politicalbetting.com » Blog Archive » Let this week be a lesson for us all: Don’t get too excited

I agree with Edmund that something seemed to be happening. But it wasn't a significant drop in Labour's share, rather a temporary rise in the Tory share. My guess is that this reflected some rallying to the Tory colours with more publicity on what Labour would do, and they can reasonably expect to get this effect again when the election campaign comes round.

What we're not seeing is any shift in the very longstanding 37+ Labour vote share. I think that punters should work on the basis that Labour's result will be at that level, and bet according to whether they think the Tories can beat it. I doubt it myself, but um, I might be biased.

(Or just jetlagged - 20-hour journey yesterday, and feel tiresomely bright and awake after 3 hours' sleep.)

Casino_Royale said:

NickPalmer said:

I agree with Edmund that something seemed to be happening. But it wasn't a significant drop in Labour's share, rather a temporary rise in the Tory share. My guess is that this reflected some rallying to the Tory colours with more publicity on what Labour would do, and they can reasonably expect to get this effect again when the election campaign comes round.

What we're not seeing is any shift in the very longstanding 37+ Labour vote share. I think that punters should work on the basis that Labour's result will be at that level, and bet according to whether they think the Tories can beat it. I doubt it myself, but um, I might be biased.

(Or just jetlagged - 20-hour journey yesterday, and feel tiresomely bright and awake after 3 hours' sleep.)

There is no way Labour will poll 37%+ next year - that'd be an increase of 8% points from 2010. Put my money where my mouth is? I have. A four figure sum.

Labour will poll in the 33-35% range.

Isn't that almost exactly the percentage that has switched from the LD's as a result of the coalition?

So, in the end, is there actually any advantage to the daily polls?

Would a weekly poll with a sample 5 times as large be more useful?

I bet someone on here has done the sums. Are the daily ones more volatile? And if so, is it noise, or can they resolve trends that more widely spaced polls cannot?

OldKingCole said:

Casino_Royale said:

NickPalmer said:

I agree with Edmund that something seemed to be happening. But it wasn't a significant drop in Labour's share, rather a temporary rise in the Tory share. My guess is that this reflected some rallying to the Tory colours with more publicity on what Labour would do, and they can reasonably expect to get this effect again when the election campaign comes round.

What we're not seeing is any shift in the very longstanding 37+ Labour vote share. I think that punters should work on the basis that Labour's result will be at that level, and bet according to whether they think the Tories can beat it. I doubt it myself, but um, I might be biased.

(Or just jetlagged - 20-hour journey yesterday, and feel tiresomely bright and awake after 3 hours' sleep.)

There is no way Labour will poll 37%+ next year - that'd be an increase of 8% points from 2010. Put my money where my mouth is? I have. A four figure sum.

Labour will poll in the 33-35% range.

Isn't that almost exactly the percentage that has switched from the LD's as a result of the coalition?

If you believe that all of them currently saying they've switched are not protesting, have done so permanently and will all actually vote and all actively cast a vote for a Miliband government next year. Some definitely will, but nowhere near all. Although the trend is definitely there and sustained to believe otherwise would be naive. It's like a (prolonged) Cleggasm in reverse and it will have a disappointing climax.

There's a lot of wishful thinking going on here amongst Labour supporters.

Outliergasms!

Well done NPXMP for inventing a cool and funny new word! I like.

Daily polls outside an election campaign are just noise. No one should get too excited. The pollsters use weightings to try to get the results to match what they expected, but we can only assess the accuracy when there is an election, and they will have tinkered with the methodology by then.

I think that there has been hardly any real movement in months. Just noise.

Innocent_Abroad said:

dugarbandier said:

So, in the end, is there actually any advantage to the daily polls?

Would a weekly poll with a sample 5 times as large be more useful?

I bet someone on here has done the sums. Are the daily ones more volatile? And if so, is it noise, or can they resolve trends that more widely spaced polls cannot?

Don't assume that accuracy and marketability are the same thing when it comes to selling polls to clients.

Hmm Just had a quick look at the internals - Labour weighted up from 594 -> 628. OK - But they ain't on 42%. ~38.5% still it is a good poll for Labour.

The REAL internal shockers come from the Lib Dems - even worse than the headline figure:


2010 vote: 404 -> 406 (Weighted)

Current VI: 101 -> 117 (Weighted)

Talk about falling off a cliff !

SouthamObserver said:

The Labour vote share is ridiculously stable. The time to get excited is when/if it starts to dip below 37%. If that happens a few times then something really will be happening. If it starts dipping below 35% then it's game on for the Tories.

