The top pick win percentage is still better than the bottom in general, but it's an issue of sample size and inconsistency. The way I was betting, I was maxing out (to a point) on the top picks - so one bad beat week, and it could sink the season. I went back to the 'ole Modern Portfolio Theory strategy of diversification lowers volatility. The backtests bear it out so far - no crash and burns, no huge deviations... and also no huge wins. So while 2007 was a stellar year with the top pick model, 2009 sucked balls. Now 2007 and 2009 are largely comparable, which means more quan to spread around year over year.
I'll be sparse with commentary this year again, just too busy. If I get the time, I'll throw down, though. Without further ado, here are this week's picks. NOTE: I NEVER bet on weeks 1-2 (or week 17), unless I'm doing it just for fun. There's money to be had, I know the sharps love the first 2 weeks because Vegas is more inefficient, but it's just too volatile for me. So I'm not betting, but the system makes picks anyway (just without bet amounts).
And the rankings, which are a 100% mirror of Aaron Schatz and the team at FootballOutsider.com for the first week (until I get more data). Schatz is basically my go-to source for solid data - I alter it during the season, and my picks are often very different than theirs, but they are awesome if you haven't read them before: