Running Back Ratings — Season 2025
A descriptive composite of every qualified RB’s season: workload (touches per game), rushing volume (yards per game), passing-down value (receiving yards + 6×TDs per game), and ground efficiency (yards per carry) — each z-scored within the season and equally weighted. Descriptive ranking by design, not a projection. Goal-line specialists and 4-game cameos are hidden from the default view.
| # | Score | Player | Team | G | Tch | Ry | T/G | YPC |
|---|
What this rating is (and isn’t)
What it is: a transparent four-component composite that scores each qualified RB’s season on workload (touches per game), rushing production (rushing yards per game), passing-down value (receiving yards + 6×receiving TDs per game, the 3DRB anchor), and ground efficiency (yards per carry). Each component is z-scored within the season and equally weighted — so league-wide changes in scheme, pace, or league-wide pass rate don’t pull scores around year to year.
What it isn’t: a betting projection. The composite shows measurable lift over naive baselines on next-game rushing yards MAE (+7.6%, 95% CI excludes zero) and boom-probability Brier — but that lift is insufficient to clear the −110 vig on sportsbook prop lines, and Kalshi RB rushing-yards sub-markets are dormant. We publish it as a descriptive ranking for readers, not a tradeable model.
Why no QB-decoupling? Unlike WRs, RBs aren’t dependent on the QB’s arm or accuracy in the same way. RB efficiency is offensive-line, box-count, and game-script bound. We tested team rushing-EPA as a residualization target and confirmed it absorbs the RB’s own signal — so the composite stays pure rate-and-volume.
Qualified floor: ≥150 touches over the season is the default display threshold (≈ 10 games of starter workload). Players with 100–149 touches are aggregated for the z-score pool so the distribution stays sane, but hidden from the default view. Toggle “All RBs” to see everyone with ≥100 touches.