Thursday, October 11, 2012

Put The Bunny Back In The Box

My thanks, per usual, to the fine Nate Silver and his FiveThirtyEight Blog for talking me off ledge today.  

The difference is that we, and HuffPost Pollster, are looking at the Electoral College and the popular vote in a holistic way. The evidence is ambiguous enough that it’s hard to know for sure, but the fact that Mr. Obama appears to hold a lead in the Electoral College is reason to be suspicious that Mr. Romney leads in the popular vote.

Two of the three hypothesis yield an Obama win. It’s something of a coincidence that our model now shows Mr. Obama with almost exactly a 2-in-3 chance of winning (as do Vegas betting lines), but it isn’t the worst way to think about the election.

Also, thank you Las Vegas!

Just wish I had read it yesterday when it was actually published.  Would have saved me 24 hours of tsuris. 


  1. Hey Schmoker,

    Just wanted to follow up on the Romer 4th down piece. While I am not an economics Phd, I do understand a little about game theory, (Think Nash Equilibrium, from "A Beautiful Mind") because I studied it a bit in school. The linch pin of game theory is that each "game" has to be equal/identical. Not identical out come, but identical game. To achieve this for his study Romer must then use an average outcome for each "game". For his purposes a "game" is considered a particular down and distance. From 30,000 ft. this may seem unflawed, but the closer you get to the ground, the obvious cracks begin to appear. In a situation where human ability is the sole determinig factor, the variablity from yr to yr, team to team and player to player, the "game", i.e (a particular down and distance) can never be the same. Whatever value he assigns to a particular down and distance, it's obvious it won't be the same for all teams in a particular season, or the same team across different seasons, because of the variance in human ability. Because baseball is a more individualized sport it is easier to see. Let's say a "game" is a particular count, with a particular number of outs and a runner on a particular base. Given a certain time frame (seasons), it would be easy to come up with a (run) value for that particular "game" (at bat scemario). But the result of any particular "game" that number can differ drastically depending on who the hitter, pitcher and fielders are and their individual abilities. So a manager would be foolish to make a decision for that particular "game", on a rule for that general "game". So as theory Romers idea might be sound, but from a practical standpoint, it is only directionally sound at best, for any particular "game" because of human variablity. If you'd like to discuss more. I can be reached at

  2. Interesting place to put this comment. Whatever flaws there are in the Romer theory, they don't include 4th and 1 on the opponent's 40 with 5 minutes left and down by a TD. So we can argue the basic theory, and whether or not the extreme examples are warranted, but the basic premise is sound when it comes to many, many, many, many more situations than what coaches currently believe is warranted.