GAINESVILLE, Fla. — To start the season, I used advanced statistics to determine whether Florida or Tennessee would win the SEC East. The analysis clearly favored the Vols, with coaching being the main advantage for Florida.
Halfway through the season, has that assessment changed?
Florida leads the East Division with a 5-1 overall record (3-1 SEC), but with the postponement and relocation of the LSU game, the quality of their opponents is backloaded (Georgia, Arkansas, South Carolina, and LSU).
Tennessee is second in the East with a 5-2 overall record (3-2 SEC), but they are coming off a bye week after the gauntlet of Florida, at Georgia, at Texas A&M and Alabama. The backside of their SEC schedule includes barnburners against South Carolina, Kentucky and Missouri.
Based on those schedules, the common thought process is Tennessee is unlikely to lose an SEC game again and so Florida needs to win all its conference games. But is that actually true?
The Pythagorean expectation — a statistic invented by famed sabermatrician Bill James — can be used to give a true measure of how a team performs. This statistic looks at points scored and points allowed and predicts what a team’s record should have been based on those numbers. I looked at the first five games of the previous 11 Gators seasons to see if this statistic is predictive.
What the above chart tells us is that the predicted win values just from the scores of the first five games are remarkably close to the actual win totals that those teams accumulated over the full season. This means that we can be fairly confident that there is something to be gleaned from the fact point differential predicts the Gators are a 10-win team this season.
As a side note, it also tells us Will Muschamp was a really bad coach. He consistently underperformed against the meat of the schedule compared to the first five games by an average of 2.5 games.
Admittedly, these calculations do tend to predict more wins than actually produced (1.3 wins on average), but this is due to two reasons: First, 2013 is a clear outlier where Florida started out the season 4-1 and then lost 7 straight as QBs Jeff Driskel and Tyler Murphy were injured. Second, typically the first game of the season is against an inferior opponent. If you subtract the first opponent and examine only Games 2 through 5, then the calculations only over predict the win total, on average, by 0.5 wins. However, even subtracting the UMass result from the Gators ledger does not change the expectation that this is a 10-win team.
So, how can we use this to predict the SEC East? The figure below shows the actual wins, expected wins based on score differential (Pythagorean expectation) and the difference between the two for each team in the SEC.
The teams are ordered not based on their records, but based on their expected wins based on score differential. This reveals some interesting trends:
First, there is Alabama and then there is everybody else. This is probably obvious, but it is striking how much better the Crimson Tide are than everybody else in the conference. Auburn also grades out pretty well using this metric, indicating that its season opener against Clemson wasn’t a fluke.
LSU and Florida are next, and then there is a definitive falloff in expected win totals after Texas A&M with Tennessee, Georgia, Vanderbilt and Arkansas at various stages of average.
The bottom three teams on this list are particularly terrible. These just happen to be the three teams left on Tennessee’s schedule, and so the expectation should be the Vols win the rest of its SEC games.
But Florida’s schedule isn’t as grueling as it might appear either. Georgia, Arkansas and South Carolina are all in the bottom half of the conference using this metric, with a significant separation between them and the Gators.
This means the key game on the schedule will be the rescheduled visit to Baton Rouge to take on LSU. Butch Jones was right to complain about the game needing to be made up. It’s going to be the key to Florida’s ability to get to the SEC championship game this season.
Of course, that is contingent upon Tennessee taking care of business. My previous criticisms of the Vols have focused on two things: Butch Jones and Joshua Dobbs.
Jones has won only 54 percent of his games as head coach decided by seven points or less, indicating he isn’t really giving his team a strategic advantage. Add to that Dobbs’ proclivity for turnovers (9 INTs in seven games) and it’s easy to imagine a lesser opponent taking them to the wire. Indeed, this has already happened against Appalachian State.
Florida isn’t perfect either. It must contend with significant offensive line issues. The Gators seem to average five or six false start penalties per game. But they also don’t turn the ball over as much as Tennessee, which is why I think most experts see them as more consistent.
If you made me choose, I’d go with the Gators. This is both because I’m an enormous homer but also because I think they’re going to be ticked off going into Baton Rouge after listening to the Clay Travises of the world accuse them of dodging LSU.
Regardless, it’s going to be a fun second half.