Optimising Spatial Teamwork Under Uncertainty
A new method for optimising team defence in football using Monte Carlo Tree Search and linear programming — reducing opponent threat by 21% in simulation.
I'm a data scientist specialising in AI and machine learning, helping improve team performance within football clubs, from analysing matches to identifying recruitment targets. My work is backed by peer-reviewed research and shaped by working inside a professional club every day.
A new method for optimising team defence in football using Monte Carlo Tree Search and linear programming — reducing opponent threat by 21% in simulation.
A team-selection model that balances performance with injury risk — demonstrating ~13% fewer first-team injuries and ~11% reduction in inefficiently spent wages.
GAPP: a graph attention network for predicting pass reception and quantifying off-ball defender contributions with interpretable metrics.
I'm a data scientist at Southampton FC with a PhD from the University of Southampton. My research focused on using AI and machine learning to solve real problems in football, from optimising team selection to evaluating player performance using techniques like reinforcement learning and graph neural networks.
Now I apply that research day to day inside a professional club, contributing to projects across recruitment, player evaluation and tactical decision support.
Full bio