We’re halfway through college football’s regular season. Chalk, for the most part, has prevailed so far, but as the field trims itself, the biggest matchups on the horizon grow more and more consequential.
Here’s what happened this week, with insights from our model:
Georgia Loses
South Carolina isn’t a bad team. Entering last week, the aggregate ratings our model uses had the Gamecocks listed as the 33rd-best in the country. Still, those ratings had the traditionally lesser USC as a 20-point underdog in Athens, and they beat Georgia, formerly the fifth-most likely team to make the playoff. It was surprising. It was a bad look for Georgia. And after not looking particularly good at times against Tennessee and even Notre Dame, it felt like confirmation, in a way, that this year’s UGA is not quite what it was the past two seasons.
But what does it really mean?
Fortunately for Georgia, its path to the SEC title game is still clear: Beat Florida, let the rest of the year transpire as expected. The Dawgs are our model’s SEC East favorite. But they’re down to 16.6% likely to win the SEC after entering the week 27.9% likely. That’s a big difference, stemming not only from their smaller margin for error but from the realization that, yes, Georgia isn’t as good as the ratings thought. To put a number on it, they’re measured as 2.7 points worse compared to top-rated Alabama than they were prior to Saturday.
The bigger danger for Georgia is one that should cause some unease in the SEC as a whole. South Carolina, though a solid team, does not project to have a solid résumé. They’ve already lost to UNC. Their other losses—Alabama and Missouri—are both to be expected from even a top-15 team, but those are two more tallies against them. With Florida, Appalachian State, and Clemson all yet to come to Columbia, and visits to Tennessee and Texas A&M still on the schedule, the only game Will Muschamp should feel comfortable about is Vanderbilt’s visit at the beginning of November. The Cocks are favored in three of the six games, but three losses would leave them a .500 team, making the home loss a dark mark on Georgia’s team sheet no matter how well South Carolina plays the rest of the way, especially with no clear injury to which the committee can point in order to justify forgiveness.
This is dangerous for the SEC because Georgia is still a very talented team, and as was mentioned, remains the most likely SEC East team to represent the division in the league’s championship. A 12-1 Georgia that won the SEC would be expected to make the field, but what if they lose at Auburn, still win the SEC, and leave the league with a two-loss champion? And what if Oklahoma and Clemson both win out while Ohio State and Wisconsin split their contests? It’s a narrow group of hypotheticals, but they’re impactful: the probability of the SEC being completely left out of the playoff jumped from 2.9% to 7.6% after Saturday’s events.
Separation at the Top
All season, we’ve been breaking down the contenders into various tiers. Now, we’re at a point where they don’t need to be treated with much nuance. There are two top tiers, then a collection of teams that could pull off a playoff bid but face long odds, then everyone who’s already eliminated. The top two tiers, for now, are the ones that matter. All seven teams in these situations control their fate as far as making the playoff goes. Three of them—the top tier—are expected to control that fate successfully. Four of them—the second tier—are each more likely than not to slip up.
In the top tier, we have Ohio State (84.8% likely to make the field), Alabama (76.2%), and Clemson (67.4%). In the second, we have Wisconsin (43.0%), Oklahoma (38.9%), LSU (33.1%), and Penn State (29.8%). Wisconsin, LSU, and Penn State all have one or more games remaining in which they currently project as underdogs. Oklahoma is favored in each of their remaining contests but has fewer layups and a slimmer margin for error, thanks to the Big 12’s more centralized distribution of talent.
Wisconsin vs. Oklahoma—Who’s the Current Fourth Team?
To be clear, our probabilities shouldn’t be used to gauge how good teams are, or how good their résumés are so far. They’re a measurement of how likely each team is to make the playoff (and win the conference and win the national championship)—nothing more.
Still, since four teams make the playoff, it’s worth looking at which teams are, at any given point, the four most likely to be selected. In other words, our model’s placement of Wisconsin above Oklahoma is worth examining.
We covered some of this a few paragraphs back: At the easier end of each team’s remaining schedule, Oklahoma’s opponents are more difficult than those of Wisconsin. West Virginia is better than Illinois, even when adjusting for home field advantage. TCU is much better than Purdue. Kansas State is better than Nebraska. This hurts Oklahoma’s chances because it’s more likely they’ll suffer what could be termed a “damaging loss” compared to the Badgers.
At the top end of the schedule, the opposite is true: Wisconsin has it tougher. They play Ohio State on the road in two weeks. If all goes well and chaos doesn’t erupt in the Big Ten East, they’ll have to either play the Buckeyes again in the Big Ten Championship or deal with Penn State. For Oklahoma, the toughest remaining contests figure to be against teams outside the fifteen best: Iowa State at home, and most likely Texas in Arlington for the Big 12 Championship.
