Knowledge of the characteristics of elite water polo players can be used by coaches in appropriate selection and preparation methods. Several studies have measured the fitness and physiological characteristics of male players (1,3,14,17,21,25), but information about their female counterparts is scarce (12). Previously, studies have used laboratory-based cycle, run, and swim ergometer tests to assess aerobic and anaerobic capabilities of players (1,9,12), whereas others have used various swimming tests that may not necessarily be specific to the game (e.g., 200-m step tests; 6 × 30 m; 2 × 400 m; 4 × 50 m) (3,15,21,25). Thus, the validity and relevance of these tests is questionable if they are not performed in the water or do not simulate the specific nature of the sport.
Existing literature suggests that there are moderate demands on all 3 energy systems during a water polo match, which typically lasts about 1 hour (24). A well-developed aerobic system has been shown to be an important requirement for elite players (6,16). Consequently, Rechichi et al. (19) developed the multistage shuttle swim test to assess the aerobic fitness of water polo players in the field. This test is based on the multistage shuttle run test developed for land-based intermittent-type sports (10). Additionally, previous match analysis studies have shown that speed over a short distance (<10 m) is of prime importance during the game (11,20), and the ability to elevate the body out of the water is a crucial skill for the purpose of passing, shooting, and defending (22).
Positional differences exist in the skills and abilities required of field players (24). Research on match demands has reported differences between the 2 main positional groups, namely, center and perimeter players (4,23). Specifically, center players spent more time in contact with opponents (i.e., wrestling), and perimeter players performed more sprint swimming (4,23). Differences in match demands between positions may be reflected in distinct physical and fitness characteristics of the players.
To date, there is a lack of research on the physical attributes of contemporary elite female water polo players, both generally (24) and particularly between positions. Such data would assist coaches in profiling players and evaluating adaptations to training. Therefore, the purpose of this study was to investigate the anthropometric and fitness characteristics of elite female water polo players and examine the differences in characteristics between players of different competition levels and playing positions.
Experimental Approach to the Problem
This study investigated the anthropometric and fitness characteristics of elite national- and international-level female players, who were members of an Australian National Water Polo League Club and the Australian Women's National Squad, respectively. The National Water Polo League is the premier domestic competition in Australia, and the National Squad was the second ranked national team in the world at the time of this study. Testing was conducted during the mid season of the National Water Polo League competition for the National League players and at the end of the general preparation phase for the National Squad players. The 2 main positional groups examined were center and perimeter players. All subjects performed measurements of standard anthropometry, lower-body muscular power, speed, and aerobic fitness. Tests were selected and modified based on the specificity and reliability of the protocols.
Twelve female National League players (center: n = 4 and perimeter: n = 8) from 1 club and 14 players from the Australian Women's National Squad (center: n = 5 and perimeter: n = 9) participated in the study. No goalkeepers were involved. The University of Western Australia's Human Research Ethics Committee approved the study design, and written informed consent was obtained from all players who were familiarized with the study procedures before actual testing.
Fitness Testing Battery
All players were in good physical condition at the time of testing. The order of tests was as follows: standard anthropometry, in-water vertical jump, 10-m maximal sprint swim, and multistage shuttle swim test. Anthropometric measurements were obtained in one session, and all other tests were conducted in another session on a separate day. Test-retest reliability was evaluated with all the National Squad players. Players performed the in-water vertical jump, 10-m maximal sprint swim, and multistage shuttle swim test on 2 occasions, one week apart. Tests were performed at the same time of the day on each occasion to minimize diurnal variations. Additionally, as time was determined from video recordings, video analysis for one 10-m maximal sprint swim trial for each player was performed twice to establish the intratester reliability. Water temperature of the pool was kept at 27.3 ± 0.3° C during all tests. Players were instructed to refrain from strenuous exercise on the day before testing and to avoid smoking and drinking alcohol, tea, and coffee on the day of testing. They were also asked not to exercise in the 3 hours leading up to the test and consume their normal pretraining diet, which was standardized for each testing session.
