Tuesday night’s rankings didn’t much change the CFP picture. The games today and Saturday…should? Among the top twelve teams in Tuesday’s rankings, there’s a greater chance five lose this weekend than there is that all twelve win. But we’ll get to that. First, we’re going to take a quick walk through Tuesday night’s rankings. Then, we’re going to look at what those rankings did to the bracketology situation. After that, we’ll examine the different levels of chaos we could achieve in these next 48 hours.
Tuesday’s Rankings: What to Think
Our model has a variable called FPA (“Forgiveness/Punishment Adjustment”). Our model also produces a number we’ve been calling Ranking Score. Ranking Score is scaled from 0.0 to 100.0, with 0.0 the score of the worst résumé in the FBS (UMass) and 100.0 the top-ranked team in the CFP rankings. FPA feeds into Ranking Score and is used to reconcile our model’s expected rankings with the committee’s real rankings. Teams with a high FPA are ranked more favorably than precedent suggests. Teams with a low FPA are ranked less favorably.
Right now, among what our model calls the 30 top-ranked teams in the country, SEC teams have the highest average FPA. Fair enough. The SEC has nine teams in that top 30. It’s a good conference. In second place, well…look at the list:
- SEC: +2.0 average FPA
- ACC: +1.8 average FPA
- American: –0.4 average FPA
- Big Ten: –1.2 average FPA
- Big 12: –1.6 average FPA
- Notre Dame: –2.3 FPA
- Sun Belt: –6.9 FPA
Our model’s baseline assumption is that the committee will treat all Power Four conferences and Notre Dame the same, and that it will treat all mid-major conferences the same. So again, a favorable FPA for the SEC makes sense. The SEC is better than the average Power Four league. A negative FPA for the Big Ten does not make sense. The Big Ten is at least average among the Power Four leagues. A favorable FPA for the ACC really does not make sense. The ACC is not on the same level as the SEC or the Big Ten. (Or the Big 12, but there are at least angles for the ACC in that argument if you really, really squint.)
Still, these are small numbers, and they vary a lot from team to team. Here’s that top 30. The first number in parentheses is the ranking score. The second is their total FPA right now. We’ll intersperse some commentary.
1. Ohio State (100.0, +1.0)
2. Indiana (98.8, –2.4)
3. Texas A&M (97.5, +0.8)
These FPA’s are a little funky because we don’t have a good way to know the gaps between these teams in the committee’s eyes, nor do we know the size of the gap between Texas A&M and Georgia. Take these three with a grain of salt. That said, our model still thinks Indiana’s résumé is more impressive than Ohio State’s. Our model’s right about that. But the committee sometimes ranks the best teams and sometimes ranks the best résumés. Sorry, computers.
4. Georgia (93.0, +3.1)
5. Texas Tech (92.1, –4.3)
6. Oregon (91.3, –0.5)
7. Mississippi (90.6, +2.3)
The committee finally started respecting Oregon, or at least not actively disrespecting them. It took beating USC comfortably to get the Ducks back close to their baseline.
Texas Tech is still waiting for a reevaluation. Their only loss came to Arizona State in a game where their current starting quarterback was inactive. This week, the committee’s impression of Arizona State evidently really improved. Texas Tech doesn’t appear to have gotten any of that improvement themselves.
8. Oklahoma (89.1, +4.8)
9. Notre Dame (87.9, –2.3)
10. Alabama (86.9, +1.3)
11. BYU (85.7, –6.4)
12. Miami (84.5, +1.9)
13. Utah (83.4, –2.9)
14. Vanderbilt (82.4, +0.8)
15. Michigan (81.2, 0.0)
16. Texas (80.0, +2.9)
More negative FPA for the Big 12 here. Taken alone, this would be fine. Maybe the committee doesn’t think highly of the Big 12. If so, that’s prerogative. Taken in conjunction with Miami getting a better FPA than any non-SEC team we’ve seen so far, it gets a little weird. It’ll get weirder in a moment.
