OK thanks. How many pollsters is it? The number of polls is redundant since only the most recent from each pollster actually effects the numbers - it is the number of pollsters that matters.
EG if there are 5 YouGov, 2 Populus, 1 ICM and 1 Panelbase polls in your SPUD then that's 9 polls (5+2+1+1) but it's only 4 pollsters (YouGov, Populus, ICM and Panelbase).
SPUD only takes into account the most recent YouGov and Populus polls, so to get an average per pollster we need to divide by 4 not 9. The first for YouGovs and the first Populus are disregarded by the formula used.
Given the polls stubbornly refuse to move in the direction of the Tories and insist on at best showing what is basically a tie, on average, I think we'd already moved on to the 'the polls are all wrong' stage of the losing campaign message.
Paddy Power has still not settled its "Size of the House of Commons" market which was a big talking point on here back in 2012 before the boundary changes were scrapped.
Many people (including me) backed 650 seats at prices a bit better than evens.
Seems ridiculous that this still hasn't been settled.
Paddy Power has still not settled its "Size of the House of Commons" market which was a big talking point on here back in 2012 before the boundary changes were scrapped.
Many people (including me) backed 650 seats at prices a bit better than evens.
Seems ridiculous that this still hasn't been settled.
You never know, HM may decide to dissolve the council permanently...*winks at Sunil*
OK thanks. How many pollsters is it? The number of polls is redundant since only the most recent from each pollster actually effects the numbers - it is the number of pollsters that matters.
EG if there are 5 YouGov, 2 Populus, 1 ICM and 1 Panelbase polls in your SPUD then that's 9 polls (5+2+1+1) but it's only 4 pollsters (YouGov, Populus, ICM and Panelbase).
SPUD only takes into account the most recent YouGov and Populus polls, so to get an average per pollster we need to divide by 4 not 9. The first for YouGovs and the first Populus are disregarded by the formula used.
Make sense?
You are asking me questions that are easily answered by this
Paddy Power has still not settled its "Size of the House of Commons" market which was a big talking point on here back in 2012 before the boundary changes were scrapped.
Many people (including me) backed 650 seats at prices a bit better than evens.
Seems ridiculous that this still hasn't been settled.
Oh please everyone accept my most humble apologies for not updating SPUD last night... I had a date would you believe?!
After todays Panelbase, the scores are
Con -9 Lab -4 UKIP +8 LD +1 Green -4
But the UKIP fade/Tory campaign success "feeling" is the PB meme
I'm curious how you've got this figure as I can't think of any poll that's had the Tories down 9 or 7, how are you calculating this?
Its the aggregate changes on the week calculated by comparing each poll to the previous one from that pollster.
Right so if a 1% change is shown by 6 different pollsters then that shows a 6% change rather than 1%?
One suggestion if you're open to it would be to divide the totals by the number of pollsters in your sample.
Yes that's it
And, yes I agree. That's why I try to mention how many polls it is based on so people can divide by that if they like, to show an average per poll
Is it the latest YouGov that get's counted? Or the first one that week?
And if it's the latest, is it the difference between that one and the previous one this week, or that one and the one from last week?
Honestly there are no traps! It is as simple as this
I measure the difference between scores from each poll and the last poll carried out by the pollster.. I haven't had occasion to worry about same pollster/diff newspaper yet, but will prob compare to the last one from that pollster and newspaper
Yeah, I only asked between those two options because they both seemed equally legitimate to me. Wasn't implying any sleight of hand
Wish I had started to do this earlier.. it would be possible, but time consuming to work backwards and compare the line with the betting moves I think
If you want I can add this to my PB polling google doc, it shouldn't take too long to add a column to compare with the previous result from the same pollster.
Or I can do this offline, and send you the results, don't want to steal your thunder!
That would be great, I am happy to share my thunder!
I only offer because I can back-calculate it to 2010 I'll work on it offline, and send you a PM when it's ready.
Righto yes that'd be great.
Then we can look at the trends and see how they compared with the thread/betting moves etc
OK thanks. How many pollsters is it? The number of polls is redundant since only the most recent from each pollster actually effects the numbers - it is the number of pollsters that matters.
