The Victors

Obviously, no preseason projection is perfect. He explains that his projections were within a win of the overall totals last year for a lot of teams with some overachieving and others underachieving:

***

Preseason rankings therefore have two main purposes: First, they get us talking about the sport and help us determine which games to focus on early in the season. That's useful when there are a million games each weekend. I know people complain about how polls therefore set a narrative regarding a given team, but, well, Indiana got the No. 1 seed in the CFP last year despite a preseason No. 20 ranking, and Texas started out No. 1 but missed the playoff. The results will speak for themselves by the end.

In the case of projection systems like SP+ and FPI, meanwhile, they are pretty damn good at making predictions. You might not like that information from prior seasons -- plus bits 'n' bobs such as returning production and a dash of recruiting rankings -- can create accurate projections, and it sure would be a lot more fun if, say, SMU, Texas, Texas State and UTEP all started out with equal chances of being awesome in 2026.

Priors are predictive, however, even in the transfer portal era. And preseason rankings like SP+ paint an interesting and reasonably accurate picture; the formula doesn't get every team right -- and it wouldn't be any fun whatsoever if it did -- but it can tell us quite a bit. Of last year's projected top-15 teams in SP+, seven finished within one (regular-season) win of their August projections. And for the eight that didn't, close games (and, therefore, some randomness) played an enormous role.

• Six teams -- Georgia, Ohio State, Oklahoma, Ole Miss, Oregon and Texas A&M -- overachieved their projected win total by at least 1.4 regular-season wins; they were projected to win 52.1 combined games but went 66-6 instead, primarily because of a combined 21-3 record in one-score games.

• Two teams -- Clemson and Penn State -- significantly underachieved their projections. They were expected to win 19.3 combined games on average but instead went 13-11 in the regular season. They certainly spent much of the season looking like anything but top-15 teams, but they also went a combined 2-8 in one-score finishes.

There will always be some Clemson-style whiffs, and trying to guess which ones they'll be is part of the fun. But it's useful to find out what the computers have to say, and when it's time for teams to be judged on their CFP bona fides in a couple more months, they'll stand mostly on their own merits. Win-win.

***

It's useful to start with basic guesses and then see who does better or worse. We certainly know enough about these teams to go into the season with some basic assumptions. Otherwise, you could say that Ohio State and Bowling Green are equivalent until we get on-field results to prove otherwise.

Also, the optimistic assumptions that you are making about Michigan are probably being made with all of our rivals. Penn State fans are excited about the practice reports and think that their team will be much better coached this year. (Optimism is especially pronounced with a fan base where the prior coach has been fired and a new guy is in charge.) Iowa fans think that their offense is improving and they will have their normal strength on defense and special teams. Oklahoma fans think that their offense got worse when Mateer hurt his throwing hand and that they will be back this year with an elite defense and a better offense. It's a subjective bias to assume that our team will be better while everyone else remains static.