Applying an Expected Pass Turnovers model to inform pressing strategies in professional football

The aim of this study was to apply the Expected Pass Turnovers (xPT) model and demonstrate its practical applications within a practitioner setting. The model was implemented using event data and positional data from all 380 matches in the 2020/21 English Premier League season, comprising 256,433 passes in the final dataset. Players exhibiting lower pass retention ability, as quantified by xPT, can be identified as pressing targets for opposing teams and subsequently can be targeted through deliberate pressing strategies to regain possession. Furthermore, the defensive quality of players was evaluated using a complementary metric derived from xPT, termed Expected Pass Turnovers Forced (xPTF). As xPT quantifies the probability of turnovers for each pass based on contextual features, xPTF extends this by attributing these probabilities to the defending player applying pressure. Hence, the xPTF value was calculated for all passes made under pressure and attributed to the closest player executing the pressure action. Team-level xPTF performance correlated with Passes per Defensive Action (PPDA; R = -0.52, p < 0.05), supporting the metric’s validity. Overall, the xPT model provides actionable insight into potential pressing targets, while xPTF serves as an evaluative tool for assessing individual and team pressing performance.