What D/R/I model is Ras using?
Although I doubt Rasmussen is deliberately oversampling Ds, if Romney leads by 12 among Independents and is still tied with Obama then maybe Rasmussen needs to adjust his model. Morris says Ras is using a model that combines the 2008 and 2004 models, but that's still likely to D friendly because of the unusually good year Dems had in 2008.
How does his 2012 model compare to his 2004 & 2008 turnout assumptions?
He should be using the ‘Chick-fil-A’ model...