The state-by-state presidential election went EXACTLY as expected, which I guess is possibly unexpected. The polls were perfectly good (if you analyzed them rigorously), and Intrade's state markets had Obama at 364 (his likely total once the dust settles) for the last 3 weeks. My simulator's max. likelihood estimator was similarly 364 for the better part of the last month. I haven't run all the numbers yet, but there will be a final column on Intrade soon.
While a lot of people on both sides were pooh-poohing the polls and the prediction markets, I had pretty good faith based on...everything that has ever happened in recent memory...that pollsters and prediction market investors were not complete fools. As had been my thesis for quite some time, Obama had such a commanding lead in the electoral college, even giving McCain a 6 point handicap in each state would not get him to victory. The swing states were not about to swing, and were only labelled as such so the media would have something to talk about. In the end, the swing state of PA was won by 10 points, the swing state of Nevada 13, the swing state of New Mexico 14, etc.... This election hadn't been close for a long time, but it was good to see a vindication that pollsters and prediction market investors (and me) were not all complete fools.
No comments:
Post a Comment