Fake News: Leave.EU publishes junk poll calling Stoke by-election for UKIP

Polling experts warn journalists not to report results


This morning, seedy Brexit sideshow Leave.EU has published a new ‘poll’, which shows UKIP leader Paul Nuttall winning the Stoke central by-election with 39 per cent of the vote.

However, they have published no methodology beyond claiming an implausibly large sample size. The research director of YouGov has advised scepticism pointing out that a phone poll of 4,000 people in a constituency election is ‘getting on for impossible’.

Another leading psephologist, Will Jennings, has dismissed the poll, warning journalists that to report its findings with further corroboration would be ‘as bad as fake news’.

Back in June, Leave.EU published an poll predicting the results of the EU referendum which, although it included no convincing methodology, accurately predicted a 52-48 result. Some have suggested that this lends credence to the ‘data’ published today.

That fails to recognise that given the tight margin predicted in the referendum, and the fact that there were only two possible options, there was a high probability of guessing the correct outcome.

Given that Leave.EU founder Arron Banks — also UKIP’s largest donor — openly admits that their MO was getting media attention through sensationalist content, we should assume that their ‘polling predictions’ are no more than stabs in the dark.

This is further evidence that Banks and his cronies are deliberately creating fake news in an attempt to influence the electorate.

We should call them out for what they are — liars.

Niamh Ní Mhaoileoin is editor of Left Foot Forward. Follow her on Twitter

See: Dear Stoke-on-Trent: Here’s what Paul Nuttall really stands for

Like this article? Sign up to Left Foot Forward's weekday email for the latest progressive news and comment - and support campaigning journalism by making a donation today.

8 Responses to “Fake News: Leave.EU publishes junk poll calling Stoke by-election for UKIP”

  1. Richard

    this article is biased. Is there was a highly probable outcome but they would correctly guess the referendum result why did no one else. This article is fake news.

  2. Ces

    This article is garbage fake news.

  3. Millie

    I did a google and Leave.EU say that they are using a method of collecting a very large randomised sample, that is much bigger than would be used in more normal methods, which tend to control their samples to reflect the population make up. Instead this method randomly selects a very large sample, and works on the basis that the much larger sample size will reflect the make up of the population better than a smaller one. It’s pretty much a big data technique rather than one based on smaller sample selection. But anyway, we’ll find out soon enough whether or not it was accurate.

  4. ted francis

    Given the well established, historical fact that UKIP has made a practice of lying, I couldn’t believe Nutall & co., if they told me night follows day. The high probability has to be that any “news” coming out UKIP is fake.

  5. Damian Hockney

    The problem with endlessly using the words ‘fake news’ and ‘liar’ about all that opponents do is these words they can, if one wishes, cover pretty much everything done by all parties in promoting their own interests – unscientific polls without full details of demographics, weightings and data should simply be marked as such and commented on in that way: the poll is exactly that it says it is – someone’s own research built up to ‘prove’ that a candidate can win. It’s not a ‘lie’ or indeed fake news, as it exists and has been put into the public domain. It’s part of the hype in a seat which we all know, let us face it, that UKIP can win, however much the establishment tries to stop it, often using what can be described as lies and fake news in the process.

    The problem with being too extreme in response to this type of thing is that it in fact has the effect of giving polling without methodology some credence, which itself is not a good thing: you have rightly reported that polling along these lines by this same body got the result of the referendum completely accurate (whilst all other polling did NOT). Given the characterisation (often very unfair and which itself could be described as fake news) of all polling now as totally inaccurate and useless, voters looking in a detached and non partisan way might wrongly take from your article “well, at least the supposedly unscientific one was right while all the other were wrong”. It is not enough to use the ‘fake news’ that the polling itself was nothing more than ‘stabs in the dark’. You can neither know this, nor assume it reasonably given the fact that they alone got the result spot on. In any event, in life and business and politics, someone’s ‘stab in the dark’ that is right has more value than all the flawed and tendentious methodology in the world. But in any event, you must know it was not a stab in the dark – there was no one single reason for any side in the referendum campaign to prove that they were or were not going to win, as no-one in the campaign properly calculated the potential outcomes of voters thinking ‘we are gonna win, let’s not bother going out to vote’ against ‘hurray we are gonna win, we are now all inspired to all go out and put the boot in and make it a big win’. On either side Anyone who has spent hours knocking on doors and campaigning knows that this type of thing is impossible to call.

    The best answer to this is for a reputable polling body to do its own poll – with the caveat and recent memory to the fore that the Richmond poll during the by-election late last year suggested a conclusive win for the Tory (56-29)…and the LibDem won it conclusively. Was that fake news? Any more than the ‘stab in the dark’. If not, it was real news that was woefully wrong and inaccurate, and the ‘fake news’ was spot on.

  6. Andrew

    Come on, let’s be fair. A larger sample obviously provides a more acurate picture than a smaller one. To say otherwise is irrational, and is a repudiation of elementary scientific rigour. A larger poll is self-evidently the main reason why the actual votes on polling day provides the most illuminating picture – because it is a full complement. The above comments disparaging large samples, simply lack a sense of objectivity. Do those same commenters read the press/ internet merely seeking a comforting echo chamber? If we can keep the dialogue in this forum on a more reasonable footing; I’m sure that would better promote readers’ trust.

  7. Paul Icini

    Given that there are in fact now two Labour parties, with utterly different policies, it’s hard to know which will end up being implemented if they are elected. Labour members are subject to whips just like other parties. Having studied the UKIP manifesto and the comments from their new spokesmen I can see nothing objectionable and since most members are no longer ex Tories they appear to be a better bet for the working man than anyone else’s, especially for those of us who wish to leave the EU.

  8. Peter Kenny

    It seems unlikely that leave eu have discovered some miraculous new technique which professional polling organisations have missed.

    As for Labout being two parties UKIP strike me as being barely a party at all at the top – witness people rolling around fighting at the EU parliament and their amazing leadership elections. More of a desperate cabal looking for another source of income now the EU parliament is over for us.

    Farage himself called them ‘low grade people’ and who are we to argue – he knows them better than us.

Leave a Reply