PFEI Preseason Projections
updated August 17, 2024
It is tempting to think that college football, with almost constant roster turnover, might be susceptible to wild swings in performance among teams. But we know that's mostly not the case. How good your team was last year actually correlates very strongly with how good your team will be this year.
Previous year data is a solid starting point for projecting the season ahead, and data over several prior seasons is even better. Though data from five years (and an entirely different roster) ago may not intuitively inform a projection model, I've found that including it does make my projections stronger. PFEI ratings are produced from the results of games and possessions played in each of the previous five seasons, significantly weighted for the most recent results. PFEI ratings have consistently outperformed other model projections I have tested in terms of correlation to next-season FEI ratings.
The 2020 college football season was an outlier for many reasons and featured an extraordinarily disconnected network of teams. For many programs, efficiency results in 2021 "bounced back" to levels more closely aligned with 2019 data than to the year in between. For these reasons, an adjustment for unreliable 2020 data has been included in the PFEI formula.
It is also notable that the year-to-year correlation of FEI ratings peaked in 2023. Despite transfer eligibility rules allowing for much more player movement than in years past, team performances were relatively stable between 2022 and 2023. This also could be an outlier rather than a trend, but it further supports why the FEI preseason projection model is built on priors.
Correlation of Preseasion Projection Data to Team and Unit Ratings |
||||
| Rating | Definition | LY Data | PFEI Data | |
| FEI | Opponent-adjusted overall possession efficiency | .730 | .760 | |
| OFEI | Opponent-adjusted offensive efficiency | .585 | .664 | |
| DFEI | Opponent-adjusted defensive efficiency | .597 | .663 | |
| SFEI | Opponent-adjusted special teams efficiency | .324 | .403 | |
A projection model built exclusively from prior performance data admittedly has blind spots, but I expect that the PFEI model will continue to be an effective starting point for team season ratings as they have proven to be in the past. Preseason projected data is filtered out in weekly FEI ratings updates as new game results roll in throughout the year.