NFL · Empirical Bayes

True-Talent Leaderboards

A raw single-season rate is a noisy guess at a player’s real ability, and the noise is worst for the smallest samples — so a naïve leaderboard is topped by whoever got lucky in the fewest tries. This one shrinks every rate toward its position-group prior by how much the sample can be trusted, and shows a 90% credible interval instead of a point. Toggle raw vs. shrunk to watch the flukes fall back to the pack.

Rush success rate · 2023

Share of carries with positive EPA. Blocking & scheme are large, un-controlled inputs.

Beats raw by
19.4%
lower out-of-sample error
RMSE raw → shrunk
9.1% → 7.3%
odd vs. even weeks
Split-half reliability
0.20
how repeatable the raw stat is
Stabilizes at
183
carrys to trust the number
shrunkrawthe shrink90% CI
30%35%40%45%50%avg1De'Von AchaneMIA · 10344.5%2Kyren WilliamsLA · 22943.5%3Christian McCaffr…SF · 27243.3%4Raheem MostertMIA · 20942.4%5Aaron JonesGB · 14242.2%6David MontgomeryDET · 22041.7%7Gus EdwardsBAL · 19941.7%8D'Andre SwiftPHI · 22941.3%9Keaton MitchellBAL · 4741.0%10Nick ChubbCLE · 2840.8%11James CookBUF · 23740.5%12D'Onta ForemanCHI · 10940.5%13James ConnerARI · 21040.3%14Jaleel McLaughlinDEN · 7640.2%15Rhamondre Stevens…NE · 15640.2%16Latavius MurrayBUF · 7940.2%17Royce FreemanLA · 7740.1%18Jordan MasonSF · 4040.0%19Tony PollardDAL · 25339.7%20Ty ChandlerMIN · 10239.7%21Ronnie RiversLA · 3239.6%22Zach CharbonnetSEA · 10839.6%23Ty JohnsonBUF · 3039.5%24Damien HarrisBUF · 2339.4%25Justice HillBAL · 8439.4%26Zack MossIND · 18339.4%27Tony JonesARI · 2639.3%28Alexander MattisonMIN · 18039.2%29Joe MixonCIN · 25739.1%30Isiah PachecoKC · 20639.1%31Michael CarterARI · 3039.1%32Ameer AbdullahLV · 1539.0%33Chris Rodriguez J…WAS · 5139.0%34Kendre MillerNO · 4138.9%

Each row is a player: the solid dot is the shrunk estimate, the hollow dot the raw rate, joined by the red pull of regression; the grey bar is the 90% credible interval. A hollow shrunk dot means the sample is below the stabilization line — the number is mostly the position prior. Switch to Raw rank and watch the small-sample names climb.

