The Poll.ai

2025 season · through week 2

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. 1California2-02-0Gap +61.81 in 9key 0.96Ranked 1. The model replayed this schedule 1,000 times, and this team finished between 56 and 134 of 136 in 90% of them.
  2. 2South Florida2-02-0Gap +65.31 in 5key 0.69Ranked 2. The model replayed this schedule 1,000 times, and this team finished between 12 and 127 of 136 in 90% of them.
  3. 3Illinois2-02-0Gap +64.01 in 5key 0.68Ranked 3. The model replayed this schedule 1,000 times, and this team finished between 17 and 133 of 136 in 90% of them.
  4. 4Temple2-02-0Gap +60.21 in 5key 0.67Ranked 4. The model replayed this schedule 1,000 times, and this team finished between 31 and 133 of 136 in 90% of them.
  5. 5UNLV3-03-0Gap +66.71 in 5key 0.67Ranked 5. The model replayed this schedule 1,000 times, and this team finished between 48 and 136 of 136 in 90% of them.
  6. 6Texas State2-02-0Gap +67.11 in 5key 0.66Ranked 6. The model replayed this schedule 1,000 times, and this team finished between 29 and 135 of 136 in 90% of them.
  7. 7Memphis2-02-0Gap +61.11 in 4key 0.63Ranked 7. The model replayed this schedule 1,000 times, and this team finished between 24 and 132 of 136 in 90% of them.
  8. 8Utah2-02-0Gap +65.81 in 4key 0.62Ranked 8. The model replayed this schedule 1,000 times, and this team finished between 22 and 127 of 136 in 90% of them.
  9. 9App State2-02-0Gap +56.11 in 4key 0.60Ranked 9. The model replayed this schedule 1,000 times, and this team finished between 34 and 135 of 136 in 90% of them.
  10. 10Houston2-02-0Gap +68.31 in 4key 0.60Ranked 10. The model replayed this schedule 1,000 times, and this team finished between 15 and 130 of 136 in 90% of them.
  11. 11Tulane2-02-0Gap +67.61 in 4key 0.57Ranked 11. The model replayed this schedule 1,000 times, and this team finished between 17 and 129 of 136 in 90% of them.
  12. 12Ole Miss2-02-0Gap +63.21 in 4key 0.56Ranked 12. The model replayed this schedule 1,000 times, and this team finished between 14 and 134 of 136 in 90% of them.
  13. 13Iowa State3-03-0Gap +64.61 in 3key 0.53Ranked 13. The model replayed this schedule 1,000 times, and this team finished between 24 and 133 of 136 in 90% of them.
  14. 14Vanderbilt2-02-0Gap +51.71 in 3key 0.51Ranked 14. The model replayed this schedule 1,000 times, and this team finished between 20 and 130 of 136 in 90% of them.
  15. 15Wyoming2-02-0Gap +63.11 in 3key 0.50Ranked 15. The model replayed this schedule 1,000 times, and this team finished between 23 and 132 of 136 in 90% of them.
  16. 16Mississippi State2-02-0Gap +64.61 in 3key 0.48Ranked 16. The model replayed this schedule 1,000 times, and this team finished between 17 and 129 of 136 in 90% of them.
  17. 17Auburn2-02-0Gap +65.91 in 3key 0.46Ranked 17. The model replayed this schedule 1,000 times, and this team finished between 11 and 132 of 136 in 90% of them.
  18. 18Georgia Tech2-02-0Gap +62.71 in 3key 0.45Ranked 18. The model replayed this schedule 1,000 times, and this team finished between 10 and 126 of 136 in 90% of them.
  19. 19LSU2-02-0Gap +74.21 in 3key 0.42Ranked 19. The model replayed this schedule 1,000 times, and this team finished between 26 and 129 of 136 in 90% of them.
  20. 20North Texas2-02-0Gap +54.41 in 2key 0.39Ranked 20. The model replayed this schedule 1,000 times, and this team finished between 15 and 129 of 136 in 90% of them.
  21. 21Fresno State2-12-1Gap +32.31 in 2key 0.35Ranked 21. The model replayed this schedule 1,000 times, and this team finished between 48 and 135 of 136 in 90% of them.
  22. 22South Carolina2-02-0Gap +62.01 in 2key 0.35Ranked 22. The model replayed this schedule 1,000 times, and this team finished between 36 and 135 of 136 in 90% of them.
  23. 23Florida State2-02-0Gap +52.11 in 2key 0.31Ranked 23. The model replayed this schedule 1,000 times, and this team finished between 15 and 128 of 136 in 90% of them.
