We aren’t really a quarter of the way through the season.
Not everyone’s played three games.
And more teams play thirteen or more games than twelve, anyway.
But we’re about a quarter of the way through, and in ways, the state of the playoff race looks the way a single game would through one quarter of play. No one all that good is out of it, though they may be limping. We can stop paying attention to a lot of teams. The dominators are dominating. Leverage is rising.
With the help of our playoff/championship model, here are the biggest changes following Week 3’s games:
Penn State Pulled a Michigan, Meaning…
Our model had been higher than the public on Penn State. After Saturday’s performance, that gap has shrunk. If you missed it, Penn State nearly lost at home to Pitt, who, for context, is rated the 49th-best FBS team by the aggregate ratings we use even after they took the Nittany Lions to the wire. Penn State won, so no real harm is done to their résumé, but their path forward looks worse, as the computers now see them as roughly two points worse than they did a few days ago. This shift was enough to push the Nittany Lions down from 17.4% likely to make the playoff to 8.2% likely. Two points is a lot.
Just as we saw with Michigan last week, losses aren’t the only thing that can hurt our perception of a team’s chances. Our measure of a team’s ability is recalibrated each week, and with so little margin for error in the playoff race, those recalibrations make an impact. It doesn’t mean the team’s any worse than it was. It just means they’re probably worse than we thought they were.
The Gates Opened Wider for Ohio State
The Buckeyes took care of business in Bloomington, but more importantly for their hopes, three of their division-mates looked bad.
We talked about Penn State, but Maryland and Michigan State’s impotent losses to Temple and Arizona State (respectively) also took their toll on the Big Ten’s likely also-ran’s. The result was Ohio State’s Big Ten Championship chances rising from 38.5% to 50.7%, and their playoff hopes rising from 41.4% to 58.5%. It also didn’t hurt that Iowa’s narrow victory over Iowa State was sloppy enough to push the Hawkeyes down two and a half points in the aggregate ratings.
With this the current outlook, Ohio State is squarely in its own tier of playoff contention behind Clemson and Alabama (though Georgia is right there with the Buckeyes in championship probability).
Georgia Climbed Back Up, But…
As I wrote last week, Georgia’s drop following Week 2 was odd, and seemed to be the product of a strange confluence of events. The Dawgs erased those doubts this week, though, bouncing back up to 35.1% likely to make the playoff after throttling Arkansas State.
LSU Suffered Because of It
LSU fell to 25.0% likely to make the playoff, down from 35.4% entering the week. They effectively traded spots with Georgia, through little fault of their own. They weren’t strikingly impressive against Northwestern State, so their aggregate rating dropped by less than a third of a point following the performance. Georgia’s, in turn, rose by over three and a half points.
A Good Weekend for the Big 12, But…
The three biggest movers in the aggregate ratings, in a positive sense, were Kansas, West Virginia, and TCU, each of whom got a nice win over a mediocre Power Five foe. With the Oklahoma’s also both climbing, half the Big 12 was among the 14 teams whose aggregate rating jumped more than two points. And that’s without getting into Kansas State’s victory over Mississippi State, which pushed the Wildcats up by a point and a third.
It Didn’t Mean Much
Oklahoma did climb after all this, from 27.4% likely to make the playoff to 33.1%. But behind the Sooners, it’s still just Texas at 1.1% likely to make the playoff and a handful of teams at and below 0.5%. Oklahoma’s Big 12 title chances sit at 65.5%. When a Power Five title isn’t enough to get a team in the playoff half the time, that Power Five conference shouldn’t feel too good about its place in the world.
UCF Took Care of Business—Their Chances Dropped
After declaring total victory over Stanford at home, UCF rolls into this week’s playoff probabilities down at 4.0% from 5.1%. This is right on the edge of statistical significance given our model’s sample (a priority for us in the near future is increasing the sample each week from 1,000 simulations to something closer to 10,000, or ideally a million, but we’re not there yet). But it’s likely a real phenomenon, and it likely has something to do with Ohio State and Georgia’s paths getting easier.
It’s hard to imagine UCF getting in over any one-loss Power Five champion. It’s hard to imagine them getting in over any one-loss SEC or Big Ten team. They need chaos, and they need the SEC and Big Ten specifically to cannibalize themselves.
I’ll run more numbers on their specific favorable scenarios later in the week, but the overall message is clear: even a solid win over a solid team didn’t do much for the Knights.
Notre Dame Took Care of Business—Their Chances Rose
New Mexico is, by the ratings, worse than eleven FCS teams. Still, Notre Dame played a good game Saturday, and found themselves rewarded by the ratings, which bumped them up two points. The result of this, even with Georgia’s perception correcting itself, is that the Irish were one of the biggest positive movers this week, jumping past Penn State as they went from 14.1% likely to 19.5% likely to make the playoff.
