The influence of situational variables on ball possession in the English Premier League.

Abstract The aims of this study were twofold: (1) to examine the influence of situational variables on ball possession in elite soccer and (2) to quantify the variables that discriminate between high or low percentage ball possession teams (HPBPT and LPBPT) across different playing positions. Match performance data were collected from English Premier League matches using a multiple-camera system. Data were examined using linear regression, a 2 × 5 factorial analysis of variance and discriminant analysis. Playing against weak opposition was associated with an increase (P < 0.01) in time spent in possession while playing away decreased (P < 0.01) the time spent in possession by ~3%. Possession was increased (P < 0.01) when losing than winning or drawing. Finally, the better the ranking of a team, the higher (P < 0.01) the time spent in possession. The playing position effect was significant for all variables (P < 0.05); however, there were only interactions with team ball possession in some cases. The discriminant analysis identified functions for all five playing positions (P < 0.01). The variables that discriminated performance between HPBPT and LPBPT were different for various playing positions, although the number of successful passes was the most common discriminating variable. The results demonstrate that HPBPT and LPBPT developed different possession strategies during matches and that selected variables such as successful passes were identified to explain these data trends across various playing positions. Combinations of variables could be used to develop a probabilistic model for predicting time spent in possession by teams.