SPORT-SPECIFIC CONDITIONING VARIABLES PREDICT

OFFENSIVE AND DEFENSIVE PERFORMANCE IN

HIGH-LEVEL YOUTH WATER POLO ATHLETES
DAMIR SEKULIC,1 DEAN KONTIC,2 MICHAEL R. ESCO,3 NATASA ZENIC,1 ZORAN MILANOVIC,4
MILAN ZVAN5

AND

Faculty of Kinesiology, University of Split, Split, Croatia; 2University of Dubrovnik, Dubrovnik, Croatia; 3Department of
Kinesiology, University of Alabama, Mongomery, Alabama; 4Faculty of Sport and Physical Education, University of Nis, Nis,
Serbia; and 5Faculty of Sport, University of Ljubljana, Ljubljana, Slovenia

ABSTRACT
Sekulic, D, Kontic, D, Esco, MR, Zenic, N, Milanovic, Z,
and Zvan, M. Sport-specific conditioning variables predict
offensive and defensive performance in high-level youth water
polo athletes. J Strength Cond Res 30(5): 13161324, 2016
Specific-conditioning capacities (SCC) are known to be generally important in water polo (WP), yet the independent associations to offensive and defensive performance is unknown.
This study aimed to determine whether offense and defense
abilities in WP were independently associated with SCC and
anthropometrics. The participants were 82 high-level male
youth WP players (all 1719 years of age; body height,
186.3 6 6.07 cm; body mass, 84.8 6 9.6 kg). The independent variables were body height and body mass, and 5 sportspecific fitness tests: sprint swimming over 15 meters; 4 3
50-meter anaerobic-endurance test; vertical in-water-jump;
maximum intensity isometric force in upright swimming using
an eggbeater kick; and test of throwing velocity. The 6 dependent variables comprised parameters of defensive and offensive performance, such as polyvalence, i.e., ability to play on
different positions in defensive tasks (PD) and offensive tasks
(PO), efficacy in primary playing position in defensive (ED) and
offensive (EO) tasks, and agility in defensive (AD) and offensive
(AO) tasks. Analyses showed appropriate reliability for independent (intraclass coefficient of 0.820.91) and dependent
variables (Cronbach alpha of 0.810.95). Multiple regressions
were significant for ED (R2 = 0.25; p , 0.01), EO (R2 = 0.21;
p , 0.01), AD (R2 = 0.40; p , 0.01), and AO (R2 = 0.35;
p , 0.01). Anaerobic-swimming performance was positively
related to AD (b = 20.26; p # 0.05), whereas advanced sprint
swimming was related to better AO (b = 20.38; p # 0.05).

Address correspondence to Damir Sekulic, dado@pmfst.hr.

30(5)/13161324
Journal of Strength and Conditioning Research
2015 National Strength and Conditioning Association

1316

the

In-water-jumping performance held the significant positive relationship to EO (b = 0.31; p # 0.05), ED (b = 0.33; p # 0.05),
and AD (b = 0.37; p # 0.05). Strength and conditioning professionals working in WP should be aware of established
importance of SCC in performing unique duties in WP. The
SCC should be specifically developed to meet the needs of
offensive and defensive performance in young WP athletes.

KEY WORDS reliability, multiple regression, ecological validity,

sport-specific test
INTRODUCTION

ater polo (WP) is a team sport that has

been present in the Olympic Games since
1896. Although the rules have changed considerably over time, the game has remained
a highly demanding physical activity (2,24). During the
course of the match that lasts 4 3 89 minutes, athletes
constantly swim, while wrestling with opponents, passing
and shooting the ball in an accurate manner, and completing
complex tactical tasks throughout the game (14,29). Therefore, specific strength and conditioning capacities such
as sprint swimming, swimming endurance, dynamometric
force, shooting ability, and in-water-jumping are commonly
suggested to be important factors of overall success in this
sport (26,29).
As in most other team sports, the success of the whole
team and final result of the games are results of a complex
interaction of different factors, including team tactics,
technical skills of the athletes, individual fitness capacities,
psychological qualities, etc. (17,30). Previous studies have
focused on the performance in WP, mostly by observing
statistical indicators of the game (7,8) or general performance variables of a game such as throwing velocity, sum
of goals, throw, block, etc. (1). Although valuable with regard to overall explanation of the sport, there remains some
important information that is difficult to gather and often not
statistically analyzed.

