Most of the last polls before tomorrow’s election are now out, and they seem to point to a result along the lines of Con 33-37, LD 26-29, Lab 24-28.
There’s also a Scottish poll predicting Con 17, LD 22, Lab 37, SNP 21.
If I put the average of the figures into the Electoral Calculus Regional Predictor and shift LibDem support slightly towards the north of England, I end up with a seat prediction of Con 288, LD 98, Lab 232.
The Tories seem to be doing better in Lab-Con marginals, but this is to some extent cancelled out by the LibDems doing better in Con-Lib marginals.
Let’s therefore assume that Labour lose another 20 seats to the Conservatives, and that the LibDems win another 10 seats from the Tories, and we end up with my final prediction for tomorrow: Con 298, LD 108, Lab 212.
In many ways this would be an excellent result: The Tories would probably end up in power, but without being able to do whatever they fancy. Labour would have to find a new leader, and the LibDems would have demonstrated how bad the electoral system is if they got more votes but fewer seats than Labour.