I’m afraid in depth analysis of the result will have to wait until I’ve slept for just about the first time in 48 hours. I’ll just observe that that BludgerTrack thing on the sidebar isn’t looking too flash right now, to which the best defence I can offer is that aggregators gonna aggregate. Basically every poll at the end of the campaign showed Labor with a lead of 51.5-48.5, and so therefore did BludgerTrack – whereas it looks like the final result will end up being more like the other way around. The much maligned seat polling actually wound up looking better than the national ones, though it was all too tempting at the time to relate their pecularities to a past record of leaning in favour of the Coalition. However, even the seat polls likely overstated Labor’s position, though the number crunching required to measure how much by will have to wait for later.
Probably the sharpest piece of polling analysis to emerge before the event was provided by Mark the Ballot, who offered a prescient look at the all too obvious fact that the polling industry was guilty of herding – and, in this case, it was herding to the wrong place. In this the result carries echoes of the 2015 election in Britain, when polling spoke in one voice of an even money bet between the Conservatives and Labour, when the latter’s vote share on the day proved to be fully 6% higher. This resulted in a period of soul-searching in the British polling industry that will hopefully be reflected in Australia, where pollsters are far too secretive about their methods and provide none of the breakdowns and weighting information that are standard for the more respected pollsters internationally. More on that at a later time.