The Futility Monster

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

Posts Tagged ‘opinion pollsters’

Final Poll Round-Up: What Happened?

Posted by The Futility Monster on May 9, 2010 @ 10:31

Two issues to round up. How did the Monster’s Poll do, and how did all the other pollsters do?

For the answer to the second question, I defer to the human calculator that is Nate Silver of FiveThirtyEight.com.

First thing to say is I’m delighted that MORI did well, as they’ve been something of a whipping boy over the past few years. ICM too retains their crown as being seen as the “gold standard”. And credit to the newbie RNB Research, an Indian based polling firm. It seems they all deserve high praise for getting the Tories and Labour spot on.

The Lib Dem figure, however, is cause for embarrassment for some firms. How did they get them so wildly wrong, appears to be the universally asked question.

But let’s not be too harsh. Polls are +/- 3%. Those who had them on 26% could say they were within the margin of error. Even 27% is not hugely wrong. But 29%? 32?!

What seems to have happened is certainly that the LD waverers did not go to the Conservatives. The table above suggests that, in fact, around 1.5% of the total vote moved from the LD to Labour, and the rest either didn’t vote at all or went to another party (as “others” seems to have been underestimated). Perhaps that is a reflection of the fact that most Lib Dem support seemed to be coming from sections of the electorate (i.e. the youth) that are totally unreliable.

But the chunk that moved to Labour… my gut feeling is that it was LD sympathisers that nevertheless chose to rally back behind Labour in order to stop a Cameron advance. Sometimes this occurred regardless of the actual situation on the ground. That seems to account for the LDs bizarre inability to take the low-hanging fruits off Labour in Edinburgh South, Islington South, Watford, Aberdeen South, Hampstead & Kilburn and Edinburgh North.

Then there was the movement to the others. If we use RNB as our new gold standard, there was a 0.8% increase in Others. Now it seems that barely any of that went to the SNP or Plaid. Perhaps the more likely explanation is that from the general rise of turnout some of it went to the other others… UKIP and the BNP both had decent elections when examined on a GB basis. It would seem odd that LD support would go in this direction though.

Either way, the pollsters all have methodological revisions to do. The era of Labour overstatement appears to be over – and I now feel vindicated in dropping the Mike Smithson Golden Rule from my own polling prediction. Angus Reid on 24% (ignore the nonsense OnePoll) was clearly very wrong. I suspect they have a lot to learn about how to poll British elections.

Likewise, the era of Lib Dem understatement is also in jeopardy.

As for the final Monster’s Poll prediction, it doesn’t seem too bad. You know what they say, garbage in, garbage out.

The Tories over-performed their median figure by nearly 2%. Labour also over-performed by the same amount. The LDs underperformed their median by just over 4%. If I’d stuck to the Labour minimum and LD maximum, however, I would have been way, way out.

With that in mind, what might the prediction have been if I’d used Con MAX, Lab MAX and LD max over the final 10 days of polling…

CONSERVATIVES: 38%

LABOUR: 30%

LIB DEMS: 24%

The answer is that it would have given us Labour and LD almost spot on, and the Tories slightly over-stated. The total of them is 92%, which is too high. Approximately 2% needs to be removed from them and given to others. How about deducting that straight from the Tories and giving it to Others? Doing that would give an estimate that is pretty much exactly right.

Food for thought for future predictions. But would it still apply in future elections? Surely the conditions for this one were unique?

Probably. But hey, it was fun analysing it!

Advertisements

Posted in Monster's Poll | Tagged: , , , , , | 1 Comment »

The Battle Of The Pollsters

Posted by The Futility Monster on May 5, 2010 @ 23:25

Let’s face it, there are too many polling organisations doing political work.

But this year, one of them is definitely going to get it right (and by that I mean within 1% on each party) because of the wide spread of results. You pays your money, and you takes your choice. As you can see below…

These are the final polls for the named pollster.

