Sometimes, it’s easy to know how good a college football team is. We knew last winter, for example, that Ohio State was about nine points better than Notre Dame. With less specificity, we knew in 2020 that Alabama was a heck of a lot better than everybody else.
Other times, this is an impossible question. Usually because of inconsistency, whether that be short-term inconsistency like Alabama’s or season-over-season inconsistency, the kind we’re now seeing from Florida State.
How good is Florida State? I don’t know. But it’s a big question, and with Week 3 idle for FSU and Week 4 a formality against Kent State, it’s going to remain a big question. We won’t know much more about this Florida State team until they play Virginia. We might not know the whole story until they’ve played Miami the first weekend in October.
Partly as a thought exercise and partly because I desperately want to know the answer, let’s talk through four methods of evaluating the Seminoles. Each has its flaws. Each gives only a partial answer. But maybe through them, we can piece together an estimate.
Movelor
We have a college football model, and it has a power rating system named Movelor. Movelor says Florida State is the 56th-best team in the country. We don’t appreciate this from Movelor. We saw the Alabama game. We saw FSU hang 77 on East Texas A&M, a bad Division I team but still a Division I team. Why can’t Movelor see these things too?
Movelor’s big strength and big weakness is its simplicity. It doesn’t overthink itself. When it’s wrong, we can usually see how and we can usually see why. In this case, it isn’t reacting fast enough. We would love to build in a variable that measures a team’s unpredictability and instructs Movelor to react more quickly when that unpredictability is higher. Movelor is simple. We have not yet built that variable. To be honest, it would be hard to build that variable in a way that even captures Florida State. This team is bizarre, and in ways we did not see coming. That makes them even more bizarre.
The best way to check whether Movelor is properly evaluating a team is to look at that team’s record against Movelor’s spread. Unfortunately, it’s only Week 2. Plenty of teams are 2–0 against Movelor’s spread. In lieu of that, then, here are the five teams whose Movelor rating has changed the most since the season began. Theoretically, they should be the five Movelor is still at most risk of being wrong about:
- Tarleton State (up 11.7 points)
- West Georgia (up 9.4 points)
- Florida State (up 9.3 points)
- USF (up 9.1 points)
- UTRGV (up 9.0 points)
On that list, you’ve got a team in its sixth year of FCS football, a team in its second year of FCS football, a team in its first year of any sort of football, USF, and Florida State. Yes. Movelor is likelier to be underrating Florida State than it is to be underrating any other FBS team. It’s a game of catchup, and we’re holding Movelor back because if we let it race to catch FSU, it would start chasing every single result and ultimately self-destruct. But Movelor is wrong about Florida State. The question is more about when it’ll catch up than whether it needs to.
Those 9.3 points are useful, though. Across Weeks 4 and 5—the next two weeks with FSU games—Movelor can only react 75% as strongly as it did across Week 1 and Week 2 combined. Accounting for that, then, let’s say Movelor pushes Florida State up another 7.0 points, 75% of how far they’ve risen so far.
At this moment in time, 7.0 extra points would leave Florida State 34th in Division I in the eyes of our model. Closer to accurate, but we’re still skeptical. That’s right in front of Iowa, Vanderbilt, and Rutgers. FSU’s better than those guys, right?
From there, a good rule of thumb is that Movelor will move teams a maximum of about three points per week. How this shakes out is that if FSU’s really around the edge of the top 25, Movelor will catch up after Week 6. If FSU’s really around the edge of the top 12, Movelor will catch up after Week 7. If FSU’s really around the edge of the top 10, Movelor will catch up after Week 8. This is a long time away. This is a big Movelor weakness. In our defense, Movelor caught up even more slowly to Indiana last year, and there’s a good chance Florida State isn’t a top-12 team. But this is bad stuff from us, and part of why we’re writing this post is to say that we know we’re wrong and to explain why.
First Principles
The big reason Movelor is so wrong about FSU is that we’ve found a great indicator of a team’s strength is how strong that team was the prior season. How strong was FSU last season? Not strong at all!
Let’s see what happens if we were to scrap last season and come in blind.
