This is Week 5. Week 12 games are out →
Fri Oct 2, 6:00 PM CT · Delaware Stadium
Our call: Liberty by 8.1 ✓ right
Final: Liberty 30, Delaware 14
Closest: Elo (Liberty by 13.4, missed by 2.6). 4 of 4 models named the winner. Our call was right.
Records on this page are each team's record entering the game; the team pages and the scores page count this result.
Model comparison
3 of 3 models favored Liberty
| Model | Says | Pregame win chance | Missed by |
|---|---|---|---|
| Elo Closest Liberty 1725 vs Delaware 1328 | Liberty by 13.4 ✓ | Liberty 83% | 2.6 |
| SP+ Liberty +2.5 vs Delaware -10.7 | Liberty by 10.7 ✓ | Liberty 78% | 5.3 |
| Our model · experimental Liberty -6.7 vs Delaware -9.3 | Liberty by 0.1 ✓ | Liberty 50% | 15.9 |
| Record Delaware 2-2, Liberty 3-1 | Liberty (better record) ✓ | – | – |
| Market no market margin | n/a | – | – |
| Our call average of SP+, Elo and Our model | Liberty by 8.1 ✓ | – | 7.9 |
Each margin is from the same weekly ratings the team pages use: SP+ and Elo from their rating gap, the home team given 2.5 points (none at a neutral site); Our model is this site's experimental rating; Market is the margin the market expected before kickoff; ESPN FPI, SRS, Sagarin and Sonny Moore are rating gaps like SP+'s (each with its own home field), The Margin is that site's own model, and Record names the team with the better record (the home team on equal records). ✓ when a model named the winner, ✗ when it did not; Closest marks the one nearest the final margin (an exact tie goes to the first listed). How these work ›
How the models read it
All three models favored Liberty (the records too).
SP+ had Liberty by 10.7, Elo had Liberty by 13.4, and Our model had Liberty by 0.1. They were within 13.3 points of each other.
This page was first built after kickoff, so these are the week's projections as recorded then, from the same weekly ratings the season record scores.