Austrian Parliamentary Election - Sept. 29

[peterould]

Is it just me or do the SPÖ videos look like a "core vote" strategy?

[peterould]

Not just me then!

A little introduction folks. I'm a British Citizen (technically subject of the Crown) but from a large Austrian family. Grew up between Yorkshire, Austria and bits of the Rheinland Pfalz.

I work as a consultant statistician and am also a Church of England priest (as you do). Always had a healthy interest in psephology and am currently working on a Nate Silver style model for the UK electoral scene, but will be testing out a few things from that methodoloy on the Austrian and German elections (and I'll post links when they're up).

Enough for now.

[peterould]

Not just me then!

A little introduction folks. I'm a British Citizen (technically subject of the Crown) but from a large Austrian family. Grew up between Yorkshire, Austria and bits of the Rheinland Pfalz.

I work as a consultant statistician and am also a Church of England priest (as you do). Always had a healthy interest in psephology and am currently working on a Nate Silver style model for the UK electoral scene, but will be testing out a few things from that methodoloy on the Austrian and German elections (and I'll post links when they're up).

Enough for now.

Cool. Welcome to the Forum !



In which parts of Austria did you grow up ?

Family is from Bad Aussee (where my Omi still is) - we spent every summer there and some winters.

[peterould]

Always had a healthy interest in psephology and am currently working on a Nate Silver style model for the UK electoral scene, but will be testing out a few things from that methodoloy on the Austrian and German elections (and I'll post links when they're up).

Very good. Please complete your model and become the "Nate Silver" of the UK and Europe. The Silver style models are great, but their restriction to US politics is a huge disadvantage.

I would post a link now but apparently as a newbie I'm not allowed to until I've done twenty posts. Grrrrrrr.....

[peterould]

Always had a healthy interest in psephology and am currently working on a Nate Silver style model for the UK electoral scene, but will be testing out a few things from that methodoloy on the Austrian and German elections (and I'll post links when they're up).

Very good. Please complete your model and become the "Nate Silver" of the UK and Europe. The Silver style models are great, but their restriction to US politics is a huge disadvantage.

I would post a link now but apparently as a newbie I'm not allowed to until I've done twenty posts. Grrrrrrr.....

You can do it like this:

www (point) link (point) uk/com whatever ...

www (dot) peter-ould (dot) net/austria-election-predict/

[peterould]

Always had a healthy interest in psephology and am currently working on a Nate Silver style model for the UK electoral scene, but will be testing out a few things from that methodoloy on the Austrian and German elections (and I'll post links when they're up).

Very good. Please complete your model and become the "Nate Silver" of the UK and Europe. The Silver style models are great, but their restriction to US politics is a huge disadvantage.

I would post a link now but apparently as a newbie I'm not allowed to until I've done twenty posts. Grrrrrrr.....

You can do it like this:

www (point) link (point) uk/com whatever ...

www (dot) peter-ould (dot) net/austria-election-predict/

Looks good.

I have used some D'Hondt calculators as well, which have produced roughly the same seat estimates.

It depends though if the polls are right and the BZÖ is currently out, or if we assume that the BZÖ picks up some support during the "real" campaign that starts in the last August-week and gets above 4%. We also have to watch NEOS, starting from early September.

Yes, if the polls are out (Can anyone say "BZÖ" and "2008") then the model is wrong! At the moment I'm running at around 25% chance that the BZÖ get past the 4% mark - I don't think they're going to do well enough to win a seat outright anywhere. What's remarkable in the polls at the moment is that things are broadly static - 100% chance SPÖ will come first and 100% chance ÖVP will be in second place.

[peterould]

Always had a healthy interest in psephology and am currently working on a Nate Silver style model for the UK electoral scene, but will be testing out a few things from that methodoloy on the Austrian and German elections (and I'll post links when they're up).

Very good. Please complete your model and become the "Nate Silver" of the UK and Europe. The Silver style models are great, but their restriction to US politics is a huge disadvantage.

