Urban75 Home About Offline BrixtonBuzz Contact

Political polling

redsquirrel said:
Piece in the Guardian arguing that there could be a even bigger Troy majority (though pleasingly also that Farron might lose his seat). It's a little handwavium TBH, but the general thrust, that the increase is the Labour vote may be in seats that aren't really in play isn't totally daft.
Click to expand...

The general point that geographic distribution is true, of course. A couple of things, though. The article seems to be based on a YouGov poll from a fortnight ago which had Labour on 31%. The latest YouGov poll has Labour on 38%. So the data they are using is stale, including the regional variances. Also, it would surely cut both ways. For example, the article points to a startling 7% variance in the Tory vote in the North East, with the implication that this means they're going to be picking up seats left right and centre. But most of the NE constituencies are solidly Labour, so they are surely actually going to be piling on votes in places where they can't win, in just the same way as Labour might do down south.
 
redsquirrel said:
Piece in the Guardian arguing that there could be a even bigger Troy majority (though pleasingly also that Farron might lose his seat). It's a little handwavium TBH, but the general thrust, that the increase is the Labour vote may be in seats that aren't really in play isn't totally daft.
Click to expand...

It's just a sneaky voting reform / progressive alliance plea.
 
I think, as I've said all along*, that not many seats will actually change hands, so it will be about how the % share of the vote can be spun.


*I should remind everyone that I confidently predicted wins for Miliband in 2015 and Remain, so treat my predictions with the scorn they deserve!
 
newbie said:
the Sun attributes the rather startling 9% support the Greens attract in Wiki to 'others' but doesn't produce any analysis.
Click to expand...
I think that's actually green, nats and others, but yeah, not sure why it isn't disaggregated. Adds to the idea there was a slightly dodgy methodology in play here, even if the 8% gap was consistent with the direction of travel in the polls more generally.
 
Wilf said:
Latest poll has 8% gap:
Opinion polling for the United Kingdom general election, 2017 - Wikipedia
It's a very large sample (6,000), but done by Survey Monkey for the Sun. :hmm: I'm not minded to click through for the details.
Click to expand...

There doesn't seem to be any published data or methodology for this apart from what is in the article, so i think it can probably be considered junk. AFAICT, SurveyMonkey will find you people to complete your online survey by attaching it to competitions and things like that. Maybe it's possible that you can get a decent poll that way, but if they won't tell us how they did it, we have no way of knowing.
 
Raheem said:
There doesn't seem to be any published data or methodology for this apart from what is in the article, so i think it can probably be considered junk. AFAICT, SurveyMonkey will find you people to complete your online survey by attaching it to competitions and things like that. Maybe it's possible that you can get a decent poll that way, but if they won't tell us how they did it, we have no way of knowing.
Click to expand...
Yeah, survey monkey is usually just a quick and dirty commercial tool for doing surveys. Not sure whether survey monkey themselves have set something up with all the normal methodological safety nets in place for this survey. The fact that they had all the 'smaller' parties bundled under one figure suggests not.
 
Interesting analysis of the changes to the weighting given since the last election by the different polling companies. This confirms what I'd been thinking, and means that IMO they're going to have swung too far the other way and are going to be over predicting the Tory vote and under predicting the Labour vote in their headline figures because the youth vote is so much more engaged this time around, and will almost certainly vote in much higher numbers than in 2015.
  • ComRes have changed their turnout model, so it is based more on respondents’ demographics rather than how likely they claim they are to vote. The effect of this is essentially to downweight people who are younger and more working class on the assumption that the pattern of turnout that we’ve seen at past elections remains pretty steady. ICM have a method that seems very similar in its aim (I’m not sure of the technicalities) – weighting the data so that the pattern of turnout by age & social grade is the same as in 2015.
  • Kantar (TNS) have a turnout model that is partially based on respondents age (so again, assuming that younger people are less likely to vote) and partially on their self-reported likelihood.
  • ORB weight their data by education and age so that it matches not the electorate as a whole, but the profile of people who the 2015 British Election Study who actually voted (they also use the usual self-reported likelihood to vote weighting on top of this).
  • Opinium, MORI and YouGov still base their turnout models on people’s answers rather than their demographics, but they have all made changes. YouGov and MORI now weight down people who didn’t vote in the past, Opinium downweight people who say they will vote for a party but disapprove of its leader.
Click to expand...

