Sunday, June 17, 2007

Postseason Odds on Baseball Prospectus vs Pinnacle

Baseball Prospectus is a good site to pick up information. Although they are mostly a pay-site, there are some quality free material for those that aren't subscribers. These three links I find interesting (I was reminded about these links from a poster on who posted the link in a thread):

Postseason Odds using statistics compiled so far this season

Postseason Odds using PECOTA - player projections formula made up by Nate Silver

Postseason Odds using ELO

The short description of ELO from is: "Their goal is very simple: to provide an objective and reliable measure by which the strength of two opponents can be compared"

For the most part, I don't care to look at their first estimate using statistics compiled so far this season. There is more information than just this season. Why not use data from last season as well? To just use this season's number so far seems like an arbitrary cutoff date. I do think that this season's numbers are a better reflection of a player's current skill than previous season, but it is also a very small sample size. I think one could do a better job by including past seasons (last year and maybe the year before as well) and weighing them appropriately. So I am left with looking at their Pecota and ELO projections since I think they are more meaningful.

I would always trust my own projections more than any of these, but I would look at them carefully if my projections differed substantially and ask myself the question: "what am I looking at differently than they are?" If I can answer that question by pinpointing to a specific reason, then I would be happy to stick to my projection. If I can't answer that question reasonably, then I would look much more carefully at how I calculated those projections and think twice or thrice about making changes.

The Yankees are always an interesting team to look at from a betting perspective. One could argue that lines are inflated on the Yankees because there is more money betting on them (probably true). BP obviously doesn't factor that in since they have nothing to do with sports betting.

One could also argue that futures lines are correctly adjusting for possible trades. The Yankees aren't afraid to go out and get one or two or even three free agents that they need down the stretch (whereas other teams may not do it for money reasons). Those potential new players are not calculated in any of BP's projections.

The postseason projections for the Yankees are shown here because it is the one team that Pinnaclesports had odds to make the playoffs. As of the morning of Sunday, June 17, here were the following estimates on the Yankees making the playoffs:

Pinnacle (mid-market): 65%
BP current stats: 30%
BP Pecota: 39%
BP ELO: 51%

Because BP does not factor in potential trades (which should favor the Yankees), I believe their estimates are too low. But at the same time, I believe Pinnacle's number is probably too high because they have a different agenda in that they are taking bets on the prop as well as on individual games.

I'm not recommending a wager. I'm just pointing out the differences.

On another note, my publisher just told me that Weighing the Odds in Sports Betting is expected at their offices on July 16. Hopefully that means it will get to the bookstores a week or two after that.

No comments: