This morning I finished the final seat three seat profiles for the Victorian state election. You can access the entire guide, or view the seats in alphabetical order, by Legislative Council region, or by viewing the pendulum. Each Legislative Council regional profile includes links to the 11 Legislative Assembly seats in that region.
In about a quarter of seats I haven’t had a chance to update the candidate list to the final candidate order after the close of nominations, I am rushing to do those today.
Hey Ben, are there any better models than the classical “uniform swing” model avaliable for this election? Surely there must be a better way than assuming that the greens will get the same swing whether they were on 25%+ or 5%? Has anyone produced a model which takes these things into account?
Yes, there’s an alternative called “proportional swing”, where swings are bigger in areas where the existing vote is bigger. I built a spreadsheet for NSW using this system and someone else made a version for Victoria. It also allows you to alter preference flows to see the effect, but I haven’t inputted the latest primary vote figures and seen what effect the Liberal preference decision could have on it.
The proportional swing model seems to be dubious from what I can see. The results suggest that it’s weakly advantaged over a uniform swing at best.
Perhaps it would be sensible to try and work out a curve for the distrubtion of swing across seats based on previous results ( perhaps the last two elections in Victoria, and other states with a relevantly similar political system- i.e. single member districts)?
Any chance the spreadsheet file could be uploaded so people can download it and play around with it themselves?
My tips for Green marginals after the election, whether they win this time or next is another matter:
Grn v ALP:
Melbourne
Richmond
Brunswick
Northcote
Footscray
Preston
Albert Park
Grn v Lib
Prahran
Hawthorn
Kew
Comments are closed.