International Journal of Performance Analysis in Sport

Published by Informa UK (Taylor & Francis)
Online ISSN: 1474-8185
Print ISSN: 2474-8668
The aim of this study was to design and implement a Player Impact Ranking (PIR) matrix to objectively quantify the impact of players on team performance and game outcomes in rugby union. Game data from the Crusaders, Force and Lions in the 2006, 2007 and 2008 Super 14 competition seasons were used. A framework of game actions (GA) was developed for individual players and allocated positive and negative weightings to reflect each GA's relative importance to a team's winning performance. The weightings were applied to the frequencies of each game action to derive a total net game performance (NGP) score, which was then used to assess the teams' performances, utilising game results (i.e. win/loss), competition points and score margins (i.e. the difference in scores between two teams). NGP scores showed significant relationships with wins/losses, team scores and competition points (p<0.05). Moderate, positive correlations were recorded with NGP scores and score margins (r=0.66-0.76; p<0.01). Results suggested that the derived PIR matrix accurately reflected game results and team performances, making it a useful tool for post match analysis of performance. Further research into applying the PIR matrix developed here to in-match situations for real time analysis is recommended.
Historically, athletes attempting the 800m and 1500m double have been world class athletes with a genuine medal chance in one or both events. However, an athlete would not be expected to attempt the double if this jeopardised their chance of winning a medal in a single event. Therefore, the approach to attempting the double must allow the athlete to race 6 times during a games and recover sufficiently between races. The purpose of the current investigation was to compare the strategies of female athletes attempting the 800m and 1500m double with those entering a single event. A manual race analysis system was devised to analyse the performances of these two types of athlete, finding that the double event athletes preferred to run at the back of the leading group during the early stages. A supporting qualitative analysis revealed that the reason for not leading in the early stages was to preserve energy. The decision to attempt the double had to consider the risks involved and strategy was influenced by the opponents, the order of the events, the athlete's best event and confidence in the athletes ability to take up the challenge of competing in two events.
The present study was designed to examine the offensive tactic of elite women's football teams. The sample of the study constituted of 20 games of the four top(quarter finalists) teams of the 3rd Women's World Cup (USA, China, Brazil, Norway). Totally, 749 offensive phases were analyzed. Teams' offensive patterns during games were coded using the following categories: (1) team, (2) zone of scoring attempt, (3) source of scoring attempt, (4) ball possession and (5) type of passes. Correspondence analysis showed that the four teams had different offensive profile. USA's scoring attempts were mainly executed from the central zones using combinations of small and medium passes with low possession and set plays. China's scoring attempts were mainly performed from the central and left zone after a combination of two or three small and intermediate passes. The source of Brazil's scoring attempts was frequently executed from the box area and the right offensive zone, using combination with small passes and individual actions. The source of Norway's final scoring attempts was mainly crosses and opponents mistake, executed from the box area. In addition, Norway frequently used small possession with long balls beginning from the defend area directed to the opponents' box area.
The differences in playing patterns of soccer teams has been, and remains to be, one of the largest areas of post-match performance analysis within association football. Many studies try to encapsulate the multiple facets of play and consequently only provide limited detail in any one area (Low et al., 2002; Hughes Robertson and Nicholson, 1988). This study aimed to provide a more detailed insight into passing during the 2002 World Cup whereby post event analysis of six group matches was undertaken (3 each for a successful and an unsuccessful team, based on qualification through the group stages, Stanhope, 2001). The resulting passes (>4000) were analysed for start and finish positions (for length of pass), whether they were played to a player or space, played first time or after a dribble and some outcome was recorded e.g.possession lost or retained. The coding process was designed and recorded using a computerised notational analysis software package, the Noldus Observer Video-Pro (Noldus Information Technology, 2001). The data was then transferred to SPSS v11.01 (SPSS Inc.) for statistical analysis (Chi Square). Reliability measures were conducted at the level of the subsequent analysis, as suggested by Hughes, Cooper and Nevill (2002). In this respect all defined performance measures provided good reliability estimates (<5% error) with a further assessment of the time coding process (built into the Observer) measuring the time from when the player received the ball to when the pass was made. This resulted in a low average error of 0.08 seconds but a relatively large spread (± 1.08 s). The passing strategy for the successful and unsuccessful teams did not differ (types of pass employed) suggesting that this performance indicator does not easily discriminate teams at this level. The only apparent difference existed in the pre-defensive area of the field (near the halfway line, James, Mellalieu & Holley, 2002) where the unsuccessful team tended to play significantly more passes. This observed difference was thought suggestive of a difference in total possession in this area rather than indicative of a strategy difference. The findings thus suggested that there was no difference in passing ability between the two teams within this study even though they differed markedly in terms of success in the tournament. Consequently it is suggested that either the criteria used were not sensitive enough to detect differences in passing or the teams were of a similar standard and other factors determined match outcomes.
Summary of the Stability Profile calculations from the IRB Rugby World Cup 2003 performance indicators displayed on the website, for South Africa (SA), England (Eng), Australia (Aus) and New Zealand (NZ).
A retrospective analysis of performance during the Rugby Union World Cup 2003 was conducted by comparing data from the top three nations with data from South Africa (SA). Differences were observed between the four teams in three performance variables; number of penalty kicks and drop goals scored, and percentage possession. England scored more points from penalty kicks and drop goals. SA had less percentage possession than either England or New Zealand. SA's scoring opportunities began more frequently from inside the opposition's territory (86% of all scoring movements). Possession was most frequently lost when the movement began inside SA's territory (55% of all movements in which possession was lost). The duration for which possession was retained was greater for movements that resulted in points (21.9 ± 14.5 s), than where possession was turned over (14.8 ± 5.2 s) (P< 0.05). Points were scored against SA 24 times, 58% in the second half and 71% of the scores against SA occurred the next play after SA had lost possession. These data suggest that superior performance in World Cup rugby is linked to possession retained, the number of points scored in the second half and the propensity to lose possession in areas of the field from which the opposition is likely to score.
The purpose of this study was to find the efficacy values in the playing micro-situations in numerical equality with or without ball possession and to analyze the relation between these and the condition of winner or loser. We analysed the matches of the X World Championship of Water polo which did not end in a draw. Playing micro-situations in numerical equality were evaluated by means of coefficients, which obtain efficacy values. Some differences were revealed, in male category, in the following coefficients with possession: concretion, definition, resolution, accuracy and blocked shots received (p<.001) and precision (p=.001). Without possession: concretion, definition, resolution, accuracy and blocked shots made (p<.001) and precision (p=.001). In female category with the following coefficients with possession: concretion, definition and precision (p<.001), resolution (p=.001), possibility (p=.005) and accuracy (p=.017). Without possession: concretion, definition and precision (p<.001), resolution (p=.001), possibility (p=.005) and accuracy (p=.017); taking as a reference a value of p<.05. To conclude with, we can say that in twelve out of the fourteen efficacy coefficients proposed for evaluating the playing micro-situations in numerical equality with or without ball possession in male and female water polo there are significant differences between the condition of winner or loser.