  24. 24Nebraska2-02-0Gap +62.71 in 2key 0.30Ranked 24. The model replayed this schedule 1,000 times, and this team finished between 18 and 135 of 136 in 90% of them.
  25. 25TCU1-01-0Gap +60.41 in 2key 0.29Ranked 25. The model replayed this schedule 1,000 times, and this team finished between 12 and 122 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 9 put California 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 113.5 places.

2025 week 2 · e524ef50 · 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
1California2-01 in 90.8+1.2-1.8 (81)+61.856 to 134-39.0
2South Florida2-01 in 51.0+1.0-5.3 (116)+65.312 to 127-1.0
3Illinois2-01 in 51.0+1.0-4.0 (104)+64.017 to 133-11.0
4Temple2-01 in 51.1+0.9-0.2 (64)+60.231 to 133-67.0
5UNLV3-01 in 51.8+1.2-6.7 (122)+66.748 to 136-23.0
6Texas State2-01 in 51.1+0.9-7.1 (125)+67.129 to 135-9.0
7Memphis2-01 in 41.1+0.9-1.1 (76)+61.124 to 132-42.0
8Utah2-01 in 41.1+0.9-5.8 (118)+65.822 to 127-18.0
9App State2-01 in 41.0+1.03.9 (34)+56.134 to 135-49.0
10Houston2-01 in 41.1+0.9-8.3 (130)+68.315 to 130-10.0
11Tulane2-01 in 41.1+0.9-7.6 (127)+67.617 to 129-10.0
12Ole Miss2-01 in 41.1+0.9-3.2 (99)+63.214 to 134-4.0
13Iowa State3-01 in 32.1+0.9-4.6 (109)+64.624 to 133+12.0
14Vanderbilt2-01 in 31.2+0.88.3 (15)+51.720 to 130-20.0
15Wyoming2-01 in 31.2+0.8-3.1 (97)+63.123 to 132-30.0
16Mississippi State2-01 in 31.2+0.8-4.6 (110)+64.617 to 129+6.0
17Auburn2-01 in 31.3+0.7-5.9 (119)+65.911 to 132-5.0
18Georgia Tech2-01 in 31.3+0.7-2.7 (93)+62.710 to 126-9.0
19LSU2-01 in 31.3+0.7-14.2 (136)+74.226 to 129+13.0
20North Texas2-01 in 21.3+0.75.6 (29)+54.415 to 129+9.0
21Fresno State2-11 in 21.4+0.6-11.4 (133)+32.348 to 135-36.0
22South Carolina2-01 in 21.3+0.7-2.0 (82)+62.036 to 135-9.0
23Florida State2-01 in 21.4+0.67.9 (18)+52.115 to 128+16.0
24Nebraska2-01 in 21.4+0.6-2.7 (91)+62.718 to 135+5.0
25TCU1-01 in 20.5+0.5-0.4 (67)+60.412 to 122-19.0
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 113.5how unlikely1 in this many
1CaliforniaACC2-0·Power -1.81 (81)·Gap +61.81·▼39Ranked 1. The model replayed this schedule 1,000 times, and this team finished between 56 and 134 of 136 in 90% of them.0.9621 in 9
2South FloridaAmerican Athletic2-0·Power -5.34 (116)·Gap +65.34·▼1Ranked 2. The model replayed this schedule 1,000 times, and this team finished between 12 and 127 of 136 in 90% of them.0.6871 in 5
3IllinoisBig Ten2-0·Power -4.03 (104)·Gap +64.03·▼11Ranked 3. The model replayed this schedule 1,000 times, and this team finished between 17 and 133 of 136 in 90% of them.0.6771 in 5
4TempleAmerican Athletic2-0·Power -0.17 (64)·Gap +60.17·▼67Ranked 4. The model replayed this schedule 1,000 times, and this team finished between 31 and 133 of 136 in 90% of them.0.6731 in 5
5UNLVMountain West3-0·Power -6.65 (122)·Gap +66.65·▼23Ranked 5. The model replayed this schedule 1,000 times, and this team finished between 48 and 136 of 136 in 90% of them.0.6681 in 5
6Texas StateSun Belt2-0·Power -7.07 (125)·Gap +67.07·▼9Ranked 6. The model replayed this schedule 1,000 times, and this team finished between 29 and 135 of 136 in 90% of them.0.6651 in 5
7MemphisAmerican Athletic2-0·Power -1.08 (76)·Gap +61.08·▼42Ranked 7. The model replayed this schedule 1,000 times, and this team finished between 24 and 132 of 136 in 90% of them.0.6311 in 4
8UtahBig 122-0·Power -5.79 (118)·Gap +65.79·▼18Ranked 8. The model replayed this schedule 1,000 times, and this team finished between 22 and 127 of 136 in 90% of them.0.6211 in 4
