I'd done the simple calculation ages ago: there was no way that the LibDem was going to overtake either of the others. The 2005 figures (rounded) were: Labour 19000 (45%), Cons 14000 (34%), LibDem 7000 (16%), on a turnout of 55% of an electorate of 72000. So I had to vote tactically - Labour in order to keep out the Tory. But it stuck in my craw. I was about to search out a suitable clothespeg.
But this afternoon I decided to have a closer look (well, it was that or wash the kitchen floor). I made a few not unreasonable assumptions, and came up with a surprising conclusion.
The Labour vote in 05 was 8% down on 01, largely as an anti-war reaction. That it wasn't down further was due to the popularity of our great constituency MP Martin Salter, who has now retired. So I thought it reasonable to assume that the new Labour candidate would lose this personal vote, especially as he's been helicoptered in from Essex. So, I thought, reduce let's Labour's share by another 12%.
Assume that the Tory vote here is more or less unchanged. 34% seems about right. The candidate is local, and seems unobjectionable (apart from his politics).
Apply the national swing to the LibDems, so up from 16 to say 28%. The candidate is, again, local.
All the above is based on the 2005 turnout of 55%.
But then, assume that turnout is going to increase dramatically. This is a politically aware constituency (personal knowledge confirms this). I have factored in a 75% turnout. I've also, controversially, applied the increase in the ratio of 25% Labour, 25% Tory, 50% LibDem. (I think this is ungenerous to the LibDems).
Running all that through the calculator, it comes out at near enough 17000 votes for each of the three parties. Would you believe it?
So, guess what I'm doing tomorrow morning? That's right, looking again at my maths, checking out the relative health of my head and my heart, then trolling across to Cranberry Road to cast my vote.