Don't vote Tory

[arcade fireman]

2 minutes ago, Mr.Tease said:

Firstly, the poll is bollocks! It will never happen! And secondly May warned a week ago that if she lost just 6 seats it would lead to a  'coalition of chaos', then suddenly this poll appears in a Murdock newspaper. I'm not saying Yougov are corrupt but what they choose to research can be ideologically driven (or even commissioned by a vested interest) 

It would be great if it were true though, because the tories would be unable to make a workable coalition (smaller parties are all pro Europe so she would either have to completely alter her brexit strategy), which would mean a Labour, SNP and libdem coalition, which I'd be quite pleased with. Might even result in Proportional representation. 

I don't think this poll is accurate either but there's nothing nefarious going on. They're just wildly differing in how they perceive the youth turnout to be. For example ICM showed a lead of 12% but that was after turnout adjustment. Without turnout adjustment that lead was only 3% - which would lead to similar results here.

This difference has existed throughout the campaign. It's only now it's starting to show a potential difference to the result since the narrowing which has been apparent in all polls. The coalition of chaos line has been around since she called the election pretty much.

On the face of it YouGov's methodology would seem to be quite sound - interviewing 50,000 people and doing a constituency based profile. It's mainly my gut instinct which is overriding that and saying it can't be right. There's little chance this has vested interests - it's a fairly simple voting intention poll and they're doing one every few days just like the other companies. All that's changed is the methodology which is a scientific not political decision. 

[eFestivals]

16 minutes ago, Mr.Tease said:

Firstly, the poll is bollocks! It will never happen! And secondly May warned a week ago that if she lost just 6 seats it would lead to a  'coalition of chaos', then suddenly this poll appears in a Murdock newspaper. I'm not saying Yougov are corrupt but what they choose to research can be ideologically driven (or even commissioned by a vested interest) 

The Times is spinning the data - that's what the tweet is - but YouGov itself is making clear its experimental nature.

I doubt there's anything of the conspiracy you're suggesting, and it's simply the case that The Times wanted a poll and the results happen to be what they are. The Times then used those results in a manner which suited The Times.

(I will point out tho that The Times is not a Murdoch sock-puppet in the way The Sun is. How it's been used is more likely to be driven by people within The Times than by Murdoch).

 

Quote

It would be great if it were true though, because the tories would be unable to make a workable coalition (smaller parties are all pro Europe so she would either have to completely alter her brexit strategy), which would mean a Labour, SNP and libdem coalition, which I'd be quite pleased with.

There would be that coalition, maybe ... tho it's quite possible, for example, that the SNP might get Labour to back down on (say) it's higher income tax plans because those would hugely embarrass the lesser tax-the-rich plans of the SNP (who's supporters like to claim them as more left than Labour).

And then there's everything about how that coalition might operate, which might prove May's claims true and ensure such a thing never ever happens again.

 

Quote

Might even result in Proportional representation. 

PMSL

I wet myself every time anyone suggests Labour might back PR. You might as well say the Tories will double NHS spending.

Edited May 31, 2017 by eFestivals

[Mr.Tease]

12 minutes ago, eFestivals said:

The Times is spinning the data - that's what the tweet is - but YouGov itself is making clear its experimental nature.

I doubt there's anything of the conspiracy you're suggesting, and it's simply the case that The Times wanted a poll and the results happen to be what they are. The Times then used those results in a manner which suited The Times.

(I will point out tho that The Times is not a Murdoch sock-puppet in the way The Sun is. How it's been used is more likely to be driven by people within The Times than by Murdoch).

 

There would be that coalition, maybe ... tho it's quite possible, for example, that the SNP might get Labour to back down on (say) it's higher income tax plans because those would hugely embarrass the lesser tax-the-rich plans of the SNP (who's supporters like to claim them as more left than Labour).

And then there's everything about how that coalition might operate, which might prove May's claims true and ensure such a thing never ever happens again.

 

PMSL

I wet myself every time anyone suggests Labour might back PR. You might as well say the Tories will double NHS spending.

Re PR :Things change- you have to hope that the proposed boundary changes will have woken a lot of them up. Certainly there's a cross section of the party that are starting to push the idea again

[eFestivals]

Just now, Mr.Tease said:

Re PR :Things change- you have to hope that the proposed boundary changes will have woken a lot of them up. Certainly there's a cross section of the party that are starting to push the idea again

I don't doubt the possibility of a change in the future, tho that will only be an admission by Labour that it no longer a party which can gain power via FPTP.

We're still a million miles away from Labour being willing to face up to that possibility. It'll probably take 30 years of out-of-power before it happened, given that Blair got them back in power after 18 years.

[Mr.Tease]

7 minutes ago, eFestivals said:

I don't doubt the possibility of a change in the future, tho that will only be an admission by Labour that it no longer a party which can gain power via FPTP.

We're still a million miles away from Labour being willing to face up to that possibility. It'll probably take 30 years of out-of-power before it happened, given that Blair got them back in power after 18 years.

