Development and Validation of Team Functional Degrees of Freedom (TfDoF): An Eco-Physical Variable for Functional Collective Off-ball Movements in Football

In football, limited research has considered both the processes underlying collective behavior and its functional outcomes, constraining its practical applicability. In this study, we developed and validated an eco-physical variable, Team Functional Degrees of Freedom (TfDoF), to quantify how off-ball players dynamically coordinate their movements with teammates, opponents, and the ball to afford functional passing opportunities for a team in football. TfDoF was developed through a “glocal” spatial framework integrating centroid dynamics, Voronoi diagrams, dynamic pitch zoning, and a rule-based degrees-of-freedom approach. Validation comprised two complementary stages. First, expert-based validation involved seven football experts who evaluated the computational framework and identified passing opportunities in 30 match clips at the instant of passing. TfDoF showed a very strong correlation with expert-perceived passing opportunities (r = .98, p < .001), while all content validity indices exceeded .86, providing preliminary evidence of its content and convergent validity. Second, performance-based validation examined the relationship between TfDoF and pass outcomes using synchronized tracking and event data from 35 Italian Serie A matches (20,531 passes). Among six machine learning algorithms, XGBoost achieved the best classification performance (AUC = 0.909). SHAP analysis identified TfDoF as the most influential predictor of pass success (FI = 1.453), providing evidence of its predictive validity. Notably, TfDoF’s contribution to pass success was non-linear, with the greatest marginal effect occurring between TfDoF values of 5 and 7. Finally, we proposed a multi-timescale analytical framework demonstrating how TfDoF can support tactical analysis and coaching practice.