This repo stores my code and plots for sports-related plots.
-
Stability of goal-scoring rate and/or xG over-performance
-
Madden rating EDA with UMAP + kmeans clustering (for optimal # of UMAP groups).
-
In which sport is the leading team most likely to further their lead near the end of the game due to an opposition error, overly-aggressive tactics by the opposition, etc.? (End of game depends on the sport. Ideally, we could use a win probability model to help idenitfy end-of-game context, although it would still need some adjustment.) Extension: In which sport are gamblers most likely to lose/win due to a back-door cover?
-
Decision tree to filter down players in a sport to just one player who has achieved a very specific set of career totals/averages (i.e. https://twitter.com/JacobEGoldstein/status/1296570734047686658)
-
“Standardized” soccer shot data locations: https://kloppy.pysport.org/changelog/#110-2020-08-07
-
xG soccer model: https://github.com/ML-KULeuven/soccer_xg
-
Transfer market data: https://github.com/ewenme/transfers
-
FIFA data: https://github.com/RobWHickman/fifadb