Pollster Labour Conservative Lib Dem
Angus Reid 24% 36% 29%
BPIX 27% 34% 30%
ComRes 28% 37% 28%
Harris 29% 35% 27%
ICM 28% 36% 26%
MORI 29% 36% 27%
Opinium 27% 35% 26%
Populus 28% 37% 27%
TNS 27% 33% 29%
YouGov 28% 35% 28%
RANGE 24-29% 33-37% 26-30%

On top of all this, we have the monster exit poll to come tomorrow, which is going to involve NOP, who haven’t done any political polling for anyone for years, despite getting the 2005 election spot on.

There has been some convergence, so they’re either all going to look good, or all bad. I’m supremely confident that each party will get a figure within the range indicated. Why bother with polling otherwise?

Of all of them, the one that stands out the most is Angus Reid. They have a lot on the line tomorrow. Are they going to be rewarded for bucking the trend all year? Or are they going to find British polling a lot tougher than they thought?

I’ll be taking my final shot at the numbers tomorrow. With and without YouGov. Not because I don’t like them, just because my sample is so heavily biased towards their polls cos there are so bloody many of them.

Good luck to them all, anyway. But we may be wishing or or two farewell come May 7th…

Posted in Musings | Tagged: , , | Leave a Comment »

Pollwatch: T-Minus 2 Days

Posted by The Futility Monster on May 4, 2010 @ 22:39

CONSERVATIVES: 34% (N/C)

LIB DEMS: 28% (-1%)

LABOUR: 28% (N/C)

Changes based upon last time (yesterday). Sample consists of all polls with mid-point fieldwork dates within the last 10 days, including today (n=29). Includes all British Polling Council registered pollsters. The results above are the median figure for each party.

This election has been like a rollercoaster. One day up, one day down.

The problem is that rollercoasters are under the influence of gravity. And, in the end, it returns to the Earth.

Just like the Lib Dems. Tonight’s polls continue to show their steady decline back to the norm. Bear in mind that good old Charles Kennedy notched up some 23% in very favourable circumstances last time. Now that expectations have been raised immeasurably, anything in this region would be very disappointing for Nick Clegg.

So what are we to make of tonight’s YouGov putting them on 24%. It feels like an outlier. We might find out tomorrow. But what if it’s not? Oh, I can’t take the stress any more.

This is what makes this election so cruel to us poor downtrodden Lib Dems. We thought we might be in with a shot. The polls were even mean enough to confirm our gut feeling that something special had happened on the night of that first debate.

And then it gets snatched away from us.

I’m trying to stay positive though, especially as it would be fickle of me to change my mind again after my post yesterday, in which I tried to cheer myself up.

The election, however, is even more wide open than before. If YouGov is right, and Labour get 30%, that will definitely be enough to ensure a hung parliament. I’m confident of that, because it will mean Labour don’t lose too many Lab-Con marginals; and if the LDs outperform their rating in the LD-Con marginals, it will make for one hell of scrap for power on May 7th.

The final polls tomorrow will be the ones to watch out for. Will there be the usual convergence? Even if there is, will it even mean anything, because 1% here and there when the polls are this tight, and in our electoral system, can be the difference between hung parliament and majority?

I love elections.

There will be no Pollwatch instalment at this time tomorrow, as it will be shifted forward into Thursday instead to make sure I don’t miss any of them being released on Wednesday night.

Be lucky.

Posted in Monster's Poll | Tagged: , , , , , , , | 1 Comment »

Pollwatch: T-Minus 6 Days

Posted by The Futility Monster on April 30, 2010 @ 23:59

CONSERVATIVES: 34% (N/C)

LIB DEMS: 29% (N/C)

LABOUR: 27% (N/C)

Changes based upon last time (yesterday). Sample consists of all polls with mid-point fieldwork dates within the last 10 days, including today (n=28). Includes all British Polling Council registered pollsters. The results above are the median figure for each party.

