Urban75 Home About Offline BrixtonBuzz Contact

The London elections - your predictions

scifisam said:
Boris at the helm would make me want to move, but I wouldn't really, of course.

I bet it'll go to the second round, and Ken will win that, just about. But I'm so uncertain of that that there's no way I'd put money on it.
Click to expand...
Pedantic gambler's point, but this post is exactly why you should put money on Ken at the odds being offered. If you are right (and I think you are), the bookies' odds on Ken are way too high - when you're betting the end result is irrelevant, it's the odds that matter.

A punt on Ken at 2:1 is a bargain. 1.5:1 is very profitable. This is because he's probably, in truth, no worse than evens, so they're offering to pay you more than the bet is really worth if he wins. If you keep making bets that are theoretically profitable, you will eventually turn a real profit. :)
 
Jografer said:
Unlikely imho, not unless Labour want to kiss goodbye to even more seats than they already look likely to lose....

If they just pissed off 1 in 50 Labour supporters, it would be an additional 2% swing, which on current polls would mean losing 13 seats...

...piss off 1 in 20 they'd lose 33 :eek:
Click to expand...

Indeed, but they are already doing it in Scotland, and I doubt Labour lose any sleep over the effect of their decisions on their core vote because, lets face it, they havent given two hoots thus far - the vote either doesnt turn out, or finds the alternative too unpalatable even after all the bollocks since 1997 and votes Labour anyway.
 
agricola said:
Indeed, but they are already doing it in Scotland, and I doubt Labour lose any sleep over the effect of their decisions on their core vote because, lets face it, they havent given two hoots thus far
Click to expand...

yes, but as polls go down more MPs start jumping about....
 
Jografer said:
yes, but as polls go down more MPs start jumping about....
Click to expand...
Yes. Their willingness to challenge Brown will go up as the likelihood of hanging onto their seats goes down. Blair got away with so much shit for so long because he was seen as their ticket to an easy election victory.
 
Jografer said:
yes, but as polls go down more MPs start jumping about....
Click to expand...

Indeed (that is of course the main reason for the 10p tax "rebellion" and its swift end) - so perhaps a Boris victory would not be so bad, given that it would be one nail in Gordon's coffin (a failure on 42 days would probably be the other).
 
agricola said:
Indeed (that is of course the main reason for the 10p tax "rebellion" and its swift end) - so perhaps a Boris victory would not be so bad, given that it would be one nail in Gordon's coffin (a failure on 42 days would probably be the other).
Click to expand...

... conversely, it will show everyone what an old-etonian tory will do with a bit of power, which is why some tory commentators are hoping for a Ken victory....
 
Jografer said:
... conversely, it will show everyone what an old-etonian tory will do with a bit of power, which is why some tory commentators are hoping for a Ken victory....
Click to expand...

This is true, and in addition when Boris was being encouraged to stand there was an aspect of that it might deflect him away from Cameron (who, lest we forget was rated very poorly only 12 months ago).

Of course, whether there is any identifiable difference between an old-Etonian Tory and the likes of Brown is another issue - certainly, with hindsight Major's government was a lot better than Blair's was, or Brown's thus far.
 
jæd said:
I wouldn't trust any of those polls since they're only for first choices anyway... I think that Ken will win, but will only scrape in.
Click to expand...
If you follow the link in the source on that page, you can get the second round poll results:

Who For, Date, Boris, Ken, Boris Lead
Code: 
YouGov/Evening Standard	30/04/08		53	47	+6
YouGov/Evening Standard	25/04/08		55	45	+10
Ipsos MORI/Unison	24/04/08		48	52	-4
mruk/Sunday Times	??/04/08		49	51	-2
YouGov/Evening Standard	18/04/08		53	47	+6
mruk Cello/Sunday Times	14/04/08		50	50	0
YouGov/Evening Standard	11/04/08		54	46	+8
Ipsos MORI/Observer	09/04/08		50	48	+2
Ipsos MORI/Unison	07/04/08		51	49	-2
YouGov/Evening Standard	04/04/08		56	44	+12
ICM/Guardian		01/04/08		51	49	+2
 
Crispy said:
If you follow the link in the source on that page, you can get the second round poll results:

Who For, Date, Boris, Ken, Boris Lead
Code: 
YouGov/Evening Standard	30/04/08		53	47	+6
YouGov/Evening Standard	25/04/08		55	45	+10
Ipsos MORI/Unison	24/04/08		48	52	-4
mruk/Sunday Times	??/04/08		49	51	-2
YouGov/Evening Standard	18/04/08		53	47	+6
mruk Cello/Sunday Times	14/04/08		50	50	0
YouGov/Evening Standard	11/04/08		54	46	+8
Ipsos MORI/Observer	09/04/08		50	48	+2
Ipsos MORI/Unison	07/04/08		51	49	-2
YouGov/Evening Standard	04/04/08		56	44	+12
ICM/Guardian		01/04/08		51	49	+2
Click to expand...

