It's always a good day when the Pomeroy's are released. This is the most heavily utilized site on the web for stats geeks and gamblers. But – as with any ranking system (see AP and USA Today having identical top 10s) – there are flaws. Here is how KP describes his.
These rankings might not match what you’ve seen in any other venue. The uniqueness is due to two general reasons. First, my computer doesn’t see everything humans see, and for the most part, humans have an advantage here. I generally think humans do a good job of assimilating data this time of year, with perhaps the exception of overvaluing a long tournament run fueled by close wins or a favorable draw. Even then, it’s just a hunch on my part that people overvalue that. I could be the one undervaluing postseason performance.
These – again, like all rankings – will become more valuable as the season wears on. And by mid-season there's nothing even close (besides Vegas).
(As an aside, the best value is everything that is hidden behind that front page. The stats are impressive. Just sign up and give the guy your $20).
Here are a few things that jump out:
– while nearly everyone has NC State in their top 5, I've repeatedly gone on record stating that they're a couple of bad bounces from being a 4-5 seed, rather than a team competing for a No. 1. I have NC State at 15, and Pomeroy is even less bullish at No. 24.
– the Wisconsin effect is back. Pomeroy has them at No. 5. Last year they finished at No. 5, though to be fair, he has stated that his model overvalues them.
– UK is No. 1.
– he predicts that Ohio State and Louisville will have the best two defenses in the nation for the 2nd straight season.
– the Big Ten is once again the best overall conference.
Don't agree? Want to vent? Well, he's inviting it.
Tomorrow, I’ll have some comments on particular teams and devote a few words to Dan Hanner’s awesome new (and more sophisticated) system. In the meantime, if you need to vent you can flame away on twitter to @kenpomeroy.