Here are my updated calculations on the state of early voting in the swing states of Colorado, Florida, Iowa, Nevada, and North Carolina, Please check my prior diary (http://www.dailykos.com/...) for the approach I'm using to calculate these values, but the bottom line is that I'm applying the current percentages of support by party from the most recent five polls in each of these swing states to the early voting totals as captured in the US Elections Project site.
The interesting part of this is that allows me to estimate how well Obama needs to perform on election day in order to win the state. Below are the current values:
CO: 53.3-46.7% Obama; 47.6% needed by Obama on Election Day
FL: 49.1-50.9% Romney; 50.5% needed by Obama on Election Day
IA: 57-43% Obama; 46.9% needed by Obama on Election Day
NV: 54.3-45.7% Obama; 45.6% needed by Obama on Election Day
NC: 55.3-44.7% Obama; 46.6% needed by Obama on Election Day
A couple of points.
1) Obama has dropped a bit in FL since yesterday largely because of the new Gravis poll which has had a significant influence on the overall average support of Democrats for Obama.
2) The percentage of early vote in NV continues to grow dramatically: the current early vote represents more than half of the actual TOTAL votes cast in 2008. (This makes me think that a lot of these votes are net new voters.)
3) North Carolina's early voters also represent 43.2% of all votes cast in 2008. While polling shows a significant lead for Romney of those who haven't voted yet (i.e., 2:1 according to SurveyUSA), my sense is that these results are reflecting a lot of net new voters that could actually give that 2:1 Romney edge a real run for its money. North Carolina feels very much like it could be a real surprise win for Obama next Tuesday...we'll see how much more voting occurs over the next few days.
Another update tomorrow.