I flipped through my normal cohort of political polling sites this morning, and stood a bit flabbergasted at the changes that happened in the last 24 hours. Rasmussen and Opinion Research both put out wide polls today for several key states, and several non-key states that I none the less have interest in.
Alaska: On the 1st we were 54/35 McCain. Now we're 64/33 McCain. All this from what used to be a statistical tie. not surprising - many of the likely to vote have rallied around Palin.
Michigan: 49/45 Obama. No one's had a commanding lead in this state, but if they were looking (either side) to widen the gap, now was the time. If it's tight now, it'll be tight to the end. This is statistically a tie.
New Mexico: 49/47 McCain. I won't say this state has been solidly Obama since May, but it's definitely leaned towards him. Now, though in a statistical Tie, McCain has the edge. Considered by some a must-win state for him.
North Dakota, Montana: 55/41; 53/42 McCain. Wait, you say. These states were ever anything but shoe-ins for McCain? Yup. Up until a week ago, they were all statistical Ties. So was South Dakota. That's right, they were actually competitive. Amazing, huh? Not so much anymore. Baring something amazing, the question is now `how much will he win them (+SD) by?`
Now, how much is this driven by short-term factors? That's the 270 electoral vote question, isn't it? Some of this is no doubt due to McCain's RNC performances, but others - I think notably MT, ND and SD - are probably driven by the choice of his running mate alone. The old axiom that we vote on the top of the ticket might not be so true this-go-round.