This is definitely a “Stats Geek!” post.
Over at Statistically Speaking, Matt Melton uses a statistical tool he labels SDPI to evaluate the strength of a particular team in the context of its conference opponents. In essence, he measures a school’s points scored and points allowed against the conference average in order to determine some measure of performance.
Here’s a fleshed out example from a 2006 SEC team, Arkansas:
The mean points scored and allowed for all SEC teams in conference play (championship game not included) was 166.83 points. The standard deviation for points scored was 36.46. The standard deviation for points allowed was 33.20. Arkansas scored 221 and allowed 134 points. Arkansas’ offensive SDPI was 1.49 = ([221-166.83]/36.46). Their defensive SDPI was .99 = ([166.83-134]/33.20). Their total SDPI was 2.47 (difference due to rounding). In the 2006 SEC, that was good second, just a hair behind LSU.
Got that? It’s not a perfect formula, as Matt admits. For one thing, in a conference like the SEC, you don’t have a round robin schedule, so schedule strength is ignored in the calculations. That is a bigger factor in looking at a particular season than in evaluating a program over a period of several years, I would think.
The latter is what he’s done for the SEC in his post here, covering the BCS era (1998-2006). It’s worth a look, both for what it tells about historical trends at a particular school and also what it shows about the impact coaches have on a program.
A few things I noticed:
- Arkansas’ history under Nutt is definitely cyclical. Things build to a peak, then a cycle ends, performance tumbles and builds up again. I would assume this is talent driven.
- Over time, Tuberville has definitely built the Auburn program up. Auburn’s numbers were the best in the conference in ’04 and ’05. But the trend has been downward over the last two seasons, with the drop in ’06 being pretty substantial.
- As Matt notes, it’s not hard to tell when Spurrier left Florida.
- I’m impressed with how consistent Georgia has been under Richt. One thing the numbers indicate is how important an experienced quarterback is for Richt: Georgia’s two worst years with him as the head coach came with a redshirt freshman and a true freshman starting the lion’s share of the games.
- Many folks like to point to Kentucky’s ’06 season as being the start of a program on the rise, but its SDPI is actually consistent with its historical numbers (although it’s been climbing out of a very deep hole from a production standpoint over the last two years). The schedule is marginally tougher in ’07 – eight home games, but the ‘Cats drop Mississippi for Arkansas. But the big thing, as I’ve mentioned before, is whether UK can match its stunning +15 turnover margin from ’06 again. I doubt it.
- It’s hard to say which chart is more depressing – Vandy’s or Mississippi State’s.