How good have you been, how well do you recruit, and who do you return? These three questions are the building blocks for how we set expectations for a college football team in a given year, and they are how I go about putting together preseason projections for my SP+ ratings.
Here's how I address each question:
How good have you been? For this, I look at a weighted average of recent performances. Yes, last year's SP+ ratings are heavily involved, but adding in your ratings from previous seasons with diminishing weight (so, what happened five years ago doesn't matter nearly as much as last year) gives us a pretty solid way to measure overall program health. The more you've been good of late, the most likely you are to be good moving forward.
How well do you recruit? You can find a "Stars don't matter!" take lighting up the internet at any moment of any day, but guess what: Stars matter. Teams will forever over- or underachieve compared to their recruiting rankings, but recruiting averages correlate well with performance on average, whether you like it or not. I look at a weighted average of recent seasons to get a good idea of the caliber of talent you are working with.
Who do you return? For decades, we have used returning starters as a proxy for measuring experience ... mainly because we haven't had anything better to use. But for a few years now I've been attempting to expand how we measure returning production.
Full projected SP+ rankings will come out in a few days, once recruiting is finalized for the 2021 cycle. But for now, let's talk about returning production.