2020 · After week 15
Projected AP Top 25 after week 15
Where we expect AP voters to put every team, from last week’s poll and how each team just played. The real poll is out: we named 24 of its 25 teams and missed their spots by 1 place on average, with 13 exactly right.
This week is worked out from the results with the same model as our Saturday-night projections, so each team gets a one-line summary instead of a write-up.
- Won 52-3 at Arkansas as a 30-point favorite · Was No. 1
- Off this week · Was No. 2
- Off this week · Was No. 3
- Off this week · Was No. 4
- Off this week · Was No. 5
- Off this week · Was No. 7
- Off this week · Was No. 8
- Off this week · Was No. 10
- Won 42-38 at Troy as a 12.5-point favorite · Was No. 11
- Won 49-14 at Missouri as a 14-point favorite · Was No. 12
- Off this week · Was No. 13
- Beat San Diego State 28-14 at home as a 16.5-point favorite · Was No. 14
- Beat Illinois 28-10 at home as a 13.5-point favorite · Was No. 15
- Won 62-26 at No. 9 Miami as a 3-point underdog · Was No. 20
- Won 43-38 at UCLA as a 3.5-point favorite · Was No. 16
- Lost 37-34 at home to LSU as a 23-point favorite · Was No. 6
- Off this week · Was No. 17
- Beat No. 25 Wisconsin 28-7 at home as a 2-point underdog · Was No. 19
- Off this week · Was No. 18
- Lost 62-26 at home to No. 20 North Carolina as a 3-point favorite · Was No. 9
- Beat Akron 56-7 at home as a 33-point favorite · Was No. 24
- Off this week · Was No. 22
- Beat Nevada 30-20 at home as a 2.5-point underdog · Was unranked
- Off this week · Was No. 23
- Won 17-9 at Wyoming as a 9.5-point favorite · Was unranked
Dropping out
Colorado (was No. 21): Lost 38-21 at home to Utah as a 2.5-point underdog.
Wisconsin (was No. 25): Lost 28-7 at No. 19 Iowa as a 2-point favorite.
The projection uses the same vote model as our season simulator, fitted to every AP poll from 2014 to 2025: wins over ranked teams and lopsided wins lift a team, losses cost more the weaker the winner and the bigger the margin. Write-ups are generated from the game data and checked against it before they publish.