Thursday, September 18, 2014

Last night's Survation poll was better for Yes on the unrounded numbers

A small piece of good polling news as we all wait (or hopefully as we get on with the Get Out The Vote operation) - last night's Survation poll in the Record, which was one of only two polls over the last few days to report a Yes vote of lower than 48%, was actually slightly better for Yes on the unrounded figures -

Yes 47.3% (+0.8)
No 52.7% (-0.8)

Changes are from the directly comparable Survation telephone poll published on Saturday.

Meanwhile, because of the rumours that were swirling around about last night's YouGov poll in advance of publication, I was a bit worried that Yes may have only been rounded up to 48% by the skin of their teeth, but in fact that isn't the case, and it turns out that they've made a small gain on the last YouGov poll -

Yes 48.0% (+0.4)
No 52.0% (-0.4)

UPDATE : The datasets for today's Ipsos-Mori's poll have finally been released, and just like Survation, it turns out that Yes did slightly better than the rounded numbers suggested -

Yes 47.4% (-1.5)
No 52.6% (+1.5)

The above numbers should certainly ease any concerns that there was any detectable swing back to No on Tuesday or Wednesday - all of the changes are well within the margin of error, and in any case two of the three polls are showing small shifts towards Yes.

Lastly, I have a new article at the Fair Observer website, on a similar theme to my last-but-one article at the IBTimes.  You can read it HERE.  I was under pressure of time when I wrote this one, as you might just be able to tell!

71 comments:

  1. It is all a bit academic now ,as the real poll grinds on ,however there is quite a bit of reporting around which suggests that No voters switching to yes as they reach the polling booth .Hope over fear ? or when the chips are down they vote with their heart for a better Scotland ? or is it they want to be a part of the best party in town ...
    We may see an enormous upset coming down the pike

    ReplyDelete
  2. Hello, Mr Bravely Anonymous No Troll.

    ReplyDelete
  3. So there has been a last minute swing to Yes.

    ReplyDelete
  4. ylee coyote,

    On my and wife's Facebook it's the No's who are doubtful, the undecideds look to be breaking heavily to ayes.

    Yes voters seem determined and certain.

    ReplyDelete
  5. Also hearing that soft no's are now going Yes.

    Let's hope it's being UNDER reported.

    ReplyDelete
  6. We should be careful about anecdotes.

    ReplyDelete
  7. Sky news on that rumour;

    " And as the campaigning moved into its final hours, police said they were investigating a complaint that somebody involved in the referendum count had been making public comments about postal vote results in Edinburgh.

    Police Scotland gave no further details, but the allegation reportedly came to light via a tweet."

    ReplyDelete
  8. There are also tweets that Iain McKenzie Labour MP for Inverclyde has been been told to leave a polling station for screaming at people to vote No.

    ReplyDelete
  9. "There are also tweets that Iain McKenzie Labour MP for Inverclyde has been been told to leave a polling station for screaming at people to vote No."

    That brings back memories of the worst TV by-election debate in history (plus a hugely satisfying slap-down from Ms Sophie Bridger.)

    ReplyDelete
  10. So he has taken Jim Murphy's approach to campaigning.

    ReplyDelete
  11. See Labour, see electoral fraud.

    They go together like Brown and lies.

    Murphy and Eggs.

    Blair and Bombs.
    Do your own jokes. I've already voted. IN PEN!

    ReplyDelete
  12. Following twitter and lots of shy yes out there;

    " Felix Adamson @FliXeRoCk107 · 3m

    More of my friends who were undecided are telling me they voted Yes; when it came down to it, they wanted to control their "

    ReplyDelete
  13. No no, shy yes doesn't exist ; )


    Turnout in Castlemilk - 60%

    With just under 7 hours to go.

    ReplyDelete
  14. I voted in pen despite the guy saying the pencils were indellible? I spoke to an English yes woman at the door. I have never heard someone so passionate about Scottish Independence. Her husban is in the police and he said 70% of the force were yes. This tells you how they feel about our Scottish government. She said the respect the SNP have for the police has been humbling to them.

    ReplyDelete
  15. Castlemilk 60% if we get that to 80% and replicate the turnout it's game on guys!

    ReplyDelete
  16. @chalk, @November13, All I can say is a hats off WOW.

    If thats being replicated in other similiar situated areas, well you can not blame the pollsters. Turn out in such areas in prior elections would have been well below 60%.

    So if the ground game keeps up and they hit 80%.....

    You can not do do a model projection to forcast with that.

    I can take a wee bit of comfort sitting in my chair hundreds of miles South, that I followed up with donations for the SNP the past several weeks and even on Monday. :)

    That update is comforting.

