France 2012: the official thread

[DL]

Actually I don't think a leftwing candidate for the presidency in France has ever has less than 46-47% of the vote in the second round and that would have been Jospin in 1995.

France at the presidential level in that barring a fluke like LePen being in the second round in 2002 - each side has a floor of about 45% - even though most polls give Hollande a 12-16% lead in the end if he wins by 7 or 8% it will be a landslide by French standards. 

[Colbert]

i speaked about the first round

[Californian Tony]

How many polls do we have this week ? Has there been others apart from the IFOP one ?

[big bad fab]

2012 Big Bad Tracker #31 - 5 December 2011      
            
Each opinion poll is weighted with its sample and loses a ponderation of 15% each week: it may change after the autumn.            

I haven't given a ponderation based on the quality of pollsters, as big changes have occurred in their teams, without any certainty that the 2007 rating is still OK.      
And this tracker is built more out of fun than out of scientific seriousness...       

      

Hollande   31,98
Sarkozy   26,41
Le Pen   17,78
Bayrou   7,01
Mélenchon   6,55
Joly   4,60
Villepin   1,62
Morin   0,73
Lepage   0,40
Dupont-Aignan   0,88
Boutin   0,39
Nihous   0,07
Arthaud   0,45
Poutou   0,39
Chevènement   0,76
   
Hollande   58,49
Sarkozy   41,51

Only one new poll this week. A rather bad one for Hollande, but, on the other hand, popularity is again down for Sarkozy and Fillon.

The only notable feat is that Bayrou is at 7 again.
Le Pen is very resilient and really strong.
Villepin is almost out.

[DL]

2012 Big Bad Tracker #31 - 5 December 2011      
            
Each opinion poll is ponderated with its sample and loses a ponderation of 15% each week: it may change after the autumn.            

I haven't given a ponderation based on the quality of pollsters, as big changes have occurred in their teams, without any certainty that the 2007 rating is still OK.      
And this tracker is built more out of fun than out of scientific seriousness...       

FYI: In English we don't say "ponderated" or "ponderation" - we say "weighted" and "weighting"

[Californian Tony]

Fab, you're gonna love these graphs.

1st Round :



2nd Round :




Anyways, the good thing is that polls start to become regular (we tend to have 1-2 polls per week rather than 4 one week and nothing the next one), and as a result, the lines are smoother. No your tracker's style becomes effective.

[big bad fab]

Hold on... there is another IPSOS poll, made on 2nd and 3rd of December, but published only this morning... sigh...
I'll update the tracker tonight. Fortunately, it's very, very close to the tracker !

[Californian Tony]

Hold on... there is another IPSOS poll, made on 2nd and 3rd of December, but published only this morning... sigh...
I'll update the tracker tonight. Fortunately, it's very, very close to the tracker !

What are the numbers ?

[big bad fab]

2012 Big Bad Tracker #31 - 5 December 2011
   
            
Each opinion poll is weighted with its sample and loses a ponderation of 15% each week: it may change after the autumn.            

I haven't given a ponderation based on the quality of pollsters, as big changes have occurred in their teams, without any certainty that the 2007 rating is still OK.      
And this tracker is built more out of fun than out of scientific seriousness...       


BEWARE ! UPDATED WITH IPSOS AND BVA POLLS !      

Hollande   32,26
Sarkozy   26,13
Le Pen   17,62
Bayrou   7,20
Mélenchon   6,70
Joly   4,61
Villepin   1,60
Morin   0,63
Lepage   0,46
Dupont-Aignan   0,80
Boutin   0,40
Nihous   0,05
Arthaud   0,47
Poutou   0,31
Chevènement   0,75
   
Hollande 58,73
Sarkozy   41,27

I'm forced to include now the latest polls, so that recent polls aren't too undervalued.
I don't intend to change the way I decrease the weight of old polls: I'll do it from the 1st of January. I know French medias too well: the first weeks of December seem wonderful, but the last 2 weeks will see no poll at all. But, afterwards, it'll be OK and I'll fasten the rhythm in January, a lil' more in February, even more in March, a lot more in the first 2 weeks of April and the 3rd week of April will be based only on the last 2 weeks.

