I’d like to believe that a bad performance would damage the LDs generally – but I don’t think it’s really that crucial because they are not expected to win it anyway.

I think Boris can win this and agree with some of Richard’s points – a lot of recognition – but I also think Ken Livingstone will be hard to shift.

However, I’m more interested in how a disasterous showing by the Lib Dem candidate might damage Nick Clegg. Any thoughts?

The large weighting downwards of over 55s means only that more over 55s responded than expected and fewer younger respondents replied than expected (or possibly that too many over 55s were invited, but that’s the same thing really: if too many were invited it would be because the response rate was underestimated).

Its always easy to judge whether weighting on demographics is sufficient – look at the weighted figures and compare them to reality. Political weighting by past vote and things is controversial. Weighting on things like age is incredibly simple and (unless its a universe where the age profile is unknown, like party membership) very difficult to get wrong.

We need a poll that shows both first and second prefrences.

The questions should be, will you vote for the labour party candidate Ken tory party candidate Boris etc in the telephone polls. Although both figures have managed to get themselves recognised beyond their parties.

I think Ken will win because more Libdems will always vote for the Labour candidate in the second prefrence.

What if on the first prefrence one candidate gets more than 50%, do they still count the second prefrences. (Can I vote for Boris twice and mess up my second prfrence)