Who will win in Colorado?

[KCDem]

Udall. Colorado polls have a history of under sampling Democrats, and now that the state uses all-mail ballots the traditionally Democratic constituencies with midterm turnout issues should vote in larger numbers. Basically, if the Democratic candidate leads, is tied with the Republican, or trails the Republican by 2-3 points, odds are they'll win.

http://en.wikipedia.org/wiki/United_States_Senate_election_in_Colorado,_2010#Polling_3

You're proving his point. No poll in the page you linked after October 2 showed Bennet leading Buck. Bennet won by 1.7.

[KCDem]

Democrats have outperformed their polling in Colorado for two or maybe three election cycles.  That doesn't make it a standard rule of performance.  If the polling shows Udall and Gardner tied, then that means that they very likely are tied.  Hickenlooper's implosion, significantly improved GOP GOTV, a highly motivated GOP electorate, a Democratic base weakened over divisions on fracking and education reform, and a generally good GOP national environment will all help Gardner win by three or four points this year.

Ok, you're actually nuts.

[KCDem]

Has anyone bothered looking at the voter samples in the polling this year in Colorado?  Pollsters do this for a living, and they know just as many of you have pointed out that they often underestimate Democratic strength in Colorado.  So most of the polling I've seen in Colorado this year has actually over sampled Democrats in polls.  There is absolutely no way that more Democrats than Republicans will vote this year in Colorado.  None.  Yet most polls have had samples with more Democrats than Republicans.  Why?  Because they don't want to mess the state up again.

Moreover--and I can speak to this as someone intimately involved with the GOP's GOTV effort this year in Colorado--the GOP will have a much improved ground game.  Republicans have actually outperformed the Democrats on the ground since 2012, even when severely outfunded in recalls and statewide ballot races.

Gardner is in a very good position, and he would not have run if he didn't feel very, very confident about this opportunity.

You will have to excuse us out-of-staters for relying on polling data from good firms as the least-unreliable predictor of the outcome, rather than anecdotal evidence from a partisan from said state. While you may feel confident regarding Gardner's chances in this race, I cannot share that same confidence so long as Udall clings to slight leads in almost all available public polling. While the possibility of Democratic bias in the polls is certainly not improbable, I cannot pretend that it is more likely than not.

Which pollsters would that be?

PPP?

Udall +1

Rassy?

Udall +1

Quinnipiac--easily the state's best pollster?

Gardner +2

That is a tied race.  Period.  If you want to use YouGov to discern how a race will go, then you are, of course, entitled to do so, but I think it would be a mistake.

And, to be clear, I am not saying that the polls have Democratic biases.  I think the polling from the standard pollsters like PPP, Rassy, and Q, have been very, very good, actually.  The Democratic oversampling is probably even necessary to compensate for the Democratic advantages spoken of here.  

But that's my point.  Even with that built-in compensation, the race is basically dead even, and Quinnipiac even has Gardner up by two.

Quinnipiac always has overestimated Republicans' standing in Colorado.