I would post a link now but apparently as a newbie I'm not allowed to until I've done twenty posts. Grrrrrrr.....

You can do it like this:

www (point) link (point) uk/com whatever ...

www (dot) peter-ould (dot) net/austria-election-predict/

Looks good.

I have used some D'Hondt calculators as well, which have produced roughly the same seat estimates.

It depends though if the polls are right and the BZÖ is currently out, or if we assume that the BZÖ picks up some support during the "real" campaign that starts in the last August-week and gets above 4%. We also have to watch NEOS, starting from early September.

Yes, if the polls are out (Can anyone say "BZÖ" and "2008") then the model is wrong! At the moment I'm running at around 25% chance that the BZÖ get past the 4% mark - I don't think they're going to do well enough to win a seat outright anywhere. What's remarkable in the polls at the moment is that things are broadly static - 100% chance SPÖ will come first and 100% chance ÖVP will be in second place.

But sometimes Austrian polls are off the mark (usually not at federal elections, more in the state elections - like Salzburg, Tyrol and Carinthia, where they failed big time).

Many polls that are released such as Gallup only have samples of 400 people, with a 5% margin of error. The ÖVP is only 3% behind. So, there's still everything possible and the ÖVP could still overtake the SPÖ on election day.

The Greens will not overtake the FPÖ though and TS is somewhere in between 5-10%.

The other interesting thing will be if the BZÖ and NEOS get beyond the 4% threshold and what turnout will be like.

Considering all the graft and corruption scandals in the past 5 years from ÖVP/FPÖ/FPK/BZÖ and the examined scandals from the Schüssel-time, as well as a more-or-less boring election campaign, I expect turnout to drop from 79% in 2008 to around 70-75% in this election.

True, the polls have small samples, but I am very clever (he said humbly) so what I do is get as close to the raw data as I can, standardise it and produce a master poll together with variances.

[peterould]

Have we seen the videos at neuwal (dot) com (slash) index.php/2013/07/19/and-the-wahlkampfvideo-oscar-goes-to/ ?

The Greens one is by far the best!

[peterould]

Added the new polls into my model and for the first time there's a smidgeon of a hint of the slightest possibility of an ÖVP win (as in largest party). Probability? 0.01%

Let's see what tomorrow's polls bring.

[peterould]

Added the new polls into my model and for the first time there's a smidgeon of a hint of the slightest possibility of an ÖVP win (as in largest party). Probability? 0.01%

Let's see what tomorrow's polls bring.

Don't think there are any polls tomorrow. Gallup will release one next week again and the Profil one is only monthly.

I'm actually more interested in the turnout figure right now.

The Spectra poll said that 76% are "certain to vote" this time, with another 14% who "are likely to vote".

Turnout in 2008 was 79%.

Historically, polls with these models were very accurate, meaning that the actual turnout on election day mirrored the "certain"-share in the polls, plus a few from the "likely" category.

I actually thought turnout will drop significantly this year, mostly because the state elections already showed a turnout drop.

The poll could just be wrong as well ...

I thought Gallup were doing weekly polls up to the vote?

[peterould]

No, every 2 weeks.

They just always publish the results of the previous week and say they are "new" ...

***Looks back through poll database***

Gosh, you're right. How cheap is that?

[peterould]

Assuming no more polls tomorrow, I have updated my predictions for this weekend.

[peterould]

Assuming no more polls tomorrow, I have updated my predictions for this weekend.

My calculation with a D'Hondt calculator would add a seat to the FPÖ from TS.

All other parties' seats are identical with your estimate.

I used the average of the July polls, excl. the earlier Market poll.

The difference between you and me is

i) The cunning way I weight the polls
ii) I'm running 10,000 simulations and reporting the most likely outcome!

My outcome occurs 2.12% of the time, yours 1.19%

[peterould]

I've updated my weekend's prediction based on including the IMAS poll. Definitely beginning to see Team Stronach picking up steam. Although I'm not showing them on the page, I'm starting to get scenarios at both fringes of the Greens' vote where they beat the FPÖ into third place (0.01%) or conversely they get pushed by Team Stronach into fifth place (0.22%). Expect that scenario to become more likely as Team Stronach pick up steam.

