Survation

[Andrew_S]

LAB 35 (-1)
CON 31 (+4)
UKIP 19 (-3)
LD 8 (+1)
OTHER 8 (-2)

[Devonian]

As well as doing a Scottish poll Survation also have an England & Wales poll this weekend

Labour 33%
Conservative 30%
UKIP 23%
Lib Dem 8%
Green 3%

survation.com/wp-content/uploads/2014/09/Referendum-Reactions-Results.pdf

[Pimpernal]

Latest Opinion Poll (Survation):
CON - 31% (=)
LAB - 31% (-4)
UKIP - 25% (+6)
LDEM - 8% (=)

[Deleted]

Oct 11, 2014 20:56:38 GMT Pimpernal said:

Latest Opinion Poll (Survation):
CON - 31% (=)
LAB - 31% (-4)
UKIP - 25% (+6)
LDEM - 8% (=)

Interesting...

[Devonian]

Full tables here

survation.com/wp-content/uploads/2014/10/Political-Survey-12-October-MoS-Tables.pdf

Seems to be much interest in the regional sub samples and in particular the sub sample for the South of England which is 

UKIP 37.6%
Con 33.3%
Lab 19.5%
Lib Dem 6.8%
Green 2.4%

Comment on this in the Mail

Labour and the Tories are both on 31% while the Liberal Democrats are on 8%, according to the research for the newspaper.

Experts suggest that the ratings would give Labour 253 MPs, Conservatives 187, Ukip 128, Lib Dems 11, and other parties, such as the SNP, 71.

John Curtice, professor of politics at Strathclyde University, told the Mail on Sunday : "If Ukip are to turn votes into Commons seats in Britain's first-past-the-post system, they need to build up bastions of local strength. Today's poll suggests they may have begun to do that.

"T he 25% level represents a 22-point increase on the 3% the party won in 2010.

"If that increase were to occur evenly in every constituency, they could still fail to pick up a single seat.

"But today's poll suggests Ukip's support has increased much more in the south of England outside London than it has elsewhere in the UK - by a staggering 34 points.

"If that level was recorded throughout the South, Ukip could win as many as 128 seats, with no less than 102 of them coming from the Conservatives, whose vote in the region is down 14 points.

"In that event, Cameron would be left with just 187 seats, almost as weak a position as the Conservatives were in after their calamitous defeat in 1997.

"Mr Farage would achieve his ambition of holding the balance of power at Westminster. Any poll estimate of what is going on in an individual region is inevitably not as robust as that for the country as a whole.

www.dailymail.co.uk/wires/pa/article-2788816/Bullish-Ukip-swoops-Kent-target.html

[Devonian]

The Scottish sub sample (just for fun, I know all the usual caveats about sub samples)

SNP 50.0%
Lab 30.0%
Con 9.4%
Lib Dem 4.8%
Green 2.9%
UKIP 2.8%

[Deleted]

Oct 11, 2014 22:07:07 GMT Devonian said:

Full tables here

survation.com/wp-content/uploads/2014/10/Political-Survey-12-October-MoS-Tables.pdf

Seems to be much interest in the regional sub samples and in particular the sub sample for the South of England which is 

UKIP 37.6%
Con 33.3%
Lab 19.5%
Lib Dem 6.8%
Green 2.4%

Comment on this in the Mail

Labour and the Tories are both on 31% while the Liberal Democrats are on 8%, according to the research for the newspaper.

Experts suggest that the ratings would give Labour 253 MPs, Conservatives 187, Ukip 128, Lib Dems 11, and other parties, such as the SNP, 71.

John Curtice, professor of politics at Strathclyde University, told the Mail on Sunday : "If Ukip are to turn votes into Commons seats in Britain's first-past-the-post system, they need to build up bastions of local strength. Today's poll suggests they may have begun to do that.

"T he 25% level represents a 22-point increase on the 3% the party won in 2010.

"If that increase were to occur evenly in every constituency, they could still fail to pick up a single seat.

"But today's poll suggests Ukip's support has increased much more in the south of England outside London than it has elsewhere in the UK - by a staggering 34 points.

"If that level was recorded throughout the South, Ukip could win as many as 128 seats, with no less than 102 of them coming from the Conservatives, whose vote in the region is down 14 points.

"In that event, Cameron would be left with just 187 seats, almost as weak a position as the Conservatives were in after their calamitous defeat in 1997.

"Mr Farage would achieve his ambition of holding the balance of power at Westminster. Any poll estimate of what is going on in an individual region is inevitably not as robust as that for the country as a whole.

www.dailymail.co.uk/wires/pa/article-2788816/Bullish-Ukip-swoops-Kent-target.html

Subsample=meaningless data.

