5 ! ? gxf5 19 i.c2 a6

!Cc3-b5 is a potentially unpleasant threat.

20 ttJde4 i.xcl 2 1 ttJxf6+ lixf6 22 l:.fxcl I;taf8 23 llb6 .tc8

The simplification resulting from the exchanges has given White nothing
special on the queen ' s flank-and only potential worries on the other side
of the board.

24 ttJe2 f4 !
Tightening the noose around the white king.

25 i.e4 ttJf5 26 l:c6 "illg 7 27 l:tb 1 ttJh4 ! 28 'ii'd 3 .tf5!

2
a

1 06 Variation 9 li)d2

Removing the piece that protects the g2 and f3 squares.

29 h1 f3! 30 lL!g3 fxg2+ 31 <tt>g 1 i.xe4 32 'i/Vxe4 lL!f3+ 33 xg2

lLld2 White resigned.
As recent practice shows, Black players have decided on 9 . . . a5 as the best
solution, and this is by far the most frequently played move:
Game 2 8

Van den Sterren - Gelfand

Munich 1 994
1 d4 lLlf6 2 c4 g6 3 lL!c3 i.g7 4 e4 d6 5 lL!f3 0-0 6 i.e2 e5 7 0-0
lL!c6 8 d5 lL!e7 9 lL!d2 a5
White' s intention is b2-b4, c4-c5, li)d2-c4 . With his pawn foray on the
queenside Black tries to slow White' s plan down as much as possible.

1 0 a3 lL!d7

Black tries to activate the pawn mass on the kingside as quickly as poss
ible. Instead of this frequently played move perhaps more precise is
10 . . . d7 with the idea of gaining a tempo for development by threatening a
blockade with . . . a5-a4 (see the examples below). In case of 1 0 . . . li)e8 1 1
l:tb l f5 1 2 b4 axb4 1 3 axb4 li)f6 1 4 c5 h8 1 5 f3 f4 1 6 li)c4 li)e8 1 7 li)a4 !
l:tb8 1 8 b5 b6 1 9 cxb6 cxb6 20 i.d2 d7 2 1 lL!cxb6! White is clearly
superior, Sakaev-Tseitlin, St. Petersburg 1 997.

1 1 l:.b 1 f5 1 2 b4 'it>h8
Passive is 12 . . . b6 13 f3 f4 1 4 li)a4 axb4 1 5 axb4 g5 1 6 c5 li)f6 1 7 cxd6
cxd6 1 8 b5 d7 1 9 li)c4 li)c8 20 .ta3 li)e8 2 1 g4 ! fxg3 22 hxg3 g4 ! 23
.te l ! gxf3 24 .txf3 li)f6 25 .tg5 ! with the better game, Kasparov-Smirin,
USSR 1 98 8 .

Variation 9 lDd2 1 0 7

13 f3

If 13 'ii'c 2 (see diagram) 13 ... lDf6

7
7
( 1 3 . . . lDg8 [ 1 3 . . . b6 1 4 lDb3 axb4 1 5 axb4
6
6
fxe4 1 6 lDxe4 lDf6 1 7 i.d3 lDxe4 1 8 i.xe4
5
5
lDf5 1 9 'ii'd 3 'ifh4 20 g3 'iff6 2 1 f3 i.d7
4
4
with chances for both sides, Gavrikov-Kas3
3
earov, USSR 1 988, or 1 3 . . . axb4 1 4 axb4
2
2
ttJf6 1 5 f3 g5 ! ?-a new conception- 1 6 c5
lD g6 1 7 lDc4 lDf4 with a good game, Vaga
a
b
c
d
e
f g h
nian-Kasparov, Olympiad, Manila 1 992] 1 4
exf5 gxf5 1 5 f4 axb4 [or 1 5 . . . lDe7 1 6 lDf3
e4 1 7 lDe 1 axb4 1 8 axb4 c6 1 9 i.e3 with the preferable chances, L_putian
Van Wely, Helsinki 1 992] 1 6 axb4 exf4 1 7 lDf3 lDdf6 ! 1 8 i.xf4 ttJe4 1 9
lDxe4 fxe4 2 0 lDd2 i. f5 2 1 lDxe4 'ife8 22 i.d3 :a3 2 3 :be l 'ii'g 6 24 i.c 1
:xd3 25 'ii'x d3 lDf6 26 :xf5 'ifxf5 27 lDc5 'ifxd3 28 lDxd3 b5 29 c5 1/2-1/2,
Lputian-Piket, Olympiad, Manila 1 992.) 14 i.b2 (or 1 4 f3 axb4 1 5 axb4 c6
I 6 lDb3 fxe4 1 7 fxe4 cxd5 1 8 cxd5 'iib 6+ 1 9 'ifi>h 1 i.d7 20 lDa5? lDg4 ! with
countefJ2lay, Lputian-Piket, Wij k aan Zee 2000) 14 ... axb4 15 axb4 fxe4 1 6
lDcxe4 ttJxe4 17 lDxe4 lDf5 18 :a1 :xa1 19 :xa 1 lDd4 20 i.xd4 exd4 2 1
:as 'ii'e 7 22 i.d3 i.f5 2 3 :xrs+ i.xf8 24 g 3 i.xe4 25 i.xe4 b 6 26 h 4 with
a permanent initiative for White due to his better bishop and kingside pawn
maj ority, Gelfand-Ye, FIDE world championship 2000.
a

2
a

1 3 ttJg8
...

Another continuation is 13 ... f4 14 lDa4 (if 1 4 lDb3 ? ! axb4 1 5 axb4 g5 1 6

i.d2?! lDg6 1 7 :a i :xa 1 1 8 'ifxa 1 lDf6 1 9 'ifa7 g4 20 fxg4 lDxg4 2 1 h3
lDh6 22 i.e 1 :g8 23 lDd2 .tf6 ! with an attack, Ljubojevic-Kasparov,
Linares 1 993) 14 axb4 15 axb4 c6! 16 c5 ! ? cxd5 17 cxd6 lDc6 18 exd5
lDd4 Salov-Nunn, Rotterdam 1 989.
..

An interesting plan, starting from the 1 51h move, is 13 ... axb4 14 axb4 c6

15 :n (instead of the standard 1 5 'ith 1 lDf6 1 6 lDb3 cxd5 1 7 cxd5 f4 1 8

lDa5 g5 1 9 lDc4 lDg6 20 lDb5 :a6 2 1 i.d2 :g8 22 :c 1 i.f8 23 'ifc2 i.e7

1 08 Variation 9 ti:Jd2

24 tl::J a 5 g4 25 tl::J c 7 gx3 26 .i.x3 .lia7 27 tl::J e 6 .txe6 28 dxe6 ti:::J h4 29 .te l
ti:::J x3 30 gx3 .lia8 with unclear chances, Chuchelov-Nataf, France 2000)
15 ti:::J f6 16 ti:::J fl cxd5 17 cxd5 f4 18 .i.d2 g5 19 .tel .lig8 20 .lial .lixal 2 1
'ii'x al g 4 22 'i'a5 b 6 2 3 'i'a7 .t f8 24 .i d l tl::J g 6 25 .lic2 gxf3 26 .i.xf3 .i.g4
27 ti:::J d 2 'i'c8 28 'i'a2 'ii' d 7 29 'ii'c 4 'ii'g7 30 ti:::Jb 5 'ii'h 6 31 'ii'e 2 tl::J h 4 32
.txh4 'ii'x h4 33 .lic6 i-d7 34 .lixb6 tl::J g4 35 ti:::J fl 'ii'd 8 36 .lib7 'ii'c 8 37 .lic7
'ii'b8 38 .lixd7 'ii'b 6+ 39 tl::J e3 tl::J x e3 40 hl .lig7 41 .lixg7 xg7 42 g3
i.. e7 43 gxf4 exf4 44 tl::J c3 'ii'x b4 45 ti:::J d l tl::J x dl 46 'ii'x dl 'ii'c 5 47 .tg2
'ii'f2 48 'Wg4+ 1/2-1/2, Beliavsky-Antic, Yugoslavia 200 1 .
.

1 4 'ifc2
a

8
7
6

This is the standard position, seen in many top-level encounters.

14 . . . liJgf6
In the game Chernin-J.Polgar, Groningen 1 993, Black tried 1 4 . . . ti:::J h6 1 5
ti:::J b 5 axb4 1 6 axb4 ti:::J f6 1 7 c5 ti:::J h 5 1 8 cxd6 cxd6 1 9 tl::J c4 .lia6 20 g3 with
unclear chances.

1 5 i.. d3
8
7

6
4

6
5
4

Forcing Black to commit himself with the f

pawn, so that White can have a free hand to
exert pressure on the queenside. This method
costs him two tempi but Black too is maneuvering slowly, improving the position of his knight
on the kingside.

If 1 5 ti:::J b 5 axb4 16 axb4 ti:::J h 5 ! 17 g3 ti:::J df6

1 8 c5 (see diagram) 1 8 ... fxe4 (unclear is
1 8 . . . .i.d7 19 .lib3 i-h6 20 .lic3 i.. f4 ! ? 20 cxd6
1 tl::J x g3 2 1 hxg3 ti:::J h 5 22 gxf4 and Black' s attack

b
"
d e f g h
did not succeed, K.amsky-K.asparov, Dortmund
1 992, or 1 9 . . . tl::J x g3 ? ! 20 hxg3 ti:::Jh 5 2 1 f4 ! Karpov-Kasparov, Tilburg 1 99 1 ,
and Epishin-Piket, Wij k aan Zee 1 992) 1 9 fxe4 .ih3 2 0 .lif2 'ir'd7 with
equal chances, Epishin-Van Wely, Wijk aan Zee 1 992, or 20 . . . tl::J g4 2 1
2

Variation 9 tiJd2 1 09

xf8+ .ixf8 22 tiJf3 Polugaevsky-J.Polgar, Aruba 1 99 1 . Also 20 e l t?Jg4

2 1 t?Jf3 h6 22 cxd6 cxd6 23 'illc 7 "ilxc7 24 t?Jxc7 lta2 25 b2 lta3 26 t?Jh4
h7 27 tiJb5 ! a6 28 c2 f7 29 tiJf3 .if6 with chances for both sides,
Gelfand-Van Wely, Wijk aan Zee 1 992.

1 5 ... f4 1 6 i.e2
Also playable is 16 tiJb5 b6 17 c5 dxc5 1 8 bxa5 xa5 19 t?Jc4 (premature
is 1 9 d6? ! t?Je8 ! 20 dxc7 t?Jxc7 2 1 tiJd6 a8 etc.) 1 9 . . . a8 20 a4 tlJ e8 21 a5
i.a6 22 i.d2 .ixb5 23 .ixb5 tiJd6 24 ltb2 with compensation for the pawn,
lvanchuk-Kasparov, Linares 1 992.

1 6 .. Jlg8 1 7 tt:Ja4 i.f8 1 8 c5 axb4 1 9 axb4 i.e7 20 tt:Jc4 !

This provides more lasting pressure than 20 c6 bxc6 2 1 dxc6 t?Jf8 22 b5
t'Lle6 23 .ib2 g5 24 .ic4 g4 25 "ifb3 g6 ! 26 .ixe6 .ixe6 27 'ifxe6 xa4
Chernin-Ehlvest, Podolsk 1 99 3 .

2 0 g 5 2 1 b5 .1Ixa4 ! ?

B y sacrificing the exchange for a pawn, Black prevents the queenside

from collapsing.

22 'ii'x a4 tt:Jxc5 23 "ii'c 2 g4 24 i.a3 g3

8

2
a

25 i.xc5
Correct was 25 h3 ! 'ir'e8 (either 25 . . . b6 26 .ixc5 bxc5 27 b6, or 25 . . . 'iff8
26 .ixc5 .ixh3 27 gxh3 g2 28 f2 is winning for White) 26 t?Jxe5 ! dxe5 27
i.xc5 with the better chances.

25 ... gxh2+
Not 25 . . . "ile8 26 h3 'ifh5 27 .id 1 ! dxc5 28 b6 etc.

26 xh2 'fi'f8
If 26 . . . tiJh5 27 lth 1 ! .

1 1 0 Variation 9 ti)d2

27 lLlxe5?
Correct was 27 :th i ! dxc5 28 <ifi>g i with the advantage.

27 dxe5 28 xe7 'ifxe7

..

28 . . . :xg2+ would lead to perpetual check.

29 l:tfc l llJhS 30 fl
Not 30 'ikxc7? Wh4+ 3 I g i :txg2+ 32 xg2 h3+ 33 h i ti)g3+ and
Black wins.

30 lLlg3 31 g1 'ilfh4 ! 32 'ii' x c7! lLle2+! 33 xe2 l:r.xg2+! 34

xg2 h3+ 35 h 1 g4+ 36 g2 'ii'g 3+ 37 h 1 'iih 3+ th- 1/z.
..

Let us now take a look at an example with the development I O . . . i.d7.

Game 29

Sakaev - Nevednichy
Olympiad, Elista 1 998
1 d4 llJf6 2 c4 g6 3 lLlc3 g7 4 e4 0-0 5 lLlfJ d6 6 e2 e5 7 0-0
lLlc6 8 d5 lLle7 9 lLld2 aS 1 0 a3 d7 1 1 b3
8

3
2

2
a
8

3
2

=s-.....a

White accepts the loss of a tempo on this

preparatory move so as not to allow his queen
side pawns to become disconnected. Another
possibility is to save time at any cost by continuing 1 1 :tbl a4 12 b4 axb3 13 ti)xb3 (see
diagram) 13 ... b6 (or 1 3 . . . c5 14 dxc6 xc6 I S
'i'd3 ti)e8 I 6 :td i 5 I 7 f3 <ifi>h8 I 8 a4 b6 I 9
i.e3 ti)c8 2 0 c 5 bxc5 2 I ti)xc5 with some
initiative, Sakaev-Baklan, Panormo 200 I ) 1 4
:t a l ti)e8 (or I 4 . . . 'i'e8 I S 'i'c2?! ti)hS ! I 6
ti)bS :tc8 I 7 a4 5 I 8 f3 ti)f4 I 9 aS bxa5 20

Variation 9 c!Lld2 I l l

l:f.xa5 g5 2 1 c 5 'ifh5 ! with strong counterplay, Komarov-M.Pavlovic,

Yugoslavia 1 998, but better was 1 5 'fid3 c!Llh5 1 6 xh5 gxh5 1 7 c!Llb5 ,
Ioseliani-Hort, 1 995. Instead of 1 5 . . . c!Llh5 a n interesting manoeuvre is
1 5 . . h8 with the idea . . . c!Llfg8, . . f5) 15 d2 (also possible is 15 a4 f5 or
perhaps 1 5 . . . c5 ! ?) 15 f5 16 f3 fxe4 17 fxe4 ( 1 7 c!Llxe4 allows a favourable
1 7 . . . c!Llf5) 17 ... l:f.xfl + 18 'ii x fl h8 19 g5 h6 20 .i.e3 c!Llg8 21 a4 f6
(2 l . . .h5 ! would be more consistent, intending . . . .i.g7-h6) 22 'it'cl ! g5 23
.i.xg5 'iix g5 24 'it'xg5 hxg5 25 aS with a somewhat better endgame,
Komarov-Smirin, Paris 1 998.
.

..

l l c6
...

There are several playable alternatives:

ll ... c5 1 2 l:f.b1 c!Lle8 1 3 b4 (see diagram)
axb4 (or 1 3 . . . b6 1 4 c!Llb3 f5 1 5 bxa5
bxa5 16 d2 c!Df6 1 7 f3 c!Llh5 1 8 g3 c!Llf6 1 9
c!Dc 1 h8 2 0 c!Dd3 c!Deg8 2 1 c!D f2 .i.h6 22
'iic 1 .i.xd2 23 'iix d2 f4 24 gxf4 c!Llh5 25 fxe5
c!Llf4 26 h i 'i'g5 27 l:f.g l 'iix e5 28 .i.d 1
c!Df6 29 c!Lle2 c!Llxe2 3 0 xe2 l:f.ab8 and Black
had compensation for the sacrificed pawn
along the dark squares, Ftacnik-Hjartarson,
Olympiad, Manila 1 992) 14 axb4 b6 1 5
bxc5 (better i s 1 5 'ifb3 f5 1 6 bxc5 bxc5 1 7
'iib6 h6 1 8 'iix d8 .:.xd8 1 9 l:f.b6 c!Llc8 20
l:f.b7 Portisch-Paragua, Olympiad, Istanbul
2000) 15 ... bxc5 16 c!Llb3 f5 17 f3 c!Df6 1 8
d2 f4 1 9 c!Llb5 c!Llc8 20 l:f.a1 l:f.xal 2 1 'iix a1
g5 22 'ii'a 6 'it'e7 (the position is very
double-edged) 23 c!DaS g4 (see diagram) 24
c!Llc6 'ii'f7 25 c!Llb8 g3 26 c!Dxd7 gxh2+ 27
hl 'ifxd7 28 'iic 6 'it'd8 29 l:f.al c!Llh5 30
.tel c!Llg3+ 3 1 .i.xg3 fxg3 32 c!Llc7 c!Lle7 33
'iib 7 'iib 8 34 .:.b1 'ifxb7 35 l:f.xb7 c!Llg6 36
.i.d1 ..tf6 37 .i.c2 c!Llf4 and Black realized his
material
advantage,
Vaganian-Gelfand,
Odessa 1 989.
13

...

Another option is to abandon the blockade

of the queenside with l l ... c!Lle8 (see diagram)
1 2 l:f.b1 f5 1 3 b4 axb4 1 4 axb4 c!Df6 1 5 c5
h8 (if 15 . . . fxe4 1 6 c!Lldxe4 c!Llf5 1 7 .i.g5 h6
1 8 ..txf6 .i.xf6 19 b5 b6 20 c6 .i.c8 2 1 .i.d3
.i.g5 22 :a t with better chances for White,
Kamsky-Nunn, Monte Carlo (blindfold)
1 995) 16 f3 c!LlhS 1 7 c!Llc4 c!Llf4 18 e3 c!Llc8
19 hl g5 20 c!DaS 'i'e8 21 c!Llxb7 .:.a3 ! 22
c6 .:.xc3 23 cxd7 'ifxd7 24 .i.d2 .:.a3 25
c!Lla5 g4 26 .tel c!Dxe2 27 'iix e2 gxf3 28

1 1 2 Variation 9 l?Jd2

gxf3 ::r.a4 29 g5 h6 30 e3 l?Je7 3 1 l:tg1 f4 32 f2 f6 33 ::r.b3 ::r.bs

with equal chances, M .Gurevich-Gelfand, Munich 1 993.
8

7
5

Also interesting is the manoeuvre l l ... l?Jc8

(see diagram) 12 ::r.b1 (or 12 'ifc2 l?Jb6 1 3 a4
::r.e8 14 a3 i.f8 1 5 l?Jb5 c5 1 6 dxc6 xc6 1 7
:rd 1 'iib 8 ! 1 8 l?Jc3 ! l?Jbd7 1 9 l?Jfl l?Jc5 20
i.xc5 dxc5 21 l?Je3 h5 22 l?Jed5 with initiative
to White, Komarov-Kir.Georgiev, Lyon 1 995)

1 2 ...l?Jb6 13 'ifc2 ike7 14 b4 axb4 1 5 axb4

l?Ja4 16 l?Jxa4 i.xa4 17 'iVc3 i.h6 18 'it'd3 c6
1 19 l?Jf3 i.g7! 20 i.g5 cxd5 21 cxd5 h6 22 d2
-.u
i.d7 23 :be l 'iid 8 24 'iib 1 ! l?Jh5 25 g3 i.g4
"
b
c
d
c
r
g
"
26 g2 'ifd7 27 l?Jg1 i.xe2 28 l?Jxe2 f5 29
l?Jc3 l?Jf6 30 f3 :res 3 1 'iid 3 with a slight advantage for White,
M.Gurevich-Smirin, Burgas 1 994.