Well, that was exciting, wasn’t it?

There has only been one post-debate poll released, and that’s the daily YouGov. That showed… nought but a 1% rise in the Labour share. Barely anything. And, though we still need more data, I’m going to stick my neck firmly on the block now and say I’m feeling pretty vindicated about predicting nothing will come of “bigotgate” .

Sunday is definitely the big day. There is bound to have been lots of polling in the field today, and more tomorrow, all ready for the final Sunday of the campaign. There will be much talk of whether these polls on a bank holiday weekend are accurate, but I trust the pollsters to get round any sampling difficulties. They know what they’re doing.

And expect next week to be a continued deluge. Be aware too, that of all the numbers coming out over the next few days, one of them is bound to be exactly right. And maybe more.

Scary thought, no?

The endgame approaches. Are you ready for it?

Posted in Monster's Poll | Tagged: , , , , , | 5 Comments »

That Elusive Labour Share

Posted by The Futility Monster on April 14, 2010 @ 09:15

It’s time to stick my finger in the air and sense the mood. Here’s the latest Monster’s Poll…

CONSERVATIVES: 38% (N/C)

LABOUR: 26% (N/C)

LIB DEMS: 23% (+2%)

On the surface, there’s something not right here. These figures are now being taken across a 30 day window. The changes listed above are from last time, which was over 20 days ago.

That means, because over 2/3rds of the polls in this sample are new, that the polls are still not showing much movement. Except to the Lib Dems, but one expects that during a campaign. That’s why we always take their maximum, even if many intervening polls show them lower.

If we look instead at medians (bearing in mind that the Tory share above is already the median) the Labour Party are on 31% and the Lib Dems on 19%.

That is closer to the mark, but it hides a huge degree of volatility, particularly in the Labour share. Within the past week, they have been as low as 26% and as high as 33%. Yes, we’re comparing apples with oranges, as they are from different pollsters with different formulae, but when you look at the Tory share in the last week, their range is narrower, from 36% to 40%. As a proportion of their actual share, the Labour volatility is much higher.

So, as I have agonised about why it’s so hard to get a sensible reading for the Labour figure, the pollsters are in the same boat. They all have their own weird ways of assigning “don’t knows” based on why they voted for last time, or who they say they lean towards, or how certain they say they are going to vote… but whichever way you look at it, the Labour share of the vote is a true toss up. 26% will result in a Tory landslide. 33%, depending on how the Tories do, might just mean Labour remains the largest party in a hung parliament.

Perhaps Labour support is indeed solidifying, and the perceived closeness of the election will encourage more and more of their supporters to the polls. But what if all of those Labour loyalists live in solid Labour seats anyway? That is no use to them.

The marginals are all that matter, and as the specific marginal polls keep showing, the Tories are doing better than the national trend there. In some respects then, that distortion would have a similar effect to a lower national share for Labour, and projecting the result based on that may be more accurate after all.

Here’s hoping we see the classic poll convergence in the coming weeks. Cos right now, I ain’t a betting man.

Posted in Monster's Poll | Tagged: , , , , , | 1 Comment »

Why Other Polls Don’t Matter

Posted by The Futility Monster on March 29, 2010 @ 10:01

We Love Polling!

While browsing through the newspapers websites this morning looking for some inspiration, one thing rather surprised me:

Since last week’s Budget, polls have suggested that Labour has regained support on its handling the economy.

It turns out that the BBC, no less, commissioned a poll to find out what the people’s verdict is on the respective Treasury teams of the front benches. From the same newspaper, but a different article, we get the numbers:

The poll showed Gordon Brown and Alistair Darling have overturned David Cameron and George Osborne’s lead on economic issues.

Mr Brown and Mr Darling were backed by 33 per cent of voters on the economy; Mr Cameron and Mr Osborne were supported by 27 per cent.

So, the Budget has turned things around for Labour, it seems. Hurrah for Alistair Darling: Saviour of New, New Labour.

