Odds 150

This entry was posted in Australian Politics. Bookmark the permalink.

19 Responses to Odds 150

  1. derrida derider says:

    Interesting. So if all the favourites are returned the Coalition will have a 4 seat majority.

  2. a stewart says:

    Re your election forecast for Wide Bay. I supose every seat has special issues that generally get submerged in the federal campaign tsunamai….but Wide Bay has some special issues. 1/ thr fedefraland stater, Labor and Cosalition stance in the seat’s biggest isssue, the Traveston Crosing Dam. If Howard opposes the Dam and Rudd decideds not to trash his old fioe Beattie, then the coalition willpick up a few percent – and preferences from indepondents and even Greens. 2/ The stea’s boundaries now extend from the traditional farming (kingaroy, Gympie) and retirement (hervey Bay) areas pro Howard with Labour’s Maryborough strionghold into the Noosa area where there are lots of Libs who hate voting National, and lots of champagne socialists who’ll lift Labour not just in votes but some active campaign dollars if they can be engaged (notas yet since ALP running the campaign bizarrely out of Maryborough).

    I’d put this as a sleeper if Kevin Rudd wants a long shot from his old boyhood area and where he holidays if Labourt oputs in some real effor
    t. The coalition is falling apart and Truss very worried. He’s actually beenseen in the electorate and trying toengage the local media. But evenlocal Nationals say he is not the man for this newly configured electorate.

  3. JJ says:

    Not sure about this… the betting agency has shorter odds for ALP winning Deakin than La Trobe yet you have La Trobe a higher probability for ALP to win???

  4. Luke says:

    And of the nine close seats (3% betting difference between the major parties), 6 are for Labor and 3 for Lib/Nat. Note that a couple of seats are “too close to call”. So it would appear even more favourable for the Howard government.
    Andrew – it would be an interesting academic exercise to see how these betting markets change over time – perhaps do a snapshot like this once per week and check the movement. Does the betting market match the telephone polls?

  5. Pingback: Club Troppo » Missing Link, Friday 6 July

  6. Andrew Leigh says:

    DD, you’re right. Their opening media release said “On current prices, Portlandbet forecasts the Coalition to return to government with a reduced majority, winning 77 seats to 71 for the Australia Labor Party. Independents should win 2 seats.” Potentially there’s an arbitrage opportunity here, since some bookies have Labor as favourites to win government. But you’ve gotta work out which seats are ‘wrong’ in the Portlandbet market.

    JJ, Deakin’s odds have changed over the past 24 hours. The ALP was $3.65 there yesterday – now it’s $2.40.

    Luke, there are definitely things that sensible academics could do with these data. I’m hoping that Portlandbet will make them publicly available to anyone who wants to use them.

  7. Nick says:

    Surprise result!

    I’ve crunched the numbers….

    * The bookies say that the Libs are clearly ahead (greater than 60% chance to win) in 69 seats.
    * The ALP is clearly ahead (greater than 60% chance) in only 54 seats.
    * There are 26 marginal seats, according to the bookies (with either party a 40-60% chance of a win)
    * Two independents are slated to win (New England, NSW and Kennedy, Qld) to total 150 HR seats.

    To win the election, the ALP will have to win 21 of the 26 seats that the bookies say are on a knife-edge – and will need to vastly improve their vote in Qld and SA.

    No wonder Family First preferences are so important to the ALP, who are strongest in those states. Family Fist polled 7% in contested Qld seats at the last election, and poll about 6% in SA elections. Family First voters follow their How to Vote Cards like zombies so imagine 6-7% going straight off the Lib vote and onto the ALP vote. F1 preferences could be the thing that swings the election, believe it or not.

  8. Nick,

    I’m not sure how you arrived at the conclusion that “Family First voters follow their how-to-vote cards like zombies”.

    That doesn’t appear to have been the case at the 2006 Victorian election.

    Have a look at the 2006 preference distributions for Ferntree Gully and Macedon, where Family First preferenced Liberal over Labor:


    Unfortunately the VEC don’t distribute preferences for all seats. But a cursory glance suggests that at best 65% of Victoria’s Family First voters followed their HTVs.

    Note also the significant preference flow from Family First to the Greens. The Greens were not only last on the Family First HTVs but the subject of repeated Family First attacks during the campaign, eg:



  9. kiwipundit says:

    I notice that Portlandbet seems to be the only agency with odds narrowly favouring the Coalition winning the Federal Election, while the other betting agencies running odds on the Federal Election narrowly favour the ALP winning.