EG if there are 5 YouGov, 2 Populus, 1 ICM and 1 Panelbase polls in your SPUD then that's 9 polls (5+2+1+1) but it's only 4 pollsters (YouGov, Populus, ICM and Panelbase).
SPUD only takes into account the most recent YouGov and Populus polls, so to get an average per pollster we need to divide by 4 not 9. The first for YouGovs and the first Populus are disregarded by the formula used.
Make sense?
You are asking me questions that are easily answered by this
I have no ownership of the numbers.. they are as I presented and you can dissect whichever way makes you happy
Not really as the date completed isn't the date published. I don't know which polls are included in your SPUD, there are 3 with a date completed of 20/4 onwards while eg ICM is 9 polls ago on that list.
Without knowing which polls you've included, I'm not sure how many pollsters you've included. I didn't think it'd be a difficult question to answer. To get an average figure, it'd be interesting to divide SPUD by pollsters (NOT polls).
Paddy Power has still not settled its "Size of the House of Commons" market which was a big talking point on here back in 2012 before the boundary changes were scrapped.
Many people (including me) backed 650 seats at prices a bit better than evens.
Seems ridiculous that this still hasn't been settled.
You never know, HM may decide to dissolve the council permanently...*winks at Sunil*
Will you be re-organizing the UK into the First Galactic Empire - for a safe and secure society?
I've been wondering if that's an illusion, and come the day women will vote the same as the men.
We'll find out soon enough.
It could be that men in low/ unskilled work are feeling the squeeze from the immigrants (or those voting UKIP are) and they haven't told the missus yet. Just a thought.
I've been wondering if that's an illusion, and come the day women will vote the same as the men.
We'll find out soon enough.
You mean they should do as they're told? That's a bit off.
No, I mean they're answering differently to men from the same demographic.
During the Scottish referendum there was a marked gender split in the polling. The conclusion offered by the pollsters was that women are more 'risk averse'.
Given the polls stubbornly refuse to move in the direction of the Tories and insist on at best showing what is basically a tie, on average, I think we'd already moved on to the 'the polls are all wrong' stage of the losing campaign message.
That all depends on which polls are wrong!
If you vote Tory, then Panelbase, TNS, Populus and Ipsos are wrong.
If you vote Labour, then ICM, Ashcroft, Comres, Survation and Opinium are wrong.
Yougov swing both ways depending on the mood their methodolgy wakes up in each morning. They're a love 'em and leave 'em kinda poll.
With the recent signs of how poorly the Lib Dems are doing in the South West, I've dragged my predictions for Lib Dem seats down to below 20.
Bristol West will be a ray of sunshine for Labour in the Southwest.
I'm interested to see how Labour will do in Stroud. The Ashcroft poll a while back gave them a healthy lead, but the Kingswood/NESomerset polls are a little at odds with that.
I've been wondering if that's an illusion, and come the day women will vote the same as the men.
We'll find out soon enough.
You mean they should do as they're told? That's a bit off.
No, I mean they're answering differently to men from the same demographic.
During the Scottish referendum there was a marked gender split in the polling. The conclusion offered by the pollsters was that women are more 'risk averse'.
I'm translating 'risk averse' to 'shy UKIP'.
Right example, wrong conclusions. Yes there was a marked gender split but the polls were correct in picking this up, because the final answer matched quite closely to the polls. Its not a case that the pollsters incorrectly picked up the women's answers.
Women probably are more "risk averse", so they're simply disproportionately not UKIP voters just as they were disproportionately not Yes voters.
I've been wondering if that's an illusion, and come the day women will vote the same as the men.
We'll find out soon enough.
You mean they should do as they're told? That's a bit off.
No, I mean they're answering differently to men from the same demographic.
During the Scottish referendum there was a marked gender split in the polling. The conclusion offered by the pollsters was that women are more 'risk averse'.
I'm translating 'risk averse' to 'shy UKIP'.
I was pulling your leg. Should have added one of these:
True on the risk averse. Historically women were more likely to be Tory voters...
My concern about the legitimacy and stability of the new UK government from 8th May - of whatever hue - grows daily.
A period without legislation might do the country good :-)
To be fair, you could say a Parliament unable to pass major legislation is actually what the country wants. Clearly people don't have much enthusiasm about either parties' offerings right now, so maybe it's for the best that we essentially just have civil servants running things (just doing the basics and keeping the lights on) until one or other of the parties has actually come up with something good and which the public is prepared to give a mandate to.