I don't think it will take that long as that as I don't think the younger MPs are as married to the current system as the older guard are/were.

When Blair won they suddenly decided 'wow, we can win under this system - we have no need for PR', which is like suddenly getting a big feast after starving for days and declaring 'I'll never be hungry again! '- couldn't believe the short sightedness at the time! 

[arcade fireman]

50 minutes ago, eFestivals said:

I can see YouGov getting slated for the "results" of this latest poll, tho that'll be much-more due to the many false narratives being used around data that Yougov itself is happily saying is as uncertain as it's possible to be.

It's essentially an experiment, and YouGov is making that clear with the huge range its saying the outcome might fall between.

It's a wide range alright, but interestingly they're claiming the Tories have as much chance of ending up with 274 seats (the lower limit) as they do 345 seats (upper). I don't think that can be right but that's what their data concludes.

The fundamental problem the polling companies have at this election is that they're trying to correct the errors of the last election where one major error was overestimating the turnout amongst Labour's core vote. However they're now confronted with a Labour leader whose whole MO is getting that vote mobilised and who has seen a big spike in voter registrations amongst that group.

So no one really knows how this will turn out. People who are being paid big sums to estimate these things for a living are disagreeing on this, so I doubt we'll come to a proper conclusion. I'd guess somewhere in between though, which would may leave Labour in a worse position in the long term than if it were either way. 

Edited May 31, 2017 by arcade fireman

[Mr.Tease]

Both May and Hunt this morning said it's not right that younger people should have to fund older people's social care. I do wish Labour would seize the opportunity and push the line that older people have contributed via tax their whole lives, and it's insulting to them etc we really need to drive a wedge between older voters and the tories. 

[Mr.Tease]

13 minutes ago, arcade fireman said:

It's a wide range alright, but interestingly they're claiming the Tories have as much chance of ending up with 274 seats (the lower limit) as they do 345 seats (upper). I don't think that can be right but that's what their data concludes.

The fundamental problem the polling companies have at this election is that they're trying to correct the errors of the last election where one major error was overestimating the turnout amongst Labour's core vote. However they're now confronted with a Labour leader whose whole MO is getting that vote mobilised and who has seen a big spike in voter registrations amongst that group.

So no one really knows how this will turn out. People who are being paid big sums to estimate these things for a living are disagreeing on this, so I doubt we'll come to a proper conclusion. I'd guess somewhere in between though, which would may leave Labour in a worse position in the long term than if it were either way. 

So essentially they've spent a lot of time and money to conclude they have no idea what will happen! 

[arcade fireman]

3 minutes ago, Mr.Tease said:

So essentially they've spent a lot of time and money to conclude they have no idea what will happen! 

They're different polling companies. This election is very different to others for them - recent results have led them to develop quite different approaches to how they conduct their polling. One or two of them will probably be right. It's just a question of which one. 

Also "herding" was seen as a problem in the last election. There was a poll conducted a couple of days from the last election which got things pretty close to the end result but wasn't published as it didn't fall into line with the others. They're trying to move away from that.now, though I suspect at the next election pollsters will follow the lead of whichever company gets things right this time.

Edited May 31, 2017 by arcade fireman

[eFestivals]

17 minutes ago, arcade fireman said:

It's a wide range alright, but interestingly they're claiming the Tories have as much chance of ending up with 274 seats (the lower limit) as they do 345 seats (upper). I don't think that can be right but that's what their data concludes.

which only shows how little faith they have in that data and methodology at the moment. Which is fine. A new method has to start somewhere.

But just about everywhere I'm now seeing that Times tweet, which is a big lie and a big lie that people are swallowing as meaningful, when YouGov make clear it's not.

[eFestivals]

5 minutes ago, Mr.Tease said:

So essentially they've spent a lot of time and money to conclude they have no idea what will happen! 

Not really. They knew at the start of getting into that new method that no reasonable conclusion can be drawn from it until *after* the election. It's an experiment until after the election.

Meanwhile it suits others to spread lies about what they're really doing.

[eFestivals]

22 minutes ago, arcade fireman said:

The fundamental problem the polling companies have at this election is that they're trying to correct the errors of the last election

This poll is not that. It's something completely different to everything YouGov have done before.

It's their first attempt at using constituency data, probably being done because they now have enough people inputting data that they feel they can get something sensible out of the other end of it.

BUT .... because they've not done the process before, they have no benchmarks to work from, which is why the conclusions are so wide.

 

[Mr.Tease]

2 minutes ago, eFestivals said:

Not really. They knew at the start of getting into that new method that no reasonable conclusion can be drawn from it until *after* the election. It's an experiment until after the election.

Meanwhile it suits others to spread lies about what they're really doing.

I said that the people who commissioned the research and also the times have an agenda (as well as possible researcher bias which is a real thing- most research finds what the researcher expected) 

[eFestivals]

4 minutes ago, Mr.Tease said:

I said that the people who commissioned the research and also the times have an agenda

there's no commissioner of this research as such. It's YouGov's own experiment, of an area they feel they need to expand their commissioned research into.