247’s Talent Composite doesn’t break out by age or experience, but it’s a useful tool in figuring out a team’s ceiling. At a raw level, Florida State’s got a ceiling somewhere among the top 20 ceilings in college football. Could it be higher than that? Yes. 247 ranked FSU’s transfer class sixth this offseason. Programs that bring in a lot of quality through transfer generally have a higher ceiling than the Talent Composite implies, because their roster overindexes on players who are exceeding their original expectations. High school recruiting rankings added up across a roster (this is what the Talent Composite does) are a great tool for estimating the total amount of speed, size, and strength on said roster. But transfers into high-transfer programs are likelier to be late bloomers than the average player.
Still, FSU’s transfer class was only sixth. FSU’s only 19th in that talent composite. FSU returned an average share of production from last year’s roster. Even giving bonuses for quarterback moxie (a questionable but understandable practice), it’s hard to see how FSU’s ceiling ranks higher than 15th. FSU might be better than the 15th-best team, but their ceiling isn’t one of the ten best in the country, and we’ve seen no indication in their Mike Norvell era that they have a numbers-defying culture or a numbers-defying ability to develop certain position groups. The closest thing they have to a numbers-defier is Gus Malzahn, and it’s hard to believe he’s such a good offensive coordinator that he could elevate FSU to top-ten-ceiling status, given how his tenure went at UCF.
Floor? We saw this program’s floor last year. It’s low.
Where you can land, based on this—and this is where I do land—is that Florida State has probably been playing close to its ceiling so far, and that it did catch Alabama on a terrible Alabama day (or it induced Alabama into a terrible Alabama day). FSU blew out East Texas A&M to a degree that isn’t easy to do, but we’ve seen Mississippi and USC turn in performances like that over the last few years and wind up merely good, not great. Given how few teams play to their ceilings this early, I think you could be justified in calling FSU a top-ten team so far. But the expectation should be that questionable depth, distractions, and some known limitations catch up to these guys. For Florida State, the question isn’t whether they’re good enough to win the ACC. It’s whether Miami is bad enough to let them. And, you know, how that Alabama win will age. We could conceivably see a CFP committee snub Florida State at 11–1. Wouldn’t that be something?
SP+, FPI
We’ve mentioned how much we respect SP+ and FPI, so in the interests of this case study, they’re worth a look. SP+ started the Seminoles at 39th and has only moved them up to 31st. All systems struggle to adjust quickly enough to the really odd teams, but it’s worth noting that SP+ has dropped Clemson from 10th to 30th. It can move teams a long ways. It hasn’t moved Florida State very far.
I don’t have access to any sort of FPI archive, but we do know that FPI had FSU a 17-point underdog against Alabama and a 35-point favorite against East Texas A&M. SP+ had those numbers at 14 and 53. Movelor had them at 27 and 35. That seems to imply that FPI was somewhere between SP+ and Movelor at the start of the year, though we don’t know what it thought of Alabama. Either way, it might be reacting more quickly than SP+, and it still doesn’t have these guys in the top 25. Maybe my First Principles estimation was too optimistic. Maybe that Alabama game was just another round of Kalen DeBoer playing 2024 Vanderbilt.
Betting Markets
You might wonder why we’re not talking more about betting markets, since—as we often say—they’re the most efficient predictor of college football results. The reason is partly that even as the best in the business, they too struggle this early in the season. More than that, though, it’s that we don’t have a line on FSU vs. Kent State or FSU at Virginia, and the FSU vs. Miami line exists but is probably misleading. This far out, those forward-looking lines are promotional more than they’re serious markets. Betting limits are a lot lower. The number doesn’t tell you as much.
That said…
We’re seeing FSU listed as a 4.5-point underdog at home against a team ranked 12th/27th/21st by Movelor/SP+/FPI. If betting markets are more efficient here, it’s Miami they think SP+ and FPI are undervaluing.
If you enjoy these posts and want to receive them more directly, we have two options for you. The first is to subscribe to our Substack, where we currently exclusively publish college football blog posts roughly six days a week. The second is to subscribe to our daily newsletter, a morning rundown of everything going on at The Barking Crow.
**