A comparative study of English, Australian, New Zealand and South African performances in the 2003 Rugby World Cup was performed from a video evaluation of all matches played by the four teams during the tournament. Each team averaged 43 movements per match comprising 20% point scoring, 56% running play and 24% kicked play movements. South Africa completed fewer point scoring (16%) and more kicked play movements (29%). Differences between the mean movement times for the four teams were not significant. All teams had significantly longer time in possession during point scoring movements than in movements that resulted in a turnover (P<0.05). All teams with the exception of England had significantly longer time in possession of the ball during running play than during kicked play movements (P<0.05). There were significant differences between the locations on the field where the teams scored points. England and Australia predominantly scored between the posts, South Africa scored out wide, but New Zealand scored with equal frequency across the try line. This study also found that semi-final teams (England, Australia and New Zealand) transferred 10% more possession from the defending to the attacking half of the field than did South Africa. In summary this study has shown clear differences between teams reaching the semi finals of the 2003 Rugby World Cup and those knocked out at the quarterfinal stage (South Africa). These differences can be summarized as a greater ability to retain the ball for longer and to move possession from the defensive to the attacking half of the field.
Frequency of goal scoring / 45 min. 
Type of offensive style when goal scored. 
Area of materialization of final attempt. 
The aim of the present study was to study the characteristics of goal scoring patterns in top leveled soccer matches. The sample the study constituted of 32 games of the European Championship (Euro 2004). Cross tabulation and chi-square methods were used for the data analysis. The results revealed that more goals were achieved in the second half (57.4%) than in the first half (42.6%, p<0.05). As far as the type of offense concerned, goals achieved through organized offence presented the higher frequency (44.1%) following goals after a set play (35.6%) and counter-attacks (20.3%). Regarding the actions that occurred prior to the goal, long passes presented the higher frequency (34.1%). More specifically, the kind of dead-ball situations was examined, and the conclusion is that corners and free kicks showed bigger frequency of appearance in the game. Finally, regarding the zone of scoring attempt, the following percentages were presented: 44.4% penalty area, 32.2% goal area, and 20.4% outside the penalty area. The results reveal that coaches should focus on train of the dead-ball situations. Also attention must be given to the fatigue that players appear towards the end of a game, which consequently leads to goal scoring by the opponent team, and to its confrontation through training.
While Beach Volley fares practically the same popularity as Volleyball, its tactical and technical elements have not as yet been submitted to the same systematic analysis. The aim of the present study was to compare the most important technical elements of Beach Volley between men and women, as this was expressed in its highest level at the final games of the Athens 2004 Olympics, using standard methodological approach, previously applied in the classic game of Volleyball. The analysis of the results revealed that the game of Beach Volley is played quite differently by men and by women. Although the relative proportion of points won by the serving and receiving team are practically the same in both genders, this comes as a result of actions, which vary distinctly in power in all skills, starting from the jump power serve, to the smashing spike at the net, forcing one of the players of the opposing team to make a block defense. Consequently the evident difference in muscular power between the two genders is the major factor discriminating the game tactics of Beach Volley. This differentiating factor manifests itself in several sports, such as volleyball, handball and tennis.
The study compares the effectiveness of the five principal skills in men's Volleyball (serve, reception, attack, block and dig) between the Sydney 2000 and the Athens 2004 Olympic Games and examines to what extent the observed changes are connected with the implementation of the new rules in Volleyball. The findings revealed a universal tendency of the elite men's volleyball teams to enhance their defence by reducing their block and dig faults. On the contrary there was an increase of the reception faults as a result of the improvement of the service effectiveness. The above changes reflect the teams' shift of tactics to win more points from their own serve. This tactical development became more imperative after the introduction of the rally-point system, which reduced the number of points played per set, and hence the teams' opportunities of gaining the minimum two-point advantage in order to win the set. Nonetheless, the Athens gold-medalist team of Brazil showed, in addition to the above, remarkable reception effectiveness which led to an outstanding attack capability, thus reestablishing the attack as the most important skill in volleyball
The British Olympic Association's Biomechanics Steering Group was redefined, over a period of time, to include notational analysis, and later motor control, to become the Performance Analysis Steering Group. Where are we now with Performance Analysis? As this Steering Group reaches the end of its natural span, let us examine how notational analysis has changed over the last 20 years, and how developments over the last 5 years have materially effected this area of Performance Analysis. Current technology has the potential of bringing us all, biomechanists, notational analysts and motor control scientists, together but it is sometimes difficult to explore a new modus operandi, especially away from the safe confines of a laboratory. This could be a unique opportunity to implement all the promise offered by digital technology. The challenge to the new structures in the BOA support mechanisms is then not just to enable, but also to prompt these crucial areas of sports science support to work closely together.
The South African regional teams participating in the Super 12 rugby union competition seem to under-perform compared to the performance of the national team in the Tri-Nations competition against similar Australasian opposition. The aim of this study was to compare various performance parameters between the four South African Super 12 teams (Bulls, 3rd position; Stormers, 9th position; Cats, 11th position; and Sharks, 12th position) and the eventual winners, the Crusaders from New Zealand. Nine games from the 2005 Super 12 season were analysed. These nine games included the four South African teams playing against the Crusaders and against each other (one game, Stormers v Sharks, was lost due to technical problems). All game analyses were performed after the tournament using a digital analysis software program (Sportscode Elite version 5.4.24, Sportstec, Australia). Ball possession, tries scored and various performance indicators associated with successful ball retention and attacking effectiveness in the tackle situation (as obtained through a panel of elite coaches and analysts) were quantified. There were no statistically significant differences between the teams for the total amount of ball possession per match (Crusaders 1057±216s, Bulls 1048±158s, Cats 852±73s, Sharks 1078±84s and Stormers 984±186s) or time of each movement involving ball possession (Crusaders 15.2±3.0s, Bulls 15.5±1.9s, Cats 13.0±1.0s, Sharks 16.9±2.3s and Stormers 15.0±3.8s). There were no significant differences between the number of tries scored (Crusaders 6±4, Bulls 4±3, Cats 3±1, Sharks 3±2 and Stormers 2±1 tries), number of offloads (Crusaders 19±12, Bulls 12±4, Cats 13±5, Sharks 15±4 and Stormers 17±5 offloads), turnovers won (Crusaders 3±3, Bulls 4±2, Cats 5±1, Sharks 2±1 and Stormers 5±1 turnovers) or conceded (Crusaders 4±1, Bulls 3±2, Cats 4±2, Sharks 5±1 and Stormers 6±2 turnovers). There were no statistically significant differences between the number of times the different teams committed 0, 1, 2 or 3 support players when they were tackled. There were also no statistical differences in the number of times the various teams committed 0, 1, 2 and 3 support players to counter-ruck once they had made a tackle. Due to the relatively small sample size, there is a risk of making a type II error (i.e. missing significant differences) between the variables analysed. However when these data are inspected using a visually striking box and whisker plot, there are noticeable differences between the styles of play of the teams. Future studies should investigate whether this form of data analysis is valid, particularly when having to give feedback to coaches of elite teams, involving analyses in which the sample size may always lack statistical power.