What’s going on with Miami? Well, their résumé isn’t very good. There’s been a lot of public pressure on the committee to ignore this and rank Miami ahead of Notre Dame because Miami beat Notre Dame by a field goal at home. The committee seems to be splitting the difference? That’s my best guess.
Texas has a good FPA but it used to be better. They dropped behind Michigan this week, which was surprising. What happened? One theory is that the committee didn’t want to explain why it had Vanderbilt (9–2) exactly one spot ahead of Texas (8–3) when Vanderbilt beat Texas head-to-head. This is the world the head-to-head obsession creates. Thankfully for Texas, the ranking compared to Michigan’s probably doesn’t actually matter that much. Should Michigan beat Ohio State, they’d have jumped Texas no matter what. Should Ohio State beat Michigan, Texas will either also lose (making their ranking irrelevant) or will beat the #3 team in the country (separating them from Michigan by a long distance). Our model is making something of that ranking, but it probably doesn’t need to.
17. USC (77.5, –2.9)
18. Virginia (76.7, +0.5)
19. Tennessee (76.1, +1.8)
20. Arizona State (75.5, +4.8)
21. SMU (74.9, +2.6)
22. Pitt (74.3, +3.4)
23. Georgia Tech (73.6, +0.6)
24. Tulane (72.9, +4.2)
25. Arizona (72.2, +1.0)
NR. North Texas (71.9, –5.0)
NR. Missouri (71.4, –0.2)
NR. Washington (71.0, –2.3)
NR. James Madison (70.5, –6.9)
NR. Iowa (70.0, –1.6)
This is a big group, but it’s all related. FPA is zero-sum within our model, and most of the positive FPA’s from Tennessee through Tulane are coming from negative FPA’s out on the fringe.
If any ACC team has a precedented case for a positive FPA, it’s Virginia. We’ve seen quarterback injuries in losses matter before. Chandler Morris was hurt against Wake Forest. SMU and Pitt are odder cases, probably driven by a combination of 1) Georgia Tech deserving to be ranked but Pitt blowing the doors off them and 2) SMU beating Miami, whom the committee has been convinced has a better résumé than their eleven games indicate.
The committee did tone down the love for Tennessee this week, and for Tulane. It’s possible they saw the reaction to the Tulane ranking and felt the need to backpedal. It’s not their fault! How could they have known Tulane didn’t deserve that? And don’t say, “Spend fifteen minutes researching mid-major football.” Because that is impossible.
USC continues to get slapped around, the largest driver of the Big Ten’s negative average FPA. USC wasn’t expected to be a good team preseason. The AP Poll’s consistently been lower on them than both résumé metrics (SOR, etc.) and predictive metrics (like Movelor, our own power ratings) would suggest. Sometimes narrative can be a killer.
As for Arizona State and Arizona…I don’t know, man. It’s not that ranking them is egregious, but it doesn’t mesh at all with the treatment of Texas Tech, BYU, and Utah. Again, narrative’s probably playing into that. Weird situation.
Tuesday’s Rankings: What to Think (continued)
Overall, the biggest takeaways from these rankings:
- The ACC and Big 12 champions will both make the playoff, no matter who they are. (Except for Duke.) All of them comfortably lead Tulane.
- Alabama is in a lot of danger, as we highlighted on Sunday. They should be just as concerned about Miami’s rise as Notre Dame is.
- Oregon could still get a bye. The most straightforward path to that is one loss each by Texas A&M (to Texas) and Texas Tech (in the Big 12 Championship).
- Mississippi’s hypothetical 10–2 résumé is bubbly. Bubbly enough that Lane Kiffin’s decision really could shape this playoff field.
CFP Bracketology
What did our model think? Up top, nothing moved. The bracket is the exact same as we had it on Sunday. Further down:
- Miami rose from 42% playoff-likely to 47%, which means their chance at an at-large bid at 10–2 rose from 40% to 45%. That might be a little juiced—our model assumes a 25% probability that conference championship losses don’t count, and speculation has that probability even higher (this would benefit Texas Tech, BYU, and Alabama). But that’s the magnitude of benefit these rankings delivered the Hurricanes. Only five percentage points.
