Posts by bmk
Last ←Newer Page 1 2 3 4 5 Older→ First
-
I just tried running a simulation using the numbers from the latest bias adjusted tracking poll rather than the latest poll. This gave it as 45.9% chance of National and NZF deciding 54.1%.
I think unless something major happens between now and the election once again Winston will be deciding where the country goes - and I think the left will be disappointed if they expect him to support them. While I would want him to go with Labour; I really don't think he will. If it were just Labour then it would be a realistic possibility but he won't get along with the Greens and he doesn't want to be the third party in the coalition.
The only way I can see a left-leaning government at the moment (this is assuming no significant shift in support) would be with a Labour-NZF coalition that excludes the Greens but somehow they buy their support on supply and confidence. This would lead once again to the Greens being shafted by Labour (see both 2002 and 2005). And with some parties wouldn't be possible but with Labour saying 'If you don't agree to it then it's three more years of National and anything is better than nothing.' - they may cave and roll over. It's still a very unlikely outcome compared to a National-NZF coalition.
-
Thank you for producing and sharing the code David and Jonathan. I'm installing R now on my home laptop (already got it at work and am familiar with R although hardly fluent in it) so that I can play out some scenarios on it.
-
Hard News: Gower Speaks, in reply to Russell Brown,
Just to note that I'm very keen to host and promote anything that emerges from these exercises.
Maybe closer to the election there could be a weekly post with updated graphs as David says:
Well, we can probably get it making pretty graphs of the result for inclusion in posts about election things.
I know I'd find this really interesting and it could be a regular thing up to the election. Ideally it would be a stand-alone post with the latest graphs and then you could add your thoughts or someone else could add some analysis. But even if you just had the graphs it would be a great service and be sure to start some interesting conversations.
-
Hard News: Friday Music: Free and Legal, in reply to Howard Edwards,
Details here.
-
Hard News: Polls: news you can own, in reply to Pete George,
Assuming that National and Green variances last election will be repeated this election are risky.
This is true. And if it were only 2011 that followed this pattern then I wouldn't be so sure. But it's also based on 2008 where the same thing. Most polls had National 2-3% higher than their final result. While the Greens I believe have always polled higher than their actual election tally.
I also can't see people breaking to National unless National drop from their current point. I think there is a natural ceiling for parties and for National I think it's about 48%. In fact the polls showing over 50% support for National probably hurt them more than it helps them as some people won't like the idea of a one-party majority.
-
Hard News: Polls: news you can own, in reply to Pete George,
Roy Morgan include cellphones.
Exactly. I was aware of this but as the results you list above show there isn't really a significant difference. I think the non-polling cell phones isn't a big deal.
I think there is some flaw in the methodology that the pollsters use which means National and Greens seem to be typically overstated while NZF understated. But I don't believe it's a result of cellphones.
And still this difference is only around 3% I believe not as significant as some people like to imagine. I think the adjusted tracking poll talked about above is the best of the kind we have available in NZ. That's the aggregated poll I would predict would be most likely to call an election correctly.
-
Hard News: Polls: news you can own, in reply to ThoughtSpur,
A phone poll via land line is almost laughable.
This is only a factor if the people who don't own a land line vote differently from those who are in the same demographic and do. So say half the people between 18-30 don't have a land line if the other half vote the same way then it doesn't really matter. It just makes it harder for the polling company to get their pool of 18-30 year olds so they are more likely to weight the answers they do get - this can lead to higher variance from poll to poll. From what I can remember from (I think it was on 538?) in the US they found that polls that included mobiles weren't any more accurate than land line only polls.
I get tired with some people on the left saying "but polls only count land lines" as if that will magically provide them with victory. I remember this delusion at the last election too.
On your fundamental point I do agree though that pollsters will need to change and adapt their methods for the future; there will come a time when polling only land lines will be too unrepresentative.
-
Hard News: Polls: news you can own, in reply to BenWilson,
But these are actually questions that possibly could be statistically answered. I’m curious whether polling companies would attempt that, considering that they do get competitive advantage out of having better results. By better I mean “results that more accurately predict what customers really want to know – what the actual election results will be”.
Agreed. They already weight by other factors. You would think if they statistically observed that say there were consistently a 3% discrepancy between those who said they would vote National and those who actually did then you would weight for it.
In the same way as (in the US anyway) they weight for ethnicity etc and match it up for typical turnout figures.
-
What I find interesting is that the polls in the US and Australia when aggregated prove very accurate but in NZ this isn't the case at all.
One factor may be the lack of exit polls. Pollsters can use the demographic data obtained from exit polls to help weight their polling via demographics.
I'm not sure whether NZ polling companies do exit polls and don't publish the data for this case - or if this is legal. As a keen political observer I do wish exit polls were allowed here and I really don't understand why they aren't.
-
Up Front: The Kids are All Right, in reply to TracyMac,
Agreed totally - if you're going to have a democratic process you want as few barriers as possible. While I can see the attraction of the idea of a voting test when you start to consider the implications - you can easily envision some ugly scenarios.
Where do you draw the line etc. And people will say that's a slippery slope argument but slippery slopes often do occur. Often once people agree to the principle of something then it gets extended and extended - each step being a logical progression of the last.
First people who have dementia shouldn't be allowed to vote. Then people who have a sub-80 IQ, then people who are mentally unwell (which can then be defined as anyone who doesn't agree with the state). All people who have committed a crime etc. Then all people who don't pay taxes, then people who aren't net tax payers.
Sure I'm doing a reductio ad absurdum, but again while it's listed as a logical fallacy it can be handy to explore the conclusions of a policy.
I understand in the US last election some states were already using strict voter-registration laws as an attempt to dis-enfranchise the black population.