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 · 2022

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
563
carrys to trust the number
shrunkrawthe shrink90% CI
36%38%40%42%44%46%avg1Miles SandersPHI · 26043.3%2Isiah PachecoKC · 17042.6%3J.K. DobbinsBAL · 9242.5%4Elijah MitchellSF · 4542.5%5AJ DillonGB · 18642.4%6Aaron JonesGB · 21342.3%7Cordarrelle Patte…ATL · 14442.2%8Josh JacobsLV · 34142.1%9Kenneth GainwellPHI · 5341.9%10Nick ChubbCLE · 30341.7%11Caleb HuntleyATL · 7641.5%12Jamaal WilliamsDET · 26341.4%13Jeff WilsonMIA · 17641.4%14Benny SnellPIT · 2041.4%15Tyler AllgeierATL · 21141.4%16Matt BreidaNYG · 5441.4%17Alexander MattisonMIN · 7441.3%18James CookBUF · 8941.3%19Jordan MasonSF · 4341.3%20James ConnerARI · 18341.2%21Raheem MostertMIA · 18141.2%22Darrel WilliamsARI · 2141.1%23Gus EdwardsBAL · 8741.1%24Ezekiel ElliottDAL · 23141.1%25Chuba HubbardCAR · 9541.1%26Brian RobinsonWAS · 20541.0%27Joshua KelleyLAC · 6941.0%28Latavius MurrayDEN · 17241.0%29Devin SingletaryBUF · 17741.0%30Austin EkelerLAC · 20441.0%31Ronald JonesKC · 1740.9%32Justice HillBAL · 4940.9%33Corey ClementARI · 1540.9%34DeeJay DallasSEA · 3540.8%35Raheem BlackshearCAR · 2340.8%36Marlon MackDEN · 1640.8%

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 · 2022 · full board

Rush success rate leaderboard for the 2022 NFL season, shrunk by empirical Bayes with 90% credible intervals, ranked by the shrunk estimate.
#PlayerTeamPoscarrysRawShrunk90% interval
1Miles Sanders◦ provisionalPHIRB26048.9%43.3%40.5%46.1%
2Isiah Pacheco◦ provisionalKCRB17048.8%42.6%39.6%45.6%
3J.K. Dobbins◦ provisionalBALRB9253.3%42.5%39.3%45.7%
4Elijah Mitchell◦ provisionalSFRB4564.4%42.5%39.2%45.8%
5AJ Dillon◦ provisionalGBRB18647.3%42.4%39.4%45.3%
6Aaron Jones◦ provisionalGBRB21346.5%42.3%39.4%45.2%
7Cordarrelle Patterson◦ provisionalATLRB14447.9%42.2%39.1%45.3%
8Josh Jacobs◦ provisionalLVRB34144.3%42.1%39.4%44.8%
9Kenneth Gainwell◦ provisionalPHIRB5354.7%41.9%38.7%45.2%
10Nick Chubb◦ provisionalCLERB30343.6%41.7%39.0%44.5%
11Caleb Huntley◦ provisionalATLRB7647.4%41.5%38.3%44.7%
12Jamaal Williams◦ provisionalDETRB26343.0%41.4%38.6%44.3%
13Jeff Wilson◦ provisionalMIARB17643.8%41.4%38.5%44.4%
14Benny Snell◦ provisionalPITRB2060.0%41.4%38.0%44.8%
15Tyler Allgeier◦ provisionalATLRB21143.1%41.4%38.5%44.3%
16Matt Breida◦ provisionalNYGRB5448.1%41.4%38.1%44.6%
17Alexander Mattison◦ provisionalMINRB7446.0%41.3%38.1%44.5%
18James Cook◦ provisionalBUFRB8944.9%41.3%38.1%44.5%
19Jordan Mason◦ provisionalSFRB4348.8%41.3%38.0%44.6%
