Politics, elections and piffle plinking

Newspoll Quarterly – Election Simulation

With our new quarterly Newspoll data we can get a bit of an idea of what it would mean were an election held in the last 3 months that produced results the same as the Quarterly Newspoll. It uses a monte carlo simulation to run 10,000 hypothetical elections based on this Newspoll data where the State results are calibrated to the historical standard deviation of their swing at previous elections. Read the probability of the ALP (left axis) gaining at least X many seats (bottom axis) in both absolute terms and in terms of what a hypothetical new parliament would look like.

And the histogram for those interested

We wont spend a lot of time going over it as it will become completely redundant come the next Nielsen poll release as I’ve just about finished building the new simulation that aggregates both Newspoll quarterly and Nielsen monthly data, as well as accounts for margin of error issues in the polling. Explaining how it all works might be good for a Nerdy Sunday over the Christmas period.

6 Comments

  1. 1
    Posted December 13, 2008 at 12:50 pm | Permalink

    ...] Newspoll Quarterly – Election Simulation – Pollytics We wont spend a lot of time going over it as it will become completely redundant come the next Nielsen poll release as I’ve just about finished building the new simulation that aggregates both Newspoll quarterly and Nielsen monthly data … [...

  2. 2
    Ron
    Posted December 13, 2008 at 1:47 pm | Permalink

    Who but O’Ronly Factor to christian from Pendulims to Sims , and wondering ar there any Mr 38 aussie flavoured graphs to follow , and maybe past electons swings deviations data

  3. 3
    Aristotle
    Posted December 13, 2008 at 2:33 pm | Permalink

    “It uses a monte carlo simulation to run 10,000 hypothetical elections”

    Possum, perhaps to suit Australian circumstances more closely it should be
    re-named as the Iced-Vovo simulation!

  4. 4
    fmark
    Posted December 14, 2008 at 11:39 am | Permalink

    Bring on the nerdy Sunday :)

    As much as I like the notion of an Iced-Vovo simulation, perhaps a more accurate neologism would be a “Packer prediction,” given the etymology of the Monte Carlo method.

  5. 5
    Evil Gromit
    Posted December 14, 2008 at 11:16 pm | Permalink

    ALP Landslide, Next!

  6. 6
    David Richards
    Posted December 15, 2008 at 1:00 pm | Permalink

    Great work as usual Poss!

    Yes, Evil Gromit… kind of Sybil Fawlty bleeding obvious territory that the Libs have no chance, unless things go horribly wrong in a definitely Rudd blameworthy way, and the Libs suddenly become a united, coherent, viable alternative. So long as Australia stays in a better position than most of the comparable OECD countries, blaming Rudd simply doesn’t fly. Should other countries bounce back, and Australia flounders… that would be the only hope for the Libs. Not that I’m a card-carrying member of the Ruddophile Society, (far from it – I am grossly disappointed by him… maybe I expected more of a turn to port to change the course of SS Australia to avoid the Icebergs from Global Warming, Randism, Globalisation and the whole mess of the last 20 years.

Post a Comment

You must be logged in to post a comment.