Chris Hanretty tells #newsnight: This will be the hardest post-war election to predict
— BBC Newsnight (@BBCNewsnight) October 28, 2014
Newsnight ran a good piece yesterday explaining some of the problems with forecasting the next general election, including a very interesting contribution from Chris Hanretty from the UEA.
One of the best features of Chris’s model is that it produces precise probabilities for each party to win each seat, which we can compare to Ladbrokes’ constituency odds. I haven’t been through every seat, although I have had a good look at the Scottish ones already, because his forecasts produce some startling results there. If you like his approach, then there are some fantastic bets to be had, all on the SNP. Here are just a few seats I picked out with the Ladbrokes’ SNP odds and those suggested by the model.
|Argyll and Bute||Evens||1/10|
|Caithness, Sutherland and Easter Ross||2/1||1/16|
|Fife North East||9/4||2/5|
|Ross, Skye and Lochaber||3/1||1/4|
|Paisley and Renfrewshire South||5/1||2/1|
|Orkney and Shetland||10/1||11/8|
Before piling in to the prices, it’s worth having a good look at the FAQs on Chris’s site. They are completely open that the local performance of the Liberal Democrats is a major source of uncertainty, and I think that has caused a lot of the differences above. For example, in 2010 the SNP finished 51% behind the Lib Dems in Orkney & Shetland. I don’t think many people would be in a hurry to take 11/8 about them turning that around. As they point out on the site: “We are not gambling on the basis of these predictions”.