DISTRIBUTION
Live board. Empirical EWM + Poisson (at least one). The daily board runs an empirical estimator: exponentially weighted game-log average, blended with a usage-adjusted rate, scaled by an opponent-allowed factor, with a negative-binomial or normal spread fitted from the league's own game-to-game variance. NFL means are calibrated on 2021-2025 outcomes (accuracy only: the pool of quoted players is over-projected by 3-7%). It does not model injuries, lineups, minutes, weather or teammate absences, and it has not been validated against prices.
Research model. XGBoost (Poisson) + NegBin
WHAT THE LIVE BOARD LOOKS AT
- Exponentially weighted game-log average
- Usage-adjusted rate (carries + targets)
- Opponent-allowed factor, shrunk and clipped
LIVE BOARD ACCURACY
SAMPLE
2021-2025 regular seasons
MAE (ENGINE VS L5)
0.486 vs 0.492
MAE CHANGE
-1.2% (se 0.4%)
Full serving-code replay (nfl.props.engine_eval), opponent factor included, as of each regular-season week, 2021-2025. Accuracy of the projection only; it says nothing about beating a price.
RESEARCH MODEL VALIDATION
SAMPLE
Walk-forward 2021-2025
MAE (MODEL VS BASELINE)
--
CRPS (MODEL VS BASELINE)
--
Result. Beats trailing-mean baselines
Against prices. No edge shown. On the banked weeks (main line, 3+ books) the model-minus-market Brier gap was -0.005 +/- 0.013 (n=83), indistinguishable from zero.
Record. No graded history is published for NFL projections: they are not logged to the public ledger, so there is no win rate or ROI to show.
Source: Walk-forward and banked-market study, 2026-10-14 (NFL prop systems handoff). MAE and CRPS: lower is better. The live board does not run the research model.
Probability of at least one rushing or receiving touchdown. Return and defensive touchdowns are not counted; the books' own scope is unverified.