in 2008

and McCain actually came out ahead among voters in some of those states on election day voting.

We're 3 weeks away from the election still.

In Iowa in 2008 only 32% were cast early. In 2010 35% and they're estimating 40% this year. That would mean that the estimates of what % of the electorate will vote early are way way off OR that 85% of those who are going to vote early already have.

http://www.radioiowa.com/2012/10/16/pace-of-early-voting-in-iowa-off-the-charts/

so it is entirely possible.

have already been cast....with 3 weeks left to go in the election???

Doesn't seem like it's likely 34% of the electorate has already voted early based on your info

...Romney would need to win the remaining voters by 20 points. Roughly.

Kep in mind if Obama's ground game is so much more successful at mobilizing their early voters we pretty much *expect* Romney voters to be showing up on election day more and him to be winning the remainder by some amount... but 20 points seems like a stretch..

34 points are sealed now. 22.7 go to Obama. Assuming no one votes for a third party candidate:

50.1 - 22.7 = 27.4

27.4 / 66 = 41.5% of the remaining vote needed for Obama to win

Also assuming I didn't make any math mistakes.

Assume 100 votes to be had...

EV (who have already voted)


Obama has 67% of 34, which is 22.78.

Romney has 32% of 34, which is 10.88.

EV (who have yet to vote)


Obama has 55% of 11, which is 6.05.

Romney has 39% of 11, which is 4.29

Election Day Voters

Obama has 39% of 55, which is 21.45

Romney has 55% of 55, which is 30.25.

Add them up...

Obama: 50.28
Romney: 45.52

The early votes are already in the bag 2-1 for Obama. The not-yet-voted majority favoring Romney can still 1) switch preference, or 2) not get around to voting. I sense that the not-yet-voted component is too heavily weighted toward Romney, to some extent.

Iowa knows Romney very well by now. He spent millions of dollars there for the caucuses in the last two cycles and lost both times.

I do not see a dramatic change in the numbers coming.

Which is probably 19% of the total, not 34% (see my post below).

of the entire population of the state.

And about 5% of the number of pigs in the state.

When there is real data to corroborate or dismiss poll results, they should be included and the poll should be reskewed if need be.

Take, in this case, the actual numbers, current as of yesterday: http://sos.iowa.gov/elections/pdf/2012/general/absenteestats.pdf

284,000 absentee ballots have been received. There were 1,500,000 total votes in 2008. That means, 19% have voted as yet, not 34%. Sure, some may have been sent in and are pending, but nearly double the amount? No.

More likely, people are saying they have voted when they really have not, or thought they did.

In my town in NJ we usually have anywhere from a 60-80 percent turnout in a presidential election year.

So based on these numbers I guess we shoot root that it rains like Hell in Iowa on November 6

is that people think that this will get them off the call lists.

And aren't those official figures just for absentee voting?

So 34% is probly accurate