Welcome to our college football newsletter. If you like what you read and want similar posts delivered to your email inbox, subscribe to our Substack. It’s free!
**
Our final Movelor power ratings are posted. 1,665 games ago, they said Georgia was the best team in Division I football and that Stetson, who plays in the FCS’s non-scholarship Pioneer League, was the second-worst. Today, they say Ohio State finished as the best, and that Stetson was actually the worst.
Movelor does not try to reflect how good teams were over the entire season. Teams get better and worse as months go on. Movelor instead predicts how good a team will be in its next game, or would be if there were another game on the schedule.
This is an important distinction, because while part of what we’re looking at is how wrong Movelor was, we’re also looking at teams who got better and worse as the season went along. Movelor was wrong about Indiana in August. Arizona State got better over the course of 2024.
We’re going to go start by exploring a few noteworthy groups of teams. Rankings (the numbers accompanied by a number sign) are where each team ended the season within Movelor’s view of all of Division I. Ratings (the number to the right of the team, unless something’s otherwise specified) are the margin by which the team would be expected to beat or lose to an average Division I opponent on a hypothetical game played next week on a neutral field.
The Best Teams
#1 Ohio State: +49.5
#2 Notre Dame: +42.5
#3 Oregon: +42.0
#4 Penn State: +41.0
#5 Mississippi: +38.6
#6 Alabama: +37.9
#7 Texas: +36.6
#8 Georgia: +36.3
#9 Michigan: +34.7
#10 Florida: +33.2
#11 Tennessee: +32.3
#12 Indiana: +32.2
#13 LSU: +31.2
#14 Arizona State: +29.7
#15 Minnesota: +29.0
#16 South Carolina: +28.9
#17 Missouri: +28.7
#18 Clemson: +28.7
#19 USC: +28.2
#20 SMU: +28.1
#21 Miami (FL): +27.6
#22 Louisville: +26.5
#23 Auburn: +25.9
#24 Iowa State: +25.4
#25 Iowa: +25.4
It’s easiest, I think to divide the top 25 into four sections: First, you have Ohio State. One touchdown behind them, you have Notre Dame, Oregon, and Penn State. One field goal behind them, you have a pack of SEC teams featuring Mississippi, Alabama, Texas, and Georgia. After that, you get to everybody else.
The first two groups are straightforward. It’s interesting that Movelor thinks Penn State was better than the whole SEC at the end of the year, but that doesn’t exactly ring any sirens.
The third group tells an important story. Georgia and Texas might not have been the two best teams in the SEC. They earned their SEC Championship berths, but Alabama and Mississippi were probably better, as each showed in beating Georgia (who beat Texas twice). It’s very tight, so we aren’t trying to make any huge claims here, but at the very least, Alabama and Mississippi were comparable football teams to Georgia and Texas. The difference in outcomes stems from each losing one or two games they absolutely should not have lost. Were Alabama and Mississippi better than Georgia and Texas? Probably. But they weren’t consistent enough to make it count.
This idea of seizing opportunities shows up in the fourth group as well: Everybody else. So does the concept of luck. Clemson and SMU were lucky to play in the ACC. Michigan and Florida started the season dreadfully but finished as good teams. Tennessee, Indiana, and Arizona State made the most of their opportunities, each around the edge of the top twelve but not meaningfully better than Michigan, Florida, LSU, or even Minnesota.
Other teams who jump out in the top 25:
Minnesota, firstly, who beat USC and nearly beat both Penn State and Michigan. USC, who beat LSU and lost to four top-15 teams. The Big Ten was better this year, and a lot of this was because of Ohio State, Oregon, and Penn State. The focus on those three and Indiana obscures that Minnesota made big steps and USC was sneakily respectable. We’ll talk about each of these teams’ 2025 outlooks in the next few weeks.
Miami ended the year only a little better than Louisville, right there in that bunch of ACC teams. Louisville did lose to Stanford, but the Cards had the toughest schedule of those four by far and should be considered a near-equal to the three above it. Miami was in line for the playoff’s 2-seed towards the end of the season. I wonder what would have happened to them had they played a top-8 team.
Auburn didn’t make a bowl game, but the Tigers were competitive, better on paper than Texas A&M, Oklahoma, and Vanderbilt by the end of the year.