But instead of helping Oklahoma, this hurts their cause. Because while the committee does value holistic strength of schedule, they also focus on teams’ best wins and best losses. And a 12-1 Oklahoma with a conference championship doesn’t necessarily beat out a 12-1 Wisconsin that split with Ohio State (or, for that matter, a 12-1 Ohio State that split with Wisconsin, or possibly even an 11-1 Penn State whose only loss came in Columbus). Yes, Oklahoma might win out in those scenarios, but it’s far from assured. Then there’s the unique risk they have of seeing their “game against a common opponent” evaluated in a comparison with a one-loss LSU. In scenarios in which Oklahoma wins the Big 12 at 12-1, they stack up more poorly against an 11-1 LSU than a 12-1 Big Ten Champion Wisconsin does, because there’s the direct result the committee can point to in which LSU beat Texas by seven points on the road, but Oklahoma did so at a neutral location. It’s splitting hairs, but in that sort of scenario, hairs will have to be split.
Our model’s only working with five years of data, and it’s yet to see any rankings from this year’s committee, which will impact things. But for right now, it seems reasonable that Oklahoma and Wisconsin are neck and neck, and reasonable that Wisconsin’s ahead.
The Pac-12 Passes Notre Dame
For the first time in well over a month, the Pac-12 cumulatively has a better chance of making the playoff than Notre Dame does.
Notre Dame, of course, just had a potentially stellar loss swept down a few notches with South Carolina’s upset of Georgia. Similarly, their hope of Virginia fooling everyone into thinking that eleven wins by an ACC team meant something has evaporated with Virginia’s loss to Miami. And they’re in that double-edged boat in which a victory over Michigan in two weeks would likely result in Michigan only finding itself on the fringes of the top 25 at season’s end, and thereby not a résumé cornerstone, while a loss would all but eliminate the Irish from playoff consideration, if not destroy those chances entirely.
For the Pac-12, though, hope has arisen in the forms of Oregon, Utah, and in an odd way, Arizona State.
Oregon’s been the Pac-12’s best hope ever since Washington lost to Cal in the season’s second week. The Ducks are a good team (rated the ninth-best, as you’ll see a bit further down). They’ve got what should be a respectable, if not actively good, loss to Auburn—narrow and neutrally located. They’re favored in each of their remaining games (though only by about half a point at Washington this coming Saturday). And while it’s doubtful the committee would choose them over Clemson should both finish 12-1 with a conference title, the ascent of Arizona State means that between the Sun Devils, Washington, USC, and possibly Utah, a one-loss Oregon should be able to find at least one and possibly four best wins better than Clemson’s best, which will likely go down as the one over Texas A&M. The committee judges quality of team—not just résumé. But Oregon’s within striking distance for both and is in the best position of anyone outside the top seven to take advantage of any chaos that unfolds above them.
Utah’s basically Oregon Lite. They, too, have yet to deal with the iceberg that is a two-bad-loss Washington in Seattle. They, too, have a loss that could turn out to be ok when all’s said and done (on the road against USC). They, too, have Arizona State to potentially help pad their résumé. They aren’t as good as Oregon. Their résumé isn’t as good as Oregon’s. But they’re a worthy sidekick, enough to place Notre Dame firmly in the Pac-12’s rearview, at least for this week.
Who’s the Best:
In the interests of interest, here’s how the ratings our model uses break down the top teams in college football, with the number in parentheses indicating by how many points each would be expected to lose to Alabama on a neutral field. As usual, we cut the field off at North Dakota State, partially to highlight that the Bison would be the second-best team in this year’s ACC, and partially because we just find it interesting.
1. Alabama (0.0)
2. Ohio State (1.3)
3. Clemson (4.8)
4. Oklahoma (6.3)
5. LSU (6.4)
6. Wisconsin (8.0)
7. Georgia (9.0)
8. Penn State (9.0)
9. Oregon (11.2)
10. Auburn (11.3)
11. Florida (13.2)
12. Notre Dame (13.8)
13. Washington (14.6)
14. Utah (14.7)
15. Missouri (15.5)
16. Michigan (15.8)
17. Texas (15.9)
18. Iowa State (16.4)
19. UCF (16.4)
20. Iowa (17.5)
21. Texas A&M (18.0)
22. Michigan State (19.2)
23. Oklahoma State (19.5)
24. Baylor (19.6)
25. USC (19.6)
26. Minnesota (20.6)
27. South Carolina (20.9)
28. Cincinnati (21.5)
29. TCU (21.7)
30. Memphis (21.8)
31. Boise State (22.2)
32. Arizona State (22.7)
33. Washington State (23.1)
34. North Dakota State (23.1)
