When they were good

You would think that over the course of a season that there should be some sort of broad correlation between a team’s statistics and its record. Notice that I’m not talking about a causal relationship between these, nor am I trying to analyze this in any narrower context. All I’m saying is that I’d expect a highly ranked team to generate good stats.

So, how true is it? To get some idea, I rambled over here: NCAA rankings.  The NCAA tracks individual teams in seventeen separate statistical categories.  Here’s Georgia’s story.  The categories comprise a pretty big breadbasket that includes performances on offense, defense and special teams.

What I’ve done is look at the top 10 teams in the final polls and counted the number of NCAA statistical categories in which each school placed a finish in the top ten and top twenty percentiles (with 119 D-1 teams, I’ve made each ten percent equal 12) . Additionally, I’ve tried to filter this through strength of schedule, so that there would be better context for the rankings. I used two SOS lists: Steele strength of schedule (which doesn’t include the postseason) and GBE strength of schedule (which does).

Here’s what I came up with.

  • LSU (Steele #35, GBE #13): 7 top 10; 1 top 20
  • Georgia (Steele #41, GBE #7): 1 top 10; 7 top 20
  • Southern California (Steele #54, GBE #83): 6 top 10; 2 top 20
  • Missouri (Steele #15; GBE #28): 4 top 10; 2 top 20
  • Ohio State (Steele #51, GBE #54): 8 top 10; 0 top 20
  • West Virginia (Steele #53, GBE #41): 7 top 10; 6 top 20
  • Kansas (Steele #101, GBE #80): 10 top 10; 1 top 20
  • Oklahoma (Steele #59, GBE #42): 4 top 10; 5 top 20
  • Virginia Tech (Steele #36, GBE #12): 6 top 10; 1 top 20
  • Texas (Steele #56, GBE #59): 1 top 10; 3 top 20

As a point of comparison, 8-5 Kentucky (Steele #19, GBE #8) had zero appearances in the top 10 percentile of any of the seventeen categories with four in the top 20 and 6-6 South Carolina (Steele #7, GBE #3) had two in the top 10 and zero in the top 20.

Again, I wasn’t looking for anything more than a correlation, so I wouldn’t read too much into these numbers, but I do think it’s credible to argue that West Virginia, with a strength of schedule that’s as good or better than five other schools on that list and being top 10 or top 2o in a whopping thirteen out of seventeen categories (only one other school had at least ten) deserves a little more love.  And Texas clearly looks to have the weakest resume on the list.  But, yeah, bottom line, good teams have good stats.

About these ads

6 Comments

Filed under Stats Geek!

6 responses to “When they were good

  1. kckd

    I like Steele, but there is something really screwed up with how he does SOS. I’ve seen Mizzou’s schedule and outside of OU and ILL., there is no one on it worth a hooey.

  2. Kansas wound up in the top 10.

  3. Hobnail_Boot

    So.. that says that while UGA didn’t do much at at elite level, it did most things at a very high level.

  4. JasonC

    SOSs?
    Obviously Steele and the other guy use different formulas, but I found it interesting that LSU had a tougher schedule that UGA before the bowls per Steele, but in the other guy’s rankings UGA had a tougher schedule post-bowls. I know you can’t make this correlation because it is 2 different ranking systems, but UGA’s schedule got tougher after playing Hawaii, than LSU’s did playing tOSU.
    Also, not that looking at the 2 different SOSs, that tOSU’s schedule got “easier” after playing LSU.
    Like I said, trying to connect them is faulty reason, but it is a little funny to look at it.

    I agree with Hobnail’s comment and expect that to change over the next season as the defense progresses from really good to (hopefully) lights-out dominant.

  5. It would be kind of neat to see what the stat rankings would look like for the last seven games of the season.

  6. Pingback: Signs of success « Get The Picture