With the recent signs of how poorly the Lib Dems are doing in the South West, I've dragged my predictions for Lib Dem seats down to below 20.
Bristol West will be a ray of sunshine for Labour in the Southwest.
I'm interested to see how Labour will do in Stroud. The Ashcroft poll a while back gave them a healthy lead, but the Kingswood/NESomerset polls are a little at odds with that.
Stroud looks like it has a mass of soft yellows. It's a seat Labour should have been working very, very hard.
My concern about the legitimacy and stability of the new UK government from 8th May - of whatever hue - grows daily.
A period without legislation might do the country good :-)
To be fair, you could say a Parliament unable to pass major legislation is actually what the country wants. Clearly people don't have much enthusiasm about either parties' offerings right now, so maybe it's for the best that we essentially just have civil servants running things (just doing the basics and keeping the lights on) until one or other of the parties has actually come up with something good and which the public is prepared to give a mandate to.
An interesting point. I bet it won't stop whoever wins (or whichever combination of parties who win) talking about the people having spoken clearly for change/more of the same and all that cobbler though.
Reporting of the IFS report is getting ridiculous. Sky news are spending their time claiming that falls in Spending as a percentage of GDP are "cuts". The utter idiocy and numerical illiteracy of their reporting is bad even for Sky.
With the recent signs of how poorly the Lib Dems are doing in the South West, I've dragged my predictions for Lib Dem seats down to below 20.
Bristol West will be a ray of sunshine for Labour in the Southwest.
I'm interested to see how Labour will do in Stroud. The Ashcroft poll a while back gave them a healthy lead, but the Kingswood/NESomerset polls are a little at odds with that.
Stroud stands out as being a bit different - they have a fair few Green councillors.
Seems like ages since the Easter Sunday ELBOW that put the Tories 0.4% ahead.
Wait until the Tory voters come home from their Easter break, Tory leads AOTS.
Perhaps they are all now so well off after the economic recovery that has swept the nation that they have taken the whole month off. Just you wait till they are all back in May then we'll see the Tories 10% ahead!
From memory I think in the last Presidential Election in 2012 every state went to the favourite- which probably reflects better state by state polling than constituency polling in GB.
With the recent signs of how poorly the Lib Dems are doing in the South West, I've dragged my predictions for Lib Dem seats down to below 20.
Bristol West will be a ray of sunshine for Labour in the Southwest.
I'm interested to see how Labour will do in Stroud. The Ashcroft poll a while back gave them a healthy lead, but the Kingswood/NESomerset polls are a little at odds with that.
Stroud stands out as being a bit different - they have a fair few Green councillors.
Mike - if you are not going to defend political markets, who is? They can be better than the alternatives, particularly pundits, even expert's political judgement!
Just flicking through, some marvellous little details along the way:
Para592...
This will prove quite the read, I have no doubt.
Rotherham.. people said for years that Asian men were abusing underage girls They were ignored, called racist... they were right
Tower Hamlets... people said for years that the voting was fraudulent and Rahman was corrupt They were ignored, called racist... they were right
The judge in this case seems to consider such reticence a major concern to be addressed, given several references in the afterword, which seems more likely to be read than much of the underlying detail in this case. I suspect the final paragraph will be much quoted:
para686 Events of recent months in contexts very different from electoral malpractice have starkly demonstrated what happens when those in authority are afraid to confront wrongdoing for fear of allegations of racism and Islamophobia. Even in the multicultural society which is 21st century Britain, the law must be applied fairly and equally to everyone. Otherwise we are lost.
Anyone in the real world with an open mind has known this for years.
Dogmatic multiculturalists just refused to see it. Trevor Phillips, the leading Multiculturalist for many years, should be knighted for having the guts to admit it was all bollocks
I will freely admit I was probably one of them. The first time I recall hearing someone talk about grooming gangs of asian men it was the leader of the EDL I believe, and obviously he was presented as and I took him to be, a racist nutbag. Now, I'm not about to start taking what they say as gospel, or think in anyway their 'solutions' are the way to go, but I like to think now if such claims emerged, I would not instinctively dismiss them just because of the source.
