British election to be tight, but not that tight…

Nick Cohen with the best run down on election pitfalls (now Gordon has officially called the start) of the morning…

…political polls are loss leaders whose main purpose is to generate publicity for the brand. Often newspapers receive them free of charge. They are not proper random samples, and historically have always had a bias towards Labour. Mike Smithson, of has a golden rule that “whenever polls have been tested against real election results it’s been the survey with Labour in the least favourable position that has been the most accurate”.

In other words, Labour supporters should be more worried than they are about their low poll showing. Their party’s real performance could well be worse. One company normally does what it promises and produces an accurate prediction.

  • articles

    Five questions before you fill in that nomination form

    In your business/professional activities do you
    A seek a middle ground with your opponents
    B take the opposition into account
    C battle against your competitors
    D fight your competitors by any means
    E eliminate competitors for the common good

    Do future events good or bad leave you
    A nervous but you have made sensible precautionary plans
    B eager in anticipation
    C excited in anticipation
    D convinced you can deal with anything
    E unworried because you have fixed everything and everybody in advance

    In regard to opponents do you
    A try win them over
    B compromise good humouredly
    C try and square them
    D try and square them using financial incentives
    E try to destroy them

    In social settings do you want to
    A mix and socialise and network without offence
    B use them as occasions to air your ideas
    C surround yourself with disciples
    D be respected and acknowledged by all present
    E address people from a platform

    Dress wise do you
    A remain content with what’s in your wardrobe
    B dress stylishly but urbanely
    C dress loudly
    D dress late 70s
    E await the day when everyone has a uniform

    If it’s all As and Bs you’re a loser.

  • Drumlins Rock

    listened to Gordons speech in which he said he would be “campaigning the length and breadth of Britian in the next few weeks” look forwards to his visit to Fermanagh in that case.

  • Billy Pilgrim


    But Fermanagh isn’t in Britain.

    Exciting few weeks ahead for we anoraks. It’s our World Cup. Enjoy, everyone!


    UUP Win Belfast East, Belfast South, Fermanagh & South Tyrone, Lagan Valley, South Antrim, Strangford (6)
    DUP Win Belfast North, East Antrim, East Londonderry, North Antrim, Upper Bann (6)
    SF Win Belfast West, Mid Ulster, Newry and Armagh, West Tyrone (4)
    SDLP Win Foyle, South Down (2)
    Other win North Down (1)

  • Sean Og

    6 seats for UCNF? A bit early to be on the drink.

    They may get either F&ST; or South Belfast if they do a deal with DUP. They can’t win both.

    The other 4 you mention are well out of reach.


    depends how many more scandals hit the DUP before election and just how sick people are of SF/DUP Sleeze and negitive politics.

  • Mike Smithson, of has a golden rule that “whenever polls have been tested against real election results it’s been the survey with Labour in the least favourable position that has been the most accurate.

    Ah! a neat generalisation! The horse with the pretty face should always win! And, with certainties like that, being a bookie must be a mug’s game.

    Clearly there has been a systematic fault, over decades, in over-estimating the Labour vote. Take, for example, the mythical “late swing” of 1970. That was invented to get pollsters off the hook: “our findings were spot on until the voters mucked them up.”

    The result is that pollsters have built in more and more caveats. Again, an example: the much-touted YouGov/Sun poll overnight is puffed as:

    YouGov show 10 point Conservative lead

    So this was an increase in the Tory lead, huh?

    Well, perhaps not. Here’s Anthony Wells for UK Polling Report:

    This poll has a significant methodological change, albeit one which has made hardly any difference to the topline figures…

    At this election YouGov have decided to factor [likelihood to vote] in from the start of the campaign, so on the assumption that Brown will call the election tomorrow, they are now in election mode and the figures are weighted by likelihood to vote in the same way that Populus do.

    The all-of-1% “increase” in Tory support, then, hailed so loudly by ConHome and elsewhere, was a statistical “adjustment”. It makes a better story, of course, than “no change”.

