Friday, November 30, 2018

Not-A-Poll: Best State Premiers Of The Last 40 Years: Round 2 Results And Runoffs

For the past few months the mostly hopelessly socialist heathen who come here to admire the colour scheme have been involved in the selection of best state premiers of the last 40 years.  This month saw the runoff stage for those states that were not resolved in round 1 by absolute majority, and also the start of the consolation prize round for Coalition premiers.

We are not yet ready to proceed to the grand final stage because some young chap called Andrews was involved in a real election and I feared this could contaminate the vote.  The Victorian runoff will be in February to get a little distance from this result and meanwhile we can continue eliminating Coalition premiers.

In the meantime the winners of four of the remaining states and territories have been decided:

NSW Neville Wran 152 defeated Bob Carr 50
Queensland Wayne Goss 122 defeated Peter Beattie 73
Tasmania Jim Bacon 105 defeated Lara Giddings 84
ACT Katy Gallagher 119 defeated Jon Stanhope 52

These join Don Dunstan and Clare Martin in the final runoff series, which is scheduled to begin in March.

For Western Australia, Geoff Gallop and Carmen Lawrence tied on 97 votes apiece.  Normally I declare a tiebreaking procedure (which in the past has been firstly the leader on primaries at the last stage at which there wasn't a tie, and failing that the leader who was least recently in office wins).  However in this case I did not do this, so there is a runoff for Western Australia, and if there is a tie again the above will be applied.  The WA runoff will run for two months for sample size reasons, this being a quiet time of year.

As for the Coalition premiers and Chief Ministers, the following was the Round1 result for their runoff:

 
 
 
 
 
 
 
 

Total Votes: 236

Candidates Barnett, Marshall and Everingham are all eliminated for failing to reach the 8% threshhold and the rest continue to the next round.   The same tiebreak as noted above applies for the Coalition runoffs.

Sunday, November 25, 2018

2018 Victorian Postcount: Other Indie Challenges (Pascoe Vale, South-West Coast etc)

On this page - Pascoe Vale, South-West Coast, Geelong, Ovens Valley, Werribee, Mildura

Link to state tally and main postcount thread

Link to upper house coverage

I've already posted threads on the interesting post-counts in Morwell, Benambra and Melton.  There are more seats I could post threads of their own on but I shouldn't put too many on the front page!  This seat covers all remaining seats I am aware of where there are interesting issues involving independent candidates creating problems for either major party.  Often in election leadups this is spoken about as a factor but then most of the indie challenges fizzle.  In this case the Coalition's performance has been so bad that it has opened many doors to independents to either beat the Coalition on Labor preferences or beat Labor on Coalition preferences.  Suzanna Sheed has easily retained, Ali Cupper appears to have won Mildura (see below),  Russell Northe is in a fairly good looking position in Morwell, and there are a bunch of others who either can't be written off, or who can be written off but have come close.  Here we go then.  All seats will be updated from time to time unless I have already called them.

2018 Victorian Lower House Postcount: Melton

Melton (Labor vs probably Birchall (IND), 2014 ALP vs Lib (11.2))
2014-elected ALP member Don Nardella quit party, sat as independent and did not recontest
Assessment: It's complicated [update: fairly close to an upset but Labor has won.]

(Link to main postcount page and state summary)

Melton? What is this?

The Daniel Andrews Labor government has crushed the Coalition opposition in the state election, but it's had a bit of bother in a few seats from independents, and these create the only real threat to its 2014 seat collection apart from the fairly likely and widely expected loss of Brunswick to the Greens.

One that sticks out like a sore thumb on the 2PP swingometer is Melton, the former home of Deputy Speaker Don Nardella, who resigned from the party and declined to recontest his seat after being caught up in an expenses claim scandal.  In an election where the swings are a sea of red everywhere except a few safe rural Nationals seats, Melton has produced a 2PP swing to the Liberals of 7.2%.  Currently, the Liberals are getting 58% of all preferences in a safe Labor seat where last time they got 42.4%.  There is the in-theory prospect of a bizarre boilover in this seat, and while someone out there might have information to prove it won't happen, I don't.  Even if it doesn't happen, it is worth keeping an eye on in case such a contest happens again in the future.

2018 Victorian Lower House Postcount: Benambra

Benambra (Lib vs probably Hawkins (IND), 2014 Nat vs ALP (9.7%)
Current 2PP Lib vs ALP figure is irrelevant
Assessment: Probable Liberal retain (update: retained)

(Link to state tally and main postcount page)

The seat of Benambra has been held by conservatives for 141 years but is under siege from independents inspired by the Cathy McGowan victory in Indi.  Bill Tilley's primary has fallen well below 50% leaving him in the danger zone. Here are the current primaries:

Tilley (Lib) 40.29%
Tait (ALP) 17.61
Hawkins (IND) 16.77
O'Connor (IND) 12.91
Knight (Shooters) 8.97
Bardsley (Green) 3.44

Jacqui Hawkins is a McGowan staffer and Jenny O'Connor is a local mayor who was a Greens candidate for the federal seat of Indi.

The Green how-to-vote card preferences O'Connor then Hawkins.  The Shooters registered two cards, one of which preferences Tilley then Tait and the other preferences Tait then Tilley.  O'Connor registered an open preference card (as did Hawkins.) Labor's card preferenced O'Connor then Hawkins with Tilley last.

2018 Victorian Lower House Postcount: Morwell

Morwell (IND vs ALP, Ind Held, 2014 Nat vs ALP 1.8%)
Nat vs ALP two-party figure is irrelevant
Assessment: Northe (IND) wins subject to being 2nd after preferences which is overwhelmingly likely
(update: confirmed, Northe has won)

(Link to main postcount thread and tally)

This is the first of my indie-seat postcounts.  The 2018 Victorian state election has thrown up a very large number of seats where independents have some sort of chance in the postcount and are likely to finish in the top two.  The count in Morwell may be more straightforward than in Melton and Benambra but it is nonetheless still messy.  Perhaps not as messy, however, as many thought it might be.

Russell Northe held the seat narrowly in 2014 despite a monster swing to Labor.  He has been a very much embattled incumbent (including in the final days of the campaign when there was more adverse media coverage of debt issues) but also one who has received plenty of sympathy for his struggles with the unusual pressures of political life in this seat.  He's polled a primary of around 20%, which normally wouldn't be enough, but he may have been saved by the collapse in the Coalition vote.  Here's how the primaries currently line up:

2018 Victorian Postcount: Greens Vs Labor (Prahran, Brunswick, Melbourne)

Link to main postcount thread including state summary

This thread covers late counting in seats being contested between the Greens and Labor.  The Greens went into the election holding Melbourne, Northcote (which they won from Labor in a mid-term by-election) and Prahran (which they won in a ridiculously close three-cornered contest in 2014) and hoped to pick up Brunswick (ALP vacancy) and Richmond (where there is perennial opposition to their candidate Kathleen Maltzahn from sections of the left on account of her support for the Nordic model of criminalising paying for sex).

The Liberals tried to stoke the pot in Richmond by not running a candidate at all, the strategic point of which remains elusive.  Former Prime Minister Paul Keating waded in by accusing the Liberals of piking on the contest to try to dislodge Planning Minister Richard Wynne in order to assist Liberal-linked property developers, while Maltzahn issues were another distraction for the Greens in a campaign full of them.  In the end Wynne has won Richmond with a commanding swing in his favour, and Labor has also comfortably recaptured Northcote.