Kle - don't beat yourself up. In my line of work, it's quite common to find people dismissing a message precisely because of who delivers it. They refuse to pay attention to what is being said because of who is saying it and, while someone may well have quite malicious or mixed motives for saying something, that is emphatically not a reason for refusing to look properly at what is being said.
Very hard to distinguish between message and messenger.
There again, there are none so blind as those that don't want to see.
"On election day 2010 the betting markets had CON with a 100 seat lead – it finished up at 49 seats."
So this time, then, it's looking like a CON majority or a CON-LD coalition.
It's remarkable how so many pundits are assuming that the difference between polls and the actual results will be tiny. Why are they doing that? Can't they bear the thought of another CON-LD coalition?
If LAB had any sense, they'd send all their heavyweights to Scotland. People talk about Holyrood 2016, but we're still only 6 months away from the Sep 2014 indyref in which hundreds of thousands of people who wouldn't otherwise have voted trekked to polling stations to stop the SNP.
Comments
EG if there are 5 YouGov, 2 Populus, 1 ICM and 1 Panelbase polls in your SPUD then that's 9 polls (5+2+1+1) but it's only 4 pollsters (YouGov, Populus, ICM and Panelbase).
SPUD only takes into account the most recent YouGov and Populus polls, so to get an average per pollster we need to divide by 4 not 9. The first for YouGovs and the first Populus are disregarded by the formula used.
Make sense?
Many people (including me) backed 650 seats at prices a bit better than evens.
Seems ridiculous that this still hasn't been settled.
YG-only = Lab lead 0.6%
non-YG = Lab lead 0.9%
Official ELBOW = Lab lead 0.7%
UKIP seem OK with men - but not with women.
Undecided voters (130 sub sample) in forced vote:
"Q4b. If the election was today, and you were standing in the polling booth right now, how would you vote?"
Con 23%, Lab 15%, UKIP 13%, Green 8%, LD 5%.
http://en.wikipedia.org/wiki/Opinion_polling_for_the_2015_United_Kingdom_general_election
I have no ownership of the numbers.. they are as I presented and you can dissect whichever way makes you happy
We'll find out soon enough.
Then we can look at the trends and see how they compared with the thread/betting moves etc
Without knowing which polls you've included, I'm not sure how many pollsters you've included. I didn't think it'd be a difficult question to answer. To get an average figure, it'd be interesting to divide SPUD by pollsters (NOT polls).
During the Scottish referendum there was a marked gender split in the polling. The conclusion offered by the pollsters was that women are more 'risk averse'.
I'm translating 'risk averse' to 'shy UKIP'.
BRAN (titters)
If you vote Tory, then Panelbase, TNS, Populus and Ipsos are wrong.
If you vote Labour, then ICM, Ashcroft, Comres, Survation and Opinium are wrong.
Yougov swing both ways depending on the mood their methodolgy wakes up in each morning. They're a love 'em and leave 'em kinda poll.
Women probably are more "risk averse", so they're simply disproportionately not UKIP voters just as they were disproportionately not Yes voters.
True on the risk averse. Historically women were more likely to be Tory voters...
https://t.co/shX75IL3Kt
(via @TimMorton2 )
(Yeah, i know it's a joke account!)
Rumours of the forthcoming enormo-haddock coup are mostly inaccurate. Do not be alarmed, citizens.
Hat. Coat. Taxi for one!
Stroud National VI= C 33 L 34 (Ashcroft national that week Lab +4)
Stroud Constituency VI = C30 L41
Looking at the data it seems that a lot of D/K and refused split to Labour on the second question. This turned neck and neck into a big lead.
There's little sign of additional switching from LD/UKIP/Grn.
Very hard to distinguish between message and messenger.
There again, there are none so blind as those that don't want to see.
So this time, then, it's looking like a CON majority or a CON-LD coalition.
It's remarkable how so many pundits are assuming that the difference between polls and the actual results will be tiny. Why are they doing that? Can't they bear the thought of another CON-LD coalition?
If LAB had any sense, they'd send all their heavyweights to Scotland. People talk about Holyrood 2016, but we're still only 6 months away from the Sep 2014 indyref in which hundreds of thousands of people who wouldn't otherwise have voted trekked to polling stations to stop the SNP.