The good news for them is that they have a very good chance of repeating their 2010 vote share. The bad news is that this is unlikely to be anywhere near enough to win a majority; which, of course, is great news for the rest of us.

Hah - Yes a Conservative majority is my essential Lay for the next General Election.

A continuation of the coalition could be a brilliant result financially for me ! (NOM, Lib-Con Coalition, and a fair few seats bets should come in)

I think the ONLY weighting should be to past vote VI, the various demographic meanderings should be accounted for in past voting...

Just a personal view mind.

Yougov tends to underestimate Libdem and overstate Labour. Other than that it seemly broadly ok next to other pollsters.

Following my thought - is there anywhere to get the YouGov Data historically in Data format? On the YouGov website all I can find is the individual pdf's of each poll which for data manipulation is worse than the proverbial chocolate teapot?

Well said, Fraser Nelson:

The Tories’ loop of vengeance could sink their election hopes
Many Conservative MPs are more fixated on internal battles over Europe than on winning the public vote in 2015

http://www.telegraph.co.uk/news/politics/conservative/10607626/The-Tories-loop-of-vengeance-could-sink-their-election-hopes.html

Mick_Pork said:

Labour, the tories and the kippers have been basically stable since September last year. Before that there were very telling moves in VI which will likely be repeated come the EU elections.

The only completely reliable and utterly stable polling is the lib dems flatlining on 10% since late 2010. That's not a blip. That's locked in.

They are getting no benefit, boost, drop, surge no matter what has happened politically since 2010. It will take a bomb or truly massive political event to shift them off of that 10% right now. Anyone pinning their hopes on a big lib dem surge, that would get a mass of lib dem MPs that would actually make a hung parliament likely, really should have woken up by now.

If you want to cut out the spikes, more volatile polling and quick moves (because opinion can and does change from week to week or month to month let us never forget) then look at ALL the polls and look at them since 2010. Not just a week or a month.

http://upload.wikimedia.org/wikipedia/en/9/96/UK_opinion_polling_2010-2015.png

To counterbalance though - The Lib Dems won in Eastleigh, the SNP is polling well north of the border and the left looks as if it will vote more tactically this time round. In addition there is a 4th party in the mix that could take add unpredictability, particularly in the east of England. I still think No Overall Majority is a real (Most likely) possibility.

TOPPING said:

CarlottaVance said:

Well said, Fraser Nelson:

The Tories’ loop of vengeance could sink their election hopes
Many Conservative MPs are more fixated on internal battles over Europe than on winning the public vote in 2015

http://www.telegraph.co.uk/news/politics/conservative/10607626/The-Tories-loop-of-vengeance-could-sink-their-election-hopes.html

Is becoming quite a recurring theme but you have to wonder when and if and why those numpty rebels will listen.

I'll go out on a limb here and predict that if Cammie doesn't want any more rebellions then caving in to the tory rebels might not be the best way to ensure that happens. It may even make things worse. Hard to believe I know.

Nor is it just the Cameroon leadership that emerged from yesterday's immigration bill shambles looking clueless and incompetent. This really does bear repeating as it is an object lesson in how not to do things

"My understanding this morning was that the LDs had agreed to the home secretary’s power to remove citizenship only to see off the unacceptable rebel amendments. But the Government has just announced it won’t oppose the Raab amendment"

http://www.libdemvoice.org/why-has-nick-clegg-backed-plan-to-deprive-terror-suspects-of-citizenship-38025.html#comments

Master strategising 'genius'. Cammie was so terrified of his backbenches he begged Clegg to accept all the concessions to the tory rebels (which unbelievably Clegg agrees to) then when it becomes clear the tory rebels have told Cammie where to stick his concessions he caves in anyway and leaves Clegg to vote it out and holding the bag as angry lib dems realise just what he's signed them all up for.

TOPPING said:

In their own minds they are fighting the good fight for freedom and conscience. In actual fact they are proving themselves to be supremely misguided and, which is more surprising, extraordinarily naive. No one is going to thank them for such a fabricated point of principle and, as all the commentaries point out, many will penalise them.

They also think they are in power right now and want to do something with that power while they still have it. Leaders who can only think in terms of what is politically expedient for the next election are no less deluded than rebels who only ever think in the very short term.
The tory rebels didn't all just suddenly pop up one day. This has been long in the making.

