Exploring advanced Bayesian statistical modeling and high-performance open-source tools
for sports analytics. Home of the penaltyblog Python library,
supporting researchers and analysts in quantifying the unpredictability of the game.
Advanced Bayesian methods for incorporating uncertainty and prior knowledge into football analysis, providing probabilistic insights for decision-making.
Leveraging Rust, Cython and C-extensions to build high-performance statistical samplers that bridge the gap between Python's flexibility and C's speed.
Development of match outcome models, player performance metrics, and tactical analysis systems using machine learning and statistical approaches.
Building production-ready data pipelines and open-source Python libraries used by analysts and researchers at all levels of professional football.
Why most finishing metrics are flawed and how a Bayesian approach gives us a truer picture of a player's finishing ability.
Comprehensive evaluation of different modeling approaches for football prediction, with performance metrics and optimization strategies.
A mathematical approach to calculating Expected Threat (xT) in Python using linear algebra, providing a more efficient alternative to the standard iterative method.
A comprehensive Python library for football analytics used by analysts, researchers, and consultants globally. Features advanced statistical models, data collection modules, betting market analysis tools, data visualization utilities, and implementations of cutting-edge metrics. The codebase has been adopted for professional work at all levels of the game.
I am always interested in how penaltyblog is being applied to real-world challenges.
Whether you are a researcher, an analyst at a club, or a developer working on sports data,
I’m open to discussing new ideas, experimental features, or potential research collaborations.
Get in Touch.