TM

Journal of Strength and Conditioning Research

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

In brief, there are athletes on a specific team who may
have important roles, yet do not directly contribute to the
overall statistics of that team. For instance, some players may
not shoot a goal or pass the ball before scoring, who are
contributors of game-related statistics. However, these
athletes may be constantly involved with other specific
and important duties, such as pressing or interrupting the
opponents playmakers, and have specific multitask positions
within the offense, all while never being substituted. Such
indicators are difficult to measure and analyze statistically,
even though they largely contribute to the overall team
success.
Furthermore, game statistics generally favor offensive
performance, such as shooting, passing, and earning exclusions of the opponent. Therefore, the achievement and
performance of those players who are highly involved in
defensive duties are regularly underestimated if observed by
the official statistics of WP. It seems reasonable to conclude
that accepted statistical approaches of WP performance do
not present real-sport situations and therefore have limited
ecological validity. Indeed, authors of the previously cited
studies were evidently aware of these problems and observed
team statistics rather than individual statistics as indicators of
performance (7,8).
These issues are most likely the main reason why there is
evident lack of studies that examined associations between
fitness status and performance in WP. More specifically, the
studies that have been performed so far established differences between playing positions (16,18,29) or between different qualitative levels of WP players (20,25,28). Factors
that directly contribute to specific performance in WP are
rarely reported. There seems to be no previous studies that
have examined the association between strength and conditioning capacities and real-game performance in WP. One
of the likely reasons for this is the limited amount of data
on reliable and applicable tests that will allow accurate
and objective identification of the athletes WP-specific
performance.
The first aim of this study was to construct and evaluate
methods for determining WP-specific offense and defense
performance. Additionally, we aimed to define associations
that exist between sport-specific strength and conditioning
capacities and general anthropometric characteristics, separately for offense and defense. The main hypothesis of the
study was that the strength and conditioning and anthropometric indices would be specifically and differentially
related to offensive and defensive WP performance.

METHODS
Experimental Approach to the Problem

The main rationale for this study was the evident lack of
information of the association that may exist between
strength and conditioning capacities and specific performance in offense and defense during WP game play. The
study intended to determine whether a newly proposed

| www.nsca.com

method to evaluate performance achievement will be related

to WP-specific strength and conditioning capacities.
The experimental approach consisted of several phases.
Throughout the first phase, we consulted 5 WP coaches
and asked them to define the most important performance
indicators of the WP athletes, irrespective of the final
achievement observed with standard game statistics (i.e.,
scores, personal fouls, etc.). The coaches were offered
a wider list of performance indicators that included the
following: agility (mobility), sprinting capacity, endurance
capacity, wrestling (tackle) capacity, pressure over opponent, game intelligence, technical knowledge, tactical
knowledge, playmaking, passing, shooting capacity, polyvalence in game (i.e., ability to play on different positions),
and specific efficacy on players primary playing position
(primary position efficacy). Coaches were asked to rank
performance indicators from the most important to least
important. Specifically, they were asked to make such
a ranking of performance indicators separately for offense
and defense. Within the 5 most important indicators in
offense and defense, 4 of the 5 coaches specified the following components: (a) agility, (b) polyvalence, and (c)
primary position efficacy. Consequently, the set of performance indicators in this study included these 3 variables,
which were analyzed separately for offensive and defensive
performance. In the second phase of the experiment,
anthropometrics and strength and conditioning capacities
were measured, which were considered the independent
variables. At the same time, the players were evaluated
on performance indicators, which were considered the
dependent variables (complete details on evaluation explained below). The third phase of the experiment included
reliability analyses. In the fourth phase, the association
between the independent and dependent variables was
established.
Subjects

The sample of subjects consisted of 82 high-level youth WP

players (aged 1719 years; body height, 186.3 6 6.07 cm;
body mass, 84.8 6 9.6 kg). All players had been trained in
WP for at least 7 years. At the time of testing, they were
participating in 810 training sessions per week (plus 1
game) with each session lasting approximately 2 hours.
Morning training usually consisted of swimming, and gym
and technical-tactical exercising, whereas the afternoon
training consists of tactical exercises.
All subjects were informed of the purpose, benefits, and
risks of the investigation. Subjects voluntarily participated
in the testing after obtaining parents written consent for
those subjects younger than 18 years. Subjects above the
age of 18 years provided their own written informed consent. The study conforms to the Code of Ethics of the
World Medical Association (approved by the ethics
advisory board of the Faculty of Kinesiology, University
of Split).
VOLUME 30 | NUMBER 5 | MAY 2016 |

1317

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Predictors of Game Performance in Water Polo

Procedures

The sample of variables included body height and mass,

5 specific physical fitness tests, and 6 performance-related
tests.
Body height was measured in centimeters by a stadiometer, whereas body mass was measured in kilograms using
a weight scale (Seca Instruments Ltd., Hamburg, Germany).
Water polo fitness tests consisted of the following: (a) WP
sprint swimming over 15 meters, (b) 4 intervals of specific
anaerobic-endurance test consisting of swimming 50 m, (c)
vertical in-water-jump test, (d) semitethered dynamometric
test of isometric force, and (e) test of throwing velocity. The
procedure for each test is described below:
 The sprint swimming started on a sound signal and the
subjects were not allowed to push off the pool wall. The
WP crawl position was performed during the test with
the head remaining out of the water throughout and the
athlete sprinted over a 15 m distance (28). A Longines
(Saint-Imier, Switzerland) swimming timing apparatus
was used. The best of 3 trials was retained as a final
result for each participant.
 During the anaerobic-swimming test, each subject
swam at their maximum exertion over a 50-m distance
4 times with a 30-second recovery period between each
interval. As a final result, the average time of the four
50-m trials was used (15).
 The in-water-jump test was measured as 1-arm vertical
thrust. The participant started from a standard WP
defensive position, using previously described procedures (28). Maximal reach height over 3 testing trials
was measured. After reliability analyses, the best score
was retained as a final result for each athlete.
 The semitethered dynamometric test of isometric force
consisted of maximum intensity upright swimming
using an eggbeater kick with an elastic band fastened
to a special belt (29). Swimming force was recorded
using a tensiometric dynamometer coupled to
a MAX-5 device (JBA Staniak, Poland) through
a WTP 003 amplifier and Max_5.1 computer software.
The subjects were instructed to perform the eggbeater
kick as hard as possible and to achieve the maximal
possible dragging force over 10 seconds. The maximal
force achieved throughout 3 testing trials was recorded
and included in data analysis.
 Throwing velocity was measured by a drive-shot test.
This test involved each subject lifting the ball from the
water and shooting it maximally into a 60 3 60-cm
square goal net. An investigator was positioned with
a velocity-detecting radar device (Speedster Radar
Gun; Bushnell, Overland Park, KS, USA) behind the
net. Only successful throws that made the 60 3
60-cm square goal were analyzed (28,29). The official
WP ball size 5 (6871 cm in circumference) was used
and the maximal achievement of 3 trials was recorded
as final result.