Surely the difference is that the election has a margin of error of 0%, if only by definition. To suggest that peoples' views are really changing on an almost daily basis because of the almost random noise produced by polling organisations is a bit of a delusion.

Not quite as delusional as the belief England can play cricket, but close. I hope suicide is no longer illegal in Australia either.

richardDodd said:

Labour still soaring..should be completely out of sight by now...

It doesn't matter. On these figures, it's still PM Milliband.

Opinion polls individually very rarely tell us anything of importance at all (though curiously when they do, they are usually completely overlooked). Even in aggregate they don't tell us very much.

tpfkar said:

The really scary thing? The general election is itself a 'single survey.'. If we accept accuracy of both polls, imagine the difference between a GE run on Thursday and one on Friday. Would give a radically different result from a poll just a day later, but would change complexion of Govt for four / five years.

The scatter in the polls is caused by sampling error (i.e. you are not properly sampling the population). If you had polls the size of the general election on a daily basis, there would be very small scatter.

NickPalmer said:

tpfkar said:

The really scary thing? The general election is itself a 'single survey.'

I remember after one election with a surprise result contradicting the polls, there was a reader's letter in The Times suggesting that in view of the low turnout, the polls were right and the result was wrong :-).

rcs' idea of gradual election is seen in various forms in other countries, e.g. the Senate in the USA. A problem is that it puts the parties into permanent election frenzy, since there's always a crucial election coming up. We've got used to our pattern of governments doing unpopular stuff early in the 5 years, and cynical though that is, perhaps it does meet a real need to give governments space to do difficult things and show the results before having to face voters again.

lennon's idea of rolling averages has also been seen in US polls, where it doesn't have a very successful record. I suspect a reason may be that the subsamples that are being added together are not balanced overall, i.e. you may get (say) an older-voter bit of the Tues-Thur sample combined with an older-voter bit of the Friday sample.

Some very good points there Nick.

How about developing the American approach of separating the legislature from the executive? If some way could be developed of maintaining a stable long term government, supported by a variable HoC with regular elections, we could get a situation where the long term planning is possible.

NickPalmer said:

tpfkar said:

The really scary thing? The general election is itself a 'single survey.'

I remember after one election with a surprise result contradicting the polls, there was a reader's letter in The Times suggesting that in view of the low turnout, the polls were right and the result was wrong :-).

rcs' idea of gradual election is seen in various forms in other countries, e.g. the Senate in the USA. A problem is that it puts the parties into permanent election frenzy, since there's always a crucial election coming up. We've got used to our pattern of governments doing unpopular stuff early in the 5 years, and cynical though that is, perhaps it does meet a real need to give governments space to do difficult things and show the results before having to face voters again.

lennon's idea of rolling averages has also been seen in US polls, where it doesn't have a very successful record. I suspect a reason may be that the subsamples that are being added together are not balanced overall, i.e. you may get (say) an older-voter bit of the Tues-Thur sample combined with an older-voter bit of the Friday sample.

I meant take a rolling average of properly weighted genuine polls (ie the YouGov daily), rather than average unweighted subsamples as they do in the US.

I like the idea of rolling elections. Rather than it being a drawback that governments would need to do unpopular stuff in an election year, it would help force the politicians to treat the electorate like adults.

The single biggest problem that we have in Britain is that our politicians do not trust the electorate. If they started doing so, they might find in turn that the electorate might start trusting them.

Pulpstar said:

Mick_Pork said:

Labour, the tories and the kippers have been basically stable since September last year. Before that there were very telling moves in VI which will likely be repeated come the EU elections.

The only completely reliable and utterly stable polling is the lib dems flatlining on 10% since late 2010. That's not a blip. That's locked in.

They are getting no benefit, boost, drop, surge no matter what has happened politically since 2010. It will take a bomb or truly massive political event to shift them off of that 10% right now. Anyone pinning their hopes on a big lib dem surge, that would get a mass of lib dem MPs that would actually make a hung parliament likely, really should have woken up by now.

If you want to cut out the spikes, more volatile polling and quick moves (because opinion can and does change from week to week or month to month let us never forget) then look at ALL the polls and look at them since 2010. Not just a week or a month.

http://upload.wikimedia.org/wikipedia/en/9/96/UK_opinion_polling_2010-2015.png

To counterbalance though - The Lib Dems won in Eastleigh.

They only just won in Eastleigh but yes, in small tests like one-off by-elections and individual local by-elections they can marshall what strength remains usually pretty effectively. On a much bigger scale like a mass of local elections every year or EU elections this year? Not so much. To say the least.

