http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=457571&mesg_id=457723The numbers need to be taken with a grain of salt, as they are subject to rounding errors, but they give a ballpark idea of the relative discrepancy between the close-of-poll cross-tabs (which probably reflect estimates made on the basis of exit poll responses, telephone polls of absentee/early voters, and pre-election estimates, but not vote-returns, because there wouldn't be any at that stage)
Contrary to posts elsewhere, Montana and Virginia don't look especially anonymous - Virginia not, although OTOH's download seems to differ from TIAs, and Montana's less than some other Senate races, including Nebraska, which was anomalous in the other direction.
It is, of course, all consistent with some kind of participation bias, which we would certainly expect, give the recent poll showing that Republicans were less likely to indicate that they would be willing to participate in an exit poll than Democratic voters. So I'd (yet again) strongly suggest that these exit polls tell you very little about where fraud might have occurred, if at all, or its likely magnitude.
But this is NOT a plea to stop looking - it's a plea to stop chasing after spurious inferences from exit polls and to start looking for hard evidence, of which there is a disturbingly large supply.