Five sleeps to go and the Psephos are going to town.
Things are looking good. Very very good.
For those of my readers who remain unconvinced and keep muttering about marginals and the like, feast your eyes on this...
Possum's Big Prediction (Anthony Green in disguise?)
So how good is [my] model using previous elections?
In 1998, the model predicted an ALP TPP of 50.82 whereas the actual result was 50.91
In 2001 the model predicted an ALP TPP of 49.15 whereas the actual result was 49.07.
In 2004 the model predicted an ALP TPP of 47.23 whereas the actual result was 47.20.
So what is the forecast for the election?
An ALP two party preferred result of 55.15%
And again.
The message is clear - the game is over.
That is what makes it so dangerous.
The polls are consistent, the fantasy of “Liberal strategists” being able to hide under the petticoat of fictitious marginal seat polling because “they’re closer than the national polls suggest” now looks like the façade it actually always was. The media have picked their winner, Uncle Rupert has moved behind Rudd in The Oz to match what has in reality been happening with his Tabloids and the Smage for weeks. This very morning Centrebet blew out to $4.60 - reflecting that even the punters are starting to get it, punters which haven’t got very much at all over the last 5 months.
There is risk and plenty of it over the next 7 days to E-Day, but it’s not downside risk for the ALP vote, it’s the risk of a collapse in Coalition support.
Simon Jackman - Professor of Political Science at Stanford University
We’re looking at almost a 7% swing, in 2PP terms, which would make it one of the biggest “swing” elections in Australian political history as well (Labor got a 7.1pp swing in 1969, but failed to win office; Labor suffered a 7.4pp swing against it in 1975; Howard won office with 5.1pp swing in 1996).
...
We see evidence of a trend away from Labor from September 1 onwards, reaching its peak in the first week of the campaign, at which point Labor has shedding about 0.05pp of 2PP vote share, or about a percentage point every 3 weeks. It looks like that trend is continuing, but the evidence for it is weaker later in the campaign. And in any event, 1 percentage point every 3 weeks, or 2 percentage points over the course of the campaign isn’t enough to bring the election back for the government.
Barring something amazing in this last week, or a marginal seats miracle, Labor will win, and win comfortably.
And finally, Geoff Lambert - Medical researcher and statistical gun. [pdf]
History shows the polls don’t lie. That was a title of an article I wrote for the Australian Financial Review in the lead-up to the 1996 election, in which I predicted a50-seat majority for John Howard. After this duly occurred, the AFR asked me to explain how I did it and this duly appeared as Maligned opinion polls got it right.
...
The perception that the polls lie arises from a selective reading of immediate pre-election polls at past elections. Although the vagaries of sampling error are routinely acknowledged, they are often forgotten when pundits look at the polls which emerge on the Thursday before the election.
...
In the 24 elections from 1946-2004, projections made from the aggregated polls produced an average overestimate in predictions for the ALP TPP of 0.6% (graph below). In the worst case (1987), the error in the TPP was 3.8%. In half the elections, the error was less than 1%. There are differences among pollsters (the “house effect”), but these tend to cancel one another out, so that the average of all pollsters has nearly always proved to be the best estimator.
...
...“elections are won and lost in the marginals”. In theory this again is true but, in practice the average swing in the marginal seats rarely differs from the nation-wide swing. ... Thus, if campaign effort has been concentrated in the marginal seats over the years, then the efforts of both sides must have cancelled at almost every election. Only 1998 bucks this trend.
...
The bottom line is that the weighted national swing is likely to be about 8.2% and the final national TPP about 55.5%- this would give a most likely number of seats for the ALP of 97. The different methods contributing to the weighted results give projected TPPs of 53.8% to 57.3%, and seat numbers ranging from 94 to 102, which is not symmetric about 97 seats, mainly because the clustering of seats on this part of the pendulum is not homogenous. In this region of the pendulum, every 1% swing can produce a 14-seat majority change.
Monte Carlo simulations based on the weighted data show that about 19 out of 20 elections conducted under these conditions would produce a TPP of between 54.7% and 56.3% (55.5% is in the middle of this range) and a 95% confidence limit for the number of ALP seats of 87-104.
...
It would take a political sensation of biblical proportions for the Coalition to win from here or, as Antony Green has said, the greatest come-from-behind victory in history. The chance of the Coalition pulling the ALP TPP vote back to Tiger Territory in the region of 50.6% in the absence of such a sensation, or in the absence of the statistics going pear-shaped, is ridiculously small.
Come on Psephos! Don't let me down. I've come to love you all over the last few months and if you're wrong it will be utterly devastating. This isn't just the election you're predicting here - it's our relationship!!!!
No comments:
Post a Comment