Statisticians Needed

[Goldeneye]

My statistics teacher said today, that the supreme court uses z-scores as criteria for determining whether or not to hear a case (assuming there is a reason to use statistics in the case). They won't even look at a case where the z-score is under 2. I think we should look at each state individually and find the average and standard deviation of the difference between past exit polls and the actual outcome. We can then (for each state) find the probability that the outcome of this years election had of occurring. For most states it should be fine, but for Ohio, Florida and other "special" states this years election results should have a high z-score and thus a low probability of occurring. If this is the case, and if it is a good enough reason for the supreme court to consider looking at it further, it is good enough for me too.

[Ranec]

1. More info about the exit polls is necessary.

In the end, nobody has access to sufficiently detailed information about the exit polls to do the sort of analysis you are suggesting.
The organization that did the research has locked down all the internals of their polling.

This analysis is the best I've seen

http://www.ilcaonline.org/freeman.pdf

He says that something is definitely wrong, but there is no way to know how without more info.

Well, thats I how I see it anyway.