ESPN's FPI predicts the remaining games on USC's schedule
USC arrives at its bye week with a 6-1 record, but probably a bit of disappointment after Saturday’s 43-42 loss to Utah.
The Trojans had a 42-35 lead with 6:15 to play after a 20-yard touchdown pass from Caleb Williams to Michael Jackson III, but couldn’t hold it. Utah drove 75 yards in 15 plays and 5:27 to score the game-winning touchdown, successfully going for a two-point conversion to take a one-point lead with just 48 seconds left. Penalties derailed USC’s final possession, which ended well shy of field goal range.
“We didn’t play as clean on all three sides as we wanted to,” coach Lincoln Riley said after the game. “It came down to they made one more play, or we made one more mistake. We were one inch away on several occasions and, honestly, several times very close to running away with it.”
But USC is still the favorite to win the Pac-12 championship, according to ESPN’s Football Power Index, and the Pac-12’s best shot at getting to the College Football Playoff.
FPI gives the Trojans a 31.4% chance to win the league and an 8.4% chance of making the CFP even after the loss to Utah. USC’s Playoff chances are the ninth-best of any team.
USC will be favored in every game the rest of the way, according to FPI.
Here’s how FPI is projecting the rest of the Trojans’ schedule, along with what the model was predicting back in the preseason:
- Oct. 29 at Arizona — 88.1% chance of winning (preseason: 77.4%)
- Nov. 5 vs. Cal — 91.8% chance of winning (preseason: 79.8%)
- Nov. 11 vs. Colorado — 98.0% chance of winning (preseason: 85.3%)
- Nov. 19 at UCLA — 61.3% chance of winning (preseason: 45.7%)
- Nov. 26 vs. Notre Dame — 75.1% chance of winning (preseason: 33.6%)
That Nov. 19 clash with UCLA at the Rose Bowl could very well decide a spot in the Pac-12 title game. FPI is slow to take on the Bruins. Even with a win over FPI’s No. 9 team (Utah) and an unbeaten record, the Bruins rank 26th in FPI.
FPI has the Trojans’ predicted record at 11-2. They rank 10th nationally in the model.