• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

May 7th | UK General Election 2015 OT - Please go vote!

Status
Not open for further replies.
M

Miles Quaritch

Member
pulsemyne said:
OMG at the torygraphs fuck up today. The following is tomorrows front page
CDjCHg9WgAIen4q.jpg

Trouble is the forgot about the letters PDF properties.
https://twitter.com/andyjameshicks/status/592445020227141634
Turns out it was from CCHQ and it's been running for weeks on their website.
CDjGsRWWIAEpYUk.png

So lets get this straight, The telegraph makes a big headline about a letter which turns out to be from CCHQ which was comprised of names taken from the tories website. They really do have some shitty journos over there.
Click to expand...

pulsemyne said:
Oh it gets funnier as apparently some of those businesses aren't even running anymore and at least one of the names is an ex employee of CCHQ! Hahahahaha!
Click to expand...

Man, where does the Telegraph go from here. They've haven't bothered trying to hide their obvious bias, their reputation as a broadsheet seems to be in tatters with this being the last straw.

This is pure redtop territory from the Telegraph.

Such a sad sight... Is what I would be saying if KT wasn't so hilarious.
 
K

kmag

Member
CyclopsRock said:
I heard a suggestion that, thanks to the Lib Dem collapse and the fact that mostly all Tory marginals in Scotland are with the LD's that les bleus could end up actually picking up a seat in Scotland. That'd be pretty lolsome.
Click to expand...

Yeah sort of, the most likely result is that they stay at 1 but the constituency they win changes. They actually have a punchers chance in 3 seats. potentially 4 (although that's a 3 way long shot). The Borders are essentially almost exactly the sort of seat in terms of demographics that Tories easily win in rural Cumbria.
 
K

kmag

Member
27/04 Populus UK VI

With changes from 24 April Populus Online Poll
LAB 36% (+1) CON 33% (+1) LDEM 8% (=) UKIP 14% (=) GRNS 5% (=)
 
K

Kathian

Banned
I think think they might take 55/56. Scottish politics really is why FPTP is rubbish. Tories and Labour killed by majorities in each constituency.
 
K

kmag

Member
CyclopsRock said:
Tasty, digestible format:

CDm3vOKWEAEaQUq.png:large
Click to expand...

Those Ashcroft national numbers barely seem credible. Especially given his constituency polling. He's got Labour 6% ahead of the Tories in Cannock Chase (a seat which required a 3.5% swing to Labour for them to win), a 6.5% swing doesn't seem feasible in terms of his national figures, and Labour two points behind in Great Yarmouth (74th on Labours target list) which requires a near 4% Con to Lab swing. And a 8% swing in Great Grimsby. A 2% national swing doesn't really make much sense, especially given other national polling. On the other hand the ICM figures I can being pretty close the final scores, maybe falling between ICM and YouGov.
 
K

kmag

Member
sampling_error.png


Interesting graph from LSE.

What is striking is just how wide the band is. A poll with a sample size of 500 has an interval that ranges from -3.9 to 7.9, while the interval for a sample size of 1500 still ranges from -1.4 to 5.4. The difference between a 5 point Labour lead and a 1 point Conservative lead is politically substantial, but even with 1500 respondents that is the level of variation we can expect to see from poll to poll.

The plot also shows the average poll size for all the major pollsters. While the pollsters ComRes, Opinium, Populus and YouGov record preferences from a relatively large number of respondents in their online surveys, this is not the case for polls done over the telephone. Notably, ICM and TNS have some of the lowest sample sizes of respondents stating a preference, the two pollsters which showed such divergent results last week. The implication of this is that polls from these companies are likely to display higher levels of volatility than those with larger effective sample sizes. When it comes to sampling variation, the online pollsters have a clear advantage.
Click to expand...

pollster_forecasts_Conservatives.png


pollster_forecasts_Labour.png


Given this, and in order to help interpret new polls as they are released, the plots below depict the 90% predictive intervals for each party, for each pollster, given their typical sample size (of voters that express a preference), their house effects, and our current national pooled poll estimates. These are estimates of what polls from these pollsters might show if they ran a survey over the last few days. The vertical lines in the graph are the prediction intervals, and the points are actual poll results from the past 10 days.

Two things stand out from these plots. First, we can see the variances for each pollster. As we saw in the plot of the Labour-Conservative margin, the prediction intervals are wide, particularly for the larger parties. This reflects the fact that there is a good deal of uncertainty in individual polls. Second, we can see the relative biases of each pollster. Almost all recent polls fall within the prediction intervals for the pollster that conducted them. This suggests that even polls that seem to show relatively large changes have been within the expected range of estimates for that pollster.

How should these prediction intervals inform our interpretation of new polls? First, if a new poll comes out, and it is inside the range we show for that pollster for all parties, this is an indication that the poll is not surprising given that pollster’s house effects and where the polls are more generally. To reiterate: this can occur even if the pollster shows a substantial swing from their previous poll, as there is substantial uncertainty in every poll.

Second, if a new poll comes out, and it is outside the range we show for that pollster for one or more parties, there are two possible interpretations. Either this is one of the 1 in 10 polls that will inevitably fall outside a 90% prediction interval, or there is some recent movement in the support for the parties that our model has not yet incorporated. While the latter is the more exciting interpretation, the former will usually be the correct one.

