The Poll.ai

2025 season · through week 3

Too early to call this a ranking.

Ranked by how hard each record was to get against the schedule that produced it. No preseason poll, no reputation, no conference name.

Switch below and watch which teams move.

ProvisionalPROVISIONAL. THE POLL OPENS IN WEEK 5. SEE docs/constraints.md.
  1. 1Georgia3-03-0Gap +51.21 in 5key 0.73Ranked 1. The model replayed this schedule 1,000 times, and this team finished between 2 and 110 of 136 in 90% of them.
  2. 2Oklahoma3-03-0Gap +47.81 in 5key 0.67Ranked 2. The model replayed this schedule 1,000 times, and this team finished between 3 and 92 of 136 in 90% of them.
  3. 3Utah3-03-0Gap +39.51 in 5key 0.66Ranked 3. The model replayed this schedule 1,000 times, and this team finished between 4 and 87 of 136 in 90% of them.
  4. 4NC State3-03-0Gap +55.61 in 4key 0.65Ranked 4. The model replayed this schedule 1,000 times, and this team finished between 6 and 128 of 136 in 90% of them.
  5. 5Memphis3-03-0Gap +42.21 in 4key 0.65Ranked 5. The model replayed this schedule 1,000 times, and this team finished between 5 and 98 of 136 in 90% of them.
  6. 6LSU3-03-0Gap +60.51 in 4key 0.56Ranked 6. The model replayed this schedule 1,000 times, and this team finished between 11 and 131 of 136 in 90% of them.
  7. 7Ole Miss3-03-0Gap +56.21 in 4key 0.55Ranked 7. The model replayed this schedule 1,000 times, and this team finished between 9 and 130 of 136 in 90% of them.
  8. 8Iowa State4-04-0Gap +55.71 in 4key 0.55Ranked 8. The model replayed this schedule 1,000 times, and this team finished between 13 and 120 of 136 in 90% of them.
  9. 9Houston3-03-0Gap +53.21 in 3key 0.54Ranked 9. The model replayed this schedule 1,000 times, and this team finished between 9 and 113 of 136 in 90% of them.
  10. 10Texas A&M3-03-0Gap +52.71 in 3key 0.53Ranked 10. The model replayed this schedule 1,000 times, and this team finished between 9 and 111 of 136 in 90% of them.
  11. 11Mississippi State3-03-0Gap +48.91 in 3key 0.52Ranked 11. The model replayed this schedule 1,000 times, and this team finished between 8 and 114 of 136 in 90% of them.
  12. 12Tulane3-03-0Gap +61.01 in 3key 0.48Ranked 12. The model replayed this schedule 1,000 times, and this team finished between 13 and 133 of 136 in 90% of them.
  13. 13California3-03-0Gap +56.11 in 3key 0.47Ranked 13. The model replayed this schedule 1,000 times, and this team finished between 14 and 133 of 136 in 90% of them.
  14. 14Navy3-03-0Gap +46.51 in 3key 0.46Ranked 14. The model replayed this schedule 1,000 times, and this team finished between 11 and 114 of 136 in 90% of them.
  15. 15Auburn3-03-0Gap +53.51 in 3key 0.46Ranked 15. The model replayed this schedule 1,000 times, and this team finished between 10 and 120 of 136 in 90% of them.
  16. 16Illinois3-03-0Gap +50.61 in 3key 0.45Ranked 16. The model replayed this schedule 1,000 times, and this team finished between 10 and 119 of 136 in 90% of them.
  17. 17Vanderbilt3-03-0Gap +40.11 in 3key 0.44Ranked 17. The model replayed this schedule 1,000 times, and this team finished between 13 and 109 of 136 in 90% of them.
  18. 18Georgia Tech3-03-0Gap +56.51 in 3key 0.43Ranked 18. The model replayed this schedule 1,000 times, and this team finished between 12 and 121 of 136 in 90% of them.
  19. 19USC3-03-0Gap +39.91 in 3key 0.41Ranked 19. The model replayed this schedule 1,000 times, and this team finished between 11 and 90 of 136 in 90% of them.
  20. 20Nebraska3-03-0Gap +46.11 in 2key 0.39Ranked 20. The model replayed this schedule 1,000 times, and this team finished between 14 and 116 of 136 in 90% of them.
  21. 21Indiana3-03-0Gap +49.01 in 2key 0.39Ranked 21. The model replayed this schedule 1,000 times, and this team finished between 14 and 119 of 136 in 90% of them.
  22. 22Oregon3-03-0Gap +40.11 in 2key 0.37Ranked 22. The model replayed this schedule 1,000 times, and this team finished between 16 and 112 of 136 in 90% of them.
  23. 23North Texas3-03-0Gap +43.11 in 2key 0.35Ranked 23. The model replayed this schedule 1,000 times, and this team finished between 18 and 86 of 136 in 90% of them.
  24. 24Ohio State3-03-0Gap +52.41 in 2key 0.35Ranked 24. The model replayed this schedule 1,000 times, and this team finished between 20 and 118 of 136 in 90% of them.
  25. 25Maryland3-03-0Gap +51.31 in 2key 0.33Ranked 25. The model replayed this schedule 1,000 times, and this team finished between 21 and 117 of 136 in 90% of them.