Rush success rate · 2023 · full board

Rush success rate leaderboard for the 2023 NFL season, shrunk by empirical Bayes with 90% credible intervals, ranked by the shrunk estimate.
#PlayerTeamPoscarrysRawShrunk90% interval
1De'Von Achane◦ provisionalMIARB10355.3%44.5%39.7%49.3%
2Kyren WilliamsLARB22947.6%43.5%39.5%47.5%
3Christian McCaffreySFRB27246.7%43.3%39.5%47.2%
4Raheem MostertMIARB20945.9%42.4%38.3%46.5%
5Aaron Jones◦ provisionalGBRB14247.2%42.2%37.7%46.7%
6David MontgomeryDETRB22044.5%41.7%37.7%45.8%
7Gus EdwardsBALRB19944.7%41.7%37.6%45.8%
8D'Andre SwiftPHIRB22943.7%41.3%37.4%45.3%
9Keaton Mitchell◦ provisionalBALRB4751.1%41.0%35.7%46.3%
10Nick Chubb◦ provisionalCLERB2857.1%40.8%35.4%46.5%
11James CookBUFRB23742.2%40.5%36.6%44.5%
12D'Onta Foreman◦ provisionalCHIRB10944.0%40.5%35.8%45.2%
13James ConnerARIRB21041.9%40.3%36.2%44.4%
14Jaleel McLaughlin◦ provisionalDENRB7644.7%40.2%35.3%45.3%
15Rhamondre Stevenson◦ provisionalNERB15642.3%40.2%35.8%44.6%
16Latavius Murray◦ provisionalBUFRB7944.3%40.2%35.2%45.2%
17Royce Freeman◦ provisionalLARB7744.2%40.1%35.1%45.1%
18Jordan Mason◦ provisionalSFRB4047.5%40.0%34.7%45.4%
19Tony PollardDALRB25340.7%39.7%35.9%43.6%
20Ty Chandler◦ provisionalMINRB10242.2%39.7%35.0%44.5%
21Ronnie Rivers◦ provisionalLARB3246.9%39.6%34.2%45.2%
22Zach Charbonnet◦ provisionalSEARB10841.7%39.6%34.9%44.3%
23Ty Johnson◦ provisionalBUFRB3046.7%39.5%34.1%45.1%
24Damien Harris◦ provisionalBUFRB2347.8%39.4%33.9%45.1%
25Justice Hill◦ provisionalBALRB8441.7%39.4%34.5%44.4%
26Zack Moss◦ provisionalINDRB18340.4%39.4%35.2%43.6%
27Tony Jones◦ provisionalARIRB2646.2%39.3%33.9%44.9%
28Alexander Mattison◦ provisionalMINRB18040.0%39.2%35.0%43.4%
29Joe MixonCINRB25739.7%39.1%35.3%43.0%
30Isiah PachecoKCRB20639.8%39.1%35.1%43.2%
31Michael Carter◦ provisionalARIRB3043.3%39.1%33.6%44.6%
32Ameer Abdullah◦ provisionalLVRB1546.7%39.0%33.4%44.7%
33Chris Rodriguez Jr.◦ provisionalWASRB5141.2%39.0%33.8%44.3%
34Pierre Strong◦ provisionalCLERB6440.6%39.0%33.9%44.1%
35Tyjae Spears◦ provisionalTENRB10040.0%38.9%34.2%43.7%
36Kendre Miller◦ provisionalNORB4141.5%38.9%33.6%44.3%
37Jaylen Warren◦ provisionalPITRB14939.6%38.9%34.6%43.4%
38Elijah Mitchell◦ provisionalSFRB7540.0%38.8%33.9%43.9%
39Cam Akers◦ provisionalMINRB6040.0%38.8%33.7%44.0%
40Najee HarrisPITRB25639.1%38.8%35.0%42.6%
41Kareem Hunt◦ provisionalCLERB13539.3%38.7%34.3%43.3%
42Roschon Johnson◦ provisionalCHIRB8139.5%38.7%33.8%43.7%
43Patrick Taylor◦ provisionalGBRB3240.6%38.7%33.3%44.2%
44Darrynton Evans◦ provisionalMIARB3240.6%38.7%33.3%44.2%
45Alvin Kamara◦ provisionalNORB18239.0%38.7%34.5%42.9%
46Jonathan Taylor◦ provisionalINDRB16939.1%38.7%34.5%43.0%
47Emari Demercado◦ provisionalARIRB5839.7%38.7%33.6%43.9%
48Trey Sermon◦ provisionalINDRB3540.0%38.6%33.3%44.1%
49Craig Reynolds◦ provisionalDETRB4139.0%38.5%33.2%43.9%
50Antonio Gibson◦ provisionalWASRB6538.5%38.4%33.4%43.5%
51Kenneth Walker IIISEARB21938.4%38.4%34.4%42.4%
52Chuba HubbardCARRB23838.2%38.3%34.4%42.2%
53Kevin Harris◦ provisionalNERB1637.5%38.3%32.7%44.0%
54Samaje Perine◦ provisionalDENRB5337.7%38.2%33.1%43.5%
55Salvon Ahmed◦ provisionalMIARB2236.4%38.1%32.6%43.8%
56La'Mical Perine◦ provisionalKCRB2236.4%38.1%32.6%43.8%
57Israel Abanikanda◦ provisionalNYJRB2236.4%38.1%32.6%43.8%
58Kenneth Gainwell◦ provisionalPHIRB8537.6%38.1%33.3%43.1%
59Jeff Wilson◦ provisionalMIARB4136.6%38.0%32.8%43.4%