This paper explains the effects of performance and chance in the results obtained by teams in the matches of the FIFA World Cup Germany 2006. The research is based on linear regression analysis, mean comparison test and logit multinomial. Results point out that performance is a relevant variable for explaining the points obtained by teams in the first round of the competition, when the system of play is the league system. However, in the second round of the competition, when the system of play is the knockout system, the role of performance is less important: there are not statistically significant differences in the performance obtained by winners and losers.
General game variables (± SD) of winning and non-winning teams.
Bowling variables (± sd) of winning and non-winning teams
Knowledge of the relative importance of team performance indicators in cricket helps determine team strategy and tactics. We analysed team, batting and bowling performances at the 2007 ICC Cricket World Cup Tournament to determine the magnitudes of differences between winning and losing teams. We compared the magnitudes of differences in key batting and bowling indicators between the qualifying Round Robin and the final Super 8 games. Magnitude of difference between teams was established with a standardised (Cohens) effect size (ES) with 90% confidence limits. The difference in performance indicators between winning and losing teams were smaller in the later (Super 8) stages of the tournament. The two performance indicators most highly correlated with winning in the Super 8 stage were taking wickets (ES=1.79±0.04 90%CL) and run rate (ES=1.39±0.02 90%CL). Hitting sixes had a greater influence during the earlier stage of the tournament; while bowling maiden overs was more important as the tournament progressed. The main contribution of this paper is that winning teams capture more wickets and have more 50-plus partnerships while maintaining a higher run rate primarily through hitting a higher percentage of runs in boundaries. Team selection should exploit these performance factors through selection of players most capable of delivering these objectives.
Magnitude of difference (reported as effect size ± 90% CL) in selected performance variables between the winning and losing teams during the 2008 IPL. The positive effect size values indicate the winning team had a more favourable outcome, while the negative effect size values indicate the losing team were better.
Team strategy and tactics in cricket can be influenced by knowledge of the relative importance of team performance indicators. We analysed team, batting and bowling performances at the 2008 Indian Premier League Twenty/20 Tournament (IPL) by comparing the magnitudes of differences in key batting and bowling indicators between winning and losing teams. Magnitude of difference between teams was established with a standardised (Cohen's) effect size (ES:0.2) with 90% confidence limits. There were several moderate or large differences in performance indicators between winning and losing teams in the IPL. The three best indicators of success in the IPL were taking more wickets in the game (ES=1.93±0.04 90%CL), taking more wickets in the last 6 overs (ES=1.01±0.03) and having a higher run rate (ES=0.96±0.02). Winning teams captured more wickets particularly in the first and last six overs, and were more effective in containing the opposition teams' run scoring in the middle eight overs. When batting, winning teams tended to face less dot balls and score more runs from 25+ run partnerships. Team tactics should focus on wicket-taking bowling and field placements in the first and last six overs, and run restrictive field placing and bowling in the middle eight overs.
Batting variables (±SD) of winning and non-winning teams.
Bowling variables (±SD) effecting the outcome of T20 cricket.
Fielding variables(±SD) effecting the outcome of T20 cricket.
Cricket has evolved in recent years and has resulted in the emergence of Twenty20 cricket. We examined the batting, bowling and fielding variables associated with success in cricket in the recent Twenty20 World Cup. We compared several key batting and bowling variables of winning and non-winning teams by comparing the magnitudes of differences (Cohen's effect size). We established several moderate or large differences between winning and losing teams with respect to batting, bowling and fielding variables. The best indicators of success in the tournament can be broken down into general match, batting and fielding variables. The top 5 indicators for success in the tournament were losing less wickets in the game (ES= −1.66), losing less wickets in the powerplay while batting (ES= −1.22), scoring more runs per over (ES= 1.23), scoring more runs in the middle eight overs (ES= 0.86) and bowling more dot balls (ES= 1.15). Thus it could be concluded that for overall success in Twenty20 cricket, teams should focus on taking wickets and bowling dot balls whilst fielding, and implementing tactics that encourage 50+ partnerships and boundary hitting batsmen whilst batting.
Mental training aims to improve the foundation for cognitive regulation of complex movement performance. When mental representation of a sport skill is improved with the aid of systematic techniques, performance is enhanced considerably. The computer-aided tool "Gymnastic-Mental" should be tested during training practice to make mental training more effective and to improve motor performance. As a result of the testing, we looked for a training means that could be used in rhythmic gymnastics to optimize the training technique. In the "Gymnastic-Mental"-programme are selected model examples of sport-specific body techniques and combination techniques as figures. This system can influence the internal movement representations over the visual and acoustic analyzer. The pictures of the technique are to be arranged in the correct order for the movement. Indicators for quantifying the effects of practising and of cognitive learning gain are defined using the "error rate" as a measurement for accuracy of action and "cognitive time consumption" as a measurement for time taken to make decisions and hence availability of representation pattern. A group of 8 top-level gymnasts were tested. The cognitive and motor components were concurrently examined in pre- and post-tests. Mental Training was carried out on a daily basis over a one-week period. Practice was being done during this week, as well as for another seven weeks. The athletes were video recorded before and after the period of practicing. Data were collected and analysed with SIMI Motion. We looked for biomechanical parameters which are demonstrated the quality of the movement implementation. For our selected technique "double diving turn backward" a low back bend and big amplitude of the legs are important parameters. A significant improvement in learning was found. After one week of mental practice, 95% of the pictures were assigned correctly during the post-test conditions. The time required to select the pictures decreased on average about 80.6%. A significant improvement of the accomplishment could be proved after a period of seven weeks. The back bend increased on average about 15° and the angle between the legs about 8°. After one-week period of mental training exercise, it can be assumed that the programme supports mental learning (basic structure of the movement) and allows the control of movement association. As a result, mistakes associated with an incorrect movement can be assessed and removed. It becomes clear that the motor implementation still shows considerable deficienies. At the time a quantitative evaluation of the biomecanical parameters takes place.
