Categories
Options
politicalbetting.com » Blog Archive » The Scottish Independence Referendum becomes a clash of pol
politicalbetting.com » Blog Archive » The Scottish Independence Referendum becomes a clash of polling methodologies
There was a time when you asked YouGov’s Peter Kellner why his figures were very different from other firms he would respond by saying that he never commented about how other firms operated. No more.
0
This discussion has been closed.
Comments
Let me be clear what I am NOT saying. I do not accuse Survation, or any other company, of any intention to bend the truth. On the contrary, I am certain that they try their hardest to publish data that is as accurate as possible. They – and we – know that any systematic bias is likely to result in the embarrassment of inaccurate final polls, published during the week of the referendum.
The comments after the blog post, as ever, say more about the poster than the article itself......you can feel one former member of this parish' self importance radiating from the page.....
Is that self imporkance?
http://www.scotsman.com/news/politics/top-stories/labour-candidate-resigns-over-hitler-youth-tweet-1-3462378
His final paragraph is a total cop out: Scotland will vote decisively to stay in the UK, unless it doesn't.
The proper full post is now up
"I have been divorced more than once. Trust me it is never ever amicable, whatever anybody tells you. But you can make a deal. You can give the partner who is walking out on you all the CDs the DVDs, the dog, the car – you can give them everything, but the one thing you will never ever give them is the right to continue to use the joint credit card.
And that is what their plan A – and they have no plan B – amounts to.
They want to use a currency issued by the Bank of England – the clue being in the name; they want to continue to use it and they imagine that the people that issue it will allow them to do so; to use the joint credit card, even though and as they are walking out the door.
So this is the first time ever that people in a small country, where everyone speaks the same language, are being asked to break up and break up on the basis that they don't have a currency to use.
There will be no pound. Trust me on that. I came yesterday from Parliament (where) the leaders of the mainstream parties have not changed their minds. An independent Scotland will not have the pound. "
In 2011 SLAB were led by a sub Miliband character called Gray who was spectacularly unimpressive. Salmond, in contrast, had led a minority administration with considerable skill, not rocking the boat unduly, working for consensus in the Parliament and generally, along with his Ministerial team coming over as competent. Given a choice of Salmond or Gray many Scots who would not normally vote SNP went Salmond. One of the reasons they would not normally vote SNP is that they did not support independence.
OTOH I think it is equally clear that to seek to weight results based on 2010 is to make the same mistake in reverse. The question SLAB put to the Scottish people in 2010 was do you want a Tory government in Westminster and Scotland gave its usual resounding answer to that. Many SNP inclined supporters will no doubt have lent their votes to the cause. So weighting on 2010 is, in my view, equally suspect.
Doing anything useful with the Euro results is practically impossible given the level of turnout.
For me, this shows the difficulty in applying party support weighting to a one off question like Independence. People vote for a whole range of reasons at each election. As I have said on here before the Independence campaign is much bigger and broader than Salmond. There is, I think, a strong strand in this campaign of putting it to the English and a determination not to be bullied. There is also the anti tory strand so carefully developed by SLAB over a generation.
These movements make for strange bedfellows. In my limited canvassing for Better Together I came across a surprising number of 2011 SNP voters who were voting no, supporting Kellner's premise. But there is support across the parties for Independence, particularly amongst the have nots and the hard done bys who might be thought to be traditional Labour supporters.
In short whilst I understand Kellner's argument I think traditional weighting is even more problematic for this question than he is acknowledging. It frankly worries me that ICM are in the closer result camp. They are the Gold Standard and I would love to read their response.
If the L/Ds don't realise they are going down,down,down-to quote Status Quo-they should now.The only arrow left in their bow is to sack the manager and try for a new start.
Punting-wise,I am happy with 6/4 over 250 L/D lost deposits.
Bermondsey will be tough for Labour.Hughes has shown himself to stand up to criminality in his constituency,including threats and intimidation towards him.He is a man of some courage.
Nevertheless,Labour seem to be doing especially well in London.If the local focus is on the brand rather than the individual people will vote for the party more than the man.Politics is very cruel as Peter Tatchell discovered.Hughes may still suffer in return.
As a result of all this,Bermondsey is a Lab gain at 2-1 with Ladbrokes.My first constituency bet to small stakes to go with the Popcorn and Kebabs on election night.
He is correct in believing that not everyone who voted SNP in 2011 wants independence. He ignores the fact that 2011 was the watershed election in which the SNP moved from being a party which generally took Tory and LibDem votes and for the first time took wholesale control of many former Labour voters. With the odd exception, in traditional working class areas since 2011 the SNP vote has held up incredibly well. The Aberdeen Donside by-election is a perfect example.
Those of us who are Scots have always known that the fate of the Referendum will be determined by how the traditional white working class Labour voters who live in the sprawling housing estates on the edge of Glasgow and Edinburgh and the 5 new towns across Central Scotland vote.
The majority of LibDem voters and the overwhelming majority of Tory voters are committed NO voters. A majority of SNP voters are committed YES voters. It is for the Labour party to deliver its traditional constituency for the NO camp. I fear they are failing miserably. In addition the pollsters are simply not reaching these people. Few now have landlines and even fewer are members of online voting panels. Face to face polling is the only method which is likely to be anywhere near successful in assessing how these people will vote.
Anyway, today's YouGov has "sticks to what he believes in" for Cameron shooting up from 23% to 31%. Not a bad thing for a politician these days, particularly Cameron who some people think is all over the place.
Should help turnout - which helps no.
Yes 355 No 417 54% Adjusts to 53%
Survation Feb Yes 369 No 438
Yougov
16 June Yes 348 No 586 71% Adjusts to 60% (Excl DK)
Interesting that assuming a neutral midpoint of 63% No, 37% Yes (DKs are neccesarily zero in the actual poll) the sample bias of Yougov is so far to No and Survation so far to Yes in their latest polls.
Finger in the air using a midpoint of both the raw and the cooked:
60% No 40% Yes
The YF finding that Cameron has bounced on "sticks to what he believes in" with a sample which actually has an increased Labour lead is interesting - presumably showing people trying to be fair but not attaching especial weight to that (hey, I think George Galloway sticks to what he believes in, but have no plans to vote Respect). His main problem is the mere 5% who think he's in touch with the concerns of ordinary people (Miliband 20%). Everyone has poor ratings on that, hence the general disillusion, but the Conservatives do have a particular problem with it, linked to the other frequent finding that 50% think them mainly focused on the needs of the wealthy.
I think we are in agreement that weighting by party is problematic though.
I am in agreement with YouGov based on election data from the last couple of years. The Donside byelection actually confirms YouGov case. While the SNP held the seat they lost about a fifth of their vote from the Holyrood elections. Some of those went back to Labour, some to the Lib Dems and some to UKIP. The euros were not great for the SNP and recent council by-elections have been very disappointing for them.
Almost all the opinion polls show there will be a strong No vote by Labour. I dont know EasterRoss if you have some great insight on this but I run a factory in Central Belt Scotland. Heartland Labour. The guys on our shopfloor are all No supporters both Catholic and Protestant.
While I would have Labour as favourites here, I don't think they should be a 66+% IP. People have missed a couple of things:
1. Highgate used to be part of of the Hampstead constituency, and is now part of H&WG. That's a lot of not very left leaning voters in the constituency that were not there in 2005.
2. Gospel Oak, Muswell Hill, and Alexandra Palace are increasingly affluent suburbs. These (and Highgate) are now naturally Conservative areas.
At the locals in the constituency it was about 39% Labour to 33% LibDem. While that's a gap, it is no more than in Bermondsey, and there is a larger Conservative vote to squeeze. I would suggest Labour is value in Bermondsey at 2-1, and the LibDems are the value in H&WG.
Centre and Number One Courts start play an hour earlier than usual (at midday).
Hard to say on polling methodology, as independence votes don't happen very often so it's not like the pollsters can tweak their models based on recent past results, as they can for elections.
Vote for status quo/no - not necessarily 100% committed to vote
Therefore no is likely to be overstated to some extent.
Lower the turnover, better the chances for yes
I don't see why down-weighting the SNP would help, because your sample of SNP-voters should include those who voted for them in 2011 but would not vote for Independence. You would only be justified in down-weighting the SNP if you thought those people were uniquely unlikely to be part of your polling sample - but the polls picked up the swing of these people support from Labour to SNP before the Holyrood elections in 2011, so they obviously are people who talk to opinion pollsters.
Until we know the official outcome next Sept, it’s hard to say which pollster is more right than wrong, they all agree on the same result; it’s just a matter of how big the gap is.
The only thing of interest is when comparing the two extremes: YouGov & Panelbase, both predicting a NO vote, but there’s quite a difference in projected outcome between the two – which pollster will end up with egg on their face? I suspect it will be the latter.
