From the best resource (early voting database via CNN)
http://elections.gmu.edu/early_vote_2008.htmlLooking at the four states where there's party or race data, and substantial increase in early voting over previous years (so it's evidence of the GOTV, not of a mail-in tradition like in Oregon), looks like Obama could really piling up tremendous leads with large portions of the votes already in. I didn't realize how many early votes there are, but this looks like it could be the equivalent of the caucuses, where Obama studied how to play the game to maximum effect and surprised everyone with the near sweep.
Florida - early votes already 44% of 2004 total votes; 45.6% D vs. 38.2% R
Florida includes a lot of both in-person and absentee, where absentee includes a lot of retirees and military); 2004: early voting was 43.5% R / 40.7% D. Assuming absentees are about the same as last time, in-person early votes (where the new effort is) would be about breaking 50-35 Obama - and about 28% of the total vote from 2004.
Georgia - early votes are 53% of 2004 total votes (20% of 2004 votes were early, by the way), of these 35% black, 56.4% women vs. 40.4% men, 1% unk (which leaves 2% unaccounted for, but never mind).
Nevada - early voting in Washoe & Clark counties (about 90% of the state) about 62% of total vote in 2004, 51% D, 32% R.
North Carolina - early voting is 58.5% of 2004 total votes, 52.6% D, 29.6% R
For the last three, if we just ballpark a 20% Obama lead with 60% of votes in, that means McCain would need to win the remaining 40% by 30%, i.e., he'd need a 50% swing.
I think it really looks this effective, and it's going under the radar screen because it's kind of hard to analyze and explain. Am I getting this right? Could it really be this good?