Well, what do you know, I go away for a few days, and...not much changes.
The Biden-Ryan debate was entertaining, probably a draw overall, and didn't have much of an impact on the race, though Biden did well enough to calm Democrat nerves.
Mitt Romney's bounce from the first debate turns out to have been something more than a bounce, because the ball stayed up: Romney is now equal with Obama or leading him very slightly in most national polls. The race's equilibrium has been reset. Certainly, the lead that Obama built up after the conventions has been wiped out.
The incumbent is still favourite, for the time being at least, mainly because he is doing better in the swing states than in the national race (Nate Silver calculates a 15% chance that Obama wins in the electoral college and loses the national vote, which for liberals, outraged by the 2000 election, would be, well, awkward). The best explanation for this discrepancy is that the swingers have been soaked in advertising for several months and thus more likely to have made up their minds about the candidates, whereas, for other voters, the evidence of debate night was given more weight.
One thing that political scientists will look at in depth after this campaign is over is why the first debate had such an unusually big effect on the polls. They'll have to start with the question of why the debate was watched by so many people live on TV, when everyone has been saying that fewer people were likely to watch than ever. My theory is that it's down to the rise of Twitter and social media, which seem to be making these big live TV events even bigger than they were. That might help to explain the big impact too - the consensus about who won and who lost is spread with a speed and intensity that far exceeds anything in previous elections.
Which brings us to Tuesday night, and the second presidential debate. It's a town hall format, with the candidates taking questions from voters in the round. I'm sure Obama will show more signs of life than he did in the first debate, but I'm far from sure he'll come off best. His strength in this format is that he can be genuinely warm and charming around voters when he's relaxed. His vulnerability is that when he's nervous or uncertain his default is to act like a professor addressing a seminar. If you hear him use the word "instructive" then you'll know he's in trouble.
Romney will probably do well again. Despite his excellent showing last time, expectations may be comfortably low again for him, because he isn't exactly known for his folksy charm, improvisational ability or emotional empathy. But he's had much more practice at this over the last year than the president has, so I expect he'll be competent enough to put the spotlight back on Obama.
What does Obama need to do? He can't just say all the stuff he wishes he'd said in the first debate; the town hall format doesn't really lend itself to aggressive debating or forensic dissection of an opponent's positions. So he will have to be tougher on Romney's evasions without losing sight of the questions and the questioners - constructing his answers precisely while remaining attuned to the emotional atmospherics of the room. Above all, Obama has to convey energy, conviction and passion for the cause. He has to look like he believes. Because if he doesn't, nobody else will.