[Tony Otim]

Experts suggest that the ratings would give Labour 253 MPs, Conservatives 187, Ukip 128, Lib Dems 11, and other parties, such as the SNP, 71.

Read more: vote-2012.proboards.com/post/new/1244#ixzz3Fw9lM16A

Ok - 18 Nortern Ireland
PC - maybe 4 or 5 on a good night
Green - 1
SNP - 47 ? I don't think so.

[iain]

Galloway too, so actually the SNP would only need a very achievable 46

[Sibboleth]

I think it's fairly clear that those seat numbers were just made up for sensation.

[Richard Allen]

Oct 12, 2014 15:00:16 GMT iain said:

Galloway too, so actually the SNP would only need a very achievable 46

Indeed.

Besides, if we have a UKIP surge to the extent that some polls are suggesting it really is going to throw the seat calculations out of the window.

[manchesterman]

Can anyone point me in the direction of the analysis which shows UKIP gaining 128 seats based on 25% of the vote?

I put the figures (with the Lab/Con/LD etc percentages) in the electoralcalculus predictor and they only came out with 2 seats - this assumes even swings across the country of course. Was there a more detailed breakdown which would account for the huge discrepancy? Is there a list of the "gained" seats based on this poll?

Thanks guys.

[Devonian]

Oct 13, 2014 23:14:50 GMT manchesterman said:

Can anyone point me in the direction of the analysis which shows UKIP gaining 128 seats based on 25% of the vote?

I put the figures (with the Lab/Con/LD etc percentages) in the electoralcalculus predictor and they only came out with 2 seats - this assumes even swings across the country of course. Was there a more detailed breakdown which would account for the huge discrepancy? Is there a list of the "gained" seats based on this poll?

Thanks guys.

It was based on the regional sub samples. The subsample for the South (outside London) showed UKIP in first place.

[manchesterman]

Thanks Devonian, but is there an actual breakdown stated anywhere. Do we know which 128 seats were the predicted "wins"?

[carlton43]

Oct 13, 2014 23:14:50 GMT manchesterman said:

Can anyone point me in the direction of the analysis which shows UKIP gaining 128 seats based on 25% of the vote?

I put the figures (with the Lab/Con/LD etc percentages) in the electoralcalculus predictor and they only came out with 2 seats - this assumes even swings across the country of course. Was there a more detailed breakdown which would account for the huge discrepancy? Is there a list of the "gained" seats based on this poll?

Thanks guys.

I don't think there could be any generalist model for a calculator to be based on if UKIP were to reach that level of support in the final lead-up to the GE. Such models are far too generalist with an inflator across all seats, whereas there are weak and strong areas and hot spots that could well develop very late in the campaign. Any forecasts predicated on a potential UKIP turnout of 25% or higher would suggest differential vote movement across many parties, whilst at the same time local influences could be working in contrarian directions, and the existence of local constituency polling would probably lead to massive incidence of tactical voting, mainly of a negative variety. I assume there will be considerable tactical voting against Labour, Conservative and UKIP candidates, but far less against LD and Green candidates unless they are incumbents. 

[manchesterman]

Thanks Carlton. I take on board all that you said as a general rule of thumb.... but does anyone know about these "128 seats" Survation was projecting they would/could win on 25% share of vote?

[carlton43]

Oct 15, 2014 21:45:19 GMT manchesterman said:

Thanks Carlton. I take on board all that you said as a general rule of thumb.... but does anyone know about these "128 seats" Survation was projecting they would/could win on 25% share of vote?

Even if you find it I would suggest that it is likely to be deeply flawed unless it has been most carefully nuanced to reflect differential voting switch in differing areas for differing reasons. 

[Devonian]

Tonight's Survation poll for the Mirror

Lab 31%
Con 27%
UKIP 24%
LD 9%
Other 9%

[Devonian]

Details of the Survation poll here

survation.com/wp-content/uploads/2014/11/Voting-Intention-Poll-Tables-6-November-2014.pdf


One noticale detail is the big difference in the UKIP vote between men and women. UKIP is in first place amongst men but some way behind amongst women. As might be expected the other party with a large disproportion of men/women voters is the Greens with many more women than men.

Men
UKIP 28.7%
Lab 27.4%
Con 25.8%
LD 9.0%
SNP 5.4%
Grn 2.5%

Women
Lab 34.8%
Con 29.3%
UKIP 17.6%
LD 8.8%
Grn 4.2%
SNP 3.9%

[YL]

Do we know why the questions are Q9 and Q10? I would hope that they didn't ask the EU referendum question (which they must have asked, because there are crossbreaks for it) or indeed the 2010 vote question before the voting question.