Antony Green goes through the history of swings in NSW by-elections.
The Australian Parliamentary Library has released a research paper titled “The potential for renewable energy to provide baseload power in Australia” which is bound to cause some argument… although not here please! 
Mark Blumenthal of Pollster.com fame has a piece in NationalJournal.com about the “cell phone effect” that you might hear some of us talk about occasionally – and how it’s likely to underestimate the Democrat vote in US polling.
Nate Silver at 538 highlights the most interesting piece of US political polling done in the last 6 months, SurveyUSA polling early voters in five States. The results are profound considering the context (make sure you read the whole article about how early voter splits went in 200 and 2004) – go and read it!
On the right in the sidebar you might notice that the old Battleground 131 charts and data have disappeared; that’s because they’ve become irrelevant with the Obama surge. Obama now has 338 Electoral College Votes from States where he has over a 70% probability of winning on the Intrade markets, there’s 60 Electoral College votes worth of contestable States which sit between 30% and 70% probability of a Democrat victory and McCain has 140 EV’s worth of States where the Republican’s are greater than a 70% probability of victory.
There is no longer an Intrade battleground – just a fight over the size of the Democrat victory. Over at the US Election page here you can see how the State Intrade simulation numbers have jacked up to over 95% probability of a Democrat win (and have over the last week closely followed 538’s polling simulations for the first time).
15 Comments
Great post, valuable links. Thanks Poss.
More than welcome!
Thankee Mr P
Gotta go, Da Debate is on
Obama had a crushing debate win according to Insta polls, and the Dems’ win prob on Intrade is now up to 86%.
It’s also the first debate where Republican money didn’t launch a push to up McCains price.
Wasn’t the ‘cell phone effect’ tediously trotted out in our election campaign to explain the Libs poor polling?
Can’t remember that it was Ltep – although if some tool tried to trot that line out in defense of the Libs position last year it would have been laughable since the cell phone effect is mostly a function of age, and most young voters in Australia vote for the ALP and the Greens (in the same way that most young voters in the US are voting for the Dems)
Thanks for the Links poss. I’ll give the parliamentary research paper a close read.
You clever marsupial!
You and Nate were right.
Trader Drove Up Price of McCain ‘Stock’ in Online Market
http://www.cqpolitics.com/wmspage.cfm?parm1=5&docID=news-000002976265
Antony Green in his piece says:
“What is interesting on Saturday is that if the Liberal Party can win Ryde, it will be the first time the Liberal Party has taken a seat off the Labor Party in Sydney for two decades. Not since the election of the Greiner government in 1988 has the Liberal Party managed to dislodge a Labor MP from a seat the Labor Party held.”
Didn’t Labor lose Port Stephens to the Liberal Party at the last election?
I think Anthony Green is referring to seats within the Sydney area
Possum
Does the verification that Intrade has been targetted and manipulated for political purposes make you doubt it’s “integrity” as a data source for punditry or are the amounts too small to corrupt the data?
Doesn’t make me doubt it at all Dio. The money that the whole market has is always far far greater than the amount that any single player has, so when some group pushes up the price deliberately it only takes a few hours for any effect to be completely washed away by everyone elses money.
Trying to make narrative plays on Intrade is an exercise in impotence.
Though it does make it important for analysts to choose carefully the time they take snapshots of the numbers to analyse. Those deliberate price manipulations only last a few hours, so sometimes it’s a good idea to wait a couple of hours after they’ve washed out to get a true snapshot of Intrade prices and implied probabilities.
Poss
Any chance the same could happen in Oz? Our betting markets are much smaller than the US and a lot more susceptible to “tinkering”. A few well placed bets by either side with one of the two big sites might create a “win expectation” for about $100,000, which you might even get to collect on. It’s quite possibly a better use of hidden contributions than laundering the money for advertising.
It’s great to see you’ve confirmed Old Timey Man is DOOMED!