danny la rouge
More like *fanny* la rouge!
Really?Yes. We had the same conversation with the same guy. Because we went there together.
I'm not stalking you, honest!
Political polling
- Thread starter brogdale
- Start date
I'm not stalking you, honest!
Vintage Paw
dead stare and computer glare
Really?
LIES!
It's sometimes a race to see who can get the talking point out on twitter first.
During question time I've been known to utter "don't you dare tweet that, I said it."
Pickman's model
Starry Wisdom
Not to mention why do they even bother
bemused
Well-Known Member
To be fair they had to hedge their bets giving they were on their own. I'm sure now they are trolling all other polling companies.
they didn't adjust their model, just their final poll.
Fez909
toilet expert
Half right.
They were running two polling methods in parallel. The new one, that seems to have been correct, and did not change throughout. And the trad one - which they did adjust the modelling on for the final poll. And was wrong with or without the change - actually the change made the prediction worse.
ska invita
back on the other side
The lesson from all this should be that there are now only two reliable polling methods: a multi-dimensional regressional analysis of demographics based on 50,000+ people and a thorough or a statistically controlled exit poll. Anything else is just pissing around.
chilango
Hypothetical Wanker
Dare we go there?
OK...here goes (dons kevlar flak jacket)...
Here's the 'baseline' polling for the 2017 GE:-
Anyone else got the stomach for this?
Yes even 'gut feeling of pessimistic urbanites' needs a bit of fine tuning.
brogdale
Coming to terms with late onset Anarchism
47 days; a long time in politics.
free spirit
more tea vicar?
or the the raw data actually has at least as much validity as the adjusted data, because the raw data on most of the polls has been showing labour pretty close to the tories for weeks.
Vintage Paw
dead stare and computer glare
NEW POLLING
(Thank god, I was getting withdrawal symptoms.)
On who would make best Prime Minister:
T. May: 39% (-4)
J. Corbyn: 39% (+7)
(via @YouGov & @ShippersUnbound / 09 - 10 Jun)
Chgs. w/ 07 Jun
(Thank god, I was getting withdrawal symptoms.)
On who would make best Prime Minister:
T. May: 39% (-4)
J. Corbyn: 39% (+7)
(via @YouGov & @ShippersUnbound / 09 - 10 Jun)
Chgs. w/ 07 Jun
Vintage Paw
dead stare and computer glare
And more:
moochedit
I'm not feeding seagulls
I read somewhere that you govs new method correctly predicted brexit and trump as well but they didnt release them at the time as they were testing it out.
agricola
a genuine importer of owls
Exactly - this was a failure of interpretation, in that everyone bar Yougov and Survation (and Ashcroft, to an extent - his polling did show that the more turnout was up the closer it would be) didn't take into account what would happen if turnout was up, especially the youth turnout. There were plenty of reasons to suggest that it might have been (hot weather, Corbyn deliberately going after the youth vote, there being an actual choice at this election etc) and yet many firms persisted in assuming it would be a "normal" election for which the normal rules would apply.
free spirit
more tea vicar?
yep, as I pointed out multiple times in the election run up.
The key bit I didn't expect was the scale of the SNP losses in Scotland.
free spirit
more tea vicar?
That survation poll has loads of detail, including a statistic that 45% of people expect that Labour would win a 2nd election vs 33% for the tories.
Also 47% oppose a tory/DUP government vs 33% approve. 38.5% approve of a lab / lib / snp / green alliance to govern with 47% opposing it. (not sure why they've not got plaid in there).
Also 47% oppose a tory/DUP government vs 33% approve. 38.5% approve of a lab / lib / snp / green alliance to govern with 47% opposing it. (not sure why they've not got plaid in there).
Vintage Paw
dead stare and computer glare
That's the problem I've got with these polls. It grinds my gears when I do a YouGov poll and it asks something and I want to be able to qualify it: "yes, but I'm dissenting from a left-wing point of view, not from the right as this will be spun in the press."
newbie
undisambiguated
Uncertainty doesn't appeal to everyone: as a sentiment "I don't care who wins so long as it's decisive" has some merit as a vote driver for those that aren't ideological and just want a quiet life.
A lot of voters want strength and stability from the party they vote for, rather than looking to back a winner. Which of the two main parties most fulfils that brief has reversed completely in the last week.
Much criticism I've encountered of Corbyn is of his incompetence and inability to control his party - I think the popular view of that will have been changed substantially by Thursday's result - this is what we're seeing reflected in the polls (as well as a change in weighting to reflect the reality of new voting patterns).
Much criticism I've encountered of Corbyn is of his incompetence and inability to control his party - I think the popular view of that will have been changed substantially by Thursday's result - this is what we're seeing reflected in the polls (as well as a change in weighting to reflect the reality of new voting patterns).
strung out
💩 🤣 🍆 💦 🧐 👻 🐝 🐈⬛
Is this a good thread to point out that the tories probably didn't blow a 20+ point lead? The fact is that their lead was never as big as people thought it was in the first place, because the polls were all being weighted incorrectly.
The final gap in the election was a couple of percentage points, and although Labour significantly closed it over the course of the campaign, the polls being so misleading in the the first place are probably to thank for May thinking she had it sewn up and calling the election so early.
The final gap in the election was a couple of percentage points, and although Labour significantly closed it over the course of the campaign, the polls being so misleading in the the first place are probably to thank for May thinking she had it sewn up and calling the election so early.
Fez909
toilet expert
Although you have a point here, of sorts, your comment is a bit contradictory.
Either they have a huge lead and Labour "significantly closed it" or they didn't and the polls were wrong. It can't be both.
Unless you're saying they had something like a 10pt lead, not 20? Think the final result was 2.5% lead to Tories.
Even if that's true, blowing a 10 pt lead in 6 weeks is ridiculous.
Raheem
Well-Known Member
The difference between the polls and the reality on election day was pretty much purely turnout. If the people that made the difference were largely only motivated to vote by the campaigns, then that does indeed mean that a gap of 20-odd points was closed. We'll never know, of course, but Corbyn's personal ratings shot up massively in the weeks following the manifesto leak, which suggests that a large number of people were inspired to support Labour after that.
Similar threads
