Can we wallop the PH100 again on a by-election prediction?
Three months ago the endlessly irritating PH100 panel of so-called “experts and insiders” on the PoliticsHome PH100 produced their prediction for the Glasgow East by election. They got it wildly wrong while a PB online poll got it precisely right.
Now the PH100 is back again in the prediction business and their projection for next week’s Glenrothes by election appears above. The panel clearly are not gamblers and their projection has had zero affect on the by election betting.
For what it’s worth I think this is 50-50 situation and will bet on Labour as long as you can get a price that is substantially better than evens.
But can the PB community once again produce a better prediction? Please take a couple of seconds to do the online poll below.
UPDATE 1110 am
I have received the following email from Freddie Sayers of Politics Home:-
Hi Mike,
Re your blog on Glenrothes, there was an error in our graph title which we have now fixed – I wonder if you could possibly exchange/update it urgently as it is misleading? I include the corrected version below.
As is made clear in Andrew’s piece, that line graph does NOT show a prediction of the percentage result. The panel were asked to choose from a multiple choice which expressed the result “in terms of a percentage majority†– the spread of predictions is shown in the bar graph below. The line graph shows overall aggregate Westminster sentiment about the eventual victor in Glenrothes over time.
As I keep on saying, this has never been a PoliticsHome prediction: it is a very accurate measure, from the very best sample there is, of what the consensus feeling is in Westminster at any one time. It is just as useful and interesting whether they turn out to be wrong or right: if the consensus expectation is that Labour are set to win Glenrothes and they then lose, for example, the backlash can be expected to be that much more severe.
Below is the amended headline and graphic –
Mike Smithson