The article deals with two closely related topics: A general conceptual framework and a corresponding mathematical model for observational systems designed to support the analysis of sport games. Focussing on volleyball the framework is based on the notion of action chain and its description by parameters. The central aspect of the ensuing analysis is the search for relations among the included components of the observation. Rates are introduced to provide a quantitative measure. As an application the main constituents of AZuR are explained, a program for the analysis of setter decisions in volleyball.
The penalty corner is one of the strategic actions that determine the result in a match in field hockey. In that investigation we want to show if differences exist between international men's and women's teams in the development of this strategy when it is successful. We propose the following aims in our study: to see if a relationship exists between gender and the following variables: used skill, number of players who intervene, number of passes, zone of shot and direction of the sucessful shot at the goal in the penalty corner. The sample is formed by 59 games of international level of top level female (n=21) and male (n=38). In these games 128 penalty corners that finished in goals were registered. We used the data base notational "OBANGOFH" for the capture of information. The analysis of the data was carried out through crosstabs, and Chi-square (p<0,05), where the statistical SPSS package was used. The results obtained in this study determine that the action of the goal in the penalty corner, suggests significant differences between men and women, in the skill of use, the number of players who intervene, the number of passes and the zone of shot.
Ahead of the 2001 season FIVB (Fédération Internationale de Volleyball) reduced the beach volleyball court dimension from 9×18 m to 8×16 m, and changed the scoring system from side out scoring (scoring only from one's own serve) to rally point scoring. The purpose of this study was to investigate the quality and efficiency of game actions in top international beach volleyball for men before and after these changes in constitutive rules. The winning teams in 17 semi-finals and finals in three World Tour tournaments in 1999 and 2000 (sample A) were compared with the winning teams in 17 semi-finals and finals in the same tournaments in 2001 and 2002 (sample B). Serve efficiency, reception efficiency, set efficiency, attack efficiency, block efficiency and defense efficiency in the two samples were analyzed, using established evaluation scales. Mann-Whitney U-test was used to test any significant changes in efficiency values on the game actions in the two samples. The results showed a significant reduction in serve efficiency and attack efficiency after the change of regulations, as well as a significant increase in block actions and block efficiency. The results indicate that the change of rules and court dimension only to a certain extent have led to the expected intensions of FIVB, as the field defense actions nor showed an increase in number nor in efficiency in this study.
Cohesion of he opponents' serve reception with the type of setter's set.
Cohesion of the opponents' serve reception with the serve and the setter's set effectiveness.
Cohesion of the opponents' serve reception with the line-up of opponent's block
Cohesion of the opponents' serve reception withthe reception of the serve and the attack's effectiveness.
The purpose of this study was to evaluate the influence of the reception of the opponents' serve on the offensive actions and strategy of the A1 National Division setters and, by extension, of their team. The subject of the study consisted in Greek volleyball teams of A1 Men's National Division. Thirty-six videotaped matches from the period 1998-1999 were observed (3 for each team). In order to collect data, a method of indirect observation through Vicas analysis system was used. The following parameters were evaluated: a) team b) service reception, c) type of set, d) tempo of set, e) set's area, f) setter's set effectiveness, g) type of attack h) line-up of opponent's block, and i) attack's effectiveness. Cross-tabular statistical analysis showed that the quality of the reception of the opponents' serve formed the offensive strategy of the Greek setters of A1 National Division, though it did not affect the effectiveness of their sets to the attackers and the attacking strategy of the team. However, it should be pointed out that, one could not draw reliable conclusions with respect to the team's attacking strategy and effectiveness, from simply studying the reception of the opponents' serve.
Duration, speed and heart rate values during driving bouts.
Objectives: we aimed to evaluate heart rate (HR) response in pilots to repeated bouts of race car driving. Design and methods: Eight young male student pilots (18.75 ± 3.41 years) participated to a training session consisting in 5 successive bouts (27.81 ± 1.50 min) of driving a “Formula academy“ open-wheel race car on the “Bugatti“ Speedway of Le Mans (France) . Mean and peak speeds were calculated after lap duration measurement using a telemetric infrared timing device. HR was recorded continuously on 5-second intervals using a portable cardiometric device. Results: when driving at a mean 134.94 ± 2.96 km/h speed, mean HR was 132.71 ± 10.71 bpm corresponding in 67.88 ± 5.37 %HRmax intensity. No significant differences were found between the different driving bouts whatever the parameters considered. Conclusions: these results showed good repeatability of the measurements and they suggested that HR monitoring is a valid method to evaluate racecar pilot adaptation to driving . Moreover, the lack of relationships between duration variation, mean or peak speed and HR all along the experiment confirms the major role of muscle isometric loads on energy expenditure and therefore on HR when driving a “Formula academy“ open-wheel racecar.
The Bloomfield Movement Classification (BMC) allows speed agility quickness requirements as well as injury risk of activity in team games to be characterised. A limitation of previous applications of the BMC is that frequencies of movement types can be over-estimated while duration of movement instances can be under-estimated. This is because a movement instance composed of segments performed in different directions and / or turning activity will be presented as separate movement instances even if the same locomotive movement type is being performed. The current paper proposes a method of processing data captured using the BMC to address this. The method not only recognises movement instances composed of multiple segments but also allows movements to be characterised by the number of turns and direction changes performed within the movement instances. The netball movement data used in the current investigation has limited reliability and the results should only be considered in the knowledge that reliability is limited. However, the way in which the results are presented here are a good example of how the BMC can be used in future investigations of movement in different sports where a greater level of reliability is achieved.
This study has pursued to find the efficacy values in the playing microsituations in the counterattack and in the defensive adjustment and to analyse the relation between these and the condition of winner or loser. We analysed the matches of the 10th World Championship of Water polo which did not ended in a draw. Playing microsituations in the counterattack and in the defensive adjustment were evaluated by means of coefficients, obtaining efficacy values. Differences were revealed in male category in the following coefficients: possibility (p=.025), concretion (p=.006), definition (p=.029) and precision (p=.047). In female category differences were found in the following coefficients: possibility, concretion, definition, resolution and precision (p=.000) and accuracy (p=.001); taking as reference a value of p <.05. To conclude with, we can say that in eight, in male category, and twelve, in female category, out of the fourteen efficacy coefficients proposed for evaluating the playing microsituations in the counterattack and defensive adjustment in water polo, there are significant differences between the condition of winner or loser.
