Statistically Speaking: UCLA
Utah has won the south. That's what my math says. Not a 98% chance, or a 99% chance. Not basically a sure thing. The model I've designed runs through the season ten thousand times. In those iterations, Utah took ten thousand trips to the championship game. In all but thirty of them, Utah headed to that game as an eleven win team.
That's a crazy place to be. It strains my willingness to trust the numbers, and I'm sure it strains yours. Surely there is SOME chance that Utah falls apart in the next three weeks. My model has performed very well over the last five years- nearly two games better on average than FPI, a game and a half better than SP+. Still... a sure thing is a hard story to swallow.
I'm going to stand by it, and here's why. During Kyle Whittingham's coaching tenure Utah has been, on average, about six and a half points better than FPI has predicted. If we take the simplified step of simply adding 6.5 to Utah's FPI, Utah would be +25, the fifth best team in the country. That's fairly close to where they rank in ESPN's efficiency metric- 6th overall. Where FPI attempts to predict the future, Efficiency describes the past. This season, Utah has been that good. If Utah's FPI was +25, they'd be slotted between Georgia and LSU.
Now imagine it's not your Utes' schedule. Imagine you are taking a cursory glance at the SEC, trying to decide if Georgia is going to be 11-1. Imagine Georgia's next three opponents are Ole Miss, Southern Miss, and Wyoming. Imagine someone says 'I've done the math- Georgia has a 99.7% chance of being 11-1'. I don't know about you, but I wouldn't bat an eye over that number.
That's Utah the rest of the way. The Utes are playing playoff-caliber football and have been all year. They are fresh and healthy. Meanwhile the three teams remaining on their schedule have been playing some of their worst football in recent memory. UCLA has managed to build a nice narrative for themselves in the last few weeks- but they are not remotely in Utah's league.
"Any given Saturday" is a mantra in college football. UCLA certainly has loads of excellent athletes on their roster that could, in theory, make plays against Utah's defense and slow down Utah's offense. But in four years of modeling PAC 12 football, I've identified fifty or so 99.9% model-breaking conference games. The inferior team has never won one, not once. Utah has three 99.9% games on their schedule, and I see no reason the trend will break. Utah is LSU. UCLA is Ole Miss. This one will be over when the Utes get off the bus.
Utah 49, UCLA 27