A Machine Learning Framework for Off Ball Defensive Role and Performance Evaluation in Football

TL;DR

This paper presents a novel machine learning framework using covariate-dependent Hidden Markov Models to evaluate off-ball defensive roles and performance in football, addressing limitations of traditional metrics and existing models.

Contribution

It introduces a label-free, context-aware model for inferring defensive roles and a new framework for credit attribution and counterfactual analysis in football defense.

Findings

Effective inference of player roles from tracking data

Enhanced evaluation of defensive contributions

Improved interpretability of defensive performance

Abstract

Evaluating off-ball defensive performance in football is challenging, as traditional metrics do not capture the nuanced coordinated movements that limit opponent action selection and success probabilities. Although widely used possession value models excel at appraising on-ball actions, their application to defense remains limited. Existing counterfactual methods, such as ghosting models, help extend these analyses but often rely on simulating "average" behavior that lacks tactical context. To address this, we introduce a covariate-dependent Hidden Markov Model (CDHMM) tailored to corner kicks, a highly structured aspect of football games. Our label-free model infers time-resolved man-marking and zonal assignments directly from player tracking data. We leverage these assignments to propose a novel framework for defensive credit attribution and a role-conditioned ghosting method for counterfactual analysis of off-ball defensive performance. We show how these contributions provide a interpretable evaluation of defensive contributions against context-aware baselines.