What is CFBPRS?

CFBPRS is a predictive ranking system for college football. It rates teams based on what they have done on the field and uses those ratings to estimate what is likely to happen in future games. This site is built to let people explore those rankings, current and past game projections, and hypothetical matchups between teams.

The main pages are:

• Rankings: Shows where teams stand in the model and how they grade out in the main areas of the game.
• Games: Shows scheduled and completed games, the final result when a game is over, what the model expected in games that have already been played, and what the model expects for games that are still scheduled.
• Matchups: Lets you select any two teams and see what the model thinks would happen on average on a neutral field.

Any time you see a team name on the site, you can click it to go to that team’s page. There you can see all of its games, a more detailed breakdown of where it ranks across different metrics, and what the model thinks the Vegas line should be against every other FBS team on a neutral field. To see Indiana’s 2025 page, click Here.

How it works

The backbone of CFBPRS is an analytics system that studies every play from FBS vs FBS games across the season. It looks at how efficient teams are, how explosive they are, how quickly they play, and how those traits hold up once opponent strength is taken into account. The play-by-play data comes from CollegeFootballData.com's API.

That opponent adjustment matters a lot. A strong performance against a top team should not be treated the same way as the same raw performance against a weak team. By recursively adjusting for opponent quality, the system builds a much clearer picture of what each team really is.

On top of that, the game prediction side of the model was trained on historical data from the 2010s so it can compare two team profiles and estimate a result. That includes the effect of home field advantage when it is looking at real scheduled games.

Accuracy

The model is strongest as a team rating and winner prediction system. Over nearly 2,500 FBS matchups from the 2023 through 2025 seasons, it picked the outright winner 71.7% of the time.

Against the spread, the model was 51.8% over that same sample. That is respectable, but it should be treated as a secondary check rather than the main purpose of the system.

A useful benchmark is the sportsbook favorite. From 2023 through 2025, the market favorite won 72.6% of the time, while the model picked the winner 71.7% of the time. That gap is not surprising because sportsbooks are absorbing more information, including things like injuries and weather.

One encouraging sign is that the model’s confidence behaves the way you would want it to. As confidence rises, winner accuracy rises with it.