So, if a new poll is inside our prediction interval for that pollster, it probably indicates nothing has changed, and if it is outside our prediction interval for that pollster, it still probably indicates nothing has changed. This reflects the fact that, in general, no single poll is going to give a very strong indication that public opinion has actually changed. As these plots show, individual polls are too uncertain, relative to the magnitude of real swings in voting intention. The only way we can be confident that change has occurred is if we see evidence from many polls, each taken in the context of what the corresponding pollsters had found in their previous polls.
Click to expand...

http://blogs.lse.ac.uk/generalelection/interpreting-new-polls/
 
Q

QuicheFontaine

Member
Just as I walk in the door, Radio Wiltshire on in the kitchen:

"Earlier today Ed Miliband revealed that when he was seven years old his dream job was nothing to do with politics. He wanted to be a bus conductor".

What is this?
 
P

pulsemyne

Member
I think that phone pollsters are a bit old hat considering that there could be quite a few shy UKiP voters who don't want to say on the phone what they intend to vote. Online gives there much more anonymity in voting. It's been interesting to note just how low UKiP is in phone polls compared to online.
 
C

CyclopsRock

Member
pulsemyne said:
I think that phone pollsters are a bit old hat considering that there could be quite a few shy UKiP voters who don't want to say on the phone what they intend to vote. Online gives there much more anonymity in voting. It's been interesting to note just how low UKiP is in phone polls compared to online.
Click to expand...

It's also possibly because they're second maybe only to the SNP nationalists in having a barmy, out-of-proportion gang of retards online.
 
G

giant_frying_pan

Member
David Coburn says UKIP aren't right-of-centre and are a libertarian party. Lawl. Maybe in the backward, American sense of the word perhaps.

Still, brave of him to walk about Parkhead instead of Bridgeton so fair play to him.
 
M

mclem

Member
CyclopsRock said:
I heard a suggestion that, thanks to the Lib Dem collapse and the fact that mostly all Tory marginals in Scotland are with the LD's that les bleus could end up actually picking up a seat in Scotland. That'd be pretty lolsome.
Click to expand...

I'm not quite clear why, if you're pissed off with the LDs for entering into a coalition, you'd vote for the other party in that coalition. Are these polls basically dependent on the LD voters just not being arsed to go vote?
 
C

CyclopsRock

Member
mclem said:
I'm not quite clear why, if you're pissed off with the LDs for entering into a coalition, you'd vote for the other party in that coalition. Are these polls basically dependent on the LD voters just not being arsed to go vote?
Click to expand...

Partly, but also the fact that in many of those sorts of seats, the 3rd placed party (be it Labour or SNP) is simply so far behind that even if a full half of LD support went to them, it still wouldn't be enough. In all likelihood it'll be far less than that.
 
I

IdreamofHIME

Member
Dambrosi said:
Er, no. I'll just fix that image up for you, just a sec.
.
Click to expand...



Fixed it for you
eCSKT9a.png


Miles Quaritch said:
Man, where does the Telegraph go from here. They've haven't bothered trying to hide their obvious bias, their reputation as a broadsheet seems to be in tatters with this being the last straw.

This is pure redtop territory from the Telegraph.

Such a sad sight... Is what I would be saying if KT wasn't so hilarious.
Click to expand...


They don't do anything, sadly no one cares. Both the BBC and ITV ran stories in their early evening news about the small businesses begging the tories to stay in power, but not a word that it was the Tories begging the small businesses to sign up.
 
K

kmag

Member
Had a look at Ashcroft's poll. It's almost certainly a bit of a outlier. It's got a massive oversample of AB's and has females voting Tory over Labour by a massive degree which isn't picked up in any other poll including ones with Tory leads.
 
M

Miles Quaritch

Member
IdreamofHIME said:
They don't do anything, sadly no one cares. Both the BBC and ITV ran stories in their early evening news about the small businesses begging the tories to stay in power, but not a word that it was the Tories begging the small businesses to sign up.
Click to expand...

Well that's depressing. Seems Marr is the only one to call them on it.

Anyway, not sure if this has been posted...

http://fivethirtyeight.com/interactives/uk-general-election-predictions/

fivethirtyeight have this up. Was referenced in the panorama special.
 
P

pulsemyne

Member
kmag said:
Had a look at Ashcroft's poll. It's almost certainly a bit of a outlier. It's got a massive oversample of AB's and has females voting Tory over Labour by a massive degree which isn't picked up in any other poll including ones with Tory leads.
Click to expand...

Yeah I just looked at it. Really iffy poll. Anything reporting a big lead I think needs to be treated as an outliner.
 
I

IdreamofHIME

Member
K

kmag

Member
New Survation poll for Daily Record
SNP 51% (+4);
LAB 26% (NC);
CON 14% (-2);
LD 5% (+1);
OTHER 4%
Seats by Uniform National Swing: SNP 55; LAB 3; LD 1; Tories 0

Second poll today putting SNP above 50%
 
P

pulsemyne

Member
kmag said:
YouGov/Sun:
CON 35 (+2)
LAB 34 (=)
LIB 9 (+1)
UKIP 12 (-2)
GRN 5 (=)
Click to expand...

Looks like more polldrums. Labours total seems very steady on yougov. There does seem to be a small shift to the tories in the polls but nothing major (I'm discounting ashcrofts poll as that screams outliner). Still all within MoE.
 
Status
Not open for further replies.

Similar threads

  • Locked
MI5 warning over 'Chinese agent' in Parliament
News
2
1K
Jose92 replied
  • Locked
[TNYT] Brazil Blocks X After Musk Ignores Court Orders
Social News Community 2
69
5K
ManaByte replied
Bobby Kotick, former CEO of Activision, Interested in Buying Tik-Tok
2 3
143
11K
Gameplay Gods Bless replied
  • Locked
UK General Election - 8th June 2017 |OT| - The Red Wedding
396 397 398 399 400
20K
2M
chromatic9 replied
Austrian Elections 2017 |OT| Voting right wing parties into parliament since 1949
4 5 6 7 8
377
44K
Red Arremer replied
Top Bottom