2
1
a

12 i.b2
An interesting idea is 12 :a2 (see diagram) 12 ...iih8 (alternatives are
1 2 . . . 6 1 3 :c2 l?Je8 1 4 dxc6 l?Jxc6 1 5 l?Jd5 'iid 8 1 6 l?Jf3 M.Gurevich
Stryjecki, Pardubice 2000; 1 2 . . . l?Je8 1 3 l?Jdb l cxd5 14 cxd5 f5 1 5 a4 ::r.c8
1 6 i.d2 l?Jf6 1 7 l?Ja3 fxe4 1 8 l?Jc4 l?Jf5 1 9
"
c
e
1'
d
f
1'
.::t e l Portisch-Nijboer, Olympiad, Istanbul
8
8
2000; 1 2 . . . i.h6 1 3 'iic 2 c5 1 4 l?Jf3 i.xc 1 1 5
7
7
'ifxc 1 l?Jh5 Guseinov-A.Kuzmin, Dubai 1 999;
6
6 1 2 . . . c5 1 3 l?Jdb 1 f5 1 4 l?Jb5 'itr>h8 1 5 l?J l c3
5
5
l?Jg8 1 6 f4 Portisch-Babula, Olympiad,
4
4 Istanbul 2000) 13 :c2 :cs 14 i.d3 i.e8 1 5
3 l?Jdb1 b5 1 6 cxb5 cxb5 1 7 b 4 l?Jh5 1 8 g 3 f5
3
2 1 9 bxa5 l?Jf6 20 ::r.b2 with a strong initiative
2
=-..a 1
on the queenside, Portisch-Iordachescu, Berlin
1 997.
a
b
c
d
e
f g h

Variation 9 d2 1 1 3

Or 1 2 .l::. b l b5 (see diagram) 13 b4 (if 1 3

dxc6 b4 ! 1 4 d5 xc6 with counterplay)
1 3 axb4 14 axb4 bxc4 1 5 xc4 cxd5 1 6
cxd5 f5 1 7 b5 i.c8! 18 i.a3 i.b7 19 i.b4
li.Je8 20 e4 d4 Ftacnik-Babula, Prague
1 999, when White should have continued 2 1
li.Jcxd6 xd6 (if 2 l . . .i.xd5? 22 i.c4 xd6
23 i.xd5, or 2 l . . .xe2+ 22 'ii'x e2 i.xd5 23
J::!. fd l .l::.a 2 24 'ii' e 1 with advantage to White)
22 xd6 i.xd5 23 i.c4 i.xc4 24 xc4 l:le8
25 b6 'ii'd 5! 26 b7 .l::. a b8 27 a5 with slightly better chances.

...

8
7

6
5

1 2 . c5
..

A new idea is 1 2 ... i.h6 (see diagram) 13

J::!. b 1 (more enterprising is 1 3 c5 ! ? dxc5 14
c4 cxd5 1 5 exd5 with an unclear game,
R. Jansen-Nijboer, Holland 1 999) 13 ... c8 14
dxc6 i.xc6 1 5 i.d3 b6 1 6 b4 axb4 1 7
axb4 h5 18 g3 i.d7 19 'ii'e 2 .l::. c 8 20 b3
a4 2 1 xa4 i.xa4 22 i.cl i.xcl 23 .l::. fxc l
i.xb3 24 .l::. xb3 g7 25 c5 e6 26 .l::.b c3
dxc5 27 bxc5 'ifa5 with counterplay,
Ftacnik-Topalov, Polanica Zdroj 1 99 5 .

1 3 b5 lbe8

1 4 b4

6
5

Less efficient is 1 4 f4 i.xb5 1 5 cxb5 exf4 1 6 i.xg7 xg7 1 7 .l::.x f4 g5 1 8

.l::. 2 g6 1 9 'ii'c 2 'fie7 20 i.fl .l::. a e8 2 1 c4 e5 22 .l::. e 1 xc4 23 i.xc4
'ii' e 5 and Black had excellent control of the dark squares, Gelfand-Shirov,
Horgen 1 994.

14 b6
...

1 1 4 Variation 9 Ci::Jd2

Or 14 axb4 1 5 axb4 :xa1 1 6 'ii'x a1 i.b6 1 7 Ci::J x d6 (more energetic than

1 7 Ci::Jb3 i.xb5 1 8 cxb5 cxb4 1 9 VaS b6 20 'it'xb4 f5 2 1 exf5 Ci::Jxf5 22 Ci::Jd 2
Ci::J f6 23 l"i::Je4 l"i::Jx e4 24 \i'xe4 fie? 25 g3 'it'c5 26 g2 i.e3 27 f4 i.xf4 2 8
Axf4 \i'e3 29 l:tg4 'ii'd 2 30 i.xe5 dxe5 3 1 Vxe5 Ci::J e 3+ 32 h3 l"i::Jx g4 3 3
i.xg4 'ii'h 6+ 34 g2 'ifd2+ 3 5 h 3 and White managed t o save the game,
Lalic-Kupreichik, Olympiad, Moscow 1 994) 17 ... Ci::J x d6 18 bxcS i.xd2 1 9
cxd6 Ci::J c 8 20 c S f6 2 1 i.a3 VaS 22 'ii'b 2 b S 2 3 c 6 i.c3 24 'ii'b 3 b 4 25 i.cl
Ci::J x d6 26 cxd7 Af7 ! with equal chances, Ulibin-Galkin, Russian ch. 1 996.
..

15 bxc5 bxc5 1 6 f4 !
Emphasizing the usefulness of the bishop on b2. Less dangerous is 1 6 a4
f5 1 7 l:.a3 i.h6 1 8 f3 i.c8 1 9 :e 1 Ci::J f6 20 i.d3 fxe4 2 1 l"i::Jxe4 Ci::J f5 22 i.e 1
i.f4 23 Ci::Jxf6+ 'ii'x f6 24 i.xf4 exf4 25 i.xf5 i.xf5 I/2-Ih, Piket-Gelfand,
Tilburg 1 996.

16 ... exf4 17 xg7 xg7 1 8 l:.xf4 g5 1 9 lif2 lbg6

a

20 'ilb3

If 20 i.g4 i.xb5 2 1 cxb5 Ci::J f6 (weaker is 2 l . . .Ci::Je 5 22 i.e2 Ci::J f6 23 'ii'b 3

Ci::J fd7 24 'il'c3 f6 25 l:tb 1 a4 26 Ci::J fl c4 27 l"i::Jg 3 'ii'b6 28 Ci::J f5 etc., Sakaev
Iuldachev, Dubai 1 999) 22 i.f5 Ci::J e 5 23 a4 with unclear chances. 20 Ci::J fl ! ?
i s another possible attempt to exploit the weakness o f the f5 square.

20 .. .'ii' e7 2 1 l:.afl g8 22 'ti'g3 'ti'e5 23 'ifxe5 lbxe5 24 h3 a4? !

An error in a somewhat weaker position. Black should have continued
24 . . . f6 25 Ci::J f3 Ci::Jx f3+ 26 i.xf3 g7 .

25 h2 f6 26 lbf3 lbxfJ+ ? !
Correct was t o retain the knight b y 26 . . . Ci::Jg 6 27 g 3 etc.

27 gxfJ t7 28 f4 gxf4 29 l:r.xf4 e7 30 l:r.h4 l:tt7 3 1 l:.g1 .ixb5

32 cxb5 f8 33 l:r.f4 lib8 34 i.h5 l'lg7 35 l:r.b l e7 36 b6 l:tg5 37
i.d1 f5 38 h4 l:g7 39 exf5 lbf6 40 .txa4 lbxd5 4 1 f6+ Black
resigned.

Variation 9 d2
This continuation adds t o the development o f the white pieces, retains the
' status quo ' and doesn' t make premature commitments regarding whether
the knight will go to el or perhaps to g5.
Game 3 0

Gligoric - Quinteros
Linares 1 981
1 d4 lbf6 2 c4 g6 3 lbc3 J.g7 4 e4 d6 5 ltJO 0-0 6 J.e2 eS 7 0-0
lbc6 8 dS lbe7 9 J.d2
a

9 ... lbd7

1
a

Surprised by my 9'h move, my opponent, for

the first time, spent some time thinking about
his move which ' copies ' various other continuations. I was not worried by 9 li::l h 5, (see
diagram) although Black was successful with it
in the game Vogt-Gufeld, Baku 1 980, after 10
g3 ! ? f5 II exf5 lt::l xf5 1 2 lt::l e 4 li::l f6 13 ..tgS h6
(or 1 3 . . .'ife8 1 4 ..ixf6 .i.xf6 1 5 b4 a5 1 6 a3
.i.g7 1 7 c5 axb4 1 8 axb4 l:.xa 1 1 9 Wxa 1 li::ld4
20 lt::lxd4 exd4 2 1 ..id3 .i. f5 22 l:.e I and White
was better in Salinnikov-Smimov, Tomsk
1 998) 14 ..ixf6 ..ixf6 15 ..id3 ..ig7 16 h4 c6!
...

4
3

1 1 6 Variation 9 i.d2

1 7 g2 i.d7 18 'ii'd 2 cxd5 19 cxd5 't!Vb6 20 l:tadl l:.f7 21 i.bl ?! l:taf8 22

g4 ? ! ti::J e 7 etc., because I considered the manoeuvre on the flank (9 . . . ti::J h 5)

premature at a moment when White was better developed, his maj or pieces
connected, and when he was not weakened on the long diagonal (as after 9
b4).
On 9 ...tLlh5 White can also _play 1 0 l:tcl f5 1 1 tLlg5 (or I I exf5 tLlxf5 I 2
tLle4 tLlf4 1 3 l:.e I tLlxe2+ I 4 'ifxe2 b6 I 5 b4 h6 I6 i.c3 i.d7 I 7 'ii'd3 g5 I 8
tLlfd2 g4 I 9 tLlfl 'ike7 20 c5 'ii' f7 2 I cxd6 cxd6 22 b5 tLld4 23 i.xd4 exd4
with good counterplay, Taimanov-Spassky, Moscow I 973) 1 1 ... tLlf4 1 2

i.xf4 exf4 13 i. f3 fxe4 1 4 i.xe4 tLlf5 1 5 tLle6 i.xe6 1 6 dxe6 c 6 1 7 l:Iel

'ike7 1 8 'i!Vg4 h8 1 9 l:tcd l l:.ae8 20 i.b l tLlh6 and Black 's position is
8

3
2

good, Korchnoi-Spraggett, Montpellier I 98 5 .

Also possible i s 1 0 !t e l h 6 1 1 !t e l (This strategy resembles the 'bayonet attack' variation
discussed below) 1 1 ... lll f4 1 2 i.fl g5 1 3 h4
(see diagram) 13 ... gxh4 14 tLlxh4 f5 15 g3

ltJfg6 16 tLlxg6 tLlxg6 17 exf5 i.xf5 18 ltJe4

tLle7 'ii' e 8 with the better position for White,
Geller-Van Wely, Tilburg I 99 3 , or 13 ... g4 (no
better is 13 ... f6 I 4 ltJh2 gxh4 I 5 i.xf4 exf4

I 6 'i!Vh5 f5 I 7 c5 fxe4 I 8 cxd6 cxd6 I 9 tLlxe4

tLlf5 20 tLlg4 i.d7 2 I i.d3 etc., Piket-Nunn,
Monte Carlo (rapidplay) I 995) 14 tLlb2 h5 1 5
c 5 ! dxc5 1 6 i.e3 tLleg6 1 7 ..txc5 .l:Ie8 1 8 g 3 a 6 (the piece o n f4 i s taboo,
but, generally speaking, Black is strategically weaker) 19 'ii'b 3 b6 20 i.e3
.i.f8 21 tLldl i.d7 2 2 i.d2 i.c5 23 ltJe3 a5 24 a4 with a clear positional
advantage, Piket-Gelfand, Dos Hermanas
a
b
c
d e f g h
I 995.
a

4
3

8
7

6
4

9 ... tLle8 (see diagram) is interesting

(compare Part Two of the book), because the
knight does not block the diagonal of the
bishop on c8 in controlling the sensitive e6
square. In the third game of the match,
Taimanov-Fischer, Vancouver I 97 I , followed 10 .l:tcl f5 (see diagram) (weaker is
I O . . . c5 1 1 dxc6 bxc6 I 2 b4 f5 1 3 c5 fxe4 I 4
cxd6 tLlxd6 I 5 tLlg5 tLlef5 I 6 tLlcxe4 tLlxe4
I 7 i.c4+ h8 I 8 tLlxe4 h6 .1 9 i.c3 'ii' h4 20
'ike 1 tLld4 2 I f3 'ii'x e I 22 .l:tfxe I with a
superior endgame, Geller-Sax, Hilversum
I 973) 1 1 'it'b3 (an improvement on the first
game, where the players continued I I exf5
gxf5 I 2 tLlg5 h6 1 3 tLle6 ..txe6 I 4 dxe6 'iVc8
I 5 'ikb3 c6 I6 i.h5 'ii'x e6 I 7 'i!Vxb7 tLlf6 I 8
i.e2 l:.fb8 I 9 'i!Va6 l:.xb2) 1 1 ...b 6
lso
possible was I l . . .h8) 12 exf5 gxf5 1 3 g5
tLlf6 (if I 3 . . . h6 I4 tLle6 i.xe6 I 5 dxe6 'ii'c 8

Variation 9 i.. d2 1 1 7

1 6 c 5 ! bxc5 1 7 lt:Jb5) I 4 f4 h6 (the alternative is 1 4. . . e4) I 5 fxe5 dxe5 I 6

c5! lt:Jfxd5 ( i f 1 6 . . . hxg5 1 7 d6+ h8 1 8 dxe7 fixe? 1 9 cxb6 axb6 2 0
i.xg5) I 7 lt:Jxd5 lt:Jxd5 I S cxb6 axb6 I 9 l:t c 6 ! h8 ( i f 1 9 . . . hxg5 20 i.. c4

h8 2 1 i.. x d5 f4 22 l:tc3 with the threat 23 l:th3) and now Taimanov,

instead of 20 lt:Jf3? i.. b 7 2 1 l:tg6 lt:Jf4 ! 22 i..x f4 exf4 23 l:td 1 "i/e7 24 l:te6
c5+ 25 fl l:tfd8 with advantage to Black, should have tried 20 'ii'h 3 !
lt:Jf6 2 I i.c3 i.. d 7! 22 lt:Je6 i.. x e6 23 l:txe6 "ild7 24 l:txe5 lt:Jg4 25 l:tdi "ilf7
26 l';led5 with the initiative.
On 9 ... lt:Je8 there was also the continuation

IO b4 f5 1 1 "ilb3 lt:Jf6 I2 exf5 gxf5 I3 c5

(see diagram) 13 ... h8 I 4 cxd6 cxd6 I 5

.l:!acl i.d7 I 6 a4 lt:Jg6 I 7 i.b5 h6 I S i.xd7

xd7 I 9 lt:Jb5 lt:Je8 20 lt:Jei h7 2 I 'it'h3 !
e4 22 lt:Jd4 ! lt:Jc7 23 l:r.xc7 'ii'x c7 24 lt:Je6
W/e7 and White has a slight initiative,
Korchnoi-Geller, Moscow 1 97 1 , or I O lt:Jei
f5 I I lt:Jd3 lt:Jf6 I 2 f3 (see diagram) 1 2 . . . f4

(Or 1 2 . . .'h8 1 3 l';lc 1 c5 1 4 g4 i.. d7 1 5 a3 a6

1 6 b3 lt:Jc8 1 7 h 1 lt:Jb6 1 8 tiJt2 l:tf7 1 9 gxf5
gxf5 20 f4 exf4 2 1 i.. x f4 'ile7 22 i. g5 'i'f8
23 exf5 i.. xf5 24 i.. d 3 with a strong initiative
for White, Taimanov-Geller, Moscow 1 973)

h
8

7
5

how managed to draw a passive endgame,

Korchnoi-Giigoric, Leningrad 1 97 3 .

13 l:tc l g5 I4 c5 lt:Jg6 15 cxd6 cxd6 16 lt:Jb5

Ilf7 1 7 tiJf2 i.. f8 I S "ilc2 i.. d 7 1 9 "i/c7 lt:Je8
2 0 xd8 Jixd8 21 lt:Jxa7 l:ta8 2 2 lt:Jb5 l:txa2
23 i.. c 3 l:ta8 24 l:t a i l:tb8 and Black some-

3
2

Premature i s 9 . . . c5 1 0 dxc6 bxc6 1 1 i.. g 5 !

a
b
c
d
e
f g h
i.. e6 1 2 c5 lt:Je8 1 3 cxd6 lt:Jxd6 1 4 'i!Va4 f6 1 5
i.e3 'ilc7 1 6 l:tac 1 .l:.fb8 1 7 l:tfd 1 i.. f8 1 8 l:td2 with a clear positional
advantage, Korchnoi-R.Byrne, Leningrad 1 97 3 .

8
7
6
5
4
3

1f:--l--J'.o;!,ll
a

1 1 8 Variation 9 J.d2

1 0 b4
This seemed to me more appropriate for a qu eenside advance than I 0
llc l .
1 0 ... h6
The opponents fight for the initiative on opposite sides and this prophy
lactic move means a loss of a tempo in the race, but Black did not like
I O . . f5 because of I l lbg5 .
.

1 1 'ii'b3 h8
Another, perhaps unnecessary, loss of time on taking precautions, now on
the b3 -g8 diagonal, and another 'point' for White in the battle for the
initiative.

1 2 ltac l
Sounder was to play 1 2 lbe l first in order to strengthen the pawn chain as
quickly as possible.

12 f5 13 lLle1 tiJf6
.

Now Black could have also tried 1 3 . . . fxe4 .

14 0 f4 1 5 c5 g5 16 lLlb5 lLle8
Black has to defend the c7 point from the forays of the white knight, but
this is after taking ' one step forward-two steps back' .

1 7 lLld3 h5 1 8 lLlf2 lLlg8

1
a

Black's counterplay lies in the breakthrough . . . g5-g4 and that' s why he

mobilizes all his forces to achieve it.

19 .l:ic2 lLlh6 20 h3 l:lti 21 l:tfcl

Variation 9 i..d2 1 1 9

Compared to the continuations 9 ltle l and 9 ltld2, this i s a great gain for
because he has managed to include even the king's rook in exerting
pressure on the queenside. Black is now on the defensive.
White

2 1 . a6 22 lla3 ..tf8 23 llc4

..

Threatening 24 cxd6 cxd6 25 ltlb6.

23 ..td7 24 a4
.

White harmoniously includes the last reserve on the side where he is

superior. He threatens b4-b5-b6, so Black reacts with a pawn sacrifice in
order to get more air on the other flank.
a

2
a

24 g4 25 fxg4 hxg4 26 hxg4 l:lh7 ! ?

