Head coaches · Analytics
Coaching decisions
Every 4th down since 2010, and every 4th-quarter try after a touchdown, scored by win probability: the chance of winning with each option, against the call the coach made. The grade is how little win probability a coach gives away per game, compared with every coach-season since 2010.
Stats through Week 10 and part of Week 11updated Sat 8:20 pm ET
Head coaches 2026 season to date
Sorted by this season's grade. Grades are 0–100: the share of coach-seasons since 2010 that gave away more win probability per game on 4th downs.
How to read
One game's calls say little (most of the gap between two coaches in a game is luck of the situations), so every season is blended with 24 games of league-average play before it is ranked: early-season grades sit near 50 and only a long run of calls moves a coach to the ends.
| Coach | Team | 2026 record | 2026 grade | Calls | Went when right | Win prob. lost / game | 2025 grade | 2025 went when right | 2025 win prob. lost / game | Career rank |
|---|---|---|---|---|---|---|---|---|---|---|
| 6-4 | 97 | 65 | 83% 5 of 6 | 1.0 pp | 85 | 63% | 2.6 pp | 2nd of 73 | ||
| 8-1 | 92 | 55 | 78% 7 of 9 | 1.5 pp | – | – | – | new | ||
| 3-6 | 91 | 67 | 83% 5 of 6 | 1.5 pp | 74 | 73% | 2.9 pp | 39th of 73 | ||
| 4-6 | 85 | 62 | 75% 6 of 8 | 2.0 pp | – | – | – | new | ||
| 6-4 | 80 | 81 | 63% 5 of 8 | 2.3 pp | – | – | – | new | ||
| 2-7 | 74 | 69 | 60% 3 of 5 | 2.5 pp | 93 | 70% | 2.2 pp | 3rd of 73 | ||
| 7-2 | 72 | 56 | 73% 8 of 11 | 2.5 pp | 58 | 75% | 3.4 pp | new | ||
| 7-2 | 72 | 59 | 83% 5 of 6 | 2.5 pp | 95 | 76% | 2.1 pp | 4th of 73 | ||
| 5-5 | 70 | 56 | 69% 9 of 13 | 2.7 pp | 96 | 71% | 2.2 pp | 15th of 73 | ||
| 6-4 | 67 | 61 | 44% 4 of 9 | 2.8 pp | 82 | 72% | 2.8 pp | 34th of 73 | ||
| 4-5 | 67 | 67 | 54% 7 of 13 | 2.8 pp | 64 | 67% | 3.2 pp | new | ||
| 3-6 | 67 | 66 | 64% 9 of 14 | 2.8 pp | 83 | 74% | 2.6 pp | new | ||
| 1-8 | 65 | 61 | 33% 3 of 9 | 2.9 pp | – | – | – | 27th of 73 | ||
| 4-6 | 63 | 68 | 40% 4 of 10 | 3.0 pp | 9 | 40% | 5.2 pp | 42nd of 73 | ||
| 3-7 | 60 | 62 | 40% 2 of 5 | 3.2 pp | 31 | 65% | 4.2 pp | 10th of 73 | ||
| 8-1 | 60 | 41 | 57% 4 of 7 | 3.1 pp | 73 | 56% | 3.0 pp | 11th of 73 | ||
| 4-6 | 60 | 70 | 57% 8 of 14 | 3.2 pp | 99 | 77% | 1.6 pp | new | ||
| 7-2 | 59 | 71 | 50% 5 of 10 | 3.2 pp | 54 | 53% | 3.5 pp | 22nd of 73 | ||
| 7-2 | 58 | 45 | 56% 5 of 9 | 3.3 pp | 92 | 68% | 2.3 pp | new | ||
| 7-2 | 56 | 54 | 69% 9 of 13 | 3.3 pp | 52 | 70% | 3.6 pp | 26th of 73 | ||
| 7-2 | 53 | 44 | 43% 3 of 7 | 3.4 pp | 59 | 61% | 3.4 pp | 12th of 73 | ||
| 6-3 | 47 | 63 | 50% 3 of 6 | 3.7 pp | – | – | – | 24th of 73 | ||