2018 Victorian Lower House Postcount: Summary And Classic Seats

STATE LOWER HOUSE SUMMARY

SEATS WON:
Labor 55, Coalition 27, Green 3, IND 3

Seats covered on this page:
Bayswater
Ripon
Hawthorn
Nepean
Caulfield

Links to other postcount threads (links to be added as completed):

Green vs Labor (Brunswick, Prahran)
Morwell
Benambra
Melton
Other indie challenges (Pascoe Vale, Mildura, South-West Coast, Geelong, Ovens Valley, Werribee)

Link to Upper House coverage

Victorian Upper House Live

Go to new button press thread for final results and discussion. 

Button presses to occur on Tuesday at 10-minute intervals commencing 2:10 pm.  Very close results (if any) could still be subject to recount beyond that.  ABC Calculator seat "results" (actually output of a flawed but useful model) are not final and some are not likely to be correct.

Warning: The North Metro count section has been rated Wonk Factor 5/5.  Some of the rest aren't too far behind.


Current estimate:

Labor 18 Coalition 11 Greens 1 Transport Matters 1-2 Hinch Justice Party 3-4, Lib Dems 1-2, Shooters 1, Sustainable Aus 1, Animal Justice 1, Reason 0-1

At present the Greens with around 8.5% of the vote will win only 1 seat while either Transport Matters (0.6%) or Liberal Democrats (2.7%) appear likely to win two.

General Considerations: Put The Calculator Away!

Welcome to my Victorian upper house comments.  Apologies for the delay getting on to these but the Lower House count is fascinating.  The Upper House count should be equally so.

The initial results based on the ABC Calculator look like a depressing number of mostly unworthy micro-parties will win.  This may still be the case, but fortunately there has been a large increase in below-the-line voting which is running at around 10%.  This will generally count against micro-parties that are snowballing up from low primaries to beat parties that start on 14%.  Below the line voting is likely to destroy at least some apparent calculator wins, and analysing this will be very complicated.  My suggested rule for this analysis: the calculator is only a tool, use it as a guide but in some cases you may have to  put the calculator away. It's a model based on 100% assumed preference flows, but a lot of the flows will actually be only 80% or 90%, and that will greatly affect the results.  However, it's still likely that the micro-parties with really good preference flows will win seats - and humans trying to work out where the calculator might be wrong can easily make errors too (see Northern Metro Thursday update).

The other important thing is that as I start this article, most of the counts are at only 40-45% counted, which means the prepolls aren't yet in.  As we have seen in the Lower House these are likely to favour the Coalition.  So in seats where the Coalition is narrowly missing out, that may not stay the case.  We may also see that micro-parties generally do worse with more votes added (at present their vote is generally very high.)  Changes in the votes may bring scenarios into view that are not readily apparent at an early look.  These counts are also extremely complex to model and generally micro-party seats can't be called for sure until the button is pressed.

As usual small party voters have tended to vote BTL more than large party voters, and left party voters have done so more than right party voters (spectacularly so in some cases). 

This thread will have a section for each district which will be updated from time to time until the button is pressed.  If one district becomes especially fiendish I may move it to its own thread (if I have time).  In general when I have a look at the districts (every day or so) I won't update those where I don't detect any change in the prognosis.

As I start the thread the districts will be progressively unrolled through the night until all have an initial post up.  The projections are not to be treated as calls of any kind and it is entirely possible other parties will come into the mix that I have not yet considered.

----------------------------------------------------------------------------------------------------------------------
Donations welcome!

If you find my coverage useful please consider donating to support the large amount of time I spend working on this site.  Donations can be made by the paypal button in the sidebar or email me via the address in my profile for my account details.  Please only donate if you are sure you can afford to do so.
----------------------------------------------------------------------------------------------------------------------

Eastern Metro (2014: 3 Liberal, 1 Labor, 1 Green)
Projection 2 Liberal 2 Labor 1 Transport Matters

2:10 am Sunday: As I start this thread Labor is on 39.2%, Liberal 34.5%, Greens 9.2% and a gaggle of micros.  The calculator winner of the final seat is Rodney Brian Barton (Transport Matters) who was a usual suspect in pre-election projections.  His primary vote is 0.61%.

Barton can lose early if he falls behind both Aussie Battler and Australian Liberty, but the laws of small numbers are such that this is highly unlikely.  At present, Barton finishes well over quota at the final count, carrying preferences of every party in his battle with the Greens except for the Victorian Socialists.  Indeed the calculator count doesn't even throw all the votes.

Below the lines don't make a difference here because the flow is way too strong.  Unless there is some way that shifts can muck about with the exclusion order and propel a different winner my initial view is that Barton will win.

9:20 Tuesday: I have not seen anything to change the assessment in this seat.

10:10 Thursday: I have done a manual data entry off the manual count Word document which shows 85.8% counted.  Labor 37.1 Liberal 36.4 Green 8.9 LDP 4.1 and so on with Transport Matters still 0.61.  I am still seeing nothing that stops the Barton (Transport Matters) spiral.

----------------------------------------------------------------------------------------------------------

Eastern Vic (2014: 2 Coalition, 2 Labor, 1 Shooters Fishers and Farmers)
Projection 2 Coalition, 2 Labor, 1 SF+F (Aussie Battler looks highly unlikely)

2:35 am Sunday: As I start this thread Labor is on 34.8%, Coalition 33.2%, Greens 7%, Shooters 5.2, DHJP 4.0, LDP 3.4, AJP 3.3, Labour DLP 1.7, Aussie Battler 1.4%, and according to the calculator Vern Hughes (Aussie Battler) wins.

However, Hughes is vulnerable at an early point where, after receiving ALA, ACP, Health Australia and Labor preferences, he leads Animal Justice and Hinch Justice by less than 2%.  Hughes is coming up from a lower primary than DHJP and AJP so can lose votes to below the line "leakage", but a bigger problem for him here is that the addition of prepolls and postals is likely to drop Labor to around 2 quotas (33.3%) if not slightly below.  If I knock out Hughes at this stage, which at the moment is a fairly likely outcome, the Shooters, Fishers and Farmers become the likely winners.  If I knock Labor slightly below two quotas, Animal Justice can win Labor's seat, but this is an artificial result that will be foiled by below the lines unless Labor drop well below two quotas, which isn't likely.  At the moment I think there is a fair chance of a no-change result in this district.

Monday 4:30: Hughes' position has continued to weaken and has done so rapidly for the small volume of added votes.  He now trails Animal Justice on the calculator at the first crucial point, where he leads DHJP on the calculator by 1.07%.  He is vulnerable to (i) Labor's primary continuing to fall (it is currently on 34.43) (ii) "leakage" of BTLs from his feeder parties (iii) being outperformed by DHJP on the BTLs of parties whose ATLs flow to neither.

Tuesday 8:30: Bourman (SFF) has now moved into the calculator lead with Hughes now eliminated at the cutoff with DHJP even assuming all votes are ATLs.  Labor are only very marginally above two quotas so the calculator might start showing AJP as beating them soon, but I wouldn't take that seriously unless Labor dropped at least 1% below.

Thursday 10:25: 81.1% counted.  Coalition 35 Labor 33.6 Greens 6.5 Shooters, Fishers and Farmers 4.9 DHJP 4.47 etc.  The Shooters are the calculator winners again, although only by 0.01% at the DHJP cutoff.  As noted above this is effectively higher because of BTLs.

Tuesday 6:33: 90.69% counted now and showing on the ABC calculator.  Aussie Battler is over DHJP by 0.11% (534 votes) at the cutoff.  However Hughes is carrying over 2000 votes of below the line liability as well as the issue with DHJP being more likely to snag BTLs as mentioned above.  So I don't have any real doubt the calculator result will be overturned here.

Friday 1:00 91.05% counted.  No real change to the above.  Also note that Labor are holding steady at over 2 quotas.

