[Longbow1969]

You realize that Gallup was wildly inaccurate in 2008 and Pew was nearly dead on. Gallup’s final tracking poll had Obama winning the 2008 election by 11 points and wasn’t even close. Meanwhile Pew nailed the final election results in 2004 and 2008. I’m sorry, but you shouldn’t be so quick to dismiss polls you don’t like - especially ones with a decent track record like Pew.

[trappedincanuckistan]

If the internals are off they’re off. Period. I don’t care about they’re track record. They are counting on a +6 D turnout that most other pollsters think is ridiculous. Also take into account early voting numbers. D turnout down across the board.

[MNJohnnie]

You realize Ras was everybit as accurate in 2008 and he shows a tie race.

So why should we believe a PEW poll with an absurd +6 Dem oversampled and not the every bit as historically accurate Ras poll which shows the race tied?

Let not fool ourselves by clinging to our knee jerk Eyeore opinions that lead us to believe a +6 Democrat sample in the Pew is a “real poll”.

Correct that oversampled D to a rational +4 Dem sample and you get Romney 49, 0bama 47 which is right in line with all the other polling out there

[fortheDeclaration]

I dismiss polls that are based on nonsense data, like a D+6, which isn't going to happen!

Now, M.Barone has stated that this is going to be a Romney win, but I guess he doesn't know anything about election data?

[fortheDeclaration]

As for 2008, everyone knew Obama was going to win, so the fact that they were closer to the actual numbers is really not that impressive.

How did Pew do on 2010-predict the GOP taking the House?