ANOTHER ELECTION SIMULATOR
Prof. Sam Wang of Princeton is also running a simulation and agrees with TruthIsAll that as of today, John Kerry has a very high probability of winning the election. He pegs the number at 98%. Wow, that is big, but he acknowledges that it is just a snapshot. I'll be interested to see his numbers on Monday, November 1.

But I have problems with some of his calculated probabilities, and his description doesn't make it clear (to me) of how he did it. The great thing about statistics, is if specified correctly, a model like this can help to eliminate the bias of the person analyzing the race. The bad thing is an incorrectly formed methodology that presents probabilities that are unreasonable to anyone actually following the race and studying the dynamics beyond the straight polling numbers.

For instance, Prof. Wang constructs probabilities for Kerry wins in PA (89%), AZ (0%), FL (98%), MN (100%), OH (26%), NH (96%). No matter how many simulations he runs, if these are unreasonable probabilities, then his results are unreasonable. While I think that a probability of Kerry winning PA of 89% is not unreasonable on its face, I do think that a probability of him winning Florida of 98% or of winning Arizona of 0% is prima facie unreasonable.

Here are my probabilities for a Kerry win for the above listed states (in my still being beta tested simulation). PA (67%), AZ (38%), FL (54%), MN (73%), OH (44%), NH (53%). Which looks more reasonable?

Thanks to commentor "BushNoMore" for pointing me to Prof. Wang's site.