The purposes of the present investigation were the analysis of aggressive behaviors and their relationship with performance according to the person-context interaction model. All squads that participated in the soccer World Cups for U-17, U-21 and adults were selected. Predictor variables were squad characteristics including players' body mass index, age, number of tournaments participated in, squads' place, and play round, and criterion variables were aggressive behavior as reflected by cards, fouls, and performance. The results of path analysis and MANOVA mixed factorial design revealed that the aggressive behaviors are associated with number of tournaments participated in and play round. In addition, the preliminary performance is related to aggression, but these results didn't show for knock-out performance. In conclusion, the aggressive behavior in the elite level of soccer is related to the interaction of squad characteristics and stress and aggression causes deterioration in the performance.
Eventing is a three phased equestrian sport. It involves the competition of a dressage test, round of show jumps and completion of a cross-country course. The cross-country phase is described by the sports governing body as the most important and influential element of the competition. Within the sport various levels of competition are run and equality of competition should be maintained at within these levels. The paper compares the relative effect of the cross-country phase of competition between 10 advanced events (sub populations) run in during 2003. Non-normality was apparent within the whole population (Kolmogorv Smirnov K statistic = 5.178 (p<0.001); Shapiro Wilks W = 0.711 (P<0.001). The use of Log (x+1) transformation was applied, positive skew (+2.853) was reduced after transformation to -0.543 (Kolmogorov K statistic = 0.754; p>0.05). The influence of extreme kurtosis was removed after transformation (+10.581 reduced to +1.288). The paper suggests that there is considerable variation observed between some sub populations. Measures of central tendency are discussed as comparative measures between events. It is apparent that arithmetic means may be misleading and the uses of other measures of central tendency are discussed including maximum likelihood estimators. Analysis of data distribution using box and whisker plots indicated outliers and extreme outliers were exerting a considerable effect on the population.
Home advantage in the NBA has recently been reported to be "frontloaded," that is, it is largest in the 1st quarter, smallest in the 4th quarter, and in-between in the 2nd and 3rd quarters. Home advantage in the 1st quarter was 60% of home advantage at the end of the game. The present study was intended to determine if NHL hockey, a very different sport, is also frontloaded. The results were that home advantage in the first goal was 53.01% (home wins) and 52.37% in goals scored after the first. Home advantage in complete games was 56.05%. Hence, the advantage accumulated in the 1st quarter equaled 49.75% of the advantage in complete games. Theories of "psychological momentum" have been advanced according to which scoring first gives a team an advantage in subsequent play. In the NHL, however, no such advantage obtains. In fact, the team scored against does slightly better in goals scored after the first than the team which scores first. Momentum theory also claims that scoring first markedly increases the likelihood of winning. It does, but the effect is due mainly to the fact that the first goal is included in the final score.
The purpose of the current paper is to propose a movement classification scheme to be used when investigating injury risk from movement and agility demands of sporting activities. The method was tested and found to be operated with acceptable reliability by the authors. However, the method can be used more reliably for the analysis of soccer than it can for sports like tennis and netball where more direction changes and braking movements occur. The validity of the method was explored by applying it to 8 different cases from 7 different sports to ensure it covered the key events associated with agility demands and injury risk from movement within those sports. The method has over the advantage the Bloomfield Movement Classification is that different path change and turn types can be entered directly by the observer rather than having to be inferred through temporal analysis of the data. This also applies to acceleration and deceleration events although there are accelerations from non-stationary situations and decelerations to non-stationary situations not covered by the method. The method allows different sports and, indeed, different conditions within the same sport to be compared in terms of injury risk from movement and agility demands.
Four Observation areas Another type of grid approach consists in considering four observation areas: (a) defensive; (b) pre defensive; (c) pre offensive; and (d) offensive.
The central corridor One could also consider the central corridor of the pitch and the two bordering corridors. This would make it possible to note play actions conducted in the axis "attacked goal / defended goal" and others carried in the peripheral channels.
An example of nomogram with two performance scores according to the Table 1.
Over the last 10 years, there has been a growing interest in matchanalysis of soccer. Usually, notational analysis uses numerical data to study and assess the quality of a match. But as far as the analysis of the tactical aspects of the game is concerned, there is a dearth of published research with regards to their theoretical bases. The purpose of this presentation is to contribute to the construction of a knowledge base about soccer using some qualitative observational tools. In a soccer match, structures and configurations of play should be considered as a whole rather than examined a piece at a time. Systems with many dynamically interacting elements can produce of rich and varied patterns of behaviours which are clearly different from the behaviour of each component considered separately. To that effect, effective space game, action zone, and configurations of play will be examined to show that this type of analysis is complementary more than opposed to the numerical data analysis systems. We shall also pay special attention to the study data bases about the evolution between two or more configurations of play. To deal with such data, we used a methodology that makes possible the qualitative study of configurations of play. Observation tools for the analysis of team sports, such as the one discussed in this presentation seem to offer a viable and pertinent basis for explaining the evolution of the rapport of strength during a match.
The level of reliability of performance indicators used in an investigation should be related to the analytical goals of the investigation. This paper uses the example of analysing athletes using different tactical performances within the 2004 Olympic women's 800m running final. The percentage error statistic should not use the whole of the mean performance but consider a meaningful sub-range of values. There are different split times within the 800m that characterise the difference between the performances of different types of athlete such as leaders, waiters, breakers and pacers. A comparison of inter-operator errors with inter-sample differences for these split times is a useful means of relating reliability to analytical goals. The main contribution of this paper is the proposal of an alternative %error equation that considers absolute error as a percentage of the mean value recorded above some theoretical minimum value.
The speed and angle of ball throwing in handball penalty were measured and the effects of the muscular fatigue and the differential tasks implied in the different field positions on these variables were studied. The ball thrown by ten players of the female Uruguayan national handball team were 2-D kinematically analysed in the lateral plane. Ball speed and output angle were analysed taking into account the match time and the player's position, either back or wing, using paired sample t-test (p<0.05). A significantly higher speed (p=0.0006) during the first half of match time, compared to the second, was found. Significant difference in the speed according to the position in the field was also warranted (p=0.00015), with backs throwing faster than wings. The throwing angle showed no differences for match time and position (p=0.43 and p=0.63, respectively). The variation in speed during the match could be due to the effect of the increase in muscular fatigue. The difference found for this variable according to the field position is congruent with the function players carried out during the match. The small variations in the shot angles are interpreted as indicative of technique predominance not affected by muscular fatigue.
