James Madison thumped Texas State, and there’s a new favorite for mid-majordom’s guaranteed College Football Playoff bid.
Kind of.
Two lists from our model:
In probability, James Madison leads, USF is second, and Memphis is third. In average final ranking, USF leads, JMU is second, and North Texas is third. Probability’s more important—every one of these teams wants to win out, so “average” is a little irrelevant—but we look at average for a reason. In this case, that reason is that it shows us JMU doesn’t control its fate. USF has a worse chance than JMU of winning out and winning its conference. USF has a better average final résumé. It’s going to be hard for James Madison to finish ahead of whoever wins the American.
That all said, if JMU does win out, the Dukes are probably in. That’s true for a lot of these teams:
85% isn’t a bad probability. Neither is 69%. 85% is roughly the likelihood of a 14-point favorite winning a college football game, or the Eagles winning the NFC East. 69% is roughly the likelihood of a 7-point favorite winning, or the Dodgers beating the Nationals on an average day. Still, 14-point favorites lose. The Nationals on average take one of three games from the Dodgers. Nobody on this list controls their fate. USF, Memphis, and North Texas don’t even control whether they make the AAC Championship. Tulane at least has that piece working in its favor—the Green Wave haven’t lost a conference game yet, partly because they’ve only played three—but Tulane still has to worry about finishing ahead of James Madison.
One more list, and then a caveat that’s going to mess this all up even further:
This is the same list except instead of probability it’s where each team lands in the rankings if they do win out and win their conference. The implication here is that Tulane leads JMU, who then leads any other AAC champion plus Boise State. There are two problems, though. First, there’s usually hardly any space at all between 14th and 17th in the final rankings. That is well within our model’s uncertainty margin, especially when we haven’t seen any rankings yet from the CFP committee. At this point, it’s almost meaningless. Calling this a five-way tie is closer to accurate than saying Tulane’s in a comfortable position.
Second, the committee is most unpredictable when it comes to ranking mid-majors. There are so few ranked every year that we don’t have a lot of data on how the committee treats them. We mostly know where the committee will rank undefeated and one-loss teams from the SEC and the Big Ten. We don’t know how they’ll handle Memphis. Add in conference realignment and the reality* that committee members know less about mid-major teams than they do about high-majors, and it’s a cloudy picture.
What we do know bodes just fine for the American but poorly for JMU. Over the last four years, teams from the American have seen their rankings vary widely relative to our formula. 2021 Cincinnati was ranked where our model expected. 2022 Tulane got a little benefit of the doubt. 2023 SMU was surprisingly close to unranked. Memphis got inexplicable love last year while Army got surprising disdain. In the Sun Belt, the story’s shorter. 2021 Louisiana finished ranked 23rd when our model’s formula expected them to be ranked 16th. 2022 Troy was ranked 24th when our model expected them to be ranked 22nd.
That Sun Belt piece is a sample size of two, but it’s the only sample size we have, and it’s not encouraging for the Dukes. Couple that with JMU’s recent transition from the FCS (committee members probably don’t know that North Dakota State could be favored against any of these teams) and the Sun Belt’s history as more of a MAC-adjacent league, and it’s easy to see a scenario where so long as the eventual American champion doesn’t get beat again along the way, they get the bid and even a dominant–from–here, 12–1 James Madison does not.
Thankfully, we’ll get some answers along the way, starting on Tuesday night when the committee releases its first rankings of the season. For the mid-majors, those rankings are a bigger deal than they are for everybody else. Not so much in terms of who lines up ahead of whom—the AAC teams play better remaining competition than JMU or Boise State, so they have more room to rise but are also likelier to lose—but in terms of where each team lands relative to where our model expects them to land. If Tuesday comes and JMU’s unranked, that is a bad sign for JMU. Especially if at least one team is ranked from the American.
We’ll have more next week when this weekend’s results are in. And by the way, you’re right. Those first probabilities we shared don’t add up to 100%. Our model sees about a 1-in-8 chance two mid-majors get in. What a world we live in.
*I also dislike that the committee doesn’t know as much about mid-majors, but I’m not sure there’s a reasonable way to change it outside of going to a pure ranking formula, which would introduce its own issues (good luck finding a perfect formula—we have tried). Part of why the committee understands Ohio State with some precision is that it understands the Big Ten top to bottom. Asking committee members to divert attention on Saturdays to follow Louisiana–Monroe vs. South Alabama rather than Clemson vs. Virginia Tech isn’t realistic, and it wouldn’t help make the overall rankings better. More ranked teams play Clemson and Virginia Tech than play ULM.
**
