As the season approaches and Dan Mullen embarks on his flagship season as the Gators Head Coach, I thought it would be interesting starting point to think about how to set reasonable and predictive expectations for the Gators. But I wanted to try to take things a little further.
Instead of looking just at individual years’ recruiting rankings, I want to try to estimate the overall talent on the roster, and compare the performance in a given year to how you’d “expect” that team to perform based solely on talent alone. So, we need to know two things: how talented is this current team, and how well does it perform
The last three years, 247 has created “Team Talent Composite” rankings based on the actual rosters for that season. Ideally, we would just use those values to set the baseline expectations–if a team is the 20th most-talented in the country, then you’d expect them to play like the 20th-best team in the country, all else equal.
Since Team Talent Composite rankings only go back to 2015, but Mullen’s tenure as a head coach stretches all the way back to 2009, I’m just using the average national ranking for the prior four recruiting classes as a proxy. For the years we *do* have Talent Composite rankings, they’re pretty close to the four-year rolling average:
So, we’ve got our rough baseline for measuring how talented the team is.
We all know that schedules vary wildly in college football, so even when two records may be similar, the quality of the teams can be very different. That makes it tough to try to evaluate how well a team like Mississippi State is actually playing, since they’re plopped into the toughest division in college football.
If Mississippi State is playing against a schedule full over juggernaut teams, they may actually be playing like a top-25 squad but end up with a pretty middling record. So how do we adjust our performance rankings for schedule? Well, we can use S&P+ rankings, which don’t rank teams based off of wins and losses, but instead look at overall performance on a per-play basis.
This ranking is consistently good at predicting who will win or lose a game, and more importantly it predicts the spread and win likelihoods about as accurately as Vegas. So we can look at these rankings to tell us how well a team played, on average, in a given season, adjusted for their schedule.
First let’s look at how Florida’s performance under McElwain compared to the talent on hand:
So we see underachievement in two of three years, and one year of roughly meeting expectations. One could make the argument that the 2017 S&P+ performance rating shouldn’t be totally credited against McElwain, since he didn’t coach the last few games. But we’re not so much worried about exact numbers here as the main idea: he badly underperformed.
Now let’s look at Mullen versus expectations:
Here we see that, on average, his teams performed *very* close to expectations based on talent. We see slightly more overachievement than underachievement, and I’m sure someone could dig into the details of these teams to explain the biggest drivers. The most obvious is that the biggest overachievements came with Dak Prescott at QB. My takeaway is that Mullen isn’t a master developer, but he’s good at deploying what he has on hand. And it’s no surprise that his system hinges on quarterback play.
One final visual that’s worth checking out: if you sort his years by over/underperformance amount, it’s spread pretty evenly among possible over/underachievement values. In a third of seasons he overachieved by 10 or more, in a third of seasons he achieved within 10 of expectations, and a third of seasons he underachieved by 10 or more.
So, what can we expect for the Gators in 2018 and beyond? Well, the average rank of our last four classes (2015, 2016, 2017, and 2018) is 14.5. It seems pretty optimistic to expect the team to immediately jump all the way to that level of performance under Mullen, but even if we match the worst underperformance of Mullen’s MSU tenure–28 spots–that would land the Gators as the roughly 43rd-best team in the country. Preliminary S&P+ ratings released today have us at #32.
My guess is that our 2018 team will flirt with the top 25 and ultimately finish right in the 20-30 range. But as we move fully into the Mullen era in 2019, I’d expect the team’s performance to start mirroring the recruiting rankings pretty closely. If Mullen can start consistently reeling in top-ten classes in the future, then we should reasonably expect top-ten performance in the fall.
There will almost certainly be some fluctuation up and down from year to year, but the biggest driver will likely be the quality of the guy standing behind center. If Mullen can get and develop the right QB for his system, then we should have the talent in place to compete with the elites of the sport.
No pressure, Emory.