Maximising the Set-Piece Return: Optimising Football Corner Tactics with Graph Reinforcement Learning

2026-06-04

AI Summary

This study addresses the limitations of traditional corner-kick tactical analysis, which relies on historical replay or predefined counterfactual scenarios and struggles to proactively discover novel strategies. The work proposes a general-purpose tactical optimization framework that formulates corner-kick offense as a decision-making problem, enabling optimization from arbitrary initial player formations. By integrating graph reinforcement learning with motion trajectory modeling, the method dynamically adjusts the positions and velocities of attacking players to maximize the probability of a shot on first touch. This approach marks a paradigm shift from imitation learning to reward-driven strategy generation through graph neural networks. Experimental evaluation on over 3,000 Premier League corner-kick sequences demonstrates that the proposed framework significantly outperforms existing baselines under identical inference budgets, effectively enhancing offensive efficiency.

Abstract

Machine learning is increasingly employed for the evaluation of football tactics. However, existing approaches focus on characterising historical actions or analyst-specified counterfactual scenarios. In this work, we seek to go beyond the imitation of historically observed patterns towards discovering new generalisable player configurations and strategies. To tackle this, we focus on optimising corner kick routines, and formulate a decision-making problem in which a central policy makes adjustments to attacking player positions and velocities to maximise first contact shot probability. Unlike classic optimisation that solves for isolated setups, we contribute a reinforcement learning architecture operating on graph-structured data that yields a general policy for adjusting arbitrary starting player positions. Evaluated on over 3,000 Premier League corners, our approach strongly outperforms baseline optimisation techniques under matched inference budgets. Our results suggest that graph reinforcement learning can shift set-piece analysis from historical evaluation and imitation towards reward-driven tactical discovery.

Problem

• corner kick
• tactical optimization
• set-piece
• player positioning
• shot probability

Innovation

• graph reinforcement learning
• corner kick optimization
• tactical discovery
• player positioning
• football set-piece