So last night's US Presidential election debate (above) was a rather convincing, though not entirely crushing, victory for President Obama. That's three out of four in the instant polls for the Democrats, if you (firstly) ignore the one that really counted - the President's disaster in Denver - and (secondly) the testy and unsatisfying Vice-Presidential debate.
Actually, as with the second debate, there's quite a lot of evidence buried under the topline numbers (that the President came out on top - on style, substance, gravitas and just plain knowledge) that Governor Romney may have the last laugh. Independent voters, in particular, seemed to like his new 'I'm a man of peace' pledge - a sudden pivot just as unconving as all the others, at least amongst those who've been paying attention. And the President's aggressive testiness and mocking attacks may not be the best idea for a man elected on a pledge of bipartisan change - as I've been arguing all year. But still - there it is. That part of the strategy probably can't be changed now.
In fact, there is evidence of seepage among independent voters towards Governor Romney across the board at the moment. The challenger is now up among non-aligned voters in almost every poll - sometimes by very large margins. It's beginning to look like he might become President after all.
Except, except, except... Much of the data we have from the (murky and complicated) early voting ground war favours the Democrats - especially in Nevada, and to some extent in Ohio, two states that together would probably keep Mr Obama in the White House.
And we have ample evidence, from many, many elections, that that might mean very little in two weeks' time. There was the famous Dewey-Truman turnaround at the last moment, in 1948. No-one thought that Edward Heath could possibly beat UK Prime Minister Harold Wilson in 1970 - until he did, on a last-minute swing that almost everybody missed. Al Gore seemed to be miles behind heading into the final furlong of the 2000 Presidential election, only to catch up and pass his rival in terms of the popular vote. The UK Labour Party flirted with coming third in 2010, only to find a reservoir of strength from somewhere (I'm still not sure where) in the final days of the campaign. Even polls taken in the last few days of a campaign are not always that accurate, because they can't be as people make up their minds. Even the good ones have mixed records. In a tight race like this, where almost everything is within the margin of error, they can't be accurate enough to give you the winner. Especially in state-by-state and day-by-day samples.
So the Republican probably does have his nose a little bit out in front right now. It feels like we're getting near a tipping point. But anything could still happen. History says: don't bet against it.