- Texas dropped from 5% playoff-likely to 1%. 1-in-20 to 1-in-100 is a huge shift. On Sunday, our model didn’t see much of a path for the Longhorns, but what path it did see involved a great ranking Tuesday night after a good showing last weekend. Basically, the Longhorns needed an unlikely card, and that card didn’t come out of the deck. Now, they’re almost drawing dead.
- Oregon rose from 87% playoff-likely to 93%. The fact the committee accepted they’re better than Mississippi signals there’s some chance the Ducks can survive a loss to Washington. Not that they should try it.
The readouts:
CFP Bracketology (continued)
Again, Notre Dame’s average outcome is the 9-seed. There are only seven teams who average a home game.
Similarly, Texas A&M’s not in a great space to get a bye with a loss to Texas. An average ranking of 4th when there’s roughly a 50/50 chance they lose that game doesn’t bode as well as the 3-seed implies.
Last: BYU making it is not the chalk scenario. But as we’ve been saying, a lot can go right for them and a lot can go wrong for Alabama.
Chaos Time – The Top Twelve
The top twelve teams in the final CFP rankings will not make the College Football Playoff. It’s probably going to be the top ten, plus the ACC and American champions. But for now, the top twelve are the main characters. After Alabama, BYU, and Miami, there’s a big drop-off in playoff probability before the list reaches Utah.
We checked those 10,000 simulations, and we didn’t find one where nine or more of these twelve teams lose. We did, though, find some scenarios where eight lost. We’ll build to those.
Scenario 1: Chalk (5%)
Per our model, there’s a five percent chance all of Mississippi, Georgia, Indiana, Texas A&M, Alabama, BYU, Oklahoma, Miami, Notre Dame, Oregon, Texas Tech, and Ohio State win. If this happens, everything mostly stays put. Alabama probably benefits the most—they avoid a bad loss to Auburn and get to play Texas A&M in the SEC Championship instead of Georgia—but even they would really like to see at least one of the non-locks ahead of them go down.
Scenario 2: One top-12 team loses, but it’s Texas A&M, Ohio State, Georgia, or Indiana (8%)
Texas A&M is the likeliest of the twelve to lose (Movelor has them an outright underdog), but if Texas A&M loses, it doesn’t do a whole lot. It’s better for hypothetical bubble teams than one of the other three losing—aside from Alabama, the rest of the bubble wants Alabama to either miss the SEC Championship or have to play Georgia, who’s better than A&M—but even it doesn’t open up a spot on its own.
There’s also a bad scenario in here for bubble hopefuls. It’s chaotic, but not in a good way. If Michigan beats Ohio State, it only makes the playoff race more crowded. Michigan winning would add a team to the bubble without taking one away.
Scenario 3: One top-12 team loses, and it’s not one of the locks (11%)
This scenario does a lot for Notre Dame, Oklahoma, Mississippi, Alabama, and Texas Tech. Does it help BYU and Miami? Maybe. BYU probably still needs to win the Big 12 Championship to get in. Miami has one fewer team to worry about, but still no spot. Does it help Utah and teams below them? No.
Scenario 4: Two top-12 teams lose (28%)
The median scenario this weekend is somewhere on the chaotic end of this range. What does that mean? The average amount of chaos to expect from these twelve teams is either two big favorites losing (say, both Texas Tech and Notre Dame) or three narrow favorites going down (say, Texas A&M and Miami and Mississippi). That’s decently chaotic. Not enough to make many teams completely safe, but enough to push Oklahoma and teams above them up past a 99.5% playoff probability with a win.
Scenario 5: Three top-12 teams lose (26%)
We’re getting past the median here and deeper into the more chaotic half of universes. What happens if three in the top twelve get upset? It depends which three they are. And which three they are is hard to identify. Texas A&M, Miami, and Alabama are Movelor’s three likeliest to lose. But the combination where those three lose and the other nine win only comes up 1.85% of the time. For what it’s worth, that scenario probably locks in the top twelve: Ohio State, Indiana, and Oregon from the Big Ten; Georgia, Texas A&M, Mississippi, and Oklahoma from the SEC; both Texas Tech and BYU from the Big 12; and Notre Dame, plus the two remaining automatic bids. Not exactly chaotic, but you get the point.
