Sunday, November 26, 2017

2017 Queensland Election Postcount (Main Thread)

Labor has won the election

Final result ALP 48 LNP 39 KAP 3 ONP 1 IND 1 GREEN 1 

Pumicestone won by LNP
Cook assumed won by ALP
Rockhampton won by ALP
Gaven won by ALP
Maiwar won by Greens
Hinchinbrook won by KAP
Macalister won by ALP
Burdekin won by LNP
Townsville won by Labor

This thread will follow the 2017 Queensland election post-election-day count in those seats that remain up in the air.

I'm in no real doubt now that Labor has won the election, in some form or other, but that form (majority or not, and if not with what numbers) remains to be clear.  They have 44 wins that I am regarding with some trepidation as solid, and one more ( Gaven) that I think they have probably won.  There are two more that appear to be going to "the left" in some form (Maiwar and Rockhampton), a close one in Townsville, and it is also worth noting that the indie who appears to be winning Noosa preferenced Labor. A vast amount has to go wrong from there before an LNP-ONP-KAP-indie government could be seriously considered, and even then it's not clear which way KAP would go.

This thread will follow the seats that I consider to be in significant doubt.  Where time permits and a seat greatly interests me, it may be given a breakout thread.  Seats appear in alphabetical order but when a seat is no longer considered of interest it will be moved to the bottom of the thread.  It will take me some time today to add all the seats.  Updates on specific seats will be added as time allows, but because of work commitments this is only likely to happen in the evenings, and probably not all of them.



I was including Mundingburra but I don't think it's realistic for the One Nation candidate to close 9.2% on the LNP candidate off 14.1% of Katter votes, so I'm assuming it's a Labor retain.  I was also going to include Thuringowa but we have a real 2PP count showing Labor beating One Nation in that seat and I don't see the LNP staying in second once the Katter preferences hit the One Nation pile. The ALP candidate has claimed victory in that seat.

I have not yet included Aspley (in which Labor leads by 301, 50.6-49.4) or Bonney (LNP leads by 275, 50.8-49.2) because these seats projected strongly to the leading party according to the ABC on election night.  In Bonney the count is slow with only 70% counted and some primaries yet to go to 2PP, while in Aspley the postal count is very advanced, perhaps accounting for the close margin.

The 2PP and polling

It's not entirely relevant to this article but some quick comments about the 2PP and polling are in order.

The 2PP is very difficult to estimate from current figures because of the number of non-classic and potentially non-classic seats.  We won't see a final figure for a very long time, and that may just be an estimated figure anyway.  Based on those seats being counted as Labor vs LNP I am estimating the 2PP swing at 1.3 points at present, for a 2PP of 52.4 to Labor.  However, the non-classic seats fall heavily in the regions where Labor's form was spottier, and once they are added I expect it to fall to 52.0 or maybe into the mid to high 51s.  Whatever, it is looking like the Newspoll/Galaxy 2PP picture was accurate all along, while ReachTEL only got close in the last two polls having had the LNP as 2PP winners up til then.  Even then, the second-last ReachTEL had the LNP primary much too low, and probably would have overshot the 2PP but for an excessive flow of respondent preferences.  It is looking like whatever Newspoll/Galaxy were doing with preferences (the exact basis for which remains unclear) was either accurate or at least nearly so.

On current primaries, the final Newspoll (ALP 36 LNP 34 ONP 13 Green 10 Other 7) was another stunner, with a maximum error of 0.7 points on the One Nation vote (and even that has been falling) and an average error around 0.3 points.  The final late ReachTEL is running at about 1.1 points average, which is far from bad, but is currently trailing the one-month Essential sample which is running at about 0.6 points average error (but had the 2PP for Labor a point too high it seems.)  I will have a full assessment when the results are final.

It was very strange that Sky chose to commission a final ReachTEL that didn't see release until 5 pm on Saturday, but apparently wasn't an exit poll.  Seriously Sky, what on earth's the point?  Just putting polls and those commissioning them on notice that in future I will ignore all polls not released by the opening of the polls on polling day, in considering the accuracy of a pollster's regular polling.