-------------------------------------------------------------------------------------------------------------------

Northern Metro (2014: 2 Labor, 1 Liberal, 1 Green, 1 Sex Party (now Reason))
Projection: 2 Labor, 1 Green, probably 1 Liberal, 1 Reason/DHJP (potentially both at expense of Liberal though not on current numbers). Outside chance of 3rd Labor instead of Reason or DHJP.

NOTE: My earliest projections in this seat regarding below the lines were incorrect, see Thursday update.

3:08 am Sunday: As I start this thread Labor is on 45.2%, Greens 15.8, Liberal 15.0, Victorian Socialists 4.7, Labour DLP 4.27 (they were slightly to the left of Labor on the ballot), Reason 3.46, Animal Justice 1.8, DHJP 1.67.  Dagiandis (DHJP) obviously has a great preference flow and the calculator has her snowballing and winning from 1.67% and the Liberals missing out to Labor.  However, I expect the Liberals to improve with more counting and hold their lone seat.  The question then is whether Dagiandis can beat Labor and here below-the-lines are relevant.

Currently Dagiandis snowballs from 1.67% to 18.51% on the preferences of: Hudson 4 NV, Country, Liberty Alliance, Transport Matters, Sustainable Aus, Health Aus, Aussie Battler, Shooters, LDP, Reason, DLP and the Greens' surplus after being put over quota by Victorian Socialists.  Labor goes from 11.91 to 16.48 on votes from the Socialists (via the Greens surplus), Vote 1 Local Jobs, Voluntary Euthanasia and Animal Justice.

On the calculator projection Dagiandis beats Labor by 4162, and Labor will probably drop back further.  But 7548 votes worth of below the line is being treated as above the line votes for Dagiandis, compared to only 1717 for Labor.  Moreover, the below the line votes are very probably more likely to flow to Labor than to Dagiandis (especially those coming from the Greens, Reason and Animal Justice).  What looms as a probable killer for the DHJP snowball is the massive Reason Party below the line vote of 55% below the line! (See Thursday update below)

We'll have to keep an eye on just how much Labor's primary weakens but at the moment I think the high BTL rate in this district will trash the calculator projection.

Monday morning: Commenters have hinted at a possible pathway for Fiona Patten to retain her seat.  If the Green vote rises to very close to a quota, Samantha Ratnam might be able to cross the line on BTLs before the Victorian Socialists are excluded.  This is an important tipping point because if this happens, the preferences flowing from the Victorian Socialists exclusion are entirely their own and not the Greens', meaning that the Socialist ATLs go entirely to Reason instead of the same value of votes going mostly to DHJP.  If the Labor vote has also dropped off (which is more likely than not) then it might be possible for Patten to beat Labor, although the huge BTL vote for the Socialists won't help her.  I think this is a rather difficult path for Patten, but it isn't hopeless.

Monday 4:40: The Liberals have improved to 15.57 and in my view will get a quota on primaries, so should appear with a seat in the calculator eventually.  However the Greens have dropped back to 15.26, which is not good news for Patten.  Labor has actually improved so far, to 45.4, and Dagiandis' calculator lead over them is dropping.  Dagiandis may pick up if the Liberals go over a quota; on current numbers she wouldn't win.

Monday 10:45: The Liberals have improved to 16.40 and will soon move into a seat on the calculator.  The Greens are dropping further - absents will pull them back but it is much more difficult now for them to get quota.

Tuesday 9:00: The calculator now shows the Liberals getting quota but it is on DLP preferences via DHJP.  Assuming that the Greens do not make quota before Victorian Socialists are eliminated, DHJP's margin over Labor on the calculator is currently 8278.  But I estimate DHJP's below the line liability at 11226 votes, meaning that they would have to substantially outperform the majors on the below-the-lines of the micros to beat them. It still doesn't look like Dagiandis is winning despite her calculator lead. (See Thursday update below)

Thursday 10:39: Labor 43.5 Liberal 16.7 Green 16.0 Vic Socialists 4.3 Labour DLP 4.1 Reason 3.35 DHJP 1.94 etc with 71.2% counted.  Here I have discovered that some of my previous estimates of DHJP's effective lead over Labor were underestimates, because of a peculiar artefact in the Victorian counting system.  The primary vote value votes of an excluded candidate are treated as one count in the preference distribution, and then all their remaining full-value votes are treated as a second.  This means that ATL votes from the Voluntary Euthanasia Party and Animal Justice Party, which I was treating as Labor votes in the event of a contest between Labor and DHJP, are in fact DHJP votes in that contest, because DHJP can get them if it is not already over quota on Fiona Patten's primary votes.

The consequence of this is that the effective margin of DHJP over Labor is in fact not the 3.8% I thought but more like 9.1%, meaning that the Dagiandis spiral will not be derailed by below-the-lines, and Dagiandis will win unless the Greens cross quota before the Victorian Socialists do (in which case Patten wins).

As @sorceror43 notes, the low percentage completed means it is likely there are still lots of inner-city absents (or out-of-electorate prepolls) to be added and these are likely to boost the Greens, perhaps at the expense of the Liberals.  The Liberals would have to drop a long way to not win themselves (because of BTLs), in which case both Patten and DHJP would win, but the chance of Patten beating DHJP because of the Greens going up to quota must be quite realistic.

Sunday 3:30 Commenter David J has noted that the Word documents now include district totals making them much more tractable.  On counted votes both Liberals and Greens are just a whisker shy of quota, which means they would make quota on BTLs and Patten would win.  However the question is whether the Green vote would stay high enough when the undercount in Yuroke (where they are polling only 4%) is rectified.  I am going to have a go at a projection later today.

Sunday 6:30 I have run a projection which assumes the uncounted vote in each electorate will break the same way as the counted vote and the turnout will be the same as in 2014.  The critical votes are the Liberal and Green votes and in this projection the Liberals are on 16.53% and the Greens on 16.44%.  However the current count is completely missing absents in Pascoe Vale, Northcote and Yuroke and includes an extremely small sample of early votes from Pascoe Vale.  The Greens are likely to do a little worse on the early votes than the ordinary votes, but significantly better on the absents, so it's not clear how much their tally might move around.

Based on the BTL figures compiled by David J, about 1.3% of the count will be BTLs for the 12 micros that can expect to be excluded before Victorian Socialists.  These BTLs will be splitting either six or seven ways: Labor, Green, Reason, DHJP, FPRP, exhaust and possibly Liberal.  It's plausible that if the Greens do finish about 0.23% shy of quota, they could make that (probably about 1000 votes) back on these BTLs.

However the potential killer for Patten's chances is leakage from minor Green candidate BTLs.  In 2014 the Greens dropped 1033 votes off leakage from their three excluded minor candidates while gaining only 336 from Labor surpluses and leakage from minor Labor and Liberal candidates.  This time the Greens will have four minor candidates excluded instead of three, Labor will have two instead of one, and Labor will have a somewhat larger surplus.  Even so with the increased BTL rate it is quite easy to see the Greens, and by extension Patten's prospects, being 500 votes if not more down at this point.  That on current numbers could leave them needing more like a 30% share rather than an 18% share of the BTLs escaping from other parties.  An especially bitter irony here could be that Patten is one of the common destinations of leakage from minor Green candidates (getting 20% of their leaks in 2014) and could lose because votes from minor Greens leaked to her instead of flowing to other Greens!  

At the moment this looks like one of these counts where we just won't know til the button is pressed - let me know if I've missed anything or made any obvious mistakes here!

Wednesday 12:30 am: Updated current totals again posted by David J in comments suggest the Greens have picked up again; they are on 16.54%.  The Yuroke count is up to speed with the most lagging districts being, in order: Pascoe Vale (about average for Greens), Northcote (excellent) and Broadmeadows (terrible).   So maybe the Greens can hold at about this level, at which they will probably just make quota before the Victorian Socialists cutoff.

