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.
Catch rate · 2022
Receptions per target, shrunk within position (WR vs TE separately). Depth-of-target confound; catchable-target rate not available here.
- Beats raw by
- 18.3%
- lower out-of-sample error
- RMSE raw → shrunk
- 11.1% → 9.1%
- odd vs. even weeks
- Split-half reliability
- 0.35
- how repeatable the raw stat is
- Stabilizes at
- 145
- targets to trust the number
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.
Catch rate · 2022 · full board
| # | Player | Team | Pos | targets | Raw | Shrunk | 90% interval |
|---|---|---|---|---|---|---|---|
| 1 | Will Dissly◦ provisional | SEA | TE | 38 | 89.5% | 72.6% | 67.1% – 77.9% |
| 2 | Dallas Goedert◦ provisional | PHI | TE | 69 | 79.7% | 71.9% | 66.8% – 76.8% |
| 3 | Robert Tonyan◦ provisional | GB | TE | 67 | 79.1% | 71.7% | 66.5% – 76.6% |
| 4 | Noah Fant◦ provisional | SEA | TE | 63 | 79.4% | 71.6% | 66.3% – 76.6% |
| 5 | Daniel Bellinger◦ provisional | NYG | TE | 36 | 83.3% | 71.2% | 65.6% – 76.6% |
| 6 | Mitchell Wilcox◦ provisional | CIN | TE | 18 | 94.4% | 71.1% | 65.2% – 76.8% |
| 7 | Johnny Mundt◦ provisional | MIN | TE | 21 | 90.5% | 71.0% | 65.1% – 76.7% |
| 8 | Noah Gray◦ provisional | KC | TE | 34 | 82.3% | 70.9% | 65.2% – 76.3% |
| 9 | Hayden Hurst◦ provisional | CIN | TE | 68 | 76.5% | 70.9% | 65.6% – 75.8% |
| 10 | Jake Ferguson◦ provisional | DAL | TE | 22 | 86.4% | 70.6% | 64.7% – 76.3% |
| 11 | Kylen Granson◦ provisional | IND | TE | 40 | 77.5% | 70.2% | 64.6% – 75.6% |
| 12 | Zach Gentry◦ provisional | PIT | TE | 23 | 82.6% | 70.2% | 64.3% – 75.8% |
| 13 | Travis Kelce | KC | TE | 153 | 71.9% | 70.1% | 65.7% – 74.4% |
| 14 | Adam Trautman◦ provisional | NO | TE | 22 | 81.8% | 70.0% | 64.1% – 75.7% |
| 15 | Dawson Knox◦ provisional | BUF | TE | 65 | 73.9% | 70.0% | 64.7% – 75.0% |
| 16 | Josiah Deguara◦ provisional | GB | TE | 15 | 86.7% | 70.0% | 63.9% – 75.8% |
| 17 | Evan Engram◦ provisional | JAX | TE | 101 | 72.3% | 69.9% | 65.0% – 74.6% |
| 18 | David Njoku◦ provisional | CLE | TE | 80 | 72.5% | 69.8% | 64.6% – 74.7% |
| 19 | C.J. Uzomah◦ provisional | NYJ | TE | 27 | 77.8% | 69.7% | 63.9% – 75.3% |
| 20 | Trey McBride◦ provisional | ARI | TE | 39 | 74.4% | 69.5% | 63.8% – 75.0% |
| 21 | Donald Parham◦ provisional | LAC | TE | 12 | 83.3% | 69.4% | 63.2% – 75.3% |
| 22 | Cole Kmet◦ provisional | CHI | TE | 70 | 71.4% | 69.3% | 64.0% – 74.3% |
| 23 | MyCole Pruitt◦ provisional | ATL | TE | 21 | 76.2% | 69.2% | 63.2% – 75.0% |
| 24 | Colby Parkinson◦ provisional | SEA | TE | 34 | 73.5% | 69.2% | 63.5% – 74.8% |
| 25 | Brock Wright◦ provisional | DET | TE | 24 | 75.0% | 69.2% | 63.2% – 74.9% |
| 26 | Jack Stoll◦ provisional | PHI | TE | 14 | 78.6% | 69.1% | 63.0% – 75.0% |
| 27 | Harrison Bryant◦ provisional | CLE | TE | 43 | 72.1% | 69.1% | 63.5% – 74.5% |
| 28 | Durham Smythe◦ provisional | MIA | TE | 20 | 75.0% | 69.0% | 63.0% – 74.8% |
| 29 | Geoff Swaim◦ provisional | TEN | TE | 16 | 75.0% | 68.9% | 62.8% – 74.8% |
| 30 | Jonnu Smith◦ provisional | NE | TE | 38 | 71.0% | 68.8% | 63.1% – 74.3% |
| 31 | George Kittle◦ provisional | SF | TE | 86 | 69.8% | 68.8% | 63.7% – 73.7% |
| 32 | Shane Zylstra◦ provisional | DET | TE | 15 | 73.3% | 68.7% | 62.6% – 74.6% |
| 33 | Hunter Henry◦ provisional | NE | TE | 59 | 69.5% | 68.6% | 63.2% – 73.8% |
| 34 | Ian Thomas◦ provisional | CAR | TE | 30 | 70.0% | 68.5% | 62.7% – 74.2% |
| 35 | Connor Heyward◦ provisional | PIT | TE | 17 | 70.6% | 68.5% | 62.4% – 74.3% |