One of the last bastions of the Orange Lodge and Labour donkeys, what a combination.
Because in America, multiple elections are held on the same day, Senate, Gubernatorial, Congress and Presidential, and some people voted for different parties.
So how do you classify someone who voted Democratic Party for Governor and Congress but Republican for President and Senator on the same day.
Are they a republican or dem?
You could class someone who voted Lab in 2010 GE but 2011 SNP (but is planning to vote Lab in 2015) as a split ticketer.
But dealing incorrectly with split ticketers was a problem for US pollsters.
http://en.wikipedia.org/wiki/Split-ticket_voting
So false employment stats, Jedi speech,Red Len threat to damage the whole country while funding Labour's campaign, Livingstone demanding a place in a Labour cabinet ( last two not unconnected) and the paymasters demanding a return to Pleistocene Labour Practices.
Great 36 hours in convincing business Labour can be trusted with the economy.
Almost by definition YES should get their vote out whereas some No may be apathetic
So my projection:
No 54
Yes 46
Salmond beaten but not humiliated
Part of the problem is that place names in London can be a bit elastic - witness the way in which West Dulwich now suspiciously borders West Norwood - presumably the people of "East Norwood" preferred to be part of a larger West Dulwich...
That said, his mistake was not actually Bank of England vs Treasury. The Bank of England issues the currency.
His mistake (which he may have deliberately conflated to make the message simple to understand) was to suggest that the rUK cares whether Scots continue to use sterling specie after a yes vote. They don't. What they care about - very strongly - is that iScot won't be able to *borrow* in sterling secured (implicitly or otherwise) against the credit of the UK Treasury.
Point of order: there is no such district as Alexandra Palace. There is Muswell Hill on one side and Wood Green on the other.
Yes, in short.
http://www.bbc.co.uk/sport/0/formula1/28116193
Won't be Mercedes or Ferrari, as Horner suggests. If not Renault that means, probably, either Honda or making their own engine. The latter's been raised before and it could happen.
I'll get me coat.
Some of the more acutely touched in the head posters have departed, so that a blessed near-silence from this tendency now reigns.
My own feeling though is that this isn't the only reason for the silence. The Anglophobes have had to recognise that they have comprehensively lost on every single point. It's not been an inconclusive or even a close debate, in the sense that they won on some points, lost on others, and honours are only just not even. On the contrary, the Nits have lost every aspect of every argument.
They've conceded.
It's like that scene towards the end of Zulu when the plucky English soldiers are waiting, exhausted, for the next onrush of the bizarrely-clad heathen horde. And gradually it dawns on them that the silence bespeaks total victory.
Who would have predicted the rise of a self-declared caliphate this year? Or the Arab Spring?
Sure, it can be done, but would require iScot to sign up to tight restrictions on fiscal and monetary policy.
And that's not really independence, is it?
Postcode districts don't save you from London place name confusion I'm afraid.
I'll follow you out the door
Some years back a friend of mine lived in a village near Stamford and had a Stamford and Lincolnshire post code and address despite the fact the village was located in Rutland.
He insisted his post was delivered to his Rutland address with the final line the Stamford postcode. The world still turned.
While this Tinder story is really rather amusing, why on earth has the Daily Mail chosen to asterisk out the word "whore"?
http://www.dailymail.co.uk/news/article-2677289/A-heartless-w-flirts-Muslim-pigs-The-texts-lay-bare-ugly-break-heart-Tinder-sexual-harassment-suit.html
Off topic - the Metro claims that Victoria Derbyshire leaving 5Live has led to speculation she is going to anchor Newsnight...
So at 45% for Yes I'd pay you £20, and at 39% you'd pay me £40.
Or £5 a point or £1 a point, just a bit of fun really.
"It's like that scene towards the end of Zulu when the plucky English soldiers are waiting, exhausted, for the next onrush of the bizarrely-clad heathen horde. And gradually it dawns on them that the silence bespeaks total victory."
Ah, but they weren't English soldiers.
It was a Welsh regiment The soldiers were mainly Welsh. That is why they sing Men of Harlech in the movie.
Your post neatly demonstrates one of the very things that the Welsh and Scottish find so irritating about the English.
Edit: Minimum of £10 to charity sound OK ?
Edit 2: £3/point 43% Unders/Overs line Minimum £10 Maximum £30
People often get battles, and those who fight them, wrong. Some battles between 'England' and 'Scotland' largely involved Scots on both sides.
