I think that Scott Rasmussen is correct in his projection of voting by ‘party affiliation’. He hedges down as low as +2 democrat advantage. What the polling is not revealing clearly yet is the number of crossover democrats who will vote Romney. I have seen recent projections of 90%+ republican affiliated who are a firm vote for Romney. I have seen the democrat affiliated lower, in the mid-80’s for Obama.
How many Reagan democrats will there be? I believe the debates will turn this to us. Romney has very wisely spent a lot of time preparing his message. My gut tells me the first debate will be hard facts and straight talk vs. unicorns, rainbows, blame and assorted nonsense. I believe we start to see the turn after Wednesday. With these polls so close in spite of the illogical partisan samples (all but Rasmussen), Romney doesn’t have to persuade that many democrats.
Is it conceivable that he can convince 1 out of every 20 or 40 democrats to change their vote? Is it conceivable that he will take at least 3 of every 5 undecided? Is it conceivable that Ryan is going to make Biden look like such a boob in their debate that it only adds more anxiety to the fringes of Obama’s support? I don’t think it’s only conceivable, I think it’s likely. This race is Romney / Ryan’s to lose and I don’t think they are going to lose.
Check out post #7, it’s the most detail I’ve ever seen from a Rasmussen poll. It must be what you see if you’re a subscriber.