...

Also possible was 2 6 . . J::tg 7. With his next move, White, at the right
moment, parries the threats along the h-file, dealing with the foray of the
queen to h4.

27 .tel ! l:r.b8
Black has been stopped on the kingside and now he tries to prevent the
threat 28 cxd6 cxd6 29 aS and 30 ltlb6.

28 b5 axb5 29 axb5 b6
He mustn 't allow 30 b6 cxb6 3 1 cxd6 ! .

30 c6!
Black's position would be satisfactory after 3 0 cxb6 cxb6.

30 .-tcs 31 'ii a 3 !
..

The double threat of 32 'iVa7 and 32 ltlxeS spells the beginning of the
end.

120 Variation 9 i.. d2

3 1 . l:tg7
.

To meet 32 'iVa7 with 32 . i.. xg4, but he cannot parry the other threat as
well.
. .

32 lLlxe5 lLlf6 33 lLld7 !

In a winning position White does not allow his opponent even a small
respite.

33 ... lLlxd7 34 cxd7 l:.xd7 35 i.. c 3+ g8 36 'fia7 i.b7 37 g5 c5 38

bxc6 ita8 39 'i!Vxa8 i.. x a8 40 cxd7 'ifxg5 4 1 i.d4 i.b7 42 l:tc8 lLlf7
43 lLlg4 Black resigned.

Variation 9 g5
Unlike the previous variation, here the white bishop acts much more
aggressively on the 9'h move. Because of the prejudice that the bishop will
have nothing to do on a diagonal crowded with black pieces, it was only
later on that the continuation attracted the attention of White players. The
bishop at g5 nevertheless hinders Black in his attempt to carry out
efficiently the thematic breakthrough . . . f7-f5 . According to the available
data, some 1 50 tournament games were played in which White players were
quite successful, so the variation has been sporadically accepted in grand
master practice as well.
Game 3 1

Yermolinsky - Topalov
Wijk aan Zee 1 999
1 f3 f6 2 c4 g6 3 c3 j_g7 4 e4 d6 5 d4 0-0 6 j_e2 e5 7 0-0
c6 8 d5 e7 9 i.g5
a

h
8

8
7

2
1
a

In

this position Black has a greater choice of possible replies.

9 h5
...

The most frequent reaction of Black players.

1 22 Variation 9 .i.g5
a

There are also many games with 9 ... h6 1 0

.i.xf6 .i.xf6 1 1 b4 (see diagram) l l ... .i.g7
7
7
(possible is 1 l . . ...t?g7 1 2 cS lbg8 1 3 lbd2 .i.e7
6
6
1 4 c6 bxc6 1 S dxc6 .i.gS 1 6 .i.c4 fS 1 7 l:te 1
5
lbf6 1 8 exfS .i.xf5 1 9 tll de4 lbxe4 20 lbxe4
hS 2 1 bS .i.f4 22 g3 .i.h6 23 a4 l:tb8 24 lbc3
4
4
h4 2S aS .i.gS 26 l:ta2 hxg3 27 hxg3 l:th8 with
3
3
2
counterplay on the opposite flank-the
kingside, Eingorn-Wojtkiewicz, Tilburg 1 992,
or 1 3 l:tc 1 ! .i.gS 14 lbxgS hxgS 1 S 'ifc2 lbf6
1 6 a4 l:th8 1 7 f3 aS 1 8 Wb2 c6 1 9 bS dxcS 20
bxc6 bxc6 21 l:tfd 1 with the initiative, S .Cvetkovic-Tsarouhas, Chania
1 999) 12 cS (White can build up his game on the light squares by 1 2 lbd2
f5 13 .i.f3 Wh8 14 cS lbg8 1 S l:tc 1 a6 16 lbc4 lbf6 17 a4 gS 1 8 exf5 .i.xf5
19 l:te 1 g4 20 .i.e2 e4 2 1 lbe3 'ifd7 22 c6 bxc6 23 dxc6 'ifc8 24 bS axbS 2S
axbS l:tb8 26 .i.c4 .i.e6 27 'ifh3 with a clear advantage, N.NikcevicD .Antic, Vrnjacka Banja 1 999) 1 2 .. f5 (see
a
b
c
d e f
h
diagram) 13 lbd2 .i.d7 (or 1 3 . . . fxe4 14 lbc4
g
8 lbf5 1 S lbxe4 dxcS 1 6 bxcS lbd4 1 7 lbe3 ..t?h8
8
7 1 8 .i.d3 .i.fS 1 9 l:tc 1 with better play, Legky7
6 Vuruna, Vrnjacka Banja 1 989) 14 .i.bS .i.c8
6
5 1 5 .i.e2 <l;>hS 1 6 l:tcl a6 1 7 a4 lbg8 18 lbc4
4 lbf6 19 cxd6 cxd6 20 f3 hS 21 aS .i.h6 22
4
3 l:t c2 h4 23 'ii' e 1 lbhS with counter-chances,
3
2
Kohler-I. Sokolov, Leeuwarden 1 993 .
2
8

A solid alternative is the restrained 9 ...lbd7

(after 9 . . . lbe8 10 lbd2 h6 1 1 .i.e3 fS 12 f3 f4
1 3 .i.f2 gS 1 4 cS l:tt7 1 S lbc4 White has
everything as in the variation ' 9 lbe 1 lbd7 1 0
a
b
c
d e r 8 h
8 .i.e3 ' , plus a more active king 's knight at d2,
8
going to c4, Legky-Lornje, St. Petersburg
7
7
1 989) 1 0 lbd2 (see diagram) 10 ... f5 (if 1 0 . . . f6
6
6
1 1 .i.e3 f5 1 2 f3 f4 1 3 .i.f2 gS 1 4 b4 lbf6 1 S
5
5
cS lbe8 1 6 a4 lbg6 1 7 lbc4 with the initiative,
4
4
Legky-Reyes, Vrnjacka Banja 1 989, or 1 0 . . . aS
3
3
1 1 a3 f5 1 2 f3 cS 1 3 dxc6 bxc6 1 4 b4 lbf6 1 S
2
2 'iii> h 1 .i.e6 1 6 lbb3 axb4 1 7 axb4 l:txa 1 1 8
1
'ifxa 1 'ii'b 6 1 9 'ifa3 fxe4 20 fxe4 lbg4 2 1
a
b
c
d e -=f1..h..a
l:txf8+ .i.xf8 22 cS 'ifd8 23 h3 lbh6 with
g
chances for both sides, Eingorn-Zapata, Mos
cow 1 989) 11 f3 f4 12 .i.h4 'iii> h 8 13 b4 hS 14 l:tcl with the better chances,
N.Nikcevic-Lanka, Cannes 1 99 S .
a

Variation 9 i..g5 123

1 0 t'Llel
More accurate than 1 0 g3 f6 (if 10 ... f5 1 1 exf5 .i.xf5 12 li)d2 i.. h 3 1 3
.U.e 1 li)f6 1 4 li)de4 with the better position, Legky-Milos, Aosta 1 990;
White also held the advantage after 10 ... h6 1 1 .i.d2 .i.h3 1 2 Ae 1 f5 13 exf5
li)xf5 14 li)e4 'Wd7 1 5 .i.c3 li)f6 1 6 li)fd2 li)xe4 1 7 li)xe4 N.Nikcevic
Brustman, Koszalin 1 998, or 1 3 li:)h4 li)f6 14 exf5 gxf5 1 5 'ii'c 1 f4 1 6 gxf4,
Legky-Mestel, Groningen 1 989) 1 1 .i.d2 fS
a
b
c
d e f g h
1 2 exfS li)xfS (see diagram) 13 li)e4 (Black
8
8
has counterplay after both 1 3 li)h4 ! ? .i.h3 1 4
7
7
li)g2 lt:)f5 1 5 .i.d3 li)d4 1 6 f3 li)f6 1 7 li)e4
6
li)xe4 1 8 .i.xe4 .i.f5 1 9 .i.xf5 Axf5, Timo- 6
shenko-Vitolinsh, Podolsk 1 990, and 1 3 i.. d 3 5
4
li)f6 1 4 li)g5 li)d4 1 5 f3 c6, Eingom-Kuzmin, 4
3
Moscow 1 99 1 ) 13 .i.h6 (or 1 3 . . . li)f6 14 i.. d3 3
2
.id7 [also possible is 1 4 . . . .i.h6 1 5 l:lc 1 .i.xd2 2
'h- 'h, D.Rajkovic- Vukic, Cetinje 1 992] 1 5 1
3 l:lb8 1 6 a4 b6 1 7 'ii'c 2 a5 1 8 g2 h6 1 9
a
b
c
f g h
d c
li)xf6+ 'ii'x f6 20 .i.e4 g5 2 1 h3 .U.f7 22 1i'd3
.U.bf8 with a strong position, Eingom-Smirin, Odessa 1 989) 14 li)fgS li)f6
1 S li)xf6+ 'ii'xf6 1 6 tt:le4 'ii' g7 17 .i.c3 'We7 18 .U.e1 .td7 19 i.. d 3 Aae8 with
chances for both sides, Ftacnik-Kr.Georgiev, Palma de Mallorca 1 989.
...

1 0 t'Llf4 11 t'Lld3 t'Llxe2+

Or 1 I . . .li)xd3 12 .i.xd3 f6 13 .i.e3 f5 14 f3 c5 1 5 Ab 1 f4 16 i.. f2 g5 1 7 b4

b6 1 8 bxc5 bxc5 1 9 .i.e2 h5 ! ? 20 .ixc5 dxc5 2 1 d6 .i.e6 22 dxe7 'flxe7 23
li)d5 1i'f7 24 l:lb5 l:lac8 25 'iib 3 with the initiative, Yermolinsky
Fedorowicz, Philadelphia 1 993 .

1 2 'ii'x e2 h6 13 .i.d2

124 Variation 9 i..g5

8
7

1 3 ... f5

13 c6 (see diagram) 14 .l:r.acl (or 1 4 'itt h 1

cxd5 1 5 cxd5 g 5 1 6 g4 g6 1 7 f3 i.. d7 1 8 a4
a6 1 9 i.. e 3 .l:r.c8 20 'iWd2 'iWa5 2 1 .l:r.fc 1 .l:r.c4 22
b3 .l:r.cc8 23 e2 'iWxd2 24 i.. xd2 h4 25
g 1 with better chances on the queenside,
N.Nikcevic-P.Popovic, Niksic 1 996) 14 ... f5
1 5 f3 cxd5 1 6 cxd5 g5? ( 1 6 . . . i.. d7 ! ) 1 7 b5
i.. d 7 18 f2 'i!Vb6 19 a4 g6 20 'ir'e3 f4
2 1 .l:r.fe1 fxe4 22 fxe4 i.. x b5 23 'iWxb6 axb6
24 axb5 .l:r.fc8 25 .l:r.xc8+ .l:r.xc8 26 i.. c 3 .l:r.c5
27 g3 g6 28 i.. b 4! .l:r.xb5 29 d3 i.. f8 30
.l:r.c1 with a winning endgame, Ftacnik
Wojtkiewicz, Budapest 1 99 3 .
..

I t i s also possible t o play 13 . . .g 5 (see

diagram) 14 g4 (if 14 h4 g4 1 5 f4 gxf3 1 6
'iWxf3 f5 1 7 'iWh5 .l:r.f6 1 8 g4 f4 1 9 g 5 hxg5 20
hxg5 .l:r.g6 21 'itt f2 c6 22 .l:r.h 1 cxd5 23 cxd5
'ifb6+ 24 'itt e2 i.. d7 with sufficient counter
play, 011-Shirov, Tilburg 1 992) 14 ... g6 1 5
f3 f4 1 6 xf4 exf4 1 7 d 1 c6 1 8 i.. c 3

cxd5 19 i.. x g7 'itt x g7 20 cxd5 i.. d 7 21 f2

.l:r.c8 22 'iWd2 'iWb6 23 .l:r.fe l .l:r.c5 24 'itt g 2 .l:r.fc8
25 .l:r.ac l with an equal position, 011Dolmatov, Rostov-on-Don 1 99 3 .

1 4 f4 exf4 1 5 lDxf4 g 5 1 6 lDe6? !

The popular and more challenging continu

ation 16 h5! (see diagram) leads to a high
ly tactical game. For example 1 6 . . . i..d4+
( 1 6 . . . i.. e 5 1 7 exf5 favours White according
to Shirov) 1 7 'itth 1 f4 1 8 g3 fxg3 1 9 .l:r.xf8+

Variation 9 ii.g5 125

'ifxf8 20 lLlb5 (An improvement on 20 l:lfl Ji.h3 ! ! [20 . . . 'ii'd 8?! 21 hxg3 ? !
i.h3 2 2 l:lf6 ii'd7 23 e 5 ii.xe5 2 4 l:lxh6 ii' f5 25 ii'e4 ii.xc3 0- 1 Stanec
Nataf, Fuerstenfeld 200 1 ; 2 1 lLlb5 ! ] 2 1 l:lxf8+ l:lxf8 22 lLlxg3 l:lt2 23 'ii'xt2
i.xt2 24 lLld 1 Ji.d4 with a slightly preferable position for Black, Prakash
Konguvel, Nagpur 1 999) 20 . . . ii't2 2 1 ii'xt2 ii.xt2 22 lLlxc7 ! (Better than 22
hxg3 Ji.b6 23 c3 a6 24 lLld4 r/;f] 25 r/;g2 r/;g6, Prakash-Konguvel, Chen
nai 2000) 22 . . . Ji.h3 23 hxg3 l:lc8 24 r/;h2 ! l:lxc7 25 r/;xh3 l:lxc4 26 Ji.c3
i.d4 with an equal game, Lukacs-Rajlich, Budapest 200 1 .

1 6 ... ..txe6 1 7 dxe6 f4

A possible line is 1 7 . . . c6 1 8 exf5 lLlxf5 1 9 l:lae 1 'ii'e7 20 ii'd3 lLld4 2 1
.l:i.xf8+ l:lxf8 2 2 lLle4 with chances for both sides.

18 g3

3
2
a

1 8 ... fxg3 ?
Correct was 1 8 . . . lLlg6 ! 1 9 gxf4 (if 1 9 'li'h5 'iie 8 20 lLld5 'ifxe6 2 1 l:.ae 1

fxg3) 1 9 . . . ii.xc3 20 bxc3 ! (weaker is 20 iLxc3 lLlxf4 with advantage to

Black) 20 . . . gxf4 2 1 r/;h 1 (2 1 l:lt2) 2 l . . .'iif6 and Black was better, BleesKlarenbeek, Irakl io 1 993 .

1 9 hxg3 lbg6 20 lbd5 ! c6

Not 20 . . . Ji.xb2 2 1 l:lxf8+ ! 'ii'x f8 22 l:lfl ii'g7 23 l:lfl 'iie 5 24 lLlf6+ 'iVxf6
.l:i.xf6 i.xf6 26 'ii' t2 , or 20 . . . l:le8 2 1 'ii'f3 t"De5 22 'i!Vf5 c6 23 e7 'ii'd 7 24
xd7 lLlxd7 25 lLlc7 .
25

2 1 'ifh5
After this move Black faces an unpleasant choice.

2 l . cxd5
..

The best chance. Either 2 1 . ..lLle7 22 l:lxf8+ 'ii'x f8 23 l:lfl or 2 l . . .h7 22

l'lxf8 ii'xf8 23 l:lfl 'ii'd 8 24 e7 lLlxe7 25 ltJf6+ r/;h8 26 ii.xg5 is unsatisfac

l o ry

for Black.

126 Variation 9 ii.g5

22 'fixg6 'ti'b6+ 23 h 1 'it'xb2 24 e7 l:r.fe8

A better chance was 24 . . . %txfl + ! 25 %txfl Vxd2 26 e8='if+ %txe8 27
'ifxe8+ h7 2 8 exd5 'ifd3 .
a

3
2
a

2S i.xgS ! hxgS 26 'iNti+ h8

Or 26 .. /h7 27 %tae 1 %txe7 28 'ifxe7 %th8 29 'iff?.

27 :ae t ! dxc4
So that on 28 %tf2 he could try 28 . . . c3, but. . .

2 8 'fibS+! g8 29 l:le2 'iix e2

If 29 . . . c3 30 %txb2 cxb2 3 1 'iff?+ h8 32 g2 ! wins.

30 1!fxe2 l:r.xe7 3 1 1Wxc4+ h7 32 'ii'd S !tae8 33 llti lilxfi 34

'fixti I:r.eS 35 'iix b7 l':taS ? !
H e could put up better resistance by 3 5 . . . a 5 3 6 'ifd7 g6 3 7 'ii'x d6+ .1i.f6
3 8 'ifc6 etc.

36 'iib3 i.eS 37 g4 i.f4 38 a4 g6 39 'ii'g8+ h6 40 'ti'd8 ! lieS

Not 40 . . . %txa4 4 1 'ii'h 8 + g6 42 'ii'e 8+ and 43 'Wxa4.

4 1 'fif6+ h7 42 'iiti + h6 43 1!fxa7 g6 44 aS f6 4S a6 l:r.cl+

46 g2 l:.c2+ 47 13 Black resigned.

Variation 9 b4 "Bayonet Attack"

In this variation White, in the standard position after eight moves and
with maximum economy of time and without any delay or preparations,
grabs space on the queenside by 9 b4.

One of Black' s first reactions was to exploit the fact that the e2-h5 diag
onal is presently covered and to carry out the active manoeuvre 9 .!Li h5 .
White players considered it necessary to prevent . . . .!Lih5-f4, so in the first
decades of this practice they played 10 g3 (for 1 0 .!Lid2 see the game
Petrosian-Giigoric, given below) 10 ... f5 1 1 ll:)g5 .!Lif6 1 2 f3 f4, with the idea
of exploiting the weakening of the white kingside and the far advanced
k n ight on g5 .
..

2
a

In th e game Pachman-Taimanov, Havana 1 967, after 13 b5 (in order to

i ncrease his control of the light squares and his future pawn at e6) Black
made use of a tempo to extort a draw by perpetual check after 1 3 ... h6 1 4
li'le6 .txe6 1 5 dxe6 fxg3 1 6 hxg3 'ii'c 8 1 7 .!LidS 'ii'x e6 1 8 .!Lixc7 ,.h3 1 9
l,i)xa8 "ii'x g3+ 20 h 1 ll:)xe4 21 fxe4 'ii' h 3+ 22 g 1 'ii' g 3+.
In earlier games of this variation White players tried 13 c5 but without
much success. For example:

13

...

l J SSR

fxg3 14 hxg3 h6 (14 ... .!Lih5 15 'ii'e 1 .!Lif4 16 gxf4 exf4 Stein-Gufeld,
1 967) 1 5 ltle6 i.xe6 1 6 dxe6 d5 Taimanov-Bilek, Leningrad 1 957.

1 3 dxc5 1 4 bxc5 (14 i.c4 cxb4 15 d6+ h8 16 .!Lib5 h6 1 7 .!Lif7+ :xf7

1 8 .txf7 cxd6 1 9 .!Lixd6 i.h3 20 i.b3 h7 2 1 :t2 .!Lic6 and Black had a
dear advantage, C.Hansen-Nunn, Wijk aan Zee 1 99 1 ) 1 4 . . . h6 1 5 .!Lie6
...

