When the 2026 ESPN Football Power Index (FPI) was released, Alabama football fans responded with a mix of anger and dismay, as the computer model projected an over-under of 8.6 wins and 3.6 losses.
ESPN's algorithm is understandably a mostly closely guarded secret. As a predictive model that requires in-season game results, it takes almost half of a season to gain credible accuracy.
On the other hand, even in preseason form, the model is not tainted by subjectivity. The same model provides win probabilities for individual games. In 10 of Alabama's regular-season games, the Crimson Tide has a higher win probability than its opponent.
In college football, the difference between 8.6 wins and 10 wins is huge. It can be the difference between making the College Football Playoff field and not.
ESPN Analytics Alabama Football Game Probabilities
- Alabama vs. East Carolina - Tuscaloosa - Sept. 5 - 94.6% Crimson Tide Win
- Alabama at Kentucky - Sept. 12 - 82.5% Win
- Alabama vs. Florida State - Tuscaloosa - Sept. 19 - 81.7% Win
- Alabama vs. South Carolina - Tuscaloosa - Sept. 26 - 78.8% Win
- Alabama at Mississippi State - Oct. 3 - 85.2%
- Alabama vs. Georgia - Tuscaloosa - Oct. 10 - 41.9% Win
- Alabama at Tennessee - Oct. 17 - 59.1% Win
- Alabama vs, Texas A&M - Tuscaloosa - Oct. 24 - 58.2% Win
- Alabama at LSU - Nov. 7 - 43.8% Win
- Alabama at Vanderbilt - Nov. 14 - 75.6% Win
- Alabama vs. Tennessee-Chattanooga - Tuscaloosa - Nov. 21 - 99% Win
- Alabama vs. Auburn - Tuscaloosa - Nov. 28 - 77.6% Win
For comparison, Kenneth Massey's computer model projects the Crimson Tide with a 49% probability of beating Georgia, a 68% probability of beating Texas A&M, and a 65% probability of beating LSU in Baton Rouge. If Massey's preseason calculation is correct, Alabama will finish the regular season at 11-1.
Add us as a preferred source on Google