www.fgks.org   »   [go: up one dir, main page]

Baseball nerd who predicted Obama's win foresees Labour meltdown

Statistical genius Nate Silver has applied his insights to the concept of a uniform swing and suggests large Labour losses

US election campaign: Barack Obama and john Mccain

Nate Silver predicted the correct result between Barack Obama and John McCain in 49 of America's states. Photograph: Getty Images

The baseball nerd who used his genius for statistics to make startlingly accurate predictions in the 2008 US presidential race has weighed into the British election – and his conclusions make chilling reading for Labour.

Nate Silver, who correctly predicted the result between Barack Obama and John McCain in 49 out of America's 50 states, argues that the most popular method for translating opinion poll results into numbers of seats in parliament greatly overstates how well Labour will perform, giving the false impression that the party "has a fairly large buffer zone before facing total Armageddon".

The concept of uniform swing assumes that the projected national swing to or away from any given party will be manifested in identical vote swings in every constituency. If an opinion poll suggests that the Conservatives are up six points on 2005, for example, the assumption is that they will do six points better in each constituency. The method has long been criticised as flawed, especially in elections with strong third parties. But on his blog FiveThirtyEight.com, Silver goes one step further, presenting his own alternative methodology that suggests a disastrous May 6 performance for Labour.

Silver's method breaks up the monolithic uniform swing and instead assigns specific percentages of the parties' votes in 2005 to other parties in 2010. Using a recent polling average of the three main parties – with the Conservatives on 34%, the Lib Dems on 29.1%, and Labour on 26.9% – the differences between the two methods become stark. Using uniform swing, those percentages translate into 253 Labour MPs, 271 Conservatives, and 93 Lib Dems. Using Silver's method, Labour ends up with just 214, with the Conservatives surging to 304, and the Lib Dems on 101.

Another scenario considers what would happen if 10% of those who voted Labour in 2005 switched to the Conservatives and 15% to the Lib Dems, while 10% of 2005 Tory voters also switched to the Lib Dems. Under uniform swing, Labour would have 254 seats, nearly neck-and-neck with the Tories on 256. Silver's method gives the Conservatives 276 to Labour's 218.

Last year, Time magazine named Silver one of the world's 100 most influential people for the accuracy and impact of his presidential predictions. The Labour party may will be hoping that he has lost his touch.


Your IP address will be logged

Comments in chronological order

Comments are now closed for this entry.
  • This symbol indicates that that person is The Guardian's staffStaff
  • This symbol indicates that that person is a contributorContributor
  • Mickeyleon Mickeyleon

    28 Apr 2010, 9:33AM

    Conservative will win as they have the right wing press owned by Murdoch as their propoganda machine. I am hoping the Lib Dems can shock the world, but I know that is not going to be the reality.

    Conservative will win, they will do nothing positive for years, they will blame all their failings on the previous government, Cameron will be ousted and a new person will come in. Nothing will change, same old lies and BS.

  • FrankLittle FrankLittle

    28 Apr 2010, 9:41AM

    EmpireHancock

    I wonder what Nostradamus had to say about it?

    He said that Clegg would do well in a box like object, that he would get browned off with Labour and would regret getting his hands sticky by showing support for a toffee nosed schoolboy. He also predicted that some Yank would try to predict the election result by basing it on a children's game called rounders.

  • kvlx387 kvlx387

    28 Apr 2010, 9:43AM

    Don't we already have polls of marginals in the UK? These are showing a somewhat higher swing to Conservatives (I believe about 5.5% from the last general election) than the national polls, but not a large enough swing to give the Tories a majority, let alone a 'Labour meltdown'.