    ReplyDelete
  17. George : Peter Kellner was adamant on the radio yesterday that it's easier to poll accurately for a high-turnout election than it is for a low-turnout election. Time will tell, I suppose.

    ReplyDelete
  18. I've often heard that one of the greatest dangers for my side (The No side, that is) is differential turnout. Huge turnout reduces that possibility.

    ReplyDelete
  19. Reduces, but not eliminates, of course - it would have to be a Quebec-style turnout before you could be sure there wasn't much in the way of differential turnout.

    ReplyDelete
  20. Why is that? Is it because it is easier to anticipate the demographic composition of the electorate?

    ReplyDelete
  21. Xabi - it's because you don't need to bother anticipating the demographic composition of the electorate, you can get away with mostly just using the demographic composition of the country (available from very reliable census and marketing database information).

    ReplyDelete
  22. The odds on NO have drifted down from yesterday. Are people assuming that high turnout favours NO? I wonder if any people with deep pockets have commissioned their own exit polls so as to clean up on the betting?

    Slightly worried by the way the odds look but amazing to see such a big turnout.

    There are indeed NO people on twitter by YES control of social media, but this is really nothing new.

    ReplyDelete
  23. There is a YouGov exit poll of sorts underway, apparently. But there are any number of things that can cause movement in odds.

    ReplyDelete
  24. @James, registration is at 97%, a poll where 97% all show up would be accurate.

    Anything beneath the full registration means there is a differential turnout.

    Lets assume turnout is 80%, we do not really know who is turning up?

    My hope is that C2DE are outperforming ABC in turning up, thats the ground game. Who can get more YES or NO voters to the polls.

    IF the polls are in fact a statistical dead heat which is what they are, then it shoudl be possible for a 35,000 person ground team to squeek the extra points needed to win.

    ReplyDelete
  25. "My hope is that C2DE are outperforming ABC in turning up"

    That's the problem with this theory, because in my view it would be a great achievement if C2DEs only slightly underperformed against ABC1s.

    ReplyDelete
  26. What I've just said doesn't exclude the possibility of differential turnout, of course, but you'd need more Yes voters in each social class to come out than No voters.

    ReplyDelete
  27. "IF the polls are in fact a statistical dead heat which is what they are"

    In a more pessimistic note, all polls show the very same result. There is not much variation. Do we apply the same 3% MOE when all polls show exactly the same result? It seems to me that this convergence can be caused by the fact that they are expecting a nearly total turnout.

    Perhaps we should be considering whether Yes would in fact be favoured by low turnout (say in the 70s or low 80s).

    ReplyDelete
  28. Odds on No have probably drifted due to the Betfair story.

    ReplyDelete
  29. Some utter rubbish getting posted up here about result predictability based on high turnout. I mean utter rubbish including quoting arch Unionist Kellner. The measure for high turnout is usually General Election v local election. However, the fact the former is 'higher' than the latter doesn't conflate to 'high turnout'. GE polling prediction theory just becomes the norm. Today's turnout will not be the norm; it is therefore far harder to model and predict. There will be Scottish constituencies with turnout increase of 30%+. Unless you build models that fully factor that phenomenon and adjust samples accordingly you simply CANNOT predict the likely outcome. I have seen no polling evidence to state that 'don't normally votes' and 'new to register' have been satisfactorily accounted for in prediction models. Be in no doubt - extremely high turnout today is great for Yes and bad for No. Kellner knows it, Curtice knows it, MacDougall knows it, Shorthouse knows it, every screaming Labour MP getting removing from their local polling station today knows it. Just gauge by their tweets closer to 10. And if you know anyone who has yet to vote tell them to get to their arses down to their local PS well before 9.30.

    Cheers
    James

    ReplyDelete
  30. Where is this movement to no where is the proof. It's all speculation. The odds are speculation as well!

    ReplyDelete
  31. "Do we apply the same 3% MOE when all polls show exactly the same result?"

    Not at all. No clearly have a small lead in the polls, so for Yes to win, there has to be either -

    a) A late swing of about 2% or so on polling day itself.

    or

    b) A systemic flaw in the polls (but not necessarily a large one) that has exaggerated the No vote.

    or

    c) A differential turnout favouring Yes.

    ReplyDelete
  32. @James

    I hope something of that happens. I just think differential turnout with 80%+ of turnout will not be what wins us the referendum.

    ReplyDelete
  33. Tom Newton Dunn just tweeted BT confident of win because of big turnout in NO areas.

    ReplyDelete
  34. Some polling stations are now closing due to having reached 100%turnout.

    Not sure where.

    ReplyDelete
  35. Anon...that'll be the same Tom Newton Dunn that talked Murdoch out of coming out for Yes...