Once it's no longer the experiment it currently is, only then will they be seriously trying to punt it out to people as something they should look at commissioning.

The fact that they found a buyer for their experimental data is something different.

Edited May 31, 2017 by eFestivals

[eFestivals]

4 minutes ago, Mr.Tease said:

 (as well as possible researcher bias which is a real thing- most research finds what the researcher expected) 

PMSL

That's just the idjuts at uni, or those who remain idjuts after uni.

Edited May 31, 2017 by eFestivals

[Mr.Tease]

3 minutes ago, eFestivals said:

PMSL

That's just the idjuts at uni, or those who remain idjuts after uni.

Its a fact as most NHS workers can tell you

[DeanoL]

 

1 hour ago, eFestivals said:

I can see YouGov getting slated for the "results" of this latest poll, tho that'll be much-more due to the many false narratives being used around data that Yougov itself is happily saying is as uncertain as it's possible to be.

It's essentially an experiment, and YouGov is making that clear with the huge range its saying the outcome might fall between.

Yeah, The Times are reporting the "worst case" scenario as that's a better story. Arguably also because it benefits the Tories.

45 minutes ago, arcade fireman said:

The fundamental problem the polling companies have at this election is that they're trying to correct the errors of the last election where one major error was overestimating the turnout amongst Labour's core vote. However they're now confronted with a Labour leader whose whole MO is getting that vote mobilised and who has seen a big spike in voter registrations amongst that group.

And it's not part of the regular electoral cycle. So it's questionable how much traditional GE methodologies work. And on top of that it's a year after a major and successful referendum. 

[eFestivals]

18 minutes ago, Mr.Tease said:

Its a fact as most NHS workers can tell you

which might explain some of the state of the NHS, just perhaps? Or is it only money that can make brains work there? 

[Kashkin]

There are some further details of this modelling exercise (it's not a poll, but additional analysis based on multiple polls) on the YouGov website:

https://yougov.co.uk/news/2017/05/31/how-yougov-model-2017-general-election-works/

[eFestivals]

17 minutes ago, Kashkin said:

There are some further details of this modelling exercise (it's not a poll, but additional analysis based on multiple polls) on the YouGov website:

https://yougov.co.uk/news/2017/05/31/how-yougov-model-2017-general-election-works/

wow - far too small a sample size per constituency, and particularly when sign-ups are open and so gaming of the yougov system is possible.

They're attempting to mitigate those issues, but as they say themselves it's a 'best guess' and not an accurate statistical model (tho they do attempt to talk it up, tho they would do as it's likely to be their future business).

[Kashkin]

55 minutes ago, eFestivals said:

wow - far too small a sample size per constituency, and particularly when sign-ups are open and so gaming of the yougov system is possible.

They're attempting to mitigate those issues, but as they say themselves it's a 'best guess' and not an accurate statistical model (tho they do attempt to talk it up, tho they would do as it's likely to be their future business).

Yes - but polling individual constituencies is difficult online, and telephone polling would be prohibitively expensive on a large scale (hence why 'proper' constituency polling is generally only used for specific marginal seats of interest, rather than on a wider scale). 

[Mr.Tease]

2 hours ago, eFestivals said:

which might explain some of the state of the NHS, just perhaps? Or is it only money that can make brains work there? 

It's a basic tenant of all research and methodology as anyone who has done any will tell you. Knowingly or unknowingly a researcher will tend to frame their research to achieve the outcome they expect- the questions they ask, the measurements they use etc etc. I don't get your hostility to this basic fact. 

[zahidf]

Vote Tory for 5 more years of heart warming stories like this

Edited May 31, 2017 by zahidf

[waterfalls212434]

squeaky bum time efests.....you can call this poll inaccurate all you like and theres noone saying polls ARE all accurate, however this simply mirrors a trend of many polls over the last month or so which show labour making large gains. 

The graphs of the `poll of polls` as it were shows this, a steep rise in labour support over the period since the election was called a steep drop in tory support in the same period.  I may not trust polls by themselves but I do trust trend as trends happen for a reason, one or 2 rogue polls can be just that, and meaningless but when the median of `all` the polls is showing a trend in one direction, I take it more seriously.

`edit` love the fact ukips support has just fucking tanked as well....that purple drop is good to see...if theres one band of w*nkers I detest more then torys its the racist torys.

Edited May 31, 2017 by waterfalls212434

[eFestivals]

17 minutes ago, Mr.Tease said:

It's a basic tenant of all research and methodology as anyone who has done any will tell you. Knowingly or unknowingly a researcher will tend to frame their research to achieve the outcome they expect- the questions they ask, the measurements they use etc etc. I don't get your hostility to this basic fact. 

Nope, that's a basic tenet of bad researchers. And you only know bad ones from what you say.

I know with certainty that internal govt-level research wasn't that shit 2 decades ago.