David also raises an issue I mentioned briefly further up: that even if Dagiandis doesn't get the Green preferences, there is another chance for Dagiandis down the track - Patten must stay ahead of Labor.  Although Patten is around 8150 votes ahead of Labor late in the count, Patten is carrying 8126 votes of BTL liability, mostly from the Victorian Socialists.  Labor has a small amount of BTL liability from leakage from exhaust and also from its minor candidates, which in 2014 cost it 718 votes with 75 going to Patten.  With Labor's vote increased this liability could be close to 1000 this time.  On neutral BTLs (those not for a candidate whose HTV preferences flowed to either party) as best I can tell Patten did somewhat better than the minor ALP candidates in 2014. Whether she would do better off Victorian Socialists BTLs than Labor I am unsure of.  I think Patten would probably survive at this point, but it isn't clear.

If Patten is out at this point, Dagiandis then has a notional calculator lead over Labor which I get at about 12600 votes (though this may be a bit inaccurate as I have had to make some calculator fiddles to bring up this scenario).  BTL liabilities are about 12600 for Dagiandis (mostly Patten) and 7500 for Labor (mostly Victorian Socialists).  I would think Dagiandis wins here but I can't write Labor off entirely.

Wednesday 8:45 David J's latest figures have the Greens over quota and the Liberals falling to 16.11%.  I will be having a serious look at what sort of primary the Liberals need here to be safe from losing as they get almost no ATL preferences (a tiny parcel from Vote 1 Local Jobs as part of a surplus from Patten).

Wednesday 10:43 On the current calculator (based on votes as relayed by David J) the Greens reach quota on primaries, the Victorian Socialist preferences flow to Patten, and Patten beats Labor by about 13700 which seems easily enough.   Patten's preferences then almost all flow to Dagiandis putting Dagiandis over the Liberals by about 4,000.  However, Dagiandis carries BTL liabilities worth at least 4600 even before she gets to Patten's preferences.  After that, she is vulnerable to whatever net loss Patten has sustained on BTLs on her way up the ladder, while the Liberals pick up leakage and, except for small losses from their own minor candidates, can only improve (perhaps to quota).  It's not quite as simple as saying that every one of the 6300-odd Vic Socialist BTLs that doesn't go to Patten hurts Dagiandis, because these can be compensated by whatever BTLs Patten got from someone else.

Another curious thing is that at this point another ugly and undemocratic feature of the election system makes an appearance - unweighted inclusive Gregory distortion.  Unweighted Gregory is a ludicrous leftover from the days of hand-tallying in which all ballots in a surplus are given the same value (whatever their previous value) so that electoral officials don't have to keep track of countless transfer values and multiply 10-digit decimals by each other.  These days all this is easily done by computer so what unweighted Gregory does is distort vote-values in a way that can cause votes to increase in value unfairly, or lose value unfairly, or not lose as much as they should.  In this case the effect of it is that Labor preferences swamp Patten's surplus meaning that her own votes are devalued as surpluses and Dagiandis becomes much less exposed to damage from Patten's huge pile of BTLs.  The Patten votes which were worth 1 vote get cut down to 0.08 apiece, while the Labor votes that were worth only .22 apiece get cut down to the same.

Projecting just how much the flow from Patten to Dagiandis might get affected by BTL issues is getting even beyond my pay grade.  It might be only a few thousand votes.  On current primaries, I think the Liberals are fine here, but if they lose, say, another 0.5%, this won't be so clear-cut.  Yes, North Metro has massive below-the-line vote rates, but nearly all of those are for Greens, Labor, Patten and Socialists, and the first three get cut down in surpluses if we get Dagiandis v Ondarchie (Lib) for the final seat.

Friday 1:00 The ABC calculator figures are still not useful as the calculator is still lagging the actual count here.  However it should catch up soon.

Friday 7:00 The ABC calculator has finally caught up in North Metro with 87.9% counted.  Currently the Greens have 16.74%.  Based on that Ratnam won't make quota on primaries, and whether the Greens' pickup on BTLs from other parties is enough to compensate for leakage from their other candidates is not clear - one would say probably, but any further drop back of the Greens total makes things precarious.  The Liberals are currently OK, having picked up to 16.46%.

---------------------------------------------------------------------------------------------------------------------

Northern Vic (2014: 2 Coalition, 2 Labor, 1 Shooters, Fishers and Farmers)
Projection: 2 Labor, 1 Coalition, 1 Liberal Democrat, 1 DHJP (2 Coalition instead of 2 Labor now looks unlikely)

3:45 am Sunday: As I start this thread Labor is on 32.5%, Coalition 30.8%, Shooters 8.3, Greens 6.9, DHJP 4.3, Liberal Democrats 3.4, AJP 2.22, VEP 1.87 etc.  The calculator has Labor winning 2, Coalition 1, DHJP and Liberal Democrats 1.

DHJP's Tania Maxwell however, has a juncture where she only just get over the Greens.  Currently 715 more votes are being treated as above the line in DHJP's pile to this point than for the Greens, and that exceeds the margin.  It's quite possible Maxwell gets eliminated here (indeed minor changes could eliminate her even earlier.)  Even if she does get over this point there is a later point where she has snowballed to just over 1% ahead of Labor.  That might expand with more counting, but even so the Greens' below the line vote is 1.6% by itself.  So I have a lot of doubts about this snowball and think it more likely both majors will win two.  (See some comments from Henry Schlechta in comments section - at the moment I still think the Coalition are best placed on current numbers, but it could shift pretty easily.)

The Tim Quilty (LDP) preference spiral, on the other hand, is not so easily disposed of.  Presently Quilty has a calculator lead of 5684 votes over the Shooters at the key point.  His feeder parties have only 2915 below the lines so I suspect that the Shooters won't bridge that gap and that Quilty will take their seat - unless the vote for all these micros declines sharply or the Shooters lift.

Monday 4:45: There has been quite a turnaround in the primaries here already with the Greens dropping back to 6.45 and DHJP climbing to 4.68.  As a result Maxwell's calculator lead has shot out to a whopping 2819 at that point.  Also, Maxwell's calculator lead over the majors is now just over 4%, which seems like plenty.  If both micros get up this pitches Labor and the Coalition into a battle for the final seat, which Labor is currently just winning but I don't believe that would stick. Absents might yet turn this around but there are still a lot of prepolls to be added.

Monday 5:40: On cue the calculator has flipped the final seat to the Nationals.

Monday 10:35: @sorceror43 has noted that the prepolls counted so far appear atypical in that the Liberal to Labor ratio is even higher than in postals, so they may be from an unrepresentative area.  That ratio should moderate but whether things can be dragged back in Labor's favour is another question.

Tuesday 3:30: Labor's #2 candidate Jaclyn Symes has just been promoted to Cabinet, which is being taken as a sign that Labor think they will win.  The likelihood here is that there is a disproportionately uncounted vote in Yan Yean where Labor did very well off the back of a Liberal candidate disendorsement in the Lower House.

Tuesday 9:00: The calculator has reverted to Labor for the final seat.  However its margins are very close.  At the moment the DHJP seat looks reasonably comfortable but I'll wait for further counting there.

Thursday 11:25: Coalition 32.1 Labor 31.3 Shooters 7.9 Greens 6.3 DHJP 4.9 LDP 3.8 AJP 2.17 VEP 1.96 etc (81.8% counted).  The calculator still has both Quilty (LDP) and Maxwell (DHJP) winning and currently has Labor just winning the final seat but by a margin I wouldn't go placing confidence in.

Saturday 5:20: @sorceror43:

"I've copied the district totals into Excel for North Vic, and calculated a grand total, then percentages. With count at 87.4%, Labor 31.9%, L/NP 31.4%, Shooters 7.8%, Greens 6.5%, Hinch 4.8%, Lib Dems 3.8%. With primary vote lead, think Labor wins final seat"  

Wednesday 1:00 am: This count is very well advanced now at 90.66% and Labor is looking good for the final seat.