128 Variation 9 b4 "Bayon et A ttack "

.ixe6 1 6 dxe6 'iid4+ 1 7 'iixd4 exd4 1 8 tt:)b5 fxg3 1 9 l:tb 1 Sieglen-Duester,

Germany 1 989.
8

More precise is 13 g2 (see diagram)

when Black has several continuations at his
disposal :

1 3 ... h8 1 4 c5 h6 1 5 cxd6! 'iix d6 1 6 tt:)b5

'iib 6 1 7 a4 ! hxg5 ( 1 7 . . . tt:)e8 1 8 tt:)e6 .ixe6 1 9
a5 'ii' a 6 20 tt:)xc7 'iix e2+ 2 1 'it'xe2 tt:)xc7 22
dxe6) 1 8 a5 'iia 6 1 9 tt:)xc7 'iid6 20 tt:)xa8
i.d7 2 1 b5 l:txa8 22 .ia3 'ii'b 8 23 .ixe7 1 -0
Van Wely-Nunn, Wijk aan Zee 1 992.

13 . h6 1 4 tt:)e6 i.xe6 1 5 dxe6 c6 1 6 b5

'ii'c 7 1 7 bxc6 bxc6 1 8 .ia3 l:tfd8 1 9 'ii a4 'it'c8 20 l:tab 1 h8 ! ? 2 1 l:tfd 1
'ii'x e6 2 2 'it'a6 h5 ! 2 3 l:tb7 g 5 2 4 l:txa7 l:tab8 2 5 'it'a5 tt:)e8 2 6 l:tb 1 l:tbc8 !
with a strong counterattack on the kingside, Van Wely-Ye Jiangchuan,
Interzonal, Biel 1 99 3 .
..

13 ... a5 1 4 bxa5 l:txa5 1 5 'iib 3 tt:) e 8 1 6 .id2 l:ta8 1 7 c5 h 8 1 8 cxd6

tt:)xd6 1 9 tt:)e6 i.xe6 20 dxe6 lDc6 2 1 tt:)d5 l:te8 22 .ic3 tt:)d4 23 i.xd4
exd4 24 tt:)xf4 c6 25 i.d3 'it'e7 26 h4 l:ta5 27 a4 l:tea8 with counterplay,
Lobron-Gelfand, Munich 1 992.
13 ... tt:)h5 14 g4 i.f6 15 lDe6 i.xe6 16 dxe6 lDg7 17 c5 h8 ! 1 8 cxd6
cxd6 1 9 tt:)b5 tt:)c8 20 i.c4 l:te8 2 1 i.. d 5 as in Van W ely-Zapata, Matanzas
1 995, and Black missed the chance to gain somewhat better chances by
2 l . . .'it'e7 22 i.b2 tt:)b6 23 a4 a6 24 tt:)c3 tt:)xe6.
13 ... c6 14 'iib 3 ! ? h6 1 5 tt:)e6 i.xe6 16 dxe6 'it'c8 1 7 l:td 1 l:td8 (Not
1 7 . . .'xe6? 1 8 l:txd6 ! 'it'fl 1 9 c5) 1 8 tt:)d5 ! cxd5 1 9 cxd5 g5 20 .id2 with
compensation for the sacrificed piece, Van Wely-Fedorov, Batumi 1 999.
Much later, White players (Hort?) came up with the right idea after 9 b4
tt:)h5-to make the potential arrival of the knight on f4 less effective by the
manoeuvre 10 l:te1 ! , which releases the fl square for the bishop.
a

2
1
a

Variation 9 b4 "Bayon et A ttack " 1 2 9

I n this position, Black players searched for their best chances in several
different continuations, as for instance 10 a5, which was also tried a move
earlier, or to accept the challenge with 10 .!i:Jf4 1 1 .i.fl , or to undermine
t he white centre at once by 10 f5 when after 1 1 .!i:Jg5 there were two different conceptions: 1 1 .!i:Jf4 or 1 1 .!i:Jf6.
.

...

...

...

...

There follow examples from characteristic games that illustrate all these
options for both sides.
Game 3 2

Petrosian - Gligoric
Rovinj/Zagreb 1 9 70

1 c4 g6 2 f3 g7 3 d4 f6 4 c3 0-0 5 e4 d6 6 i.e2 e5 7 0-0

c6 8 d5
White does not have much use for retaining the tension in the centre by 8
i.e3 l:te8 ! 9 dxe5 dxe5 1 0 'i'xd8 l:txd8 ! (more active than the frequently
played 1 0 . . . .!i:Jxd8) 1 1 .i.g5 l:td7 ! with a good game, Pomar-Giigoric,
Olympiad, Lugano 1 968.

8 e7 9 b4 h5
...

2
1
a

1 0 d2 !
The exclamation mark represents the surprise with which I was
confronted. The ex-world champion once admitted to me that he had a
notebook full of his new ideas in various openings-which he started using
only after he had lost his world championship title. Up to the present
encounter he had kept this idea secret.
With the text move White saves the time necessary to increase the
pressure on the queenside, whereby his king ' s knight can have an important
role, and Black is no longer able to block that side as after 9 .!i:Jd2 c5.

1 3 0 Variation 9 b4 "Bayon et A ttack "

White doesn 't achieve anything with the

continuation 1 0 c5 lt:lf4 1 1 i.. xf4 exf4 (see
diagram) 1 2 .l:.cl (if 1 2 'i'd2 .tg4 ! 1 3 .l:.ac 1
7
7
.txf3 1 4 .txf3 g5 1 5 .l:.fd 1 lt:\g6 with very
6
6
good counterplay for Black, Larsen-Gligoric,
s
Lugano 1 970) 1 2 a5 13 cxd6 cxd6 14 lt:lb5
4
4
i.. g 4 1 5 .l:.c7 axb4 1 6 'ii'd 2 i.xt3 ! 1 7 i.xt3
3
3
i.e5
18 .l:.xb7 'il'a5 1 9 lt:ld4 'ii'x a2 20 'ii'x a2
2
2
.l:.xa2 21 .l:.xb4 .l:.fa8 22 i.g4 .txd4 23 .l:.xd4
g5 24 h4 gxh4 25 h2 lt:\g6 26 'iii> h 3 .l:.b2 27
a
b
c
d
e
f g h
i.f5 lt:le5 28 'iii> x h4 h6 29 i.h3 g7 and
Black was better, Kamsky-Kasparov, New York (rapidplay) 1 994.
8

..

1 0 lDf4
.

The knight is strong here, but it can 't stay on this square for ever, and
Petrosian's idea is based on the assessment that Black has spent two tempi
on this manoeuvre and that the knight stands in the way of the black pawn
mass on the kingside. He applied this idea very successfully in the Siimisch
variation against Hort in Palma de Mallorca in 1 969 as well.

1 1 a4
The bishop can't run away to f3 at once: 1 1 i.f3 lt:ld3 12 i.. a 3 a5 so the
other white bishop doesn't have a good place to hide.

l l ... f5
At this moment I felt I was in grave danger of being outplayed via the
queenside, so all my moves were motivated by my hurry to carry out a
counter-reaction that would neutralize White 's initiative. My first idea was
to reduce White's menacing pressure after 1 2 c5 by 1 2 . . . fxe4 and 1 3 . . . lt:lf5.
However, perhaps the simple 1 1 . . .lt:\xe2+ was also playable, clearing the
way for the black pawns.

12 i.t3
Up to here Black has been fighting in the dark, not knowing exactly the
essence of White' s plan, and this move made by his opponent was a small
psychological shock for him that lasted some five minutes. Should he have
taken this bishop earlier? Because now it is too late for 12 . . . lbd3 13 i.a3
and the other bishop is active and safely hidden behind his a-pawn, which
was the idea of White's 1 1 'h move.

1 2 g5 !
.

After the initial surprise, Black spent twenty minutes searching for the
best solution in this critical moment of the battle. The move played is
probably the only sound solution. Black weakens the light squares, but
speeds up his action at the kingside, which, in this dramatic position, is
important for maintaining the balance.

Variation 9 b4 "Bayonet A ttack " 1 3 1

Here 1 2 . . . fxe4, intending t o reduce the pressure o f the white pawn mass,
was much slower. After 1 3 lDdxe4 lDf5 14 g3 the other black knight is in a
very unpleasant position. After the text move, because of the threat l 3 . . . g4,
Black reaches a similar position two tempi earlier.

13 exf5 lbxf5
Not 1 3 . . . il.xf5 because the primary task is to introduce the passive knight
into the game.

14 g3
White doesn't have time for 1 4 lDde4 because of the threat 14 . . . lDh4.
a

1
a

14 ... lt:Jd4 !
At the time this seemed to me to be the only good reaction. This is no
wild piece sacrifice but rather a positionally active continuation which,
together with the material investment, should solve the problem of
maintaining the balance. This is so because after 14 . . . lDg6 the black pieces
would be pushed back and White would not only have a spatial advantage
but would also be superior on the light squares.
Only later did Soviet grandmasters find the
continuation 14 lDh3+ 1 5 g2 1i'd7 ! . (see
diagram) This not so obvious move solves the
problem without sacrificing a piece, because it
prevents 1 6 il.g4. For instance: 1 6 lDb3 (it is
not good to play either 1 6 Jl.g4 lDxt2 ! , or 1 6
lDde4? lDd4 1 7 i.. h 5 lDf4+ 1 8 gxf4 1i'h3+ 1 9
'it>h 1 g4 with a decisive attack) 1 6 ... lDd4 17
lt:Jxd4 exd4 1 8 lDb5 c6 1 9 lDa3 :xf3 20 1i'xf3
g4 21 'Wb3 1i'e7 22 :a2 i.. f5 23 f3 d3 24 fxg4
'ife4+ 25 :n lDg1 (stronger than 25 . . . i.xg4
..

6
5

3
2

8
7

7
6

1 32 Variation 9 b4 "Bayonet A ttack "

26 'ii'xd3) 26 'ii'x d3 'ii'x d3 27 .!:txd3 i.xd3 28 xg l cxd5 29 cxd5 l:te8 h1h, Keene-Kavalek, Teesside 1 97 5 .

1 5 gxf4
Accepting the challenge because Black would feel pretty good after 1 5
i.g4 i.xg4 1 6 'ii'xg4 h5 1 7 'ii'd l .!Llh3+ 1 8 'itt g2 g4 1 9 f3 'ii'd 7.

1 5 xf3+
...

Black could also have played 1 5 . . . exf4, but, considering that he chose to
take ultimate measures by sacrificing material, he prefers to speedily
remove one of the pieces that is protecting the white king.

16 'i'xf3
A more cautious option was 16 lt'lxf3 exf4 1 7 i.b2 g4 1 8 h I ! , intending
to give back the piece in order to balance the position on the endangered
kings ide.

1 6 ... g4 !

2
a

1 7 'i'h 1
I neither anticipated nor expected such a passive move. However, the
manoeuvre is typical of Petrosian 's original style. In our encounter in
Zagreb in 1 965, that is five years earlier, having the king at h i and three
pawns on the second rank in front of it, Petrosian had surprised me by
moving his queen as far as g l , in order to protect the sensitive h2 square
and gain a free hand at the queenside, where the outcome of the game was
decided in his favour.
But, here it was probably better to play 1 7 'ii'd 3 i.f5 1 8 .!Llde4 exf4 1 9
i.. xf4, giving back the piece in order to have counterplay after 1 9 . . . i.. xe4 20
'ikxe4 i.xc3 2 1 .!:ta3 with unclear chances.
With the text move White wants to protect his king from mating threats,
but the queen remains quite out of the game.

Variation 9 b4 "Bayonet A ttack " 133

1 7 ... exf4 18 i.b2 i.f5

Bl ack could have played 1 8 . . . f3 at once, to cut off the white queen from
t he game. But he vainly tries to postpone this advance in order not to allow
strongholds to the white knights at e3 and g3 .

1 9 llfel f3
There was a threat of activating the queen by 20 'i'g2, so this was the last
chance to firmly tighten the position around the white king and queen.

20 tl:Jde4
20 h4 was also playable, but even then Black's attack remains strong.

20 ... 'it'h4 2 1 h3
White can 't allow 2 1 . . .'i'h3 which would definitely sentence the white
queen to languishing impotently in the comer at h l .

2 l . .. e5!

2
a

Now Black' s attack becomes irresistible.

22 lle3
Hopeless is 22 hxg4 'ifxg4+ 23 fl i.xc3 etc.

22 ... gxh3 23 'it'xf3 i.g4 !

This is more energetic than 23 . . . i.xe4 24 :xe4 :xf3 25 l:lxh4 i.xc3 26
.ltxc3 Ihc3 and the gain would not be clear. With the text continuation
Black consistently carries on with the attack regardless of being a piece
down.

24 'ifh l h2+ 25 g2
On 25 fl . 25 . . . :f3 ! is decisive.

25 .. .'i'h5!

134 Variation 9 b4 "Bayonet A ttack "

It took Black some time to find this fine manoeuvre which is the most
efficient way of carrying on the attack, with the battle revolving around the
light f3 and h3 squares close to the king. White ' s response is forced because
he has to protect the f3 square.

26 tbd2 i.d4 !
Black attacks the main defensive piece-the rook that protects the third
rank.

27 'i'e l
This response is again forced because White has to free the h 1 square for
his king. If 27 :ae 1 .th3+ 28 :xh3 'ii' g4+ with mate after 29 . . . :xf2+.
a

1
a

27 ...l:lae8 !
Introducing the last reserve into the battle and this is what decides the
game. White would now rather return the piece than face the continuation
28 h 1 :xe3 ! (saving the dark-squared bishop for the final attack) 29 fxe3
.tf3+ 30 lDxf3 'fi'xf3+ 3 1 xh2 .te5+ with imminent mate.

28 tbce4
The other knight has to protect the f3 square.

28 i.xb2 29 J:.g3
.

White does his best to achieve counterplay. This is the very reason why
Black postpones the capture of the exchange.

29 ... i.e5 30 l:.aa3 'itth 8 3 1 'itt h l l:.g8 32 'iffl i.xg3 33 l:xg3 ?

An error in a lost position.

33 .. Jlxe4 White resigned.

Variation 9 b4 "Bayonet A ttack " 135

Game 3 3

Gligoric - Donner
Berlin 1 9 71
1 d4 tbf6 2 c4 g6 3 tbc3 .tg7 4 e4 0-0 5 .te2 d6 6 tbf3 e5 7 0-0
tbc6 8 d5 tbe7 9 b4 a5
On the alternative 9 . . . lL!hS 1 0 l:Ie 1 aS White is clearly better after 1 1 bxaS
l:IxaS 12 lL!d2 lL!f4 1 3 J.. fl cS 14 a4 l:Ia6 1 S l:Ia3 when Black's retreat
I S . . . lL!hS 1 6 lL!bS lL!f6 1 7 g3 lL!d7 1 8 i.b2 f5 1 9 f4 resulted in an initiative
for White in Xu Jun-Magai, Olympiad, Istanbul 2000. The reader will see
more on this line (9 . . . lL!hS 1 0 l:Ie 1 aS) later.
a

1
a

10 bxa5
This is simpler than 1 0 i.a3 b6 ! 1 1 bxaS lL!hS ! 1 2 l:Ie l f5, which is the
most recent solution to obtaining successful counterplay for Black-for
more about this see the next example.

1 0 .. Jba5
Black's queenside defence would be tougher after the blockading 10 . . . cS.

1 1 tbd2 tbd7 1 2 lbb3 l:a8 1 3 a4

Or 13 i.e3 f5 1 4 f3 f4 1S i.f2 b6 16 a4 lL!cS 1 7 aS lL!xb3 18 'ii'x b3 bxaS
1 9 cS gS 20 cxd6 cxd6 2 1 J.. b 6 'fiVeS 22 lL!bS 'ii' g 6 23 l:Ifc l g4 24 l:Ic7 l:If7
2S fXg4 i.8 26 h3 with superior chances, Atalik-Socko, Saint Vincent
200 1 .

1 3 ... f5 1 4 f3 f4 1 5 .ta3 g5 1 6 c 5 tbf6 1 7 cxd6 cxd6 1 8 tbd2 tbg6

19 tbc4 tbe8

White is happy with the early opening of the position on the queen 's flank
because now he develops an initiative there even more quickly and, as
il lustrated by this forced backward step, he also hampers Black's kingside
counterattack.

1 3 6 Variation 9 b4 "Bayonet A ttack "

20 'it'b3 l:.'tf7 2 1 'ii'b 6 i.. f6 22 'ii'x d8 it.xd8

a

2
a

Without queens, Black can't attack the white king-and White has a
superior endgame.

23 l:.fb 1 l:.a6 24 it.b4 it.d7 25 a5 CiJe7 26 it.a3 it.xa5 27 CiJxa5

l:.xa5 28 ltxb7 CiJc8 29 nbb 1 h5 30 i.. b 4 lha1 3 1 lba 1 g4 32 l:la8
llf8 33 l:tb8 gxf3 34 gxf3 f7 35 l1b7 CiJf6 36 CiJb5 e7 37 CiJa3
:I.g8+ 38 f2 d8 39 CiJc4 llg6 40 i.. fl CiJh7 4 1 i.. a 5+ e8 42
l:'Lxd7! xd7 43 i.. h 3+ e7 44 i.. x c8 Black resigned.
Game 3 4

Gligoric - Simeonidis
Zonal, Panormo 1 998
1 d4 CiJf6 2 c4 g6 3 CiJc3 it.g7 4 e4 d6 5 CiJf3 0-0 6 i.. e 2 e5 7 0-0
CiJc6 8 d5 CiJe7 9 b4 a5 10 bxa5
a

7
6

3
2
a

Variation 9 b4 "Bayonet A ttack " 1 3 7

This i s perhaps more precise than 1 0 .ia3,

played in top circles, because of the recent
finesse 1 0 ... b6 ! .
On the other hand, the often played

1 0 . axb4 1 1 .ixb4 (see diagram) favours

..

"

8
7

5
4

White ' s intentions, for instance: 1 1 ... l'Lld7 3

(Better is 1 l . . .b6 1 2 a4 l:r.e8 transposing into 2
2
the line 1 0 .ta3 b6 1 1 bxaS lhaS 1 2 .tb4
l:ta8 1 3 a4 l:te8-see below. Instead of the
idea 1 2 . . . l:r.e8 weaker is 1 2 . . . .ih6 1 3 aS fS 1 4
.id3 l'Llf6 1 S cS fxe4 1 6 l'Llxe4 l'Llxe4 1 7
.ixe4 il.fS 1 8 l:r.e 1 'ii'd 7 1 9 l:r.b 1 h8 20 .ic3 l:r.a7 2 1 'ii'd 3 l'Llg8 22 .ixf5
'jVxfS 23 'ii'xfS l:r.xfS 24 cxd6 cxd6 2S l:tb6 .if8 26 l:teb 1 with the superior
endgame, Lautier-Roeder, Germany 200 1 ) 12 a4 f5 (or 1 2 . . . .ih6 1 3 aS f5
1 4 lll d2 l'Llf6 1 S cS .ixd2 1 6 'ii'x d2 l'Llxe4 1 7 l'Llxe4 fxe4 1 8 cxd6 cxd6 1 9
.tc3 l'LlfS 20 l:r.a4 l'Lld4 2 1 l:r.xd4 ! cxd4 2 2 'ii'xd4 Black resigned, Ivanisevic
Antic, Novi Sad 2000) 13 l'Llg5 l'Llc5 14 .ixc5 dxc5 15 .to l:ta6 16 aS !
Wh8 17 l'Lle6 .txe6 1 8 dxe6 f4 19 'ii'x d8 l:r.xd8 20 l:ttbl l:tb8 2 l l'Lld5 l'Llxd5
22 cxd5 .tf8 23 .ie2 l:r.a7 24 a6 .id6 25 h4 and White won the superior
endgame, Kramnik-Kasparov, Moscow 1 998.