1318

the

Before testing, each subject performed a standard warmup, which consisted of 1015 minutes of swimming combined with convenient in-pool warm-up exercises such as
thrusts, turns, and intensive strokes. All subjects were familiarized with the ball, distances of the tests, and their design
before tested on each single procedure. The subjects were
allowed 35 minutes rest between each within-test trial and
1015 minutes rest between each test. To avoid diurnal variations, all tests were performed in the morning from 8 AM to
11 AM.
Three variables were evaluated specifically for defensive
and offensive performance (6 variables total) as follows: (a)
polyvalence, (b) primary position efficacy, and (c) agility.
Polyvalence was defined as the ability of the athlete to
accomplish different playing duties while playing at different
playing positions over defense (polyvalence defense) or
offense (polyvalence offense). Primary position efficacy
was defined as ability of the athlete to efficiently perform
the position-specific tasks while playing at his primary playing position during defense (efficacy defense) or offense (efficacy offense). Agility was defined as the ability of the
athlete to reasonably, quickly, and efficiently change the
playing position during defense (agility defense) or offense
(agility offense).
For the purpose of the evaluation of the defense and
offense performance variables, 3 WP experts were interviewed. The interviewed experts were officials of the
National Water Polo Federation and have been involved in
the selection of the National team members (i.e., National
team coaches). Therefore, they were well trained and
understood all tested athletic qualities. Each expert (i.e.,
judge) independently evaluated each subject on all performance indicators using a scale from 1 (poor; player is
absolutely unreliable on a specific task and makes many
mistakes) to 5 (excellent; player is absolutely reliable on
a specific task and makes no mistakes at all). Therefore,
regardless of their main duties and game-tasks, all participating athletes were evaluated on indicators of offensive and
defensive performance. Evaluators observed all players at all
official game during the competitive period. After reliability
analyses, the final score for each subject on each performance indicator was expressed as an average value of all
3 judges. Furthermore, to establish the objectivity of the
evaluation of defense and offense performance variables,
players were additionally evaluated by their team coaches
with binomial criterion. Namely, team coaches were asked
to cluster their players into 2 qualitative groups (high quality
vs. low quality) based on their own experience. More
precisely, each player earned separate scores for his overall
offensive and overall defensive quality.
On the first day of data collection, the subjects were tested
on anthropometrics, in-water-jumping, and sprint swimming.
On the second day, subjects were tested on the dynamometric
force, throwing velocity, and anaerobic endurance. All subjects were tested throughout the period of 1 month.

TM

Journal of Strength and Conditioning Research

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

| www.nsca.com

TABLE 1. Descriptive statistics (mean and SD) and reliability (CA, CV, and ICC) of the studied variables.*

Polyvalence defense (score)

Agility defense (score)
Efficacy defense (score)
Polyvalence offense (score)
Agility offense (score)
Efficacy offense (score)
Sprint swimming (s)
In-water-jump (cm)
Drive shot (km$h21)
Dynamometric force (kg)
Anaerobic swimming (s)
Body height (cm)
Body mass (kg)

Mean

SD

CA

CV

ICC

2.87
2.51
2.11
2.44
2.11
2.52
8.99
144
72.5
34.9
30.0
186.3
84.8

1.59
1.38
0.95
1.41
1.34
1.76
0.50
6
4.55
13.0
1.7
6.1
9.6

0.82
0.81
0.85
0.88
0.95
0.88

0.05
0.06
0.05
0.05
0.03
0.03
0.05
0.02
0.04
0.05

0.82
0.91
0.89
0.84

*CA = Cronbach alpha; CV = coefficient of variation; ICC = intraclass coefficient; Polyvalence defense = athletes ability to
accomplish different playing duties in defense; agility defense = athletes ability to efficiently change the playing position during
defense; efficacy defense = athletes ability to perform his primary position tasks in defense; polyvalence offense = athletes ability
to accomplish different playing duties in offense; agility offense = athletes ability to efficiently change the playing position during
offense; efficacy offense = athletes ability to perform his primary position tasks in offense; sprint swimming = water polo sprint
swimming over 15 meters; in-water-jump = maximal vertical in-water-jump test; drive shot = test of maximal throwing velocity;
dynamometric force = isometric force test consisting semitethered swimming; anaerobic swimming = 4 times 50-meter maximal sprint
with a 30-second recovery; body height = players body height; body mass = players body mass.