No surprise about Iowa State and Iowa, but it is very fitting that those two teams ended the year as effective equals.
More Good Teams
A selection:
#26 Colorado: +25.1
#27 Illinois: +25.1
#31 Boise State: +24.5
#36 North Dakota State: +24.1
#38 South Dakota State: +23.5
#44 Montana State: +20.9
#46 Syracuse: +20.7
#49 UNLV: +20.4
#51 Memphis: +19.8
#56 Army: +18.1
This list includes all teams who finished the season ranked by the AP Poll, plus the FCS teams who were better than at least one of them. I don’t know that it’s unusual, but it’s noteworthy that three separate FCS teams were better than the Group of Five’s second-best, at least in the eyes of Movelor, whose ratings are calibrated to maximize long-term accuracy in FBS vs. FCS matchups. The Group of Five is struggling to produce good teams right now. Even this Boise State team, which earned a playoff bye, was not one of the best Boise State teams we’ve seen in recent memory. The Chris Petersen teams were better.
The Group of Five’s average rating (+5.0) is better than that of even the FCS’s top twenty teams (+4.2), so this is mostly a reflection of how good the best FCS teams are. With Montana State’s roster turning over a bit and South Dakota State’s whole program turning over, it should only be North Dakota State in this mix next year. But it’s still interesting, one of those quietly important details about the FBS/FCS landscape.
Teams Who Improved (and Whom Movelor Missed)
The number to each team’s right here is not their final rating. It’s how much better they finished the year compared to their preseason rating.
#14 Arizona State: +22.8
#12 Indiana: +20.2
#47 Vanderbilt: +16.9
#15 Minnesota: +15.8
#26 Colorado: +15.1
We’ve already mentioned four of these, but Vanderbilt’s the fifth Power Four team who dramatically overperformed preseason expectations. I do think Movelor underestimated Indiana, Vanderbilt, and Colorado, though each is a miss I can live with. I don’t think there’s a big structural flaw in Movelor that led to any of those misses. I do wish we’d built Movelor to catch up to Indiana sooner. But while I’d have to go back and check, I’m not under the impression that Movelor was missing wildly on ASU and Minnesota from the jump. Three of these teams were better than we thought. Two got better than they were to start the year.
#171 Lehigh: +18.9
#183 Stony Brook: +17.6
#217 West Georgia: +17.3
#193 McNeese: +16.2
#198 The Citadel: +15.1
#55 Marshall: +14.6
Among the rest of the eleven biggest overperformers (we stretched the line to eleven so we could get one Group of Five team in here), we’ve talked about two before. Stony Brook and McNeese both showed up in an early-season post about teams Movelor kept missing. West Georgia was new to Division I, and we treat all those teams the same, which may or may not be a good idea. We’ll revisit that. Congratulations to Lehigh and The Citadel. Big strides. I’ll try to learn something about each.
I’m curious if there’s any significance to no Group of Five teams making the list of the ten biggest overperformers, especially given what we just talked about regarding the best teams in the FCS. I don’t think there is, unless it’s to show that we didn’t get our usual breakout mid-major. Nobody really broke out. Some years, that probably happens. Other years, you probably get a handful of teams surging to the front.
Teams Who Got Worse (and Whom Movelor Missed)
Again, the number here is how much their rating changed, not what their postseason rating was. We’ll go back to postseason rating in the next section.
#85 Florida State: –24.7
#76 Arizona: –19.1
#81 Oklahoma State: –18.8
#117 Purdue: –13.1
#9 Michigan: –11.9
#73 Maryland: –10.5
#8 Georgia: –10.4
This list is more interesting. Of our seven biggest Power Four underperformers, Movelor should have seen three coming, and the industry did see two coming. I’m not sure anyone’s fully processed Georgia.
Florida State and Oklahoma State were shockers. I don’t know who could have predicted that. A little worse? Fine. Worse than Houston? Shockers.
Arizona should not have been a shocker. The staff turned over, and we said, “But Noah Fifita and Tetairoa McMillan!” Still, really bad job by Brent Brennan in his first year. Arizona did retain a lot of talent.
Purdue was a disaster. I think that’s surprising. They were supposed to be bad but not a disaster.
