NFL Player Props
HC + BET signal picks with Edge/Vol scores, logged before kickoff.
OpenAnytime-touchdown picks from a Poisson model, game-total adjusted.
OpenEvery pick logged before kickoff, graded against official NFL data.
OpenWhich defences concede most to each position, by market.
OpenCorrelated same-game-parlay stacks with a combined edge score.
OpenVegas-implied game script mapped to position-level prop modifiers.
Opens Week 1Filter the full board by market, position, opponent and edge.
OpenEvery skill-position player: season averages, recent form, red-zone share.
OpenHit rates and per-game form for every prop line, split by form and matchup. Past form, not a prediction.
OpenFactor weights, signal thresholds and walk-forward validation, in plain English.
OpenBacktested on 23,934 player-games
Five seasons of NFL data, walk-forward validated, with situation edge filters tuned per market. Below is what we found - including some non-obvious truths.
Key Findings (Counterintuitive)
See full methodologyImplied team total has r=+0.006 correlation with WR receiving yards. Zero predictive power. We removed it from the model.
Spread correlates with rushing yards (r=0.14) but adjusting for it INCREASES prediction error. Player season average already captures whether their team runs or passes.
TE receiving has the lowest win rate of any market (34.3% vs proxy lines). We require 15+ yard edge minimum before publishing any TE pick.