The Poll.ai

2025 season · through week 8

Nobody voted on this.

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.

  1. 1Indiana7-07-0Gap +40.21 in 52key 1.72Ranked 1. The model replayed this schedule 1,000 times, and this team finished between 1 and 55 of 136 in 90% of them.
  2. 2Texas A&M7-07-0Gap +46.61 in 30key 1.47Ranked 2. The model replayed this schedule 1,000 times, and this team finished between 2 and 85 of 136 in 90% of them.
  3. 3Ohio State7-07-0Gap +39.71 in 22key 1.35Ranked 3. The model replayed this schedule 1,000 times, and this team finished between 1 and 58 of 136 in 90% of them.
  4. 4BYU7-07-0Gap +47.21 in 21key 1.33Ranked 4. The model replayed this schedule 1,000 times, and this team finished between 3 and 91 of 136 in 90% of them.
  5. 5Georgia Tech7-07-0Gap +50.31 in 18key 1.26Ranked 5. The model replayed this schedule 1,000 times, and this team finished between 8 and 97 of 136 in 90% of them.
  6. 6Alabama6-16-1Gap +12.41 in 10key 1.01Ranked 6. The model replayed this schedule 1,000 times, and this team finished between 1 and 72 of 136 in 90% of them.
  7. 7Georgia6-16-1Gap +14.11 in 6key 0.78Ranked 7. The model replayed this schedule 1,000 times, and this team finished between 4 and 103 of 136 in 90% of them.
  8. 8South Florida6-16-1Gap +13.71 in 6key 0.76Ranked 8. The model replayed this schedule 1,000 times, and this team finished between 4 and 89 of 136 in 90% of them.
  9. 9Texas Tech6-16-1Gap +5.01 in 6key 0.75Ranked 9. The model replayed this schedule 1,000 times, and this team finished between 1 and 56 of 136 in 90% of them.
  10. 10Oregon6-16-1Gap +3.21 in 5key 0.66Ranked 10. The model replayed this schedule 1,000 times, and this team finished between 1 and 54 of 136 in 90% of them.
  11. 11Oklahoma6-16-1Gap +10.21 in 4key 0.64Ranked 11. The model replayed this schedule 1,000 times, and this team finished between 2 and 87 of 136 in 90% of them.
  12. 12Navy6-06-0Gap +51.91 in 4key 0.64Ranked 12. The model replayed this schedule 1,000 times, and this team finished between 8 and 107 of 136 in 90% of them.
  13. 13Louisville5-15-1Gap +10.81 in 4key 0.62Ranked 13. The model replayed this schedule 1,000 times, and this team finished between 4 and 100 of 136 in 90% of them.
  14. 14Miami5-15-1Gap +8.31 in 4key 0.59Ranked 14. The model replayed this schedule 1,000 times, and this team finished between 2 and 91 of 136 in 90% of them.
  15. 15Virginia6-16-1Gap +12.61 in 4key 0.55Ranked 15. The model replayed this schedule 1,000 times, and this team finished between 10 and 101 of 136 in 90% of them.
  16. 16Vanderbilt6-16-1Gap +4.21 in 4key 0.55Ranked 16. The model replayed this schedule 1,000 times, and this team finished between 2 and 66 of 136 in 90% of them.
  17. 17Ole Miss6-16-1Gap +15.01 in 3key 0.52Ranked 17. The model replayed this schedule 1,000 times, and this team finished between 12 and 117 of 136 in 90% of them.
  18. 18Illinois5-25-2Gap +11.11 in 3key 0.50Ranked 18. The model replayed this schedule 1,000 times, and this team finished between 6 and 102 of 136 in 90% of them.
  19. 19Houston6-16-1Gap +14.31 in 3key 0.50Ranked 19. The model replayed this schedule 1,000 times, and this team finished between 12 and 118 of 136 in 90% of them.
  20. 20Tulane6-16-1Gap +15.71 in 3key 0.50Ranked 20. The model replayed this schedule 1,000 times, and this team finished between 14 and 120 of 136 in 90% of them.
  21. 21Arizona State5-25-2Gap +10.11 in 3key 0.49Ranked 21. The model replayed this schedule 1,000 times, and this team finished between 5 and 98 of 136 in 90% of them.
  22. 22James Madison6-16-1Gap +10.01 in 3key 0.49Ranked 22. The model replayed this schedule 1,000 times, and this team finished between 4 and 102 of 136 in 90% of them.
  23. 23North Texas6-16-1Gap +8.11 in 3key 0.44Ranked 23. The model replayed this schedule 1,000 times, and this team finished between 4 and 100 of 136 in 90% of them.
  24. 24Notre Dame5-25-2Gap +4.71 in 3key 0.44Ranked 24. The model replayed this schedule 1,000 times, and this team finished between 2 and 91 of 136 in 90% of them.
  25. 25UNLV6-16-1Gap +20.31 in 3key 0.44Ranked 25. The model replayed this schedule 1,000 times, and this team finished between 36 and 129 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 52 put Indiana on top. One in 3 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 96.5 places.

