TacticAI: an AI assistant for football tactics

Zhe Wang,Petar Veličković,Daniel Hennes,Nenad Tomašev,Laurel Prince,Michael Kaisers,Yoram Bachrach,Romuald Elie,Li Kevin Wenliang,Federico Piccinini,William Spearman,Ian Graham,Jerome Connor,Yi Yang,Adrià Recasens,Mina Khan,Nathalie Beauguerlange,Pablo Sprechmann,Pol Moreno,Nicolas Heess,Michael Bowling,Demis Hassabis,Karl Tuyls

DOI: https://doi.org/10.1038/s41467-024-45965-x

IF: 16.6

2024-03-19

Nature Communications

Abstract:Abstract Identifying key patterns of tactics implemented by rival teams, and developing effective responses, lies at the heart of modern football. However, doing so algorithmically remains an open research challenge. To address this unmet need, we propose TacticAI, an AI football tactics assistant developed and evaluated in close collaboration with domain experts from Liverpool FC. We focus on analysing corner kicks, as they offer coaches the most direct opportunities for interventions and improvements. TacticAI incorporates both a predictive and a generative component, allowing the coaches to effectively sample and explore alternative player setups for each corner kick routine and to select those with the highest predicted likelihood of success. We validate TacticAI on a number of relevant benchmark tasks: predicting receivers and shot attempts and recommending player position adjustments. The utility of TacticAI is validated by a qualitative study conducted with football domain experts at Liverpool FC. We show that TacticAI’s model suggestions are not only indistinguishable from real tactics, but also favoured over existing tactics 90% of the time, and that TacticAI offers an effective corner kick retrieval system. TacticAI achieves these results despite the limited availability of gold-standard data, achieving data efficiency through geometric deep learning.

multidisciplinary sciences

What problem does this paper attempt to address?

The paper attempts to address the problem of automatically identifying opponents' tactical patterns and developing effective counter-strategies in modern football through algorithms. Specifically, the research team developed an artificial intelligence football tactics assistant named TacticAI, aimed at analyzing corner kick tactics. The primary focus on corner kick analysis is due to the fact that corner kicks provide coaches with the most direct opportunities for intervention and improvement, and they occur frequently in matches (approximately 10 times per match on average). Therefore, improvements in corner kick execution could significantly alter match outcomes.

TacticAI combines both predictive and generative components, enabling coaches to effectively explore different player configuration schemes and select the one with the highest success rate. The study validated TacticAI's performance on multiple relevant benchmark tasks, including predicting the receiving player, shot attempts, and recommending player position adjustments. Additionally, through collaboration with Liverpool Football Club, the researchers conducted a qualitative study demonstrating that the tactical suggestions provided by TacticAI are not only indistinguishable from actual tactics but also superior to existing tactics in 90% of cases. Furthermore, TacticAI offers an effective corner kick retrieval system. Despite the limited availability of high-quality data, TacticAI achieved efficient data utilization through geometric deep learning.