The state of the count

Some cautions about the state of the current count.  In 21 of the 93 seats, the ECQ does not have a current two-party preference count, because they appear not to have picked the correct candidate pair (that's not their fault).  The ABC currently has two-candidate figures for all of these except Maiwar, but the ABC's figures in these cases are not real numbers; they are estimates.  Also, in some cases the ABC may also turn out to have the wrong pair of final candidates.

I have dredged through all of these incorrect-2CP seats in search of any possible surprises (except those already on the radar) but not found anything convincing.  Often there would be a path to victory for some independent from third or fourth position if they could pull off a 75-80% preference flow, but for an inconvenient how-to-vote card allocation from some party.  Also, my experience of such cases is that postcount votes (especially postals and prepolls) have a history of snuffing them out anyway.   But it's the sort of election where there's a higher than normal chance that in some seat, somewhere, some candidate who no-one has any idea is in the hunt might pull off a freak win.

There have also been cases of seats appearing as close in the two candidate count and then blowing out again.  These have usually been caused by the prepoll two-candidate count not having been completed on the night, which is why the ABC projected them as wins.  Also beware: some parts of Queensland have high rates of postal voting which show some very distinct characteristics (eg the traditional 1.5-2 point comeback in the federal seat of Flynn).

General update Tuesday 5 pm: Currently many of the two-candidate counts are offline on the ECQ site.

------------------------Seats assumed won and no longer being followed --------------------------


COOK (Labor vs ONP and KAP)

Cynthia Liu has a massive primary lead with 39.3% and the non-Labor vote is split three ways between the LNP, ONP and KAP.  ONP's Jen Sackley and KAP's Gordon Rasmussen have cross-preferenced, making it almost certain that whichever one makes the top three will eliminate the LNP.  On current votes, Sackley leads Rasmussen by 297 (unlikely to be overturned), and the only candidate below them is the Greens whose votes will mainly go to Labor.  So it seems very likely that Sackley is in the final two and she needs 74.4% of all preferences ahead of Liu to win it.  However the 6.48% of Greens preferences will favour Labor strongly over Sackley leaving Sackley needing something like 84% of KAP and Liberal preferences to take the seat.  We don't have any data on such flows but that seems a very big order so I expect Liu to win.

Monday night: The ECQ now has a count up of Liu vs Sackley which currently shows Liu with a very large lead.  Rasmussen however is now only 97 votes behind Johnson but the LNP card (and the donkey vote) favours the One Nation candidate over the KAP one so even if Rasmussen gets into third it doesn't seem that he will overtake One Nation.

The complication in the current numbers is that there are widely uneven booth results in this seat with Liu taking 2CPs in the 80s and sometimes 90s in Indigenous communities but Sackley doing very well in certain towns and prepolls.  At this stage only about 60% of the primary votes have been transformed to the LNP vs ONP count and there are some nasty booths to come for Labor including the Mossman prepoll.  I'm pretty close to calling this but let's see how far ahead Liu is of One Nation when we have a fuller 2CP count.

Monday 9pm: We now have a fuller count with Liu up 51.5 vs Sackley and the nasty prepoll booths gone; moreover Liu has a 1500 vote boost coming her way when Mobile Teams 1 and 2 hit the 2CP, nearly quadrupling her lead.  Even supposing Rasmussen can get into third I cannot see that he will get a preference flow that much stronger than Sackley got so I am not seeing significant doubt now that Liu has won.  I may continue updates for this seat just because it's so interesting!

PUMICESTONE (LNP vs Labor)

This seat was senselessly ignored in most campaign commentary as a possible LNP pickup despite being vacated by a troubled Labor MP.  It appears to have fallen with the LNP's Simone Wilson currently on 51.05% 2PP, a lead of 499 votes.  That does include some postals so Labor might come back (as originally projected) but probably not by enough.  As with Gaven, in the absence of dramatic progress this one will be put away pretty soon.

Monday night: The gap is now out to 561 so barring anything extraordinary I'm assuming this seat won by the LNP.

