shlishi: Any poll that assumes that the democratic turnout will be equal to 2008 is skewed. In 2008 the democrats had an especially high turnout due to much higher enthusiasm. This year the pollsters are saying that republicans have much higher enthusiasm; the 2010 midterm elections have proven that.
Additionally, 8% more independents voted for obama than for mccain. This year the polls show Romney leading by 11% in independents, a massive change that isn’t applied in most of the polls. Most state polls with Obama leading are showing a lead of a small percentage which DOES NOT include the independents, or it includes too little independents, or they have the independents leaning towards Obama like in 2008. Include the independents the right way, and Romney clearly wins all the close polls.
History also states that if an incumbent isn’t polling at least 50%, then he will most likely not win the state. (For example, if Obama leads 47-45, that means that there’s another 8% which is mostly undecideds.) That is because all the undecideds are usually undecided about voting for Romney or for some other third party candidate. They are familiar with Obama and his policies, and they wouldn’t still be undecided if they end up voting for him. So Romney wins a great percentage of the undecideds, and the race isn’t going in Obama’s direction as Nate Silver wants us to think.