football Edit

Statistically Speaking: Pac-12 Championship Game

Oregon gets a second crack at the Utes this week with the Rose Bowl on the line, and one thing is for sure: the best thing you can do for your sanity is forget the last game every happened.

Utah absolutely walloped the hapless Ducks just last month. Tavion Thomas put together a backbreaking three touchdown romp and the team as a whole racked up 386 yards on just 68 plays. Meanwhile the defense would hold the Ducks under 300 yards for the first and so-far-only time this season and allow just 7 points to a team that's scored at least 3 touchdowns in every other game this year.

There was nothing weird or wacky about it either. Utah just lined up and dominated, punishing the Ducks between the tackles and stifling a usually lethal running game. Apart from a nail-in-the-coffin punt return to close out the second half, Utah got all their points and yards with old-fashioned, smash mouth football.

Now forget all of it, because it doesn't matter. The results of rematches have historically been defined by their refusal to commit to a simple storyline.

I counted 23 rematches during the FPI era between Power 5 football teams, 46 games total. Scoring margins in these games ran the gamut from contests decided by field goals to one 67 point shellacking doled out by one of the greatest football teams in history. The one constant: the first football game told you nothing about the second one.

First, the basic numbers: the scoring margin swings by an average of 5.8 points, 56% of the rematches moved by more than two touchdowns, and 35% of the rematches moved by more than the 31 points Utah won by. Home team, away team, long gap or short gap between games, none of it mattered. The statistical upshot is that the first game gives you almost no information about what to expect in the second one.

So what do you draw the data from? As always, you look to the entire body of work of both teams across the whole season. When we compare rematches to what we'd expect from an FPI score projection, we find a steady, predictable distribution. 65% of teams averaged within 10 points of their FPI score projection. As tempting as it is to predict a repeat performance by both teams, college football is inherently chaotic with a huge random element. We learned a lot about Utah and Oregon a couple weeks ago, but if we want to set our expectations reasonably, we'll consider what we've learned about them through the course of the year.

Utah is the better team, but not by much, certainly not by 4 touchdowns or more. What we saw from Utah's offense two weeks ago likely won't change- Oregon is going to struggle to stop Tavion Thomas and Cam Rising is going to feast on a sparsely defended backfield. What was shocking about the first game was how poorly Oregon's offense performed against a Utah defense that has been merely serviceable this season. I project Oregon will clean up the miscues and put Utah into a higher scoring game, but in the end, Cam Rising and the Utah offense will deliver Utah it’s first Pac-12 title.

Utah 42, Oregon 36