I did some analysis on the first round polls you linked to earlier. There's even a graph if you can tell me how to post it. EDIT Scratch that - worked it out. :)

NB: The figures for the second round follow the same pattern as those for the first round. Ken is ahead in both. :D

The output from stata is as follows:

Code: 
. metan p se , by(pollster)

           Study     |     ES    [95% Conf. Interval]     % Weight
---------------------+---------------------------------------------------
     0
2                    |   -0.010    -0.069     0.049          6.25
3                    |   -0.030    -0.089     0.029          6.25
5                    |   -0.010    -0.069     0.049          6.25
7                    |    0.060     0.001     0.119          6.25
8                    |   -0.010    -0.069     0.049          6.25
10                   |    0.010    -0.049     0.069          6.25
13                   |   -0.040    -0.099     0.019          6.25
14                   |   -0.040    -0.099     0.019          6.25
15                   |   -0.010    -0.069     0.049          6.25
16                   |   -0.060    -0.119    -0.001          6.25
 Sub-total           |
  I-V pooled ES      |   -0.014    -0.033     0.005         62.50
---------------------+---------------------------------------------------
     1
1                    |    0.110     0.051     0.169          6.25
4                    |    0.070     0.011     0.129          6.25
6                    |    0.060     0.001     0.119          6.25
9                    |    0.130     0.071     0.189          6.25
11                   |    0.100     0.041     0.159          6.25
12                   |    0.120     0.061     0.179          6.25
 Sub-total           |
  I-V pooled ES      |    0.098     0.074     0.122         37.50
---------------------+---------------------------------------------------
Overall              |           
  I-V pooled ES      |    0.028     0.013     0.043        100.00
---------------------+---------------------------------------------------

Test(s) of heterogeneity:
               Heterogeneity  degrees of
                 statistic     freedom      P    I-squared**
0                   10.93          9      0.280     17.7%
1                    4.31          5      0.505      0.0%
Overall             67.83         15      0.000     77.9%
Overall Test for heterogeneity between sub-groups: 
                    52.58          1      0.000

Polls not conducted by YouGov/Evening Standard estimate an average lead for Ken of 1.4% in the first round, with a 95%CI which almost excludes Boris getting more votes in the first round.

By contrast, the ES commissioned polls suggest an average 9.8% margin for Boris with a 95%CI that entirely excludes Ken getting more votes in the first round.

Both subgroups are internally very homogeneous (ie the non-ES polls agree with each other, and the ES polls agree with each other), but disagree vehemently with each other. This suggests very strongly that one group of pollsters is getting it very wrong.




E2A: No sample sizes given in the wiki article, so I gave all the polls the standard +/- 3% margin of error, which is approximately equivalent to assuming they sampled approx 800 people each).
 

Attachments

  • LondonMayorPolls.png
    LondonMayorPolls.png
    13.7 KB · Views: 70
agricola said:
certainly, with hindsight Major's government was a lot better than Blair's was, or Brown's thus far.
Click to expand...

... now there is a matter of opinion... :eek: & not mine.... :hmm:

.. but then again I was paying a mortgage in 1992...... never mind the rest of the shit that went on......

.... Major was probably a nicer bloke, but he didn't have a cat in hells chance of leading the tories at that time. Also they were massively more unpopular than Labour is at the moment, which took some doing....

.... rose tinted specs imho...
 
ymu said:
That they've realised how shit those polls are and they're about to take a kicking. :D
Click to expand...

Good thing I placed my bet at lunch-time... :D

Cool analysis on the polls. I certainly feel Ken might get in again. :( I'm going to vote for Brian, so at least my winnings will buy me enough beer to console me. :D
 
jæd said:
I wouldn't trust any of those polls since they're only for first choices anyway... I think that Ken will win, but will only scrape in.
Click to expand...
2nd prefs are tougher to predict from the polls because whether or not they count depends on what the first preference was (quite a few people seem to be voting for both Boris and Ken, for example).

Anyways, I've updated the analysis to include the 2nd round numbers Crispy posted. Identical picture. The first and third group are non-ES polls, the 2nd and 4th are ES ones. It's ridiculous. Ken won't get a huge margin, but he's unlikely to be beaten.
 

Attachments

  • London1+2.png
    London1+2.png
    17.1 KB · Views: 223
Jografer said:
... now there is a matter of opinion... :eek: & not mine.... :hmm:

.. but then again I was paying a mortgage in 1992...... never mind the rest of the shit that went on......

.... Major was probably a nicer bloke, but he didn't have a cat in hells chance of leading the tories at that time. Also they were massively more unpopular than Labour is at the moment, which took some doing....

