As the opinion polls get it wrong again

The pollsters called GE2015 wrong, they called the Brexit referendum vote wrong, and now they called the American Presidential election wrong. Guido Fawkes suggests, understandably, that in place of opinion polls gut instinct is back in fashion.


I think what we learn from this is not that gut instincts are better than polling evidence - heaven only knows that enough people predicting all three votes based on gut instinct also got it wrong - but that it is very difficult to do.

Voodoo polls such as those on the websites of newspapers or on twitter with a self-selecting sample are totally useless. And that's before you get to the ones which use the Sir Humphrey technique:



If a reputable series of opinion polls shows a substantial lead for a party or view, it's probably real. But if opinion polls suggest that it's close, be prepared for either the side which is shown ahead to have a rather larger lead in the ballot box than they had in the polls, or for the underdogs to win.

However we should also learn from the events of 2015 and 2016 that no result is impossible, nothing and nobody should ever be taken for granted, and nobody who cares about the result of a vote should ever risk being complacent about it.

Comments

Jim said…
Its not so bad,

the conservative majority made me a few quid, so did brexit, so did may and now donald has made me a few as well.

I like this political betting lark.

next call is for a tenner on Le Pen I think.

Popular posts from this blog

Nick Herbert on his visit to flood hit areas of Cumbria

Quotes of the day 19th August 2020

Quote of the day 24th July 2020