CO-Quinnipiac: Hick down 10

[eric82oslo]

The exit polls in 2010 in Colorado showed a party ID breakdown of

Dem 33%
GOP 28%
Ind 39%

While this poll shows

Dem 27%
GOP 34%
Ind 33%

Plus this poll is 8% Hispanic compared to 12% in 2010. So even assuming 2014=2010(no), this poll is ridiculous as is Quinn P.'s Iowa poll.

Thanks for pointing this out. That means that this sample is tweaked artificially by somewhere between 12% and 15% in favour of Republicans. In other words, this poll still shows a Hickenlooper lead.

[eric82oslo]

You know that re-weighting polls by party ID is what those "unskewing polls" guys were doing in 2012. I'd suggest not "unskewing polls;" it rather reeks of desperation.

Yet 2010 was the best possible scenario Republicans have seen for decades. It was an even greater cycle than the grand 1994. When a pollster expect Republican turnout to be much higher and Democratic turnout to be much lower than in 2010, some, or rather many, alarm bells should be going off. Especially considering the fact that demographic changes have not made life easier for Republican candidates since 2010 - and that is especially true for Colorado (Georgia & North Carolina are other examples). Of course for some other states - say West Virginia, Kentucky, Arkansas, Louisiana, Iowa, New Hampshire - demographic changes are not really important.

[eric82oslo]

You know that re-weighting polls by party ID is what those "unskewing polls" guys were doing in 2012. I'd suggest not "unskewing polls;" it rather reeks of desperation.

Yet 2010 was the best possible scenario Republicans have seen for decades. It was an even greater cycle than the grand 1994. When a pollster expect Republican turnout to be much higher and Democratic turnout to be much lower than in 2010, some, or rather many, alarm bells should be going off. Especially considering the fact that demographic changes have not made life easier for Republican candidates since 2010 - and that is especially true for Colorado (Georgia & North Carolina are other examples). Of course for some other states - say West Virginia, Kentucky, Arkansas, Louisiana, Iowa, New Hampshire - demographic changes are not really important.

And the "unskewing polls" guys were weighting party ID to 2008, claiming that it was the "best-case scenario" for Democrats.

They were actually right, even; the 2012 exit polls were similar to 2008 in terms of party ID, and the phone pollsters did have samples with a greater proportion of self-reported Democrats than the exit polls. But the "unskewed polls" were still hilariously wrong. People report their party ID differently in exit polls than they do in phone polls.

The reason why the unskewed 2012 polls were hilariously wrong was of course the very late surge Obama got after the New Jersey/New York hurricane, which showed Obama as a great national leader in times of disaster - contrasting as crazy to the incompetent W preceding him - as well as the heartfelt hugging of Christie, which didn't exactly hurt either. Well, not only that, as his final debate performance - especially his comparison of Romney's fantasy military to the Civil War era - was a performance of utter brilliance, one that will go down in the history books as a prime example of how to do presidential debates, much like Reagan's statement that he wouldn't use his opponent's inexperience against him. Before these two events, Romney had lead most polls for about a month or so - or at worst being tied. Basically every poll showed a Romney lead in Florida up until election day. Even Nate predicted Florida would ultimately go ever so barely for Romney, didn't he?

[eric82oslo]

You know that re-weighting polls by party ID is what those "unskewing polls" guys were doing in 2012. I'd suggest not "unskewing polls;" it rather reeks of desperation.

Yet 2010 was the best possible scenario Republicans have seen for decades. It was an even greater cycle than the grand 1994. When a pollster expect Republican turnout to be much higher and Democratic turnout to be much lower than in 2010, some, or rather many, alarm bells should be going off. Especially considering the fact that demographic changes have not made life easier for Republican candidates since 2010 - and that is especially true for Colorado (Georgia & North Carolina are other examples). Of course for some other states - say West Virginia, Kentucky, Arkansas, Louisiana, Iowa, New Hampshire - demographic changes are not really important.

And the "unskewing polls" guys were weighting party ID to 2008, claiming that it was the "best-case scenario" for Democrats.

They were actually right, even; the 2012 exit polls were similar to 2008 in terms of party ID, and the phone pollsters did have samples with a greater proportion of self-reported Democrats than the exit polls. But the "unskewed polls" were still hilariously wrong. People report their party ID differently in exit polls than they do in phone polls.

The reason why the unskewed 2012 polls were hilariously wrong was of course the very late surge Obama got after the New Jersey/New York hurricane, which showed Obama as a great national leader in times of disaster - contrasting as crazy to the incompetent W preceding him - as well as the heartfelt hugging of Christie, which didn't exactly hurt either. Well, not only that, as his final debate performance - especially his comparison of Romney's fantasy military to the Civil War era - was a performance of utter brilliance, one that will go down in the history books as a prime example of how to do presidential debates, much like Reagan's statement that he wouldn't use his opponent's inexperience against him. Before these two events, Romney had lead most polls for about a month or so - or at worst being tied. Basically every poll showed a Romney lead in Florida up until election day. Even Nate predicted Florida would ultimately go ever so barely for Romney, didn't he?

Nate actually said Florida was 50-50. However, he said the momentum was with Obama, so gave him a 50.01% chance of victory and Romney a 49.99% chance of victory. He did the same thing, flipped around, with Missouri in 2008.

Yeah, I think I remember 538 shifting Florida to barely Obama the last 2-3 days or so. Yet almost everyone else expected Obama to be near chanceless there.