----------------------------------------------------------------------------------------------------------------------

South East Metro (2014: 2 Coalition, 2 Labor, 1 Green)
Projection: Looking like 3 Labor, 1 Coalition, 1 Lib Dem or possibly Transport Matters

4:05 am Sunday: As I start Labor is on 52.7, Liberal 26.9, Greens 5.61, DHJP 2.75, AJP 2.12, Labour DLP 1.45, Shooters 1.31 and Transport Matters' Ali Khan 1.14.

According to the calculator Khan then snowballs to victory while the Liberals do not see another vote, and the final margin is almost 13%.  Thus far I can't see anything that might interfere with this result, not below-the-line votes alone, perhaps there is some exclusion order issue that hasn't been found yet.  This is a result that was projected as possible by commenter David J a few days ago.

1:50 Tuesday: Tim Quilty (see above) in comments has noted that Transport Matters are at some risk of being chopped off at the point where they fall to second-last (Count 20, where their calculator lead has shrunk to 3891 votes (1.49%).  Transport Matters' vulnerability to losing votes from BTLs is fairly small at this point (about 1370 votes if treating all Labor BTLs as 1 for the lead candidate, but higher in reality because the Labor vote will be disproportionately for minor candidates - see Alaric in comments) while the Greens can only lose BTLs from Vic Socialists (217).  However the Greens can also gain on BTLs from third parties, especially Reason who have 893 BTLs.  At the moment there are not enough BTLs to dislodge Transport Matters at this point.  However any vote Labor loses in post-counting takes a vote off Transport Matters' calculator lead.  Labor currently contribute 1.25% to Transport Matters so if they drop back to quota Transport Matters will lose.  In this case it appears the winner is the Liberal Democrats, off an even lower 0.79%, but I will look at this more carefully should it look more likely.

11:40 Thursday: ALP 50.6 Lib 28.8 Green 5.37 DHJP 3.01 AJP 2.15 LDLP 1.45 Shooters 1.43 Transport Matters 1.22 FPRP 0,86 LDP 0.82 with 75.9% now counted.  The lead of Transport Matters over Greens at the key point on the calculator is now only 1.03%, which is getting a little bit shaky.

5:13 Monday: Current live figures posted by David J in comments have ALP 50.18 Lib 29.02 Green 5.5 DHJP 3.01 AJP 2.17 LDLP 1.44 Shooters 1,42 TM 1.24 FPRP 0.85 LDP 0.82 with 82.3% counted.  This would give Transport Matters a calculator lead over the Greens at the key exclusion of 0.52%, which I don't think would be enough (if they fail, the Liberal Democrats win).  I haven't taken into account where these votes are from as yet.

12:05 Friday: The ABC calculator has caught up with now 86.0% counted.  Labor has fallen below three quotas so now its third candidate carries ATLs instead of its fourth candidate.  Transport Matters are over the Greens by only 0.34% on the calculator, which seemingly shouldn't be enough.  The difference between TM's BTL liabilities and the Greens' is very similar to TM's calculator lead, but the Greens are much more likely to make gains on BTLs.  However it has occurred to me that Labor falling below three quotas presents a possible small advantage for Transport Matters.  Firstly the Labor BTLs are no longer a liability for Transport Matters.  Secondly if Labor actually crosses quota on BTLs from some party that doesn't preference Transport Matters, then that creates a surplus for Labor that flows almost entirely to Transport Matters, which in the most optimistic scenario might give Transport Matters say 150 votes.  Even with this considered it still looks like Transport Matters are struggling.

5:00 Friday: Labor now on 49.98% with 88.36% counted.  Whether Labor actually hit three quotas here on primaries probably doesn't matter much because leakage and votes for minor candidates will probably mean that their third candidate does not cross on surpluses from the first two in any case, and will get across the line on BTLs or Greens preferences at some unknown stage of the count. The Transport Matters - Greens gap on the calculator has stretched to 0.40% which might be survivable but at present I still doubt it.

-----------------------------------------------------------------------------------------------------------------------

Southern Metro (2014: 3 Liberal 1 Labor 1 Green)
Projection: 2 Liberal, 2 Labor 1 Sustainable Australia

4:25 am Sunday: As I start Labor is on 37.6, Liberal 35.6, Green 13.2, Animal Justice 2.12, Reason 2.08, LDP 1.46, Labour DLP 1.37, and Clifford Hayes (Sustainable Australia) 1.33.  The calculator script is simple: Hayes gets everyone's above-the-line votes except the Socialists, and beats the Greens by 11338 votes (nearly 6%).  Hayes is never at danger of exclusion and while there might be some rival harvester, I'm not aware of one at this stage.  Assuming it is Hayes vs Greens the issue is below-the-lines.  At present Hayes is relying on votes including about 6769 votes worth of BTLs, and a lot of those BTLs (eg from Reason, Labor and AJP) are likely to flow to the Greens.  But at present the Greens need to get about two-thirds of the BTLs which is completely unrealistic.  The Greens will need to lift substantially, or have micros fall substantially, to hold their seat.  Sustainable Australia looks quite strong here.

9:20 Tuesday: I have not seen anything to change the assessment in this seat.

11:50 Thursday: 75.5% counted and it's Liberal 37.9 Labor 35.3 Green 13.2 AJP 2.2 FPRP 1.9 DHJP 1.44 LDP 1.38 LDLP 1.33 Sustainable Australia 1.32. With more or less exactly one-tenth of the Greens' vote, Sustainable Australia's Clifford Hayes appears to be beating them easily.

12:50 Friday 7th: 87.96% counted and the Greens have come up to 13.93%, but on the calculator are still losing by over 23,000 votes.  SA's below-the-line liabilities come out to about 13,000-14,000 so they are still winning rather comfortably.

---------------------------------------------------------------------------------------------------------------------

Western Metro (2014: 2 Labor 1 Liberal 1 Green 1 DLP)
Projection: 3 Labor 1 Liberal and 1 DHJP or possibly SF+F

4:55 Sunday: As I start this one the leaders are Labor 47.88 Liberal 20.02 Green 9.31 DHJP's Catherine Cumming on a stonking 6.71% and Labour DLP on 3.37.  The calculator has Cumming making gains from Sustainable Australia, Animal Justice, Reason, Voluntary Euthanasia and Hudson 4NV to get over the Greens, at which point their preferences put both her and Labor over the line defeating a rival snowball from Labour DLP.  At the moment the calculator margin at this key point between Cumming and the Greens is 3676 votes but Cumming is notionally carrying 2301 BTL votes to 504 for the Greens (which are Socialist votes that may well flow to them anyway).  Quite a lot would have to change for Cumming to drop out at this point instead of the Greens so at this stage I think she could very well win.

9:20 Tuesday: I have not seen anything new in this seat other than Cumming's position improving.

10:30 Wednesday: ...but commentator Cam Nation has! The Liberal Party are sneaking up on Labour DLP at a critical exclusion and on current numbers would beat them there because of BTLs, and the resultant preference flows appear to result in Cumming losing to the Shooters, Fishers and Farmers.  The reason is that if Labour DLP go out before the Liberals then the Liberal preferences do not flow to Labour DLP, the Shooters get Labour DLP preferences, and as a result the Shooters get Liberal preferences as well, which saves them from elimination.

It might seem that the Shooters should be more vulnerable to BTLs than DHJP but in fact the Shooters' spiral consists mostly of right-wing parties and major parties - all with low BTL rates - while DHJP's includes Animal Justice, Reason, Vic Socialists and Greens.  As a result, while there is a further cutoff point in this scenario with Shooters over 10,000 up on DHJP, DHJP in fact go backwards once BTLs are added.  If the button was pushed right now it appears the Shooters would actually win and it would not be surprising to see them soon take the calculator lead.

Cumming may however recover and still win should the Liberal vote go down, which @sorceror43 has noted is possible because Footscray is under-counted.

12:01 am Friday: 79.2% counted and I have Labor 46.1, Liberal 21.5, Greens 8.9, DHJP 6.89, LDLP 3.47, AJP 2.48, SFF 1.91, LDP 1.66.   I have the Shooters in the calculator lead as a result of the Liberals being 0.01% ahead of Labour DLP on calculator numbers, which would be a much greater lead in reality.  It will be interesting to see if this can be turned around.

3:30 Saturday: The VEC site is showing a tick for this seat as final and declared but with only 25% in the rechecked count and no other indication of finality I believe this is an error.

10:30 Tuesday: David Jeisman (David J in comments) has kindly sent me current primaries culled from the spreadsheets - as of last night ALP 46.11 Lib 21.34 Green 8.96 DHJP 6.8 LDLP 3.53 AJP 2.52 SFF 1.92 LDP 1.70.  86.34% counted.  These are minor changes but have enough impact to push Labour DLP ahead of the Liberals by 0.27% on the calculator at the key point.  Labour DLP's notional lead of 1218 votes includes 528 votes in BTLs that in general won't flow to them.  The Liberals don't get any ATL votes before this point but can lose votes from leakage from their minor candidate BTLs and their one surplus - based on 2014 figures this could cost them around 300 votes.  On these figures I don't see that the Liberals could pick up enough BTLs from other parties to overturn what is effectively a 1000 vote deficit so it looks like Labour DLP will survive this exclusion on current numbers.  (Perhaps late counting could still change this).

The other point of interest here is that Labour DLP themselves get within 10060 votes on current numbers of beating DHJP at the end.  Although DHJP are coming from a higher primary, they are in general carrying ATLs from the left-wing micros with higher below-the-line rates while Labour DLP are carrying the right-wing ones.  On my numbers DHJP have at least 3700 votes more exposure to below-the-lines than LDLP do (probably somewhat more because of issues with minor candidates for Labor and the Greens), but they are also more likely to pick up BTL leaks than Labour DLP do.  So on current numbers DHJP are not in danger here.

On current numbers it looks to me that Catherine Cumming is winning the final spot again, but we need to see if the exclusion point involving the Liberals and Labour DLP tightens as there are probably only several hundred in it at the moment.  There is a slight undercount in St Albans district but nothing major.

Thursday 1:07 Note that the ABC calculator now has Shooters winning but it is based on less complete count figures than above.

Friday 2:45 With 89.16% counted in the spreadsheets I have the calculator gap of Labour DLP over Liberal (which determines the DHJP vs Shooters contest) at 0.16%.  A calculator lead of around 700 votes, which is the sort of margin at which either side could win.

Friday 5:00 The calculator has very nearly caught up now, on 88.81% and has the gap out to a more comfortable 0.34%.  As @sorceror43 has noted it is not clear if the difference is caused by changes on rechecking or changes on votes yet to be rechecked (the former means DHJP are still winning, the latter that the seat is very line-ball.)

Saturday 2:20 The calculator now has 89.24% counted and the Labour DLP over Liberal gap on the calculator is 0.37% so it looks now like Cumming is winning.

------------------------------------------------------------------------------------------------------------------------

Western Victoria (2014: 2 Coalition 2 Labor and Vote 1 Local Jobs)
Projection:  2 Labor 1 Coalition 1 DHJP 1 AJP (2nd Coalition appears unlikely.)

5:00 Sunday: As I start this one Labor is on 39.16, Coalition 29.09, Greens 7.57, Shooters 4.75, DHJP 3.81. Animal Justice 2.69, Liberal Democrats 2.54, VEP 1.77 etc.  The calculator has both Animal Justice and DHJP winning with 2 Labor and 1 Liberal.

On current numbers, however, the Animal Justice candidate Andy Meddick is only 2550 votes on the calculator up on Labor at the point where he gets their surplus.  Notionally he is carrying 2354 below-the-lines here, which are probably going to flow significantly to Labor, so on current figures he's out at that point.  However if Labor drop back (which they probably will) he may get over that one.  If Meddick survives at this point he wins on Labor and Green preferences.  In that case, the calculator then has Grimley winning by 12047, but there are around 6500 votes worth of BTLs that can interfere with that at various stages (either by failing to reach AJP or by failing to flow from them to DHJP).

At the moment that gap is too large, but the Coalition are likely to come up and might reach a point where Grimley's win becomes in serious doubt if Meddick wins.  If Meddick loses, Grimley wins, and the Coalition seems to get the final spot on Greens preferences over the Coalition.  This one is very messy at the moment.

5:00 Monday: Meddick's position relative to Labor is now a 1.04% lead (2914 votes) which is about holding station given the change in votes counted (now 56%).  The calculator has another addition to the mess, which is it now has Grimley reaching quota before Meddick.  On current numbers this wouldn't actually happen because of BTLs.  At the moment I think Grimley has a good chance because he has multiple paths to victory, while Meddick is touch and go.

12:10 Friday: With 82.9% counted Labor 38 LNP 30.6 Green 7.5 SFF 4.38 DHJP 4.32 AJP 2.67 LDP 2.61 VEP 1.86 etc.  The calculator still has both AJP and DHJP winning.

Meddick is now 2.29% ahead of Labor at the point where he gets their surplus, which is easily enough.  There is also a large buffer now for Grimley after Meddick gets elected (assuming Grimley doesn't cross first) and on current number it looks like both micros win.

1:00 am Wednesday: 91.6% counted and I'm not seeing any reason why Grimley and Meddick don't both win.  The calculator at the moment has Grimley over first, I think as a result of Shooters going out before Greens.

Saturday, November 24, 2018

Victoria 2018 Live

The starting line: Labor 46 Coalition 37 Green 3 Ind 2 (Melton treated as Labor)
Polls have closed
Seats apparently won (some at low levels of doubt) ALP 51 Coalition 24 Green 1 Ind 2
10 seats currently in significant doubt (that I know of)

Apparent Labor gains from Coalition (some still in some doubt): Bass, Mt Waverley, Ringwood, South Barwon, Burwood, Nepean, Box Hill

Coalition seats in doubt: Bayswater, Ripon, Hawthorn

Apparent ALP gain from Greens: Northcote

Apparent IND gain from Nat: Mildura

In doubt Coalition held vs Ind: Benambra, South-West Coast (likely hold but exclusion order issue)
In doubt IND held vs ALP: Morwell (Ind favoured)
In doubt ALP vs Ind: Melton
In some doubt ALP held vs Ind: Pascoe Vale (probable ALP hold)
In doubt Green held vs ALP: Prahran (ALP ahead)
In doubt ALP held vs Green: Brunswick 

Friday, November 23, 2018

2018 Victorian Final Polls

Galaxy 53-47 to Labor, ReachTEL 54-46 to Labor
Current primary vote aggregate ALP 40.7 L-NP 39.4 Green 11.0 Other 8.9
Polls could be underestimating Labor 2PP vote slightly and may be overstating Greens primary
Seat projection estimate ALP 48 L-NP 36 Green 3 IND 1

This post will update all polling news in the final 24 hours of the Victorian campaign.

If the latest polls are right, yesterday's token post about Lower House modelling might not be quite so token after all. What we've seen in the Herald Sun's YouGov-Galaxy (53-47 to Labor) and Fairfax's ReachTEL (54-46 to Labor) suggests that the net effect of the last few weeks of campaigning has been more or less zero.  The polling would not have to be wildly wrong for an unlucky distribution of seats to leave Labor short of a majority, but it would have to be very wrong indeed for the Coalition to win the election in any way.