BTW, Bayrou seems to be stronger and stronger. Mélenchon too. And Sarkozy is stalled again.

[Californian Tony]

Fabien, I'm kind of worried by the precedent you are setting here. Shouldn't the rule be that the polls are integrated in the tracker in the first monday following their publishing ? Using the polling date rather than the publishing date will mean the possibility for retro-active editing of the tracker (like you just did). It really makes things dangerously confusing.

[big bad fab]

Fabien, I'm kind of worried by the precedent you are setting here. Shouldn't the rule be that the polls are integrated in the tracker in the first monday following their publishing ? Using the polling date rather than the publishing date will mean the possibility for retro-active editing of the tracker (like you just did). It really makes things dangerously confusing.

Maybe, I must think about it.
More and more, polls will be published sooner after the real polling date. And when I'll start to reduce the number of polls taken into account, the tracker would be too wrong.
On the other hand, you're right that it would mean that the tracker may be updated too often.

Let me think a bit (I've just had a very, very bad news in my job: some chambres régionales des comptes are going to be erased in the coming months...; and so I'm not able to think properly for the moment).

[Californian Tony]

Fabien, I'm kind of worried by the precedent you are setting here. Shouldn't the rule be that the polls are integrated in the tracker in the first monday following their publishing ? Using the polling date rather than the publishing date will mean the possibility for retro-active editing of the tracker (like you just did). It really makes things dangerously confusing.

Maybe, I must think about it.
More and more, polls will be published sooner after the real polling date. And when I'll start to reduce the number of polls taken into account, the tracker would be too wrong.
On the other hand, you're right that it would mean that the tracker may be updated too often.

Pardon my psycho-rigidity, but to me it's really important that the method we have for accounting polls is uniform and does not variate depending on the tracker's conveniences.

Let me think a bit (I've just had a very, very bad news in my job: some chambres régionales des comptes are going to be erased in the coming months...; and so I'm not able to think properly for the moment).

I'm sorry to hear that and I really hope it won't affect you. These are really tough times, it seems...

[Math]

IMO the tracker should be based not on the publishing dates but on the field dates, the only ones which make sense if you're tracking voting intentions. Maybe you can use all the polls whose fields are executed from Monday to Monday (or any other seven days round of course), and publish the tracker only two or three days after, in order to wait for the publication of all of them?

Anyway, I wish you good luck with your work. You shall overcome!

[JulioMadrid]

I wish you good luck with your work. You shall overcome!

[Californian Tony]

For tomorrow's tracker, I strongly recommend you to count the newest 2 polls at their 100% value, since they have been published after previous Monday.

[Hashemite]

The awfulness of the candidates here is really something: some indecisive wimp who is boring and predictable, a useless incumbent who associates with Vichyists, a closeted fascist, an egomaniac whose platform is that everybody sucks, a wannabe Marchais who puts on a pathetically fake scene, a useless end-is-nigh douche who sabotaged a party, an incompetent wannabe running for the sake of a personal vendetta, a useless boring hack who takes himself way too seriously, a failed wannabe actor who doesn't understand that nobody wants him to play the General's role and two boring nutjob Trots.

I wonder who I'll vote for. I'll probably make Excel generate a random number or roll a dice.

[Californian Tony]

The awfulness of the candidates here is really something: some indecisive wimp who is boring and predictable, a useless incumbent who associates with Vichyists, a closeted fascist, an egomaniac whose platform is that everybody sucks, a wannabe Marchais who puts on a pathetically fake scene, a useless end-is-nigh douche who sabotaged a party, an incompetent wannabe running for the sake of a personal vendetta, a useless boring hack who takes himself way too seriously, a failed wannabe actor who doesn't understand that nobody wants him to play the General's role and two boring nutjob Trots.

I wonder who I'll vote for. I'll probably make Excel generate a random number or roll a dice.

It's hard to give you wrong. But there are clear "lesser evil" options.

[big bad fab]

For tomorrow's tracker, I strongly recommend you to count the newest 2 polls at their 100% value, since they have been published after previous Monday.

Do not take into account my "update" for last week. And for tomorrow, I'll do as usual.
And I'll include the last LH2 poll, which is probably another crap one, but in which Bayrou is up and Le Pen (too much) down)

[big bad fab]

Again, I'm awfully busy this week.
I'll try to publish the tracker tonight...

