Week 6 is in the books, and here’s how our model’s seeing the playoff field. As a reminder: For its bracketology, our model lines teams up in order of their average final ranking. Memphis and Notre Dame have lower ceilings than Texas A&M. But Texas A&M has a worse average outcome, because their average season sees them finish something like 9–3.
The bracket:
Moving Out: Texas, Iowa State
Penn State was already out of our bracket (credit to Movelor, our model’s power-rating system, which doubted Penn State even after being forced to respect the normally–stronger preseason AP Poll), but Texas had snuck back in. No more Texas for now after the loss at Florida. The schedule is too hard; the team isn’t good enough; all of that can change.
Over in the Big 12, Iowa State is no longer the conference favorite after losing a winnable game at Cincinnati. Losing two starting cornerbacks for the season proved a smaller problem than failing to stop the run for the entire first half. Those things are probably linked, but more through gameplan than backup corner performance.
Moving In: LSU, Texas A&M
LSU gets bumped back up and in by falling Texas. The Tigers have some tough games remaining, but those are also good opportunities, and they get Texas A&M at home. Movelor currently has them likeliest to finish 9–3 but favored in every game besides their trip to Alabama.
Texas A&M beat Mississippi State comfortably, rising to 5–0. Our model again has them likeliest to only finish 9–3, but they should get to 7–0 before they have to visit LSU, Missouri, and Texas all over the final six weeks of the regular season. (Our model doesn’t make any claims about Marcel Reed’s durability, which evens out suspicions that Movelor’s slow to catch up on the Aggies thanks to those three defensive no-shows to open the season.)
Moving Up: Texas Tech
The Red Raiders have separated from the Big 12 pack, with Utah the closest team to them in Movelor and Utah also a team Tech beat in Salt Lake City. They’re up to a home game, and depending how close Curt Cignetti keeps things in the Indiana–Oregon game, Tech might jump the Hoosiers next week.
Moving Down: Georgia
The Dawgs slide out of a home game, not because they were underwhelming against Kentucky but because of Texas Tech’s rise. Movelor currently favors Georgia in all its remaining games, but the number’s only 2.5 against Mississippi and 3.5 against Texas. There’s a good chance they split those, and a better chance they lose a third than that they don’t.
Probabilities and Average Rankings
Nine teams are up above 50% in our playoff probabilities. There’s only a 3% chance all nine make the field. There is, however, an 83% chance at least six of the nine make the field.
Shoutout to Virginia, whose win at Louisville flips their situation. Movelor only views them as the fourth-best team in the ACC, but they have almost a 50% chance of making the ACC Championship, since NC State isn’t going to be a consequential head-to-head loss.
Oklahoma is not on the probabilities list, and that is not a mistake. Movelor hasn’t seen anything that indicates those guys are a top-ten team, and you have to be a top-ten team to make the playoff against a schedule like theirs.
Probabilities and Average Rankings, Continued
Pay attention to Notre Dame and Memphis in that average CFP ranking graphic. Notre Dame isn’t that likely to get a home game and Memphis isn’t that likely to finish as anything but the 12-seed. But their averages are better than those of LSU, Texas A&M, and—in Notre Dame’s case—Georgia.
Speaking of LSU and A&M: 20th is the average finish. It’s tough to win in the SEC. Movelor thinks Texas is the 7th-best team in the country and a decent favorite this weekend against OU. Our model still has the Longhorns finishing ranked 25th in the average simulation.
How Good Is Everybody?
Is Movelor sleeping on Miami? You could make a case for that, since Miami let both Notre Dame and Florida State back into those respective games when they should have been over. That also, though, is part of who Miami is: Mario Cristobal is a bad game manager (good head coach, to be clear). Miami habitually seems to struggle with focus. Clean those things up, and are they five points better? Yes.
Penn State is still on here and still above Texas A&M, and that might change but this is how good teams are right now. One horrific loss doesn’t tell the whole story. Notre Dame reached the national championship last year, and losing at home to NIU is more shocking than losing to interim-coach UCLA in California immediately after the Oregon loss, a game Penn State had turned into its national championship.
Florida’s into the top 25 again, at least as far as who the best teams are. They’re also the twelfth-best team in the SEC. The SEC has a lot of spoilers. Only nine SEC teams enjoy average final rankings among the 25 best in the country. Twelve SEC teams are among the 25 best teams.
Odds and Ends
This bracket rocks. Texas Tech vs. Texas A&M? Alabama vs. LSU? Notre Dame vs. Georgia? Ohio State vs. Notre Dame or Georgia in the Rose Bowl? Lock it in. This would be fun. That said…
There’s still a 73% chance at least one SEC team gets a bye. Alabama alone has a 23% bye probability despite a two-score loss to Florida State. Texas Tech is at 30%, so 23% isn’t great, but don’t expect the SEC to stay out of the top four the whole way through. Someone will come out on top over there.
Notre Dame is 92% playoff-likely—per our model—in instances in which they finish 10–2. They still have to get there, but enough will probably break their way if they do. Is our model high on them? Maybe. But a fluky aspect of the schedule is that when the committee builds its first rankings, Notre Dame will probably get credit for having played a top-ten Texas A&M. Texas A&M should have more losses by the time the season ends, but ranking stickiness will still favor the Irish. Our model does account for this, and it accounts for it more this year than it used to. But if Notre Dame wins out, the Irish have a very good playoff chance.
A 1-in-5 playoff chance like Texas’s is still pretty good. A 1-in-20 playoff chance like Penn State’s is not. Texas isn’t one of our first four teams out, but they’re not far behind that. BYU, Illinois, Utah, and Virginia are the first four. Then come Michigan, Missouri, Texas, and USC. Only then do we get to Tennessee, Oklahoma, TCU, and Iowa State. Penn State is not one of the first twelve teams out.
One Oklahoma note: Our model doesn’t have the Kelly Bryant Rule applied yet to the Sooners. What is that rule? It comes from the year Syracuse concussed Bryant and beat Clemson. Clemson saw the loss half-forgiven. Assuming John Mateer does not play against Texas, we’d expect Oklahoma to see that loss half-forgiven, but only if OU’s in the playoff mix. With seven games remaining and all of those coming against teams among Movelor’s 21 best, that probably isn’t going to happen. We’ll monitor, but we’ll probably wait for the first rankings and just pack it into FPA.
Our model does have two-team and three-team tiebreakers now programmed in, which should help accuracy. We’ll still monitor more complicated ties and adjust the model more manually when necessary to account for those. Hopefully the Big 12 plays nice. (The Big 12 is going to have a 10-way tie for second, isn’t it.)
Getting more relevant as the picture gets clearer: Last year, the committee departed from precedent in that it didn’t really punish teams for losing in their conference championships. This is complex. “You shouldn’t be punished for making your conference championship” sounds good at first listen, but there’s a lot to gain by winning these games. Should they be a risk-free event? Our model assumes a 25% chance they’re almost a complete free play (i.e., almost no punishment for losing), a 25% chance they’re treated like regular season games, and a 50% chance the committee treats them as something halfway between those two options. That chance is correlated within a simulation, meaning if Georgia Tech doesn’t get punished for losing to Miami in the ACC Championship in one simulation, Mississippi won’t be punished for losing to Alabama in that same simulation’s SEC Championship. This is especially relevant for Virginia and Georgia Tech, two fringe playoff contenders with decent chances to reach the ACC title game.
As always, our model’s numbers go up before this post. Look for the numbers on Sunday morning. We’ll get the rest of this to you as quickly as we can.
**