I once bought a flat in W9 which I assumed to be in Maida Vale, because Maida Vale is W9. Imagine my feelings when I later looked it up in the London A-Z, to find that my street could not be seen - because it had "West Kilburn" printed over it.
It was W9 all right, but there is W9 and there is W9. The clue was in the fact that the nearest Tube station was not Maida Vale or Warwick Avenue, but Queen's Park.
;-)
"While most of the men of the 1st Battalion, 24th Regiment of Foot (1/24) were recruited from the industrial towns and agricultural classes of England, principally from Birmingham and adjacent southwest counties, only 10 soldiers of the 1/24 that fought in the battle were Welsh. Many of the soldiers of the junior battalion, the 2/24, were Welshmen. Of the 122 soldiers of the 24th Regiment present at the Battle of Rorke's Drift, 49 are known to have been of English nationality, 32 were Welsh, 16 were Irish, 1 was a Scot, and 3 were born overseas. The nationalities of the remaining 21 are unknown"
and as for singing Men of Harlech it was artistic licence at its best but no more....
Back in the day .... well the turn of the 20th century actually Maida Vale used to the haunt of the better class of prostitutes and the homes of the mistresses of the aristocracy .... apparently.
No matter how much Scotland borrowed it could not be as profligate as London.
The commonly held view is that a Labour candidate will almost certainly be the next Mayor of London. The capital has steadily shifted towards the left as Labour control ever more councils, and in some areas the Greens are now the second largest vote-bloc, not Conservatives. Recent polling put Labour 13 points ahead in London and that lead is very steady.
But I think there is a scenario in which a Tory candidate succeeds Boris and becomes the next Mayor of London. Londoners are more likely to vote for a Mayor based on personality, charisma, competence and bold ideas than party affiliation, a truism Ken Livingstone grasped in 2000 but forgot by 2012.
Imagine that a liberal, centrist, ambitious Tory – largely untainted by associations to Cameron – comes forward as a candidate. Someone who has spoken more about the environment than going harder on the cuts. A candidate that could easily sit within the Labour party. Even highly liberal New York voted for centrist billionaire Bloomberg (a former Republican) three times before going for Bill De Blasio.
I imagine that person to be Richmond MP Zac Goldsmith. He is ambitious and well-connected, but has no future in a Tory government. He is more comfortable building cross-party alliances on issues than standing up dutifully during PMQs with praise for the PM.
Boris Johnson’s calculations are also relevant here. The current Mayor can see the electoral math and knows Cameron only has a slim chance of a majority next year. To be leader of the party Boris needs to go for a relatively safe seat, ideally in London so he can pretend to do both jobs, and position himself for the aftermath of the General Election. Richmond is ideal.
http://labourlist.org/2014/07/how-london-could-have-a-tory-mayor-in-2016-again/
Mr. X, bizarre clothing varies a lot. Not so long ago men wearing tights and tunics was the norm. Trousers were seen as 'barbarian' dress.
If Scotland were independent, then you would be borrowing against the rUK balance sheet without any democratic oversight from the rUK. That's the issue in a nutshell: it creates an unquantifiable and uncontrollable contingent liability, and that's something that even our politicians won't sign up to.
Malcolm, having joined the herd of PBers monitoring that Pop-ular site that can't be named, I see Mick Pork wanted you to be made aware that rather than leaving PB by choice he was and is still banned after a bit of anonymous clyping by one of his fellow posters.
I've sent you a PM with the finalised Bet details. I think there needs to be a slight premium on being "right" hence the minimum, the spread bet will kick in 4 -> 10 pts out, capped at 10 pts ^_~
The singing of Men of Harlech was obviously artistic license, I mentioned it simply to emphasise the point that the regiment is not English. It is a Welsh regiment, that is now headquartered at Brecon, and it always had close connections with Wales.
However, even on your figures, a majority of the soldiers were not English.
So, it is just not appropriate to describe them as "plucky English soldiers".
Is Newsnight becoming Mumsnight under Mr Mumsnet?
http://www.bbc.co.uk/news/uk-politics-28113516
Though, like Universal Credit, the DWP is claiming it is all a lie.
You have to love figures and statistics don't you? They can prove anything you want.
The over-under market offers a little value.Ladbrokes are pricing both over and under the L/Ds number of seats at 10-11 with the cut off figure 33.5.Paddy Power go 8-11& Evens respectively for 33.5 also.
After the Ashcroft polling,10-11 under, looks the right side of the curve.
Bet number 2 for GE2015.One for the portfolio.