Statistical Analyses

The reliability of the evaluation for performance scores was

established by the calculation of Cronbach alpha (CA) and
coefficient of variation (CV) between the scores obtained
from 3 evaluators (12,13).
The intraclass coefficient (ICC) and CV were calculated to
establish the reliability of the fitness testing for all variable
but anaerobic swimming test (12).
To define the differences between the qualitative groups
for the offensive and defensive performance variables,
independent samples t-test was calculated.
The association between the 6 performance variables was
evaluated by Pearsons product-moment correlation.

The main hypothesis of the study was tested by analyzing

the relationships between the studied variables with Pearsons product-moment correlations followed by 6 multiple
regression analyses for 6 criteria performance scores (22).
Because of the large number of subjects and consequently
high df, numerous significant correlations between the independent and dependent variables were expected. Therefore,
only those correlations with a coefficient of higher than 0.71,
which corresponds to 50% of the common variance, will be
interpreted (6). Before the multiple regression calculations,
all independent variables were checked for multicollinearity
by calculating the variance inflation factor or the tolerance
(VIF) (3). The VIF for most of the independent variables was

TABLE 2. Independent sample t-test differences between qualitative groups (high quality vs. low quality) in offensive
and defensive performance variables (data are presented as mean 6 SD).*
Offensive play

Polyvalence (score)
Agility (score)
Efficacy (score)

T-test

Defensive play

T-Test

High quality

Low quality

t-value (p)

High quality

Low quality

t-value (p)

3.11 6 1.80
3.19 6 1.69
3.19 6 1.64

1.80 6 1.69
1.61 6 1.77
1.64 6 1.72

2.06 (0.04)
2.07 (0.04)
3.58 (0.01)

3.28 6 1.68
3.19 6 1.59
3.22 6 1.66

1.68 6 1.72
1.59 6 1.75
1.66 6 1.76

2.01 (0.04)
2.07 (0.04)
2.30 (0.03)

*Polyvalence = athletes ability to accomplish different playing duties in offense or defense; agility = athletes ability to efficiently
change the playing position during offense or defense; efficacy = athletes ability to perform his primary position tasks in offense or
defense.

VOLUME 30 | NUMBER 5 | MAY 2016 |

1319

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Predictors of Game Performance in Water Polo

TABLE 3. Pearsons product-moment correlations coefficients between different performance scores.*

Polyvalence defense

Agility defense

Efficacy defense

Polyvalence offense

Agility offense

0.51
0.59
0.39
0.32
0.42

0.66
0.34
0.71
0.59

0.39
0.56
0.64

0.31
0.48

0.63

Agility defense
Efficacy defense
Polyvalence offense
Agility offense
Efficacy offense

*Polyvalence defense = athletes ability to accomplish different playing duties in defense; agility defense = athletes ability to
efficiently change the playing position during defense; efficacy defense = athletes ability to perform his primary position tasks in
defense; polyvalence offense = athletes ability to accomplish different playing duties in offense; agility offense = athletes ability to
efficiently change the playing position during offense; efficacy offense = athletes ability to perform his primary position tasks in offense.
Denotes significance of p # 0.05.

appropriate (i.e., lower than 2.5). However, the VIF of the

dynamometric force was 5.4 and therefore, this independent
variable was not included as a predictor in the multiple
regression calculations.

RESULTS
The reliability of the offensive and defensive performance
indicators showed appropriate CA values from 0.81 for
agility defense up to 0.95 for agility offense. The CV values
showed relatively small within-subject variations when
athletes offensive and defensive performances were evaluated by different judges, ranging from 3% for efficacy offense
to 6% for agility defense (Table 1).
The CV and ICC of physical fitness tests showed good
reliability of the measurements, with ICC ranging from
0.82 to 0.91 and relatively small within-subject variations

(CV: 25%). The strongest reliability was obtained for

in-water-jump (CV of 2% and ICC of 0.91), and throwing
velocity (CV of 4% and ICC of 0.89). Somewhat lower
reproducibility was noted for sprint swimming (CV of 5%
and ICC of 0.82) (Table 1).
Independent sample t-test showed significant differences between qualitative groups of players in all observed
performance indicators for defense and offense. The highquality group achieved significantly higher scores than
low-quality group. Therefore, the objectivity of the evaluation for the performance indicators was confirmed
(Table 2).
The correlation coefficients between the 6 performance
indicators showed that different performance variables should
be observed as generally independent qualities. Namely,
although all correlations were significant (p # 0.05), only

TABLE 4. Pearsons product-moment correlation between independent and dependent variables.*

Body height (cm)

Body mass (kg)
Sprint swimming (s)
Anaerobic swimming (s)
In-water-jump (cm)
Drive shot (km$h21)
Dynamometric force (kg)

Polyvalence
defense
(score)

Agility
defense
(score)

Efficacy
defense
(score)

Polyvalence
offense
(score)

Agility
offense
(score)

Efficacy
offense
(score)

