As I expected in my final piece (and as I repeatedly warned could happen while people took a hung parliament for granted), Labor has won this election. However, like pretty much everyone else I failed to anticipate the very late softening in the Green vote compared to the many polls that showed them performing strongly over a long period, even only days out from the election. It is probably worthwhile doing a systematic study of Australian opinion polls to see whether they generally overestimate the Green vote, because it seems to be a common issue.

The single most likely result is 14-8-3. The last seat in Franklin is the closest and there is still some doubt about Kim Booth losing in Bass and Tim Morris surviving in Lyons. However Labor has 13 seats in the bag for sure.

A seat by seat overview of the state of play at 10:30pm on election night:

Braddon

The Greens have polled a surprisingly weak 10% of the vote, which barring a very equal distribution between the other parties, was never going to be enough. In any case the swing against Labor was remarkably small so even a strong showing for the Greens would not have won them this seat. All the sitting members will be returned easily, with notably strong results for Bryan Green and Jeremy Rockliff.
Result: 3-2-0. (called).

Denison

Very little has happened in this electorate compared to last time – just a swing of a few percent from Labor to the Liberals. Despite suspicions that David Bartlett could be outpolled by Graeme Sturges (some commentators even far-fetchedly projecting Sturges to have a quota in his own right), Bartlett has topped the Labor ticket. The sitting members Bartlett, Sturges, Hodgman and Putt will all be re-elected. Lisa Singh will be the new Labor member replacing the retiring Judy Jackson. Louise Sullivan at c. 1400 votes back is too far behind to beat Singh or Sturges off the back of c. 4300 votes from Joe Ritchie and Julie Collins that will have to be distributed while Bartlett is still in the count.
Result: 3-1-1 (called).

Lyons

Small swings from both Labor to the Greens have made this seat somewhat interesting with the Green vote dipping below a quota. However it surprises me that this is being suggested as a likely gain to the Liberals in some quarters. The Greens have a notional lead of 0.12 quotas over the Liberals. There is no possibility of tricks involving multiple candidates from the same party as Rene Hidding will be elected quickly and Heather Butler and Kerry Degrassi will be too far behind Morris for both of them to get ahead of him. The Tasmania First preferences will benefit the Liberals, but not hugely (perhaps shaving 0.02 quotas off the Greens’ lead). The Greens will benefit slightly from Labor’s surplus (worth perhaps 0.007 quotas). The Greens are far less susceptible to leakage than the Liberals in this seat with 2516 votes at risk of leaking compared to 4906 for the Liberals. All these things point towards a Green win and all sitting members being returned, although it is slightly risky to call a seat with an expected margin of not much more than 0.10 quotas given that there are still postal votes outstanding.
Result: 3-1-1 (Extremely likely)

Bass

This has been a very great surprise. The Green vote has been very strong in Bass in polling but has crashed, while Labor has, quite incredibly, picked up a small swing. The Liberals have a slight lead ahead of Labor, are less vulnerable to leakage (6159 votes exposed compared to 7509) and will benefit from Tasmania First Preferences, so it is unlikely the Liberals could miss out to Labor let alone to the Greens. To even get close to holding his seat Kim Booth needs a large preference flow from anti-pulp-mill independent Les Rochester, who has polled nearly 2%. The Greens are also the least susceptible to leakage of all (only 2363 votes exposed). However even in an optimistic simulation, assuming 70% of Rochester’s preferences flow to the Greens, factoring in leakage only results in about 0.92 quotas for the Greens to 2.98 for Labor and 2.02 for the Liberals. This is closer than Lyons and postal votes may be stronger for the Greens but the most likely result is Kim Booth losing to the ALP’s Steve Reissig.
Result: 3-2-0 (probable)

Franklin

As widely expected, this is the closest. The Liberals’ Vanessa Goodwin is notionally .055 of a quota ahead in the race with Paula Wriedt for the final seat. This lead is in all likelihood real, since by the time Paul Lennon’s surplus, Ross Butler’s and Daniel Hulme’s votes and the Greens’ preferences have been distributed, Lara Giddings should have reached a quota, preventing any trick results. The Greens’ preferences should benefit Labor to the tune of around .02 quotas, but Labor is also slightly more vulnerable to leakage (6787 votes exposed compared to 5795), potentially benefiting the Liberals by up to .01 quotas. So at present Labor is probably about .045 quotas (c.450 votes) behind. This is fairly unlikely to be bridged on postal votes and Goodwin is best placed at present but it will be several days before this seat is clearer. The fairly low Green vote has made things more difficult than expected for Labor here.
Result: 2-1-1 and 1 unclear, leaning towards the Liberals at this stage.

Overall

13 Labor, 6 Liberal and 2 Green seats are clearcut. Of the remaining four, it is most likely each party will win one and the final one is between the Liberals and Labor with the Liberals slightly ahead. This makes 14-8-3 marginally the most likely outcome with 15-7-3 the next most likely.

Kevin Bonham congratulates the fellow psephologist who took advantage of Centrebet’s folly and is now several hundred dollars richer, and will have comments on the causes and implications of these results in, he mutters ominously, the fullness of time.