Even more SEC stats

Mergz, at Saurian Sagacity, puts his recruiting/results analysis on the line with this prediction for the upcoming SEC season:

His initial projection showed Florida with 14 wins, Tennessee and LSU with 12 wins, Georgia with 11 wins and Auburn with 10 wins. He recognizes that it’s unlikely that scenario would play out in reality, so the chart above is his “modified predictor” for the SEC regular season. What he doesn’t state in his post is what formula he used to make those modifications.

Meanwhile, over at Statistically Speaking, there’s a post up that takes a look at the 2006 SEC season from the perspective of teams points scored and allowed relative to the league average and standard deviation. Taking into consideration that his formula does not adjust for schedule strength for conferences where each team does not play each other and also ignores special teams, there’s still a lot of interesting information to glean from his post.

Here’s a few things I noticed:

  • As expected, LSU squandered a significant statistical advantage over the rest of the SEC. On paper, this was the team that should have emerged from the conference to play for the MNC.
  • As almost anyone with a pair of eyes could have confirmed, Florida wound up in the MNC because of its fantastic defense and because Meyer was a good enough coach to get his team through a series of squeakers.
  • Turnover margin made a huge difference for Kentucky last year.
  • We knew Georgia’s schedule was favorable last year. I just didn’t realize it would come out being as weak as it appears here.

He concludes with a look at SEC teams likely to improve (‘Bama and South Carolina) or decline (Kentucky, Arkansas and LSU). I agree with him on Kentucky (unlikely that the superb turnover margin for ’06 will be repeated), Arky (I think the offseason distractions will be a problem throughout the year, although with that schedule, wins won’t fall too much from last year’s 10) and Alabama (good talent base, weak schedule and better offensive playcalling should equal a win or two more). However, SC’s schedule is a killer and LSU has way too much talent for a falloff, even with Miles’ track record there for losing a game or two he shouldn’t.

Also, take a look at the two posts he links to near the end of his post. He suggests that having to replace a starting quarterback is more likely to be a negative for a team than a returning quarterback is a positive. If that’s the case, things at Florida and LSU could be interesting…

Comments Off on Even more SEC stats

Filed under SEC Football, Stats Geek!, The Blogosphere

Comments are closed.