Player stats

Advanced stats from every play: EPA (expected points added), success rate, CPOE and more, by player. The leaders below use a qualifying bar (PFR's: 14 dropbacks, 6.25 carries or 1.875 targets per team game). Click any name for that player's full profile, or open the explorer to cut the plays any way you like.

Season leaders

Splits that hold up

Most situational splits don't repeat: a player's gap between, say, blitzed and not blitzed one season predicts almost nothing about the next. These are the exceptions: splits where the same player's edge over the league carried over from one season to the next (2016–25, correlation r with its 95% interval clear of zero). They're the splits most worth reading as skill. How we tested it.

Biggest swings this season

The largest gaps between a player in one situation and the same player everywhere else, beyond the gap the whole league shows there. This is the kind of stat that makes a highlight clip. Read these as stories, not traits: we checked 3,073 player splits and 31 cleared |z| ≥ 2.5. Pure chance would produce about 38. So far this season, that's no more than luck alone would give. The badge says how much that kind of split repeats from one season to the next.

• CeeDee Lamb DAL: +0.92 EPA/target on normal rest, vs -0.70 the rest of the time. mostly noise · r -0.01 38 vs 23 targets · league gap -0.07 · z = +4.1 · compare everyone
• Justin Herbert LAC: +0.56 EPA/dropback in the 3rd quarter, vs -0.41 the rest of the time. mostly noise · r -0.06 47 vs 172 dropbacks · league gap -0.00 · z = +4.1 · compare everyone
• Christian Watson GB: +1.69 EPA/target leading by 1–8, vs +0.02 the rest of the time. mostly noise · r -0.03 14 vs 34 targets · league gap -0.04 · z = +3.5 · compare everyone
• Jalen Hurts PHI: +0.57 EPA/dropback in the 2nd quarter, vs -0.26 the rest of the time. mostly noise · r -0.06 50 vs 129 dropbacks · league gap +0.00 · z = +3.4 · compare everyone
• George Kittle SF: -0.34 EPA/target while the game is in doubt (win prob 20–80%), vs +1.37 the rest of the time. mostly noise · r -0.02 21 vs 15 targets · league gap -0.01 · z = -3.2 · compare everyone
• Josh Allen BUF: +0.90 EPA/dropback between the opponent's 21 and 49, vs +0.15 the rest of the time. mostly noise · r 0.08 60 vs 140 dropbacks · league gap +0.04 · z = +3.2 · compare everyone
• Jayden Daniels WAS: -0.17 EPA/dropback while the game is in doubt (win prob 20–80%), vs +0.63 the rest of the time. mostly noise · r 0.08 87 vs 56 dropbacks · league gap -0.01 · z = -3.2 · compare everyone
• Bo Nix DEN: +0.35 EPA/dropback trailing by 1–8, vs -0.31 the rest of the time. mostly noise · r 0.05 61 vs 141 dropbacks · league gap -0.02 · z = +3.1 · compare everyone
• C.J. Stroud HOU: +0.73 EPA/dropback against five rushers, vs +0.00 the rest of the time. mostly noise · r -0.05 35 vs 166 dropbacks · league gap -0.09 · z = +3.0 · compare everyone
• Ja'Marr Chase CIN: -0.19 EPA/target between his own 21 and 50, vs +1.39 the rest of the time. mostly noise · r 0.04 22 vs 16 targets · league gap -0.05 · z = -3.0 · compare everyone
• Matthew Golden GB: +0.78 EPA/target against a four-man rush, vs -0.67 the rest of the time. mostly noise · r 0.12 29 vs 15 targets · league gap +0.02 · z = +2.9 · compare everyone
• Joe Burrow CIN: -0.05 EPA/dropback on short throws, vs +0.60 the rest of the time. mostly noise · r 0.01 100 vs 89 dropbacks · league gap -0.04 · z = -2.9 · compare everyone

Where this comes from. nflverse play-by-play (rebuilt daily), FTN charting for play action, pocket, blitz, box count, screens, motion and RPOs (2022 on), and NFL Next Gen Stats participation data for pressure, coverage and time to throw (2016–2025; not published for 2026 yet). EPA per play is noisy, about ±1.4 points a play, so 100 plays still carry roughly ±0.14 of luck. Box-score totals are checked against ESPN after every rebuild: see the check.