Sat, Sep 20, 7:30 PM
| 1 | 2 | 3 | 4 | T | |
|---|---|---|---|---|---|
| PUR | 7 | 16 | 0 | 7 | 30 |
| ND | 14 | 21 | 21 | 0 | 56 |
Our model blends Elo ratings (ND Elo 2141, PUR Elo 1310) plus home-field advantage. That projects ND -35.6 (100% to win) — 11.1 points of value on ND versus the market line of -24.5.
ND up 33 entering the 4th quarter. Across 457 historically comparable game states (within ±2 pts and ±3 min, from 3,056 games):
Purdue 30, Notre Dame 56.
Yes — the model's pick (ND) was correct.
Why trust the number?Gridpex's model is walk-forward validated on 2014–24, out-of-sample: +0.28 pts mean Closing Line Value and 53.5% beat the close. We report the record — wins and losses — never fake locks.
Called it — I had ND pregame. ✓ We log the misses too, but not this one.