20James Conner◦ provisionalARIRB18342.6%41.2%38.2%44.2%
21Raheem Mostert◦ provisionalMIARB18142.5%41.2%38.2%44.1%
22Darrel Williams◦ provisionalARIRB2152.4%41.1%37.8%44.5%
23Gus Edwards◦ provisionalBALRB8743.7%41.1%38.0%44.3%
24Ezekiel Elliott◦ provisionalDALRB23142.0%41.1%38.2%44.0%
25Chuba Hubbard◦ provisionalCARRB9543.2%41.1%37.9%44.2%
26Brian Robinson◦ provisionalWASRB20541.9%41.0%38.1%44.0%
27Joshua Kelley◦ provisionalLACRB6943.5%41.0%37.8%44.3%
28Latavius Murray◦ provisionalDENRB17241.9%41.0%38.0%44.0%
29Devin Singletary◦ provisionalBUFRB17741.8%41.0%38.0%44.0%
30Austin Ekeler◦ provisionalLACRB20441.7%41.0%38.1%43.9%
31D'Onta Foreman◦ provisionalCARRB20441.7%41.0%38.1%43.9%
32Ronald Jones◦ provisionalKCRB1747.1%40.9%37.6%44.3%
33Justice Hill◦ provisionalBALRB4942.9%40.9%37.6%44.2%
34Corey Clement◦ provisionalARIRB1546.7%40.9%37.5%44.3%
35Samaje Perine◦ provisionalCINRB9641.7%40.9%37.7%44.0%
36Damien Harris◦ provisionalNERB10641.5%40.8%37.7%44.0%
37DeeJay Dallas◦ provisionalSEARB3542.9%40.8%37.6%44.2%
38Raheem Blackshear◦ provisionalCARRB2343.5%40.8%37.5%44.2%
39Jaylen Warren◦ provisionalPITRB7741.6%40.8%37.6%44.0%
40Marlon Mack◦ provisionalDENRB1643.8%40.8%37.5%44.2%
41Rex Burkhead◦ provisionalHOURB2642.3%40.8%37.5%44.1%
42Khalil Herbert◦ provisionalCHIRB12941.1%40.8%37.7%43.9%
43Breece Hall◦ provisionalNYJRB8041.3%40.8%37.6%44.0%
44D'Andre Swift◦ provisionalDETRB10041.0%40.8%37.6%43.9%
45Clyde Edwards-Helaire◦ provisionalKCRB7140.8%40.7%37.5%44.0%
46Jaret Patterson◦ provisionalWASRB1741.2%40.7%37.4%44.1%
47Travis Etienne◦ provisionalJAXRB22140.7%40.7%37.9%43.6%
48Rashaad Penny◦ provisionalSEARB5740.4%40.7%37.5%44.0%
49Joe Mixon◦ provisionalCINRB21040.5%40.7%37.8%43.6%
50Rhamondre Stevenson◦ provisionalNERB21040.5%40.7%37.8%43.6%
51Cam Akers◦ provisionalLARB18840.4%40.6%37.7%43.6%
52Dare Ogunbowale◦ provisionalHOURB4339.5%40.6%37.4%43.9%
53Tony Pollard◦ provisionalDALRB19340.4%40.6%37.7%43.6%
54Najee Harris◦ provisionalPITRB27240.4%40.6%37.9%43.4%
55Eno Benjamin◦ provisionalNORB7839.7%40.6%37.4%43.8%
56Boston Scott◦ provisionalPHIRB5438.9%40.6%37.3%43.8%
57Ty Johnson◦ provisionalNYJRB3036.7%40.5%37.2%43.9%
58Kyren Williams◦ provisionalLARB3537.1%40.5%37.2%43.8%
59Kevin Harris◦ provisionalNERB1833.3%40.5%37.2%43.9%
60Christian McCaffrey◦ provisionalSFRB24639.8%40.5%37.6%43.3%
61Sony Michel◦ provisionalLACRB3636.1%40.5%37.2%43.8%
62Gary Brightwell◦ provisionalNYGRB3135.5%40.5%37.2%43.8%
63Zack Moss◦ provisionalINDRB9338.7%40.4%37.3%43.6%
64Kareem Hunt◦ provisionalCLERB12339.0%40.4%37.4%43.5%
