Football's data analytics in python
summary
Shows a concrete Python workflow: cleaning tracking and event feeds with pandas, NumPy, training a transfer-value regressor (random forest) that achieves ~15% holdout error, and a scouting filter that narrows 10,000 prospects to 200 high-probability targets. Presents an injury-risk logistic model that flags players with 4× higher short-term reinjury odds, and an interactive dashboard built with plotly and seaborn that visualizes tactical heatmaps and model diagnostics — all code and charts run end-to-end.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 6:24Football analytics evolution
- 15:07First part of the talk
- 26:57Data insights and scouting
- 31:35Xbox and football data
- 39:48Conclusion and Q&A
related talks
19mFootball Analytics with the Soccermatics Package
33mUsing Data Science to Better Evaluate American Football Players
31mFootball is complex, but your code doesn’t have to be — meet DataBallPy and a practical deep dive into pressing
35mMental Models to Use and Avoid as a Data Scientist
39mPredicting sports winners using data analytics with pandas and scikit-learn
23mSport analysis with Python
56mSports Analytics: Big Data in the Big Time
41m"R" You Ready for Some Football? Hacking Fantasy Sports with Open Source Software