The Futility Monster

He'll pointlessly derive more enjoyment out of your resources than you

Return Of The Poll

Posted by The Futility Monster on January 11, 2010 @ 09:48

It’s been a surprisingly lively winter period as far as the polls are concerned. So much so that it’s time to see if all that wonderful political action over the last few weeks has actually achieved anything.

  • Conservatives: 40% (N/C)
  • Labour: 22% (N/C)
  • Lib Dems: 21% (-2%)

Clearly not.

I am, however, starting to get a little concerned about my polling methodology.

I am happy with what I’m currently predicting for the Conservatives and the Lib Dems. I believe the Tories will achieve a minimum of 40% (of the GB share, excludes NI) and the Lib Dems are looking good for 20% or thereabouts. That will still be poor for them, as their GB share last time was 23% – though it all depends on each individual constituency, as we all know…

But what’s giving me greatest concern is the Labour figure. 22% would be a stunningly bad performance for them. My estimate of their share is based on Labour’s worst polling result in the last 90 days. A very large window, but I believe that British opinion does not change as much as the polls would indicate. I think taking a longer-term view (because 90 days is long-term in the world of politics!) gives us a better indicator of the core numbers for each party.

Until now I had accepted this Labour methodology without question. It comes direct from the God of Polls himself, Mr Mike Smithson of Politicalbetting.com. It is his Golden Rule. Always assume the worst of Labour.

The problem now is the following graph, showing the median Labour polling figure calculated on a rolling weekly basis from all polls across the previous 90 days…

Furthermore, the last five polls have put them on 30% or 31%. We should rightly be sceptical of such a large figure (for Labour) as they have never delivered on their best polling figure at an election, but to scale that back to 22% would be mass exaggeration.

I know Mike Smithson would counter by saying his Golden Rule talks about using all the numbers from the poll with the Labour lowest figure, but it would still remain true to say that the Golden Rule would predict that very same low Labour result.

I am loathe to change my methodology without an election as justification, and I am also wary of questioning The Man, but I just wonder if even he is starting to consider that just this once our scepticism of Labour may be misplaced. Perhaps a genuinely competitive election will finally bring out the Labour vote in more force than we’ve been used to?

It’s a difficult one, but I’ll be keeping it under review…

Advertisements

2 Responses to “Return Of The Poll”

  1. vornstyle said

    You’re right. Your Labour figure just doesn’t *feel* right. But! There’s something to be said for keeping the methodology until the d(rama)-day of a General Election. Test it to destruction. Carry on the experiemnt properly (though, of course, an election is very distinct from lab conditions!) A methodology failure on election day would give you a far richer analysis of its failure than dropping it now, and you’d have an incomplete analysis of the success/failure of the substitute/mutated methodology.

  2. […] Monster’s Poll CON: 40% LAB: 22% LD: 21% Methodology Latest analysis […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: