clock menu more-arrow no yes mobile

Filed under:

2007: Pythagorean Wins Revisited

If you have read my writings for any considerable amount of time, you are probably familiar with the concept of Pythagorean Wins. If not, though, click here for a quick refresher on the concept.

After I ran the Pythagorean Projections for the 2006 season, five teams stood out. Florida, Kentucky, and Auburn were all overachievers, and thus we should have expected a decline in performance, and Vanderbilt and South Carolina where underachievers, thus we should have expected an increase in their performance. So, given those projections, let's review and see how those teams really performed in 2007 given to what was expected from their 2006 Pythagorean Projections.

Florida: The Gators won the SEC and national championship in 2006 after going 12-1, and 7-1 in conference play, but did so despite having only 5.55 Pythagorean Wins in conference play. Coming into the 2007 season, the big question mark was Tim Tebow and the Florida defense that returned only two starters. Despite Tebow winning the Heisman and having one of the greatest years in the history of college football, Florida could not come near replicating its successes of 2006. They fell to 9-3 overall, 5-3 in conference play, and finished third in the SEC East.

Auburn: The Tigers went 11-2 in 2006, and finished second in the SEC West (6-2), despite only having 4.92 Pythagorean Wins. They won every close game they played in, and the two losses came as blowouts. A decline in performance should have been rationally expected, and I went on record saying that they would experience their worst season since 2003. No surprise, that is exactly what happened, as Auburn fell to 8-4 and 5-3 in SEC play.

Kentucky: The Wildcats were a relatively big overachiever in 2006, going 4-4 in conference play despite only having 2.9 Pythagorean Wins. In 2007 they returned Andre Woodson and a host of other key skill position players, but they actually fell to 3-5 in conference play. The overall record (8-5) will be the same if they beat a Florida State team down 34 players in the Music City Bowl, but the fact that they are possibly going to remain at the same level on the overall record is a testament mainly to the collapse of Louisville and Florida State. Despite returning the best Kentucky quarterback since the days of Bear Bryant, the Wildcats couldn't improve, nor could they meet the standards they set a year ago. The only positive difference between this year and last is that they somehow, someway, beat LSU. Aside from that single game, all signs point to a decline.

Vanderbilt: The 'Dores were very much in the SEC cellar in 2006, but nevertheless was an underachiever at 1-7 with 2.02 Pythagorean Wins. Improvement was to be expected, and it came, albeit only slightly. Vanderbilt improved to 5-7, 2-6 in the SEC, and came this close to making their first bowl game in over twenty years.

South Carolina: The Gamecocks were the only team to have bucked their prediction of an improvement in performance in 2007. South Carolina only went 3-5 in SEC play in 2006, but had 4.37 Pythagorean Wins. Improvement was to be expected, and it was seemingly coming after a hot start, but nevertheless they collapsed late. Granted their conference record didn't regress, they were 3-5 in both years, but their overall record fell from 8-5 to 6-6. That said, however, South Carolina arguably fits right in. The teams that buck their prediction generally have a lot of things fall right into place, or have a lot of things go wrong, and South Carolina was certainly an example of the latter. Their head coach is arguably well past his prime, they had an incredibly tough schedule (according to the Sagarin Ratings, the eighth toughest schedule in the country and the toughest in the SEC), turnover at the quarterback position, and several key injuries on defense.

So, all told, with one notable exception, Pythagorean Projections pretty much pegged things again in 2007.

I'll have some more stuff coming on Pythagorean Wins in the next couple of days.