Originally Posted by Pseudofool
It'd be pretty remote that ALL the polls would be wrong--as they all have different models.
Early voting isn't necessarily indicative of a change in the overall electorate, it's hard to know what to do with those numbers other than see that more Republican counties are voting earlier (there's still a limited number of votes possible in such counties). What's important are those who are changing there vote from Obama in 2008 to Romney 2012, and early voting gives us no indication of any evidence of that. It'd be a mistake to suggest that early voting debunks the polls.
This is wishful thinking.
You don't think a difference in early voting trends from 4 years ago (D's down 4%, R's up 14%, total swing of 18%...) isn't huge? Wow.