tag:blogger.com,1999:blog-38600807.post6231440189546970109..comments2008-10-30T17:55:47.720-04:00Brian Burkenoreply@blogger.comBlogger6125tag:blogger.com,1999:blog-38600807.post-16089759924803864872008-10-30T17:55:00.000-04:002008-10-30T17:55:00.000-04:00I count a 36-33 record ATS for this season, around 52.2%, due to an inordinate amount of favorites not covering the spread (something around 75% of winning favorites have covered, down from the typical 82.7%).Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-56822169772546445962008-10-27T07:31:00.000-04:002008-10-27T07:31:00.000-04:00Wow Brian - I owe a rise from 22nd to 15th in my pick'em pool to you. I used your probabilities here in a straight up pick'em pool this week (with the exception of the sd-no game) and ended up going 10-3 with tennessee playing tonight.<BR/><BR/>Muchas gracias!Doug Waltershttp://www.blogger.com/profile/03064547652967055670noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-9383064586852082782008-10-23T14:17:00.000-04:002008-10-23T14:17:00.000-04:00Yeah, I was leaning on my first explanation too. Romo might be a bit overrated, but he's definitely an upgrade over Johnson, and the QB matters a lot, if not quite as much as is often supposed in the media. A significant difference between your ratings and the public's (even the betting public's) perception early in the year has got to be last year's performance. So I suspect that Dallas was getting too much credit for staying healthy last year and not enough for having a poor secondary this year.Dave Peernoreply@blogger.comtag:blogger.com,1999:blog-38600807.post-49369192750286619202008-10-23T12:32:00.000-04:002008-10-23T12:32:00.000-04:00Jeremiah-Good catch. Thanks. I had forgotten all about the London game. You're right, the Chargers would now likely have the small edge. I'll update the numbers this evening. But it's basically a toss up either way.<BR/><BR/>DP-Good point. My feel is that the QB does matter. I think explanation #1 is very true. I think the QB matters a lot, and Johnson is a significant downgrade from Romo (but the Cowboys have other problems too). Dallas was generally overvalued, and still are to some degree. I suppose I could manually de-tune the off pass eff stat for them by half a YPA or so to account for the QB change, but then I'd be just guessing.Brian Burkehttp://www.blogger.com/profile/12371470711365236987noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-45163595848036021112008-10-23T12:21:00.000-04:002008-10-23T12:21:00.000-04:00Since the Chargers-Saints game is being played in London, shouldn't the Chargers have the higher probability of winning the game on a neutral field (based on the team rating tables)?Jeremiahhttp://www.blogger.com/profile/03129956887081162676noreply@blogger.comtag:blogger.com,1999:blog-38600807.post-8645389945112375922008-10-23T12:16:00.000-04:002008-10-23T12:16:00.000-04:00I find it interesting that you and Vegas agree on the Cowboys' prospects (The Greek's true prob. for them is about 44%). Vegas accounts for the fact that Brad Johnson is playing instead of Tony Romo; you - I presume - don't, save that Johnson played in place of Romo last week. Three explanations that I can think of:<BR/><BR/>1) Vegas (or the betting population, whoever) overvalued Dallas compared to your metrics in the past, and now, with a Brad Johnson discount, they've caught up to you<BR/>2) Romo isn't that good, so they're expected to play as well this week with Johnson as they have in the past with Romo (mostly), which wasn't THAT good (0.59 GWP)<BR/>3) The QB doesn't matter that much, regardless of the gap between Johnson and Romo<BR/><BR/>Thoughts? I assumed that a significant change in personnel is one factor that could undermine the predictive ability of your model.Dave Peernoreply@blogger.com