[Californian Tony]

Again, I'm awfully busy this week.
I'll try to publish the tracker tonight...

Don't worry. Do what you have to do before all.

[big bad fab]

2012 Big Bad Tracker #33 - 12 December 2011      
            
Each opinion poll is weighted with its sample and loses a ponderation of 15% each week: it may change after the autumn.            

I haven't given a ponderation based on the quality of pollsters, as big changes have occurred in their teams, without any certainty that the 2007 rating is still OK.      
And this tracker is built more out of fun than out of scientific seriousness...       

      

Hollande   31,91
Sarkozy   26,21
Le Pen   17,09
Bayrou   7,98
Mélenchon   6,76
Joly   4,59
Villepin   1,50
Morin   0,69
Lepage   0,42
Dupont-Aignan   0,85
Boutin   0,49
Nihous   0,04
Arthaud   0,46
Poutou   0,26
Chevènement   0,77
   
Hollande   58,34
Sarkozy   41,66

Hollande is still very strong.
Bayrou is surging, as anticipated, and he will surge even more next week, with the new CSA poll.
While he may have been, only 3 weeks ago, a threat for Hollande, he is now clearly a danger for... Sarkozy. When France lose its AAA rating, Sarkozy will probably lose voters to Bayrou, while still being under Le Pen's threat.

I've written about Sarkozy being crushed between Le Pen and a moderate socialist (first DSK, then Hollande), but he may well be crushed only between Le Pen and Bayrou...
This campaign will be excruciating for him...

And even if he can gain some ground in the 2nd round, it's only a transition: Hollande is much too high in polls and.... Bayrou may well threaten him too, either directly, or by weakening him enough to allow Le Pen to make it to the second round...

Mélenchon is in relative good shape but he doesn't manage to reach 10%, even 8%.
The other candidates are low or very low.

[big bad fab]

There is another CSA poll today, with Bayrou at 11 (after 13 in the latest LH2).
See my blog.

[Sibboleth]

Must we put up with a surge of militant tepidity at every Presidential election from 2007 onwards?

[big bad fab]

Bayrou isn't exactly the embodiment of tepidity, I think.

Let me be clear: deep inside, he has a weak personality and he is a coward and he doesn't stand for anything really consistent.

But the support he is receiving isn't about "moderation", "consensus", "centre", in the scandinavian meaning of these words, whatever the results of the CSA poll on French people wanting a "national union" government.
This isn't the French tradition at all (see my post in the Essays thread on homely's work on France since 1750).

Bayrou isn't Adolfo Suarez, Romano Prodi, Frederik Reinfeldt or Jean Chrétien ().

Bayrou is just another little proud politician trying to be the "Saviour", like Royal, Sarkozy, Villepin, Le Pen, Villiers, Mélenchon were or are (DSK was also a "Saviour", but not voluntarily...).
He has benefited from the big mistake of the French right: creating a "parti unique", crushing the centre-right.

His only interest is in the complexity of his electorate:
ecologists disappointed with ideological Joly,
real centrists (heirs of the MRP),
soft rightists (old and "kind" people),
über-realistic social-democrats who are afraid of Hamon-Emmanuelli-Montebourg-Aubry inside the PS,
neo-poujadists who were behind Le Pen once but who are not especially racists,
young people who think they can re-create politics and ideas and the world and who think every idea of the past is dead (well, in a way, some Pirate spirit here... i.e. selfish people who think they are really rebellious and they are freedom fighters just because they want to consume everything, everywhere, whenever they want and for free).

He is more the candidate of those who are politically "lost" or orphans (he is even gaining ground among workers in the CSA poll !!!) and of those who are (or think they are) "idealists".
Nowadays, with a dull Hollande, a worked-up Sarkozy and radical Le Pen and Mélenchon, he is still, like in 2007, an easy outlet, a not-so-bad receptacle, for the disappointed voters.

Unfortunately, he is the only way to have fun and surprises during this election, which is, as predited, a boring one for the moment (as Hollande is a shoo-in, for the moment, despite all his own problems).

[Sibboleth]

That's why I added 'militant'. He aggressively favours warm ice cream.