free range statistics

The polls did worse than usual in predicting the Australian federal election, but the pundits were far worse than the polls.

A few small explorations of idiosyncracies with polls, a couple of days before the Australian federal election.

I familiarise myself with the 2016 Australian Election Study, a wonderful source of individual level data on attitudes and behaviour relating to voting.

I ponder the limitations of animation, and how much work is needed to turn a fun animation into a great story.

svglite is a great way for producing good graphics for the web but there's a small hack needed if you want to use web fonts with it.

I've put up a page with my current Australian federal election forecasts for the House of Representatives, which I'll keep up to date until the election comes.

I improve my AFL predictions by adjusting Elo ratings for home team advantage (varying by home team) and with a more data-driven approach to parameters for the actual ratings.

I use Elo ratings from 12 months or from 120 years of AFL results to predict the results in the next round. Predictions based on just the past 12 months do better than those using the full history.

I explore the data on two-party-preferred voting swings in Australian federal elections and tentatively introduce the ozfedelect R package.

I tidy up Australian polling data back to 2007 and produce a statistical model of two-party-preferred vote for the coming election.