In my most recent blogpost on the British election, I noted that it’s difficult to translate poll numbers into projections of seats won, because the various models for doing so assume a uniform swing in each district—something that is unlikely to happen. Now in an excellent blogpost, ace numbers cruncher Nate Silver comes up with another formula. To generalize his results: it suggests that Conservatives and Liberal Democrats will win more seats and Labour fewer seats than the uniform-swing formulas suggest. This reduces the possibility that Labour will win fewer popular votes than one or both of the other parties but still end up with more seats. Silver’s blogpost is very much worth reading and pondering.
