My colleague James Macintyre touched on this yesterday, but I thought it was worth revisiting, as it brings home just how ineffectual very many votes will be on 6 May.
The Electoral Reform Society has identified 382 constituencies across the country that are completely safe -- a "conservative estimate", it adds for good measure. That's three in every five of the 649 seats up for grabs next month.
By party, 172 Conservative, 165 Labour and 29 Liberal Democrat seats are deemed so secure that a vote for anybody else would merely be a protest. Meanwhile, 71 per cent of the east of England and 72 per cent of all constituency battles in the north-east are foregone conclusions.
But if you like your politics volatile (and remotely democratic), try the south-west, where 56 per cent of seats promise a proper fight.
Safe: a seat where a change of hands is most unlikely
Marginal 1: a Labour seat that was marginal in 2005 but probably won't be in 2010 because it is highly vulnerable and will probably fall easily
Marginal 2: the real front line between Labour and the Conservatives
Marginal 3: a seat that Labour could lose if the election goes badly, but at the moment is probably likely to remain Labour
Lib Dem long shot: Often Lib Dem seat gains are not closely related to the apparent vulnerability on the numbers from last time. These seats are ones where a "left-field'"gain cannot be ruled out, or seats where, despite a fairly small majority over the Liberal Democrats, their chances of winning are not significant enough to qualify the seat as a true marginal
Non-Labour marginal: a marginal seat held by a party other than Labour, including Con/LD seats, a few Con/Lab seats, SNP seats, LD seats, NI, etc
3-way: a three-way marginal