ROCKHAMPTON (ALP vs IND)

In Rockhampton, normally a very safe Labor seat, Labor's Barry O'Rourke (32%) leads local mayor Margaret Strelow (IND, 23.7%).  Strelow unsuccessfully sought Labor preselection.  If Strelow stays second she should win.  The problem is that One Nation's Wade Rothery is on 21.2% and the LNP's Douglas Rodgers has 17.7%.  The Greens with 5.4% have preferenced O'Rourke but there's bound to be some flow to Strelow.  If the LNP ticket flows strongly enough to Rothery this will put him over Strelow, put her out of business and deliver the seat to Labor.  On current numbers, that seems quite plausible to me.  I've also seen some suggestion that if One Nation can get into second they might win, but it would take a near-100% preference flow to get them equal with Labor assuming the Greens preferences pool strongly with Labor (including via Strelow).  Strelow did not distribute preferences herself.  I don't see realistic chances for One Nation here.

Monday night: No significant changes and we probably won't know the outcome here until the preference flow.

Wednesday: According to Strelow, the scenario of her being jumped by One Nation (who then lose) will come to pass as the LNP to One Nation flow is strong.  It seems that a provisional preference distribution has been conducted.  Strelow has conceded.

GAVEN (Labor vs LNP)

Labor's Meaghan Scanlon featured prominently in the party's campaign launch and now whether or not she unseats the LNP's Sid Cramp could be crucial to whether her party wins a majority.  Scanlon currently leads by 462 (51.1% 2PP) which William Bowe projects to close but not by much.  Given that many of the postals are already counted to 2PP and that Scanlon emerged unscathed from a nasty-looking lot of prepoll 2PPs today, this looks very difficult for Cramp to pull back and if there is no progress I will call it pretty quickly.'

Monday night: No change on the ECQ website though the ABC has slightly more advanced figures that may be a projection.

Tuesday afternoon: The ABC has Scanlon 382 votes ahead (50.8% 2PP).

Wednesday: The ABC has Scanlon 353 votes ahead.  This may be a projection.

Friday: With the ABC now projecting Scanlon more than 400 ahead I'm treating this as an ALP gain.

MAIWAR (ALP vs GRN)

I had tickets on Maiwar to possibly pull off a repeat of the Prahran situation in Victoria and if anything it seems to have overkilled its mission.  Scott Emerson's primary of 42.2% is a little bit shy of enough and there is a lineball fight between Labor's Ali-Breeze King and the Greens' Michael Berkman for second.  However it ends, Berkman's 27.6% primary vote is huge.  At the moment King leads for second by 27 votes but there is also the matter of 678 votes from Anita Diamond, candidate for yet another of these direct democracy parties that your host likes to mutter darkly about.  I've seen a claim that her preferences are favouring the Greens.  Labor have been gaining through the postcount so far and would probably want to pad that lead to something like 300 to keep the Diamond factor under control, unless they have better scrutineering information.  Absent votes, which tend to favour the Greens, have not been counted yet so that might not be easy.

The other thing we don't know for sure is how strongly the Labor preferences flow to Berkman if Berkman is second, but experience of similar contests elsewhere suggests they will be as strong as the other way round if not stronger, and therefore the 74% preference flow Berkman will need is likely to be a doddle.  I think King is also the kind of Labor candidate likely to generate a good flow to Berkman so whichever of King and Berkman finishes second after preferences appears to have it in the bag.

On present numbers, Emerson needs 25.7% of preferences to survive.  As noted in comments, in Everton Tim Mander has a similar figure, but the Green vote there is lower.  The lead for Emerson over King on the night was based on real numbers up to a point so I believe it will now hold, but it is not absolutely certain that Emerson is gone.

Monday night: Addition of some prepolls has Berkman back in front in the race for second by 19 votes on primaries (noting the possible Diamond bonus as mentioned above).  Meanwhile we have a new 2PP count between Emerson and King (51.6% to King) that makes it clear Emerson is gone if King is his opponent and doesn't look at all promising for his chances against Berkman either.

Tuesday afternoon: Berkman leads King by 37 votes after today's postals broke fairly evenly between the two.  I have had word from an experienced scrutineer that (i) Berkman gains on the Diamond preferences by what sounds like at least .25 votes/preference (which should amount to a gain of at least 200) (ii) Berkman is smashing the 2CP vs Emerson, with the flow from Labor considerably stronger than the flow to them.  Although it is a tight contest, with absents still to come it seems that Berkman holds all the cards and should win unless there have been significant counting errors.  Emerson's vote has probably been trashed by the combined forces of the redistribution, Adani and inner-city anti-One-Nation sentiment.

Wednesday: Berkman now leads by 43 votes.  A large number of absents will be counted tomorrow.