Body mass was recorded using a calibrated digital scale to the nearest 0.1 kg (A&D, Tokyo, Japan). Standing height was measured with a stadiometer to the nearest 0.1 cm (Holtain, Pembrokeshire, UK). Skinfold measurements were taken by a level 2 certified anthropometrist with the International Society for the Advancement of Kinanthropometry using Harpenden Skinfold Calipers (British Indicators, Hertfordshire, UK). In keeping with current recommended standards for athletes, skinfold measurements were taken from the triceps, subscapular, biceps, supraspinale, abdominal, thigh, and calf sites, and the sum of 7 skinfolds (SUM7SF) was reported (13). The anthropometrist's typical error of measurement (TE) and coefficient of variation (CV) for the SUM7SF were 0.6 mm and 0.9%, respectively.
In-Water Vertical Jump
The in-water vertical jump was assessed using methods modified from Platanou (17), who used the board method and a video camera to record the jump height. The player's body had to be immersed to the acromion level on the surface of the water before the jump. The jump height reported was the difference between the highest point reached and the length of the upper limb. In the present study, 2 standard Yardsticks (Swift Performance Equipment, Lismore, New South Wales, Australia) were assembled with a connecting sleeve, extending arm and elbow joint and then positioned overhead (Figure 1). The overhanging vane stack was positioned 1 m (horizontally) from the pool wall. Players assumed a self-selected start position directly beneath the device and jumped as high as possible to displace the vanes. Testing continued until they failed to improve over 2 consecutive trials, with the highest jump recorded for analysis. Jump height was reported in absolute terms (height above water surface level) and also as a percentage of the distance from hip joint (trochanterion) to fingertip (dactylion) with the arm fully raised overhead. This measurement is termed “Hip to Tip” and allows for comparison between players with different trunk and arm lengths. Therefore, relative jump height can be reported as a percentage ([absolute jump height/hip to tip] × 100). A score of 100% indicates that the player raised her hip to water surface level in the jump. Test-retest reliability yielded TE values of 1.6 cm and 1.1% for absolute and relative jump heights, respectively. The associated CV values were 1.2 and 1.1%, respectively.
Ten-Meter Maximal Sprint Swim
The 10-m maximal sprint swim was assessed using methods modified from Rechichi et al. (20). In the cited case study, the subject started with her head 1 m behind the “start” line and a 15-m distance was used with split times obtained every 5 m. Maximum velocity was defined as the average velocity from 0 to 10 m. In the present study, sprints were filmed on digital video (Sony DCR-VX2000E; Tokyo, Japan) at 50 Hz and later analyzed using a commercially available video analysis software package (Dartfish ProSuite, Fribourg, Switzerland). The video camera was positioned on the opposite side of the pool and perpendicular to the swimming path, such that there was a clear view of the start and finish. Before the test, a thick calibration rope was filmed so that virtual lines (at the 0-, 3-, 7- and 10-m marks) could be overlaid onto the video analysis program (Figure 2). During the test, a start rope was attached to an anchor point at the 0-m mark on the opposite side of the pool. On the main side, it was handheld tautly by a tester so that it was level with the water surface. Players started in front of the start rope, with some part of their head touching it. They were allowed to start from either a front-on or side-on position, as long as they maintained contact with the rope. The start line was at least 2 m from the end of the pool to ensure that players could not push off the wall. Players started in their own time. As soon as they started moving, the rope was quickly pulled up to avoid interfering with their kick. They continued swimming as fast as possible until they passed the 10-m marker. Timing commenced when the participant's head left the start line, with split/finish times calculated as the head passed through each of the 3-, 7-, and 10-m marks. Players performed 3 trials, and the fastest scores were recorded. The speed for the 0-3, 7-10, and 0-10 m splits was determined as the acceleration, speed, and combined “sprint” scores, respectively. Test-retest reliability yielded TE values of 0.06, 0.11, and 0.06 seconds for 0-3, 0-10, and 7-10 m, respectively. The associated CV values were 3.1, 1.8, and 3.2%, respectively. Intratester reliability yielded TE values of 0.02, 0.01, and 0.02 seconds for 0-3, 0-10, and 7-10 m, respectively. The associated CV values were 0.8, 0.2, and 1.0%, respectively.