Artificial neural networks are tools, which - similar to natural neural networks - can learn to recognize and classify patterns, and so can help to optimise context depending acting. These abilities, which are very useful in a lot of technical approaches, seem to be as well useful in particular in analysing and planning tactical patterns in sport games or patterns of learning behaviour in training processes. In a first attempt, in co-operation with LAMES from the University of Rostock, tactical structures in volleyball could successfully be analysed using neural networks. However, the problem is that the special type of network that has to be used for such analyses (i.e. the so called Kohonen Feature Map or KFM) needs a huge amount of data and lacks the necessary dynamic in continuous learning. So in order to describe, analyse, and evaluate continuous learning processes in sports a dynamically controlled network (“DYCON”) has been developed, which consists of a conventional KFM combined with a time-independent neurone-driven control: Each neurone is imbedded in a dynamic performance potential control system, which had been developed for analysis and control of physiological adaptation processes in sport. Two main advantages of DYCON are: Its learning efficiency is very high. In practice, it needs only some hundred data to coin a pattern, where a conventional KFM normally needs about 10.000 to 20.000. Moreover, it can learn continuously and so can recognise and analyse time depending pattern changes. So, DYCON can support the study of processes in sport games in an easier and more efficient way. Moreover, it can help to analyse tactical changes of a team during a season or even during a tournament, as has been done with squash in co-operation with MCGARRY, University of Fredericton. Finally, in a co-operation with RAAB, University of Heidelberg, we try to find out if and how DYCON can be used for analysis and optimisation of training processes in sport.
This paper reviews developments in the use of Artificial Intelligence in technique analysis over about the last ten years. I outline the potential uses of Expert Systems as diagnostic tools for evaluating technique 'errors' and present some example knowledge rules for such an expert system. I then compare technique analysis, in which Expert Systems appear to have found no place to date, with gait analysis, in which they are routinely used. Consideration is then given to technique analysis using Artificial Neural Networks, focusing on Kohonen self-organizing maps, which have been most widely used in technique analysis, and multi-layer networks, which have been far more widely used in biomechanics in general. Examples of the use of Kohonen maps in technique analysis are presented in javelin and discus throwing and in football kicking. Shot putting is the sole technique analysis presented using multi-layer networks. An example is given of the use of Evolutionary Computation in technique optimization, rather than technique analysis, in the soccer throw in, which predicted an optimal technique close to that in the coaching literature. I conclude with some speculations about the future uses of Artificial Intelligence in technique analysis.
Quantifying tactics in team sports is of major interest for reliable diagnostics and goal oriented training but it is usually accompanied by the problem of considering several athletes at once. The purpose of this study is to analyze changes of patterns of configurations and the team specific variability of the underlying tactical concepts. This is done by means of self organizing maps that are trained with the configurations (i.e. players' positions) of female volleyball players from the world championship 2002. A time discrete (target configurations) as well as time continuous (trajectories of configuration changes) oriented approach is chosen and compared. Results show that the classification of constellations is possible and may support a qualitative analysis of structural interactions within a game. Furthermore, a detailed analysis of the rally from Germany vs. Italy shows, that the German team uses fewer types of configurations more often whereas the Italian team chooses from more constellations but less frequently. Since the world champion of 2002 - Italy - shows a higher variability of configuration patterns, these variable tactics seem to correlate with success in team sports.
The purpose of this srudy was to present a method for Assessing a reliability based upon Bland and Altman's (1999) suggestions for the non-parametric treatment of comparison data and the proposals of Nevill et al. (2001) who recommended that 95% of differences should be recorded within a reference value thought to be of 'no practical importance'. In this study, it is shown how the analyst can apply the method when individual sport performance indicators are treated as individual variables. A simple notation analysis test-retest 'experiment' is described, and the intraobserver data collected by one subject was used for illustrative purposes to explain how the necessary graphical, point estimate and confidence interval computations are made. The notational analyses of a further three volunteer subjects are considered in order to investigate the sensitivity of the method. The method is a simple measure of absolute agreement that is sensitive enough to distinguish between the successes and errors made by expert, experienced and improving sport performance analysts. The method is also free from the assumptions of normality, does not depend upon high within-systems variance and it can be applied successfully to individual performance indicators in the development of sport performance analysts and analysis systems.
Many climbers believe flexibility to be a key performance component, but this remains unsubstantiated under experimental conditions. The need for sport-specific measures of flexibility has been highlighted. The purpose of our research was to assess the validity and reliability of four novel tests of climbing flexibility. The four tests, completed on a purposebuilt climbaflex board, were the adapted Grant foot raise, climbing-specific foot raise, lateral foot reach and the foot-loading flexibility test. In addition, for comparative purposes, the participants completed two existing measures, the sit-and-reach test and Grant foot raise. With the exception of the climbing-specific foot raise all measures had good reliability (ICC = 0.90 - 0.97). The existing flexibility measures had a poor correlation with climbing ability. The lateral foot reach and the adapted Grant foot raise were correlated with climbing ability (r = 0.30; r = 0.34) and used together represent good field measures of flexibility. The foot-loading flexibility test was had the strongest correlation with climbing ability (r = 0.65) and could differentiate between climbing abilities (F3,42 = 8.38, p < 0.001) in a laboratory setting. Our findings indicate that flexibility is a key performance component for the sport when a climbing-specific test is used.
The purpose of this study was to use the data collected in three previous studies (Stretch et al 1998/99, Stretch et al 2000 and Stretch et al 2002) to determine variations, if any, between the position of impact on the cricket bat for cricketers of different batting skill levels, as well as for various strokes played off the front foot. The cricket bat was divided into a Cartesian grid (width - 0 to 110 mm; height - 0 to 555mm) and instrumented to identify the ball impact point, while the software stored biographical and stroke data (McKellar et al. 1998). All three studies used a similar data collection method with a bowling machine project the ball at a velocity of 100 to 105 km h-1 on a line about 0.1 m outside the batsman's off stump and on a length that enabled the batsman to play all the strokes off the front foot. Typically each player took part in three testing sessions of 60 deliveries (10 overs), with their normal training programmes supplemented with some additional form of intervention. Players were divided into two groups (Batsmen and Bowlers), as well as classified at the highest level at which they had played at the time of data collection (Provincial and Club). The results of the 60 cricketers and 9052 recorded impacts showed non-significant variations in the impact location for the Provincial and Club cricketers. Significant differences (p < 0.05) occurred between the impact points for the batsmen and bowlers, as well as between all the strokes with the exception of the off-drive. As the strokes were played wider of the pitch, the off- and cover drives on the off-side and the on-drive and leg glance on the on-side, the impact point was further from the midline of the bat thus increasing the risk of being dismissed. The instrumented bat was devised to assess the accuracy and consistency of stroke reproduction and the effects of various intervention modalities. It has shown that it has implications for performance analysis of cricket technique and skill levels.
