Yet conventional wisdom is often wrong. For a start, as any property analyst can attest, it tends to be self-affirming. The media has leapt on recent polls that show Mr Obama with double-digit margins. But until Friday, when the conservative Drudge Report led on the much narrower two-point lead that Gallup gave Mr Obama, those polls that have not hinted at a landslide have been downplayed. And there have been quite a few.  

The RealClear Politics website’s average of polls, which gives Mr Obama a lead of 6.8 per cent over Mr McCain, offers a better guide to the situation.  It compares to John Kerry’s lead just a few weeks before he lost the 2004 election to Mr Bush. ~Edward Luce

Having been duly chastened about poll averages in the past, I probably ought to say nothing about this one way or the other, but there is a very simple point to be made here.  At a comparable time in the ’04 election, the same RCP average showed Mr. Bush ahead by approximately 2 points, which was fairly close to the final result.  It is true that there were some pollsters, most notably Zogby, who were declaring a massive Kerry victory up through the night of the election, but they were spectacularly wrong and were not representative of what the most reliable polls showed.  Those who expect Obama to win with an Electoral College margin similar to that of Clinton ’96 and a comparable 5-6 point lead in the national popular vote seem to have more evidence on their side.  (Fivethirtyeight.com currently projects Obama winning 51.8-46.5 with 347 electoral votes.)  Could that lead widen to become a proper double-digit landslide?  Perhaps, but it seems more likely that the gap will narrow slightly and we will still see Obama winning the vast majority of electoral votes.