The continual drive for success in sport has led coaches to search for the best possible means of improving player and team performance. Indeed, finding that "little bit extra" may often be the difference between success and failure especially in top-level team sport. Information Technology is advancing at a very fast rate, affecting all aspects of our lives and sport is no exception to the rule. The need for objective, accurate and relevant feedback on the performances of players from both training and competition has led to the development of high-tech match analysis systems based on computer and video technology. Such systems help in evaluating positive and negative player performance and in devising specific training programmes and match strategies on the basis of the results obtained (Carling, 2001a). Video and computerised match analysis provide a means of quantifying and qualifying various characteristics of individual and team performance. Indeed, this is one of the most important aspects of the coaching process which enables improvement in performance through providing the coach and player with objective and relevant feedback (Carling, 2001b). In general, current performance analysis systems aim to supply both quantitative and qualitative information to provide: rapid feedback on physical, technical, tactical performance, a mechanism for selective searching through a video recording of the game, statistical evaluation, development of a database, the creation of performance models, help in identifying areas requiring improvement and aid player choice and scouting methods. Different game analysis systems have gradually been adopted over recent years by various professional team sports throughout the world looking to improve their preparation and results. Current examples include: Pinnacle Sports (Basketball, American Football), StadeXpert (Rugby), Handivision (Handball) Observer and Mastercoach (Soccer) and SportsCode (Soccer, Rugby, Australian Football). These software often combine digital video with statistical analysis either in realtime or post-match. The principle is based on a simple match video recording which is digitised and player actions (shots, passes, headers, duels, tackles…) are inputted into the computer. As digitised video contains a time-code, these events can subsequently be directly accessed and visualised at the click of a button. Most of these systems also offer either video montage (video match compilation of the most important actions chosen by the coach), statistical analysis and graphical output of the results. The StadeXpert and SportsCode systems also allow the personalisation of the system interface (e.g. adding or removing match actions) and provide a database. Indeed, as databases are created, a clearer understanding of each sport will follow (Franks & Hughes (1997). However, these particular systems have gradually shown their limitations as they do not provide information on the physical performances of players, nor allow computer simulations of match actions, the input of spatial data is subject to inaccuracies and the coach is limited to analysing what he sees on the cassette video which generally concentrates on the player in possession. This latter point can be placed into perspective as soccer players spend only 2% of the match in possession of the ball (Reilly, 1996). These particular limitations have led to the development of various state-of-the-art systems based on player tracking technology (e.g. Trakus, AMISCO, Orad, Soccerman). The Trakus system designed for American Football and Ice Hockey uses missile and radar technology to track the movements of players in real-time. Players are localised and their movements tracked through telemetric signals from microprocessors placed in their helmets which are detected by transceivers strategically positioned around the stadium. However, this system is currently forbidden for certain team sports, notably soccer, due to the electronic material carried by players. The AMISCO football system also detects the positions and movements of players but only requires standard video and computer equipment and therefore does not affect the playing conditions in anyway (Brulé et Coll, 1998). A multi-camera system combined with a high-tech computer vision program based on digital image processing techniques and mathematical algorithms provides over 4 million bits of information per match (Carling, 2002). These particular systems provide an exhaustive summary of game performance and help create individual and team match-play profiles. The physical performances of players (fig 1 shows an international footballer) such as the distance run, % distance walking, jogging and high-intensity efforts, capacity to accelerate, frequency of sprints and the recovery time between these actions can all be evaluated and for any moment of the game. The data may also help in medical issues by aiding the prediction of injury and analysing the subsequent return to competition through trends in physical performance. Computerised 2D/3D match simulations which recreate match play can help to better evaluate team strategies and accurate tactical analysis of both individual and collective attacking and defending actions is possible through detailed graphical output. Future computerised match analysis systems are being created using newer and improved technology. Quicker data input techniques via voice recognition and touch-screen systems incorporated into smaller more practical computers such as "PDAs" will no doubt continue to improve game analysis. "GPS" satellite technology for tracking movement is currently being adapted for team sports and expert-systems using artificial intelligence to process data will help make real-time decisions and aid in the prediction and modelling of sports performance.
© Copyright 2002 Expertise in Elite sport. 2nd International Days of Sport Sciences, 12.-15. November 2002, INSEP, Paris (France). Veröffentlicht von INSEP. Alle Rechte vorbehalten.