| 36 | Irv Smith◦ provisional | MIN | TE | 36 | 69.4% | 68.5% | 62.7% – 74.0% |
| 37 | Eric Tomlinson◦ provisional | DEN | TE | 13 | 69.2% | 68.3% | 62.1% – 74.2% |
| 38 | Taysom Hill◦ provisional | NO | TE | 13 | 69.2% | 68.3% | 62.1% – 74.2% |
| 39 | Anthony Firkser◦ provisional | ATL | TE | 13 | 69.2% | 68.3% | 62.1% – 74.2% |
| 40 | Dan Arnold◦ provisional | JAX | TE | 13 | 69.2% | 68.3% | 62.1% – 74.2% |
| 41 | Joseph Fortson◦ provisional | KC | TE | 13 | 69.2% | 68.3% | 62.1% – 74.2% |
| 42 | Peyton Hendershot◦ provisional | DAL | TE | 16 | 68.8% | 68.3% | 62.1% – 74.2% |
| 43 | Austin Hooper◦ provisional | TEN | TE | 60 | 68.3% | 68.3% | 62.8% – 73.5% |
| 44 | Chig Okonkwo◦ provisional | TEN | TE | 47 | 68.1% | 68.2% | 62.6% – 73.6% |
| 45 | Mo Alie-Cox◦ provisional | IND | TE | 28 | 67.9% | 68.2% | 62.3% – 73.9% |
| 46 | Tanner Hudson◦ provisional | NYG | TE | 15 | 66.7% | 68.1% | 61.9% – 74.0% |
| 47 | Lawrence Cager◦ provisional | NYG | TE | 20 | 65.0% | 67.8% | 61.8% – 73.7% |
| 48 | Tyler Conklin◦ provisional | NYJ | TE | 87 | 66.7% | 67.6% | 62.5% – 72.6% |
| 49 | Jordan Akins◦ provisional | HOU | TE | 56 | 66.1% | 67.6% | 62.1% – 72.9% |
| 50 | John Bates◦ provisional | WAS | TE | 22 | 63.6% | 67.6% | 61.6% – 73.4% |
| 51 | Zach Ertz◦ provisional | ARI | TE | 71 | 66.2% | 67.6% | 62.3% – 72.7% |
| 52 | Tyler Higbee◦ provisional | LA | TE | 108 | 66.7% | 67.6% | 62.7% – 72.3% |
| 53 | Eric Saubert◦ provisional | DEN | TE | 24 | 62.5% | 67.4% | 61.4% – 73.2% |
| 54 | Gerald Everett◦ provisional | LAC | TE | 88 | 65.9% | 67.3% | 62.2% – 72.3% |
| 55 | Pharaoh Brown◦ provisional | CLE | TE | 20 | 60.0% | 67.2% | 61.1% – 73.1% |
| 56 | Darren Waller◦ provisional | LV | TE | 44 | 63.6% | 67.2% | 61.5% – 72.7% |
| 57 | Juwan Johnson◦ provisional | NO | TE | 65 | 64.6% | 67.1% | 61.7% – 72.3% |
| 58 | Cade Otton◦ provisional | TB | TE | 65 | 64.6% | 67.1% | 61.7% – 72.3% |
| 59 | T.J. Hockenson◦ provisional | MIN | TE | 131 | 65.6% | 67.0% | 62.3% – 71.6% |
| 60 | Jelani Woods◦ provisional | IND | TE | 40 | 62.5% | 67.0% | 61.2% – 72.6% |
| 61 | Logan Thomas◦ provisional | WAS | TE | 61 | 63.9% | 67.0% | 61.5% – 72.2% |
| 62 | Albert Okwuegbunam◦ provisional | DEN | TE | 18 | 55.6% | 66.8% | 60.7% – 72.8% |
| 63 | Tre' McKitty◦ provisional | LAC | TE | 18 | 55.6% | 66.8% | 60.7% – 72.8% |
| 64 | Dalton Schultz◦ provisional | DAL | TE | 89 | 64.0% | 66.6% | 61.5% – 71.6% |
| 65 | Pat Freiermuth◦ provisional | PIT | TE | 98 | 64.3% | 66.6% | 61.6% – 71.5% |
| 66 | Tommy Tremble◦ provisional | CAR | TE | 32 | 59.4% | 66.6% | 60.7% – 72.3% |
| 67 | Teagan Quitoriano◦ provisional | HOU | TE | 14 | 50.0% | 66.6% | 60.4% – 72.6% |
| 68 | Josh Oliver◦ provisional | BAL | TE | 25 | 56.0% | 66.4% | 60.4% – 72.3% |
| 69 | Mark Andrews◦ provisional | BAL | TE | 114 | 64.0% | 66.4% | 61.5% – 71.1% |
| 70 | Foster Moreau◦ provisional | LV | TE | 54 | 61.1% | 66.3% | 60.7% – 71.7% |
| 71 | Mike Gesicki◦ provisional | MIA | TE | 53 | 60.4% | 66.1% | 60.5% – 71.5% |
| 72 | Greg Dulcich◦ provisional | DEN | TE | 55 | 60.0% | 66.0% | 60.4% – 71.4% |
| 73 | Isaiah Likely◦ provisional | BAL | TE | 60 | 60.0% | 65.8% | 60.3% – 71.2% |
| 74 | Brevin Jordan◦ provisional | HOU | TE | 28 | 50.0% | 65.3% | 59.3% – 71.1% |
| 75 | Cameron Brate◦ provisional | TB | TE | 38 | 52.6% | 65.0% | 59.1% – 70.7% |
| 76 | O.J. Howard◦ provisional | HOU | TE | 23 | 43.5% | 64.8% | 58.7% – 70.8% |
| 77 | Kyle Pitts◦ provisional | ATL | TE | 59 | 47.5% | 62.2% | 56.6% – 67.7% |
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).