Multistage Shuttle Swim Test
The multistage shuttle swim test was conducted as previously described (19). Briefly, players swam continuously back and forth between 2 lane ropes 10 m (4 lane widths) apart. Each lane rope was at least 2 m away from the nearest wall. A compact disc (10-m Multistage Shuttle Swim Test, Western Australian Institute of Sport and University of Western Australia) emitted a series of audible signals that set the pace. The starting speed was 0.90 m·s−1 and increased by 0.05 m·s−1 every level (~1 minute). At the end of each shuttle, players touched and immediately released the lane rope. If they arrived before the signal, they remained stationary while awaiting the cue for the commencement of the next shuttle. Players were given a warning if they failed to be within one arm stroke of the lane rope as the audio cue sounded and were removed from the test when they failed to be within one arm stroke of the lane rope on 2 consecutive shuttles. The final level and shuttle completed before being eliminated was converted to distance in meters and recorded as the performance score. The same investigator judged the test for all players. Test-retest reliability yielded a TE of 2.7 shuttles and a CV of 4.8%.
Differences between the National League and National Squad players and differences between center and perimeter players were evaluated using independent t-tests. Differences were also analyzed using magnitude-based Cohen's effect size (ES) statistic with modified qualitative descriptors (8). Effect sizes were assessed using these criteria: <0.2 = trivial, 0.2-0.6 = small, >0.6-1.2 = moderate, >1.2-2.0 = large, and >2.0 = very large. Ninety percent confidence limits (±90% CL) were calculated to indicate the precision of the estimate of observed effects. Reliability measures were calculated using a spreadsheet for TE and the CV from log-transformed raw data (7). Statistical significance was set at p ≤ 0.05.
National League Vs. National Squad Players
Table 1 shows the comparison between the National League and National Squad players. Differences between these 2 groups for age and skinfold levels were small to moderate in magnitude but nonsignificant (ES = 0.64 and 0.59, respectively, p > 0.05). Similarly, differences for 7-10 m sprint time and relative jump height were trivial and small in magnitude, respectively, and nonsignificant (ES = 0.12 and 0.48, respectively, p > 0.05). However, the National Squad players were taller and heavier (ES = 0.82 and 1.07, respectively, p < 0.05) and demonstrated greater absolute jump height (ES = 1.57, p < 0.001) and 0-3 m (ES = 0.74, p > 0.05) and 0-10 m (ES = 1.06, p < 0.05) sprint times when compared with the National League players. The largest difference was in their aerobic fitness as measured by the multistage shuttle swim test (ES = 1.92-1.95, p < 0.001).
Center Vs. Perimeter Players
Table 2 shows the comparison between center and perimeter players. Among the National League players, center players were taller (ES = 1.56, p < 0.05) and tended to be heavier (ES = 0.92, p > 0.05) and demonstrated greater absolute jump height (ES = 1.32, p > 0.05) when compared with perimeter players. For the National Squad players, center players were heavier (ES = 2.09, p < 0.001) and had higher skinfold levels (ES = 1.61, p < 0.01) than perimeter players. Additionally, center players demonstrated lower relative jump height (ES = 1.43, p < 0.05) and tended to have slower 0-3 m (ES = 0.83, p > 0.05) and 0-10 m (ES = 1.16, p > 0.05) sprint times and lower multistage shuttle swim test scores (ES = 0.83-0.84, p > 0.05) than perimeter players.
The present study is the first to concurrently investigate the anthropometric and fitness characteristics of elite female water polo players and to compare the differences between different competition levels and playing positions. The results demonstrated significant differences between national- and international-level and center and perimeter players for height, body mass, skinfold levels, jumping, sprinting, and endurance swimming abilities. Additionally, positional differences were greater among international-level players.
