Political polling

[bi0boy]

Political polling is even less of a science than economic forecasting.

 

[brogdale]

So, again, the polls were wrong.

Yes, with a few 'honourable' exceptions.
Though, this was of course something of an exceptional event.

 

[weltweit]

The polls got the numbers right, they just transposed the numbers onto the wrong positions!

Will political polling survive this debacle?

 

[Libertad]

The polls got the numbers right, they just transposed the numbers onto the wrong positions!

Will political polling survive this debacle?

Who gives a fuck?

 

[killer b]

The polls got the numbers right, they just transposed the numbers onto the wrong positions!

Will political polling survive this debacle?

Of course it will.

 

[weltweit]

Of course it will.

They got it wrong at the last general election, they got it wrong at the referendum .. will you trust them in future? I doubt I will.

 

[killer b]

Bullshit. You'll be poring over them, same as ever. What's the alternative?

 

[Fez909]

Bookies got it wrong as well. So much for their wisdom.

Obviously they still made boat loads of cash

 

[weltweit]

Bullshit. You'll be poring over them, same as ever. What's the alternative?

Waiting for the actual result !

 

[kabbes]

This all still holds, and the leave camp have done nothing to address it.

The economy is always the primary thing that drives the population in the closed ballot booth, where nobody can see our selfish, scared, short-term choices. People are entirely understandably worried about their world collapsing around them. It's easy to be declamatory and principled when making arguments but when decision time hits and the X has to go into the box, it all gets very real. This referendum is no different.

Of course, I am wrong a lot when I make predictions about elections, so you should bear that in mind

well I was right about one thing: I get it wrong a lot.

 

[The Boy]

Bookies got it wrong as well.

No they didn't. The punters did.

 

[Fez909]

No they didn't. The punters did.

Bookies don't lose cash. They don't predict outcomes based on odds of success, they hedge the odds against their potential losses. They will be up overall.

 

[The Boy]

Bookies don't lose cash. They don't predict outcomes based on odds of success, they hedge the odds against their potential losses. They will be up overall.

Which was point.

 

[The Boy]

And more specifically it was the big punters who got it wrong - according to William Hill, almost 70% of bets placed were for leave.

 

[Fez909]

Which was point.

whoops. misread your post, sorry.

 

[killer b]

well I was right about one thing: I get it wrong a lot.

Innit. I've called every major election wrong for the past 10 years.

 

[discokermit]

i won ninety two quid!

 

[Fez909]

This isn't a Brexit thread

 

[killer b]

everything is a brexit thread today, soz.

 

[The Boy]

And more specifically it was the big punters who got it wrong - according to William Hill, almost 70% of bets placed were for leave.

Just to come back to this for a minute. As I understand it, one of the reasons analysts like using betting markets as a bellweather is that people are actually risking money to back up their opinion.

It would appear, if the guy from William Hill that I follow on twitter is correct, that the overwhelming majority of punters were backing leave but that this was being skewed by the fact that large investors were backing remain and warping the markets. I saw a figure that 69% of bets were backing leave, but 69% of total value was for remain.

 

[brogdale]

They got it wrong at the last general election, they got it wrong at the referendum .. will you trust them in future? I doubt I will.

Did they?

If I were TNS or Opinium I would be rather annoyed today. Looking through social media, twitter and so forth there are lots of comments about the polls all being wrong and it being a terrible night for the pollsters, etc, etc. Both TNS and Opinium had final call figures of REMAIN 49%, LEAVE 51% – within a point of the actual result. Far from being a terrible night, they got it pretty much spot on, and should be getting congratulated.

Since the error in the polls in 2015 I’ve said that the problems won’t be solved overnight. Pollsters are experimenting with different methods. Some of those things will work, some will not – it is a learning process. The record of polls conducted online is getting more promising – the performance of the mostly online polls at the May elections was mostly good, and most of the online polls for the EU referendum were either good, or at least only a few points out. While the problems of 2015 are probably not entirely cured yet, online companies are showing clear progress, for some phone polls there is clearly still work to be done.

 

[bi0boy]

Both TNS and Opinium had final call figures of REMAIN 49%, LEAVE 51% – within a point of the actual result. Far from being a terrible night, they got it pretty much spot on, and should be getting congratulated.

Should they be? With so many pollsters giving such a broad range of results, it was inevitable that some would be closer to the result than others. This doesn't make them more competent or mean that their methodology is better than others. Were these two pollsters' turnout, regional and demographic predictions were also spot on? If not then it makes it more likely they fluked it.

 

[Louis MacNeice]

Ipsos Mori gives Labour a 5% lead...surely an outlier!

Cheers - Louis MacNeice

 

[killer b]

They didn't really did they? I think the unweighted figures had a 5 point lead, the weighted figures had them level.

 

[free spirit]

They didn't really did they? I think the unweighted figures had a 5 point lead, the weighted figures had them level.

the weight total is also 38% to 33% labour leading., page 2

eta - ah, that includes all voters, the headline figure for it only includes those who're most likely to vote.

 

[killer b]

It's not the final figure though is It? The final figure has them level (actually tories slightly ahead), once likelihood to vote is factored in.

 

[killer b]

The figures the polling companies produce are predictions based on the data collected and loads of other variables - you can't just point at the figures they've based their prediction on as if they mean much by themselves, just because you want them to be true.

 

[brogdale]

Mods, can we please now move this thread back to the Politics forum?
editor
Thanks.

 

[editor]

Mods, can we please now move this thread back to the Politics forum?
editor
Thanks.

Please use the report post function if you wish to get the attention of the mods. Thanks.

 

[Louis MacNeice]

The figures the polling companies produce are predictions based on the data collected and loads of other variables - you can't just point at the figures they've based their prediction on as if they mean much by themselves, just because you want them to be true.

I certainly wasn't posting it because I wanted it to be true. I was just surprised at the raw data given what's in the news at the moment. Also Mori provide a caveat for their adjusted figure; i.e. they're in the middle of a review of their own practices. So all in all 'they're very close' seems to be the best bet at present.

Cheers - Louis MacNeice