The right-hand number is how hard that season was to pull off against that exact schedule. One in 5 put Georgia on top. One in 2 was enough for 25th. Below about 1 in 5 it stops separating teams, which is what the smaller key underneath is for. The bar under each name is where the model kept landing that team out of 136, every bar on one scale. A typical bar this week spans 110.5 places.

2025 week 3 · 5ec6d1fe · code e215160 · The House Poll · method 061a6d4e · 136 teams ranked

Every column

The board above is the thirty-second version. Below it is everything the run published: the résumé, the power rating, the gap, the rank interval, and where the finished season puts each team.

Every figure the cards carry, for the 25 on the board, with nothing to click.
#teamrecordhow unlikely1 in this manyexpectedwins this schedule asked forbeat that bypowerrating (rank)gapplaces from the pollrangein 90% of replayshindsight
1Georgia3-01 in 51.8+1.28.8 (31)+51.22 to 110-4.0
2Oklahoma3-01 in 51.8+1.212.2 (15)+47.83 to 92-4.0
3Utah3-01 in 51.8+1.220.5 (1)+39.54 to 87-20.0
4NC State3-01 in 41.8+1.24.4 (53)+55.66 to 1280.0
5Memphis3-01 in 41.9+1.117.8 (6)+42.25 to 98-24.0
6LSU3-01 in 42.0+1.0-0.5 (85)+60.511 to 131-2.0
7Ole Miss3-01 in 42.0+1.03.8 (58)+56.29 to 130-6.0
8Iowa State4-01 in 42.9+1.14.3 (54)+55.713 to 120+5.0
9Houston3-01 in 32.0+1.06.8 (42)+53.29 to 113-13.0
10Texas A&M3-01 in 32.0+1.07.3 (39)+52.79 to 111+9.0
11Mississippi State3-01 in 32.1+0.911.1 (19)+48.98 to 114-7.0
12Tulane3-01 in 32.1+0.9-1.0 (88)+61.013 to 133-4.0
13California3-01 in 32.1+0.93.9 (56)+56.114 to 133-17.0
14Navy3-01 in 32.1+0.913.5 (11)+46.511 to 114-26.0
15Auburn3-01 in 32.2+0.86.5 (44)+53.510 to 120-13.0
16Illinois3-01 in 32.2+0.89.4 (29)+50.610 to 119+6.0
17Vanderbilt3-01 in 32.2+0.819.9 (4)+40.113 to 109+2.0
18Georgia Tech3-01 in 32.2+0.83.5 (60)+56.512 to 121-2.0
19USC3-01 in 32.3+0.720.1 (3)+39.911 to 90-6.0
20Nebraska3-01 in 22.3+0.713.9 (10)+46.114 to 116-7.0
21Indiana3-01 in 22.3+0.711.0 (21)+49.014 to 119+9.0
22Oregon3-01 in 22.3+0.719.9 (5)+40.116 to 112+15.0
23North Texas3-01 in 22.3+0.716.9 (7)+43.118 to 86+14.0
24Ohio State3-01 in 22.3+0.77.6 (38)+52.420 to 118+13.0
25Maryland3-01 in 22.3+0.78.7 (33)+51.321 to 117-21.0
Show the full table: top 25 of 136, every columnthe published poll
ProvisionalPROVISIONAL. THE POLL OPENS IN WEEK 5. SEE docs/constraints.md.
#teamunderline: 90% rank interval · league median width 110.5how unlikely1 in this many
1GeorgiaSEC3-0·Power 8.85 (31)·Gap +51.15·▼4Ranked 1. The model replayed this schedule 1,000 times, and this team finished between 2 and 110 of 136 in 90% of them.0.7311 in 5
2OklahomaSEC3-0·Power 12.17 (15)·Gap +47.83·▼4Ranked 2. The model replayed this schedule 1,000 times, and this team finished between 3 and 92 of 136 in 90% of them.0.6661 in 5
3UtahBig 123-0·Power 20.51 (1)·Gap +39.49·▼20Ranked 3. The model replayed this schedule 1,000 times, and this team finished between 4 and 87 of 136 in 90% of them.0.6581 in 5
4NC StateACC3-0·Power 4.44 (53)·Gap +55.56Ranked 4. The model replayed this schedule 1,000 times, and this team finished between 6 and 128 of 136 in 90% of them.0.6511 in 4
5MemphisAmerican Athletic3-0·Power 17.79 (6)·Gap +42.21·▼24Ranked 5. The model replayed this schedule 1,000 times, and this team finished between 5 and 98 of 136 in 90% of them.0.6481 in 4
6LSUSEC3-0·Power -0.48 (85)·Gap +60.48·▼2Ranked 6. The model replayed this schedule 1,000 times, and this team finished between 11 and 131 of 136 in 90% of them.0.5581 in 4