So for example Yougov are weighting down people who didn't vote in 2015, but there's something like 25% of the younger age range who didn't vote in the referendum who're saying they will definitely (10/10) vote this time, and they're mostly voting for Labour.

Without the weighting the last Yougov polling only had labour 2% behind with 15% don't knows still to play for that includes 11% of those who voted Labour in 2015.
 
Given that they have been overstating the Tories since forever, I don't see why this election should suddenly be the one in which their methodology changes actually work. They'll just have more excuses like "oops Tory voters were even more shy this time than previously so 90% of the DKs went blue, oh well maybe next time we can fix that"
 
Last edited:
I've been reading quite a few interesting pieces from NumbrCrunchrPolitics (@NCPoliticsUK) on Twitter today. They predicted that the 2015 polls were wrong, iirc.

My feeling is their general gist seems to be that the Tories will still win by a decent margin and that most polling atm is likely incorrect to at least some degree because of changes of weighting and just how febrile and uncertain everything is.
 
free spirit said:
Interesting analysis of the changes to the weighting given since the last election by the different polling companies. This confirms what I'd been thinking, and means that IMO they're going to have swung too far the other way and are going to be over predicting the Tory vote and under predicting the Labour vote in their headline figures because the youth vote is so much more engaged this time around, and will almost certainly vote in much higher numbers than in 2015.


So for example Yougov are weighting down people who didn't vote in 2015, but there's something like 25% of the younger age range who didn't vote in the referendum who're saying they will definitely (10/10) vote this time, and they're mostly voting for Labour.

Without the weighting the last Yougov polling only had labour 2% behind with 15% don't knows still to play for that includes 11% of those who voted Labour in 2015.
Click to expand...

Young people never turn out on the day as much as they say they will, and I don't see that being any different this year. I reckon only 7 in 10 of the 10/10 definitely will vote under 25s will bother.
 
Last edited:
On who is most likely to protect the interests of older people dependent on the social care system: T. May: 20% J. Corbyn: 43%

ComRes.
 
Full list:

DA2XeJ9XoAQDYN4.jpg
 
The bookies are getting worried , I put a tenner on Labour getting the most seats at 15/1 , (my mum asked me to put a tenner on ) they are offering me a cash out option of about £12. And my £3 at 28/1 on Labour getting over 251 seats ,cash out is currently £6.
 
ItWillNeverWork said:
Not so inspiring, that one.
Click to expand...

chilango said:
Still trending in the right directions
Click to expand...

That's the thing. It's about movement in each poll. The ComRes one had Labour lower than the YouGov one to begin with, so while YouGov showed a smaller margin the movement in both directions are largely consistent between each poll. If anyone was expecting all the other polls to show a 5 point margin they were deluding themselves. It's the movement that matters when showing what's happening - and the final numbers will differ from poll to poll.
 
Vintage Paw said:
That's the thing. It's about movement in each poll. The ComRes one had Labour lower than the YouGov one to begin with, so while YouGov showed a smaller margin the movement in both directions are largely consistent between each poll. If anyone was expecting all the other polls to show a 5 point margin they were deluding themselves. It's the movement that matters when showing what's happening - and the final numbers will differ from poll to poll.
Click to expand...
Some interesting findings in that ComRes..

upload_2017-5-27_20-52-36.png
 
You must log in or register to reply here.

Similar threads

Veronika94
What's happened to classic,old-fashioned political scandals?
Replies
27
Views
968
tim
tim
quimcunx
  • Poll
What is your polling station usually?
2 3
Replies
60
Views
2K
ska invita
ska invita
Pickman's model
yvette cooper's pitiful politics
Replies
0
Views
197
Pickman's model
Pickman's model
N
Ever had a negative reaction to a political T-shirt?
2 3 4 5 6
Replies
163
Views
8K
Chz
Chz
Silas Loom
Practical republican politics
2
Replies
59
Views
2K
Silas Loom
Silas Loom
Back
Top Bottom