In comparison with elite female swimmers (body mass: 64.9 ± 9.0 kg; SUM7SF: 67.6 ± 12.0 mm) (18), the National League players had similar body mass (ES = 0.10, p > 0.05) but significantly higher SUM7SF (ES = 1.90, p < 0.001), and the National Squad players were significantly heavier (ES = 1.14, p < 0.01) and had higher SUM7SF (ES = 1.78, p < 0.001). Higher body mass and skinfold levels of players in the present study when compared with elite female swimmers may represent a physical advantage in terms of buoyancy and physical contact (e.g., wrestling) in water polo.
The present study showed that the National Squad players were significantly taller and heavier than the National League players, with moderate and nonsignificant differences in skinfold levels between these 2 groups. This finding reiterates that body size may play an important role in water polo by enabling players to obtain better position in physical contests and allowing taller players to reach and control the ball (25). In comparison with elite international-level female water polo players from the 1991 World Championships (height: 171.3 ± 5.9 cm; body mass: 64.8 ± 7.2 kg) (5), the National Squad players were similar in height (ES = 0.42, p > 0.05) but heavier (ES = 1.29, p < 0.001). Similarly, when compared with recent data from the Scottish Women's National Squad (height: 168.7 ± 7.9 cm; body mass: 65.9 ± 6.1 kg) (12), the National Squad players were moderately taller (ES = 0.73, p = 0.06) and heavier (ES = 1.22, p < 0.01). Based on their world ranking, the Australian National Squad players are likely to train and compete more regularly at the highest level. Although a direct cause-effect relationship cannot be established, it is possible that greater body mass may present players with some advantage at higher levels of competition.
Among the National League players, center players were taller and heavier than perimeter players but their skinfold levels did not differ. However, among the National Squad players, center players were similar in height but significantly heavier with higher skinfold levels than perimeter players. Height seems to be an important physical attribute for all international-level players. Additionally, center players were markedly bigger than perimeter players as demonstrated by large to very large differences in body mass and skinfold levels. Because center players need to maintain their position in front of goal and are involved in more wrestling activities (4,23), the results of this study demonstrate that body mass is a key attribute to assist them in their role during the game (i.e., generate greater forces). These findings are in agreement with previous analysis of elite international-level players (2).
This study found that the National Squad players had greater absolute jump height than the National League players. This can be attributed primarily to differences in stature between the 2 groups because the difference in their relative jump heights was small and nonsignificant. Due to some important methodological differences between the vertical jump test that was described by Platanou (17) and the one used in this study, comparison of data between these 2 studies cannot be made. The in-water vertical jump test used in this study was thought to be less cumbersome and more reliable. The CV of 1.2% compares favorably with the 3.7-4.6% reported for other jump tests (26). It is also noteworthy that Young et al. (26) showed that the board method recorded significantly lower jump heights when compared with the vane method in the assessment of maximum jump performance on land.
Differences in 0-3 and 0-10 m sprint times between the National League and National Squad players were moderate (although nonsignificant) and large in magnitude (ES), respectively. Difference in 7-10 m sprint times between the 2 groups was trivial and nonsignificant. Accordingly, it suggests that the National Squad players had better acceleration but a similar peak velocity when compared with the National League players.
Large differences between the National League and National Squad players in the multistage shuttle swim test indicate that aerobic fitness was the largest discriminator between these 2 groups. This finding agrees with previous investigators who identified that aerobic fitness is an integral requirement of competitive water polo players (6,16). Using the multistage shuttle swim test, Marrin and Bampouras (12) showed that the Scottish Women's National Squad achieved 330 ± 87 m. This result was markedly lower than those achieved by the National League (449 ± 124 m; ES = 1.11, p < 0.01) and National Squad players (652 ± 84 m; ES = 3.77, p < 0.001) in the present study. It is likely that greater aerobic fitness may significantly contribute to the higher playing standards of the players in the present study.
