COMMENT ON EXIT POLLS

This is elevated from the comments to an earlier post about FISHY EXIT POLLS
The following from Michelle:

I heard a guy on MPR a couple of weeks ago who was an exit poll designer.
He said the discrepancy was most likely due to a design error in the model
of the exit poll, which placed too much emphasis on a younger voter turn
out, a turn out that would likely favor Kerry. The young voter turn out was
quite a bit smaller than the exit poll model had predicted -- thus
explaining why Kerry was always favored in the exit poll and Bush always had
the advantage in the actual vote. I don't know if this is correct or not,
but it sounded (sadly) plausible. I have also heard that although Kerry
conceded, it is not legally binding, and he and his team of lawyers are
working fast and furiously behind the scenes to see if they can assemble a
legitimate case of enough voter fraud to turn the election. They would wait
to announce it until after the holidays, choosing wisely not to open
themselves to claims of "disrupting the nation while it is engaged at war,"
with all the big Fallujah battles. I don't know. But it seems a smart way
to work, if true.
-Michelle

The part about "concession" is true. There's nothing in the Constitution about concessions. However, what Kerry's very quick concession did do was take the eyes and noses of CorpMedia off of the vote count and the election itself. It's all done as far as they're concerned.

As far as the exit polls. This still doesn't explain why the polls were so dramatically off in only specific counties. And these were counties with the newest kinds of voting machines, none of which have paper trails. If the model, emphasizing the word "the" in the singular, it would have been consistently off across the board. And it wasn't. In most areas it was spot on, but those areas were not hotly contested.