OR-SurveyUSA: Obama by only 4

[Oakvale]

Rasmussen's tracker has Romney up by 7, and if we factor in a 3.9-point house effect as determined by Nate Silver (and I think Rassy has partially fixed their model since 2010), this result would still be consistent with Romney up by 3 nationally.

Yes, but all other polls have Obama winning nationwide.

All other polls weren't conducted after Obama's gay marriage announcement (which, sadly, seems to have hurt substantially).

What evidence is there for this?

EDIT: Unless I'm much mistaken there's a Rasmussen tracker, which.... no, and a Dick Morris poll, which people should be banned for posting. Not exactly a wealth of evidence, wormy.

[Oakvale]

They had me at 21% of Tea Partiers in Oregon being Democrats.

SUSA's crosstabs have always been weird, but generally the result added up is very good.

SUSA is a superior pollster to PPP in Nate Silver's rankings. It's very funny to see the PPP fanboys dismiss them.

So tell us how 21% of Oregon Democrats supposedly being tea party supporters makes you so trusting of SUSA here? Beyond simply liking the end result of course.

It doesn't. Many polls have odd crosstabs but generally correct toplines. The Siena poll showing Obama +20 in New York has some odd crosstabs, but Obama +20 in New York is a perfectly valid and reasonable result.


If anything should make you trusting of SUSA it is their superior track record, as being 2nd behind Quinnipiac in average error.

You can't necessarily judge a poll's accuraccy entirely byit's track record. Trying to his SUSA's supposedly strong record to defend the reliability of a poll claiming over a fifth of Oregon Democrats are tea party supporters is just plain wrong-headed. Sometimes even artists put out pieces of $%*t.

The poll doesn't make that claim; those polled make that claim. If you wanted to poll the percentage of Oregon Democrats that are Tea Partiers then you would need a much larger sample size.

It's supposed to be a representantive sample, though...