Among the National League players, differences in fitness characteristics between center and perimeter players were mostly trivial to small in magnitude, except for absolute jump height. Greater absolute jump height in center players may be attributed to their greater stature, as there was no difference in relative jump height between these groups. In contrast, among the National Squad players, differences in fitness characteristics between center and perimeter players were mostly moderate to large in magnitude. Greater relative jump height indicates that perimeter players were able to elevate their hips higher above the water surface level than center players. Additionally, perimeter players were also quicker and had better aerobic fitness as demonstrated by the 10-m maximal sprint swim and multistage shuttle swim test scores, respectively. The latter findings are logical and appropriate for the role of perimeter players because they tend to perform more sprint swimming during the game when compared with center players (4,23).
Besides player selection, the anthropometric and physical performance differences found between the National League and National Squad players could be due to differences in training and competition loads between the 2 groups, thus resulting in larger and leaner body types, and superior fitness in the National Squad players. With more consistent monitoring of players over time, coaches may gain a better understanding of the relationships between the amount and types of workloads and the resulting adaptations.
Differences in characteristics between center and perimeter players were greater among the National Squad than the National League players. This finding is not surprising as one would expect players to be more specialized in their positional roles at higher levels of competition. It is noteworthy that although there are presumably positional differences in skills and abilities, and coaches apply position-specific conditioning exercises, the physiological implications of playing in the center and perimeter positions have not been presented (24). The findings of the present study provide some indication of the differences in physical characteristics between these 2 main positional groups. A systematic analysis of the positional demands of the game would enable prescription of specific training programs for individual athletes (4,24).
In summary, this study showed that some anthropometric and fitness characteristics may discriminate players of different competition levels and playing positions. The National Squad players were taller and heavier and had better jumping, sprinting, and endurance swimming abilities than the National League players. Aerobic fitness was the largest discriminator between these 2 groups. Additionally, perimeter players had lower-body mass and skinfold levels and better jumping, sprinting, and endurance swimming abilities than center players. Body mass was the largest discriminator between these 2 groups. Positional differences in characteristics were greater among international-level players. The battery of tests described in this study can be potentially implemented by coaches to assist in profiling players and evaluating adaptations to training.
The battery of field tests described in this study is reliable and specific to water polo. Coaches may use these tests to assist in profiling players and evaluating adaptations to training. Significant differences in some anthropometric and fitness characteristics between elite national- and international-level players demonstrate the need to consider these physical attributes when identifying potentially talented players (i.e., in addition to players' technical and decision-making skills). Because the positional demands of the game are different (4,23), coaches should apply position-specific conditioning exercises and evaluate players accordingly, in order that players may receive appropriate training stimuli to match the physiological demands of their playing position.
We express gratitude to the players who participated in this study and I. Mujika, N. Etxebarria, C. O'Hagan, and S. Clark for their assistance in data collection.
1. Aziz, AR, Lee, HC, and Teh, KC. Physiological characteristics of Singapore national water polo team players. J Sports Med Phys Fitness
42: 315-319, 2002.
2. Carter, JEL and Ackland, TR, eds. Kinanthropometry in Aquatic Sports: A Study of World Class Athletes
. Champaign, IL: Human Kinetics, 1994.
3. Cazorla, G and Montpetit, RR. Metabolic and cardiac responses of swimmers, modern pentathletes, and water polo players during freestyle swimming to a maximum. In: Swimming Science
. Ungerechts, VB, Wilke, K, and Reischle, K, eds. Champaign, IL: Human Kinetics, 1988. pp. 251-257.
4. D'Auria, S and Gabbett, T. A time-motion analysis of international women's water polo match play. Int J Sports Physiol Perform
3: 305-319, 2008.
5. Drinkwater, DT and Mazza, JC. Body composition. In: Kinanthropometry in Aquatic Sports: A Study of World Class Athletes
. Carter, JEL and Ackland, TR, eds. Champaign, IL: Human Kinetics, 1994. pp. 102-137.
6. Hollander, AP, Dupont, SHJ, and Volkerijk, SM. Physiological strain during competitive water polo games and training. In: Medicine and Science in Aquatic Sports
. Miyashita, M, Mutoh, Y, and Richardson, AB, eds. Basel, Switzerland: Karger, 1994. pp. 178-185.