| 5-3-1 | 44 | 68 | 45% 5 of 11 | 3.8 pp | 76 | 56% | 2.9 pp | 29th of 73 | ||
| 1-8 | 39 | 56 | 80% 4 of 5 | 4.0 pp | – | – | – | new | ||
| 3-7 | 39 | 77 | 50% 7 of 14 | 4.0 pp | – | – | – | new | ||
| 4-6 | 36 | 66 | 38% 5 of 13 | 4.1 pp | – | – | – | new | ||
| 2-8 | 35 | 74 | 50% 7 of 14 | 4.2 pp | 67 | 62% | 3.1 pp | new | ||
| 6-4 | 31 | 67 | 31% 4 of 13 | 4.4 pp | 90 | 33% | 2.5 pp | new | ||
| 7-2 | 27 | 69 | 25% 3 of 12 | 4.7 pp | 47 | 53% | 3.7 pp | 41st of 73 | ||
| 3-5-1 | 22 | 72 | 46% 6 of 13 | 5.0 pp | 79 | 56% | 2.7 pp | 32nd of 73 | ||
| 1-8 | 22 | 64 | 25% 2 of 8 | 5.0 pp | 30 | 24% | 4.2 pp | 55th of 73 | ||
| 3-8 | 11 | 76 | 47% 9 of 19 | 5.7 pp | 93 | 71% | 2.2 pp | 36th of 73 |
The league learned to go for it
Share of 4th downs where going was clearly the better call that teams actually went for it, by season. 2026 is the season so far.
And stopped giving games away
Win probability lost per team-game on 4th-down calls, in percentage points (pp), by season. 2026 is the season so far.
All-time, since 2010
Head coaches with at least 48 graded games (about three seasons) of games, ranked by win probability lost per game on 4th downs (lower is better).
How to read
Earlier eras were more conservative, so long careers that started before 2018 carry that cost.
| # | Coach | Seasons | Games | Calls | Went when right | Win prob. lost / game |
|---|---|---|---|---|---|---|
| 1 | Jonathan GannonJ. Gannon | 2023–2025 | 51 | 312 | 62% | 2.2 pp |
| 2 | Matt LaFleurM. LaFleur | 2019–2026 | 136 | 778 | 62% | 2.3 pp |
| 3 | Nick SirianniN. Sirianni | 2021–2026 | 104 | 654 | 63% | 2.5 pp |
| 4 | Dan CampbellD. Campbell | 2021–2026 | 98 | 609 | 72% | 2.5 pp |
| 5 | Sean McDermottS. McDermott | 2017–2025 | 164 | 964 | 59% | 2.6 pp |
| 6 | Kliff KingsburyK. Kingsbury | 2019–2022 | 67 | 440 | 65% | 2.7 pp |
| 7 | Arthur SmithA. Smith | 2021–2023 | 51 | 343 | 57% | 2.7 pp |
| 8 | Brian FloresB. Flores | 2019–2021 | 49 | 317 | 41% | 2.8 pp |
| 9 | Mike McDanielM. McDaniel | 2022–2025 | 70 | 425 | 53% | 2.9 pp |
| 10 | Kevin StefanskiK. Stefanski | 2020–2026 | 114 | 804 | 66% | 2.9 pp |
| 11 | Kyle ShanahanK. Shanahan | 2017–2026 | 172 | 957 | 45% | 3.1 pp |
| 12 | Zac TaylorZ. Taylor | 2019–2026 | 132 | 804 | 50% | 3.2 pp |
| 13 | Frank ReichF. Reich | 2018–2023 | 88 | 573 | 64% | 3.2 pp |
| 14 | Matt NagyM. Nagy | 2018–2021 | 67 | 452 | 54% | 3.2 pp |
| 15 | Mike VrabelM. Vrabel | 2018–2026 | 135 | 848 | 49% | 3.2 pp |
| 16 | Jon GrudenJ. Gruden | 2018–2021 | 53 | 337 | 60% | 3.3 pp |
| 17 | Norv TurnerN. Turner | 2010–2012 | 48 | 275 | 34% | 3.3 pp |
| 18 | Brandon StaleyB. Staley | 2021–2023 | 49 | 346 | 62% | 3.3 pp |
