I’m going to progressively add a region-by-region post mortem to this thread as I go, so be sure to hit “refresh” every now and then if you’re hovering around the site and you think that might of use to you (UPDATE: Scratch that – I’ll do it tomorrow). I’ll start by answering a question that will be on the lips of many: how did the polls (and by implication the poll aggregators see the sidebar) get it so wrong? The answer to this is simple: by allocating preferences as they flowed at the 2012 election. In fact, the pollsters did well at predicting the primary vote. As noted two posts back, all three late-reporting pollsters essentially had the LNP on 42% and Labor on 37%. The present vote totals are 40.8% and 38.1%, and they will probably edge closer in the direction of the poll results during late counting. This tends to argue against the notion that there was a violent late surge to Labor and also the notion that there is likely to be a particularly big move back to the LNP on late counting.
The real key to the surprise is that Labor’s share of minor party and independent preferences went from 27% to 45%, the LNP’s went from 22% to 15%, and the exhaustion rate fell from 51% to 39%. This is based on slightly incomplete data, but it should be probably be near enough. Applying those preference flows to the current primary vote totals, it’s Labor and not the LNP that comes out 52-48 ahead, contrary to what all three of the aforementioned polls said. If the pollsters had been operating on accurate assumptions concerning preferences, they would have come out at 51-49 to Labor. Which gives them a lot to think about going forward, particularly with another optional preferential election on the way in New South Wales in a little under two months.
Anyhow, stay tuned.
UPDATE: Or I could leave it for tomorrow. Yes, I think that’s the go actually. Here’s a table that should set you a long way towards working out what I’ll end up saying. Seats are rated in doubt if Labor is ahead by less than 1.8% or the LNP is ahead by less than 1.2%, according to a crude 0.3% estimate of how much late counting generally favours the conservatives. The six in question are Mansfield (Labor leads by 0.1%), Mount Ommaney (LNP by 0.9%) and Redlands (LNP by 0.9%), in Southern Brisbane; Ferny Grove (Labor by 1.6%), in Northern Brisbane; and Whitsunday (LNP by 0.2%) and Maryborough (a special case, in that it will come down to who finishes second out of Labor and independent Chris Foley) on the Central Coast. Since Maryborough is in doubt between Labor and an independent, the ceiling for the LNP is 43, which includes the long shot of Ferny Grove.