Political polling

[danny la rouge]

Anti-Clinton attack ads, customised and targetted using crafty algorithms, are credited by many with helping trump to victory. The tories' version seems much more ham-fisted. The way their ad is edited makes it obvious that it's a selective-quoting hatchet job, and as others have pointed out the still image of Corbyn with the phrase 'this man could be prime minister' could just as easily have the opposite effect to the one they're after.

There doesn't seem to be any targetting going on. There's no positive message from the tories to counterbalance the negative anti-Corbyn stuff. Remember Cameron's 2010 billboards? He was everywhere, telling people he was gonna do this that and the other. All bullshit of course, but he was at least looking like a man who wanted the job rather than a man who just didn't want someone else to have it.

Yup, you have to take it as a part of the campaign, and May has been utterly dismal. Despite her starting position. Despite the media being natural supporters. Despite an opposition with a parliamentary party that doesn't support its leader.

I still think they'll win. But May is damaged and will only limp on until whatever the 1922 committee thinks is a decent interval.

But then, at this point in the US elections I still thought Clinton would squeak it.

 

[danny la rouge]

Yes, suspect Labour haven't done enough there to recover - 25% is only marginally up on two years ago. Could hold on to Edinburgh South I guess, suppose there could be some churn - SNP got 50% in 2015 so a drop to 43-45 may see a couple of seats go elsewhere.

LD down again, from 7.5% in 2015. However, if 2015 couldn't shift Carmichael from the islands then I doubt he'll go now either. Orkney & Shetland Liberal since 1950 & only 17 years without a Liberal/Whig MP in 200 years

I think the three non SNP seats will remain the way they are. And I think the Tories may add a couple in the borders and possibly North East. But if this this right, nothing like they were hoping for.

 

[newbie]

In trying to find out if deliberately watching their ad costs them money, which I haven't as yet, I did find this round-up of attack ads from the beeb. Given the spending brouhaha after the last election this is probably going to get quite a lot of analysis after the election. They make the point

A word of caution is in order though: our methodology is limited by the fact that we can only see what you, our audience, send us.
...
If you spot an advert in your social media feed which looks political then please send us a screengrab or a link to the content.

There doesn't seem to be any targetting going on.

I certainly haven't been targetted * but I'm not sure any individual can draw conclusions based only on their own experience. It's not what the BBC have reported

Our crowdsourced experiment has also revealed more about the local ambitions of the Conservatives.

We found that the Conservatives are targeting Labour or Liberal Democrat seats in the Midlands and the North, where a significant proportion of the electorate voted for UKIP in 2015. We have seen adverts from 12 constituencies that match this profile. The aim seems to be to not only attract UKIP voters, but also Labour supporters with reservations about Corbyn.

"The geographical targeting on Facebook is pretty granular," says Carl Miller of the Centre for the Analysis of Social Media at Demos. "This allows political parties to focus their adverts on target seats."

But i still don't know if deliberately seeking out the ads costs the tories money. I don't particularly want to make extra profit for Facebook or Google, but...

* but I have been wondering why something called BlightyTV has been popping up on Youtube. Turns out it's linked on the tory facebook main page. hmm.

 

[littlebabyjesus]

Anti-Clinton attack ads, customised and targetted using crafty algorithms, are credited by many with helping trump to victory. The tories' version seems much more ham-fisted. The way their ad is edited makes it obvious that it's a selective-quoting hatchet job, and as others have pointed out the still image of Corbyn with the phrase 'this man could be prime minister' could just as easily have the opposite effect to the one they're after.

There doesn't seem to be any targetting going on. There's no positive message from the tories to counterbalance the negative anti-Corbyn stuff. Remember Cameron's 2010 billboards? He was everywhere, telling people he was gonna do this that and the other. All bullshit of course, but he was at least looking like a man who wanted the job rather than a man who just didn't want someone else to have it.

There is one important difference between 2010 and 2017, which is that Cameron was running to oust the incumbent, while May is the incumbent. Her biggest weakness as the incumbent is that she wasn't elected PM last election and the record she's running on is flimsy at best. She called Article 50. That's about it in terms of what she's done in the last year - that's all she has to run on.

 

[littlebabyjesus]

Yup, you have to take it as a part of the campaign, and May has been utterly dismal. Despite her starting position. Despite the media being natural supporters. Despite an opposition with a parliamentary party that doesn't support its leader.

I still think they'll win. But May is damaged and will only limp on until whatever the 1922 committee thinks is a decent interval.

But then, at this point in the US elections I still thought Clinton would squeak it.

I still thought Clinton would squeak it when I went to bed on US election night.

 

[kabbes]

Really? And yet the Tories seem to have decimated* their lead. Doesn't seem to have worked.


(*I know).