In the game of Volleyball, at the beginning of every set, the coach is called to declare the starting line-up of his players. Every time the team wins a rally served by the opponent all its players make a clockwise position rotation and the team serves the rally. The primary aim of the present study was to study whether the six possible line-ups that appear during position rotations have the same efficiency for the serving team. All the games of the A1 division of the men's Greek Championship of Volleyball for the 2005-2006 period were used. More specifically all the points (N=21469) of the two rounds of the regular season from 484 sets in 132 games, were taken into account. The results showed that the teams performed two and a half full rotations per set and that the serving team had the disadvantage to win the rally. The χ2 test showed that, with regards to the efficiency of the teams when serving, there are significant differences between the six line-ups. Also the efficiency when serving for home and away games depends on the setter's position and there was a statistically significant decrease for the setter at position 4.
Pitch divisions and associated terminology utilised to identify and describe the spatial distribution of passes performed by a professional association football team. 
Passes executed by a professional football team as a function of playing season, match location and match status (incidence of passes expressed as a percentage of the total passes performed within each playing season).
Pitch surface distribution of passes performed by a professional football team as a function of competitive season (incidence of passes expressed as a percentage of the total passes performed within each playing season ). Note: „D‟, „M‟ and „A‟ represent th e defensive, midfield and attacking thirds of the pitch respectively while „L‟, „C‟ and „R‟ correspond to the left, central and right areas of the pitch (i.e. DR signifies the right side of the defensive third of the pitch). 
Pitch surface distribution of passes performed by a professional football team as a function of match location (incidence of passes expressed as a percentage of the total passes performed during home matches and away matches respectively). Note: „D‟, „M‟ and „A‟ represent the defensive, midfield and attacking thirds of the pitch respectively while „L‟, „C‟ and „R‟ correspond to the left, central and right areas of the pitch (i.e. DR signifies the right side of the defensive third of the pitch). 
The purpose of this study was to examine the influence of match location (i.e. home vs. away), match status (i.e. winning vs. drawing vs. losing) and competitive season (i.e. season one vs. season two) upon tactics-related performance indicators of a professional association football team. The Noldus Observer Video Pro 4.1 behavioural measurement package (Noldus Information Technology, 2002) was employed to observe and record relevant tactical information (i.e. pass incidence and spatial distribution of pass origins upon the pitch surface) from 47 pre-recorded matches sampled from two consecutive domestic league seasons (season one = 25 matches, season two = 22 matches). Log-linear modelling procedures indicated that the occurrence of passes performed by the team varied as an interactive function of match location, match status and the competitive season. Distribution of passes across the pitch surface was also influenced by match location, match status and the competitive season but the effects were independent as opposed to interactive. The findings highlight the complex nature of football performance under differing contexts and outline the need for sports scientists and coaches working within football to consider the influence of situation variables upon tactical performance indicators. The presence of season associated effects also suggests care should be taken when extrapolating findings from one time period to another. Future research should examine alternative methods for assessing the impact of situation variables upon football performance and verify the extent to which the findings presented in the current study can be generalised to other teams and footballspecific populations (e.g. amateur, women and youths).
The purpose of this study was to characterize the attack of the opposite player in Elite Volleyball Female, according to the opposition block characteristics. Six games of the 2004 Olympics Games (female, senior teams) were analyzed, which corresponds to 437 attack actions. In order to test the association between variables, descriptive and inferential statistics were used namely the Chi-Square and Monte Carlo test. The reliability showed that the observations are reliable enough to be used as scientific tool, based on the percentage of accordance and according to the Kappa de Cohen statistic. By zone 1, the opposite player's attack is slow because it permits the opposition of the triple block and the formation of the compact block. It is also effective in as much as the continuity effect prevails. On the contrary, the attack by zone 2 is fast because it accentuates the use of the simple block in relation to zone 1 and the double and triple block is predominantly badly formed. It is also more effective since the point effect prevails when there is no contact with the ball and the block. These results prove that the 2nd line attack by zone 1 is not sufficiently developed in the female Volleyball.
This study analyze the coach's instruction during practice sessions on the task presentation and feedback, studying the coherency between coach information and the athlete retention in relation to the athletes' school level. The participants were forty-two gymnastics' athletes from different academic levels (elementary, middle and high school). The results showed that the coherency between coach information and athlete retention is similar in both instructional events and a substantive part, almost 40%, is not retained by athletes. In the feedback, we didn't verify differences in the information retained according to academic levels. In task presentation the athletes belonging to elementary schools showed the lowest values of retention, demonstrating a significant difference between middle and high school athletes. These results suggest that athletes retention tends to be more difficult when the information is longer, less contextualized and doesn't refer specifically to any specific motor task.
Weight room 1 repetition maximums (1RM) have been shown to be associated with performance in the throwing events. The purpose of this study was to investigate the primary variables that contributed to maximum throwing distance in the indoor weight throw event in track and field. A survey instrument was developed to collect data regarding national level collegiate weight throwers in the United States. Bivariate correlations for weight throw personal best and squat 1RM revealed significant correlations for both the male (r=0.642, p<0.001) and female participants (r=0.799, p<0.001). The height of the female athletes was also significantly correlated to weight throw personal best (r=0.710, p<0.001); however, height was not significantly correlated to personal best among male participants (r=0.173, p=0.361). Furthermore, the results may provide some insight into the obstacles and barriers limiting the development of US weight throwers.
Descriptive parameters of reaction time and the result accomplished at the competition for every male sprint event
The simple regression analyses of reaction time influence on the result in different male sprint events
The aim of each sprint race is to cover the competition sprint distance in the shortest possible time. Each sprint race commences at the moment of start signal announcement, which means, with the speed of a push off from the start blocks. The focus of this research was on reaction time, which is the first factor in time sequence of the sprint race. The sample of subjects included 250 female athletes and 360 male athletes who performed in athletic sprint events at the 2004 Olympic Games in Athens. The purpose of the research was to determine the existence of reaction time influence on the sprint result. The results showed that by the increase of sprint event distance, the average reaction time value became significantly increased too. Statistically significant correlation between the reaction time and the sprint result for female athletes was found for the sprint event 100 m hurdles. For male athletes, statistically significant correlation between the reaction time and the sprint result was established for sprint events 100 m flat and 110 m hurdles. Up until nowadays, the research studies were not able to establish any statistically significant correlation between reaction time and the sprint result in the sprint events.
– Dig Zone Bottom zone  
Dig Zone Bottom zone
– Determinants of the attack tempo 2 in association with the attack tempo 3 
This study aims to analyse possible determinants that might predict the attack tempo in the Volleyball's complex II. A total of 881 actions were analyzed from 28 games of the national male teams presented in the World Cup 2007. A multinomial logistic regression was applied, with the level of significance determined in α=0.05. The variables dig type, dig efficacy and setting zone demonstrated predictive power of the attack tempo. The dig without all attack options and the dig with fall decreased the probability of quicker attack tempos. The higher frequency of the dig without all attack options is, perhaps, an explanation to the high values of the non acceptable setting zone, whose occurrence promotes a slower attack organization. In this sense, the increasing of the dig efficacy and stabilization of the dig type will improve the dig action to acceptable and excellent setting zones, which will increase the use of quicker attack tempos.