After 1 0 .ia3 b6! 1 1 bxa5 (see diagram)

Black equally often plays 1 1 ... l:txa5 and
l l . l'Llh5
..

1 1 ... l:r.xa5 1 2 .ib4 l:r.a8 13 a4 l:r.e8 (see 6

5
diagram) (The idea of GM Lanka which 5
4
4
revived the continuation 1 1 . . .l:txaS . Black
intends to play . . . c7-cS and the move . . . l:r.e8 is 3
2
2
a useful way to protect the d6 pawn, i.e. by
neutralizing White' s pressure along the a3-f8
diagonal. This plan is better than the usual
l 3 . . l'LlhS 14 aS bxaS 1 S .ixaS l'Llf4 16 l'Lld2
'ii' d7 1 7 l:r.e 1 .th6 1 8 .tfl c6 19 cS ! ? cxdS 20
d e f g h
exdS dxcS 2 1 l'Llde4 with the advantage,
8
Anand-J.Polgar, Dos Hermanas 1 997) 14 l:r.el 8
7
(Premature is 1 4 aS cS 1 S dxc6 ltl xc6 1 6 7
6
6
.Jtxd6 l'Lld4 1 7 l'Llxd4 exd4 1 8 l'LlbS l'Llxe4 1 9
5
5
.Jic7 'ii'h4 and Black obtained a very active
4
game, Kirsanov-Gallagher, Port Erin 200 1 ) 4
3
3
1 4 .tf8 1 5 h3 (Taking away the g4 square
2
from the black light-squared bishop, which is z
more precise than 1 S .tfl cS 1 6 .ia3 [ 1 6 dxc6 1
l'Llxc6 1 7 i.a3 .tg4 1 8 h3 .ix3 1 9 'ii'x 3 l'Lld4
a
b
c
d e f .g h
20 'ii'd 1 l'Lld7 was unclear in Van Wely
Golubev, Dieren 1 999] 1 6 . . . .th6 1 7 i.b2 l:tf8 1 8 g3 l'Lle8 1 9 l'LlbS f5 20
exfS l'Llxf5 2 1 .tg2 l'Llf6 with a balanced position, Eljanov-Ponomariov,
Kharkov 200 1 ) 15 ... .ib7 (Better than 1 S . . . .td7 16 .tfl h6 1 7 l'LlbS l'Llc8 1 8
.

...

138 Variation 9 b4 "Bayonet A ttack "

'ii'c 2 i.. g7 1 9 a5 c5 20 dxc6 i.. x c6 2 1 a6 lBd7

22 l:ta3 lBc5 23 i.. x c5 dxc5 24 lBd2 l:te7 25
7
7
lBb 1 l:td7 26 lB 1 c3 and White was better,
6
6
Epishin-Lanka, Koszalin 1 999) 16 i.. fl c5
(see diagram) 17 i.. a3 (Another option is to
open the position by playing 1 7 dxc6 lBxc6 1 8
4
4
i..
a 3 lBd4 1 9 l:ta2 li)xf3+ 20 'ii'x f3 lBd7 2 1
3
l:tb 1 i.. c 6 22 'ii'd 1 lBc5 23 i.. x c5 bxc5 24 llb5
2
g7 25 lBd5 with positional pressure,
Epishin-Cherniaev, Geneva 2002) 17 ... i.. g 7
a
b
c
d
e
f g h
(Black has to a certain extent neutralized
White' s pressure on the queenside and now returns to his standard plan
active play on the kingside. Also playable is 1 7 . . . i.. c 8 1 8 lBb5 i.. g7 1 9 i.. b 2
l:tf8 20 l:ta3 lBe8 2 1 'Wa 1 i.. d 7 22 a5 bxa5 23 i.. c 3 lBc8 24 i.. x a5 'fke7 25
'ii'c 1 l:tb8 with chances for both sides, Kasimdzhanov-Ye Jiangchuan, Istan
bul 2000) 18 l:tb 1 i.. c 8 19 i.. c l i.. d 7 20 lBb5 lBc8 21 g3 l:tf8 22 i.. g 5 h8
23 h2 li:Je7 24 i.. d 3 lBe8 25 i.. c 2 f6 26 i.. d 2 f5 with a complex game,
Van Wely-Ye Jiangchuan, Moscow 200 1 .
8

After the alternative continuation l l ... lBh5 there is a further divergence:

If 1 2 lBb5 (or 12 g3 f5 1 3 i..b4 bxa5 14 i.. a 3 lBf6 1 5 c5 lBxe4 1 6 lBxe4
fxe4 1 7 lBd2 e3 ! 1 8 fxe3 lBf5 1 9 lBc4 i.. a 6 20 e4 i.. xc4 2 1 i.. xc4 lBd4 22
'ii'd 3 Ih- Ih, Lutz-Nunn, Bundesliga 2000) 12 ... l:txa5 13 i.. b 4 l:ta6 (see
diagram) 14 lBd2 ( 1 4 a4 f5 1 5 a5 lBf4 1 6
c
b

d e r 8 11
lBd2 l:tf6!-by covering the d6 square Black
8
8
prepares
the key counterstroke . . . c7-c5- 1 7
7
7
i.. f3 ! ? c5 ! 1 8 axb6?!-through necessity
6
6
White opts for an unclear piece sacrifice5
s
1 8 . . . cxb4 1 9 l:txa6 i.. x a6 20 lBc7 i..b 7 2 1
4
4
'ii'b 3 lBc8 22 exf5 gxf5 23 c5 dxc5 24 d6+
3
Cit>h8 25 i.. xb7 lBxb6 26 lBb5 'ii'd 7 27 i.. a 6
2
2
l:tg6 28 g3 'iWc6 29 'iWf3 lBbd5 3 0 h 1 e4 3 1
lBxe4 fxe4 32 'ii'xe4 l:te6 3 3 'ii'f3 'Wxa6 34
a
h
c
d e f g h
d7 i.. f6 3 5 l:td 1 'ii'x b5 36 gxf4 l:td6 and
White resigned, Ivanisevic-Antic, Novi Sad
2000) 14 ... lBf4 15 a4 (it is also possible to play 1 5 l:te 1 lBxe2+ 1 6 'ii'x e2 f5
1 7 a4 fxe4 1 8 lBxe4 lBf5 1 9 'ikd 1 llf7 20 a5 bxa5 2 1 l:txa5 h6 22 'iWa 1 l:txa5
23 'ii'x a5 i.. f8 24 'ii'a 8 lBd4 25 lBa7 with the
a
b
c
d
e
f g h
initiative, Van Wely-Topalov, Rapidplay,
8
8
Monaco 2000) 15 ... f5 1 6 l:te1 !i:Jxe2+ 1 7
7
7
'Wxe2 fxe4 18 lBxe4 lBf5 1 9 'ii'd3 lBd4 20
6
6
i.. c3 lBb3 2 1 l:ta3 lBa5 22 i.. d 2 i.. f5 23 'ii'e 2
5
5
!i:Jb7 24 i.. g 5 'Wd7 25 f3 with a slight
4
4
advantage for White, Van Wely-Smirin,
3
Olympiad, Elista 1 998)
2

Best is 1 2 l:te1 f5 (see diagram) 13 i.. b 4!

(White fixes his pawn structure on the queen
side and this allows him to carry out the

Variation 9 b4 "Bayonet A ttack " 1 3 9

typical plan with c4-c5. This i s stronger than I 3 lLld2 lLlf6 I 4 .i.fl .l:txa5 I 5
.i.b2 .i.d7 I 6 a4 .i.h6 I 7 f3 fxe4 I 8 tLldxe4 lLlf5 I 9 tLlxf6+ 'i'xf6 2 0 tLle4
'ii' h4 2 I .i.c3 .l:taa8 22 a5 bxa5 23 .l:txa5 .i.e3+ 24 h i .i.f4 25 h3 tLlg3+ 26
l.t>g i tLlxe4 27 .l:txe4 'ii'g 3 28 'iVa i .l:txa5 29 'ii'x a5 .i.f5 30 .l:te2 'i'li2+ 3 I
Wf2 .i.g3+ 3 2 e3 'ii'g l + White resigned, Tukmakov-M.Pavlovic, Crans
Montana I 999) 13 bxa5 14 ..ta3 lll f6 ( I 4 . . . tLlf4 is a quite playable
fashionable alternative. For example I 5 .i.fl fxe4 I6 tLld2 [ I 6 tLlxe4 .i.g4
1 7 lLled2 e4 I 8 .l:txe4 .i.xf3 I 9 ll:lxf3 .i.xa I 20 'i'xa I lLlh3+ 2 I gxh3 .l:txf3
22 1Lb2 with compensation, Danailov-Nijboer, Wijk aan Zee 2000]
1 6 . . . tLld3 I 7 .i.xd3 exd3 I 8 tLlde4 lLlf5 I 9 'ifxd3 ..th6 20 .i.e I .i.xc I 2 I
.l:taxc I a4 22 .l:tb I tLld4 2 3 .l:tb4 .i.f5 with chances for both sides, Delchev
Smirin, Pula 200 I ) 15 .i. d3 fxe4 ! ? 16 tLlxe4 ..tg4 17 lLlxf6+ .l:txf6 18 .i.e4
lll f5 19 'i'd3 .l:tti 20 .l:tab1 ..th6 21 c5 'iff6 22 cxd6 cxd6 23 .l:tb6 .l:td8 24
l:tfl 'ii'e7 25 h3 ..txf3 26 ..txf3 ..tf8 27 g3 tLld4 28 ..te4 'ifc7 29 .l:ta6 .l:tc8
30 g2 'il'd8 31 .i.b2 .l:tc5 32 ..txd4 exd4 33 'il'xd4 with an almost decisive
advantage although Black somehow managed to draw a pawn down in a
long endgame, Kramnik- Smirin, Belgrade I 999.
...

10 c5!
...

Better than IO . . . .l:txa5 as in the previous game.

a

2
1

1 1 a4

2
a

In case of 1 1 tLld2 the same position would arise as in the game

Gligoric-Jenni, Ziirich I 998, effectively won
a
b c
d e
r 8 h
by White after l l .l:txa5 1 2 a4 (or I 2 tLlb3
8
8
l:ta6 1 3 a4 tLld7 I4 .i.e3 f5 I 5 f3 f4 I6 .i.t2
7
7
g5 I 7 a5 .l:tf7 I 8 tLlb5 .i.8 I 9 tLlxd6 ! ? .l:txd6
6
6
20 lllx c5 lLlxc5 2 I 1Lxc5 .l:th6 22 .i.a3 .l:tg7
5
5
23 l:tf2 l%a6 24 c5 with firm compensation for
4
the piece, Epishin-Tal Shaked, New York 4
3
3
1 997) (see diagram) 1 2 tLld7 (or I 2 . . . .l:ta6
2
1 3 .:Ia3 tLle8 ! ? I 4 tLlb5 f5 I 5 exf5 gxf5 I 6 f4 2
exf4 I 7 l:txf4 tLlg6 I 8 .:Ifl lLlf6 I 9 'ii'c 2 'il'e7 1
20 tLlf3 tLle4 2 I .i.d3 h6 22 l:te i tLle5 23
...

...

140 Variation 9 b4 "Bayonet A ttack "

i.xe4 fxe4 24 lbxe5 i.xe5 25 'i'xe4 i.f5 26 'fi'e2 'fi'h4 27 g3 'i'g4 28 i.xh6
'Wxe2 29 l:.xe2 l:.fa8 30 i.f4 l:.xa4 3 1 l:.xa4 l:.xa4 32 i.xe5 dxe5 33 l:.xe5
i.h3 34 lbd6 b6 35 ..ti>t2 Black resigned, Yermolinsky-Fedorowicz, New
York 2000) 13 lbb5 l:.a6 14 l:.a3 f5 15 i.b2 i.h6 16 exf5 lbxf5 ! ? with
original manoeuvres: 1 7 lbe4 lbf6 18 lbxf6+ l:.xf6 19 i.g4 ! lbd4 20 i.xc8
'i'xc8 21 g3 ! {The point of the attack is the f4 square ! ) 2 1 . .. l:.ti 22 '1fi>g2
'ii'f5 23 i.xd4 ! cxd4 24 'ii'e 2 'ili'c8 25 f4 ! i.g7 26 l:taf3 exf4 27 l:.xf4 l:.xf4
28 gxf4 ! (not giving the e5 square to Black) 28 ... l:.xa4 29 lbxd6 'ii'd 7 30
f5 ! gxf5 31 lbxf5 l:.a8 32 'i'g4 ..ti>h8 33 'ili'xg7+ 'ili'xg7 34 lbxg7 xg7 35
l:r.f4 ! d3 36 f3 ! l:r.c8 37 '1fi>e3 b5 38 cxb5 l:r.c5 39 ..ti>xd3 l:r.xb5 40 '1fi>c4 l:4b2
41 d6! l:r.c2 42 ..ti>d5 Black resigned.
8

"

6
5

6
s

An interesting, similar strategy is 1 1 lbe1

(see diagram) ll ...'ili'xa5 1 2 i.d2 h6 13 a4
'ili'd8 14 l:r.a3 l:r.a6 (also playable is 14 . . . lbh7
and . . . f7-f5) 15 ..ti>h1 '1fi>h7 1 6 g3 lbd7 1 7 lbg2
f5 18 exf5 ! lbxf5 (if 1 8 . . . gxf5 19 f4 lbf6 20
lbe3 with the better game) 1 9 i.d3 lbb6 20 f4 !

lbd7 2 1 lbb5 exf4 22 i.xf4 ! lbe5 23 i.xe5 !

i.xe5 24 i.c2 ! l:r.g8 25 l:r.af3 'ii'g 5 26 'ii' e l i.d7
27 lbf4 i.xf4 28 l:r.xf4 l:r.e8 29 'ii'c3 l:r.e5 30 g4
lbg7 31 l:r.f6 ! Black resigned, Beliavsky-Mohr,
Grize 1 996.

ll .. .lba5 1 2 .tla3 h8
Or 12 . . . lbe8 13 lbb5 l:r.a6 14 g3 ! f5 1 5 exf5 gxf5 1 6 lbh4 lbg6 1 7 lbg2
lbf6 1 8 f4 lbe4 1 9 i.d3 exf4 20 ltJxf4 'ili'e7 2 1 'We2 i.d7 22 lbe6 i.xe6 23
dxe6 l:r.aa8 24 g4 lbe5 25 gxf5 lbxd3 26 l:r.xd3 'ii'h4 27 i.f4 lbg5 28 i.xd6
lbh3+ 29 '1ti>g2 'ii'xc4 30 i.xf8 l:r.xa4 3 1 f6 i.h6 Black resigned, Antic
Radovanovic, Leskovac 2002 .

13 tt'lel !

h
8

8
7
6

2
a

Variation 9 b4 "Bayonet A ttack " 141

This time the main point of White ' s attack is the e5 square.

13 ... d7 1 4 bS lla6 1 S d3 fS 1 6 exfS xfS 1 7 f4 ! d4 1 8

..tg4 !
After this move, Black, despite a long period o f thought, couldn't find an
adequate plan.

18 ... f6 19 .i.xc8 'ii' x c8 20 fxeS dxeS 21 xeS e4 !

Black finds the best chance for counterplay.

22 xd4 .l:r.xfl + 23 'ii' xfl cxd4 24 ti+ g8 2S gS lif6 26 l:.f3

xf3 27 xf3 d6
Black gets back the pawn, but White finds a way to obtain a decisive
i nitiative.

28 .i.f4 ! xc4 29 'ii' e 2 ! b6 30 d6 h6 3 1 aS dS 32 .i.g3 h7

33 h3 'ti'd7 34 'ii' e4 f6 3S 'ii' x d4 'LlhS 36 .tes .tf8 37 g4? ?

I n a n easily winning position (37 lt:Jd2 and 3 8 lt:Je4 o r 37 'iib 6 ! ) White, i n

t ime pressure, makes a mechanical blunder.

37 . .g7 38 .i.g3 e8 ! 39 'ii'd S .i.xd6 40 g2 'ti'e7 41 eS f6

42 'ii f7+ 'ii x ti 43 xti .tb4 44 d8 th- th.
.

Game 3 5

Bacrot - Motylev
Linares 2001
1 d4 f6 2 c4 g6 3 c3 .tg7 4 e4 d6 S f3 0-0 6 .i.e2 eS 7 0-0
CDc6 8 dS e7 9 b4 hS 1 0 liel fS l l lLlgS f6 1 2 f3
a

1
a

1 42 Variation 9 b4 "Bayonet A ttack "

The light-squared bishop is in danger of remaining closed in behind its

own pawns but, on the other hand, White ' s fortified pawn centre makes it
more difficult to introduce the passive black knight from e7 into the game,
as well as to revive the black pawn mass in the centre. A white pawn
coming to e6 might be in danger but in the meantime White exerts pressure
both in the centre and on the queenside and there is now practically no
danger of the white king being attacked.

1 2 . . . c6
8

4
3

-!!!i!!-..a
f 8 h

8
7

6
5

6
5

3
2

.._f_
g !...h...a
f

Another option is 1 2 ... h8? ! 13 i.e3 (see

diagram) 13 ... tL!eg8 (or 1 3 . . . tL!e8 14 l:tc 1 [if 1 4
c 5 f4 1 5 i. f2 tL!xd5 1 6 tL!xd5 'ii'x g5 1 7 cxd6
cxd6 1 8 l:tc 1 i.e6 1 9 i.fl l:td8 Black is superior, Korobov-Volokitin, 200 1 ] 1 4 . . . c6 ! ? 1 5
c5 ! tLlg8 1 6 exf5 gxf5 1 7 f4 tL!e7 1 8 cxd6
tL!xd6 1 9 dxc6 tL!xc6 20 i.c5 exf4 2 1 i.xd6
'it'xg5 22 i.xf8 i.xf8 23 i.f3 Black resigned,
Kramnik-Ivanchuk, blindfold game, Monaco
2000) 14 c5 i.h6 ? ! 15 exf5 gxf5 16 f4 ! tLlg4 ? !
1 7 .ixg4 fxg4 18 tL!ce4 ! with the superior
game for White, Piket-J.Polgar, Aruba 1 995.
Or 1 2 h6 13 tL!e6 i.xe6 14 dxe6 (see
diagram) 14 ... l:te8 (after 14 . . . c6 1 5 i.e3 [or 1 5
h 1 'flc7 1 6 i.b2 l:tad8 1 7 1i'b3 l:tfe8 1 8 l:tad 1
h7 1 9 i.d3 fxe4 20 tL!xe4 tL!xe4 2 1 l:txe4
tL!f5 22 f4 l:txe6 23 fxe5 dxe5 24 c5 with the
superior game, Epishin-Nemet, Biel 1 996]
1 5 . . . h7 1 6 l:tc 1 tL!e8 1 7 1i'b3 l:tf6 1 8 c5 d5 1 9
exd5 cxd5 2 0 l:tcd 1 d4 2 1 f4 ! with a strong
initiative, Bosboom-Pedzich, Willsbach 1 997)
..

