Steele and the Football Outsiders

I buy two football previews: the Football Outsiders Almanac and Phil Steele. As I was flipping through FOA last night and hoping that their projection of 5.8 wins for Michigan was a mistake, I decided to take out my 2009 editions of FOA and Phil Steele to see which did a better job. Then I remembered that Chris Stassen does my work for me. It turns out that Steele and FOA were both in the middle of the pack. So much for the numbers-based approach; both previews finished behind Sports Illustrated. I was unaware of the existence of CompughterRatings.com, but that site took the prize for most accurate predictions last year. (and cheesecake, too!) I suppose that Mr. Pugh’s mathematical model that might as well be in hieroglyphics to me is a step ahead of FOA’s FEI, although one year is not enough of a sample size to make that judgment. (Even a liberal arts major who needs help from his spouse to do regression analysis knows that.) If Pugh is the cream of the crop again, then that’s good news for fans of USC (Pugh has them #4; the consensus is that they’ll be #12) and Georgia (Pugh has the Dawgs at #12; the consensus is that they will be #20). Pugh is not as high as the consensus on North Carolina, Miami, and TCU.

It is amusing to me that College Football News, which is already being picked on for its factual inaccuracies du jour, has come in 12th of 17, 14th of 15, and 15th of 18 in the three years in which they show up in Stassen’s database. Maybe their confidence in Matt James stepping up for Notre Dame at left tackle will reverse the trend.

It’s interesting to me that the two previews that are most reliant on computer formulae – Football Outsiders and Pugh – are down on Nebraska and North Carolina, the two teams that are getting preseason buzz despite the fact that both had putrid offenses last year. I have to admit that I was surprised to see Steele rank the Huskers at #5. Nebraska was outgained in conference play last year. The old Phil would have jumped all over Nebraska as being potentially overrated.