Wednesday, November 21, 2018

A Token Post About Modelling The 2018 Victorian Lower House

VICTORIA (ALP 45 COALITION 38 IND 3 GREEN 2)
Seat modelling on assumed 2PP of 53.4 to Labor currently gives estimate around ALP 47 Green 3-4 LNP 36 IND 1-2 
On current numbers Labor are very likely to win, but at some risk of doing so in minority
Contest badly lacks sufficient recent polling data so any modelling is unreliable

State opinion polling aint quite what it used to be.  At this stage of the 2014 Victorian state election, there had been twelve statewide voting intention polls by six different pollsters released in the previous two months.  This time it's four by either two or three (depending on how you treat Newspoll/Galaxy) and the most recent one was commissioned.  Over a million voters have voted already (including those whose votes are in the mail) and yet so far this month the only statewide poll we've had is a ReachTEL for the Victorian National Parks Association.  There may well be a flood of polling in the final days, but at this stage, those of us trying to predict what might occur have not a lot to work with.

Saturday, November 17, 2018

Site colour change

From time to time I change the colour of this site, often partly or entirely for some reason connected with its content.  Examples of past colours adopted have been:

a shade of dark blue which was the subject of a ludicrous cease-and-desist letter from the Tasmanian Liberal director asking that an "independent liberal" candidate cease using the colour "Liberal Blue"
Orange, partly in amazement at Cathy McGowan's team finding 1000 votes under the proverbial table during the 2013 Indi count.
Purple, signifying neutrality between the major parties
The colour of Senate ballot papers, moving to a purely psephological colour as an expression of disgust with the federal parliament over anti-free speech provisions in the rushed-through same-sex marriage plebiscite safeguards bill.
Burnt orange, flying SA-BEST colours in protest against SA Labor and Liberal parties preferencing the Australian Conservatives.

The new colours (though I'm fiddling with the combination to try not to make it too hard on the eyes) are another protest, concerning the behaviour of nearly every party in or leading up to the current Victorian Legislative Council election.

Friday, November 16, 2018

Wentworthless: Another Epic Seat Poll Fail

The failures of seat polling have been a common subject on this site this year.  See Is Seat Polling Utterly Useless?, Why Is Seat Polling So Inaccurate and How Did The Super Saturday Seat Polls Go?

The recent Wentworth by-election was difficult to poll because of a late strategic-voting swing of probably a few to several points from Labor to the winner Kerryn Phelps.  All seven polls that polled a Liberal vs Phelps two-candidate preferred vote did actually get the right winner.  But that is all the good news that there is.  In so many other respects, the seat polls for the historic Wentworth by-election, perhaps the most polled seat in Australian history, were way wrong. And like other recent seat poll failures in such seats as Bass, Macarthur, Dobell, Lindsay and Longman, the failures were characterised not just by the polls being very wrong, but also by them tending to be wrong in the same direction.  The problems go beyond small sample size, and beyond even the tendency of seat polls to be less accurate than their sample sizes say they should be.  They point to systematic errors not random ones, and in this case, I suspect, to the oversampling of the politically engaged.

Wednesday, November 14, 2018

Fear And Loathing With Victorian Upper House Preference Flows

Following the launch of Antony Green's Legislative Council calculator I've been playing around with some possible scenarios for the Victorian upper house group ticket flows.  Quite a few people are doing this and so there are a number of different estimates about what might happen out there.  What we know from the past is to expect the unexpected - we can say that it looks like preference harvesters will win several undeserved seats, but it's hard to say which ones they will be and who.  The whole exercise is incredibly sensitive to starting assumptions - one micro-party you've never heard of might get 1% instead of 0.5% and suddenly something completely different happens.  Snowballs from very low vote shares have a higher chance of crashing because of below-the-line votes, especially as voters for micro-parties, with the exception of the Liberal Democrats, are more likely to vote below the line.  In 2014 the BTL rate for most micros was in the range 8-22%.

At the last Victorian election, five candidates won seats as a result of preference-harvesting:

* In Eastern Victoria, the Shooters, Fishers and Farmers (2.44%) beat ALP-2 (8.68% over quota) and Green (8.23%)
* In Northern Metro, the Sex Party (2.87%) beat Labor-3 (7.06% over quota)
* In Northern Victoria, the Shooters, Fishers and Farmers (3.5%) beat L-NP-3 (7.84% over quota) and Greens (7.68%)
* In Western Metro, Democratic Labour Party (2.57%) beat ALP-3 (10.65% over quota) and L-NP-2 (6.90% over quota)
* In Western Victoria, Vote 1 Local Jobs (1.28%) beat Greens (9.19%)

There weren't any cases of candidates winning from well below 1%, but based on our experience of the new Senate system since, none of the above would have won had voters made their own preferencing decisions.  These parties only won because the Group Ticket Voting system created completely fake near-100% preference flows.  Perhaps, had the Senate system been implemented in Victoria before that election, some of the minor parties would have merged into larger groups and polled higher primaries, but that doesn't seem all that likely.

With the release of the new round of Group Tickets it seems that almost all parties have been involved in backroom preference-trading.  There are again tight flows between the micro-parties, largely believed to be networked by Glenn Druery, that seem designed to elect a particular winner or choice of winners in each seat.  Labor has preferenced a range of, for progressives, dubious parties above the Greens in what looks like an attempt to replace the Greens with Druery parties:

* The Aussie Battler Party, an anti-immigration populist outfit that wants to place juries of randomly selected citizens in control of many aspects of the political system, and that promotes illiberal law and order policies including indefinite sentences. This party is the latest home for long-time conservative party-hopper Vern Hughes.
* Sustainable Australia, a seemingly environmentally focused party that is actually a home for old-style pre-Tampa immigration-cutting Dick Smithery.
* The Shooters, Fishers and Farmers Party.  Guns!
* The Liberal Democrats.  More guns!

(etc)

The Greens are far from blameless themselves, having preferenced the anti-vaccination, anti-fluoridation Health Australia Party, the aforementioned Aussie Battlers and Sussos, and also the law-and-order-loving Derryn Hinch's Justice Party all above Labor in various seats.

This basically means that if you want to vote for either Green or Labor and preference the other in the Upper House without your vote going to at least partial right-wingers or crackpots first, you have no choice but to vote below the line.  

The Liberals, for their part, seem to think that the only parties worse than Labor and the Greens are the Australian Liberty Alliance and Victorian Socialists.

Some of the micro-party preference orderings might be taken as vaguely logical (Fiona Patten's Reason Party) while others (eg Animal Justice) are simply all over the place.  I mention the Victorian Socialists (see comments) as one party that has produced an ordering that very closely reflects its likely voters' views.  The Liberal Democrat and Hinch Party orderings of the major parties and Greens jump around between individual candidates in a way designed to confuse the average voter out of having the slightest idea who is actually being preferenced and what the effect will be.

Some possible scenarios

In trying to test some ranges of possible outcomes, here were some assumptions I made:

* Labor currently seems on track to win the election with a modest swing to it, based on state polling.  I assumed this would be the case, all else being equal, and that the Labor primary might be up a shade.

* The field of micro-parties looks mostly weaker than last time, especially on the religious Right.  So even if Labor wins the election I assumed the Liberal primary would be little changed.

* Generally I gave new and unknown parties about half a percent of the vote, unless they seemed hopelessly limited in appeal (Hudson For Northern Victoria won't get a lot of votes outside Northern Victoria, if even there).

* Logos are now displayed on ballot papers.  I assumed this would reduce the strength of the link between ballot position relative to the majors and votes for the confusingly-named parties (Liberal Democrats and Labour DLP).

