At the beginning of this election cycle, Colorado seemed like a state that Obama could not only win, but that could decide the election for the Democrats. After all, in 2008, Obama won Colorado by 9 points—more than his 7.3 point national victory—and the state gave Obama his 270th electoral vote. So it was possible to expect that Obama had an Electoral College advantage: If the states all swung uniformly in Romney's direction, Obama could lose the popular vote by as much as 1.6 points and still win the Electoral College because of Colorado.
The math on Obama's 2008 performance wasn't the only reason for Democrats to be optimistic about Colorado. Michael Bennet's upset win in the 2010 Senate race was a rare bright spot for Democrats, and his victory seemed to show that Obama could put up a strong showing in Colorado. (Bennet won by emphasizing social issues and consequently winning a decisive share of college educated women, a strategy that that the Obama campaign clearly took note of when they decided to emphasize Planned Parenthood this summer.)
And Colorado's demographics seemed to work in Obama's favor. The state is among the best educated in the country and Obama's support among college educated voters has remained at elevated levels nationally, so one might have expected that Obama would have retained an unusually large share of his '08 supporters. Even if he couldn't, a growing Latino population and the influx of migrants from California and elsewhere seemed to work in Obama's favor.
There is one catch. Over the course of the campaign, the chances of Colorado playing a pivotal role seem to have faded.
First, Obama started lagging in Iowa, a prerequisite to the "Colorado route." (The state isn't necessary for winning with Virginia, Ohio, or Florida.) Then, Obama started looking good in Ohio, a state that can essentially ensure Obama's reelection, provided Obama wins states like Wisconsin and Nevada (ones that are also probably necessary for a victory through Colorado). Still, Obama continued to seem well-positioned in Colorado. He led the overwhelming majority of polls conducted prior to the DNC, roughly in line with his lead nationally.
Post-DNC, that changed. With Obama's lead in Virginia and Ohio appearing to grow following the DNC, Obama's middling showing in recent Colorado polls may have diminished the state's importance even further. SurveyUSA, Quinnipiac/NYT/CBS, and Rasmussen show Obama leading by an average of less than one percentage point, while Obama leads by about 3 or 4 points nationally. Similarly, SurveyUSA's other state polls show Obama ahead by near-'08 margins, including in Florida. Quinnipiac/NYT/CBS also shows Obama up in Florida, and Rasmussen shows Obama doing better in Florida, Ohio, and Virginia than Colorado.
If further polls confirm a very tight race while Obama leads nationally, then Colorado might not play a particularly pivotal role in the election. It's too small to singlehandedly provide Obama with victory, and there are many states—ones where he looks stronger—that give Obama the election with fewer additional states. Of course, none of this suggests Obama can't win Colorado. After all, he leads in the most recent polls, albeit narrowly. And Colorado can still play a critical role in the electoral math, perhaps by allowing Obama to compensate for a loss in Wisconsin. But Obama's narrow margin suggests that Colorado is less likely to put Obama over the top than it was four years ago, when it offered Democrats the serious possibility of victory in the Electoral College even while losing the popular vote.