  • benagyerek benagyerek

    28 Apr 2010, 9:45AM

    he should add some scenario analysis on tactical voting, given the recent findings from icm:

    http://www.guardian.co.uk/politics/2010/apr/26/labour-support-fall-icm

    party vote of which definite
    con 33 83
    lab 28 69
    ld 30 68

    seems to me there is now enormous scope for indiv swings within constituencies betw lab and ld aimed at stopping the tory candidate.

    this piece is also v interesting:

    http://www.ft.com/cms/s/0/46c7f5ca-5091-11df-bc86-00144feab49a.html

  • benagyerek benagyerek

    28 Apr 2010, 9:50AM

    cmenon - i think the point is that the lib dems may not get anywhere near as shafted as the swingometers assume - especially once you add tactical voting into the mix.

  • lonereader lonereader

    28 Apr 2010, 9:56AM

    The decline and fall of Labour started with the decline of the working class. Of course, Labour tried to rebrand itself my shifting towards the middle class and giving themselves a new name, and for a while it worked.

    But like many companies that have tried to reinvent themselves but failed, Labour has made the same mistake of not addressing the cause of its downfall - incompetence and corruption.

    They have screwed this country so spectacularly that they will loose this coming election, and I hope they won't be in office again for decades.

  • azog azog

    28 Apr 2010, 10:09AM

    Why baseball nerd Nate Silver as opposed to respected political analyst and psephologist Nate Silver.

    I wonder how much center-left FPP vote splitting in marginals and the left parties running-up-the-score in safe seats might be hiding an even larger Tory majority than the one Silver's predicting.

  • lpage lpage

    28 Apr 2010, 10:09AM

    Main Silver result: LD win is within their grasp
    -------------------------------------------------------------
    This article is another example of the subtle but continuous slant in mainstream media.

    There is in fact another major finding in his results -- but this article chose not to talk about it. In his simulation Silver shows that the LD would only need around 36% to become comfortably the largest party in the parliament. With YouGov indicating that 49% of people would vote LD if they had a chance to win, and with the LD hovering around 30%, Silver shows that a LD win is actually possible. Given the importance of tactical voting in the FPTP, this kind of information is critical for voters to make their decision, but the Guardian chose to talk about Silver's genius method while omitting one of his potentially most striking results.

    To see his simulation:
    http://tinyurl.com/36unk95

  • SplottandBrains SplottandBrains

    28 Apr 2010, 10:16AM

    There is now a clear difference in philosophies:

    Big Government (Labour) versus Big Society (Conservative).

    While the Liberal Democrats also appear to be more "Big Society" in their attitude, the people who vote for them could unwittingly let the Labour party and Gordon Brown back in.

    Watch out for unintended consequences!

  • QuetzalcoatlUK QuetzalcoatlUK

    28 Apr 2010, 10:43AM

    Cmenon
    28 Apr 2010, 9:46AM
    Well, you don't need to have second sight or be a maths genius to know that under this electoral system the Lib Dems get shafted.

    Good! Sod them. Hate the blighters. I live under Liberal rule in Islington, where they behave just like little Torys. Love teh way they have perpetuated the privatisation of services and housing...

  • ostrichhead ostrichhead

    28 Apr 2010, 11:17AM

    That's some ropey over-interpretation going on there!

    Recent poll of polls have Tory/Labour/ LibDems, on 33 v. 29 v. 28, not 34 v. 26.9 v. 29.1, (with Labour rising, LibDems falling in most polls).

    That small difference could shove Labour into first place in terms of seats, or very close to it, even under this bloke's method.

    Not great, but not exactly a meltdown.

    Burkeman didn't even mention the change in polls in recent days, which has to make you suspicious.

    Also it's funny that when Labour were in third place, Clegg and co. were going round saying, forget Labour, it's a two-horse race, destroying years of third party pleading. These snakes learn quick, that's for sure.

  • Monty2001 Monty2001

    28 Apr 2010, 11:31AM

    In the US polls are conducted for every region within each state, allowing for detailed analysis. In the UK we should have constituency wide polling for at least the most contested constituencies. Is this done/available?