9App StateSun Belt2-0·Power 3.89 (34)·Gap +56.11·▼49Ranked 9. The model replayed this schedule 1,000 times, and this team finished between 34 and 135 of 136 in 90% of them.0.6041 in 4
10HoustonBig 122-0·Power -8.29 (130)·Gap +68.29·▼10Ranked 10. The model replayed this schedule 1,000 times, and this team finished between 15 and 130 of 136 in 90% of them.0.5961 in 4
11TulaneAmerican Athletic2-0·Power -7.64 (127)·Gap +67.64·▼10Ranked 11. The model replayed this schedule 1,000 times, and this team finished between 17 and 129 of 136 in 90% of them.0.5721 in 4
12Ole MissSEC2-0·Power -3.23 (99)·Gap +63.23·▼4Ranked 12. The model replayed this schedule 1,000 times, and this team finished between 14 and 134 of 136 in 90% of them.0.5581 in 4
13Iowa StateBig 123-0·Power -4.55 (109)·Gap +64.55·▲12Ranked 13. The model replayed this schedule 1,000 times, and this team finished between 24 and 133 of 136 in 90% of them.0.5261 in 3
14VanderbiltSEC2-0·Power 8.26 (15)·Gap +51.74·▼20Ranked 14. The model replayed this schedule 1,000 times, and this team finished between 20 and 130 of 136 in 90% of them.0.5131 in 3
15WyomingMountain West2-0·Power -3.08 (97)·Gap +63.08·▼30Ranked 15. The model replayed this schedule 1,000 times, and this team finished between 23 and 132 of 136 in 90% of them.0.5041 in 3
16Mississippi StateSEC2-0·Power -4.59 (110)·Gap +64.59·▲6Ranked 16. The model replayed this schedule 1,000 times, and this team finished between 17 and 129 of 136 in 90% of them.0.4811 in 3
17AuburnSEC2-0·Power -5.94 (119)·Gap +65.94·▼5Ranked 17. The model replayed this schedule 1,000 times, and this team finished between 11 and 132 of 136 in 90% of them.0.4581 in 3
18Georgia TechACC2-0·Power -2.74 (93)·Gap +62.74·▼9Ranked 18. The model replayed this schedule 1,000 times, and this team finished between 10 and 126 of 136 in 90% of them.0.4481 in 3
19LSUSEC2-0·Power -14.17 (136)·Gap +74.17·▲13Ranked 19. The model replayed this schedule 1,000 times, and this team finished between 26 and 129 of 136 in 90% of them.0.4191 in 3
20North TexasAmerican Athletic2-0·Power 5.59 (29)·Gap +54.41·▲9Ranked 20. The model replayed this schedule 1,000 times, and this team finished between 15 and 129 of 136 in 90% of them.0.3891 in 2
21Fresno StateMountain West2-1·Power -11.41 (133)·Gap +32.35·▼36Ranked 21. The model replayed this schedule 1,000 times, and this team finished between 48 and 135 of 136 in 90% of them.0.3531 in 2
22South CarolinaSEC2-0·Power -1.98 (82)·Gap +61.98·▼9Ranked 22. The model replayed this schedule 1,000 times, and this team finished between 36 and 135 of 136 in 90% of them.0.3521 in 2
23Florida StateACC2-0·Power 7.94 (18)·Gap +52.06·▲16Ranked 23. The model replayed this schedule 1,000 times, and this team finished between 15 and 128 of 136 in 90% of them.0.3061 in 2
24NebraskaBig Ten2-0·Power -2.69 (91)·Gap +62.69·▲5Ranked 24. The model replayed this schedule 1,000 times, and this team finished between 18 and 135 of 136 in 90% of them.0.3001 in 2
25TCUBig 121-0·Power -0.40 (67)·Gap +60.40·▼19Ranked 25. The model replayed this schedule 1,000 times, and this team finished between 12 and 122 of 136 in 90% of them.0.2901 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 272 games, and the digest of that exact frame is 33d4b92fd2b6 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 e524ef50 · published 2026-08-19 23:10:51 UTC · code e215160 · config 1ef7cf23…
q_ref 6.30 (Troy) · β_w 7 · C 32 · h 12.433 · σ 23.062 · λ₁ 125 · λ₂ 2.0 · k 94.33 · w₁ 1.0348 · w₂ -1.8795
reproduce: uv run cfbpoll rank --season 2025 --through-week 2 --out out/
Close