Also consistent with this scenario: the Gallup tracking poll actually moved a point in Obama's favor after Thursday's polling, but then swung two points in Romney's favor after Friday's polling. To be sure, the Gallup Poll, being a seven-day tracking poll, is even more opaque than the Rasmussen poll, and we don't know in either case what exactly the number on Thursday or Friday was. What's more, some polls, especially at the state level, seemed to show pronounced pro-Romney impact on Thursday. Still, it's notable that the two most prominent tracking polls showed no pro-Romney movement in the numbers that included Thursday's polls but showed pronounced pro-Romney movement in the numbers that included Friday's polls.
Don't get me wrong. Romney undoubtedly outdid Obama in the debate, and you'd have expected simple objective reporting of this fact to have some impact in Romney's favor. Still, I do think it mattered that much of the media went into the debate expecting roughly the opposite outcome. And, as I wrote before the debate, this expectation was simply mistaken; after predicting the emergence of a "Romney surprisingly good in debates" meme, I wrote:
This meme, like the previous one, should by all rights be DOA. The truth is that Obama is not a great debater. Four years ago Hillary was on balance more impressive than he was in the primary debates, and then in the fall debates he had the good fortune to go up against a dim and crabby John McCain. Romney, though erratic, is a much better debater than McCain and on any given night has a good chance of outshining Obama. And since Obama enters the debates overrated, and Romney enters them underrated, a tie will go to Romney, who will have "exceeded expectations."
Well, rather than a tie being inflated into a Romney win, a clear Romney win -- one that shouldn't have shocked anyone -- was inflated into Hiroshima-level devastation. And so devastation is what happened -- though, as with Hiroshima, much of the damage seems to have been done not by the blast itself, but by the after effects.
[Update, 10/9, 2:15 p.m.: I heard someone from Gallup on TV today breaking Gallup's results down a bit more finely than they're generally broken down. Though he didn't cite a specific number, he said Romney had a very strong polling day on Thursday, testimony that counts against my conjectural inference, above, from the Gallup tracking results. Of course, a strong Romney showing on Thursday could still have been media-driven, given how fast media judgments are disseminated. Still, this testimony does count against the more gradual media effect I posited above.]