...and for the life of me, I can't come up with one statistical formula that separates the two teams, who are (if you've been living under a rock or just really don't give a crap) the University of Texas Longhorns and the University of Southern California Trojans. The game starts in about an hour, at which time the Men of Troy will be attempting to secure their third straight National Championship, and to do it, they're putting two Heisman Trophy winning teammates on the field.
Those last two things, they've never happened before.
My first stab was StatNorm, the College Football Statistical Normalizer, which, through a simple arithmetic formula, tries to get a picture of not just what a team's defense and offense has done, but what they've done compared with the other talent their opponents faced...
The Longhorn offense came out +169/+26, and the defense -89/-12... the Trojan offense was +202/+22, the defense -91/-11. The numbers themselves don't matter... what matters is that the average percentage of separation between the two teams in this system was something around 10% (when looking at just their defenses, this drops to 5%). That's essentially statistically irrelevant in this system, near as I can tell.
My second attempt, I've dubbed PLUMPS or the Plus-Minus Projection System. I admit, it's an extremely rudimentary system of "negatives" and "positives" in certain game situations, with results then averaged for a per-game number. I think I came up with it at lunch or something. Anyway, after running USC and Texas's seasons through PLUMPS, I came out with USC +6.92, Texas +6.75.
The numbers themselves don't mean anything, I know... but the difference is a measly 2.5%. What does this all mean?
I think it means that football is not as good a sport for stathounds like myself as baseball is.
Anyway, so what if I haven't found a statistically viable way to pick a winner in tonight's National Championship game? It's okay. I'll still be able to sleep at night. I think.
Oh, my prediction is USC 34 Texas 28. In OT.
Only 57 more days until Spring Training, thank god.
February 2005 March 2005 April 2005 May 2005 June 2005 July 2005 August 2005 September 2005 October 2005 November 2005 December 2005 January 2006 February 2006 March 2006