College football expectations vs. reality: What did I get right and wrong in 2016?

In today’s media landscape (sports and non-sports), self-reflection is unnecessary. You don’t really need to step back and look at what you got wrong or why; you just need to push forward with even more confidence than before.

But does Nick Saban self-reflect? Hell yes, he does! Verne Lundquist? Absolutely! Lamar Jackson? Yep!

Process is important, and now is as good a time as any to look back at what we got right and wrong in 2016.

So here are my preseason projections — by both my S&P+ analytics system and my own opinions after writing a few thousand words about each team during the offseason — compared to current rankings. That’s followed by a look at the teams S&P+ was either very high or very low on.

S&P+ performed about as expected on a week-to-week basis. After an early run of bad luck (a terrible record in games decided within three points of the spread), it rebounded to hit 51 percent against the spread for the season. Meanwhile, win probability projections were as strong as ever. You can find S&P+ performance data for 2016 here.

This table presents 2016 S&P+ preseason projections, my preseason power rankings, the end-of-regular-season S&P+ rankings, and the differences between preseason and present tense.

First things first: My rankings were closer than S&P+’s projections for 61 teams, by an average of six spots. S&P+ topped me on only 58 teams, but by an average of 6.7 spots. I came within five spots on 25 teams, within 10 on 48. S&P+ was within five on 35 teams, within 10 on 50.

That’s a slight win for S&P+ overall, but since I’m the one keeping score, I’ll call it a draw.

Thirteen teams overachieved their projections by at least 30 spots in the rankings:

What do these teams have in common? First of all, 12-of-13 are from Group of 5 conferences. Colorado is the only exception.

Second, seven of the 13 are led by coaches in either their second or third seasons in charge. Three others are in their fourth. Only Memphis (under Mike Norvell) was led by a first-year guy, and only ODU was led by a guy who’s been around at least six years.

If you’re going to take a leap, your second or third season is probably when you’re going to do it. We’ll revisit this topic soon.

Eighteen teams underachieved by at least 30 spots in S&P+.

You can find some themes here, too. Six of these teams (BGSU, Baylor, Georgia, Syracuse, Rutgers, Georgia Southern) were led by first-year coaches, as were Illinois (off by 26 spots), Virginia (26), and Maryland (25). Sometimes even good coaches just don’t get all the pieces in the right places in the first 12 or so games.

Another commonality: Most of these teams have had great seasons recently. Bowling Green won 28 games from 2013-15 before losing Dino Babers to Syracuse. Oregon was in the national title game in 2014. Georgia won 50 games from 2011-15 but decided that wasn’t good enough and made a change. Baylor won 50 from 2011-15 but headed into 2016 with an interim coach after Art Briles’ swift downfall. Marshall won 33 from 2013-15. Sometimes your slope for regression to the mean is pretty stark, especially if you also have a first-year coach.

One more correlation worth mentioning: Eight of these 18 teams were among the top 25 in 2016 recruiting, and nine were among the top 28 in two-year recruiting. There is a strong correlation between good recruiting and good play, but it’s scattershot. It doesn’t affect each team the same in the short-term, and for all the predictive value recruiting can have (the top six teams in two-year recruiting are among the top 11 in S&P+, as are seven of the top nine in five-year recruiting), there will be misses.

That goes for all projection factors. Recent history wasn’t helpful in projecting Oregon’s or Georgia’s S&P+ rating, and returning a boatload of experience didn’t save Syracuse from being mostly anonymous in Babers’ first season.

You can make pretty good projections with college football, but there will always be things you don’t see coming. It would be pretty boring if that weren’t the case.

Still, there weren’t that many surprises up top.

Of the current S&P+ top 10, nine were projected in the top 11 at the start of the season, and the only outsider (Louisville) was projected 18th. The current six best have spent exactly one week outside of the top 11. There was, and will always be, chaos and turmoil in the middle, but the top of college football fell into place mostly according to plan.