-Descriptive statistics per game variable
-Predicting the effect of the attack in transition
The aim of this study was to identify determinants of the attack tactics' effectiveness in Elite Youth Men's Volleyball. Eleven matches of the 2007 World Youth Championship were analyzed, totalling 863 actions of reception, 1191 actions of attack and 435 actions of defence. Multinomial logistic regression was applied in order to analyze the association of reception effect, defence effect, attack tempo and attack type, with the attack effectiveness. The reliability of the observation was calculated through Cohen's Kappa, with values being above 0.81. This study showed that the powerful attack increased the chances of scoring in both side-out and transition. Regarding the timing of the attack, 1st tempo increased the chances of scoring in transition. These results show that the Elite Youth male's game claims a more offensive game through powerful and quick attacks.
The purpose of this study was to analyse procedural knowledge, decision-making and game performance of the zone 4 attackers in Volleyball, according to the players' experience and competitive performance. A verbal interview protocol was implemented immediately after randomly chosen live game actions to analyze procedural knowledge. Decision-making and players' performance were analyzed by video images. Oneway Anova, Independent Samples t-Test, and Pearson Chi-Square were applied to data analysis. The results showed that more experienced and more successful players play more under goal concepts than condition concepts and concern more often with the opponent. Specifically, more experienced players mentioned more goal concepts and less sophisticated conditions, as more successful players mentioned more action concepts. More experienced players also presented fewer condition concepts than the less experienced ones. Successful players tended to make more appropriate decisions. However, as player's experience and competitive success are multidimensional variables they might not be fully well characterized by the number of years of practice and competitive results. Therefore, future research is required, and it should include other criteria to characterize both variables. Moreover, qualitative analysis is needed, since it will allow a deeper understanding of the tactical development of the players according to the specific nature of the training process, and also considering the competitive success and the player's experience as multidimensional variables.
The purpose of the current investigation was to produce models of tennis performance for ladies' and men's singles that related the overall performance goal (winning the match) to relevant aspects of tennis strategy. Match data from 83 ladies' singles matches and 82 men's singles matches was obtained from the official Australian Open web site. Break points won was the variable that most distinguished between the winning and losing player in both ladies' (t82 22.9, P < 0.001) and men's singles (t81 = 15.2, P < 0.001). The number of break points won was strongly associated with the number of points won when serving which is functionally dependent on the proportion of first and second serves that are in as well as the proportion of points won when first and second serves are in. Correlation analysis revealed that service speed had an indirect influence on the proportion of points that were won on first and second serve in both singles games. In ladies' singles aces and net ability was related to effectiveness in points emanating from first serve where as winner to unforced error ratio was a determinant of effectiveness on second serve points. In men's singles, effectiveness in points emanating from first serve was influenced by aces, winner to unforced error ratio, net approaches and ability at the net. Success in points emanating from the second serve in men's singles was influenced by winner to unforced error ration as well as ability at the net.
This study aimed to develop and validate player performance impact rankings for Australian football, considering players' time on ground and game situation. Player performance data was collected from an Australian Football League (AFL) club and their opponents in each game during the 2006 season. Individual player and team impact scores were generated by multiplying the frequency of selected game actions by allocated positive or negative numerical values. The study was divided into three phases. In phase 1 higher team impact scores were shown to have a significant correlation with winning (r=−0.69, p<0.01). A greater final points margin between the teams was also correlated with an increased impact score margin (r=0.85, p<0.001). In phase 2 one-way ANOVAs revealed individual player impact scores were significantly higher in the midfield than in the forward and defensive positional zones (p<0.001), suggesting that impact score comparisons should only be made within positional zones. In phase 3 a chi-square analysis revealed significant differences between individual players within each of the positional zones. It was concluded that the impact ranking scores provided a valid method of assessing game performance for players (within positional zones) and teams, allowing performance profiles to be created for coaching purposes.
Tracking player performance in team sports is difficult as games involve quick, agile movements, with many unpredictable changes in direction and frequent collisions between players. Previous approaches have partially solved the motion analysis problem by manually tracking player movements, using observational sheets or computerised digitising pads and rods. However, manual tracking can be a subjective and laborious process which has arguably discouraged researchers from conducting more detailed analyses of multiple player interactions within games. The aim of this investigation is to establish whether an automated, colour- recognition, motion detection system (A-Eye) is suitable as a tool for the analysis of multiple players in a sporting environment. Reliability tests show strong intra-operator reliability scores (2% TEM) and correlations. Validity tests comparing manual and automatic tracking methods at court boundaries during short game sequences demonstrate moderate to strong correlations between the methods in most areas of the court. A comparison of time spent in each zone of the court also largely shows similarities between analysis methods (0.09%- 2.53%). This automated player tracking system provides a detailed multi-disciplinary analysis and has practical benefits for coaches, practitioners and researchers by enhancing understandings of sports performance and team work.
Correlation coefficient matrix for pair-wise comparisons of the 6 swimming zones.
We measured the speed of 12 international-level swimmers (6♀ and 6♀) for 6 sections of the 50m backstroke race in a short course (25m) pool (0-5m; 5-15m; 15-25m; 25-30m; 30-40m; 40-50m) by breaking down the races using video analysis. Our assessment revealed that swimmers went faster when underwater (independently of the benefit obtained by pushing off the pool wall in the 0-5m and 25-30m zones) than when they swam at the surface. These results were supported by the per-zone speed ranking established by the vast majority of 151 sports students acting as observers. The latter viewed each race 3 times, in order to divide up the race into the 6 zones using frame-by-frame images (with an index of agreement of more than 76%, according to Pareto’s criterion). Only the per-zone speed rankings for the 3 slowest female swimmers deviated from the majority ranking (using Kendall’s L statistic), as the result of a shorter underwater segment during the 2nd length. Despite being considered (by the sport’s rules and in scientific terms) as a subsidiary and disruptive element of propulsion, underwater dolphin kicks are responsible for the highest speeds in this type of event.
Top-cited authors
Miguel Ángel Gómez Ruano
  • Universidad Politécnica de Madrid
Nic James
  • Middlesex University, UK
Peter Gerard O'Donoghue
  • Cardiff Metropolitan University
Mike Hughes
  • Institute of Technology, Carlow
Emerson Franchini
  • University of São Paulo