    The simplicity of US presidential elections (pick one of two candidates) and the pollerization of democratic/GOP politics make it pretty easy to predict whether a state is red or blue. Even though gettin 49 correct is pretty good the UK election is far more complex.

    Another issue that I've mentioned before is the quality of the polls. If they miss out certain demographics the polls could be skewed for/against a given party. This was thought to be the case in the 2008 US election in which pollsters called people on land lines during the day, thus missing young people who often only have mobile phones and those who are working. This was thought to reduce Obama's % since he was favoured amongst younger voters whose voting registration had increased dramatically.

  • Outradgie Outradgie

    28 Apr 2010, 12:11PM

    QuetzalcoatlUK

    Good! Sod them. Hate the blighters. I live under Liberal rule in Islington, where they behave just like little Torys. Love teh way they have perpetuated the privatisation of services and housing...

    And of course "little New Labs" also behave just like "little Torys" so let's hope you've a decent independent candidate.

  • fallentower fallentower

    28 Apr 2010, 1:25PM

    I followed Nate Silver's site closely throughout the US campaign and was immensely impressed. He took electoral forecasting to another level, and his predictions in the Democratic primaries were even more uncannily prescient than his calls for the November elections. However, in the US an important part of his method was to use an array of additional information about the electorate in each state: income, religion, ethnicity, even things like ratios of Starbucks to MacDonalds. So he had a lot more to go on than polls there: he basically built a profile of the electorate which in the case of strongly contrasting candidates like Clinton-Obama or McCain-Obama gave some powerful clues about how people would vote. He's not doing anything like that for his UK analysis, so we can't expect his predictions to be anything like as accurate. All he's doing is introducing the idea of "churn", looking at how many 2005 Tories/Libs/Labs are saying they're switching and to which parties, which can generate some different results from the rather simplistic "Uniform National Swing" idea. And so far as I can see he's not even basing his "churn" numbers on any polling evidence, just his own intuitions of what might happen. He's also ignoring other likely important factors like tactical voting and regional variations (and the YouGov regional polls show there are major regional variations, e.g. a huge swing to the LibDems from Labour in the North of England, but virtually no movement in Scotland). My guess is that Nate Silver's basically right and that if Labour polls in the 20s it will fare significantly worse in terms of seats than UNS suggests, but he's not gone anything like as deeply into the analysis as he did for the US so we shouldn't expect the same kind of accuracy for his seat projections.

  • leafrick leafrick

    29 Apr 2010, 12:04AM

    post bigotgate people will see sense. Labour is the only party with the mental and psychological stamina to deal with the country's challenges. We can all see how Gordon's strategies are already beginning to work. Asylum seeker induction centres now have a third of the possible number of occupiers, local people are being encouraged back into work and the less privileged are better protected under L abour. The Lib Dems know that they are better off supporting Labour than blue eyed boy David.

  • whollymoley whollymoley

    30 Apr 2010, 5:07PM

    So there is a real prospect of this spoiled, shallow PR man becoming PM.

    Cameron doesn't just use spin doctors he is one - it's the only job he's ever known.

    The Tory party have not even reformed or changed themselves - how can they possible do this for Britain?

    We will be electing an ideological hard-line party to government when what we need is reform in the interest of the greater good during this crisis.

    Europe is more improtant than ever - as the Greek crisis proves - and yet Cameron has already burned his bridges with potential partners by preferring to build alliances with right-wing extremists.

    The most important thing is to cut their majority by voting for the strongest non-Tory candidate in each constituency.

    Tactical voting could bring about a sea change in our politics.

  • whollymoley whollymoley

    30 Apr 2010, 5:08PM

    Personally I prefer a PM who is upset by bigots.

    Rather than one who squirms, unable to answer questions on gay people's freedoms due to the bigots in his own party...

Comments are now closed for this entry.

Comments

Sorry, commenting is not available at this time. Please try again later.

Politics blog weekly archives

Apr 2010
M T W T F S S

Find your constituency