* I thought the vote for Labour DLP might be down somewhat as the party has lost all the MPs it formerly had at state and federal level.

* The calculator assigns votes for the Sex Party to Fiona Patten's Reason Party.  But as with the change from Family First to Australian Conservatives, the name change switches from a name with a certain appeal to low information voters to a name that is more obscure (and I also think, in Reason's case, pretentious.)  I could be wrong but Reason might not do as well as the Sex Party did, outside of Patten's home seat.

* It's hard to predict what vote Derryn Hinch's Justice Party might get without Hinch on the ballot paper.  In states outside Victoria it polled less than 1% in the 2016 election.  I think it should do quite a bit better here and have guessed 2-3%.

On my first run of this and with some subsequent fiddling, here were some possible outcomes I got:

* Eastern Metro: In my first attempt I got Transport Matters off 0.5% beating the Greens off 10.4%.  If Transport Matters dropped out early or the Liberal Democrats polled 5% off name confusion (seems unlikely) then the Liberal Democrats took the seat instead.  Either way I seemed to get a preference harvester alongside two of each major party.

* Eastern Vic: On my first attempt the Liberal Democrats off 4% beat the Greens off 8.4%.  If I reduced the Lib Dem vote, the Shooters Fishers and Farmers won off 2.4% instead.  Again I kept getting two majors apiece and a preference harvester.

* North Metro: My first attempt elected two Labor, a Liberal, a Green and one from Hinch Justice, with the latter on 2% beating ALP-3 on 10.4%.  When I tried to elect Fiona Patten instead of Hinch Justice, I found I had to jack her vote up to around 7%.  Depending on how much of that I took from the Greens, in some cases I could leave the Greens short of quota and elect both Patten and the Hinch candidate (this seems unlikely.)

[EDIT: As commenter hoddlegrid notes, and as others have confirmed on Twitter, the Victorian Socialists in this district are well above average "socialist" ballot clutter level, and seem to be running a well organised union-backed campaign with a lot of street presence and a reasonably high-profile candidate who has polled several percent in the lower house in the past.  The Socialists have poor preference flows but a strong result for them at least improves the chances of the Greens and Reason compared to Hinch Party; it seem the Socialists would need about 6% themselves for any chance of winning.  If the Greens and Patten do well it becomes more like 9%.  Anything in this range seems unlikely.  If the Hinch Party bombs out I also found chances for Animal Justice and Liberal Democrats].

* Northern Vic: My first attempt here got two of each major and a Liberal Democrat, with the latter off even 2.5% beating the Greens (8%) and Lib-3 (the same).   I had to knock the Liberal Democrat down to below 1% before they lost to Shooters, Fishers and Farmers.  Another thing to bear in mind here is that Hudson For Northern Victoria can win as well as the Lib Dems; I found he could do this off a primary of about 3%.

* SE Metro: Here I got a sane result of 3 Labor, 2 Liberal at the first attempt, and 2 Labor, 2 Liberal 1 Green was just about as easy to get.

[EDIT: David J in comments has found, and I have replicated, that if Tarang Chawla (Group D) can poll a primary of around 2%, Chawla can possibly take the third Labor seat.  Also, if the Transport Matters party can poll around 0.6-0.7%, Transport Matters can win the seat.]

* Southern Metro: On my first attempt I got an unlikely snowball with two of each major and Sustainable Australia (0.5%) beating the Greens (15.6%).  However, see note at the top (I had a crunch point where SA would need almost every preference to get over at least one major), and the snowball is very prone to collapsing early if the primary vote is low, or failing because of below the line votes.  So probably 2-2 with 1 Green was a more realistic outcome for the range of numbers I was looking at.  However if I give Sustainable Australia a non-trivial vote, over 1%, in some cases I get them up with the Greens and with Labor getting only one.

* Western Metro: Here on my first go I had 3 Labor, 1 Liberal and then Aussie Battler Party (0.5%) beating Green and Liberal-2 (each 10.5%).  If I knocked out Aussie Battler early then the Greens or in other simulations Shooters (even off 1.3%) won the seat.

* Western Victoria: The most dramatic one til last - I got a 2-2 major split with DHJP (2.5%) easily beating Greens (9.4%).  I tried cutting the Hinch candidate down to see how low I had to go to beat them and they didn't lose on the calculator until reduced to 0.32%.  [EDIT: It is also plausible for the Animal Justice Party to beat DHJP if AJP poll about 2% and get over Reason, then over Labor and hence over the Greens, though this path to victory is fragile.]

Overall the Greens seem to struggle - while North Metro might not realistically fall, their other three chances are all shaky.  Both the Liberal Democrats and Hinch Justice seem to have some very good flows, though a note of caution is required because if it is close and comes down to Below The Line votes, then the Liberal Democrats will tend to lose.  With varying degrees of likelihood I found micros possibly winning off c. 0.5% of the vote or even less in up to four seats.

These were just my experiments and others doing the same thing (with varying levels of idea what they are doing) are getting different possible outcomes.  Check the Poll Bludger thread, and also Tim Colebatch's Age piece (written before Antony Green's calculator was unveiled).  Also Antony Green has more voting advice - though I am not so sure that some of these specific disasters will actually happen as he is.  Feel free to leave exciting finds in comments!

Just Don't Vote Above The Line. For Anyone!

Voting below the line is easy in Victoria - you only need to vote 1-5 for candidates.  If you're a major party voter you can even vote for your own party's candidates and then stop if you like.  For a minor party voter, you might need to pick a few more.  Your vote won't be as powerful as if you keep on going, but even stopping at 5 is better than voting above the line.  If your vote exhausts at 5 it still doesn't help candidates with no actual voter support beat candidates with voter support, in the way that an above the line vote often will.  The best thing to do is to keep going, ranking any micro party you do not know either last or only above parties you utterly can't stand - especially if the micro has a gimmicky-sounding name like, oh, "Aussie Battler".  Numbering all the boxes below the line - if you have time - will make your vote most powerful and can never help a candidate further down on your list beat a candidate who you prefer.  (If you're short of time, feel free to leave out the second candidates from all the micro parties, since they'll never get elected anyway.)

Voting above the line instead of taking the effort to number a mere five boxes is an act of cowardice and shirking.  Willingly handing out how to vote cards that encourage voters to misdirect their preferences with a 1 above the line is on a par with giving cigarettes to children and then lighting them.  If parties are too lazy and self-interested to fix group ticket voting in the few states that still have it, maybe voters have to do it for them by making aiding and abetting preference harvesters socially unacceptable.  Just remember, everybody: friends don't let friends vote above the line.

Tuesday, November 13, 2018

Poll Roundup: Well That Wasn't Much Of A Honeymoon

2PP Aggregate: 54.8 to Labor (+0.8 since last week) by 2016 preferences
54.2 to Labor with One Nation adjustment
Labor would win election "held now" with a very large majority 



It's been a while since my last federal poll roundup.  At that time the Coalition's polling was recovering from the shock caused by the messy and (to the public) inexplicable coup that deposed Malcolm Turnbull as Prime Minister, and it was too soon to read anything into what we were seeing.  Because the Coalition's polling was in recovery mode but the new Prime Minister was still in a polling honeymoon period it was a matter of waiting for things to settle down to get a feeling for how competitive the Coalition really was.

On my aggregate, the recovery from a post-coup low of 43.9% peaked at 46.7% after seven weeks, and since then things have been getting worse rather than better.  Furthermore, since the defeat in Wentworth, they have been getting worse faster, at least if this week's shocker Newspoll is anything to go by.  The Coalition's current position is worse than at any time with Malcolm Turnbull as Prime Minister, and also worse than all but the worst few weeks under Tony Abbott.