Thursday: Hearing that an indicative throw of the Diamond preferences has the Greens extending their break over Labor by 150+ votes, meaning that their currently precarious 16-vote lead is actually about ten times that.  We now await the absents which may well be decisive.

Thursday night: Around a thousand absents benefited Labor by three votes, which means it might not be quite over yet (especially if there might be counting errors) but their task looks very difficult.  They're 12 behind, but I make it that that's really about 174 after preferences.

Monday 4th: The ABC now shows Berkman's primary vote lead up to 51 and now shows 86.5% counted, so I expect there is at most a few percent to go.  With Berkman now effectively over 200 ahead unless there are serious counting errors this looks over.

Later: Preference throw on Tuesday according to the Courier-Mail. (Tuesday: As postal votes closed at 6 pm today, it is more likely this will unfold over coming days.)

Wednesday: Berkman's lead having moved to 71 votes, which is presumably final, I have decided the doubt level here is no longer sufficient to avoid treating this as an assumed Green win.

BURDEKIN (Labor vs LNP)

I had pretty much written this off on the night on the basis of a large lead for the LNP on the ABC coverage (assuming One Nation were third, which they are) but after seeing the comment by Peterjk23 below I think Labor now have prospects here.  One Nation are preferencing Labor who are now in the primary vote lead.  I am unsure if the large projected LNP lead on the night was based off real numbers or not.  The cause of the turnaround is mining town prepolls which was a known issue on the night but may not have been factored-in sufficiently.

Monday night: We now have a real 2PP vote count in Burdekin which has the LNP with a trivial lead.  However the 2PP count does not include the so-far counted postals based on which the LNP will surge.  The ABC site has Dale Last (LNP) with a 376 vote lead (50.7%) which would seem difficult to close.  I am unsure if the extra votes on the ABC are real or projected but it seems about right.

Tuesday: The ABC now shows the LNP with only a 161 vote lead, presumably a projection based on votes not yet counted to 2PP.

Wednesday: The ABC now shows Labor with a 96 vote lead, but these are probably not real numbers, and the ECQ is not currently showing the 2PP count.  William Bowe has noted the LNP's predictably strong performance on postals so far.

Wednesday night: ABC now projects/shows/who knows Dale Last with a 347-vote lead.

Friday: I'm only keeping this one on watch because of the spectacular lack of reliable information.

Wednesday: LNP won 50.8-49.2.

MACALISTER (ALP vs IND)

In Macalister, Labor has clearly beaten the LNP if they are second, and the only threat is independent Hetty Johnston, the Bravehearts director and anti-child-abuse campaigner who currently lies third.  The gap between the LNP and Johnston is currently 2.3 points, which Johnston would need to overturn on the preferences of the Greens (6.4%) and three minor candidates (6.3% combined).  This would appear to be unrealistic, except that the Greens have rather staggeringly preferenced Johnston, so it could happen - bearing in mind that the LNP will get almost no Greens preferences.  (Greens voters are not generally big how to vote card followers.)  One of the three others has also preferenced Johnston.  If Johnston makes second she needs 66.5% of all preferences, which also seems realistic.  Most observers are assuming this one is safe for Labor but I am not, although further primary vote drift to the LNP - common in postcounts involving independents - could settle the issue.

Update Monday morning: The gap is stretching - now 2.6 points.

Monday night: Gap now 2.7 points.  Opinions vary greatly on what size of gap Johnston might close here.  I think because of the random nature of some votes for very minor candidates, anything over five points should be too much. Could she close down 4.5?  Without scrutineering data I can't say.

Tuesday night: Johnston has done hopelessly on the nearly 300 absents counted so far, getting only 8% of them to 23% for the LNP.  Patterns for absents can vary depending on where they are from and so projecting from one lot of absents to the rest is risky, but perhaps Johnston's support base is especially unlikely to be mobile.  In any case, the gap for second widens to 2.83%.

Wednesday: Slowly but surely the gap continues stretching, now out to 2.98%.  There are still a lot of absent votes to go so I expect this to continue.

Wednesday night: A lot more absents counted (nearly 1400 in all) and the flow of them has moderated somewhat but the bulk of them has pushed that critical gap out to 3.33%. I wonder if we will get an indicative preference throw here soon.

