I am going to take one thing as a small light of hope. They have polled 38 people per constituency. 23000 in 650 seats. Am I being dumb to think that that is a pretty ****ty little poll?
I feel ashamed to be British. We fell for nationalism and xenophobia.” As long as I’m ok, who care about the rest “ seems are new motto. I feel my generation has failed the youth. Might as well reach for the scotch.
And as I've mentioned previous, the crying shame (beyond, y'know, the immediate effect) is that people will take away that Corbyn's democratic socialism was the problem. The next leader will be a New Labour type, and probably several leaders thereafter. The problem wasn't Corbyn's place on the political spectrum, it was Corbyn. He was absolutely the wrong leader for this moment in time, and sadly it has cost Labour (and the UK, IMO) dearly. Edit:
23000 is a huge poll, generally speaking. Most polls run in the ~2000 person range, because that's when the margin of error gets into the +/- 3% range. Not representative constituency-by-constituency, but it would take absolute malpractice for it to be inaccurate overall.
Have you ever seen the media turn on one of the party leaders like they have done with JC? I'm not a fan of he, but I found the constant stories and slights very strange. Never seen such a coordinated effort. Makes you wonder how far up the orders came?
38 people wouldn't be reliable for that riding; the margin of error would be insane. But when you aggregate the whole result, the MOE is going to be quite low, unless there are systemic problems with the way you're conducting it...only interviewing old white men with shouty voices, say.
Right, I get that, but how does it make regional predictions? Southampton is going totally blue according to this, but where is the Southampton exit poll? Is this number magic?
You'd model based on past results...so, if Hampshire normally runs Con + 5 or something, you'd run it through a model where you used the sample from all Hampshire-area exit polls + the nationwide sample to make assumptions about how that would move individual ridings. It's well short of exact, and harder by far than modeling in a two-party system like the US: there will be wrong calls there, but it's almost definitely going to be a Conservative majority at the end of the night.
I'm afraid you are clutching at a very frail straw. The very worst outcome for Pffefel would be a majority of 50. At the other margin of error 120. Brexit hasn't gone away though, that's the real challenge.
I take it the poll will be 95% accurate, but ITV/BBC are doing themselves favours as according to them, Boris will have a 10+ year legacy and Labour are done for my lifetime, so why stay up for the boring predictable count. I'll go to bed soon Let's get sleep done.
