/cdn.vox-cdn.com/uploads/chorus_image/image/46474824/usa-today-8514746.0.jpg)
The balance of power within the SEC East has shifted in recent years creating more uncertainty in the division. South Carolina won 33 games in three seasons before last season's underwhelming result. Tennessee and Kentucky both had head coaches that oversaw the respective programs slip, and now both have head coaches that are supervising their reconstruction. It's early, but Vanderbilt appears to be going in the opposite direction. The success of the Urban Meyer Era is a faint memory in Gainesville, while Missouri was an injection of a quality program. If one were forced to pick a SEC East program standing athwart the rising and falling of fortunes the hesitant choice would probably be Georgia, but even that feels a little weird.
Last fall's pandemonium will probably continue again this season based on the increasingly crowded middle tier of the division. If "chaos is a ladder", which teams are poised to rise above the mediocrity and ascend up a rung or two? One tool available to attempt and predict the future is the Pythagorean projection. Simply put, it takes into account the previous season's total points scored and total points scored against to determine if a team actually under- or over-achieved. Moreover, it's a strong indicator for success or failure the next season.
This is how Pythagorean projection views SEC East teams last season:
SEC East | Record | Pyth. Diff. |
Florida | 6-5 | +1.7 |
Tennessee | 6-6 | +1.0 |
Georgia | 9-3 | +0.9 |
Kentucky | 5-7 | +0.5 |
South Carolina | 6-6 | -0.5 |
Vanderbilt | 3-9 | -1.0 |
Missouri | 10-2 | -1.5 |
According to this metric, Florida should have won nearly two more games last year than it actually did. Given its 2-3 record in one score games, and the cancelled game against Idaho, that seems a logical conclusion. Tennessee also had a 2-3 record in one score games and will almost certainly improve upon its six regular season victories this season, barring injuries. The advanced stats really like Georgia this preseason, and this projection would agree they are poised to win 10 regular season SEC games which would probably have them competing for a playoff spot.
Missouri, more than any other team, looks to take a step backward based on going 4-1 in one score games last season. Meanwhile, Vanderbilt's 3 wins last season is actually considered an overachievement, and this year won't get easier with an opening game against Western Kentucky, a game at Houston, and drawing Ole Miss and Texas A&M from the SEC West.
The two teams key to making sense of the East's pecking order are Kentucky and South Carolina. Both programs appear equal in terms of this metric, and the product on the field for wide swaths of last season. It's probable that these two teams' postseason fortunes will be decided when they meet in Columbia in the second week of the season.
Adding Complexity
There are other variables to keep in mind regarding trends that may not continue this season. For example, it's hard to imagine Georgia will have a +16 turnover margin again next season, or that Vanderbilt will finish the season -16. If South Carolina and Kentucky are indeed the pivotal teams, South Carolina will probably have a better turnover margin than -2 next season, while Kentucky probably will take a step backwards from +8. These things swing wildly, as Georgia was -7 and Kentucky 0 just the season prior.
This says nothing of injuries, defensive touchdowns scored/allowed, fourth down conversion percentage, or special teams touchdowns scored/allowed. All are factors that can play a major role in an additional 1-2 wins or losses over the course of a season, but they can't be comfortably relied upon to repeat at the same rate in consecutive seasons.
Conclusion
Pythagorean projection is merely one metric, even if it tends to have a strong track record, and there are many more. It does suggest which teams are more likely to have better seasons than last year, and which ones may do even worse. In a chaotic SEC East where the gap has closed between at least 5 of the 7 programs, any shedding of light is useful.