Soccer Analytics

I’ve enjoyed playing and watching soccer since I was a little kid. Starting with my BSc thesis, I was able to combine this pashion with my interest in data science. I developed a model for predicting the outcomes of Premier League games, finding that financial features are as informative as performance-based metrics.

During my master’s degree, I joined a local startup called matchmetrics and developed models for predicting player performance across different leagues and countries.

My work culminated in my MSc thesis, where I developed a model to predict events in the video feed of a soccer game. The model was able to detect events such as passes, crosses, dribblings, and shots with high accuracy.