    It’ll be interesting to see if the other betting agencies will soon be offering odds on all the 150 seats or at least all the marginal seats.

    It’ll also be interesting, for comparisons sake, if the polling companies release any polling done in the key marginal electorates. Will these back up Portlandbet’s odds in the marginal seats?

    The ALP’s healthy opinion poll leads compared to it’s modest odds in betting markets may well reflect that the ALPs growth in support may be occurring not so much in the Coalition marginals but in ALP seats.

  10. kiwipundit says:

    Ooops.. by “it’s modest odds in betting markets” I mean the betting markets compared to the opinion polls forecast a much more modest electoral performance by the ALP.

  11. Stephen L says:

    There is also something pretty odd about some of the long shot seats. I realise that bookies margin means you can’t make a killing betting on near certainties but seriously – Labor at 90% in Batman. There’s an outside chance some independent could emerge, but giving the Libs a 7% chance of picking up a seat where they need a 25% swing when the current climate is so obviously for some sort of national swing against them?

    It is more likely the ALP would lose Batman to the Greens than the Libs, and even in my most optimisitic moment I don’t see that happening for quite a few elections.

  12. Stephen,

    As you note, the ‘Labor 90% for Batman’ figure is really just an artifact of the bookmaker’s unusually wide margin.

    If you look at the actual money on offer, they’re offering $1.00 for Labor for Batman — ie, nothing, they think it’s a certainty that Labor will win.


  13. peter Tucker says:

    The Portlandbet’s markets, surely, over-state “other”. No way, in most seats, would “other” win three elections out of 100.

  14. Stephen L says:


    Ok, I didn’t check the actual odds, just looked at what Andrew had quoted and assumed a normal margin. I guess the question then is how is that they’re able to get away with a wider margin than usual – is it because they think no one else will undercut them and those desperate to bet on a seat by seat basis will put up with such margins.

  15. Pingback: Prediction Markets and the Australian election « Knowledge Futures

  16. ray says:

    Louise Markus the Member for (Hillsong) Greenway was initially placed on the market a $2.05. Clearly the bookies had based their odds on the margin as of the last election, a mere 0.58%.
    They should have researched this more thoroughly. The redistribution in NSW caused her margin to “miraculously” become a very safe seat on an 11% margin.
    So as soon as the book was opened there was a flood of money her way to have her now firm as favourite on $1.40.
    It looks as though the entrepreneurs at Hillsong have found another way to make money.

  17. Nick says:

    Hi Nathan,

    re: your comments – I beg to differ. I can’t explain your stats for Ferntree or Macedon in the Vic election, if they say what you say they say (Greens asked 50 or so supporters to vote Family First and then ignore the HTV? – wouldn’t put it past them). But to quote from a Crikey article recently:

    Lisa Crago, not a member of any political party, writes: Re. Andrew Lewis’s comments on preference deals (yesterday, comments). You are right Andrew, the voters do hold the power to preference (or not as the case now is in NSW). Does anyone honestly believe that voting tickets printed by parties beforehand actually affect preference distribution? Is there any
    science behind this myth, any studies to support the argument? Yes,
    political scientists salivate over studies regarding voting behaviour as
    surveys are taken after every federal election by the ANU. However, as
    Andrew has rightly asked, are there any studies to support this argument? A
    study of 16 years of electoral survey data overwhelmingly shows that
    Australian Green voters DO NOT follow the how to vote cards for the House of
    Reps 76% of the time. Even more interesting is that 79% of those same voters
    preferenced the ALP anyway. In contrast, Family First Party voters follow
    their HTV cards 46% of the time. The only place this differs is in the Upper
    house where most would rather place a “1” in the party box than fill out the
    often gigantic number of boxes. Political parties like the Greens like to
    ignore these studies and pressure the ALP on policy; and the ALP falls for
    it every time. Does anyone remember the 2004 ALP forestry policy?

  18. Pingback: simon jackman’s blog » daily updates of seat-by-seat betting markets

  19. burton wu says:

    based on the odds, it seems that 2007 is going to be a 50-50 election rather than what the latest news poll is suggesting – at this stage, anyway. wondering what that means if the final result is 74 vs. 74 (with 2 ind) or 75 vs. 73 (with 2 ind)? who will be the prime minister / government since there is no clear majority?

Comments are closed.