The aim of this explorative study was to investigate the activity profile in basketball during competition, for the first time taking into account both teams at once. It is also the first application of the Sports Performance Analyzer (SPA) to analyse both teams on court, comparing winner to loser. We discuss the performance of offence and defence separately. While McInnes et al. (1995) found no changes in high intensity activities between quarters, Ben Abdelkrim et al. (2010) identified significant decreases from quarter one to two and quarter three to four. Solving this contradiction seems to be important to quantify on-court running performance, find reasonable performance requirement estimations and specify instructions for training. A sample of four games of the German Pro-A-League during the season 2010/2011 was evaluated using the automatic tracking system SPA (Wilhelm et al., 2010). For the first time we used particle-filter based tracking instead of template tracking. Running performance of eight teams (70 players; average age 25) was analysed. All players were tracked to generate individual activity profiles composed of running distances and intensities according to net and gross time. Intensities were Standing (<0.7 km/h), Walking (0.7-7.2 km/h), Jogging (7.2-14.4 km/h), Running (14.4-19.8 km/h) and Sprinting (>19.8 km/h). No significant difference between running performance of winner and loser was found. Running intensity and distance differed between offence and defence, running and sprinting efforts are greater in offence than in defence. A significant decrease of on-court performance was determined for quarter one to two and for quarter three to four. Dividing the games in two halves the performance also decreases in the second half. This study provides novel information with practical relevance for players, coaches and researchers in basketball. They increase the understanding of requirements and demands in competition. Further information of specific problems and aspects in basketball competition is added. This investigation enhances the knowledge on specific terms in basketball games, especially differences in running performance among quarters. The used particle-filter based tracking algorithm is faster and more error-prone than template tracking, thus making the Sports Performance Analyzer more reliable and easier to use.
© Copyright 2012 World Congress of Performance Analysis of Sport IX. Veröffentlicht von University of Worcester. Alle Rechte vorbehalten.