Rasmussen was #1 does that mean we should trust their polls the most?

.

notice that Gallup was so low because of how much they overestimated the vote for Obama. So if you want to hammer them for having R-money up 7 today just remember they lean left.

Wildly overestimated Bush this time of year in 2000 and 2004 (11 and 4% respectively), Clinton by 6% in 1992 and 3% in '96 and 2010's GOP by a monstrous 8% in their final poll.

Gallup's Predicted Republican 2010 Generic Ballot Advantage -15%
Actual Republican 2010 Generic Ballot Advantage -6.4%

Difference = 8.6% .

FIXED

This is going to be a mindfuck for the polls. I'm not going to say "it's all good" because if it were, a clown like Romney would be 15% down, but I am confident as one can be in what is ostensibly a close race.

Look up the UK General Election of 1992. Obama won't win the PV by such an amount, but this is that election's love letter to the United States.

I'm going to go and shave another few minutes off of my life with a cigarette now.

I don't think Gallup is doing this on purpose. But they over calculated Dem support in '08 and '04 and Repub support in 2010. Their likely voter questionnaire seems to pick up those enthusiastic about the election who have voted before but discounts those who are new voters. A new voter scores very low on their likely voter model but it counts the same in a real election against a voter who has voted before.
Rasmussen was accurate in 2008 but had issues in 2010. However, their voter ID screen is now about Dem plus 6 since they adjust it every month. For whatever reason, their likely voter model only gives Romney a plus 1 to 2 lead in their current polling.

Sheesh. It is tiresome to school low count posters.

because I don't spend all my time on the message board my opinion is invalid.

here is the Gallup final polls for presidential elections since 1980

2008 Obama +2
2004 Kerry +1
2000 Gore -2
1996 Clinton +3
1992 Clinton +6
1988 Dukakis +2
1984 Mondale +1
1980 Carter +3

Only once since 1980 have they not favored democrats, and even then that 2% didnt go to Bush it was overestimated for Nadar.

So you can claim methodology but that methodology obviously favors the democrats.

They are not favoring anyone as their have been too many different people in charge of Gallup over the years to claim that. It is not like Rasmussen where we know Scott Rasmussen leans Republican.

Their likely voter model still has issues. Perhaps they need to screen for party ID or weight demographically. Their likely voter questions seem to be leading to results that make no sense when matched against the other national polls out there right now. Anyone looking at statistics and averages would say the same thing.

They also stopped polling weeks before that Truman election because he was behind so much, that doesn't happen any more. You act like its impossible for them to lean left because they favored republicans over 3 decades ago, their methodology has favored democrats for 32 years now its hard to write it off as just an error that could favor either side.

Why are you even on DU arguing with a liberal? Why not go complain about Gallup on Free Republic? I mean, you should be happy about Gallup. Why are you complaining to me?

They overestimated the Republican generic ballot advantage in 2010 by a whopping 250% However I will give them the benefit of the doubt as I subscribe to the axiom that one should "never attribute to malice what can be attributed to incompetence."

In 2008 their incompetence favored the Democrats. In 2010 their incompetence favored the Republicans. But in neither election did their incompetence favor those who just wanted an accurate portrayal of the respective races.

Although the likely voter model does seem questionable. Swing state polls are more important.