The end of Rasmussen fetishism.

[Tetragrammaton]

Elections are the most fun thing to cover/read about. Such high stakes and high drama.

[COphinphan89]

I had Hannity (I know) on in my car I think it was the day before the election and he had Rasmussen and Barone on. And after Hannity pumped up his credibility a good bit Barone was calling a landslide for Romney, but Rasmussen was pretty adamant that he had no idea who was going to win because his polling was so close even though most of it was leaning towards Romney. I think his methodology might be off. Might need to up his sample sizes.

[JamesBW43]

According to Silver, Rasmussen doesn't call back numbers that get no answer, and those are more likely to be democratic households because they go out more often. Also Rasmussen's models typically have more Republicans than there actually are.

[Locke]

According to Silver, Rasmussen doesn't call back numbers that get no answer, and those are more likely to be democratic households because they go out more often. Also Rasmussen's models typically have more Republicans than there actually are.

And there it is. Not an even spread of ideologies. As a researcher, I almost wish it was someone lying because mistakes like this are inexcusable...

[Eshlemon]

Why would it? Polling overall was more accurate in 2012 than it has been in years.

And polling gets skewered by the momentum of events...in this case Sandy and a couple Senators. 40% of undecided stated disaster handling was the key in voting for Obama and virtually all of the ground Romney made up with women unraveled.