65Mike Boone◦ provisionalDENRB2433.3%40.4%37.1%43.8%
66Travis Homer◦ provisionalSEARB1931.6%40.4%37.1%43.8%
67Mark Ingram◦ provisionalNORB6437.5%40.4%37.2%43.6%
68Brandon Bolden◦ provisionalLVRB1729.4%40.4%37.1%43.8%
69Dontrell Hilliard◦ provisionalTENRB2231.8%40.4%37.1%43.7%
70Avery Williams◦ provisionalATLRB2231.8%40.4%37.1%43.7%
71Zamir White◦ provisionalLVRB1729.4%40.4%37.1%43.8%
72Javonte Williams◦ provisionalDENRB4736.2%40.4%37.1%43.7%
73Hassan Haskins◦ provisionalTENRB2532.0%40.4%37.0%43.7%
74Leonard Fournette◦ provisionalTBRB18939.1%40.3%37.4%43.3%
75Craig Reynolds◦ provisionalDETRB2330.4%40.3%37.0%43.7%
76Royce Freeman◦ provisionalHOURB4134.2%40.3%37.0%43.6%
77Trestan Ebner◦ provisionalCHIRB2429.2%40.3%36.9%43.6%
78J.D. McKissic◦ provisionalWASRB2227.3%40.2%36.9%43.6%
79Ke'Shawn Vaughn◦ provisionalTBRB1723.5%40.2%36.9%43.6%
80Jonathan Taylor◦ provisionalINDRB19238.5%40.2%37.3%43.1%
81Melvin Gordon◦ provisionalDENRB9036.7%40.2%37.0%43.3%
82Deon Jackson◦ provisionalINDRB6835.3%40.1%37.0%43.4%
83Malik Davis◦ provisionalDALRB3831.6%40.1%36.9%43.5%
84Phillip Lindsay◦ provisionalINDRB1618.8%40.1%36.8%43.5%
85Jonathan Williams◦ provisionalWASRB3729.7%40.1%36.8%43.4%
86Malcolm Brown◦ provisionalLARB1816.7%40.0%36.7%43.3%
87Justin Jackson◦ provisionalDETRB4330.2%40.0%36.7%43.3%
88Isaiah Spiller◦ provisionalLACRB1816.7%40.0%36.7%43.3%
89Kenyan Drake◦ provisionalBALRB10935.8%39.9%36.8%43.0%
90Jerick McKinnon◦ provisionalKCRB7233.3%39.9%36.7%43.1%
91Dameon Pierce◦ provisionalHOURB22037.7%39.9%37.0%42.8%
92Derrick Henry◦ provisionalTENRB34938.4%39.8%37.2%42.5%
93JaMycal Hasty◦ provisionalJAXRB4628.3%39.8%36.5%43.1%
94Alvin Kamara◦ provisionalNORB22337.2%39.7%36.9%42.6%
95Rachaad White◦ provisionalTBRB13035.4%39.7%36.7%42.8%
96James Robinson◦ provisionalNYJRB11034.5%39.7%36.6%42.8%
97Keaontay Ingram◦ provisionalARIRB2718.5%39.7%36.4%43.0%
98Bam Knight◦ provisionalNYJRB8532.9%39.7%36.6%42.9%
99Antonio Gibson◦ provisionalWASRB14935.6%39.6%36.6%42.7%
100Tyrion Davis-Price◦ provisionalSFRB3420.6%39.6%36.3%42.9%
101Nyheim Hines◦ provisionalBUFRB2412.5%39.6%36.3%42.9%
102David Montgomery◦ provisionalCHIRB20136.3%39.6%36.7%42.5%
103Darrell Henderson◦ provisionalLARB7030.0%39.5%36.4%42.8%
104Michael Carter◦ provisionalNYJRB11433.3%39.5%36.4%42.6%
105Chase Edmonds◦ provisionalDENRB6827.9%39.4%36.2%42.6%
106Saquon Barkley◦ provisionalNYGRB29535.9%39.1%36.4%41.8%
107Dalvin Cook◦ provisionalMINRB26434.1%38.6%35.8%41.4%
108Kenneth Walker III◦ provisionalSEARB22833.3%38.6%35.8%41.4%

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).