Replies

I like Rasmussen as a pollster and he is generally more accurate than others because he does likely voters and generally comes up with a sample whose party identification percentages are closer to the anticipated electorate in November.

But you neglect two factors in your extensive analysis. (1) Those who tell Ras (or, more precisely, his robo-call apparatus) that they are undecided are much more likely to vote Romney than Obama come November, because merely saying you're undecided is actually a vote against re-electing the incumbent. Thus the axiom that undecideds generally break against the incumbent in any race, especially this one where the incumbent is very well known and consistently has job approval or favorability ratings in the 40s. (2) The same goes for those who say they favor a third party candidate, many of whom say that out of annoyance with the robo-call poll. Third party candidate support almost always tails off as the election approaches, and the majority of their supporters eventually go with one of the major candidates, feeling that they shouldn't "waste" their vote. (Do you really think that third party candidates will get the 6% of the vote in Ohio, for example, that Ras shows today? No way!) These respondents who drift away from their third party candidate as the election approaches also tend to go with the challenger over the incumbent.

Moral of the story: don't read Ras so literally as indicator of the November result. Those "hidden" undecideds and third party candidate supporters will give Romney the extra few points he needs in many of the swing states to achieve a comfortable enough plurality in the popular vote (which he even showing now, nationally, in Ras) and with that an almost inevitable majority in the electoral college - barring an Obama fraud factor to a degree never before seen in a presidential election.

If you don't believe me, ask Dickie Morris. (LOL!)