7Ole MissSEC3-0·Power 3.76 (58)·Gap +56.24·▼6Ranked 7. The model replayed this schedule 1,000 times, and this team finished between 9 and 130 of 136 in 90% of them.0.5541 in 4
8Iowa StateBig 124-0·Power 4.32 (54)·Gap +55.68·▲5Ranked 8. The model replayed this schedule 1,000 times, and this team finished between 13 and 120 of 136 in 90% of them.0.5491 in 4
9HoustonBig 123-0·Power 6.79 (42)·Gap +53.21·▼13Ranked 9. The model replayed this schedule 1,000 times, and this team finished between 9 and 113 of 136 in 90% of them.0.5401 in 3
10Texas A&MSEC3-0·Power 7.33 (39)·Gap +52.67·▲9Ranked 10. The model replayed this schedule 1,000 times, and this team finished between 9 and 111 of 136 in 90% of them.0.5341 in 3
11Mississippi StateSEC3-0·Power 11.11 (19)·Gap +48.89·▼7Ranked 11. The model replayed this schedule 1,000 times, and this team finished between 8 and 114 of 136 in 90% of them.0.5181 in 3
12TulaneAmerican Athletic3-0·Power -1.00 (88)·Gap +61.00·▼4Ranked 12. The model replayed this schedule 1,000 times, and this team finished between 13 and 133 of 136 in 90% of them.0.4801 in 3
13CaliforniaACC3-0·Power 3.89 (56)·Gap +56.11·▼17Ranked 13. The model replayed this schedule 1,000 times, and this team finished between 14 and 133 of 136 in 90% of them.0.4691 in 3
14NavyAmerican Athletic3-0·Power 13.47 (11)·Gap +46.53·▼26Ranked 14. The model replayed this schedule 1,000 times, and this team finished between 11 and 114 of 136 in 90% of them.0.4601 in 3
15AuburnSEC3-0·Power 6.50 (44)·Gap +53.50·▼13Ranked 15. The model replayed this schedule 1,000 times, and this team finished between 10 and 120 of 136 in 90% of them.0.4571 in 3
16IllinoisBig Ten3-0·Power 9.40 (29)·Gap +50.60·▲6Ranked 16. The model replayed this schedule 1,000 times, and this team finished between 10 and 119 of 136 in 90% of them.0.4541 in 3
17VanderbiltSEC3-0·Power 19.87 (4)·Gap +40.13·▲2Ranked 17. The model replayed this schedule 1,000 times, and this team finished between 13 and 109 of 136 in 90% of them.0.4411 in 3
18Georgia TechACC3-0·Power 3.52 (60)·Gap +56.48·▼2Ranked 18. The model replayed this schedule 1,000 times, and this team finished between 12 and 121 of 136 in 90% of them.0.4281 in 3
19USCBig Ten3-0·Power 20.13 (3)·Gap +39.87·▼6Ranked 19. The model replayed this schedule 1,000 times, and this team finished between 11 and 90 of 136 in 90% of them.0.4061 in 3
20NebraskaBig Ten3-0·Power 13.89 (10)·Gap +46.11·▼7Ranked 20. The model replayed this schedule 1,000 times, and this team finished between 14 and 116 of 136 in 90% of them.0.3861 in 2
21IndianaBig Ten3-0·Power 11.00 (21)·Gap +49.00·▲9Ranked 21. The model replayed this schedule 1,000 times, and this team finished between 14 and 119 of 136 in 90% of them.0.3851 in 2
22OregonBig Ten3-0·Power 19.87 (5)·Gap +40.13·▲15Ranked 22. The model replayed this schedule 1,000 times, and this team finished between 16 and 112 of 136 in 90% of them.0.3721 in 2
23North TexasAmerican Athletic3-0·Power 16.89 (7)·Gap +43.11·▲14Ranked 23. The model replayed this schedule 1,000 times, and this team finished between 18 and 86 of 136 in 90% of them.0.3491 in 2
24Ohio StateBig Ten3-0·Power 7.58 (38)·Gap +52.42·▲13Ranked 24. The model replayed this schedule 1,000 times, and this team finished between 20 and 118 of 136 in 90% of them.0.3481 in 2
25MarylandBig Ten3-0·Power 8.68 (33)·Gap +51.32·▼21Ranked 25. The model replayed this schedule 1,000 times, and this team finished between 21 and 117 of 136 in 90% of them.0.3331 in 2