2025 week 8 · 7a38e08e · 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
1Indiana7-01 in 524.5+2.519.8 (4)+40.21 to 550.0
2Texas A&M7-01 in 304.5+2.513.4 (13)+46.62 to 850.0
3Ohio State7-01 in 224.7+2.320.3 (3)+39.71 to 580.0
4BYU7-01 in 214.7+2.312.8 (17)+47.23 to 910.0
5Georgia Tech7-01 in 184.8+2.29.7 (33)+50.38 to 97-1.0
6Alabama6-11 in 104.0+2.015.9 (7)+12.41 to 72+1.0
7Georgia6-11 in 64.3+1.711.3 (24)+14.14 to 103-1.0
8South Florida6-11 in 64.5+1.512.1 (18)+13.74 to 89-1.0
9Texas Tech6-11 in 64.5+1.521.2 (1)+5.01 to 56-5.0
10Oregon6-11 in 54.6+1.420.7 (2)+3.21 to 54+3.0
11Oklahoma6-11 in 44.6+1.413.1 (16)+10.22 to 87-1.0
12Navy6-01 in 44.8+1.28.1 (47)+51.98 to 107-14.0
13Louisville5-11 in 43.8+1.213.5 (12)+10.84 to 100+2.0
14Miami5-11 in 43.8+1.214.6 (9)+8.32 to 91+4.0
15Virginia6-11 in 44.9+1.19.9 (31)+12.610 to 101-4.0
16Vanderbilt6-11 in 44.9+1.118.5 (6)+4.22 to 66-5.0
17Ole Miss6-11 in 34.9+1.16.6 (51)+15.012 to 117+4.0
18Illinois5-21 in 33.9+1.18.3 (42)+11.16 to 102+3.0
19Houston6-11 in 34.9+1.16.8 (50)+14.312 to 118-1.0
20Tulane6-11 in 34.9+1.15.0 (62)+15.714 to 120+2.0
21Arizona State5-21 in 34.0+1.09.0 (36)+10.15 to 98-1.0
22James Madison6-11 in 35.0+1.011.5 (21)+10.04 to 102-6.0
23North Texas6-11 in 35.0+1.011.9 (19)+8.14 to 100+6.0
24Notre Dame5-21 in 34.1+0.913.2 (15)+4.72 to 91+8.0
25UNLV6-11 in 35.1+0.9-0.3 (95)+20.336 to 129-8.0
Show the full table: top 25 of 136, every columnthe published poll
#teamunderline: 90% rank interval · league median width 96.5how unlikely1 in this many
1IndianaBig Ten7-0·Power 19.79 (4)·Gap +40.21Ranked 1. The model replayed this schedule 1,000 times, and this team finished between 1 and 55 of 136 in 90% of them.1.7161 in 52
2Texas A&MSEC7-0·Power 13.40 (13)·Gap +46.60Ranked 2. The model replayed this schedule 1,000 times, and this team finished between 2 and 85 of 136 in 90% of them.1.4731 in 30
3Ohio StateBig Ten7-0·Power 20.31 (3)·Gap +39.69Ranked 3. The model replayed this schedule 1,000 times, and this team finished between 1 and 58 of 136 in 90% of them.1.3481 in 22
4BYUBig 127-0·Power 12.84 (17)·Gap +47.16Ranked 4. The model replayed this schedule 1,000 times, and this team finished between 3 and 91 of 136 in 90% of them.1.3311 in 21
5Georgia TechACC7-0·Power 9.71 (33)·Gap +50.29·▼1Ranked 5. The model replayed this schedule 1,000 times, and this team finished between 8 and 97 of 136 in 90% of them.1.2631 in 18
6AlabamaSEC6-1·Power 15.86 (7)·Gap +12.40·▲1Ranked 6. The model replayed this schedule 1,000 times, and this team finished between 1 and 72 of 136 in 90% of them.1.0071 in 10
7GeorgiaSEC6-1·Power 11.28 (24)·Gap +14.10·▼1Ranked 7. The model replayed this schedule 1,000 times, and this team finished between 4 and 103 of 136 in 90% of them.0.7841 in 6
8South FloridaAmerican Athletic6-1·Power 12.06 (18)·Gap +13.75·▼1Ranked 8. The model replayed this schedule 1,000 times, and this team finished between 4 and 89 of 136 in 90% of them.0.7601 in 6
9Texas TechBig 126-1·Power 21.16 (1)·Gap +5.04·▼5Ranked 9. The model replayed this schedule 1,000 times, and this team finished between 1 and 56 of 136 in 90% of them.0.7461 in 6
10OregonBig Ten6-1·Power 20.72 (2)·Gap +3.17·▲3Ranked 10. The model replayed this schedule 1,000 times, and this team finished between 1 and 54 of 136 in 90% of them.0.6641 in 5
11OklahomaSEC6-1·Power 13.13 (16)·Gap +10.23·▼1Ranked 11. The model replayed this schedule 1,000 times, and this team finished between 2 and 87 of 136 in 90% of them.0.6431 in 4
12NavyAmerican Athletic6-0·Power 8.14 (47)·Gap +51.86·▼14Ranked 12. The model replayed this schedule 1,000 times, and this team finished between 8 and 107 of 136 in 90% of them.0.6411 in 4