| 19 | Adam GaseA. Gase | 2016–2020 | 81 | 527 | 36% | 3.3 pp |
| 20 | Matt EberflusM. Eberflus | 2022–2024 | 51 | 352 | 56% | 3.3 pp |
| 21 | Leslie FrazierL. Frazier | 2010–2013 | 55 | 370 | 38% | 3.3 pp |
| 22 | Kevin O'ConnellK. O'Connell | 2022–2026 | 79 | 527 | 49% | 3.3 pp |
| 23 | Dirk KoetterD. Koetter | 2016–2018 | 48 | 296 | 30% | 3.4 pp |
| 24 | Mike McCarthyM. McCarthy | 2010–2026 | 252 | 1,541 | 42% | 3.4 pp |
| 25 | Doug PedersonD. Pederson | 2016–2024 | 139 | 939 | 53% | 3.4 pp |
| 26 | Andy ReidA. Reid | 2010–2026 | 297 | 1,829 | 48% | 3.4 pp |
| 27 | Robert SalehR. Saleh | 2021–2026 | 77 | 563 | 44% | 3.4 pp |
| 28 | Mike ZimmerM. Zimmer | 2014–2021 | 134 | 868 | 39% | 3.5 pp |
| 29 | John HarbaughJ. Harbaugh | 2010–2026 | 289 | 1,909 | 47% | 3.5 pp |
| 30 | Gus BradleyG. Bradley | 2013–2016 | 62 | 446 | 31% | 3.5 pp |
| 31 | Ron RiveraR. Rivera | 2011–2023 | 219 | 1,435 | 46% | 3.5 pp |
| 32 | Dan QuinnD. Quinn | 2015–2026 | 136 | 840 | 51% | 3.5 pp |
| 33 | Raheem MorrisR. Morris | 2010–2025 | 77 | 511 | 53% | 3.6 pp |
| 34 | Sean McVayS. McVay | 2017–2026 | 175 | 1,091 | 45% | 3.6 pp |
| 35 | Chan GaileyC. Gailey | 2010–2012 | 48 | 304 | 38% | 3.6 pp |
| 36 | Shane SteichenS. Steichen | 2023–2026 | 62 | 427 | 60% | 3.6 pp |
| 37 | Chip KellyC. Kelly | 2013–2016 | 64 | 454 | 37% | 3.6 pp |
| 38 | Hue JacksonH. Jackson | 2011–2018 | 56 | 393 | 43% | 3.7 pp |
| 39 | Todd BowlesT. Bowles | 2011–2026 | 148 | 1,023 | 44% | 3.7 pp |
| 40 | Chuck PaganoC. Pagano | 2012–2017 | 102 | 672 | 34% | 3.8 pp |
| 41 | Sean PaytonS. Payton | 2010–2026 | 252 | 1,561 | 42% | 3.8 pp |
| 42 | DeMeco RyansD. Ryans | 2023–2026 | 67 | 484 | 40% | 3.8 pp |
| 43 | Dennis AllenD. Allen | 2012–2024 | 87 | 587 | 36% | 3.8 pp |
| 44 | Brian DabollB. Daboll | 2022–2025 | 70 | 493 | 56% | 3.9 pp |
| 45 | Bill BelichickB. Belichick | 2010–2023 | 251 | 1,499 | 35% | 3.9 pp |
| 46 | Joe PhilbinJ. Philbin | 2012–2018 | 68 | 458 | 31% | 3.9 pp |
| 47 | Anthony LynnA. Lynn | 2016–2020 | 67 | 410 | 42% | 3.9 pp |
| 48 | Jeff FisherJ. Fisher | 2010–2016 | 96 | 654 | 30% | 4.0 pp |
| 49 | Jim CaldwellJ. Caldwell | 2010–2017 | 99 | 660 | 31% | 4.0 pp |
| 50 | Jason GarrettJ. Garrett | 2010–2019 | 157 | 1,003 | 36% | 4.0 pp |
| 51 | John FoxJ. Fox | 2010–2017 | 135 | 845 | 28% | 4.1 pp |
| 52 | Bill O'BrienB. O'Brien | 2014–2020 | 106 | 758 | 37% | 4.1 pp |
| 53 | Pete CarrollP. Carroll | 2010–2025 | 263 | 1,741 | 34% | 4.1 pp |
| 54 | Jack Del RioJ. Del Rio | 2010–2017 | 76 | 529 | 38% | 4.1 pp |
| 55 | Jim HarbaughJ. Harbaugh | 2011–2026 | 117 | 796 | 29% | 4.1 pp |
