On Thursday we are hosting some of Britain's leading psephologists - opinion pollsters to you and me - to hear them explain the trends that came out of the recent general election here.

It is remarkable that most of the polls started the campaign predicting something like a Labour majority of 60; then the odds widened, with predictions of 100 majority and more; then on the last day they closed again to predict something very like the outcome, in the 60s.

This is pretty weird. Perhaps there is an amount of dissembling going on. Politicians have lied to us for so long, perhaps we have started to lie to them, and not tell them our true voting intentions until it's too late to make a difference.

This campaign was fought more than any previous one in the marginal seats - and with the marginal voters. All the parties had software to help them identify the 600,000 or so swing voters who would make the difference, and to be able to focus specific messages direct to them. That makes it hard for national polls to pick up the tiny but vital details of what is happening with specific voters in specific seats.

All in all, I was surprised the pollsters called it so accurately, on the last day, at least. As election technology gets more and more sophisticated, though, I wonder how accurate the polls will remain.