Statistically Speaking- Utah at Washington

There’s been a lot of talk about the first college football playoff rankings, and most of it is wildly premature. There’s still a lot of football left to be played, and the number of ways these games could play out over the next four or so weeks is enormous. We don’t have a lot of data on the way the Committee thinks, and the data we do have is flawed because of turnover in membership, but we have enough to form some educated opinions about what matters to them.

What matters to the Committee

The first and most important factor to the Committee is record. Regardless of how well or poorly a team has played, their overall record dictates whether or not you have a shot at a playoff. Florida State did not have a particularly impressive resume and had not looked good against most of their opponents, but were allowed into the playoff against more deserving teams like Baylor and TCU because of that perfect record (in a power 5 conference). Teams with two losses were never really considered, and it does not appear they will be considered this year, barring a total unravelling of the status quo.

Second most important is an outright conference championship. This was a huge factor in the leapfrog of the Texas schools that happened at the end of the last season. By playing in a conference championship game, Ohio State demonstrated true conference mastery, something that TCU and Baylor couldn’t say they had done. Prior to the championship game, the Committee doesn’t appear to give any weight to this factor. They are perfectly happy to put two or three teams from one conference in the top 4, but last year would indicate that by season’s end, they’ll take those teams out in favor of ones which win a conference outright.

The third factor that goes into the decision making process is strength of schedule, particularly strength of out of conference schedule. Conferences with weaker schedules got something of a pass, but where the Committee really came after schools was if their nonconference schedule was exceptionally weak. Along with their conference championships, the stiffer out of conference slate faced by Ohio State and Florida State factored into their entrance into the playoff ahead of TCU and Baylor.

The Committee talks a lot about other factors, and they certainly discuss them at length in their meetings, but the way that the Committee made decisions in 2014 looked like a very simple rubric: "take the outright conference champion from each power 5 conference with the best record, and use their out of conference schedule as a tiebreaker if their records are the same."

Using these rubrics, there are 15 teams which could vie for a playoff spot. Memphis has had an incredible run, and if I were on the committee I’d want them in the running, but for the purposes of this analysis I’m taking them out of the running. It doesn’t help that they are only favored in two of their games going forward, and have only a 13.6% chance of winning out to their conference championship game, according to FPI.

What does this mean for the Utes?

To get into the playoffs, the Utes have to win their next five games. Given that they do that (far from a sure thing), out of the thousands of possible outcomes available to these fifteen teams, the number that will cause them to miss the playoffs is very small. There are a two main reasons for that certainty. The first is that every single one of these teams will play at least one other team ahead of them, if you include conference championship games.  There’s no way for more than one of LSU, Alabama, and Florida to win out, and there are several other similar clusters, mostly based on conference.

There’s the Ohio State, Michigan State, and Iowa cluster. Iowa’s got a soft path to the BIG 10 championship game, with a 31.1% chance of winning out, but then they’d have to take on one of the others. Only one of those three teams could conceivably make the playoff.

The SEC cluster is similar in structure. LSU and Alabama play this weekend, and Florida would play the winner of the West. FSU is also in this mix to an extent, although they could win out, win the ACC, and make the playoff while Alabama or LSU is doing the same. They aren’t properly a cluster member, though.

The biggest, baddest cluster is in the BIG 12. Oklahoma State, Oklahoma, TCU, and Baylor all play each other in the next four weeks. It’s hard to imagine a circumstance where all of these teams don’t come out with at least one conference loss, particularly considering that Oklahoma is favored against TCU by FPI. Even if one does somehow run the gauntlet, it’s certain that the others will fall behind the Utes.

The existence of these three clusters means that only three of these teams can stay ahead of the Utes. It’s hard to say which one, but from a playoff projection standpoint, the Utes are more properly described as seventh in line (it would be sixth, but if Oklahoma wins out they will jump the Utes).

So there are six teams ahead of the Utes- the SEC cluster winner, the BIG 10 cluster winner, and the BIG 12 cluster winner. There’s also Clemson, Notre Dame, and Stanford. Stanford isn’t a problem for the Utes because they’ll have to see the Cardinal lose if they win out, either by introducing a new PAC 12 championship game opponent or at the hands of Utah. For the Utes to make the playoff, they’ll need chaos to pop up in at least two areas ahead of them- a two loss cluster champion (or a one loss Iowa team) would work, a surprise loss for Clemson, or a loss for Notre Dame.

It’s easy to glance at a schedule and figure there’s no way for this to happen. Clemson is the heavy favorite in every game they play moving forward, as is Ohio State, Iowa, and others. However, heavy favorite is far from ‘sure thing’, as every college football fan knows. Despite being favored to win by at least 73% in every game they play, FPI still gives Clemson a 43.3% chance of losing at least one game, and that’s not even counting a championship game against a tough opponent. No one in the BIG 12 cluster has better than a 19% chance of winning out.

It’s one thing to say "Clemson is going to win out." That’s a smart bet and you will probably be right. It’s another thing entirely to say "Clemson, the winner of Bama/LSU, Ohio State, and the winner of Stanford/Notre Dame is going to win out." When you take that approach, you’re glossing over a whole lot of high-probability events, counting them as sure things. Something terrible is going to happen to someone, probably more than one someone. If the Utes win out, I’d put their odds at making the playoff at over 50%.

That’s a pretty big if, though. The Utes haven’t looked great lately and Washington is suddenly looking like one of the tougher matchups all season. Winning on the road in Huskies stadium is a tall order, but it’s going to really help the Utes’ chances of finishing 12-1.

Advanced Metrics Overview

Oregon State was about an ‘as expected’ performance for the Utes. The defense gave up a few more first downs than expected, which limited the ability of the offense to score points, but when they got the ball they were pretty darn efficient. I’ll still say the game against Oregon was Wilson’s best on the season, but his performance against Oregon State was close to flawless as well. The Huskies have a surprisingly stiff team from an advanced metrics perspective.  Washington is 19th in F/+ and 31st in FPI, compared to the Utes who are 18th and 20th, respectively. These numbers suggest an intense, close game.

We’ve got very similar profiles here, with excellent defenses supporting offenses that leave something to be desired. The Utes offense is a bit better and their defense a bit worse, but overall we’re looking at two defenses that shouldn’t struggle much against offenses that don’t have the firepower to overcome them.

Vegas and the Polls

This game opened as a pick-em, and the Huskies have picked up a point or two as the line has shifted to favor Washington. Still, a spread less than a field goal is about as close as it gets. One thing that’s surprising about the Vegas lines is the over/under. A 48.5 number there suggests that Vegas sees a lot more points being scored than advanced metrics would indicate. Massey hardly reacted to the Utes ho-hum win over Oregon State, shifting them back to 14th. The Huskies sit at 39th, dragged down by their inferior record.

Stat-Head Pick

Horrible weather and two excellent defenses mean a very low scoring game, my favorite kind to watch. The Huskies are going to load up to stop the run game and succeed at that, the Utes will keep the receivers covered and hurry the quarterback so he can’t throw the ball; they will succeed at that, too. The Utes are finally up against a defense as good at controlling the red zone as Utah is at scoring in it, which should limit their ability to generate touchdowns. The Huskies will hardly even get into the red zone. I see this game as a push, and ultimately I’ve got to go with the home team.

Utah 9, Washington 10