0.17
0.11
20.35
20.53
0.52
0.32
0.09

20.34
20.59
20.81
20.84
0.71
0.43
0.06

0.19
20.20
20.75
20.71
0.76
0.33
0.48

0.01
0.07
20.65
20.53
0.49
0.47
0.31

20.42
20.58
20.76
20.77
0.35
0.32
0.01

0.01
20.33
20.58
20.67
0.69
0.50
0.37

*Polyvalence defense = athletes ability to accomplish different playing duties in defense; agility defense = athletes ability to
efficiently change the playing position during defense; efficacy defense = athletes ability to perform his primary position tasks in
defense; polyvalence offense = athletes ability to accomplish different playing duties in offense; agility offense = athletes ability to
efficiently change the playing position during offense; efficacy offense = athletes ability to perform his primary position tasks in offense;
body height = players body height; body mass = players body mass; sprint swimming = water polo sprint swimming over 15 meters;
anaerobic swimming = 4 times 50-meter maximal sprint with a 30-second recovery; in-water-jump = maximal vertical in-water-jump
test; drive shot = test of maximal throwing velocity; dynamometric force = isometric force test consisting semitethered swimming.
Denotes significance of p # 0.05.

1320

the

TM

Journal of Strength and Conditioning Research

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

| www.nsca.com

TABLE 5. Multiple regressions for the defensive performance.*

Polyvalence defense
b
Intercept
Body height (cm)
Body mass (kg)
Sprint swimming (s)
Anaerobic endurance (s)
In-water-jump (cm)
Drive shot (km$h21)
R
R2
SE
p

20.08
0.02
0.09
20.23
0.21
0.03
0.30
0.09
0.73
0.27

Agility defense

20.44
20.01
0.00
0.13
20.10
0.03
0.01

0.91
0.60
0.88
0.61
0.14
0.12
0.81

b
20.21
20.20
20.17
20.26
0.37
0.00
0.63
0.40
0.60
0.01

Efficacy defense
b

6.78
20.03
20.02
20.25
20.11
0.04
0.00

0.04
0.06
0.10
0.22
0.04
0.00
0.98

0.02
20.18
20.14
20.20
0.33
20.10
0.50
0.25
0.67
0.01

0.53
0.00
20.01
20.21
20.09
0.04
20.02

0.89
0.88
0.17
0.35
0.17
0.01
0.42

*Polyvalence defense = athletes ability to accomplish different playing duties in defense; agility defense = athletes ability to
efficiently change the playing position during defense; efficacy defense = athletes ability to perform his primary position tasks in
defense; b = standardized regression coefficient; B = nonstandardized regression coefficient; p = level of significance; intercept =
interception regression coefficient; body height = players body height; body mass = players body mass; sprint swimming = water polo
sprint swimming over 15 meters; anaerobic endurance = swimming test consisting of 4 intervals swimming 50 m; in-water-jump =
maximal vertical in-water-jump test; drive shot = test of maximal throwing velocity; R = multiple correlation; R2 = common variance.

(r = 20.77), and efficacy defense (r = 20.71). The negative

values of the coefficients actually present positive influence
of the sprint swimming and anaerobic endurance on game
performances because both conditioning capacities (i.e.,
sprint swimming and anaerobic endurance) are represented
as time (i.e., in seconds). The in-water-jump significantly

agility performances in offense and defense shared 50% of the

common variance (Table 3).
Sprint swimming significantly correlated with agility
defense (r = 20.81), agility offense (r = 20.76), and efficacy
defense (r = 20.75). Anaerobic endurance significantly
correlated with agility defense (r = 20.84), agility offense

TABLE 6. Multiple regressions for the offensive performance.*

Polyvalence offense
b
Intercept
Body height (cm)
Body mass (kg)
Sprint swimming (s)
Anaerobic endurance (s)
In-water-jump (cm)
Drive shot (km$h21)
R
R2
SE
p

0.13
0.13
20.23
20.02
0.20
0.09
0.37
0.13
0.74
0.07

Agility offense

3.30
20.03
0.01
20.33
20.01
0.02
0.01

0.39
0.13
0.36
0.16
0.90
0.13
0.49

b
20.24
20.13
20.38
20.17
0.11
20.03
0.59
0.35
0.59
0.01

Efficacy offense
b

11.42
20.03
20.01
20.54
20.07
0.01
0.00

0.00
0.08
0.31
0.01
0.21
0.33
0.77

20.07
20.22
0.08
20.25
0.31
0.10
0.48
0.21
0.69
0.01

21.05
20.01
20.02
0.12
20.11
0.04
0.02

0.78
0.62
0.11
0.62
0.09
0.02
0.41

*Polyvalence offense = athletes ability to accomplish different playing duties in offense; agility offense = athletes ability to
efficiently change the playing position during offense; efficacy offense = athletes ability to perform his primary position tasks in
offense; b = standardized regression coefficient; B = nonstandardized regression coefficient; p = level of significance; intercept =
interception regression coefficient; body height = players body height; body mass = players body mass; sprint swimming = water polo
sprint swimming over 15 meters; anaerobic endurance = swimming test consisting of 4 intervals swimming 50 m; in-water-jump =
maximal vertical in-water-jump test; drive shot = test of maximal throwing velocity; R = multiple correlation; R2 = common variance.

