My prediction is that because this looks like the toughest election in the UK to predict in my lifetime, any predictions that are made will tend to reveal what the prognosticator in question wants to happen.
The US general election of 2000 is what I base this on, where Bush and Gore were "too close to call" for most of the year leading up to it and most liberal pundits predicted a narrow Gore win and most conservatives a narrow Bush victory.
I'm inclined to go for a small Conservative majority based on the notion that 6 weeks of exposure to Ed Miliband will make the electorate realise they cannot risk putting him in charge of the country.
However I might be demonstrating the bias that I refer to by that prediction- there is certainly enough polling evidence to sustain any prediction from a solid majority for either Conservatives or Labour, or a breakthrough by insurgent parties like UKIP, the SNP and perhaps the Green Party- and in any case using one piece of evidence to make a prediction (Miliband's popularity) means I'm probably overlooking other evidence.
So the only safe prediction is that the Liberal Democrats will suffer.
The Alabama result are starting to trickle in…
4 hours ago