13LouisvilleACC5-1·Power 13.47 (12)·Gap +10.84·▲2Ranked 13. The model replayed this schedule 1,000 times, and this team finished between 4 and 100 of 136 in 90% of them.0.6211 in 4
14MiamiACC5-1·Power 14.65 (9)·Gap +8.28·▲4Ranked 14. The model replayed this schedule 1,000 times, and this team finished between 2 and 91 of 136 in 90% of them.0.5851 in 4
15VirginiaACC6-1·Power 9.92 (31)·Gap +12.62·▼4Ranked 15. The model replayed this schedule 1,000 times, and this team finished between 10 and 101 of 136 in 90% of them.0.5491 in 4
16VanderbiltSEC6-1·Power 18.47 (6)·Gap +4.16·▼5Ranked 16. The model replayed this schedule 1,000 times, and this team finished between 2 and 66 of 136 in 90% of them.0.5461 in 4
17Ole MissSEC6-1·Power 6.56 (51)·Gap +14.97·▲4Ranked 17. The model replayed this schedule 1,000 times, and this team finished between 12 and 117 of 136 in 90% of them.0.5221 in 3
18IllinoisBig Ten5-2·Power 8.35 (42)·Gap +11.10·▲3Ranked 18. The model replayed this schedule 1,000 times, and this team finished between 6 and 102 of 136 in 90% of them.0.5031 in 3
19HoustonBig 126-1·Power 6.76 (50)·Gap +14.28·▼1Ranked 19. The model replayed this schedule 1,000 times, and this team finished between 12 and 118 of 136 in 90% of them.0.4981 in 3
20TulaneAmerican Athletic6-1·Power 5.02 (62)·Gap +15.65·▲2Ranked 20. The model replayed this schedule 1,000 times, and this team finished between 14 and 120 of 136 in 90% of them.0.4961 in 3
21Arizona StateBig 125-2·Power 8.99 (36)·Gap +10.14·▼1Ranked 21. The model replayed this schedule 1,000 times, and this team finished between 5 and 98 of 136 in 90% of them.0.4901 in 3
22James MadisonSun Belt6-1·Power 11.47 (21)·Gap +10.03·▼6Ranked 22. The model replayed this schedule 1,000 times, and this team finished between 4 and 102 of 136 in 90% of them.0.4901 in 3
23North TexasAmerican Athletic6-1·Power 11.88 (19)·Gap +8.06·▲6Ranked 23. The model replayed this schedule 1,000 times, and this team finished between 4 and 100 of 136 in 90% of them.0.4431 in 3
24Notre DameFBS Independents5-2·Power 13.22 (15)·Gap +4.69·▲8Ranked 24. The model replayed this schedule 1,000 times, and this team finished between 2 and 91 of 136 in 90% of them.0.4391 in 3
25UNLVMountain West6-1·Power -0.27 (95)·Gap +20.26·▼8Ranked 25. The model replayed this schedule 1,000 times, and this team finished between 36 and 129 of 136 in 90% of them.0.4361 in 3

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

Full MeritMargin counts for exactly what it is worth. Nothing is protected.alternate lensJust WinWin the game. Running the score up buys you nothing here.alternate lens

same evidence, different valuesEvery board on this page was fitted on the same 952 games, and the digest of that exact frame is d2ed101a0fca 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 7a38e08e · published 2026-08-19 23:13:04 UTC · code e215160 · config 1ef7cf23…
q_ref 11.26 (Michigan) · β_w 7 · C 32 · h 5.132 · σ 16.473 · λ₁ 150 · λ₂ 0.5 · k 76.66 · w₁ 0.6243 · w₂ 0.0025
reproduce: uv run cfbpoll rank --season 2025 --through-week 8 --out out/
Close

rank 3

Ohio State

7-0 · Big Ten. The model never reads conference names.

1 in 22how hard that record waskey 1.35
4.7wins this schedule asked for
+2.3how far past it they came

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

Ranked 3. The model replayed this schedule 1,000 times, and this team finished between 1 and 58 of 136 in 90% of them.

1 to 58 · 57 places wide

power20.3, ranked 3
gap+39.7
resume60.0, ranked 5
hindsight3 · 0.0 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.