The bad news is obvious. President Barack Obama took a serious hit in the national polling, and a serious, but smaller, hit in the battleground states. The good news is that his previous lead was big enough that only Colorado has flipped over to Mitt Romney, giving Obama a healthy 323-215 lead in the Electoral College. The other good news is that the daily trackers indicate a fading bounce. Gallup should be really interesting the next two days as last Thursday and Friday roll off their average.
Also, I've kept in the plethora of fly-by-night GOP pollsters who have proliferated late this cycle. I figure having them push the averages down lower than what credible polling suggests gives us a good worst-case scenario.
But for now, we face battleground polling mostly taken over those brutal post-debate days, and the movement was dramatic. Let's compare both candidate's numbers before and after the debate. First Romney:
A week ago there was just a single state in which Romney was above 46 percent, and about half he was in the low 40s. After the debate, he is over 46 percent in all but two. His share of support has gone up an average of 2.6 points. Ironically, his national support went up just 1.7 percent, suggesting he gained more in the States That Matter than nationally.
Obama suffered an average loss of 1.5 points in the battleground states, compared to 4.4. in the national polling. He's still above 46 percent everywhere, and of course, still has more than Romney in all of these states except for Colorado.
Interestingly, this was the best-case scenario for both individual candidates—Romney gained more where it mattered, while Obama lost least where it mattered.
Romney's problem remains the same one he's always had—he's still losing and faces a difficult electoral map. And if he can't take the lead now, after a major Obama stumble, when can he take a lead? Seemingly, only if Job Biden and Obama continue stinking it up in the debates. Absent that, the polling will show bigger Obama leads in next week's battleground polling.