1 5 b5 fxe4 1 6 fxe4 h7 1 7 i.g4 c6 1 8 i.h3 d5

19 exd5 cxd5 20 i.a3 with advantage to
White, N otkin-N evednichy, Bucharest 1 997.

Or 1 2 ... tL!e8 (see diagram) 13 c5 ! (if 1 3

>h l ! ? [more active i s 1 3 'irb3 h8 1 4 l:td l ! ?
6
6 with unclear chances, Haritakis-Kotronias,
5 Athens 1 996] 1 3 . . . h6 14 tL!e6 i.xe6 1 5 dxe6
5
4 'i!Vc8 1 6 c5 'i'xe6 1 7 i.e3 l:td8 1 8 'fla4 a6 1 9
4
3 b5 axb5 2 0 'i!Vxb5 c6 2 1 'it'xb7 d 5 2 2 exd5 cxd5
3
2 23 lLib5 h7 with a strong centre and counter2
1 play,
Epishin-J.Polgar, rapidplay, Geneva
996)
... a5 1 4 cxd6 'ii'x d6 15 tLlb5 'flb6+ 1 6
1
13
"' h..a
c
g ooo
a
b
d e ...,f"""'"
hl lLi d 6 1 7 bxa5 l:txa5 1 8 tL!xd6 cxd6 1 9
i.d2 l:ta8 20 l:tb1 'it'a7 2 1 i.c4 i.d7 22 i.e3 'ii'b 8 2 3 tL!e6 l:tc8 2 4 'ii'd 3 f4
25 i.gl l:ta4 26 i.b5 i.xb5 2 7 'ii' x b5 l:txa2 28 'ii' d 7 i.f6 29 l:txb7 'fla8 30
tL!xf4 ! .th4 3 1 tL!d3 'ii' a 3 3 2 'fle6+ Black resigned, Khalifman-Peng
Xiaomin, Russia-China match, Shanghai 200 1 .
7

Variation 9 b4 "Bayonet A ttack " 1 43

It would be interesting to repeat l2 ... lbh5 ! ?

(see diagram) 1 3 c5 tbf4 1 4 i.c4 'ili>h8 1 5 lbe6
i.. x e6 1 6 dxe6 fxe4 1 7 fxe4 lbc6 1 8 i.e3 lbd4
1 9 .l::r. c 1 dxc5 20 bxc5 lDfxe6 when Black
gained a material advantage and a satisfactory
position,
Tukmakov-Efimenko,
Lausanne
200 1 .

6
s

If l2 . . . a5?! 1 3 b5 c6 1 4 bxc6 bxc6 1 5 .l::r.b 1 2

cxd5 1 6 cxd5 h6 1 7 lbe6 i.xe6 1 8 dxe6 c8 1
1 9 lbb5 d5 20 i.a3 .l::r. e 8 2 1 lbd6 winning
material, Eingom-Golubev, Odessa 200 1 .
a

13 hl
13 i.e3 is an alternative, then:
l3 ... h6 (see diagram) 1 4 lbe6 i.xe6 1 5
dxe6 'fkc7 1 6 .l::r.b 1 .l::r. fd8 1 7 b5 'fkc8 1 8 bxc6
bxc6 1 9 'fka4 h7 20 .l::r. ed 1 'ii'x e6 2 1 .l::r.b 7 .l::r.d7
22 'fka6 lbe8 with equal chances,
Komljenovic-A.Kuzmin, Villa de Benasque
1 999,
l3 . lbe8 (see diagram) 1 4 c5 cxd5 1 5 exd5
e4 1 6 cxd6 lbxd6 1 7 i.d4 h6 1 8 i.xg7 'ili>xg7
1 9 'fkd4+ 'it>g8 20 lbe6 i.xe6 2 1 dxe6 'iib 6 22
.l::r. ad 1 .l::r. fd8 23 'fkxb6 axb6 24 fxe4 lbxe4 25
lbxe4 fxe4 26 .l::r.d 7 .l::r. x d7 27 exd7 .l::r.d 8 28
i.. c4+ <l;g7 29 .i.e6 c.tf6 3 0 l:lxe4 with the
superior endgame, Timoshchenko-Motylev,
Ohrid 200 1 .
..

8
7

3
1

a
a

c
c

d
d

h
h

8
7

3
1

144 Variation 9 b4 "Bayonet Attack "

.-..i-.._,..

7
6

6
5

13 ... i.h6 (see diagram) is the most popular

continuation:
14 i.d2 g7 (Or 1 4 . . . f4 [ 1 4 . . . cxd5 1 5 cxd5
fxe4 16 fxe4 'ifb6+ draw, Panno-Ricardi,
Buenos Aires 1 995] 1 5 h3 ! ? g5? ! -better is
1 5 . . . i.xh3- 1 6 f2 'ith8 17 a4 .l:tg8 1 8 a5
cxd5 19 cxd5 i.f8 20 .l:tc 1 h5 2 1 h3 g6 22
b5 e8 23 .l:tc3 .l:tg7 24 'ii'c 2 .id7 25 .l:tc 1
h4 26 c7 xc7 27 .:xc7 White was
quicker to undertake action on the queenside,
Lobron-Barcenilla, Olympiad, Erevan 1 996)
15 1i'h3 [also possible is 1 5 'ifc 1 eg8 1 6
dxc6 bxc6 1 7 b 5 f4 1 8 h3 ! ] 1 5 h5 1 6
.l:tad1 xd5 1 7 exd5 i.xg5 1 8 .lxg5 'ifxg5 1 9
dxc6 f4 20 g 3 bxc6 2 1 .l:txd6 h 5 22 c;#ilh1
e6 with unclear chances, Pogorelov-Avrukh,
Linares 1 997.
.

3
14 h4 cxd5 1 5 cxd5 (see diagram) 1 5 f4
2 ( 1 5 . . . h5 1 6 %te l g3 1 7 'ii'd2 .id7 1 8 .ib5
.......:o-...a 1 .ixb5 1 9 xb5 fxe4 20 fxe4 'ii'd 7 2 1 c7
a
b c
d c r g h
.l:tac8 22 ce6 was in White' s favour,
Savchenko-Volokitin, Cappelle Ia Grande
200 1 ; or 1 5 . . . .id7 1 6 .l:tc 1 a6 1 7 b5 f4 1 8 .if2
.ixg5 1 9 hxg5 h5 20 'ifb3 a5 2 1 b6 c8 22 .ib5 .l:tf7 23 .ixd7 .l:txd7 24
b5 'ii'x g5 25 c7 llb8 26 e6 i/Je7 27 .l:tc7 g3 28 .ixg3 fxg3 29 .l:tec 1
_.h4 3 0 .l:txc8+ %txc8 3 1 .l:txc8+ f7 32 i/Je3 Black resigned, Ivanov
Rudolf, Panormo 200 1 ) 16 .if2 .ixg5 17 hxg5 h5 18 l:c1 (if 1 8 .id3
.id7 19 'Wd2 c8 20 b5 a6 2 1 a3 'ifxg5 with sufficient counter
chances, Orso-Berczes, Hungary 2000) 18 g3 19 'ifd2 .id7 20 .ib5
.ic8 21 e2 .id7 22 c3 .ic8 23 .l:tc2 a6 24 .id3 .id7 2 5 e2 xe2+ 26
.ixe2 .ia4 27 l:r.c3 'Wd7 28 l:r.ecl .l:tac8 29 .in
8 l:r.xc3 30 l:r.xc3 .l:tc8 3 1 g3 fxg3 32 .ixg3 with
8
the better chances, Radjabov-Nijboer, Wijk an
7
7
Zee 200 1 .
6
6
3

..

..

5
4
3

---...a

1 4 c5 (see diagram) 1 4 f4 1 5 cxd6 fxe3 1 6

dxe7 'Wxe7 1 7 d 6 ilid8 1 8 .ic4+ g7 1 9
e6+ i.xe6 20 .ixe6 'ifb6 2 1 e2 .l:tad8 2 2
'iVd3 'ifxb4 2 3 .l:ted 1 with the initiative for the
pawn, Bareev-Balcerak, Germany 200 1 .
.

(From the previous main diagram) There is also:

1 3 b5 c5 14 exf5 xf5 15 .id3 ! ? d4 16 .ie3 h5 ! 17 ge4 f4 1 8
.ifl h5 1 9 a4 g5 2 0 .l:ta2 g4 2 1 fxg4 .ixg4 22 'Wb 1 h4 23 h3 i.f5 with
strong counterplay on the kingside for Black, Ye-Wang, Beijing 1 996.

Variation 9 b4 "Bayonet A ttack " 1 45

7
6

3
2

l 3 . . . h8
The alternative is 13 ... h6 14 li:Je6 i.xe6 15 dxe6 ll:Je8 (see diagram) 1 6
(or 1 6 'iVb3 ll:Jc7 1 7 i.e3 'it?h7 1 8 c 5 d 5 1 9
"
b
c
d
e
f g h
xd5 cxd5 20 l:tad 1 d4 2 1 li:Jb5 li:Jcd5 22 i.g 1 8
8
aS 23 a3 axb4 24 axb4 li:Jc6 25 i.c4 li:Jdxb4 26
7
7
l/)xd4 exd4 27 e7 ll:Jxe7 28 "iixb4 "iic 7 29 l:tb 1 6
6
I1ab8 30 'iVb6 l:tfc8 and Black maintained an 5
5
a pproximate balance, Degraeve-Howell, Gron
4
i n gen 1 995) 16 ... c5 17 li:Jd5 ll:Jc7 18 ll:Jxc7 4
3
jkxc7 1 9 exf5 gxf5 20 g4 'it?h8 2 1 l:tg1 l:.f6 22 3
2
2
gxf5 li:Jxf5 23 i.d3 "iie 7 24 "iie 2 e4 25 i.xe4
J:lxe6 26 l:tb1 l:tf8 27 "iid 3 li:Jd4 28 l:tb2 with 1

b
c
d
e
f g h
better chances for White, Lima-Romero, Leon
1 996.
hS

14 bS cS 1 S lDe6 i.xe6 1 6 dxe6 lDe8 1 7 lLldS ltJg8

1 7 . . . ll:Jc7

doesn't work out because of 1 8 i.g5 .

18 exfS gxfS 1 9 g4 fxg4 20 fxg4 e4 2 1 l:.b 1 "ikh4 22 i.f4 i.eS 23

xeS+ dxeS 24 :n l2Jg7 2S "ikc2 e3 ! ?

After 2 5 . . . ll:Jxe6 2 6 "iixe4 the pawn at e 5 will fall. Now Black places vain
hopes in the future of the pawn at f2.

26 'ii e4 ltf2 27 l:r.xf2 exf2 28 "ikxeS J:te8 29 lDc7 ! .:Us 30 l:.O lDe7
.H xeS lDg6 32 "ie3 "fie7 33 ltJdS 'ii' x e6 34 l:.xf2 "ikxe3 3S l:xf8+
0:'lxf8 36 lDxe3 lDge6 37 g2 lDd7 38 i.f3 lDecS 39 ltJf5 Black
resigned.

146 Variation 9 b4 "Bayonet Attack "

Game 3 6

Gligoric - Nataf
Cannes 1 998

1 d4 ti)f6 2 c4 g6 3 ti)c3 J.. g 7 4 e4 d6 5 ti)f3 0-0 6 J.. e l e5 7 0-0

ti)c6 8 d5 ti)e7 9 b4
8

9 ti)h5

..

Rarely played is 9 . . . c6 1 0 lL!d2 aS 1 1 bxa5 'ii'x a5 1 2 .i.b2 ! 'ii'd 8 1 3 a4 c5

(or 1 3 . . . lL!d7 14 .i.a3 c5 l 5 lLlb5 :a6 16 .i.b2 f5 1 7 :a3 lL!f6 1 8 exf5 lL!xf5
1 9 .i.d3 lL!d4 with chances for both sides, M.Gurevich-Bologan, 3 rd match
game, Saint-Pierre 2000) 14 :a3 lL!e8 1 5 lL!b5 f5 1 6 f4 ! exf4 1 7 i..x g7
liJxg7 1 8 exf5 gxf5? ! 1 9 :xf4 lL!g6 20 :n "fke7 2 1 :e 1 i.. d 7 22 .i.fl lL!e5
23 liJf3 lL!xf3+ 24 .:txf3 'iif6 25 Wa l l i..x b5 26 cxb5 b6 27 i.. d 3 'i'xa 1 28
:xa 1 with a slight advantage in the endgame, Yermolinsky-Al-Modiahki,
l st match game, FIDE world championship, New Delhi 2000.

10 .Ue1 f5 1 1 ti)g5 ti)f6

Black opts for an undermining of the white pawn centre.
a

Variation 9 b4 "Bayonet A ttack " 1 4 7

1 2 .to c 6

More modest is 1 2 ... fxe4 (see diagram) l3 7

l2Jgxe4 (or 1 3 lbcxe4 l2Jf5 1 4 i.b2 tll xe4 1 5
6
l2Jxe4 a5 1 6 b 5 b6 1 7 g3 i.d7 1 8 i.g2 g5 1 9
a4 Sherbakov-Balabaev, Karaganda 1 999, or
1 4 i.g5 l2Jd4 Savchenko-Golod, Vlissingen 4
1 999, with unclear chances) l3 ... l2Jf5 1 4 i.g5 5
l2Jd4 1 5 lbb5! l2Jxf3+ 1 6 'iix f3 i.f5 17 'iib3 ! 2
Kramnik-Gelfand, Vienna 1 996.
a

8
7

3
2
a

3
2
a

Discovered by Kramnik.
Unclear is l3 'iib3 ! ? (see diagram) l 3 ... h6
t 4 l2Je6 i.xe6 1 5 dxe6 fxe4 1 6 tll x e4 l2Jf5 1 7
1t.e3 lbh4 1 8 l2Jxf6+ 'iixf6 1 9 i.e4 'iix e6 20
bS 'ii g4 2 1 i.d2 l:tac8 22 l:tab1 l2Jf5 23 'ii d 3
Wh8 24 bxc6 bxc6 25 h3 'ii' h4 26 l:tb7 d5!
27 cxd5 lbd6 28 l:tb2 l2Jxe4 29 l:txe4 'iix f2+
30 'ifi>h2 cxd5 and in mutual time pressure
Black succeeded in mating his opponent after
3 1 'iix d5 l:tcd8 32 'ii' a 5 Ad3 33 'ii'b 4 'iig 3+
34 'it>g1 l:tf2, D.Antic-P.Popovic, Novi Sad
2000.

g
8

h
a

3
2

3
2
1

8
6

6
s

Perhaps 13 b5 ! ? is also premature. For

example l3 ... cxd5 1 4 cxd5 h6 1 5 l2Je6 i.xe6 6
1 6 dxe6 l:tc8! (see diagram) 1 7 'iib3? ! (more s
promising is 1 7 b2 fxe4 1 8 lll xe4 lll xe4 1 9 4
l:!xe4 ! [Black is better after 1 9 i.xe4 d5 20 3
1t.c2 'i'b6] with the idea of sacrificing the ex- 2
change after 1 9 . . . d5 20 l:txe5 i.xe5 2 1 xe5) 1
1 7 .. d5! 18 l2Jxd5 (if 1 8 exd5 e4 1 9 d6 [not
.

1 3 .te3 !

8
7

6
s

3
2

1 48 Variation 9 b4 "Bayonet A ttack "

1 9 .ta3 .l:tc3 20 xc3 ti:)fxd5 and Black wins] 1 9 . . . xd6 20 .ta3 'fic7 with
better chances for Black) 18 ... ti:)fxd5 19 exd5 e4 20 d6 .l:tc3 ! 2 1 .ta3 ! .l:txb3
22 dxe7 ! .!:txa3 23 exd8= .!:txd8 24 l:tadl .tf6 25 .te2 r,t>rs 26 .!:txd8+
.txd8 27 .!:tdl .tg5 28 .tc4 .l:tc3 29 .tb3 .!:tel 30 .!:txcl .txcl 31 r,t>n th-th,
Kir.Georgiev-Ponomariov, Olympiad, Istanbul 2000.
The bishop on e3 is most active and
provokes
. . . f5 -f4, which increases White ' s
7
control of the light squares. The previously
6
6
attempted 13 .tb2 (see diagram) is a less
5
Eromising alternative-1 3 . . h6 (or 1 3 . . . a6 1 4
4
tt:Je6 i.xe6 1 5 dxe6 fxe4 1 6 ti:)xe4 ti:)xe4 1 7
.1
3
i.xe4 d5 1 8 cxd5 cxd5 1 9 .tc2 th- th,
2
Kasimdzhanov-Nataf, Linares 1 998) 14 ti:)e6
.txe6 1 5 dxe6 fxe4 16 ..ixe4 (if 16 ti:)xe4
"
b
c
f g h
d
c
ti:)xe4 1 7 .!:txe4 ti:)f5 1 8 b5 .l:tc8 19 .l:te2 .l:te8
20 bxc6 bxc6 2 1 c5 d5 22 i.xe5 i.xe5 23
"
b
c
d
e
f g h
8
8
.!:txe5 'ii' f6 24 .!:te l
t/2- th,
Dautov7
7
Kindermann, Nussloch 1 996) (see diagram)
1 6 ... ti:)xe4 (interesting is 1 6 . . . 'ii'b 6 [if 1 6 . . . d5
6
6
1 7 cxd5 cxd5 1 8 i.c2 'itd6 1 9 i.b3 'ifxe6 20
5
ti:)a4 e4 2 1 ti:)c5 'iff5 22 f3 ! r,t>h7 23 fxe4
4
4
ti:)xe4 24 i.xg7 'if2+ 25 r,t>h l r,t>xg7 26 ti:)e6+
3
3
favours White, Tukmakov-Hulak, Tucepi
2
1 996] 1 7 3 d5 1 8 cxd5 cxd5 1 9 ..ixd5
ti:)g4 20 ti:)e4 [20 .l:te2 leads to a draw after
"
b
c
d
c
f g h
20 . . . ti:)xf2 2 1 ti:)a4 ti:)h3+ 22 r,t>h l ti:)2+]
20 . . . :ad8 2 1 W/g3 ti:)x2 22 ti:)x2 .!:txd5 23
.l:te2 'ifxe6 24 .!:tae l ti:)c6 with advantage to Black, Ivanov-Avrukh, Beershe
ba 1 998.) 17 ti:)xe4 ti:)f5 18 iVg4 r,t>h7 (or 1 8 . . . 'ife8 1 9 e7 ! 'ifxe7 20 'ii'x g6
.!:tad8 2 1 .!:tad l with unclear chances, Tukmakov-Delchev, Kastel Stari
1 997) 19 .!:tadl 'fie7 20 .l:td3 .!:tae8 21 1Ih3 d5 22 cxd5 cxd5 23 ti:)c5 e4
with mutual chances, Averkin-Zakharov, Krasnodar 1 99 8 .
8
7

