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

2026-01-02

arXiv.org

AI Summary

Existing metrics struggle to quantify the synergistic effects of off-the-ball defending in soccer—particularly its role in constraining opponents’ actions—and often lack tactical context. Addressing this gap in corner-kick scenarios, this work proposes the first unsupervised, annotation-free Covariate-Dependent Hidden Markov Model (CDHMM) to infer man-marking and zonal defending roles directly from player tracking data. Building upon these inferred roles, the study introduces a counterfactual analysis framework conditioned on defensive assignments, integrating an enhanced ghosting methodology to enable tactically aware attribution of individual defensive contributions. This approach facilitates fine-grained, interpretable evaluation of off-the-ball defending and significantly outperforms conventional baselines that rely on average behavioral patterns in corner situations.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Calibration & Uncertainty Quantification
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Tracking, profiling, and countermeasures against them

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.

Problem

• off-ball defense
• performance evaluation
• football
• tracking data
• counterfactual analysis

Innovation

• covariate-dependent Hidden Markov Model
• off-ball defense
• defensive role inference
• role-conditioned ghosting
• defensive credit attribution