Thursday: There is a confusing Courier-Mail report in which Johnston says that "They did a full recount because on the night of the election they did LNP and ALP and chose ALP to win, but they did that same process again yesterday for me and they had me 2277 in front at that point." If that very large margin is valid, that would seem to suggest that Johnston wins the two-candidate-preferred against Labor if she gets there, with around a 77% preference flow to Johnston from all sources.  This isn't impossible by any means, but what the report doesn't settle is the question of whether or not Johnston actually moves into second.  It sounds like Labor are confident this won't occur.

Friday: LNP-Johnston gap now 3.53%.  It's worth noting the LNP did surprisingly well on the preferences of the minor independent in Maiwar.

Friday again: The consensus including the combatants now appears to be that Labor will almost certainly win, thereby securing a majority.  The information basis for this still remains unclear!

Monday: The Courier-Mail reports a preference throw is coming on Tuesday though Labor is claiming the seat.  (Tuesday: As postal votes closed at 6 pm today, it is more likely this will unfold over coming days.)

Wednesday: I am seeing that Johnston conceded yesterday so am assuming the parties know something I don't.

The ECQ site now shows Labor has won the seat.  The margin for second was 483 votes according to Johnston on Facebook (see comments).

HINCHINBROOK (KAP vs LNP)

Currently Andrew Cripps (LNP) has 30%, Margaret Bell (ONP) 21.9%, Nick Dametto (KAP) 21.2%, Paul Jacobs (Labor) 19.1%. An independent who did not register a how to vote card, Peter Raffles, has 4.8% and the Greens have 3%.

Green preferences tend to flow weakly where their vote is low, so on current numbers Jacobs currently shouldn't quite pass Dametto on the Green votes, but the Raffles votes are an unknown quantity.  If Dametto does stay third he should jump One Nation on Labor preferences, then needs only 59% of all preferences to beat the LNP and should get that easily.

Monday night: No significant changes.

Tuesday night: Slight changes in the primary totals today but nothing significant.

Monday 4th: The Courier-Mail reports that this seat has firmed up for KAP following some form of preferences, but there is no further detail.

Wednesday: Called based on evidence from a KAP scrutineer that Labor does not pass them and they have very strong flows from ALP and One Nation.

TOWNSVILLE (ALP vs LNP)

This one seemed to be dusted on election night but a strong performance on prepolls and postals has seen the LNP challenger Kasie Scott storm back to a 61 vote lead.  The fact that that includes some postals is likely to keep Labor competitive.  I may have a look at a projection attempt for this seat soon but any projections will be pretty rubbery.

Monday night: Still 61 votes.

Monday midnight: William Bowe projects the LNP to win this seat by 290 votes based on past trends.  That doesn't mean they will necessarily win, as past trends might not apply for some reason, but it is enough to declare them the favourite for this postcount.

Tuesday afternoon: The ECQ 2PP is offline.  ABC has the 2PP at a very slim lead to Labor, 12,650 to 12,646.  I do not know if these are real votes or a projection but the primary vote tallies do match.

Wednesday afternoon: The ABC now projects a 146 vote lead for the LNP.

Friday night: The ABC projection now has this on the wire and in the absence of clearer information I will put it there too.  Labor's postcount has been better than expected.

Monday 4th: Projections have this incredibly close and it may well go to a recount.

Wednesday: Have seen some whispers that Labor may be winning by 100 or so but nothing firm.

Thursday: Labor won by just over 200; this should be easily enough to avoid a recount.

MUNDINGBURRA (ALP vs PHON)

A late addition to the in-doubt pile (simply because of inadequate information) as The Australian has reported internal concerns that the ALP claimed this seat prematurely.  The ABC has the margin down to 51.4% Labor vs LNP, which is decisive as far as that contest goes, compared to about 54% on the night (both were real preference counts).  The apparent concern involves One Nation potentially succeeding in closing the gap on Katter preferences (which actually is not so unrealistic given the very strong flows seen in some other seats) and then winning on LNP preferences.

Friday: Labor has won. I don't have any further detail yet.

27 comments:

  1. Yo Kev, when's your next update on the QLD results matey.
    Cheers

    ReplyDelete
    Replies
    1. Went up about the same time you hit send.