    ReplyDelete
  36. I can't imagine any would have 100% yet. Some people will be going after work.

    ReplyDelete
  37. James, Have you any more info on the YouGov exit poll you mentioned? I was under the impression there would be no exit poll.

    ReplyDelete
  38. I think the difference here to any other election or referendum is that so many soft no people are torn because deep down they want to vote yes but have concerns. I think that feeling will be strong among many Scots.

    I spoke to a no voter in my office who said he felt very torn for the first time when he actually saw the ballot paper. He was very close to changing his mind, maybe a few % will.

    ReplyDelete
  39. John : Well, it's obviously not been commissioned by the broadcasters, otherwise we'd have heard. It's not strictly an exit poll, because it seems to be asking how people have voted, but also how people plan to vote if they haven't done so yet.

    ReplyDelete
    Replies
    1. If it's a YouGov poll I don't want to see it!

      Delete
  40. Anon are you in fact TND masquerading as a concern troll? Honest to god...what else would you expect them to tweet with 5 1/2 hrs of polling remaining?? I live in a no area. I've chatted with two polling agents. No is ahead hear in ABC land, turnout is high, but Yes are showing strong. High turnout in No land is no better for the Dependency Campaign than it is in Castlemilk. People are working flat out to narrow the differential. Positive postings only at this stage please. We are in the game and ahead elsewhere.

    Best
    James

    ReplyDelete
  41. Big turnout everywhwere I'd suggest. Not sure why that helps No.

    ReplyDelete
  42. It's not so much that a big turnout helps No relative to the polls, rather that it makes us more likely to avoid the sort of differential turnout that could potentially damage our side. Though obviously the specifics of it are very much open to interpretation.

    ReplyDelete
  43. Robert, West LothianSeptember 18, 2014 at 5:46 PM

    Can we not just dispense with these intrusive exit polls?
    We've had polls ad nauseum these last two days just let the people of Scotland speak at around 5-6am tomorrow morning. Theirs' is the only poll that matters.

    ReplyDelete
  44. High turnout is brilliant for Yes. Look at this way in a GE say turnout 65% Now imagine a constituency of 100 in the Tory area and 100 in the poor area In the Tory areas turnout is about 90% but in poor areas turnout is 40% Then the Tories will overwhelm the poor by 90-40 but if the Turnout is 90% then it will be a draw. That is what's happening today and if the reports of turnout are true then Yes will win.

    ReplyDelete
  45. Meanwhile in Twitterland...

    https://twitter.com/Trendsmap/status/512580603200290818/photo/1

    Looks like we are only winning in the areas where people who can vote live.

    Taxi for Darling!

    clochoderic

    ReplyDelete
  46. @"Cheers, James"

    I am with you!! Thats the spirit.

    ReplyDelete

  47. Word from Team YES. Reported from Canada as the UK press does not want to print such info.


    http://www.thestar.com/news/world/2014/09/18/scotland_referendum_canadian_team_sees_yes_win_amid_large_voter_turnout.html?utm_medium=twitter&utm_source=twitterfeed


    GLASGOW—A Canadian team of voter-contact experts working with Scotland’s Yes team is forecasting that Scots will vote by as much as 54-46 in favour of independence when the final vote is tallied early on Friday morning.

    “I believe they’re going to win,” said Mike O’Neill of the Canadian voter-targeting firm First Contact, which has been doing data-modelling work with two academics, one of them Canadian, to profile likely Yes and No support in Thursday’s referendum.

    “I feel pretty confident,” O’Neill said.

    ReplyDelete
  48. The article was today, read the whole thing.

    ReplyDelete
  49. Another one, YES has a van fleet to take voters to the polls!!

    http://www.thescottishsun.co.uk/scotsol/homepage/news/scottishpolitics/referendum2014/5919113/Taxi-for-300000.html

    A FLEET of customised Yes Scotland cars are set to ferry 300,000 first-time voters to polling stations tomorrow — as pro-indy chiefs bid to secure a massive turnout.

    The branded motors will pick up people — whether they’re backing Yes or No — in a £150,000 move.

    And an army of supporters who’ll drive the fleet will explain the polling process to people on the way.

    ReplyDelete
  50. Brilliant! Well sourced George! There are reports of very high turnout in No areas such as Dumfries & Galloway...but with Yes outperforming expectations (exit chatter suggested 40% Yes). That would be nothing short of sensational given No are hoping for 70-75% in areas such as this. One big BT lie is their alleged canvas returns in the Glasgow area. RIC and local Yes groups have stormed huge chunks of Glasgow. Drive through Cathcart or Govanhill tonight (as examples) and you won't just see Yes window posters - you'll see entire windows and ledges draped in saltires and vote Yes banners. It looks sensational. If I was Blair McD or Rob Shorthouse reading this I would be very worried their media lead campaign had badly backfired. Yes shoe leather will win the day. Daring to dream now...

