So, it sounds like early voters are factored into the overall numbers. The fact that early voters are swung so hard toward Obama gives me a great deal of pause about early voting. Seems to me the early voting could create way too many fraud opportunities.
They’re using the poll to count early voters. If you look at party affiliation among early voters and project, it’s about 56/43/1, with Obama 56, Romney 43, Other 1.
Note: this doesn’t mean, by itself, that Romney is leading. It’s the D+8 that points that way. If you assume D37 R36 I/O 27 for likely voters, you get about Romney 52.3 Obama 46.5 and Other (more Green than Libertarian in 2012) 1.2. But keep working like it’s a 1 point race in Ohio, because the margin of error/fraud must also be considered.
Tell me how those exit polls worked in 2004.