So in the last month Obama has actually lost ground in Colorado, which is one of the states that he needs to flip to be able to win without Ohio. Third party support is impressively high at 8%, and 7% are undecided. 50% believe Obama to be too inexperienced, and his favs have dropped to just 50%. The good news for him is that he has a strong lead among independents (14% of independents prefer a third party candidate) and he does get 77% among Democrats, but the inexperience question reveals some dangers for him: 25% of Democrats think he is too inexperienced, as do 41% of independents. If those who take this view and those who aren’t sure either way swing to McCain, Colorado slips out of reach.
Rasmussen also shows Obama still trailing in Florida and basically tied in Ohio. Someone’s model of likely voters–Rasmussen or Quinnipiac’s–is badly off.
conradg 
I wouldn’t put much stock in any single poll. They have been so far off most of this season it’s a little absurd. Now we have McCain up in Georgia by only +!. WTF? In what universe is Obama losing ground in Colorado and gaining hugely in Georgia?
Well, “likely votes” is certainly a huge part of the problem, though not the only one. Voter turnout in the Dem primaries was unprecedented, and if the general election follows that pattern, what with Obama’s huge voter turnout organization, this won’t even be close. The good thing about all these conflicting polls is that they require pundits to actually THINK and reason about this election, rather than just slavishly quote certain polls while ignoring others and then fit them to their pet theory.