Because is there anybody at this point who doesn't know that Jezza has this supposed (but actually entirely reasonable) anti-violence record being spun as pro-terrorist? It's already baked into the numbers.

 

[danny la rouge]

Because is there anybody at this point who doesn't know that Jezza has this supposed (but actually entirely reasonable) anti-violence record being spun as pro-terrorist? It's already baked into the numbers.

Indeed. So the ad won't be affecting existing Labour support, but should be maintaining the Tory lead. Except it isn't.

 

[littlebabyjesus]

Indeed. So the ad won't be affecting existing Labour support, but should be maintaining the Tory lead. Except it isn't.

From what I've seen recently, including the debate last night, all the tories seem to be doing now is trying to get their vote out. Don't think they're even trying to change anybody's mind from another party, just trying to avert apathy from their side. In that sense, warnings about Corbyn might have an effect, but may be aimed at people who would never be voting for him.

 

[SpookyFrank]

There is one important difference between 2010 and 2017, which is that Cameron was running to oust the incumbent, while May is the incumbent. Her biggest weakness as the incumbent is that she wasn't elected PM last election and the record she's running on is flimsy at best. She called Article 50. That's about it in terms of what she's done in the last year - that's all she has to run on.

She could make some shit up though. Brexit is her opportunity to do this because it hasn't happened yet so her and the tories have yet to fuck it up. Even on brexit though, she's said nothing. Again it's all about why Corbyn can't do it, not why she can. And focussing on Brexit mugs off the large chunk or her support who would have voted against it, or who don't want us to be completely cauterised from Europe. This is probably why she's light on detail; she can't think of any details that wouldn't piss off more people than they won over.

 

[Dogsauce]

The Tories/Crosby are generally very good with what they're doing though - they do a lot of focus groups to work out the kind of attack that works - last election I thought the 'coalition of chaos' bollocks was weak and unlikely to be the sort of thing people were bothered about - instead it was picked up in post-election analysis as resonating strongly. And we've already seen on this board people lapping up the Corbyn/IRA bollocks. It'll work.

Separately, it's also worth remembering that the Tories gained their slim majority at the last election by cheating at the ground game, overspending in constituencies and bussing in activists where they had few volunteers. They might not be able to play the same game where they have a localised cunt deficiency this time, and Labour has a very active base, at least in the cities, which is being directed at target seats.

 

[Dogsauce]

She could make some shit up though. Brexit is her opportunity to do this because it hasn't happened yet so her and the tories have yet to fuck it up. Even on brexit though, she's said nothing. Again it's all about why Corbyn can't do it, not why she can. And focussing on Brexit mugs off the large chunk or her support who would have voted against it, or who don't want us to be completely cauterised from Europe. This is probably why she's light on detail; she can't think of any details that wouldn't piss off more people than they won over.

The prick Boris Johnson was waffling on about 'Labour not having a plan' for Brexit on the BBC this morning, yet the Tories haven't really been open about what they're offering yet. The hypocrisy needs flagging up when they do this.

 

[danny la rouge]

From what I've seen recently, including the debate last night, all the tories seem to be doing now is trying to get their vote out. Don't think they're even trying to change anybody's mind from another party, just trying to avert apathy from their side. In that sense, warnings about Corbyn might have an effect, but may be aimed at people who would never be voting for him.

That's the thing, though. If the polls are right they aren't holding their position, they're retreating.

 

[SpookyFrank]

The prick Boris Johnson was waffling on about 'Labour not having a plan' for Brexit on the BBC this morning, yet the Tories haven't really been open about what they're offering yet. The hypocrisy needs flagging up when they do this.

BBC etc have let them get away with it every time so far. A simple, 'this plan of yours, what was it again?' could derail the whole thing.

Nobody has mentioned the leak from the Juncker meeting either.

 

[SpookyFrank]

Separately, it's also worth remembering that the Tories gained their slim majority at the last election by cheating at the ground game, overspending in constituencies and bussing in activists where they had few volunteers. They might not be able to play the same game where they have a localised cunt deficiency this time, and Labour has a very active base, at least in the cities, which is being directed at target seats.

The tories have recently learned from their chums at the CPS that camapign spending rules are made to be broken.

 

[kabbes]

Indeed. So the ad won't be affecting existing Labour support, but should be maintaining the Tory lead. Except it isn't.

My retcon interpretation:

1) When they first hear something, people react to it. Then they get used to the idea and it stops having so much impact. So whatever gain it initially brought about for the Tories is leaking away.

2) People have now had the chance to hear Jezza explain himself and they find the explanations reasonable. At the end of the day, it's only the most extreme of the hardcore that think fighting is better than talking. And on the Ireland stuff, he actually has history on his side -- in the end, it WAS talking to the IRA that brought about peace. People know this.