    The other aspect of this is comparing oranges, apples and pears. Over a five year parliament five annual cohorts have been added (and we’ve lost a few as well). Since much of the arcane “weighting” of polling invites respondents to recall how they voted last time, there are two variables already: the new voters, and human memory.

    Then pollsters (though not so much as the partisan window-lickers on the various sides) fret over local elections as an indicator. Is a 15-25% poll (quite normal in local elections, alas) to be compared with the General Election turn-out? Even the London Mayoral election, when counting was slow because of the “large” turnout, involved only a 45% poll. And even those comparisons depend on the electors being sheep which follow the same track in different hillsides.

    Sorry: I’ll be watching the polls as keenly as anyone. I’m not laying bets. And I’ll wait for the obese prima donna’s final aria.

  • Comrade Stalin


    depends how many more scandals hit the DUP before election and just how sick people are of SF/DUP Sleeze and negitive politics.

    There’s a couple of problems with this simplistic analysis.

    The first is that the UUP are incompetent and incoherent, and I don’t see any evidence that this has changed. On one hand they tell us they want to attract moderate votes and Catholics. On the other hand, they adopt an anti-agreement approach to Hillsborough, and go off for corner-huddles with the Orange Order. Which is it ? It’s gotten so bad that they’ve already lost one MLA and are in danger of losing a second, along with 20 members in South Antrim. How can a party appeal to the electorate to support it when it can’t persuade it’s own membership to do so ?

    The other thing is that inter-unionist and inter-nationalist battles (and battles where Alliance takes a seat from either) within elections here are won, quite consistently, by the candidate with the best on the ground record for constituency presence and local constituency work. It’s simple stuff like getting drains unblocked (or potholes fixed), campaigning for local schools/jobs throughout the term (not just at elections). People aren’t going to vote against sleaze, or expenses scandals, if they think it will cost them their hard-working representative. Imagine you’re a voter with no particular interest in politics. Who cares if the guy had his hand in the till once or twice if he/she’s getting your local problems sorted ?

    This is UCUNF’s fatal mistake. Instead of having people in the constituency building up their profile over a period of time, they’ve been selected within a few weeks of the election, which creates the impression that they’re blow ins. They are trying to counteract this effect by running celebrities. I have real doubts that there is any prospect of this working. They’re focussing on the message, and the scandal, without focussing on the constituency work.

  • Stevie H

    I honestly think that the UCUNF are going to crash and burn this election. As mentioned Reg is throwing out too many contradictions. The latest is that Reg and good ole Dave want to “end Northern Ireland’s semi-detached political status”, now while I don’t have an instinctive grasp of how even moderate Catholics with vote I’m pretty sure statements like that will make them run a mile.

    When UCUNF come up with 0 seats to show Cameron will abandon them.

    As Comrade Stalin has pointed out above UCUNF have made very little impact and did very little ground work here in Northern Ireland, this suggests one of two things. Either the Tories aren’t providing the support they can i.e. lessons on how to run a proper election campaign or the Tories are providing this advise and the UUP are ignoring them.

  • union mack

    Reg should have issued an ultimatum to Sylvia after the Euro election, got it out of the way, and made quick progress to select candidates. The South Antrim mess if a fine example of his weak leadership qualities. He now has to gamble. Stand himself in South Antrim, take on McCrea, and unequivocally pursue the UCUNF project without a DUP pact. If it fails, he loses the leadership. If he doesn’t gamble, he loses it anyway.

  • Comrade Stalin


    The problem for the Tories is that they don’t have anyone on the ground. And they know a lot about how to run elections, but are local UUP activists just going to be quiet and do what they’re told by their Tory masters ? Of course they’re not. And we already know that deploying financial resources and blasting them at the electorate does not work. The Tories tried that in North Down in 1995, and ended up with one of their worst results anywhere in their history as a party.

    union mack, the dumb thing about Reg is that he didn’t think this through. It’s possible to be opposed to double jobbing while still running for Westminster, and not lose out. All you need to do is give a promise that you’ll stand down as MLA if you are elected to Westminster. You can then nominate a party colleague to take your Assembly seat.