An interesting continuation is 13 .ta3 (see

diagram) with the idea of exerting pressure
along the a3-f8 diagonal : 13 ... cxd5 (Original
6
6
sharp play occurred in the game Lautier
Nataf, Cannes 2002 . After 1 3 . . . a6 1 4 'ii'b 3 h6
4
4
1 5 c5 (1 5 ti:)e6) 1 5 . . . hxg5 1 6 b5 fxe4 1 7 cxd6
-
3
ti:)exd5 1 8 bxc6 bxc6 1 9 d7 exf3 20 ..ixf8
2
2
'itxf8 2 1 dxc8='if 'ifxc8 22 ti:)xd5 ti:)xd5 23
.!:tac l 'iig4 24 'itxf3 'ili'xf3 25 gxf3 .l:tc8 Black
had a material advantage in the endgame) 14
a
b c
d
e
f g h
cxd5 h6 1 5 ti:)e6 i.xe6 1 6 dxe6 fxe4 (Greater
compl ications are reached by 1 6 . . . .l:tc8 1 7 'iib 3 d5 1 8 exd5 e4 1 9 i.e2
ti:)fxd5 20 ti:)xd5 'ii'x d5 2 1 'ir'xd5 ti:)xd5 22 .!:tad 1 ti:)c3 23 :d7 b5 24 i.d 1

8
7

Variation 9 b4 "Bayonet A ttack " 149

ltJxd l 25 .l:r.exd l .l:r.fe8 26 .l:r.xa7 .l:r.xe6 with a balanced endgame, Jedy

nak-Motylev, Panormo 200 1 ) 17 ..txe4 d5 18 b5 lDxe4 19 lLlxe4 dxe4 20
Vi'xd8 ltfxd8 2 1 i.. x e7 :e8 22 i.. c 5 .l:txe6 23 i.. e3 a6 24 b6 .l:r.d8 25 .l:r.ac1
'3;f7 26 llc7+ .l:r.e7 27 i.c5 .l:r.ed7 28 g3 ..tf8 29 .l:r.xe4 .l:r.xc7 30 bxc7 .l:r.c8 3 1
i.b6 i.. d 6 3 2 'it>g2 'it>e6 3 3 .l:r.g4 g5 3 4 h 4 Draw agreed, Rashkovsky-Jenni,
B iel 200 1 .
a

3
2

1 3 ... h6

Also playable is 1 3 ... 'it>h8 14 a4 (or 14 b5 cxb5 1 5 cxb5 fxe4 1 6 lLlgxe4

lDf5 [possible is 1 6 . . . lLlxe4 1 7 lLlxe4 lLlf5 1 8 i.. g4 ! ? lLlxe3 1 9 .l:r.xe3 i.f5 20
..tf3 h- 1/z , Gligoric-M.Pavlovic, Yugoslav championship, Hercegnovi
200 1 ] 1 7 a4 ltJxe4 1 8 lLlxe4 lLlxe3 1 9 .l:r.xe3 i.f5 20 .l:r.c3 a6 2 1 .l:r.b 1 'ii'a 5 22
c4 with the better chances, Bareev-Cvitan, Bihac 1 999; if 14 'ii'b 3 lLle8 !
[stronger than 1 4 . . . h6 1 5 lLle6 ..txe6 1 6 dxe6 fxe4 1 7 lLlxe4 lLlxe4 1 8 i.. xe4
"i!Vc8 19 b5 xe6 20 .l:r.ed l with a certain initiative, Zontakh-Shulman,
Belgrade 1 998] 1 5 lLle6 ..txe6 1 6 dxe6 lLlc7 17 c5 d5 1 8 exd5 cxd5 1 9
i.xd5 f4 2 0 i.xb7 fxe3 2 1 fxe3 .l:r.b8 2 2 .l:r.ad l 'ii'e 8 2 3 i. f3 lLlf5 24 .l:r.d7
ltJxe6 25 .l:r.xa7 lLlxc5 26 'ii' c4 e4 27 bxc5 exf3 28 gxf3 .l:r.b2 29 lLld5 'ii'b 8
3 0 .l:r.c7 lLle7 White resigned, Chow-Shulman, Sioux Falls 200 1 ) 14 ... h6 1 5
ltJe6 i.xe6 1 6 dxe6 fxe4 17 ltJxe4 dS 1 8 cxdS cxdS 1 9 lLlxf6 .l:r.xf6 20 'ii'b3
l:Ixe6 2 1 .l:r.ad 1 l:Id6 22 i.. c S .l:r.d7 23 i.xe7 'iVxe7 24 .l:r.xdS .l:r.xdS 25 i. xdS
with a slight advantage, Bacrot-Nataf, Bermuda 1 999.
If 13 ... f4 (see diagram) 14 .te l h6 15 lLle6
i.xe6 1 6 dxe6 iDeS 1 7 b5 ! 'ii'e 8 1 8 bxc6 bxe6
1 9 e5 xe6 20 i.a3 dxc5 2 1 lLla4 ! lLlb6 22
ltJxc5 fl 23 'ii'c l ! 1i'c4 24 i.d l ! 'ii'x c 1 25
!he 1 with a strong initiative for the pawn,
Kramnik-Gelfand, Belgrade 1 997.

8
7

6
5

14 lt:Je6 xe6 1 5 dxe6 fxe4

a

150 Variation 9 b4 "Bayonet Attack "

If 1 5 . . . g5 1 6 exf5 ! lLlxf5 1 7 'ir'd3 lLlxe3 1 8 l:.xe3 "fie? 1 9 l:.d 1 l:.ad8 20

lLle4 with the better game, Kramnik-Nijboer, Wijk aan Zee 1 99 8 .

1 6 xe4 xe4 1 7 i.xe4 d5 1 8 cxd5 cxd5 1 9 i.c2 b6

It is necessary to take the c5 square away from the white dark-squared
bishop, and protect the knight at e7 which defends the pawn at d5 . If
19 e4? ! (see diagram) 20 .!:tel d4? 21 .id2 e3 22 fxe3 d3 23 .ib3 'ir'd6 24
'ifg4 .i.e5 25 l:.fl ! .i.xh2+ 26 'it>h1 "fig3 27 "fid4 ! 'ir'e5 28 l:.ti ! with a win
ning position, Kramnik-J.Polgar, Linares 1 997.
..

20 'ii g4
Stronger than 20 .ia4 'ir"d6 2 1 .id7 'ifxb4 22 l:.b l 'ii"h4 23 f3 l:.f5 24 .i.f2
'iff6 25 .ig3 h5 26 h3 l:.f8 27 'it>h I 'ir"g5 28 .ih2 e4 29 fxe4 l:.f2 30 l:.g 1
dxe4 Malakhatko-Jenni, Olympiad, Istanbul 2000.

20 . . ..U.f6 ! ?
8

1>

7
6

..a

More precise i s 2 0 ...e 4 (see diagram) 2 1

l:.ad1 "fic7 22 .ib3 l:.f5 2 3 l:.d2 (Or 23 'ife2
l:.af8 24 l:.fl Ih- 112 Rajabov-Moreno Camero,
Pamplona 2002) 23 ... 'ifc3 (after 23 . . . .ic3? 24
.!:te l 'ir'e5 25 l:.xd5 ! lLlxd5 26 'ifxg6+ White has
a strong attack, Sherbakov-Iskusnyh, Tula
1 999) 24 'ii' d 1 l:.d8 25 b5 "fic8 26 .id4 .ixd4
27 l:.xd4 'ir"c5 28 l:.e2 l:.df8 29 a4 l:.e5 30 h3
I/2-Ih, Xu Jun-Ye Jiangchuan, Shanghai 200 1 .

2 1 l:.ad1 'ifd6 22 i.b3 lid8 23 b5 'i!Vxe6

If 23 . . . l:.xe6 24 .te l and 25 i.a3 .

8
7

Variation 9 b4 "Bayonet A ttack " 1 5 1

2 4 'ii'a 4!
I played this move 'a tempo', telling myself "This is how Kramnik would
play too" (he is the one who has made the greatest improvements to White ' s
play i n this variation), not knowing of the recently played game Kramnik
Shirov, Linares 1 998, where Kramnik continued with the more cautious 24
'ir'xe6+ .l:.xe6 25 .te l rf;f7 26 .i.a3 and had to be satisfied with a draw in a
somewhat more favourable endgame after returning the material.

24 ...l:.d7 25 f4 !
A critical position.

25 ... e4 ! ?
Huzman recommends a s a better chance 2 5 . . . exf4 2 6 .td4 'ii'd6 2 7 .i.e5
'ir'c5+ 28 .i.d4 1i'd6, and if 27 .i.xf6 .i.xf6 28 'ike4 g7, although after 29
'ir'f3 ! the position remains unclear.

26 i.c1 'iff5
Strange, but Black has no way of defending his pawn because of the
multiple pins along the diagonals and files !

27 'ifxe4 'ti'xe4 28 l:.xe4 h7 29 lle2 d4 30 .l:te4 ! d3 3 1 i.e6 l:td6

32 i.c4 tbf5 33 g4 l1d4 34 l::I x d4 tbxd4 35 i.xd3 l:td6 36 a4 lbb3
37 i.e3 tbc5 38 .tc2 l:xd l + 39 i.xd1 i.c3 40 f2 g7 41 f3
f6 42 h4 g7 43 i.c2 i.f6 44 i.xc5! bxc5 45 a5 i.d8 46 b6 a6 47
e4 Black resigned.
Game 3 7

Kramnik - Shirov
Tilburg 1 99 7
1 tb f3 tbf6 2 c 4 g 6 3 tbc3 i.g7 4 e 4 d6 5 d 4 0-0 6 i.e2 e 5 7 0-0
l2Jc6 8 d5 l2Je7 9 b4 tbh5 10 :le1
a

1 52 Variation 9 b4 "Bayonet A ttack "

1 0 . . . f5
8

6
5

4
3

3
2

A quite often played alternative is IO a5 1 1

bxa5 (see diagram). Black has tried several
different plans but currently White players
have been more successful :

l l c5 1 2 a4 lt)f4 1 3 i.fl h6 14 lt)b5 g5 1 5

lt)d2 h8 1 6 l:.a3 lt)g8 1 7 g3 lt)g6 1 8 i.e2
lt)6e7 1 9 i.g4 with a positional advantage,
Lautier-Spasov, Leon 200 1 .
..

l l l:.xa5 1 2 lt)d2 [ 1 2 a4 c5 1 3 l:ta3 lt)f6 1 4

lt)h4 lt)e8 1 5 g3 f5 1 6 exf5 gxf5 1 7 f4 lt)g6 1 8
lt)g2 lt)f6 1 9 h3 i.d7 2 0 i.d3 h8 2 1 lt)b5 l:.a6 2 2 i.b2 with a preferable
position for White, Krasenkow-Fedorov, Leon 200 1 ] 12 ... lt)f4 13 i.fl c5
1 4 a4 lt)hs 15 l:.a3 lt)f6 16 lt)bS lt)e8 17 i.b2 f5 18 f4 exf4 19 e5 dxe5 20
i.xe5 i.xe5 21 l:.xe5 with advantage to White, Van Wely-Fedorov, Leon
200 1 .
a

l l f5 1 2 lt)d2 lt)f6 ( 1 2 . . .lt)f4 1 3 lt)b3 lt)xe2+ 1 4 l:.xe2 f4 1 5 l:.c2 f3 1 6

gxf3 h 6 1 7 c 5 g 5 1 8 cxd6 [Better than 1 8 lt)bS lt)g6 1 9 a4 lt)h4 2 0 lt)d2 g4
2 1 cxd6 lt)xf3+ 22 lt)xf3 gxf3 23 <iii>fl Wh4 with an irresistible attack,
Mikha1evski-Kantsler, Tel Aviv 1 999] 1 8 . . . cxd6 1 9 lt)a4 lt)g6 20 lt)b6 lt)h4
2 1 lt)d2 l:.b8 22 -.o l:.xf3 23 lt)xf3 2Dxf3+ 24 h 1 i.g4 25 'ii'g2 lt)e 1 26
._xg4 and White dominated the light squares, Shirov-Babula, Ostrava 2000)
13 c5! [if 1 3 f3 ! ? l:.xa5 1 4 lt)b3 :as 1 5 c5 f4 1 6 a4 g5 1 7 i.a3 lt)g6 1 8 a5?!
( 18 cxd6 cxd6 19 lt)bS lt)e8 20 lt)d2 h5 21 lt)c4 l:.a6 22 aS was the precise
2ath for a white initiative) lt)h4 1 9 cxd6 cxd6 20 i.b4 ! ? h5 2 1 lt)d2 g4 22
tll c4 l:.a6 23 h 1 lt)h7 ! 24 fxg4 lt)xg2 ! 25 xg2 hxg4 26 h 1 Wh4 27
lt)d2 lt)g5 ! 28 l:.g 1 lt)h3 29 l:.fl l:tf6 ! with a very strong attack, Radjabov
Smirin, Dos Hermanas 200 1 ] 13 ... dxc5 14 lt)bJ ! ? (solid was 14 'ifb3 <ifi>h8
1 5 i.a3 l:.xa5 1 6 lt)c4) 14 lt)xe4 15 lt)xe4 fxe4 16 i.c4 o!t)f5 17 it)xc5
lt)d6 18 i.b3 l:.xa5 19 it)e6 i.xe6 20 dxe6 h8 21 i.d2 l:.a6 22 l:.cl 'ii'e 7
23 i.. b 4 l:.f4 24 g3 .l:.f8 25 l:.xe4 l:.b6 26 Wei .:o 27 l:.ec4 l:.xb4 and Black
resigned, Bacrot-Shchekachev, France 2000.
..

..

If 10 <iti>h8 (see diagram) 1 1 lt)d2 lt)f4 1 2

i.. fl it)g8 (Or 1 2 . . . c 6 1 3 i. b2 cxd5 14 cxd5
7 lt)g8 1 5 l:.c l lt)f6 1 6 lt)c4 lt)e8 1 7 h 1 f5 1 8
7
6 g3 lt)h5 1 9 exf5 lt)hf6 20 fxg6 it)g4 with sharp
6
5 tactical play, Zhao Zong Yuan-Hebden, Scar4
4 borough 200 1 ) 1 3 c5 'i'g5 14 h 1 lt)f6 1 5 f3
3 lt)6f5 1 6 lt)c4 'ifh4 1 7 'it' g 1 g_5 1 8 i.e3 g4 1 9
3
fxg4 i.xg4 20 'i'c2 l:.g8 2 1 Wf2 'ii'e 7 22 h 1
2
2
1 i.f6 23 g3 :g7 24 cxd6 cxd6 25 lt)b5 l:.ag8
--......
and Black's attack failed after 26 it)bxd6 i.h4
a
b
c
d e r g h
27 gxh4 b5 28 lt)xb5 1i'xh4 29 it)xe5 1i'f6 30
it)xg4 .Uxg4 Black resigned, Savchenko-Santos, Panormo 200 1
8

..

Variation 9 b4 "Bayonet A ttack " 1 53

1 1 ttJg5 tbf4 12 .txf4 exf4 13 lic1

a

1
a

Black has opened the diagonal for his dark-squared bishop, but the draw
back of this line is that White will create a pawn on e6, which will not be
easy to eliminate.

1 3 ... .tf6

"
b
c
d c f g h
An attempt to improve B lack' s play. Another
8
idea is 13 fxe4 ! ? 14 lbcxe4 lbf5 (see 8
diagram) 1 5 g4 ! (it would be a loss of time 7
7
t o play 1 5 c5 lbd4 1 6 cxd6 cxd6 1 7 g4 xg4 6
6
1 8 1Wxg4 l:tf5 ! 1 9 lbe6 lbxe6 20 dxe6 1We7 2 1 s
h4 d5 22 lbg5 'i!i'xb4 23 e7 l:te8 24 l:tb 1 'ii' c 3 4
4
25 l:te6 f6 26 l:txb7 xg5 27 l:lxa7 fl 28 3
3
l:taa6 l:lxe7 White resigned, Dumitrache- 2
2
Kupreichik, Olympiad, Erevan 1 996) 1 5 ... lbd4 1
1 6 e6+! lbxe6 17 lbxe6 xe6 18 dxe6 l:tf5
a
b
c
d e f g h
(Timoshenko recommends as better 1 8 . . . e5
1 9 'ii'd 5 'ii' e 7 20 c5 with somewhat better chances for White) 1 9 c5 d5 20
lbd6 ! ! cxd6 2 1 e7 with a decisive initiative, Pelletier-Gormally, Hastings
1 996/97 .
.

Also possible is 13 . h8, tried out in

several games but favouring White.
..

14 tbe6 .txe6 15 dxe6 .txc3 16 .l:lxc3

fxe4 17 .tn
In a blindfold game between the same
opponents at Monaco 1 998, was played 17
i.g4 (see diagram) 1 7 ... lbc6 (or 1 7 . . . a5 1 8 b5
e3 1 9 fxe3 fxe3 20 l:texe3 ! c6 2 1 'ii'd4 cxb5 22
cxb5 l:tc8 23 h4 ! l:txc3 24 l:txc3 lbf5 25 .ixf5

6
4

3
2

2
1

1 54 Variation 9 b4 "Bayon et A ttack "

.l:.xf5 26 b6! l:e5 ! 27 .l:.c7 'iff6 28 'iti>h2 ! .l:.xe6 29 'ifxf6 .l:.xf6 3 0 .l:.xb7 d5
3 1 'iti>g3 with the better endgame, Veingold-Cevcenko, Olympiad, Elista
1 998) 18 .l:.xe4 e5 1 9 g3 g5 20 h4 h8 2 1 'ii'd 4 fxg3 22 .l:.xg3 gxh4 23
.l:.g2 'iff6 24 'ii' e3 .l:.ae8 25 f4 xg4 26 .l:.xg4 .l:.g8 with equal chances.

3
2

2
1
a

1 7 ... e3 ! 1 8 fxe3 fxe3

The game Ghorbani-Al Modiahki, Asian team championship 1 998,
continued 1 8 c6 19 exf4 ( 1 9 b5 is stronger) 19 . . Jlxf4 20 g3 'iff6 2 1
.l:.ce3 .l:.d4 22 'ffe 2 e5 with an unclear game.
..

Interesting is 18 a5 19 b5 c6 20 bxc6 bxc6 2 1 <t>h 1 'ffc 7 22 .l:.d3 .l:.ad8

23 exf4 .l:.xf4 24 .l:.d4 d5 25 cxd5 .l:.xd5 26 .l:.xd5 cxd5 27 'ifd2 Jh- th,
Browne-Fedorowicz, USA championship, Chandler 1 997.
.

1 9 .U.cxe3 c6
Kramnik mentioned 19 ... a5 20 b5 c6 21 bxc6 bxc6 as an unclear
possibility.

20 'it'd2 ! d5 2 1 cxd5 cxd5 22 'ii'd 4 ! 'Wd6 23 'iic 5 'ii'f4?

White would have a slight advantage after 23 . . . 'ffx c5 24 bxc5 .l:.fc8 25
.l:.b 1 l:tab8 26 l:teb3 l:tc7 .

24 l:lt3 'ii g 5 25 l:ti! lbti 26 exti+ c!lxti 27 'fic7 'ii'h 4 28 l:e3

'ii'x b4 ? !
Better resistance would have been offered b y 28 . . . l:tf8 29 g3 "ii'g 5 3 0 .l:.e2.