Well, a reasonable observer might conclude that that was the case. Only, it isn’t.

There hasn’t really been enough polling since the Budget to draw any definite trends just yet, largely because no one does any polling over the weekend. We’ll know more in a couple of days, but the early trend is the complete opposite.

Both ICM and YouGov, since the Budget, have been showing a delicate change in favour of the Tories. All within the margin of error, naturally, and one swallow doesn’t make a summer, and every other polling cliché in existence, but I’m willing to bet that polls this week continue to show the Tories have notched up just a point or two since last Wednesday, putting them strongly back in solid majority territory.

How can the polls provide such contradictory results? How can one poll tell us Labour are now more trusted, but the others are saying, “Thanks but no thanks”.

It’s all down to the weightings. And the weightings are the reason why no one should bother with non-voting intention questions.

The poll asking people “who do they trust on the economy” is highly likely to include a great deal of people who won’t vote, can’t vote, never will vote, don’t care, and all the other lazy sods out there.

Whereas, because voting intention polls are the Gold Standard by which a pollster is judged, they have developed years of detailed methodologies for seriously reducing the influence of such people on the topline figure. People who didn’t vote last time, for example, get less credence placed on their opinion than those who say they vote at every election.

Non-voting intention questions are all nice and occasionally interesting in the world of politics, but they are invariably used by someone trying to create an alternative narrative. In this case, it’s a bit disappointing that the BBC would dabble in such voodoo polling.

They don’t matter one bit.

The only thing that matters, and the only thing that will always matter, is how people say they are going to vote come Election day.

Posted in Musings | Tagged: , , , , , | Leave a Comment »

The Monster Moves

Posted by The Futility Monster on March 11, 2010 @ 11:49

Let me cut to chase. Here is the latest Monster’s Poll calculation.

CONSERVATIVES: 38%

LABOUR: 25%

LIB DEMS: 22%

No changes from last time because of one simple reason: methodological change.

But why so abrupt? Well, to get your attention, and then let you scurry off now without having to bother about explanations. Go on, on your bike!

Still here? You must be the only one.

Having glanced once more at my polling spreadsheet, and the deluge of polling data we are currently getting – which will only increase as the election approaches – the excitement of the impending election is ever apparent.

With that in mind, I have been forced to reconsider my polling methodology. I was hoping to avoid having to do this before the election where the methodology can be tested, but I believe there is good reason to.

In the past, my polls were considered over the previous 90 days. This very large window smoothed out the apparently violent fluctuations in public mood that our polls seem to show, something that I deem to be a fallacy. There aren’t millions of people shifting who they’re going to vote for daily. Most of it is statistical noise and variations between polling organisations and their respective methodologies.

But in honour of the election, it seems only fair to close that polling window, otherwise the figures (particularly the sensitive Labour MIN and Lib Dem MAX) are going to be heavily skewed towards older data when we will have vast amounts of fresh data to play with.

So until the election, the figures will now be calibrated based on the last 30 days polling. If that is less than 30 polls, then I will use the last 30 polls anyway, regardless of age. 30 is my magic number for approximating a normal distribution, don’t forget.

Labour still seems quite low for my liking, but this is because of our old friend Angus Reid, who are really putting their necks on the block with their data sitting way out of kilter with the rest of the pollsters regarding Labour’s recent performance.

In any event, I think it’s safe to say that there’s no way all these polling firms, now including: ICM, YouGov, Ipsos MORI, Populus, ComRes, Harris, TNS, Opinium, Angus Reid, the appalling BPIX, and maybe NOP will re-emerge too, will survive post-election. Only a handful will emerge unscathed. Others, especially those recording massive “Others” are going to look very silly. Let’s get real: they aren’t going to get 16% of the vote.

Not only are we getting an electoral war, but a polling one too.

My money is not on the newcomers…

Posted in Monster's Poll | Tagged: , , , | Leave a Comment »