3) Now he has had a lot of media exposure, people have a measure of his character. This just doesn't fit with the message of him being a terrifying terrorist-sympathising monster. He's a genial old man of good humour, not somebody shouting that all shall burn in purifying fire. People believe their own assessment a lot more than what they are told.

 

[littlebabyjesus]

That's the thing, though. If the polls are right they aren't holding their position, they're retreating.

Yep. Not saying it's working, just that this is what their whole campaign seems to amount to to me. If the polls are right, it absolutely isn't working.

I think they thought Corbyn was unelectable so all they needed to do was to point at him as often as possible. They believed their own propaganda, seems to me - the evidence suggests that Corbyn has a history of being rather electable.

 

[killer b]

While Trump's campaign did use relentless targeted attack ads, so did Clinton's (her campaign was considerably better funded too, IIRC). The difference was that as well as the attack ads, Trump also ran a very positive campaign. He gave people something to vote for. Clinton offered only hard work and the spectre of Trump - in many ways, a weird mirror image of this election campaign (hopefully with the same result...)

 

[The Pale King]

There is one important difference between 2010 and 2017, which is that Cameron was running to oust the incumbent, while May is the incumbent. Her biggest weakness as the incumbent is that she wasn't elected PM last election and the record she's running on is flimsy at best. She called Article 50. That's about it in terms of what she's done in the last year - that's all she has to run on.

Agreed. It also struck me watching some of the debt last night that this is a 7 year old government that has well run out of steam. They have few achievements to boast of ('taking the lowest paid out of tax' - a Lib Dem policy from the coalition days is the one mentioned most regularly) and seemingly few ideas for the future. All the positivity and futurity on the Labour side at the moment, and when Rudd said something about a 'brighter future' last night it just sounded fucking laughable.

 

[belboid]

Standard are reported that Corbyn is now more popular than May in the capital, and on a 50% share of the vote - which is the same as in 97

Shock surge for Labour as Jeremy Corbyn now more popular than PM in London

 

[killer b]

Labour on 50% in London (yougov)

 

[bi0boy]

5/8 were within 4% of the Tory vote. Isn't that their margin for error? The other 3 weren't that far away either.

So yes, they were underestimating the Tories but not by that huge a margin. I think the problem is as much about the communication and interpretation of the polls as the actual percentages. There is all kind of model and parameter error in there, which will tend to systematically under- or over-estimate certain results, as well as random result error. This type of uncertainty is really not properly communicated. And then the way this gets interpreted as uniform swing is just all kinds of wrong and should really be binned.

Interesting paper published last week suggesting that pollsters are taking samples which don't include enough non-voters (obviously non-voters can't be arsed to do political opinion surveys). When the sample is then demographically weighted, the views of non-voters get overweighted. The authors suggest this accounts for the 2010 Cleggasm, so it wouldn't surprise me that it is playing some role in the current situation.

Missing Nonvoters and Misweighted Samples | Public Opinion Quarterly | Oxford Academic

I think for the polling companies who for the rest of the year are doing market research for corporations, election polling is seen just as a way to stick their neck out and perhaps be lucky enough to catch a few plaudits if their way of getting things wrong happens to be right.

 

[J Ed]

Standard are reported that Corbyn is now more popular than May in the capital, and on a 50% share of the vote - which is the same as in 97

Shock surge for Labour as Jeremy Corbyn now more popular than PM in London

Is it me or are the media giving more legitimacy to the YouGov polling than the others? Is it because they allow them to write better headlines or because they genuinely think that they may be more authoritative?

 

[bi0boy]

or because they genuinely think that they may be more authoritative?

Pretty sure the Evening Standard don't have a team of statisticians assessing the merits of each polling company.

 

[Teaboy]

Standard are reported that Corbyn is now more popular than May in the capital, and on a 50% share of the vote - which is the same as in 97

Shock surge for Labour as Jeremy Corbyn now more popular than PM in London

Love the photo of May used in that article.

 

[killer b]

Is it me or are the media giving more legitimacy to the YouGov polling than the others? Is it because they allow them to write better headlines or because they genuinely think that they may be more authoritative?

it's because there's more of a story there.

 

[Teaboy]

it's because there's more of a story there.

Yeah, keen to talk up what could be an otherwise dull election.

 

[killer b]

Yeah, keen to talk up what could be an otherwise dull election.

I think it's stopped being dull a while ago...

 

[J Ed]

Graunid had a story where they called up four pollsters, one was Yougov. 2 said the Tories would have a 50+ seat majority and another 100+. Real horror show potential if the Yougov one isn't right.

 

[Teaboy]

Graunid had a story where they called up four pollsters, one was Yougov. 2 said the Tories would have a 50+ seat majority and another 100+. Real horror show potential if the Yougov one isn't right.

Most things do seem to point to a healthy tory majority. 100+ seats would be a shocker though.

 

[ItWillNeverWork]

I seriously don't know what to think anymore.