🐻⬇️🏀

Model Card

Two-parameter player RAPM variant with offense and defense estimated together. This is the first explicit expansion beyond a single player eta.

Each player's rating starts from a share of the player's previous-season RAPM rather than zero (NBA 50%, NCAAM 50%, NCAAW 50%; other leagues start from zero), so a few games move it less. The offense/defense split still starts from zero.

Validation here is out of sample. From two weeks into each season, every week's lineups are predicted by a fit on the season's earlier play, with the served settings; each player's point is the predicted and actual margin per 100 possessions of the held-out play the player was on the floor for.

Selected League / Season

Change league with the League selector at the top of the page.

Held-out player predictions

WNBA 2026: 184 players with at least 200 held-out possessions. Across players, predicted and actual correlate at r = 0.87 (weighted by held-out possessions). Squared error is 0.7% below predicting the league average. Each dot is a player (184 with the most held-out possessions are plotted; larger dots had more); on the dashed line, actual equals predicted. Select a dot for the player.

Output Coverage

Rows in each model's current published run (or legacy rows where no run is selected).