trailing 250 and 500 day (excluding offseason) regular season RAPM. updated nightly.
(this is a really early version right now. would love suggestions.)
why 250 and 500 day samples?
there is a ton of noise in RAPM for small samples. the single-season RAPM numbers that are floating out there suffer from this. 250 game days is about 1.5 seasons, and 500 game days is about 3 seasons.
a few technical details:
- ridge regression (use cross-validation to find “best” regularization parameter)
- some possessions played cutoff that regresses about 1/5th of the players toward a ~replacement level~ player
- please tweet / dm me @canzhiye any suggestions or questions