| 56 | Lovie SmithL. Smith | 2010–2022 | 99 | 643 | 31% | 4.1 pp |
| 57 | Vic FangioV. Fangio | 2019–2021 | 49 | 331 | 45% | 4.1 pp |
| 58 | Tom CoughlinT. Coughlin | 2010–2015 | 100 | 668 | 31% | 4.2 pp |
| 59 | Pat ShurmurP. Shurmur | 2011–2019 | 65 | 470 | 41% | 4.2 pp |
| 60 | Bruce AriansB. Arians | 2013–2021 | 138 | 885 | 28% | 4.2 pp |
| 61 | Mike TomlinM. Tomlin | 2010–2025 | 277 | 1,868 | 35% | 4.3 pp |
| 62 | Jay GrudenJ. Gruden | 2014–2019 | 86 | 563 | 29% | 4.3 pp |
| 63 | Mike McCoyM. McCoy | 2013–2016 | 66 | 420 | 34% | 4.3 pp |
| 64 | Mike ShanahanM. Shanahan | 2010–2013 | 65 | 463 | 40% | 4.3 pp |
| 65 | Mike SmithM. Smith | 2010–2014 | 84 | 520 | 24% | 4.5 pp |
| 66 | Ken WhisenhuntK. Whisenhunt | 2010–2015 | 80 | 583 | 26% | 4.5 pp |
| 67 | Doug MarroneD. Marrone | 2013–2020 | 101 | 729 | 26% | 4.5 pp |
| 68 | Marvin LewisM. Lewis | 2010–2018 | 149 | 1,017 | 33% | 4.5 pp |
| 69 | Jim SchwartzJ. Schwartz | 2010–2013 | 65 | 472 | 35% | 4.6 pp |
| 70 | Rex RyanR. Ryan | 2010–2016 | 114 | 802 | 27% | 4.7 pp |
| 71 | Mike MunchakM. Munchak | 2011–2013 | 48 | 345 | 36% | 4.7 pp |
| 72 | Mike MularkeyM. Mularkey | 2012–2017 | 50 | 350 | 27% | 4.8 pp |
| 73 | Gary KubiakG. Kubiak | 2010–2016 | 100 | 741 | 28% | 5.1 pp |
How the grades work
- Win probability. A model trained on every play since 2010 reproduces nflfastR's win probability for any situation, including the ones a coach did not choose.
- Options. Going for it uses the chance of converting at that distance and spot, and where a conversion usually leaves the ball; a field goal uses the make rate at that distance; a punt uses where punts from that spot usually leave the other team.
- Tries after touchdowns. Kick or go for two is scored with a possession model: the other team receives, possessions alternate until time runs out, and each scores at the league's current touchdown and field-goal rates. It reproduces the classic two-point chart's clear cases with today's scoring. Tries are graded in the 4th quarter, where they matter.
- Each team's roster. The pregame spread tells the model how good each team is overall. On top of that, the chance of converting reflects that offense against that defense, and win probability reflects how each side's strength splits between offense and defense (with a strong defense, giving the ball back costs less). Both use only games played before the call. They change the best call on about 4% of 4th downs.
- Cost. The best option's win probability minus the chosen one's. Calls within about one point of win probability are toss-ups and cost almost nothing.
- Not graded. Overtime (different rules) and games already decided.