Scenario 6: Four top-12 teams lose (15%)
As someone hoping for a little chaos, this is the “reasonable” amount I’m pining after. This is where Utah has a chance, and Miami is probably in with a win, and Notre Dame’s a lock with a win so long as Ohio State also wins. It’s also where teams like Oregon and Mississippi might be able to withstand a loss.
Scenario 7: Five top-12 teams lose (5%)
Cumulatively, we’ve reached 99% here. This is nearly max chaos. But remember: Five of these twelve losing is likelier than all twelve winning. This is almost enough to get Texas in the mix with a win.
Scenario 8: Six, seven, or eight top-12 teams lose (1%)
It’s actually 1.35%, not 1%. Now’s probably a good time to remind ourselves that our model’s imperfect.
Chaos Time – Conference Championship Scenarios
We’re not done! Shoutout to our conference tiebreaker explainer, which will make it into my law school application if my career prospects ever really hit the skids.
The ACC – Can Miami get in? Can Duke?
If you’re hoping for chaos in and of itself: We’re seeing an 11% chance Duke reaches the ACC Championship.
If you’re hoping for chaos because you want your favorite team to make the playoff: There’s a 41% chance Miami either loses to Pitt or makes the ACC Championship. Either of those scenarios helps bubble teams, because either of those scenarios nearly guarantees the ACC will be a one-bid league. Miami’s bubble competition doesn’t care about Miami if Miami gets in as an automatic bid. Miami’s bubble competition would love that, actually. What Miami’s bubble competition doesn’t want is Miami retaining an at-large case. Theoretically, they might retain one at 10–3, but a loss to Virginia, a loss to Duke, a loss to Georgia Tech, or a second loss to SMU would at least keep them from jumping anybody.
What to cheer for: The dream here—Miami vs. Duke—involves Miami, Virginia Tech, Cal, Duke, and NC State winning. If SMU does beat interim-coached Cal, either Duke or Miami will meet them in the event Virginia Tech beats Virginia and Miami beats Pitt.
The SEC – Will Alabama lose?
There’s a 69% chance Alabama either loses to Auburn or loses in the SEC Championship, which is pretty good. That’s Dodgers vs. Nationals, for baseball fans in the room. That’s the scenario bubble fans want.
For chaos fans, the scenario is a little less likely: Mississippi wins. Texas wins. Auburn wins. That’s the scenario that puts Mississippi in the SEC Championship against Georgia, potentially with an interim coach.
The Big Ten – How does everybody feel about Michigan?
Michigan does not necessarily make the Big Ten Championship if they beat Ohio State. Both Indiana and Oregon are ahead of them in tiebreaker-land. The True Chaos scenario here is both Indiana and Oregon also losing. It’s extremely unlikely. But. That’s what would give us a Michigan–Ohio State rematch in Indianapolis.
Bubble fans already want Oregon to lose. If Michigan does beat Ohio State, Oregon losing would at least force the Wolverines to play a 13th game. Just an added benefit.
The Big 12 – Surely, this can’t go chalk, right? It’s the Big 12!
This is probably going to go chalk. There’s an 82% chance both Texas Tech and BYU win. If only Texas Tech loses between that pair, the bubble interest is probably that Arizona beats Arizona State? In that specific combination of results, Texas Tech and BYU rematch in the Big 12 Championship, maximizing the likelihood of BYU losing and forcing Texas Tech to play a 13th game.
The American – The committee can’t help Navy now.
North Texas and Tulane are both favored by more than 20 points. They’re probably going to play each other, which means unless North Texas beats Tulane next week and the committee throws a curveball, James Madison is toast. Still, our model does give the Dukes a 5% playoff chance. That’s the same as that probability the twelve teams we focused on above go 12–0.
**