Movelor was really high on Maryland relative to the consensus after the Terps finished 2023 with a narrow loss to Michigan and comfortable wins over Rutgers and Auburn. A lot of the roster changed, though, and we’ll need to look into whether that’s something we should have better accounted for.
Michigan was a red flag for us, someone we spent a lot of time addressing right before the season started. We knew they’d be worse than last year, but they’d been so good at the end of last year—about the same in Movelor’s eyes as what Ohio State is now—that it was hard to persuade Movelor to back down, even when we told it about preseason AP polls dating back through its whole history. Michigan is the kind of miss we don’t want Movelor to make, and while a lot of that is admittedly optics—missing Maryland is just as bad as missing on a potential title contender, but no one really cares if you miss Maryland—optics are important as we seek to gain your trust.
Georgia…ten points. That’s a lot. Georgia really, really got worse this year. Bafflingly worse, given what they did to Texas in Austin and given they still won the SEC. The at least short-term Georgian downfall is a huge development for college football.
Bad Teams
#117 Purdue: +2.1
#105 Stanford: +6.0
#90 Michigan State: +9.6
The three worst Power Four teams. Tough look for Michigan State, who I’d imagine was hoping to get better, not a little worse, in Jonathan Smith’s first year.
#228 Kent State: –20.7
#191 Southern Miss: –11.8
#187 Tulsa: –11.3
Expanding to the whole FBS…some bad teams. For reference, Kent State was only better than 35 of the FCS’s 129 teams.
#263 Stetson: –46.0
#262 Northwestern State: –44.6
#261 Marist: –40.5
And lastly, the worst of the worst. Stetson and Marist play in that Pioneer League, where there aren’t football scholarships. Northwestern State is still recovering from the murder of one of its players last fall.
Movelor’s Performance This Year
2024 wasn’t as strong a year for Movelor as 2023, but the average miss was only a third of a point worse, relative to the closing spread. The closing spread missed by 12.04 points, on average, in FBS vs. FBS games. Movelor missed FBS vs. FBS margins by an average of 13.16 points. Movelor was a little better on FCS games, which is either a glorious opportunity to reap gambling profits or a meaningless accomplishment.
We like to talk about potential improvements to our models, but we don’t always make them. Our wish list, then, which we may or may not address…
1. For the purpose of offseason adjustments: Improve how we quantify the amount of talent present within a program. For talent, we’re still using a weighted, rolling average of recruiting scores that tells us next to nothing. In the transfer portal era, we need to better account for transfers. I don’t think we’re capable of going into returning production, like some systems do, and I don’t think that’s necessarily the smartest way to evaluate programs, at least if you don’t temper your returning production number with something recruiting-based. Basically, we need to better identify which teams got a lot better and a lot worse. A little better or a little worse doesn’t hurt us. Movelor reacts quickly enough to handle that just fine. A lot better and a lot worse is tough. Movelor should not have ranked Michigan second in the country to start the year.
2. Again, for offseason adjustments: Is there a way to use coaching changes to help us? Last offseason, we tried to look at this, and we generally found that teams got a little better in a coach’s first year. There was a wide range of variation, though, so I wonder if we can identify variables which predict which first-year coaches will have good years and which will have bad years.
3. For the purpose of in-season ratings changes: We should probably make Movelor react more dramatically earlier in the season. Movelor consistently thrives in the last two thirds of the year, but it stinks in August and September. Stinking in August is hard to avoid, but can we get our system caught up sooner to teams like Indiana? Part of this is just increasing how much we allow Movelor to move teams’ ratings early in the year. Part of it might be introducing an uncertainty score for teams with first-year coaches.
There are more, smaller things, but those are the three big ones. Hopefully addressing them can help us close the gap with betting markets, at least by enough to get Movelor in line with some of the best injury-blind systems in the industry. We don’t hold any aspirations of consistently beating those markets, but we would like to catch up to ESPN.
**
Enjoying what you’re reading? Subscribe to our free Substack to receive it directly in your email inbox.
**
I’m unaware of any major college football news today, but if there’s something I’ve missed, I’m sure we’ll catch up to it tomorrow. And on the topic of tomorrow: Ohio State’s 2025 outlook. Let’s find out who these guys are bringing back.
**
Thanks for being here. If you like what you just read and want similar posts delivered to your email inbox, subscribe to our free Substack.