How large a working majority will Tories get?

[oscarose]

Indeed. Its confirmation that the UK is fecked.....

[northernladyuk]

Indeed. Its confirmation that the UK is fecked.....

Couldn't happen to a nicer country.

[diseasex]

You like being in the same CUK set as diseasex?

And what set am I ?

[northernladyuk]

That YouGov model is great fun.....absolute toss though. Models aside, the pollsters that rely on self-reported turnout are in for another embarrassing day on Thurs...

I hope not, but still reckon the Tories are heading for an 80+ majority.

Good work, yougov. They even predicted Canterbury correctly.

http://yougov.co.uk/uk-general-election-2017/
http://yougov.co.uk/news/2017/06/09/the-day-after/

[BrilloPad]

And what set am I ?

The sasguru set.

HTH BISDI.

[BrilloPad]

Couldn't happen to a nicer country.

Good to see you think the UK is less nice than USA, Saudi(or any other Middle East country), China, Russia, North Korea.

[jamesbrown]

Good work, yougov. They even predicted Canterbury correctly.

http://yougov.co.uk/uk-general-election-2017/
http://yougov.co.uk/news/2017/06/09/the-day-after/

No, the model is still toss. It's not the outcome I'm looking at, but the internals of the model. Incidentally, constituency models have been used before (e.g. in 2015), and they are widely used in the US (for US states) by Nate Silver and others. Their performance is mixed, at best, and they fundamentally rely on national polls to distribute the votes at a constituency level. You didn't need the YouGov constituency model to predict a Hung Parliament on the basis of the YouGov polls. They (along with Survation) were the only companies whose polls were pointing to a Hung Parliament, although YouGov bottled it last minute by assigning the DKs in their final poll.

[northernladyuk]

Good to see you think the UK is less nice than USA, Saudi(or any other Middle East country), China, Russia, North Korea.

Boo hoo.