What this board believes football results are for

The House Poll, in its own words

This is the poll this project publishes and this recipe changes nothing about it. Margin counts, and it counts less and less as it grows: the response is C*tanh(m/C) with C at 32 points plus a 7 point win premium, so the gap between a 3 point win and a 24 point win is large and the gap between a 45 point win and a 70 point win is almost nothing. There is no cliff anywhere for a coach to aim at, which is what tanh buys over a hard cap. Then the headline does not read margin at all. Teams are ordered by schedule odds, the probability that a team of published reference quality would have gone at least this well against this exact schedule, and the harder it was to do what you did the higher you go. Margin is in the engine because it carries real information about how good a team is. It is out of the headline because the headline is about what a team earned, and those are two different questions that deserve two different columns.

what it costs

  • An unbeaten team can finish behind a one loss team, and that will need explaining every single year. It is the direct consequence of the promise. The explanation is on the page: the tail probability, the reference team it was measured against by name, and the Power column beside it.
  • One published constant the wins based resume did not need. q_ref is the Power rating of the 25th ranked team that week, the least flattering defensible reading of ESPN's average Top-25 team. The study measured the ordering across a 16 point swing in it: Kendall's tau never fell below 0.985 and at most one team entered or left the top 25.
  • It loses forward ordering accuracy to Full Merit by about two points. Accepted, and for a stated reason: forward accuracy is a prediction metric and the headline poll is not the instrument this project ships for prediction. L3 Power is, and it beats all three orderings on that axis.
  • It is a compromise and it will satisfy neither end of the argument completely. Someone who believes margin is the only honest measurement will find the compression arbitrary, and someone who believes point differential should never pay will find 32 points of it far too generous. Both are reading the position correctly.

what it changed

Nothing at all. This is the poll this project publishes and this recipe changes none of it.

configs/recipes/house.toml

same evidence, different valuesEvery board on this page was fitted on the same 397 games, and the digest of that exact frame is d0b206fb79ed under all three. Switching lens changes the constants above and nothing else. Method digest for this one: 061a6d4eda23.

Where next

  • Tune it yourself Move one setting and watch the board move.
  • The revision What this week looked like once the season ended.
  • Connectivity How much of the field was comparable, and what held it together.
  • Methodology Every constant, every baseline, and the lines that say FAIL.
run 5ec6d1fe · published 2026-08-19 23:11:07 UTC · code e215160 · config 1ef7cf23…
q_ref 10.13 (James Madison) · β_w 7 · C 32 · h 9.625 · σ 23.048 · λ₁ 200 · λ₂ 4.0 · k 102.26 · w₁ 0.7988 · w₂ -0.3873
reproduce: uv run cfbpoll rank --season 2025 --through-week 3 --out out/
Close

rank 18

Georgia Tech

3-0 · ACC. The model never reads conference names.

1 in 3how hard that record waskey 0.43
2.2wins this schedule asked for
+0.8how far past it they came

where the model can honestly pin this team, out of 136

Ranked 18. The model replayed this schedule 1,000 times, and this team finished between 12 and 121 of 136 in 90% of them.

12 to 121 · 109 places wide

power3.5, ranked 60
gap+56.5
resume60.0, ranked 7
hindsight20 · -2.0 against this rank

This team’s wins-based resume hit the bound the model puts on how much one result can be worth, so the figure above understates it.