    Cheers
    James

    ReplyDelete
  51. Just took part in a Populus poll. Usual questions on past voting history, and how I voted today.
    There was also an interesting question. "Do I think there should be another referendum including the rest of the UK if the vote is Yes?"
    Is someone getting worried?

    ReplyDelete
  52. Juteman, was that phone or online?

    ReplyDelete
  53. It sounds like a UK-wide poll.

    ReplyDelete
  54. No, it wouldn't ask how you voted today if it was UK-wide.

    ReplyDelete
  55. It sounds like a UK-wide poll.

    ReplyDelete
  56. Sorry for the second comment.

    Your objetion seems reasonable, James. But mobile phone?

    ReplyDelete
  57. @"Cheers, James"

    I have been doing twitter advanced searches, lots of voted yes traffic and its all positive happy.

    The voted no traffic, far less and they are all glum about it.

    YES supporters and voters are doing this with a warm heart looking forward into the future. There is nothing wrong with standing on your own two feet.

    I was warmed to read about the team from Canada that is working alongside Team YES.

    I also had not really known much about RICS as I have written earlier I am a "recovering political activist" from my decades in the states.

    Everyone seems to be hitting their stride right along.

    I have seen enough twitter traffic on no votes switching yes that I think there is at least a fraction of a point in that. The referendum question was WELL thought out.

    ReplyDelete
  58. Well, Survation and ICM claim to have phoned mobiles for their recent referendum polls. I'm slightly baffled as to how they did it, but admittedly the geographical distribution of mobile phone numbers isn't my area of expertise.

    ReplyDelete
  59. As I said here earlier knowing the election result is worth millions if you know what to do with that information.

    These polls might not be made public, purely for select consumption and private profit.

    Great info from the Canadians, so hope they are true, but then Glasgow is expected to vote YES, question is by how much.

    I am from Galloway and believe me if we got 40% down there I'd be running out into the street cheering and ripping my shirt off like we had won the world cup, because believe me that would be seismic! That would be a YES for sure. All farmers and tories down there. SNP has core support in west of the region (Stranraer to Newton Stewartish) but not big in numbers, and lots of traditional Labour in urban east. But it is all rural, lots of retired, big percentage of English. Only in Salmond's wildest wet dream fantasies does he foresee a YES vote down in Dumfries and Galloway.

    Anyone done any analysis of what we should be looking for in those smaller areas declaring in the 1.30-2.30 range. Obviously a good majority in Moray for example, but has anyone done modelling?

    ReplyDelete
  60. It wasn't UK wide. They asked what area of Scotland I was in, and read out a list of areas to choose from.

    ReplyDelete
  61. I've just remembered that I took part in a Populus poll about 18 months ago, so they would know I was from Scotland.

    ReplyDelete
  62. @All......

    The pollsters are now taking a walkback.....

    "
    Diarmuid Scully ‏@dscullylimerick

    IPSOS MORI confirm that their opinion polls are no longer valid if turnout in Scotland hits 80% I suspect these extra voters will go for Yes"

    ReplyDelete
  63. @George

    Why? IpsosMORI has actually predicted a higher turnout.

    ReplyDelete
  64. @xabi, I am only relaying what I found. The twitter account is from the Mayor of Limerick, IRL.

    My guess, is one a walkback, but second the sample could be very distorted.

    ReplyDelete
  65. Anonymous 6:50pm

    Well these days Moray is a bit like Galloway in that there is a high number of English people now living here - ex-RAF who have stayed on.
    Frankly I don't know which way Moray will go - obviously hoping for a big Yes vote but would settle for 50/50.

    ReplyDelete
  66. From Diarmuid Scully @dscullylimerick regarding the IpsosMORI walkback:

    "quoted on 5-7 Live Rte Radio 1. Given age class and gender divide in polls the variance from turnout likely to be significant"

    https://twitter.com/dscullylimerick/status/512659142696509441

    Retweet, discuss it if you like. I'm very happy with this news.

    ReplyDelete
  67. Close to tears reading these comments, 87% turnout in my village. I went round last night with my wife of 1 year,who had never done anything. Suddenly she had the bug that I have had for years.

    Still an hour to go and still people coming and going. Met salmond earlier for a photo, in inverurie,which is very no,but you wouldnt think it.

    Am I dreaming?

    ReplyDelete
  68. I see Kevin Maguire of the mirror has been getting told by senior No figures that No expects to win 58/42.... really?

    ReplyDelete