      Delete
  2. Of your Unclear grouping, I favour the Greens in Maiwar, the ALP in Rockhampton and the LNP in Townsville.

    So the ALP would need Macalister for a majority. I suspect they'll get it, with the LNP to hang onto second on the back of a boost from the postals and Greens voters doing their own thing as normal.

    ReplyDelete
  3. Aaron White who got Bolton's number 2 preference is a former Green candidate who split to run further to the left (or possibly because he didn't get preselected for Noosa). And Greens 3rd. Then Labor. I think that's a fairly safe check in the favors Labor box.

    ReplyDelete
  4. Always the way
    Cheers again mate ;)

    ReplyDelete
  5. Maiwar. In Everton (3-way contest), Greens preferences seem to be leaking 27% to Tim Mander. If this is repeated in Maiwar AND Libereals keep improving on the post-count, they could still be in with a chance.

    Rockhampton. Even if green preferences flow 66% to Strelow, it would still require a 37% leakage of LNP preferences for Strelow to get ahead of One Nation. ALP needs less in Mirani but this isn't being talked about, let alone counted for the ALP! Need to also bear in mind that some of the LNP leakers here will direct to the ALP, not to Strelow. Nearly impossible. This one is an almost certain ALP hold.

    MaCalister. I think this is a much more likely independent win than Rockhampton. Even if only one-third of Greens voters follow the ticket, on current figures that would be enough. Plus there are another 1500 votes for 3 other candidates which will also surely favour Johnston. She goes into the post-count with a lead over Liberal of around 500-600 I would think.

    Burdekin and Mirani. ALP still has chances in both these though they are unlikely, should certainly not be forgotten. In Burdekin, One Nation will come third and they have directed preferences to the ALP. In neighbouring Whitsunday where they did the same this appears to have been ignored, but still needs to be watched. In Mirani the ALP will need about 35% leakage from the LNP distribution, which seems too high..but Jim Pearce was supposedly popular and long-term member and also might improve on the post count.

    Cook. The only candidate here with any chance of overhauling the ALP would have to be the KAP and they are currently about 100 votes behind the LNP. That gap is the one to watch. Seems a very long shot given the postcount should favour the LNP, but green preferences should leak much more to the KAP than LNP so still a remote chance.

    ReplyDelete
  6. In Maiwar the vote numbers for preferences are ABC estimates, not real numbers. In real-number counting on the night the ALP had a projected margin of a few points. In any case we probably won't see much post-count improvement as the major source of improvement (postals) has been to a large degree already counted, but absent votes which tend to be better for the left have not.

    In Mirani the vote numbers for preferences are ABC-estimated and are definitely not real numbers. I agree that without actual sampling it's difficult to be sure but I understand both sides agree that One Nation are favoured.

    Whitsunday - the One Nation preferences are factored in as the two-party count in this seat is actually real numbers and we know Labor actually is that much behind.

    Burdekin - here the preference distribution is not real numbers and I am unsure when it stopped being so. I had pretty much given this one up on the night as the LNP appeared to have a large lead, but the question is whether that was just a projection perhaps based on incorrect numbers.

    ReplyDelete
  7. I've now put Burdekin back in my in-doubt list.

    ReplyDelete
  8. Maiwar just had prepolls updated, Green 7666, Lab 7467.

    Anita (Poeple decides) does not appear to have preference recommendations, where previously they directed preferences to minor's? People Decide - media. I assume this means a roughly random spread of preferences.

    Are there enough uncounted votes left for Labor to get back into second?

    ReplyDelete
    Replies
    1. I'm seeing 7666-7647. A candidate who polls only a few percent has little influence on preference distributions even if they issue a card, as they will not be handing out much and their vote tends to have a fair sized random component (people who know the candidate, none-of-the-above types and so on). But people who like that kind of policy are quite likely to favour the Greens. Still easily enough uncounted votes for Labor to recover second if they're good enough but absents may favour the Greens.

      Delete
    2. Yup. It looks like I transposed numbers. (copy/paste errors happen - so links are good)

      Delete
  9. Imho, Greens are favoured in Maiwar, and Labor should get the exclusion order beneficial to it in Macalister and Rockhampton. Townsville is unlikely to stay close after postals, could blow out to 51-49 or more.