VOLUME 30 | NUMBER 5 | MAY 2016 |

1321

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Predictors of Game Performance in Water Polo

correlated with agility defense (r = 0.71) and efficacy defense
(r = 0.76) (Table 4).
The multiple regressions calculated for defensive performance indicators reached statistical significance for agility
defense (R2 = 0.40; p , 0.01) and efficacy defense (R2 = 0.25;
p , 0.01). The significant beta weight for agility defense was
evidenced for anaerobic endurance (b = 20.26; p , 0.05)
and in-water-jump (b = 0.37; p , 0.01). The in-water-jump
was the single significant predictor of efficacy defense (b =
0.33; p # 0.01) (Table 5).
For the offensive performance, the significant models were
obtained for agility offense (R2 = 0.35; p , 0.01) and efficiency offense (R2 = 0.21; p , 0.01). The beta weights were
significant for sprint swimming (b = 20.38; p # 0.01; for
agility offense) and in-water-jump (b = 0.31; p # 0.05; for
efficacy offense) (Table 6).

DISCUSSION
This study aimed to identify specific fitness variables that
were associated to indicators of defense and offense performance in youth WP athletes. With regard to main study
hypothesis, several important findings will be highlighted.
First, anaerobic endurance, sprint swimming, and in-waterjump were found to be the most important factors of
performance in offense and defense. Second, strength and
conditioning capacities are to some extent differentially
associated with offensive and defensive performance. Third,
the players ability to perform multiple WP tasks and efficiently play on different positions in offense and defense was
not significantly associated with the studied physical fitness
capacities. Before discussing those findings, the reliability of
the proposed method in defining the WP offensive and
defensive performance will be briefly described.
One of the main rationales for this study was due to the
limited information related to defining true game performance in WP. Although the commonly used statistical
and performancerelated methods in WP allow identification of efficacy for the entire team, less emphasis has been
placed on defining the real-game performance of individual
players (7). Therefore, the results of appropriate reliability of
the evaluation-method proposed in this study are highly
encouraging. The level of consistency was most likely due
to the evaluators being highly objective and familiar with
each athletes qualities, which was additionally confirmed
by independent criterion (i.e., evaluation of the players overall quality by their team coaches).
Familiarity is known to be important for the reliability of
testing (21,23). Therefore, the fact that the evaluators followed
each athlete throughout the previous competitive season, and
watched them over all games, allowed them to be well aware
of their specific qualities, which consequently led to accurate
and reliable scoring. It is also important that the judges
avoided subjectivity. Because the team of evaluators consisted
of 3 coaches from the National team, they were well aware of
necessity and importance of objective evaluation.

1322

the

Water polo is a highly intensive sport with significant

anaerobic energy metabolism, as blood lactate levels range
from 5.3 to 11.2 mmol$L21 during a game depending on
playing position (19). However, there is a lack of studies to
directly relate anaerobic swimming capacity to WP performance, most likely because of the exhaustive nature of the
tests. Therefore, the current results of direct associations
between anaerobic capacity and performance in WP are
novel. However, the finding that anaerobic endurance is
similarly important for both offense and defense is consistent
with previous research. For instance, a recent study showed
no significant difference in anaerobic physiological load for
players involved in offensive and defensive duties, as blood
lactate levels comparably ranged from 4.5 to 6.4 mmol$L21
for both (5).
Anaerobic endurance is highly emphasized in agilitydefensive duties, especially as it relates to power play.
For instance, after committing a major foul, a WP player is
ejected from the game for 20 seconds. This results in a power
play with the opponent having offensive control with an
extra man. Efficiency during a power play situation is one
of the most important elements of overall success in a WP
game (8). When power play occurs, the defense needs to
quickly cover the offense, paying particular attention to the
athletes who are closest to the goal. Therefore, defensive
agility is highly challenged. In addition, because the defense
is 1 man short during a power-play occurrence, the players
have to cover a larger space. During this situation, the defensive players are not allowed to rest in between the highintense bouts of exertion, which directly contributes to the
importance of anaerobic endurance in defensive agility.
The different types of in-water-jumps are among the most
important fitness capacities of modern WP (28). The importance in defensive agility is highly specific because the
in-water-jump allows a defensive athlete to quickly cover
their immediate playing area. Because of this, the in-waterjump is far more efficient than swimming for covering short
distances. In real-sport situations, advanced jumping ability
provides an athlete with superior agility when it is necessary
to switch between adjacent playing positions and take over
the offensive.
This is one of the only studies that observed the fitness
capacities that were directly and differentially related to
offensive and defensive performance in the sport of WP. The
results to some extent confirmed the hypothesis that the
observed independent variables will be differentially associated with offensive and defensive performance. This is
mostly evident when predicting the efficacy of a player
while engaged in the primary position during the game.
Defense efficacy is positively associated to in-water-jump,
but offensive efficacy is associated with sprint swimming.
Although the importance of the in-water-jump during
defense is already discussed, the following text will focus
on sprint swimming and its importance specifically in
offense.

