We’re going to do two things here: We’ll show you what our model projects and we’ll tell you what I think is going to happen. Our model is imperfect. The twelve-team playoff is funkier than the four-team playoff. In the four-team era, conference championship games counted. In the twelve-team format, they seem to count a limited, undetermined, and potentially inconsistent amount. Add in a head-to-head situation and some really flawed résumés, and our model would like to call timeout.
As a reminder:
- Alabama lost to Florida State by two touchdowns.
- Miami lost to two unranked teams.
- Notre Dame lost a home game to Texas A&M. They’d currently be favored over the Aggies by ten.
Whoever misses the playoff will deserve to miss the playoff. The argument is over who deserves to miss it more.
One other note:
Credit to Kalen DeBoer for going for it on fourth down in the shadow of his own goalposts. There’s a trend right now where coaches would rather lose games close than risk worsening a beatdown by trying to win. Alabama played terribly yesterday, but they played to win. They didn’t give up a chance at winning just to try to backdoor a 10-point final margin. They let it be unpalatable. I don’t know that they should be in the playoff, but I appreciated that.
Now.
Man vs. Model
What our model thinks, and where I think our model’s wrong:
Explanations…
The Top Four
Indiana’s the 1-seed. That’s a given. Texas Tech’s almost definitely the 4-seed, because even if the committee did want to put Ohio State behind the Red Raiders, they might shy away to avoid an Ohio State vs. Indiana rematch in the semifinals.
As for Georgia and Ohio State…
Who knows! This isn’t an area where the committee’s going to spend a lot of time. They’re going to spend a few minutes confirming JMU over Duke, a few minutes confirming seedings, and the rest of their time debating Notre Dame, Alabama, and Miami.
It’s important who gets which seed. The 3-seed will either play Mississippi or Tulane, teams who currently, combined, have one half of their head coach. The 2-seed will either play Texas A&M, Miami, Notre Dame, or Alabama. It is preferable to be the 3-seed.
To the committee, though, this is small stuff. Flip a coin. Our model thinks Ohio State gets the 2-seed. I’d personally guess Georgia.
The Non-Bubble
Our model has Mississippi jumping Oregon, which obviously won’t happen. That’s by fractions of a point in our model’s Ranking Score, and it’s happening because Georgia won so big yesterday. Georgia won by so many points that it helped Mississippi’s résumé more than Indiana beating Ohio State helped Oregon’s.
Everything else is straightforward. JMU vs. Duke isn’t a serious debate. Duke is 8–5. Heading into last night, they were next to Wisconsin in ESPN’s Strength of Record. Our model says that if the committee ranked a top 36, Duke would be ranked 35th or 36th. Duke lost to Illinois, Tulane, Georgia Tech, UConn, and Virginia. All of those losses are comparable to JMU’s single loss at Louisville, and Duke’s wins—the next-best behind the Virginia win is that they beat NC State—aren’t anything to write home about. JMU is in as the 12-seed unless the committee feels too guilty about leaving the ACC out entirely.
The Bubble
Ok, here’s the big one.
Our model thinks there’s a 75% chance Alabama’s loss to Georgia counts. In 25% of our model’s simulations, a conference championship loss is a full, real loss. In 25% of our model’s simulations, it only affects a team’s résumé in a tiny way. In 50% of our model’s simulations, the loss lands halfway between those poles.
Based on that, our model says Miami is the 10-seed. In the average simulation, Miami ends up ahead of Alabama.
As for me!
It would seem to me that if the committee’s willing to drop BYU past Miami because they got blown out in the Big 12 Championship, they’ll be willing to drop Alabama past Notre Dame. This especially seems likely because on Tuesday, Hunter Yurachek told us that the debate between Alabama and Notre Dame has been very, very close. Based on that, I don’t see how you get Alabama as the 9-seed unless Notre Dame’s the 10-seed and BYU’s ranked 11th. Basically, I think we’re going to get one of these two lists:
9. Alabama
10. Notre Dame
11. BYU
9. Notre Dame
10. Alabama or Miami
11. Miami or Alabama
If there’s a debate, then, it’s Miami vs. Alabama. That’s who this is probably coming down to. Who wins that debate?
If the committee’s trying to select the best team right now, that team is Miami. At least in our model’s eyes, Miami would be favored over the Crimson Tide if they played next Saturday on a neutral field.
If the committee’s trying to select the best team over the course of the season, I’m unsure who it is. August Miami was a lot better than November Alabama, but October Alabama was better than October Miami.
If the committee’s trying to select the most accomplished team with bonus points for dominance (I think this is the best way to describe the effective criteria), it’s close, but it’s probably Alabama. Let’s break each team’s results into the extreme end and the middle.
On the extreme end, Alabama has the slight edge. Alabama beat Georgia, which is better than Miami’s win over Notre Dame. Alabama lost to Florida State, which is worse than Miami’s loss to Louisville. Alabama lost to Oklahoma, which is better than Miami’s loss to SMU. Alabama has an extra loss, and by a lot, but they also have a win over Vanderbilt, and Miami didn’t earn a thirteenth game.
In the middle, Miami was more dominant than Alabama. Miami had fewer close calls. But Miami did this against a weaker schedule than what Alabama played. If you line up each team’s regular season opponents from toughest to weakest and compare them side by side—toughest against toughest, second-toughest against second-toughest, etc.—Alabama has the tougher opponent in eleven of the twelve slots. (Louisiana Monroe is worse than Syracuse.)
I’m of the mindset that conference championships should count as a full game (I explained that more on Friday, but basically—if you have something to gain, you should have something to lose). Even I’m not sure Miami has a better résumé than Alabama, taking Alabama’s 21-point loss yesterday as a full, real life, 21-point loss. Miami beat Notre Dame, then didn’t do anything else positive and noteworthy in September, October, November, or—so far—December. Florida State wasn’t good. USF is worse than twelve SEC teams. Pitt was an easy team to dominate.
Add in that the committee is hesitant to punish teams for losing conference championships (or at least, they were last year), and I think it’s going to be Alabama. The strongest case I can see for why to expect the committee to choose Miami is that the committee doesn’t want to leave an entire power conference out of this twelve-team playoff. I’m not 100% confident, but I think it’s going to be Alabama.
That should give us the following bracket:
I’m not sure which quarterfinal would get the Orange Bowl and which would get the Cotton Bowl. Columbus is a little bit closer to Arlington than to Miami Gardens, but there’s some Big Ten history with the Orange Bowl, and I’d imagine the committee wants Texas Tech playing in Texas if possible. Besides that? I think this is it. Not 100% confident, but I’d estimate the probability of Notre Dame making it at 90–95%, and I’d estimate the probability of Alabama making it at 80–90%.
**
