There are still a few doubtful seats in the lower house in New South Wales (William Bowe has all the detail here: but since there’s not really much practical difference between a 41-seat majority and a 49-seat majority (unless your seat is one of those in doubt), most of the remaining interest is focusing on the legislative council.

Even there, the key result is not in doubt: as I foreshadowed last Friday, the combined parties of the right will have a clear majority, with 23 and possibly 24 seats out of 42. But it’s that “possible” that’s interesting, since of course it would be independent and former One Nation leader Pauline Hanson.

Since the council’s voting system was reformed following the notorious “tablecloth” election of 1999, preferences have never played a role. If we ignore the effect of preferences this time, the logic of the count is very simple. Everyone with whole quotas is elected (that’s 10 Coalition, five Labor and two Greens), and you just arrange the rest in descending order of their remaining fractions of a quota; the top four should get up.

As of 11 o’clock this morning, with something like half a million votes still to count, the result of doing that was as follows (you can check on the latest figures here:

The gap between the first three and the others is sufficiently large to be sure that they will be elected; the contest is for the final position, with Hanson in the lead by .0135 of a quota, or about 2000 votes.

So can she win? The short answer is to sit quietly and wait for the votes to be counted, but there are good reasons to think she will probably fall short.

As Antony Green pointed out last night, although preferences have never mattered before, neither Labor nor Greens preferences have ever had to be counted before, and this time Labor “actively directed preferences to the Greens”.

I expect that even so the large majority of Labor voters will have just voted “1” above the line (probably not realising that there is no automatic preference flow), but the deficit the Greens need to make up is so small that only a tiny proportion would need to follow the recommendation for them to do it easily and win a third seat.

That’s assuming the Greens finish ahead of Labor: they are now about 7400 votes clear; that’s not much, but the Greens typically do better on absentee votes and also on the below-the-line votes, so you would have to expect it to hold up. (Indeed it’s quite possible they will overtake Hanson without preferences, although she also can expect to do well out of the below-the-lines.)

If Labor does manage to edge ahead of the Greens, then Greens preferences would be critical: they were not directed, but it’s hard to imagine that there wouldn’t still be enough of them flowing to Labor to close the very small gap concerned.

As it happens, Labor and Greens preferences are also the important thing in the most interesting of the doubtful lower house seats, Balmain. Liberal candidate James Falk leads the poll with 32.4% of the vote, but in a two-party count against Labor incumbent Verity Firth, Antony Green’s computer predicts he would lose narrowly on Greens preferences.

The real challenge for Firth is to stay ahead of the Greens and avoid elimination: failing that, her preferences would presumably elect the Greens’ Jamie Parker. She’s just 110 votes ahead; the Greens again can expect good things from absentee votes, but they also have preferences from five smaller and mostly anti-Green candidates to contend with.

My guess is they’ll just miss out, but Balmain will be worth watching for a few days yet.