Goalball, a Paralympic sport since 1976 in Toronto, is designed for visually impaired athletes. The goal is that two opposing teams, each consisting of three players, try to score goals by rolling the ball into the opposition’s net. The court dimensions are the same as a standard volleyball court. The lines of the court are tactile, made by placing thick tape over cords. Players wear blindfolds to guarantee equal impairments. The ball contains bells allowing players to echolocate movements. Alternating between offense and defence Goalball shows a very structured game play.
So far only very few studies in the field of Goalball performance analysis can be found. The purpose of this study was to gain a better scientific understanding of performance in Goalball. This project was funded by the German Institute of Sport (IIA1-070405/12-13 and IIA1-070406/14) in order to develop state of the art software for performance analysis in Goalball. Based on the model “Systematische Spielbeobachtung”, three software tools, which are specific to Goalball were developed.
For practical scouting purposes GoalScout runs on tablet PCs. Additionally, positional data which is extracted form video sequences and provides precise information about start and goal sector. More frequently used game characteristics appear situation-related next to the video, which allows for a very intuitive and efficient way of scouting. Based on just 4 to 9 “clicks”, up to 18 game characteristics can be derived for each bowl. The interrater agreement for each game characteristic is high (Cohens Kappa > .80). GoalView is a specific data analysis tool for Goalball. Using different filter settings, each video sequence of a bowl can be analysed. Furthermore, reports (e.g. statistical, bowl-distribution, Performance Index, etc.) can be created in real-time. To determine ball trajectory, GoalTrack uses a multistage approach of computer vision. Using background subtraction as well as colour and match shaping the position of the ball is detected from the video frame. The regression line of the ball is a result of a localization of ground contact points of the ball plus a projection of ball position to ground level at these times.
Altogether 113 Goalball matches were analysed. Software tools were used to analyse each bowl (n=21058) with respect to all 18 game characteristics. Furthermore, GoalTrack was used to measure ball speed (n = 8397). No significant difference between male and female games with regards to number of bowls per game were detected. The scoring rate for males is higher than for females. Rotational bowling technique is used by male players more often. So called “bouncing balls” are used more often by male players as well. Scoring sector analyses show significance differences on sector 3 and 7. A correlation between scoring rate and ball speed is statistically significant.
In this study, an investigation on performance indicators in Goalball was conducted. Scoring sector analysis verified the opinion of coaches and players that it is easier to score on sectors where defensive responsibilities of two players overlap. A significant low scoring probability was found on sectors where defenders can use tactile court information. Significant correlation between scoring rate and ball speed, especially in women games, recommend a training of basic strength abilities. Additional research should further investigate underlying causes of observed bowling patterns as well as the impact on rule changes.
«
Goalball, a Paralympic sport since 1976 in Toronto, is designed for visually impaired athletes. The goal is that two opposing teams, each consisting of three players, try to score goals by rolling the ball into the opposition’s net. The court dimensions are the same as a standard volleyball court. The lines of the court are tactile, made by placing thick tape over cords. Players wear blindfolds to guarantee equal impairments. The ball contains bells allowing players to echolocate movements. Alte...
»