Election predictions...
A good night for:Phil, who made this prediction at A.Well's site was almost spot on.
Most of the final polls
Cuthie, with his betting.
A reasonable night for:
The betting firms (on the final day Mike Smithson wrote, "The projected overall Labour majority, based on what gamblers are doing, is now 88 seats." WHich is good. But see post below on BG & Bow result.
Some polling firms during the campaign (notably YouGov)
Less good for:
Martin Baxter: At firstI was amazed at the accuracy of his prediction, which was 68 seat Labour majority. Then I realised they've inputted the actual vote shares (which is fair enough, as they are testing the model not the inputs) and then they've chosen a 2% Labour to Libs "tactical unwind", which presumably was chosen to make the prediction as accurate as possible. In fact their election prediction was 132 on the polls, or over 100 (I think) with the correct vote shares, i.e. lacking the tactical element (though they did I think say it would need some tactical unwind).
Dsquared's model.
My £50 bet on the Tories to get more than 211 seats. My belief the Tories might get under 30% so betting some more on some marginals. Holding both views at the same time.
Gene, on Harry's Place, over George Galloway.
Cuthie, on Harry's Place, “Let me make a prediction based on Tory gains/losses now: if the Conservatives take only one seat from the Lib Dems tonight, it will be in Eastleigh.” Nope.
posted by Matthew @ 5/09/2005 10:32:00 AM
Comments: 0
