Friday, December 10, 2004

Football Picks -- Week 14: The fact that I went 5-1 last week is less remarkable than the statistical anomaly (it must be) that with the except of the one "on-tilt" bet I made in Vegas, I have made exactly one wrong pick each week. No matter how many games I have picked it has been exactly one wrong and the rest right. Averaging one wrong would be understandable, but never varying is truly bizarre.

Having said that I have now jinxed myself I'm sure.

Newbies, should see this page for an overview of the process and summary of the data used.

Our old standby of picking the favorite when they are 40+ percentage points superior in DVOA but favored by a touchdown or less fails us this week. There are no such situations. The closest situation we have is Philly 58.3 better in DVOA and favored by 9.5. You have to wonder if there isn't a point where superior DVOA doesn’t translate well into a larger margin of victory. Is a team that is 58 DVOA points superior more likely to win by 10 or more than a team that is, say, 42 points better. I don’t know. I am going to take Philly on this for two reasons. 1) I think there is a big difference between less than 10 points and 10 or more. I don’t think a touchdown and a field goal is an unreasonable margin of victory for any game, even when you take into account potential garbage time minor comebacks. If you opponent gets in striking range of ten points, it is no longer garbage time; 2) Looking at Philly's blow out of Green Bay last week, they are clearly motivated to win by a lot of points, and they certainly will be able to against Washington.

Next up we look for games where the spread is 7 or more, but there is not all that much difference in DVOA or it is not in (subjective) proportion to the spread. There are six such games that catch the eye:
  • Buffalo -12 over Cleveland, BUF + 18.6 in DVOA

  • Dallas -7 over New Orleans, DAL +2.0 in DVOA

  • Atlanta -8 over Oakland, ATL +17.3 in DVOA

  • Minnesota -7 over Seattle, MIN +6.9 in DVOA

  • Green Bay -9.5 over Detroit, GB +4.0 in DVOA

  • Arizona -7 over San Francisco, ARZ + 18.3 in DVOA
Picking underdogs here is complicated by the fact that, with the exception of Dallas, all the favorites are at home. I just can't pick Seattle or Oakland. Seattle has been spiraling down since early in the season, and Atlanta at home is too scary a concept to cross for the Raiders. After last weeks win and given the pathetic state of New Orleans, I would be more tempted to pick Dallas here, but that's way to small a difference in DVOA to justfy 7 points on the road.

I will take Cleveland and 12 simply because 12 is an awful lot of points. I will also take San Fran over Arizona because both these teams are so incredibly bad that a close game is likely. Now, what to do about the Lions? It's tempting to take the Lions but who knows. If Favre comes out with a chip on his shoulder in Lambeau the Lions are dead. He might after last week's loss, so I'm going to pass on this.

Our last check is for teams that are superior in DVOA but are even money or the favorite has the lower DVOA. There is only one such game (which suggests the spread is catching up to the DVOA, or vice versa). Kansas City is 23.9 points superior in DVOA but is getting a point and a half over Tennessee. That's too good to pass up, I'll take KC.

  • Philly -9.5 over Washington

  • Cleveland +12 over Buffalo

  • San Francisco +7 over Arizona

  • Kansas City + 1.5 over Tennessee
This is about the least confident I have been in any of my picks this year. Too bad I can’t get odds that my one loss streak is broken.