Pulpstar said:

I still think No Overall Majority is a real (Most likely) possibility.

It's always a possibility, but likely under FPTP with even less lib dem MPs this time around to make the narrow band where all the the votes must fall even narrower in 2015? I can't see it myself.

antifrank said:

I like the idea of rolling elections. Rather than it being a drawback that governments would need to do unpopular stuff in an election year, it would help force the politicians to treat the electorate like adults.

The single biggest problem that we have in Britain is that our politicians do not trust the electorate. If they started doing so, they might find in turn that the electorate might start trusting them.

You're wrong.

What we need is need more plebiscites on the big issues.

Apart from the fillip for democracy that would give, think of the betting opportunities.

antifrank said:

I like the idea of rolling elections. Rather than it being a drawback that governments would need to do unpopular stuff in an election year, it would help force the politicians to treat the electorate like adults.

The single biggest problem that we have in Britain is that our politicians do not trust the electorate. If they started doing so, they might find in turn that the electorate might start trusting them.

It is a 2 way street though - politicians (with media assistance) treat the electorate like children, and so the electorate behave like it. How you break that cycle is difficult as a single politician or party that treats the electorate like an adult gets slaughtered, so no-one else tries.

TGOHF said:

antifrank said:

I like the idea of rolling elections. Rather than it being a drawback that governments would need to do unpopular stuff in an election year, it would help force the politicians to treat the electorate like adults.

The single biggest problem that we have in Britain is that our politicians do not trust the electorate. If they started doing so, they might find in turn that the electorate might start trusting them.

It would be awful and nothing radical would get done - even less than now.

No no and thrice no.

Agreed. The short terms of Congress show what this would be like. Politicians almost totally fixed on running for election and, in safer seats, more scared of challenges from radical elements within the party than anything else. We need politicians who govern for at least part of the time.

Whether the Conservative party remains capable of the discipline necessary to form an administration remains in doubt after yesterday. No wonder some are promoting the idea of Andrew Strauss as a Conservative MP. The parallels between Conservative back benchers and English batsmen are striking.

foxinsoxuk said:

@pork

While 10% is consistent for the LDs, they are not being slaughtered in elections.

If Clegg's ostrich faction truly believe that then the lib dems are in for a world of pain. Year on year in the big mass local election tests they and their councillor and activist base are getting pounded. They were slaughtered in 2011 in scotland and now have a taxi full of MSPs. They're now behind the kippers for the EU elections in scotland. We'll see in May if there is any substance whatsoever to any hypothetical lib dem comeback.

foxinsoxuk said:

The hijinks in parliament yesterday do not matter much. Even as a political nerd it was impossible to follow. The only meme to emerge is that the Tory party is more anti immigration and pro deportation than other parties in parliament. That isn't exactly a shock.

The only thing that emerged from yesterday was what a complete and utter shambles it was. That's why it was impossible to follow and that's why there was and could be no coherent spin from that ludicrous mess.What isn't a shock is that Farage will be laughing long and hard at the extraordinary chaos and rubbing his hands in glee at the fear he clearly provokes inside the tory party.

Is that going to hit VI tomorrow or anytime soon? Nope. Does it mean that tory rebellions are just going to become more likely now that Cammie has caved in to the rebels? Oh yes.

Nor does it speak well of Clegg that he was taken for a ride yet again by Cammie.

Just got back in - not heard the cricket, going to go and hear how the boys are doing now.

Can't be worse than the rest of the tour at least.. I think we're due a convincing win!

Regardless of how terrible we are, 10/1 looks a reasonable bet from here.

TGOHF said:

TheScreamingEagles said:

TGOHF said:

Can Oz reach this total in 10 overs ?

They'll be able to reach this total from just Jade Dernbach's four overs.

England will be down to 8th in the 20/20 rankings.

Perfect timing for the t20 world cup.

Thirsk and Malton Conservative Association will count the votes for Anne McIntosh's reselection ballot today. Result expected at 3PM

isam said:

MSN UK News (@msnuknews)
31/01/2014 09:18
Final averages for all Jan's opinion polls, compared with Dec: LAB 38.0% (-0.8%); CON 32.3% (-0.6%); UKIP 12.6% (+1.5%); LIB 9.9% (-0.1%).

He's a bit premature, there's at least three more polls out with fieldworks that end in January.

TGOHF said:

TheScreamingEagles said:

TGOHF said:

TheScreamingEagles said:

TGOHF said:

Can Oz reach this total in 10 overs ?