7. Hopkins, WG. Reliability
from consecutive pairs of trials (Excel spreadsheet). Available at: http://sportsci.org/resource/stats/xrely.xls
. 2000. Accessed December 10, 2007.
8. Hopkins, WG. A scale of magnitudes for effect statistics. Available at: http://sportsci.org/resource/stats/effectmag.html
. 2002. Accessed December 10, 2007.
9. Konstantaki, M, Trowbridge, EA, and Swaine, IL. The relationship between blood lactate and heart rate responses to swim bench exercise and women's competitive water polo. J Sports Sci
16: 251-256, 1998.
10. Leger, LA, Mercier, D, Gadoury, C, and Lambert, J. The multistage 20-m shuttle run test for aerobic fitness. J Sports Sci
6: 93-101, 1998.
11. Lilley, G. A basis for the conditioning of state level water polo players. PELOPS: Studies in Physical Education, Leisure Organization, Play and Sport
3: 25-29, 1982.
12. Marrin, K and Bampouras, TM. Anthropometric and physiological characteristics of elite female water polo players. In: Kinanthropometry X
. Marfell-Jones, M and Olds, T, eds. Oxford, UK: Routledge, 2007. pp. 151-163.
13. Norton, K, Marfell-Jones, M, Whittingham, N, Kerr, D, Carter, L, Saddington, A, and Gore, C. Anthropometric assessment protocols. In: Physiological Tests for Elite Athletes
. Gore, C, ed. Lower Mitcham, South Australia: Human Kinetics, 2000. pp. 66-85.
14. Pavlik, G, Banhegyi, A, Kemeny, D, Olexo, Z, and Petridisz, L. The estimation of water polo players' physical condition by means of a swimming test. The relationship of the swimming test results with the relative aerobic power. Hung Rev Sports Med
42: 129-150, 2001.
15. Pavlik, G, Olexo, Z, and Batovszky, K. The estimation of water polo players' physical fitness by means of a simple swimming test. Hung Rev Sports Med
32: 17-30, 1991.
16. Pinnington, H, Dawson, B, and Blanksby, B. Heart rate responses and the estimated energy requirements of playing water polo. J Hum Mov Stud
15: 101-118, 1988.
17. Platanou, T. On-water and dryland vertical jump in water polo players. J Sports Med Phys Fitness
45: 26-31, 2005.
18. Pyne, DB, Anderson, ME, and Hopkins, WG. Monitoring changes in lean mass of elite male and female swimmers. Int J Sports Physiol Perform
1: 14-26, 2006.
19. Rechichi, C, Dawson, B, and Lawrence, SR. A multistage shuttle swim test
to assess aerobic fitness in competitive water polo players. J Sci Med Sport
3: 55-64, 2000.
20. Rechichi, C, Lyttle, A, Doyle, M, and Polglaze, T. Swimming velocity patterns in elite women's water polo: a case study. Int J Perform Anal Sport
5: 139-148, 2005.
21. Rodriguez, FA. Physiological testing of swimmers and water polo players in Spain. In: Medicine and Science in Aquatic Sports
. Miyashita, M, Mutoh, Y, and Richardson, AB, eds. Basel, Switzerland: Karger, 1994. pp. 172-177.
22. Sanders, RH. A model of kinematic variables determining height achieved in water polo boosts. J Appl Biomech
15: 270-283, 1999.
23. Smith, HK. Physiological fitness and energy demands of water polo; time-motion analysis of field players and goaltenders. In: Proceedings of the Federation Internationale de Natation Amateur (FINA) First World Water Polo Coaches Seminar; May 27-June 3, 1991; Athens, Greece.
24. Smith, HK. Applied physiology of water polo. Sports Med
26: 317-334, 1998.
25. 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: 35-41, 2005.
26. Young, W, Macdonald, C, Heggen, T, and Fitzpatrick, J. An evaluation of the specificity, validity and reliability
of jumping tests. J Sports Med Phys Fitness
37: 240-245, 1997.