Psephology fanatics (or election junkies, in ordinary English) will want to read Nate Silver’s excellent posts on the British polls. Here is his seat projection based on his allocations of 2005 voters (and non-voters) to the various parties this year. His formula shows the Conservatives and Lib Dems winning more seats than projected by swingometers that assume a uniform swing in each district, and Labour winning fewer. The score: Conservatives 299, Labour 199, Lib Dems 120. This leaves Conservatives 27 seats short of the 326-vote majority. But it also leaves a Labour-Lib Dem coalition 7 seats short of a majority. Without being able to threaten to coalesce with Labour, the Lib Dems will have very little leverage with the Conservatives.

Silver updated his post after the third debate Thursday night, and here he explains how he developed his model. It looks better to me than anything I’ve seen in the British press or British websites.