They'll be able to reach this total from just Jade Dernbach's four overs.

England will be down to 8th in the 20/20 rankings.

Perfect timing for the t20 world cup.

Poor old Darren Lehman - he has to coach the test, ODI and 20-20 teams - no wonder his team is struggling.

I suspect at least one, maybe two of the following will be gone by the start of the English summer, Flower, Cook (as Captain) or KP

richardDodd said:

Labour still soaring..should be completely out of sight by now...

Why?

@TSE

Indeed. Your point about the cricket is a very strong argument in favour of a daily poll that I'm certain will unite colleagues from both sides of the political divide.

TheScreamingEagles said:

isam said:

TheScreamingEagles said:

isam said:

MSN UK News (@msnuknews)
31/01/2014 09:18
Final averages for all Jan's opinion polls, compared with Dec: LAB 38.0% (-0.8%); CON 32.3% (-0.6%); UKIP 12.6% (+1.5%); LIB 9.9% (-0.1%).

He's a bit premature, there's at least three more polls out with fieldworks that end in January.

Would have thought this is the average of polls published in Jan compared with those published in Dec.

Even by that criteria, he's premature, there's another poll due out in the next few hours.

Haha ok

tpfkar said:

The really scary thing? The general election is itself a 'single survey.'. If we accept accuracy of both polls, imagine the difference between a GE run on Thursday and one on Friday. Would give a radically different result from a poll just a day later, but would change complexion of Govt for four / five years.

I've often wondered whether the council system of electing in thirds would be better for Westminster too. You lose stability, but gain an ability to kick a lost government out quicker, fewer bribes in election year, and more serious opposition as you have to be ready to take over, so less posturing.

Three by-elections every week - political betting nirvana and you'd completely renew the Commons every four years.

Populus ‏@PopulusPolls 10m

New Populus VI figures: Lab 39 (-1); Cons 32 (-1); LD 11 (=); UKIP 10 (+2); Oth 8 (=) Tables http://popu.lu/s_vi140131

@TheScreamingEagles Hmm - All MoE but UKIP up at 10% indicates the underlying support is still steady for them.

Dernbachs giving it the Barry McGuigan to White after going for two dozen in two overs...

Just a pretty face

I think the line earlier in the week from Jack W was: "tick tock tick tock"

Even a stopped clock is right twice a day, Jack.

rcs1000 said:

OK peeps,

Busy economics day today.

UK first. The Lloyds Business Barometer for January is out - and shows a 63 number, up from 48 in December. This is a very strong number for a seasonally weak time of the year. It's also the highest figure for (at least) a decade, and suggests my optimistic 2014 UK GDP number is not wildly ridiculous. The GfK consumer number also showed a much better number than expected - -7 against -12/

Retail sales numbers are out for France and Germany. Unusually, the French number is strong (up 1.4% in December, against an expected 0.9%), while the German number is weak (down 2.4%, against expected growth of 1.9%). Good news from Greece, where retail sales came out flat in December, suggesting the truly horrendous recession there might be nearing an end.

Italian unemployment numbers are also out, and show a slightly surprising reduction to a - still high - 12.7%. Eurozone unemployment edged down similarly to 12.0%.

Finally, Spain continued its recent run of current account surpluses - posting a €900m one in November, slightly down on October.

Did I miss it earlier, or did you not post the unexpectedly low Euro area Jan CPI number? 0.7% vs 0.9% expected and 0.8% previous month. Interest rate cuts from Draghi?

Lennon said:

Did I miss it earlier, or did you not post the unexpectedly low Euro area Jan CPI number? 0.7% vs 0.9% expected and 0.8% previous month. Interest rate cuts from Draghi?

You didn't miss it, I did :-)

With interest rates down at 0.25%, there is no room for further cuts to rates.

If we see continued weak inflation numbers in the EZ, I would suspect we'll probably see more unconventional measures - more cheap financing for peripheral banks, and perhaps some intervention in the corporate bond market.

However, I don't think this datapoint on its own will be enough to cause action, especially with everywhere (bar France) appearing to be on the (very gradual) mend.

Moray by election yesterday had SNP vote down around 9%
Ind Cowie 830 SNP 670 Ind Macrae 220 Con 143
after eliminations Ind Cowie 1034 SNP 710

‏@StewartWood
Update on that Conservative Party poll bounce:
-Nov 29 2012: 32%
-June 19 2013: 32%
-Oct 23 2013: 32%
-Dec 3 2013: 32%
-Today: 32%

@YouGov

Chuckle.