29 a3 'iih 4 30 'ii'x b7 .l:.e8 31 'ii x a7 d4

Black was in time pressure.

32 i.c4+ c!i>f8 33 g3 ! 'ii'g4 34 i.e2 Wc8 35 'ii x d4 Wet + 36 c!lg2

Black resigned.

Variation 9 b4 "Bayonet A ttack " 1 55

Game 3 8

Kramnik - Kasparov
Novgorod 1 99 7
1 t'i) f3 t'i)f6 2 c4 g 6 3 t'i)c3 j_g7 4 e4 d 6 5 d 4 0-0 6 j_e2 e 5 7 0-0
t'i)c6 8 d5 t'i)e7 9 b4 t'i)hS 1 0 .:tel t'i)f4 1 1 j_fl
a

The black knight is actively placed on f4, but it doesn't have a concrete
target and obstructs the black pawns in their quest to grab space around the
white king-and besides, this outpost will not be secure in the future. From
here on, Black players have chosen several quite different ideas.

l l . . . a5
Black uses this moment to break up the compact white queenside pawn
mass. There are numerous alternatives :
The developing move l l . . i.. g4, after 12 h3 i.. x f3? ! 13 'ii'x f3 f5 14 'ii'd 1 !
fxe4 1 5 ltlxe4 h 6 1 6 l:tb1 ! 'i!Vd7 1 7 'i!Vg4 ! 1i'e8 18 l:tb3 ..t;>b8 1 9 g3 ltlh5 20
'We6 ! , leaves White master of the light squares over the whole board,
Tukmakov-Cvitan, Zurich 1 99 5 .
.

Improving the position o f the knight on f4

by l l h6 (see diagram) is interesting but
unclear. Black defends himself against ltlg5 in
case he plays . . . f7-f5, but in so doing loses a
tempo. In this line Black concentrates his
pieces on the kingside, trying to mate the
white king by means of sacrifices. This plan is
very risky since Black doesn 't obstruct
White' s progress on the queenside at all . 1 2 c5
(or 1 2 ltld2 c6 ! ? 1 3 dxc6 ltlxc6 1 4 a3 a5 1 5
l:tb I axb4 1 6 axb4 ltld4 1 7 ltlb3 ltlfe6 1 8 ltlb5
i.. d 7 1 9 ltl3xd4 ltlxd4 ! 20 ltlxd4 exd4 2 1 i.. f4
..

II

,,

3
2

"

'

j,

1 5 6 Variation 9 b4 "Bayonet A ttack "

with an approximately equal position, B . Lalic-Atalik, Hastings 1 995; how

ever it is better to play 1 3 a4 h8 14 a5 cxd5 1 5 cxd5 f5 1 6 ltJc4 g5 1 7 b5
l:tf6 1 8 b6 a6 19 'ifb3 l:tg6 20 .ia3 g4 2 1 l:ta2 ltJg8 22 'ili'b4 h7 23 ltJe3
'ii'g 5 24 exf5 .ixf5 25 ltJxf5 'ili'xf5 26 ltJe4 with a superior position, Piket
Xie Jun, Monte Carlo (rapidplay) 1 996) 12 ... g5 (Weak is 1 2 . . . f5 ? ! 1 3 .ixf4
exf4 1 4 e5 g5 1 5 exd6 cxd6 1 6 ltJd4 i.e5 1 7 cxd6 ltJg6 1 8 ltJe6 .ixe6 1 9
dxe6 'ii'xd6 2 0 'ili'b 3 l:tfc8 2 1 ltJb5 'ili'b 6 2 2 l:tad 1 h8 2 3 l:txe5 ltJxe5 24
l:td6 1 -0 Delchev-Babula, Batumi 1 999) 13 tt::l d 2 ltJeg6 (or 1 3 . . . f5 14 g3 !
ltJfg6 1 5 a4 with the better game, Anand-Almasi, Groningen 1 997) 1 4 ltJc4
h5 15 .ie3 (also playable is 1 5 .ia3 a6 16 b5) 15 ... f5 ! 16 cxd6 cxd6 1 7
ltJb5 fxe4 1 8 tt::l b xd6 .ig4 1 9 'ili'b3 tt::l h 4 20 ltJxe4 (see diagram)
20 ... ltJhxg2 2 1 .ixg2 ltJxg2 22 .ixg5 (22 xg2 .i3+ etc.) 22 ... tt::l x e1 ! 23
.ixd8 ltJf3+ 24 g2 tt::l d 4 25 'ili'd3 i.e2 26 'ifg3 .if3+ 27 n .ixe4 28
ltJd2 (28 ltJd6 ! ) 28 ... .ixd5 29 .ig5 l:lac8 Draw ! , Ionov-Khalifman, Russia
1 996.
"

The thematic attack with 1 1 . .. f5 (see

diagram) after 12 .ixf4 (or 12 .ie3 h6 1 3 ltJd2
7
7 g5 14 c5 fxe4 1 5 ltJdxe4 ltJf5 1 6 g3 ltJg6 1 7
6 l:tc 1 l:t f7 1 8 i.g2 ltJxe3 1 9 l:txe3 g4 2 0 cxd6
6
s cxd6 2 1 ltJb5 i.f8 22 l:tec3 with the initiative,
4 S.Joksic-Bischoff, Biel
1 998) 12 ... exf4 (not
4
3
3 12 . . . fxe4 1 3 i.xe5 ! ) 13 e5 brings about favour2 able complications for White after 13 ... dxe5 (if
1
1 3 . . . g5 14 ltJxg5 dxe5 15 c5 ! h6 16 i.c4 hxg5
1 7 d6+ 'ito>h7 1 8 dxe7 'ili'xe7 1 9 ltJd5 with the
a
b
c
d
e
f g h
initiative, B.Lalic-Gicev, Skopje 1 995, but
1 3 . . . a5 is also playable) 14 ltJxe5 a5 15 c5! axb4 16 i.c4 ! etc, Lputian-Nal
bandian, Yerevan 1 996.
8

1 2 bxa5
12 .ia3 i.g4 (or 12 . . . axb4 13 i.xb4 i.. g4 1 4 l:te3 ! b6 1 5 h3 i.d7 ! 16 a4
f5 17 a5 with a slight advantage, Granda Zuniga-Van Wely, Amsterdam
1 996) 1 3 h3 .ix3 14 'ili'x3 axb4 1 5 .ixb4 c5 1 6 dxc6 ltJxc6 1 7 a3 ltJe6 1 8
ltJb5 ltJed4 1 9 'i!i'd3 ltJxb4 20 axb4 6 2 1 ltJxd4 exd4 22 'ii'd2 l:la7 23
l:tec 1 l:tfa8 24 l:lxa7 'h- 1h Kramnik-Gelfand, Dortmund 1 996.

1 2 .l:Ixa5
...

Or 12 . . . c5 13 a4 l:txa5 (White was better also after 1 3 . . . f5 ? ! 14 .ixf4 exf4

1 5 e5 dxe5 1 6 tt::l x e5 h8 1 7 ltJb5 l:txa5 1 8 l:ta3 .ixe5 1 9 l:txe5 ltJc6 20
l:te2 ltJb4 2 1 d6 with a decisive advantage, Krasenkow-Bologan, Reggio
Emilia 1 997 and 1 3 . . . h6 1 4 ltJb5 g5 1 5 g3 ltJfg6 1 6 ltJd2 h8 1 7 l:la3 ltJg8
1 8 .id3 ltJ6e7 19 i.b2, S.lvanov-Loginov, St. Petersburg 1 998) 14 l:ta3
ltJh5 1 5 ltJb5 ltJf6 1 6 g3 l:ta6 1 7 ltJh4 ltJd7 1 8 f4 exf4 1 9 g_xf4 ltJxd5 20
ltJxg6 hxg6 2 1 cxd5 l:te8 22 l:tg3 tt::l f6 23 e5 dxe5 24 fxe5 ltJxd5 25 l:td3
.ie6 26 .ig2 c4 27 l:txd5 .ixd5 28 i.xd5 with a decisive material advan
tage, Bacrot-Bologan, Cap d'Agde 2000.

Variation 9 b4 "Bayonet A ttack " 1 5 7

13 lbd2

Interesting is 1 3 a4 c5 (or 1 3 . . . f5 ! ? 1 4 l:ta3 h6 1 5 exf5 li)xf5 1 6 li)e4 ! ? g5

1 7 g3 li)g6 1 8 c5 ! ? dxc5 19 i.c4 li)d6 and White ' s pawn sacrifice remains
unj ustified, Bareev-Spasov, Olympiad, Elista 1 998) 14 li)b5 l:ta6 1 5 l:ta3
li)h5 16 g3 h8 1 7 ttJh4 li)g8 1 8 li)g2 li)hf6 1 9 i.b2 li)d7 20 Wa 1 f5 2 1
exf5 gxf5 22 f4 li)e7 23 li)xd6 ! l:txd6 24 fxe5 l:tb6 2 5 i.c3 li)g6 26 e6 li)f6
27 li)h4 li)e7 28 i.. h 3 li)exd5 29 cxd5 1i'xd5 30 i.g2 1i'c4 3 1 li)xf5 i.xe6
32 li)xg7 xg7 33 a5 l:td6 34 l:ta4 with a winning attack, Kobalij aSmimov, Russian championship, Elista 200 I .

1 3 ... c5 14 a4 l:a6 1 5 .lita3 !

Better than I 5 l:tb I h6 I 6 li)b3 g5 I 7 a5 f5 I 8 g3 li)fg6 I 9 exf5 li)xf5 20
i.. d 3 e4 ! ? 2I li)xe4 li)e5 22 f4 gxf4 23 i.. x f4 li)h4 with chances for both
sides, I. Sokolov-Glek, Wijk aan Zee I 997, or I 5 li)b3 h6 I 6 a5 b6 I 7 li)b5
f5 I 8 g3 li)h5 I9 i.. d2 fxe4 20 axb6 l:txa I 2 I 'ifxa I lt:)f5 22 i.g2? (If 22
.!;[xe4 'ii' f6 ! with counterplay) 22 . . . li)f4 ! with a strong attack, Epishin
Pikula, Biel I 996.

1 5 ... g5 ! ?
Black wants to save a tempo, but the pawn
sacrifice is not quite correct.
Unclear is 1 5 h6 (see diagram) I 6 li)b5
i.. d 7 (if I 6 . . . g5 I 7 g3 li)fg6 1 8 i.. e 2 f5 I 9 exf5
li)xf5 20 i.. h 5 li)ge7 2 I i.. g4 with a strategically better game, Krasenkov-Fedorov, Vilnius
1 997) 1 7 g3 li)h5 1 8 i.. b2 i.. e 8 1 9 i.. e 2 li)f6
20 f4 li)d7 2 1 'ii'a 1 f5 25 exf5 li)xf5 Topalov-Nijboer, Wijk aan Zee 1 99 8 .
...

8
7

1
3
2

3
2
a

1 58 Variation 9 b4 "Bayonet A ttack "

II

"

6
5

6
5

2
a

Playable is 1 5 . . . 'iti'h8 (see diagram) 1 6 lL!b5

lL!g8 1 7 .tb2 h5 1 8 'iii' h 1 h4 19 g3 unclear,
Kasimdzhanov-Safin, Tashkent 1 99 8 ; or
15 ... i.d7 1 6 lL!b5 'i!Vc8 1 7 'iii'h 1 g5 1 8 g3 lll h 3
1 9 'i!Ve2 lL!g6 20 .tg2 g4 2 1 :Xfl h6 22 lL!b l
lL!g5 23 .txg5 hxg5 24 lll l c3 , slightly better
for White, Xu Jun-Sasikiran, India 2000.

16 g3 ll'1h3+
If 16 . . . lL!fg6, then after 1 7 'ifh5 White is
clearly better.

17 i.xh3 i.xh3 18 'iih 5 'ii' d 7

Not 1 8 . . . g4? 1 9 lL!f3 ! f6 20 lL!h4 .

19 'it'xg5 h6 20 'it'e3 f5 2 1 'ii e 2 ! f4 22 ll'1b5 h7?

Correct was 22 . . . 'iii' h 8. If 22 . . . lL!g6 23 'iii'h 1 'iii' h7 24 Ag 1 f3 25 lL!xf3 i.g4
and instead of 26 i.b2 ! ? :Xxa4 Szeberenyi-Zimmerman, Budapest 1 999,
White should have kept his material with 26 'iVd 1 ! .

23 gxf4! exf4 24 h 1 ! i.g4 25 ll'1f3 ! ll'1g6 26 l:g1

a

3
2

26 ... i.xf3?
Played because of the threat 27 lL!g5+ but 26 . . . 'iii' h 8 ! 27 i.b2 was better.

27 'illx f3 ll'1e5 28 'ii'h 5 f!/f7 29 'i'h3 ll'1xc4 30 l:.f3 i.e5 3 1 ll'1c7 !

:.xa4
Not 3 1 . . .'ii'x c7 3 2 'Wxh6+ 'iii' x h6 33 :Xh3 mate.

32 i.xf4! Black resigned.

If 3 2 . . . i.xf4 3 3 lL!e6 :lg8 3 4 :Xxg8 'i!Vxg8 3 5 'iVf5+ 'iii' h 8 3 6 'ii' f6+ 'iii'h 7 37
lL!f8+.

Index of Variations
1

0-0

d4 ltlf6 2 c 4 g6 3 ltlc3 .i.g7 4 e 4 d6 5 .!Llf3

Part

9 ll:le I

6 .i.e2 e 5 7 0-0 .!Llc6

d5 ltle7

One

ltld7
10 ll:ld3 f5

..............................
I I f3 . . . . . . . . . . . . . . . . . . . . . . . . .
1 1 exf5 ll:lxf5 ( l l . ..gxf5) . . . . . . . .
12 f3
. .
1 2 ltle4 . .
l l .i.d2 . . . . . . . . . . . . . . . . . . . . . . .
.....................
l l . . . ll:lf6
.....................
l l . . . fxe4
10 .i.e3 f5 1 1 f3 . . . . . . . . . . . . . . . . . . . . . . . . .
l l . . . f4 1 2 .i. f2 g5
1 3 ltld3 ltlf6 1 4 c5 ll:lg6

............
............
............
............
............
............
............
............
............

13
13
25
25
29
31
31
34
17

.................
1 6 "ifb3 . . . . . . . . .
1 6 l:tc2 . . . . . . . . . .
13 b4 ll:lf6 14 c5 ll:lg6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15 cxd6 . . . . . . . . . . . . . . . . . . . . . . .
1 5 a4 ! . . . . . . . . . . . . . . . . . . . . . . . .
13 a4 ll:lg6
1 4 ltld3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14 a5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13 l:tc I ll:lg6 14 c5 ltlxc5 I 5 b4 ll:la6 16 ll:ld3 . . . . . . . . . .
1 6. . .h5 . . . . . . . . . .
1 6 . . .l:tf7 . . . . . . . . .
1 1 . . .ltlf6 1 2 cS f4 13 .i.f2 gS 1 4 a4 ll:lg6 IS aS hS 16 cxd6 cxd6
1 7 ll:lb5 g4 1 8 .i.xa7 . .

17
17
20
35
35
37

Part

9 ll:le l (9 h i ) ltlc8
1 0 f4 . . . .
1 0 lt:\ d 3 f5

...

....

..........
.
..........
. . . . . . . . . .
..........
..........
..........
..........
. . . . . . . . . .

. . . . . . . . .

.......
.......
.......
.
. . . . .
.......
.......
.......
.......
......
.

15 l:tc l l:tf7

. . . . .

. . . . . . . .
.
.. ... .. ..
l l . . .ltJ f6 . . . . . .
( l l ...'i'h8) . . . . .
l l . . .fxe4 . . . . . .
.. ... .. . ...
l l . . .c5
......
i l . .'h 8 . . . . . .
l l . . . f4 12 i.f2 h5

. .

. .
.

. . .

I O .i.e3 f5 I I f3

. . .

42

Two
.

. .

. . .

. . .

..

. .

. . .

. . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I I f4 .
i l il.d2

39
40
48
48
50

. . . . . . . . . . . . . . .
. .
. . . . . . . . .
...... ...... ....... .. .
............................
.. ........................ ..
. . .
. . . . . . . . . . . . . . . . . . .
. .
. . . .
. . . . . . . . . . . . . . . . .
. . .
.
.
.. ......................
........ ....................
1 3 cS gS 14 a4 . . . . . . . . . . .
14 ... ltlg6 . . . . . . . . . .
1 4. . .'i>h8 . . . . . .
. .
1 4 . . . ll:lf6 . . . . . . . .
1 4 . . . dxc5 . . . . . . . . .
1 4 . . . l:tf6 . . . . . . . . .
.

55
55

. . . 58
. . . . . . 62
. . . . . . . 62
. . . . . . . 63
. . . . . . .
67
. . . . . . 71
. . . . .
.
71
. . . . . . . 74
. . .
. . .
79
.

. . . . . . 81
. ..
81
.

82
. . . . . . 85
. . . . . .
87

1 60 Index of Variations

Part Three
9 ll'le l ll'ld7
I 0 f3 f5 I

9
9
9

g4 'it>h8

.. ....... .. .........
1 2 ll'lg2 . . . . . . . . . . . . .
1 2 .te3 . . . . . . . . . . . . . .
1 0 ll'ld3 f5 I I .td2 ll'lf6 1 2 f3 f4 1 3 g4 . . . .
ll'ld2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9 . . . .th6 . . . . . . . . . . . . . . . . . . . . . . . . . .
9 . . . ll'ld7 . . . . . . . . . . . . . . . . . . . . . . . . . .
9 ... c5 . . . . . . . . . . . . . . . . . . . . . . . . . . : .
9 . . . a5 . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 0 a3 . . . . . . . . . . . . . . . . . . .
I O . . . ll'ld7 ( I O . . . ll'le8) . . . . . .
I O . . . .td7 . . . . . . . . . . . . . .
.td2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.tgS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
b4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9 . . . ll'lh5 . . . . . . . . . . . . . . . . . . . . . . . . . .
I 0 g3 f5 I I ll'lg5 ll'lf6 1 2 f3

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.
.

..
..
..
..
..
..
..
..
..
..

.....
.....
.....
.....
.....
.....
.....
.....
.....
.....

.
.
.
.
.
.
.
.
.
.

...................................

...................................
. . . . . . . . . . . . . . . . . . . . . . . . . .
...................................
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

...................................
f4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 3 b5 . . . . . . . . . . . . . . . . . . . . . . . .
l 3 c5
....................
1 3 'it>g2 . . . . . . . . . . . . . . . . . . . . .
1 0 ll'ld2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
( 1 0 c5) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 0 J:l.e l . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
( I O . . . a5) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I O . . . f5 l l ll'lg5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
..............................
l i . . .ll'lf6
1 2 f3 . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 2 .t f3 . . . . . . . . . . . . . . . . . . . . . . . .
1 1 . . .ll'lf4
..............................
1 0 . . . ll'lf4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.
.
.
.
.
.
.

91
91
93
95
97
97
99
1 04
1 06
1 06
I 06
1 10
1 15
121
1 27
1 27
1 27
1 27
1 27
1 28
1 29
1 30
141
152
141
141
141
1 47
153
1 55

9 . . . a5
1 0 bxa5

IO
IO

J:I.xa5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 5
cs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 8
1 0 .ta3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 7
...

...