60Boston Scott◦ provisionalPHIRB2035.0%38.0%32.5%43.7%
61Chase Edmonds◦ provisionalTBRB4936.7%38.0%32.9%43.3%
62Trayveon Williams◦ provisionalCINRB1533.3%38.0%32.4%43.7%
63Melvin Gordon◦ provisionalBALRB2634.6%37.9%32.5%43.5%
64Devin SingletaryHOURB21637.5%37.9%33.9%41.9%
65Brian Robinson◦ provisionalWASRB17937.4%37.9%33.8%42.1%
66Jahmyr Gibbs◦ provisionalDETRB18237.4%37.9%33.7%42.1%
67Jerick McKinnon◦ provisionalKCRB2133.3%37.9%32.3%43.5%
68Derrick HenryTENRB28037.5%37.8%34.2%41.6%
69Khalil Herbert◦ provisionalCHIRB13336.8%37.7%33.3%42.2%
70Zamir White◦ provisionalLVRB10436.5%37.7%33.1%42.4%
71Rico Dowdle◦ provisionalDALRB8936.0%37.6%32.8%42.4%
72Matt Breida◦ provisionalNYGRB5534.5%37.5%32.4%42.7%
73Cordarrelle Patterson◦ provisionalATLRB5034.0%37.4%32.3%42.7%
74Tank Bigsby◦ provisionalJAXRB5034.0%37.4%32.3%42.7%
75Miles Sanders◦ provisionalCARRB12935.7%37.3%32.8%41.8%
76Saquon BarkleyNYGRB24836.3%37.2%33.4%41.0%
77Eric Gray◦ provisionalNYGRB1723.5%37.1%31.6%42.8%
78Tyler AllgeierATLRB18735.8%37.1%33.0%41.3%
79Raheem Blackshear◦ provisionalCARRB1520.0%37.0%31.4%42.7%
80Chris Brooks◦ provisionalMIARB1921.1%36.7%31.3%42.4%
81Joshua Kelley◦ provisionalLACRB10733.6%36.6%32.0%41.3%
82Travis EtienneJAXRB26935.3%36.5%32.9%40.3%
83Sean Tucker◦ provisionalTBRB1513.3%36.5%30.9%42.2%
84Javonte WilliamsDENRB21934.7%36.4%32.5%40.4%
85Josh JacobsLVRB23334.8%36.4%32.5%40.3%
86AJ Dillon◦ provisionalGBRB17834.3%36.4%32.2%40.6%
87Darrell Henderson◦ provisionalLARB4628.3%36.3%31.2%41.6%
88Ke'Shawn Vaughn◦ provisionalTBRB2420.8%36.3%30.9%41.9%
89Bijan RobinsonATLRB21434.6%36.3%32.4%40.3%
90Austin Ekeler◦ provisionalLACRB17934.1%36.3%32.1%40.4%
91Breece HallNYJRB22434.4%36.2%32.3%40.1%
92Ezekiel ElliottNERB18433.1%35.8%31.7%39.9%
93Deuce Vaughn◦ provisionalDALRB2313.0%35.5%30.1%41.1%
94Keaontay Ingram◦ provisionalARIRB3520.0%35.4%30.2%40.8%
95Chase Brown◦ provisionalCINRB4422.7%35.3%30.2%40.6%
96Isaiah Spiller◦ provisionalLACRB3718.9%35.1%29.9%40.5%
97D'Ernest Johnson◦ provisionalJAXRB4119.5%34.9%29.8%40.2%
98Dalvin Cook◦ provisionalNYJRB6725.4%34.9%30.0%39.9%
99Jerome FordCLERB20631.1%34.5%30.6%38.5%
100Clyde Edwards-Helaire◦ provisionalKCRB7024.3%34.5%29.6%39.4%
101Jamaal Williams◦ provisionalNORB10625.5%33.6%29.1%38.3%
102Rachaad WhiteTBRB27230.1%33.5%29.9%37.1%
103Dameon Pierce◦ provisionalHOURB14524.1%32.1%27.9%36.4%

How the shrinkage works

Two estimators

Rate stats (completion %, success rate, catch rate) use a beta-binomial model: a Beta(α, β) prior fit by marginal likelihood over each position group, then a Beta posterior per player. Per-play averages (EPA, CPOE, yards) use a normal-normal model with DerSimonian–Laird between-player variance. Both pull each player toward their group by exactly how thin their sample is.

Does it help? & the fine print

The trust panel’s numbers come from a leakage-free odd/even-week holdout: fit on odd weeks, predict even-week raw. Shrinkage lowers out-of-sample error for every stat. Priors are fit per season and per position (WR and TE separately), so a TE’s baseline isn’t a skill. Regular season only, 2016–2025. Rushing and receiving efficiency are heavily scheme-driven — read the wide bands as the honesty they are.

Source: nflverse play-by-play. Counted and computed deterministically — never modeled by a language model. Built by build_nfl_leaderboards.py (byte-reproducible).