[peterould]

I've updated my weekend's prediction based on including the IMAS poll. Definitely beginning to see Team Stronach picking up steam. Although I'm not showing them on the page, I'm starting to get scenarios at both fringes of the Greens' vote where they beat the FPÖ into third place (0.01%) or conversely they get pushed by Team Stronach into fifth place (0.22%). Expect that scenario to become more likely as Team Stronach pick up steam.

It would be weird somehow if the election ended up almost identical to the 2008 election.

SPÖ, ÖVP, FPÖ, Greens all with the same share as 2008 and the TS taking the place of the BZÖ, which takes the place of the 2008 Citizen Forum, while NEOS takes the LIF share ...

It would be nice to have an election campaign with some movement in it! TS seem to now be picking up a bit of the support from the BZÖ as they sink into oblivion. I wonder whether NEOS will make any impact.

[peterould]

I was in Austria for most of the election last time. I thought some of the TV debates were great. in particular Strache pretty well demolised Faymann. Even my middle of the road aunt watching it with me was saying "I don't like Strache but he's absolutely winning this one".

[peterould]

I was in Austria for most of the election last time. I thought some of the TV debates were great. in particular Strache pretty well demolised Faymann. Even my middle of the road aunt watching it with me was saying "I don't like Strache but he's absolutely winning this one".

I can't really remember the debates anymore, because there were so many of them (10-20).

But I think Haider did a good job at that time, that's why the BZÖ ended up with 11% on election day, instead of the 7-8% in the polls before the election, while the FPÖ did more or less like the polls predicted.

Strache "demolishing" Faymann is relative. Strache is a huge loudmouth and constantly throws aggressive terms against his opponents, especially against the SPÖ "who are incompetent pro-immigration Communists".

BTW, who would you vote for here in September ?

I believe you are a British citizen only, right ? Not a dual citizen - so you actually can't vote here ?

Yes, British citizenship. Avoided dual nationality to avoid any hint of doing national service (!) and then once I was old enough to not have to worry about that Austria was an EU member and our wonderfully parochial Austrian civil servants wouldn't do dual at that point for EU nationals (or at least not without an incredible amount of fuss).

Who would I vote for? I guess politically I'm somewhere between ÖVP and Stronach. Make of that what you will.

[peterould]

Shame to that newspaper.

Why shame ?

Same here. What's the problem with publishing an interview?

[peterould]

Such hatred shouldn't be published.

This is not "hatred", just very ignorant of the fact that gay people are born gay already and no therapy should be used for them, nor is there any reason why there should be.

Born gay? Are they? Really? Would you care to point this statistician who specialises in research into sexual orientation causation and mutability to the academic paper that proves that contention?

In all seriousness, the argument over what causes sexual orientation is still an open book. We have evidence of some form of genetic component but also some really interesting indications of environmental factors too. And one thing we *cannot* say, and may never be able to say, is that we know that people are "born gay".

So let's drop it, because I can argue this all afternoon from both directions. As happy to take on those who argue people are born gay as I am people who say it's all just a developmental psychological issue.

[peterould]

Anyway I'd bang her on the altar of her church anytime.

*cough*

Move on...

[peterould]

Seriously guys, let's all move on.

Please.

[peterould]

Any polls tomorrow? Gallup?

[peterould]

Condoms and cigarette lighters. Tells you everything you need to know about Austrians...

I think the FPÖ freebie is the cleverest.

Latest predictions after this week's polls at http://www.peter-ould.net/austria-election-predict/

[peterould]

Apparently, Faymann and Spindelegger have agreed to skip the "elephant debate" ahead of the election, in which all 6 main party leaders debate against each other.

F+S only want to take part in one-on-one debates against the other party leaders.

Not getting ready for another Grand Coalition are they, oh no sirree...

[peterould]

Love it!

Are they only going up next to Faymann? I'm looking forward to the fuss when they place one next to Strache...