What poll was the most accurate in 2004?

And I believe electoral-vote.com was quite accurate as well.

Real Clear Politics isn't a poll...It compiles polls...

Ras, Pew, and IBD were pretty accurate:


http://www.pollingreport.com/2004.htm

They sucked in 00...They tightened at the end to avoid looking ridiculous...

So 538's numbers already have that built-in.

note: whore not used in a sexist manner at all.

Are you a whore?

No, I'm a mathematician.

:rofl:

Yep, you don't hear that too often :D

Any serious survey that tries to not only make a descriptive map of national attitudes, but tries to predict a national election outcome... has to make some very big assumptions about voter turn-out.

That's why there's such a fuss about all the "likely voter" polls and their methodologies. They use past data and some survey questions about "excitement levels" to determine if somebody is likely to vote. Herein lies the catch-22.

:party:

Every election is different and unique... and this one is historic

Applying the same weightings from 1996, 2000, 2004... is inherently problematic. Even if they adjust based on local dynamics in the primaries, they are still making huge assumptions about turn-out among different demographics.

These people can't even agree on something as basic as a "bradley effect," yet we think they can not only accurately predict voter turn-out this november among folks under age 25 but nail their voting patterns within a statistically valid and properly weighted figure?

:)

I'm increasingly of the view that this is going to be a COMPLETE BLOWOUT. I mean... a heart-wrenching global gasp for air... true shock/awe political moment that will be a defining moment of this century. People around the world will be in tears, and not even sure which of the top ten reasons are making them cry.

I think these pollsters realize somethign crazy is going on... and "historical trends" wont help them capture it. hehehe

Also known as IBD. They poll in conjunction with TIPP. I've read that they were the most accurate polling outfit for the 2004 election. NOTE: IBD shows this as a one-point lead with Obama up 45-44. Very strange with other polls showing 10+ lead.

As for another poster, RCP is a good site but it is not a polling outfit. It averages a bunch of differnet polls, but does not do any actual polling itself. But it shows you all the polls out there.

the Fox poll was actually quite close to the final vote.

One can only hope -- they had Obama by 10 a few days ago.