OpenRAPM: bias by player type
| Player type | vs consensus | vs 2026 observed |
|---|---|---|
| Low-usage wings who stay out of the way -ftm -tov -pts -ast | +0.541 | -0.006 |
| Guards who handle the ball +ast -orb +tov -drb | +0.231 | +0.158 |
| High-usage scorers who draw fouls +ftm +pts +tov +ast | +0.435 | +0.132 |
| Busy all-round forwards +stl +drb +pf +orb | -0.090 | +0.022 |
| Two-way big men who score +ftm +blk +pts +drb | +0.197 | +0.338 |
| Rim-protecting centers +orb -fg3_miss +blk -fg3m | -0.081 | -0.100 |
| Ball-hawk defenders who rarely shoot +stl -pts -ftm -fg3m | -0.049 | +0.036 |
| Wings who shoot threes +fg3m +fg3_miss -drb -blk | +0.010 | -0.056 |
vs 2026 observed. We fit a RAPM model against the residuals of OpenRAPM's predictions. These are the average results for each player type (possession-weighted).
vs consensus. The consensus is a GLS-weighted mean of DRIP, LEBRON, EPM, DARKO DPM, Time-Decay RAPM, and Time-Decay-Luck-Adjusted RAPM.