TM

Journal of Strength and Conditioning Research

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

the

TM

Journal of Strength and Conditioning Research

Efficient transition between defense and offense is one of
the crucial elements in team sports (4,11). Although swimming is relatively a slow locomotion-form compared with
running and skating in other sports, the transition between
offense and defense is uniquely problematic in WP. Consequently, excellent sprint-swimming ability puts an offensive
player in a noticeably superior position over their opponent,
whereas this is not so apparent for defensive players. This is
mainly emphasized by the orientation of offensive and
defensive athletes during a game of WP. For instance, during
the WP game, turn overs occur for a number of reasons such
as an offensive foul, mistake made while passing, successful
defensive block, or goalkeeping. In these conditions, the athletes playing offense must quickly react and swim back to
their defensive positions. To achieve maximal sprinting,
those athletes who were in offense until the moment of turn
over must first quickly turn to the opposite direction and
then start to sprint maximally. Consequently, if the athlete
is not able to quickly change his/her position from facing the
opponents goal to facing their goal, sprinting ability would
not matter. Directional change while swimming is very difficult and relatively slow when compared to performing
a similar maneuver on dry land (27). However, when a turn
over occurs, the player who was in defense in that particular
moment must sprint to offense without the need to turn in
the opposite direction, and advanced sprinting capacity puts
him in noticeable superior position.
The players polyvalence in offense and defense was not
related to the independent variables. Although we have studied about youth athletes, there is a certain possibility that their
foundational strength and conditioning capacities were not
sufficiently developed at the time of testing. However, there
is another possibility for the lack of an association between the
studied independent variables and polyvalence. Previous studies noted significant differences between WP playing positions
with regard to specific body build and various strength and
conditioning capacities (15,16,29). In general, centers (both
points and center forwards) are the tallest and heaviest athletes, with center forwards having the highest body mass index
and body fat percentage (10,15,16). Drivers are the shortest
athletes and possess superior sprint-swimming ability, whereas
points possess the greatest level of the strength and conditioning capacities (10,15,16). Such differences are partially explained by sport selection and orientation as children with
higher body masses are oriented toward the center position
(29). However, the differences may also be somewhat related
to position-specific training and conditioning capacities (29).
The subjects of the study were already involved in the sport
of WP for more than 78 years and have specifically oriented
toward a certain playing position, which is common for the
athletes within the studied age range (1719 years). Because of
this, they already possessed certain specific fitness capacities
that allowed them to perform efficiently within their specific
position during the game. Therefore, the lack of an association
between the studied fitness capacities and polyvalence seems

| www.nsca.com

logical. Alternatively, this performance quality is most likely

related to characteristics not studied herein, such as technical
or tactical skills and game intelligence (9).
There are several limitations of the study that should be
noted. First, potentially important conditioning capacities
such as aerobic endurance, strength capacities, precision or
accuracy of shooting, were not observed in the study.
However, the intention was to test all athletes in the shortest
possible period; hence, relatively nontime consuming tests
were chosen. This allowed for optimal testing of the athletes
while in competitive form. Also, the study included players
who are still undergoing growth or maturational changes.
Therefore, the results presented and discussed herein are
probably not generalizable to older athletes. However, the
sample included high-level junior players. Therefore, the results may be generalized to similar samples.

PRACTICAL APPLICATIONS
This study confirmed the importance of anaerobic swimming in specific defensive duties in young WP athletes.
Because the study used relatively simple and applicable
means for evaluating anaerobic endurance, coaches can
easily include the methods into their regular testing routine.
Sprint swimming was found to be strongly related to
offensive performance, whereas in-water-jumping was found
to be strongly related in WP offense and defense. Therefore,
methods to develop these abilities should be specifically
tailored with regard to offensive and defensive duties of WP.
It must be noted that the most superior defensive performance is expected among players who possess advanced
anaerobic endurance and in-water-jumping ability. Strength
and conditioning professionals working with young WP
athletes should be aware of these results. Specific combination
of conditioning capacities should be developed simultaneously
to achieve the most superior expression of performance.
Furthermore, because this study included a relatively large
number of high-level athletes from the country with the
current Olympic champions of WP, the presented results
may be used as normative data aimed at evaluating fitness
capacities of young athletes. The information will be helpful
when designing training programs to improve specific
capacities important for certain duties in a WP game.

ACKNOWLEDGMENTS
Special thanks to all athletes who voluntarily participated in
the study and Croatian Water Polo Federation for their
support and help. The authors declare that they have no
conflict of interest relevant to the content of this article. The
results of this study do not constitute endorsement of the
product by the authors or the National Strength and
Conditioning Association.

REFERENCES
1. Alcaraz, PE, Abraldes, JA, Ferragut, C, Vila, H, Rodriguez, N, and
Argudo, FM. Relationship between characteristics of water polo
players and efficacy indices. J Strength Cond Res 26: 18521857, 2012.
VOLUME 30 | NUMBER 5 | MAY 2016 |

1323

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

Predictors of Game Performance in Water Polo

2. Aleksandrovic, M, Radovanovic, D, Okicic, T, Madic, D, and
Georgiev, G. Functional abilities as a predictor of specific motor
skills of young water polo players. J Hum Kinet 29: 123132, 2011.
3. Allison, PD. Multiple Regression: A Primer. Newbury Park, CA: Pine
Forge Press, 1999.
4. Alvarez, A, Ortega, E, Gomez, MA, and Salado, J. Study of the
defensive performance indicators in peak performance basketball.
Rev Psicol Deporte 18: 379384, 2009.
5. Botonis, PG, Toubekis, AG, and Platanou, TI. Physiological
responses of water-polo players under different tactical strategie.
J Sport Sci Med 14: 8490, 2015.

winning and losing teams in close and unbalanced games. J Strength

Cond Res 28: 210222, 2014.
18. Martinez, JG, Vila, MH, Ferragut, C, Noguera, MM, Abraldes, JA,
Rodriguez, N, Freeston, J, and Alcaraz, PE. Position-specific
anthropometry and throwing velocity of elite female water polo
players. J Strength Cond Res 29: 472477, 2015.
19. Melchiorri, G, Castagna, C, Sorge, R, and Bonifazi, M. Game
activity and blood lactate in mens elite water-polo players.
J Strength Cond Res 24: 26472651, 2010.

