KU Leuven’s Jesse Davis uses AI to reveal hidden soccer tactics
A data scientist shows why kicking the ball out of bounds can be a winning move
Imagine a World Cup opener where a player boots the ball straight out of bounds on the opponent’s end. Casual fans may scoff, but Jesse Davis’s lab at KU Leuven sees a data-backed scoring opportunity. Davis, a computer science professor, runs the Sports Analytics Lab, which built a training set of 1.4 million passes and 60,000 throw-ins (partly from the 2022 World Cup). Using tree ensemble models, their 2024 paper 'Boot it' found that when the ball is in the middle third, kicking it out of bounds near the opponent’s goal can lead to a goal within 10 subsequent actions. This seemingly counterintuitive play is now popping up in top leagues.
Beyond game-day insights, Davis occupies a unique niche in sports analytics. While many clubs hire internal data teams, Davis releases most of his research via open-source tools, allowing broader access. His journey started with AI in healthcare and biomechanics before shifting to soccer tactics thanks to a soccer-obsessed PhD student. The lab’s work not only helps clubs evaluate rosters and strategies but also standardizes in-game data. For tech-savvy professionals, this represents a rare fusion of academic rigor and practical sports impact, proving that open-source AI can reshape even the most traditional sports.
- Davis's lab analyzed 1.4 million passes and 60,000 throw-ins from the 2022 World Cup.
- Tree ensemble models showed that kicking the ball out of bounds in the middle third can lead to a goal within 10 actions.
- Davis makes his soccer analytics research freely available via open-source tools.
Why It Matters
Pro soccer clubs can gain competitive edge from open-source AI insights on tactics and player evaluation.