.... rose tinted specs imho...
Click to expand...

Well the economic situation in 1992 was not entirely down to him, and the rest of the shit tbf pales into comparison with what has been going on in the last seven years.

I agree that by 1997 they were hugely unpopular, but one must never forget that since 1997 this Government has - nearly always successfully - managed to con (or buy off, in the case of Murdoch) the media into painting a very positive image of it, even when they should by all rights have been hung out to dry - Iraq being the most obvious example.
 
Mitre10 said:
An article from the Guardian stating that Boris is in the lead which smacks of nothing else but trying to scare any of their not-yet-voted Ken-voters out of the house and get to the polling station.

http://www.guardian.co.uk/politics/2008/may/01/london08.localgovernment
Click to expand...

Interesting - thanks.

The final YouGov poll for the Evening Standard gave the Conservative candidate a lead of seven points on first-preference votes over the mayor, Ken Livingstone, down from an 11-point lead earlier in the week.

It said 43% of voters were going to back Johnson with their first vote, and 36% supported Livingstone. After the re-allocation of second-preference votes, it predicted a Johnson victory by 53% to 47%.

However, YouGov has consistently given Johnson higher leads than other polling organisations and many analysts believe the race is too close to call. A poll for Ipsos Mori last week put Livingstone 4% ahead and another for MRUK put him 2% in front.

Livingstone, who was predicting yesterday that he would win by 52% to 48%, voted early near his home in Cricklewood, north London.
Click to expand...
So even the very biased YouGov polls have Ken slightly in the lead on 2nd preference votes, narrowing the fictitious 7% margin to 6% at the end.

I think Ken is probably not far wrong with the 52:48 victory overall, taking account of the less obviously biased polls.
 
I think Ken will edge it, and we'll all breathe a sigh of relief. Those that voted for Boris will probably feel a bit stupid next time he comes out with something racist or homophobic.
 
the thing with polls is that they ask a random (maybe) sample of people, not a random sample of voters. theres usually a pretty poor turnout, and it seems very likely to me that the people that dont vote are probably systematically different to those that do. so without a way of working out how much of your polled sample is actually going to go, they are pretty meaningless (even with CIs, nice work though).

theres also the possibility that people say one thing and do another... personally i reckon ken will come in front, people are less risk averse when answering a poll than they are with making a real decision about who is in charge of the police and transport system, and lets be honest, boris is the risky candidate.
 
gracious said:
the thing with polls is that they ask a random (maybe) sample of people, not a random sample of voters. theres usually a pretty poor turnout, and it seems very likely to me that the people that dont vote are probably systematically different to those that do. so without a way of working out how much of your polled sample is actually going to go, they are pretty meaningless (even with CIs, nice work though).

theres also the possibility that people say one thing and do another... personally i reckon ken will come in front, people are less risk averse when answering a poll than they are with making a real decision about who is in charge of the police and transport system, and lets be honest, boris is the risky candidate.
Click to expand...
According to Peter Snow on some programme or other, the average margin of error for polls has been 2% since (I think) WWII. It's fairly rare that they get it very wrong. Most of them do adjust for voting intentions, AFAIK.

I think you can discount the Evening Standard polls because they're clearly biased. The rest are statistically very homogeneous with the scatter around the pooled value being pretty much what you'd expect from normal sampling error (unlike the scatter when you include the ES polls, which is far far greater than just random noise).
 
Polls now closed. Punters are still punting on Boris on betfair at odds as tight as 0.11 to 1...... Ken is 6 to 1

Either Ken is out, or there are gonna be some red faces (and red bank accounts because thousands has changed hands on this)

EDIT someone has just stuck 2 grand on Boris at 0.14 !
 
Conservativehome says "Based on a wide range of conversations we've had throughout the day with people in the field and with Tory and other insiders we are confident that on the basis of patterns of turnout, postal votes and canvass returns Boris Johnson has been elected Mayor of London."

Proberbly the most biased site on the interweb of course.................
 
According to Radio 4, Boris is pretty certain to win. How fucking depressing that we have elected a racist, homophobic adulterer as our mayor. Proof that the majority of Londoners are idiots :(
 
You must log in or register to reply here.

Similar threads

editor
The best 'dive bars' in London according to 'experts'
2 3 4
Replies
119
Views
7K
LeytonCatLady
LeytonCatLady
brogdale
The London Ringways
Replies
13
Views
1K
PR1Berske
PR1Berske
Cloo
London culture club - request a theatre/ dance/museum etc companion
Replies
3
Views
530
Guineveretoo
Guineveretoo
BigMoaner
South London Press Closes
Replies
9
Views
728
jimbarkanoodle
J
editor
The Old London Photos Facebook group
Replies
15
Views
973
Tankus
Tankus
Back
Top Bottom