6. Buchheit, M, Bishop, D, Haydar, B, Nakamura, FY, and Ahmaidi, S.

Physiological responses to shuttle repeated-sprint running. Int J
Sports Med 31: 402409, 2010.

20. Melchiorri, G, Viero, V, Triossi, T, De Sanctis, D, Padua, E,

Salvati, A, Galvani, C, Bonifazi, M, Del Bianco, R, and Tancredi, V.
Water polo throwing velocity and kinematics: Differences
between competitive levels in male players. J Sports Med Phys
Fitness, 51: 541546, 2014.

7. Escalante, Y, Saavedra, JM, Mansilla, M, and Tella, V. Discriminatory

power of water polo game-related statistics at the 2008 Olympic
Games. J Sport Sci 29: 291298, 2011.

21. Ozkaya, O. Familiarization effects of an elliptical all-out test and the

wingate test based on mechanical power indices. J Sport Sci Med 12:
521525, 2013.

8. Escalante, Y, Saavedra, JM, Tella, V, Mansilla, M, GarciaHermoso, A, and Dominguez, AM. Differences and discriminatory
power of water polo game-related statistics in men in international
championships and their relationship with the phase of the
competition. J Strength Cond Res 27: 893901, 2013.

22. Sattler, T, Sekulic, D, Hadzic, V, Uljevic, O, and Dervisevic, E.

Vertical jumping tests in volleyball: Reliability, validity, and playingposition specifics. J Strength Cond Res 26: 15321538, 2012.

9. Falk, B, Lidor, R, Lander, Y, and Lang, B. Talent identification and

early development of elite water-polo players: A 2-year follow-up
study. J Sports Sci 22: 347355, 2004.
10. Ferragut, C, Abraldes, JA, Vila, H, Rodriguez, N, Argudo, FM, and
Fernandes, RJ. Anthropometry and throwing velocity in elite water
polo by specific playing positions. J Hum Kinet 27: 3144, 2011.
11. Foretic, N, Rogulj, N, and Papic, V. Empirical model for evaluating
situational efficiency in top level handball. Int J Perf Anal Spor 13:
275293, 2013.
12. Hopkins, W. A New View of Statistics, 2014. Available at: http://
www.sportsci.org/resource/stats/. Aaccessed March 21, 2015.
13. Huck, SW. Reading Statistics and Research. Boston, MA: Pearson, 2012.
14. Idrizovic, K, Uljevic, O, Spasic, M, Sekulic, D, and Kondric, M. Sport
specific fitness status in junior water polo playersPlaying position
approach. J Sports Med Phys Fitness 55: 596603, 2015.
15. Kondric, M, Uljevic, O, Gabrilo, G, Kontic, D, and Sekulic, D.
General anthropometric and specific physical fitness profile of
high-level junior water polo players. J Hum Kinet 32: 157165, 2012.
16. Lozovina, M, Durovic, N, and Katic, R. Position specific
morphological characteristics of elite water polo players. Coll
Antropol 33: 781789, 2009.
17. Lupo, C, Condello, G, Capranica, L, and Tessitore, A. Womens
water polo world championships: Technical and tactical aspects of

1324

the

23. Sekulic, D, Krolo, A, Spasic, M, Uljevic, O, and Peric, M. The

development of a new stopngo reactive-agility test. J Strength Cond
Res 28: 33063312, 2014.
24. Smith, HK. Applied physiology of water polo. Sports Med 26:
317334, 1998.
25. Tan, FH, Polglaze, T, Dawson, B, and Cox, G. Anthropometric and
fitness characteristics of elite Australian female water polo players.
J Strength Cond Res 23: 15301536, 2009.
26. Tsekouras, YE, Kavouras, SA, Campagna, A, Kotsis, YP, Syntosi, SS,
Papazoglou, K, and Sidossis, LS. The anthropometrical and
physiological characteristics of elite water polo players. Eur J Appl
Physiol 95: 3541, 2005.
27. Tucher, G, Castro, FAD, Garrido, ND, and da Silva, A. The
reliability of a Functional agility test for water polo. J Hum Kinet 41:
181190, 2014.
28. Uljevic, O, Esco, MR, and Sekulic, D. Reliability, validity, and
applicability of isolated and combined sport-specific tests of
conditioning capacities in top-level junior water polo athletes.
J Strength Cond Res 28: 15951605, 2014.
29. Uljevic, O, Spasic, M, and Sekulic, D. Sport-specific motor fitness
tests in water polo: Reliability, validity and playing position
differences. J Sport Sci Med 12: 646654, 2013.
30. Vila, H, Abraldes, JA, Alcaraz, PE, Rodriguez, N, and Ferragut, C.
Tactical and shooting variables that determine win or loss in toplevel in water polo. Int J Perform Anal Sport 11: 486498, 2011.

TM

Journal of Strength and Conditioning Research

Copyright National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.