With the season about to start, Gator fan optimism seems to be peaking. In a recent poll in the Gator Country Bullgator Den, the most common answers to how many wins Florida football would have this year were nine (40.1% of votes) and eight (34.2%). That said, the combination of nine and ten (16.8%) added up to more than half of the total votes.
It’s natural to have the best feelings right before the season, but the unfeeling methods of projecting wins aren’t quite so high on the Gators. The S&P+ analytics system has UF favored in only eight games with a projected win total of 7.2. The FEI system, one of two along with the S&P+ that’s favored by the Football Outsiders crew, has more than 50% of its likelihood in the 8-4 or 7-5 range with only a 20% chance of doing better than eight wins. Only ESPN’s FPI among the most commonly cited analytics systems has the Gators solidly at as much as eight wins, and there it’s 8.2.
To put my cards on the table, I think there’s about as much of a chance Florida goes 9-3 as 8-4 this year, and I’d probably call 7-5 a disappointment even though it’d be a noted improvement over last year’s 4-7 record (5-7 minus the cancelation). I’m in the same place of being more optimistic than the algorithms.
Here are my best interpretations on why the formulas might be right about UF being solidly in the seven-to-eight win range and why the fans might be right to look for more.
Why the Formulas Might be Right
It’s worth noting first how these numerical systems come up with their preseason projections. Each does it a little differently, but most lean heavily on the team’s performance from the past four or five years, recruiting rankings, and returning starters and/or production.
The last of those is not a big worry. Florida has quite a lot of its production from last year coming back.
To continue in reverse order, recent recruiting doesn’t help a ton. The 247 Sports Team Talent Composite — which frustratingly still hasn’t been updated for 2018 — sheds some light here. Four teams on this year’s schedule were at least five spots ahead of the Gators last year: Georgia (4th), FSU (5th), LSU (6th), and Tennessee (12th). UF gained on the Vols by finishing well ahead of them in the 2018 class rankings, and it may have gained a little on LSU and FSU thanks to Florida having a good class and not losing a lot of highly rated talent to the NFL.
However, Georgia pulled ahead, and FSU and LSU will still rank substantially higher in total team talent this year than Florida will. That represents three potential losses right there if the talent rankings win out.
As for recent performance, having a pair of four-win seasons in the last five years drags UF down. Most of all, last year’s bottoming out weighs the heaviest in the formulas. When Florida went 4-8 in 2013, it still finished 55th in FEI, 36th in FPI, and 33rd in S&P+. This is because the defense was in or near the top ten in these measures in ’13. The 2017 squad had a defense that cratered to go with its bad offense, so the team ended up 50th in FPI, 83rd in FEI, and 86th in S&P+.
In short, the formulas don’t like the Gators’ chances against UGA and FSU for talent reasons. Thanks to talent not being far behind and a massive amount of returning production and better recent success, they all see Mississippi State as a top 11-15 team that is well ahead of the Gators (and it might be). LSU is in UF’s ballpark ratings-wise but with more talent, while Tennessee is closer to a tossup than I think fans think because it has similar talent on paper and the game is in Knoxville. South Carolina is on a cyclical upswing like MSU is, while Missouri is lurking around a similar rank.
With the systems seeing three likely losses and four relative tossups, 8-4 or 7-5 makes a lot of sense. Toss in the fact that even the best coaches sometimes experience first-year fiascos — never forget Nick Saban lost to Louisiana-Monroe in his first year at Alabama — and it totally checks out.
Why the Formulas Might be Wrong
Let’s hit the same factors.
Florida’s returning production doesn’t represent the actual amount of proven production that will be on the field. No one’s system counts Jordan Scarlett coming back after missing last year to suspension, nor do they account for Van Jefferson’s highly productive time at Ole Miss.
The recruiting data doesn’t account for guys who have blossomed above their ratings. Former 3-star recruits who should play like blue chips this year include Jachai Polite, Jabari Zuniga, David Reese II, Malik Davis, Kadarius Toney, Lamical Perine, Jawaan Taylor, and Brett Heggie. While other programs will have their own 3-stars who play like more, Florida seems to have a lot of them this year.
As far as recent performance goes, the ratings might not reflect reality even if they weren’t wrong. Florida was downright dreadful in the final five games of last year, so they should have gotten a poor rating. However, it’s debatable how much the team really tried in that stretch, particularly right after Jim McElwain was fired. No formulas have a way to account for a team checking out on a season.
Nor do these systems have a way to account for coaching changes. Dan Mullen should be able to improve Florida’s level of play merely by having hired a good strength coach and improving on Doug Nussmeier’s subpar play calling.
Florida also has a lot of opponents installing new offenses this year: Tennessee, Mississippi State, LSU, Missouri, South Carolina, and Florida State. While LSU will probably be fine because the new coordinator did a good job in an interim role in 2016, at least one of the other transplants will probably not take because it’s hard to change an offense in a year and a couple of them have first-time play callers. Florida’s offense could have a hard transition too, but struggling with a good scheme will probably still end up better than the struggling with a bad scheme the team’s been doing for a while.
Furthermore, I know a lot of Florida fans don’t hold the FSU and LSU head coaches in high regard for their in-game abilities. I think a lot of the reason I’ve seen the expected win total inch up from around eight months ago to nine or ten now is a rising belief that Mullen will make up for the talent gap by outcoaching Ed Orgeron and Willie Taggart. And, as it happens, Mullen did outcoach Orgeron and trucked the Tigers 37-7 in Starkville last year.
The bottom line is that no one — human or formula — is well equipped to predict what happens in a coach’s first year. There is more uncertainty than normal in those cases, and even probabilistic analytics systems can’t completely handle that. Florida complicates things by having an unusually high amount of incoming production from Scarlett and Jefferson and an unusually large number of opponents with coaching turnover.
Ultimately the reasoning is more important than the win total predictions, since someone throwing a 12-sided die could end up with the right win amount without knowing a thing about the team.
The formulas see a team with good but not elite talent returning a lot of production from a team that was bad last year and not great in the seasons before it having an okay but not great year. The humans, like me, see those same things but have adjusted the win expectation upwards based on immediate-impact changes like Scarlett and Jefferson (and Trevon Grimes, who wouldn’t affect returning production figures after missing most of last year to injury) and an upgraded coaching staff. The only way to settle this is on the field, which fortunately will begin happening very soon.