On Friday, we’ll take a look at the week’s biggest games, so we’ll have plenty on ND at UGA (as well as Michigan at Wisconsin, Auburn at Texas A&M, and maybe a Pac-12 game). But for the time being, the Irish are in the best spot anyone could ask them to be in as they enter Athens a double-digit-ish underdog.
The Pac-12’s Doing Its Thing
USC falling at BYU popped one of the Pac-12’s balloons, and while it was a small one, it’s one more team that’s almost certain not to make the playoff. The ratings were relatively impressed by Oregon (which climbed by two thirds of a point) and relatively unimpressed by Utah (which dropped by nearly a full point) after each played FCS foes of varying quality, resulting in the Ducks now being the Pac-12’s best hope to make the playoff, slotting in between Michigan and Florida at 3.2% likely to crack the field (yes, that’s behind UCF).
Tiers
Last week, I offered some tiers as a way to think of the national championship race. I’ll do it again this week, but instead of looking at the championship, I’ll look at the playoff chances (with the playoff so small, getting there is what you’re more likely focused on at this point). These tiers aren’t which teams are the best teams (I’ll lay those out below this). They’re which teams are the most likely to wind up one of the chosen four.
Tier 1: Clemson (90.7% playoff probability) – The ACC has one team among our 25 best. It’s Clemson.
Tier 2: Alabama (78.2%) – Alabama is probably narrowly the best team in the country. Their schedule also has enough high-end competition that they can afford a loss. Maybe even two.
Tier 3: Ohio State (58.5%) – The Big Ten East looked treacherous two weeks ago. It does not appear that way anymore.
Tier 4: Georgia (35.1%), Oklahoma (33.1%) – Georgia’s championship chances are right with Ohio State’s. The challenge will be getting there, but eleven or twelve wins might be enough to crack the field. Oklahoma might end up needing thirteen to do the same.
Tier 5: LSU (25.0%) – The win over Texas helps. Playing in the same division as Alabama (and Auburn) hurts.
Tier 6: Wisconsin (21.2%), Notre Dame (19.5%) – Two good teams. But probably not good enough.
Tier 7: Penn State (8.2%) – Maybe Pitt really is different when they’re playing top teams.
Tier 8: Auburn (4.7%), UCF (4.0%), Michigan (3.5%), Oregon (3.2%), Florida (3.1%) – Two good teams with tough schedules. One good team with an easy schedule. The suspicion-deserving Wolverines. The wounded Ducks.
Tier 9: Utah (1.8%), Iowa (1.7%), Texas A&M (1.7%) – Winning out would probably do the trick. But then again, Utah, the Pac-12’s not great.
Tier 10: Texas (1.1%), Washington State (1.0%), Virginia (0.9%) – Varying levels of fun. Similar levels of thin hopes.
Tier 11: We’re stopping with ten.
“Top 25”
For some context, here’s how the aggregate ratings we’re using would slot together a top 25, based on how good teams actually are with absolutely no consideration of résumé (we don’t hate the Pac-12—we’re just being realistic here):
1. Alabama (32.9 points better than median FBS team)
2. Clemson (32.2)
3. Georgia (26.6)
4. LSU (24.4)
5. Ohio State (24.1)
6. Oklahoma (22.7)
7. Notre Dame (20.6)
8. Auburn (20.0)
9. Florida (18.8)
10. Wisconsin (18.6)
11. Texas A&M (17.5)
12. Penn State (17.0)
13. UCF (15.9)
14. Oregon (15.6)
15. Michigan (15.2)
16. Washington (15.2)
17. Missouri (14.1)
18. Mississippi State (14.0)
19. Texas (13.7)
20. Utah (12.2)
21. Oklahoma State (12.2)
22. Iowa (11.9)
23. Michigan State (11.7)
24. Washington State (11.3)
25. Miami (FL) (10.0)
Next Ten, which we’re listing so we can remind you where NDSU falls: TCU (9.9), Virginia (9.1), USC (9.1), Kentucky (8.8), Boise State (8.8), South Carolina (8.6), Baylor (8.4), North Dakota State (8.3), Kansas State (8.1), Memphis (7.6)
Compared to the AP Top 25, we’re high on Wisconsin, Texas A&M, and still thinking teams are good after they do very ugly things. We’re low on Texas, Virginia, and thinking teams are good because they haven’t lost. We’re relying on ratings which in turn are still relying on preseason expectations. We are, as was said, just looking at how good teams are, and not at their résumé. But this is how our model views the top teams, which is what it then uses to go figure out everything else.
