Des Moines Register -- Obama +5 in Iowa

[J. J.]

Obama 47 Romney 42.

Think this is a pretty solid poll

http://blogs.desmoinesregister.com/dmr/index.php/2012/11/03/iowa-poll-final-stretch-in-iowa-gives-edge-to-obama/article?nclick_check=1

Newspaper polls are somewhat better than Uni polls, but not by much.  

The early voter number would indicate that Obama is carrying 80% of the independent vote, and frankly, that is hugely unlikely.  

[J. J.]

There polls shows that indies are at 41/37 Romney.

25% of the voters are independent so far, with slightly more outstanding applications.  

Assume that Romney has 40% of that indie vote and Obama 60%, these would be the early voting numbers.

Romney:  42%

Obama:   58%

O +16 points

If their independent result is correct, Romney's getting about 52% of the indie vote.  That would be:

Romney:  45%

Obama:  55%

O +10

They have early voting:

Romney:  39%

Obama:  61%

They did weigh it based on census data, and that might be a problem.

[J. J.]

When you start using crowd size to justify that your side will win, instead of hard date.....then you know you have lost.

Hard data that have incorrect turnout models.

So basically every pollster is wrong? Whatever helps you sleep at night, pal.

We know how many indies have voted and we know the requests.  Those are not guesses.  He also know if the indies are voting as the poll indicates they are, what the result would be.    It doesn't match the poll result. 

Now, no, I don't go on crowd sizes, but this is hard data. 

[J. J.]

Excellent pollster. This late, Iowa is in the bag along with Wisconsin and Nevada.

With six other states not quite certain (CO, FL, NH, NC, OH, and VA)... NH only because polling is wild...

New Hampshire doesn't matter, but President Obama gets re-elected with the combination of Colorado and New Hampshire or any one of Florida, North Carolina, Ohio, or Virginia. 

Off 8 points last time.

[J. J.]

It does seem like Republicans will attack any poll that doesn't fit their narrative. While this poll was off several points in the 2008 presidential race, it has pin-pointed many races. In any event, the Selzer poll should be considered the gold standard for Iowa.

They were off, a lot, in 2008.  So were most other polls in IA.  Also, their internal numbers make no sense. 

[J. J.]

In that one particular election, that is true. However, this poll has been very accurate in almost every other poll in the state. If anything, the only possible problem with this poll has been the overestimation of the "Other" vote.

But it was well off with Obama.  There is also the problem with its internal numbers and early voters.