Reply #183: why are we rewinding to the spring of 2005?

[OnTheOtherHand]

I actually helped to produce that report, so I feel that I have standing to be unimpressed.

Say we have two alternative hypotheses: "the exit polls were right and the results were wrong" or "the exit polls were wrong and the results were right." (It's not that simple: I'm sure both the exit polls and the results were at least somewhat wrong.) One way of trying to sort this out is to ask: which makes more sense, the results or the exit polls? For instance, does it make more sense that Kerry won New Hampshire by double digits (exit polls), or that NH was very close (the results)?