    ReplyDelete
  10. Bundaberg South? Last I heard it was still in doubt, please advise

    ReplyDelete
    Replies
    1. If you mean Bundaberg, this was just a bogus output from the ABC computer. It was showing in the ABC's in doubt list because the actual LNP lead dropped to something like 51-49 so the computer automatically put it in the in-doubt category, but I knew it was projected to increase greatly. This has now happened with the LNP up 54-46.

      Delete
  11. "the ECQ does not have a current two-party preference count, because they appear not to have picked the correct candidate pair (that's not their fault)."

    looking at the Act, everything about the notional preference count is left to the discretion of the ECQ, including whether or not to do one. The only requirement is that's done on election night, at the polling place, after the primary count and before the ballots are packaged and sealed.

    I think this is a new section, but I can;t put my hands on my copy of the old Act. Am I missing something?

    (5) If the commission considers it appropriate for gaining an indication of the candidate most likely to be elected for an electoral district, the commission may require the commission’s staff to—
    (a) count the preference votes in the way required by the commission; and
    (b) prepare and sign a statement of the number of preference votes (other than first preference votes) for each candidate; and
    (c) advise the returning officer for the electoral district of the contents of the statement.

    ReplyDelete
  12. A small question about terminology: In what circumstances do you use 2CP rather than 2PP?

    ReplyDelete
    Replies
    1. 2CP should be used when the final contest in a seat is other than ALP vs LNP, either in reality or when talking about such a contest hypothetically. (I am not always consistent in writing the right one!)

      Delete
  13. Regarding Rockhampton - isn't her concession quite a bit early? She still has the lead over the PHON candidate (albeit it is lowering) and all it takes is a couple of good batches to go her way.

    ReplyDelete
    Replies
    1. The reports of the preference flow suggest One Nation are second by hundreds of votes after LNP preferences. Indies tend not to do very well in post-counts so if that's true I don't think there's any prospect of that gap being overhauled. All a bit hard to tell exactly what is going on though given the woeful quality of reporting about the Queensland count.

      Delete
  14. Looks like I was wrong, twice.

    Postals in Townsville have come in far more Labor-friendly than I expected. Depending on how many are left, this one could even be leaning towards the ALP at the moment.

    Meanwhile, the first batch of absents (probably of two?) in Maiwar has come in almost dead even between Labor and the Greens. I would still expect the Greens to take make the cut on preferences, but it's certainly close.

    ReplyDelete
  15. The ECQs decision to turn off their notional counts is bizarre. I understand in many seats there are exclusion order issues (many of them in seats that are not in doubt) such that continuing notional counts would be pointless. But in SEQ or in a tight seat with a traditional 2pp count like Townsville, it makes little sense. As far as i can tell they haven't publicly stated a reason why.

    As Bob noted, on the raw figures Labor may be marginally ahead in Townsville. Bonney and Gavin are likely decided (?), but there has been a lot of post count votes added without a preference throw. The LNP look to be doing reasonably well in the post count in Aspley.

    Outside of a few snippets of party scrutineering info leaking out in Maiwar and Macalister, it seems the main source of information is the ABCs projections.

    It would be interesting to know what the ABC protections are based off. The redistribution surely makes it difficult to estimate the number of postal and absentee votes. Ditto to prefercing assumptions, even if they are based on the ECQs notional counts up until Tuesday morning.

    ReplyDelete
  16. Kevin, do you have any idea what's going on in Toohey? The preference flows seem bizarrely weak for Labor given the primary vote changes on 2015. Is this just a projection by the ABC?

    ReplyDelete
    Replies
    1. Don't know anything about it but that preference flow does not look right. It would be a projection, probably based on the flow on the night but something seems to have gone amiss.

      Delete
  17. Toohey now showing a more believable 59.9% to ALP.

    ReplyDelete
  18. Kevin, Hetty Johnston on her fb page said the margin was 483 votes.

    ReplyDelete
  19. Thankyou! That's close to half the primary gap then.

    ReplyDelete
  20. The Aus' latest is saying that the Greens will claim victory in Maiwar tomorrow, and that Labor is "likely" to win Townsville, with it currently having a 130-vote lead.

    Rather curiously, Labor sources are apparently worried that they claimed victory in Mundingburra too early. I assume it's based on a One Nation vs Labor scenario, but the preference flow required to make it eventuate...

    ReplyDelete