[Kingosaurus]

Yeah, I think Morris basically has the same opinion about the D-oversampled polling that a lot of people here do.

He’s got nothing to lose really. If he’s right and Romney wins with the useless polls predicting an Obama win, he looks like a clairvoyant. If he’s wrong again, it will eventually be forgotten, because he’s been wrong before and he just plows ahead as if it never happened.

[JediJones]

Morris wrote an article explaining why he doesn’t trust the polls. It has nothing to do with “sampling.” He explained how every poll has a sample skewed different ways, but the pollsters always adjust numbers based on a turnout model, predicting which groups will turn out in what proportion. And he says the 2008 model is the one the polls are using, which he thinks is wrong.

[TheRhinelander]

He’s basing it on this:

http://www.unskewedpolls.com/

Which is based on this:

http://www.rasmussenreports.com/public_content/archive/mood_of_america_archive/partisan_trends/summary_of_party_affiliation