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.
Party ID numbers from 2010 should also be taken with a big grain of salt, especially if one is claiming that they represent the "best possible scenario" for Republicans, since (due to a combination of the Tea Party and dislike of Bush) it was a bit of a fad at the time for Republican voters to say that they were independents. Even as Republicans decisively won the election, the proportion of self-reported Republicans was hovering around all-time lows.
And finally, it's pretty easy to believe that the enthusiasm gap in Colorado was more favorable to Democrats in 2010 than this election cycle. Democrats and Hispanics were a lot more fired up to vote against the Tom Tancredo and Ken Buck ticket than the Bob Beauprez and Cory Gardner one. If we extrapolated the 2013 recall results across the whole state, Hickenlooper would be losing by even more than in this poll, and he's only become more unpopular since then.
So this poll is an outlier, and I would like to see confirmation, but there's no reason to dismiss it out of hand. I think the Illinois poll showing Quinn ahead by double digits is believable as well (and in fact, even before that poll, if you put a gun to my head and asked me to pick the winner I'd have chosen Quinn).
