By popular demand, I hereby open a new thread for discussion of the extraordinarily tight three-way race in the Victorian upper house region of Southern Metropolitan. Earlier expectations that the final seat would come down to a race between Labor’s Evan Thornley and the Greens’ Sue Pennicuik have been undone by an unexpectedly strong performance by the Liberals on postals, which has strengthened the hand of their third candidate David Southwick. Remarkably, the current result in quota terms is 3.00 for the Liberals, 1.99 for Labor and 1.00 for the Greens, making it a near-perfect three-way tie in the race for the final seat. The Greens have suffered the worst in late counting, such that the possibility has emerged of the Liberals winning the seat with a tiny surplus that helps elect Thornley, who will receive it as preferences ahead of the Greens’ Sue Pennicuik. The irony of Liberal preferences delivering Labor an upper house majority is being widely remarked upon, though their decision to put the Greens last always meant it was a serious possibility. Antony Green explains in comments that this is a rare occasion where below-the-line votes will prove decisive, so that "the models where you treat below-the-line votes as ticket votes" – such as the calculators at Upperhouse.info – "are too crude in such a close count":
What you need to do now is break the count into above and below the line votes. Add the tickets of Family First and the DLP to the Liberal vote. Add the Democrat ticket to the Greens, and People Power and Group C tickets to Labor’s vote. At this point, none of these three totals reaches a full quota, though the Greens are the closest. The balance is determined by the below the line votes. Unless the relative percentage of Labor, Liberal or Green increases against the other, none of these totals will reach a quota. The real unknown is what happens if enough BTL votes drift to the Greens. If this happens, then the Democrat ticket will elect the Green, and release a small number of ticket preferences for Labor … What Labor needs to win the last spot is for as many BTL votes to drift to the Greens before the Democrat ticket is distributed. Unless the relative %’s of the party change again, on the current count Labor will need a surplus from the Greens to win the last spot.
At stumps on the 10 day the result in South Met is even closer than the cricket.
Counting all votes as ticket votes, narrowly delivers the 4th seat to the Greens, and leaves the major party contest for the 5th seat indistiguishable to up to 3 decimal places. ie. literally a handful of votes.
Bring out the coin.
In Western VIC the DLP would have to suffer considerable net leakage of BTLs to be excluded by the NATs, and I suspect BTLs will likely favour the Greens over the ALP, which should deliver the 5th seat to the DLP!
Counting must be almost done @ 92.77%.
Yes.
There must be something screwy with the front page though. It reports that only 311 out of 393 centres have recorded votes. These numbers haven’t changed for days, but the votes have
And the winner IS the DLP. (usual caveat re BTL)
In SMET, there may be an “unusual situation” Calling the Group votes Ticket votes leads to the situation that, with one seat left to go, the ALP has .975 of a quota and LIB has 0.996 and nobody else in the race. In theory, the race would stop at that point and the Lib be declared the winner without the need to cut-up ALP. But, imagine the chagrin if they got cut up with 97.5% of a quota. They probably won’t get this close of course… they probably have more leakage than the Libs.
Cheers Melbcity, when I voted I wonder who he was, I hadn’t realised you placed it on here until I had posted.
I was reading bottom up.
Geoff, the website will still be updated. However, all the media files are generated by something complex which isn’t automatically updated to produce xml. The data entry program is a dedicated set of modules designed to validate the ballot paper. Situations where ballot papers are declared informal, or where the BTL is informal and the ballot defaults to the ticket if marked but only periodically updated to the main database recording primary votes by booth.
As I keep saying, the missing booths are the 7 E-centres in each district, most of which are zero or very tiny numbers of votes. They haven’t been through and marked them all as entered.
The 5:22pm count today is mixed news for Labor. Thornley slips back, but the Greens advance. With the Democrat ticket, the Greens are now 428 beyond the quota, the Liberals 1414 short and Labor 3912 short. If the Greens continue to improve, the Green surplus which is mostly Green and Democrat tickets will flow very heavily to Labor. And before MelbCity starts on about ‘segmentation’ again, that surplus is before the second bundle of Democrat votes is distributed. There are 4,895 BTL votes unallocated on this count. Still too close to call.
Western Victoria has firmed for the D.L.P. Where last night the DLP were 471 behind the Nationals at a key count, tonight they are 816 ahead. Which means the final count is DLP, ALP, GRN, with the third place determining the result. Last night Labor was 117 votes behind the Green, tonight 98 behind. I have about 3,000 BTL votes determining the outcome.
Based on BTLs being broadly similar to Tickets, at the moment the LC is beginning to look like:
ALP 19
LIB 16
NAT 2
GRN 2
DLP 1
I think it rather unlikely that the BTL drifts will change this much. A couple of races are close, but likely leakages will in most cases make them less close. If the BTLs change the cut-up, it will more likely be the order in which things happen, rather than the ultimate result.
[the XML and Web page agree right now]
Yes, but what about the vote to be counted, of which there is clearly some to come in South Metro. I wouldn’t call South metro yet. You’d need to know which lower house districts have yet to enter postal, absents etc.
Antony,
You may remember me as “Cinderella man” that engaged you in a lenthy (but helpful for me) discussion on this blog about the merits of OPV v’s CPV.
It would appear that the DLP will now take the final seat for Western VIC, which would be as “miraculous” as Steve Fieldings rise to the Senate chamber. Maybe the Catholic God is as powerful as the Evangelical’s God?
Given your advocacy for OPV, would you like to comment on its application in this election and the relative merits v’s CPV in the Senate.
Do you yet have a measure of the % that voted BTL here v’s that in the Senate poll?
If this system were applied to the Senate, would it have made any difference with respect to Fieldings election?
Would you prescribe a different version of OPV to that which applied in this election?
It would appear that few recommendations from the JSCEM were adopted, and the ones that were seemed only to favour the encumbant. Surprise!
Care to comment?
Either way, it would appear that the GVT is alive and well.
Ray,
Antony can better respond to your questions directed to him, but you can work out the proportions of BTL/ATL by referring to the VEC record of the count. There each party’s ATL vote is referred to as its “Group” figure, then the individual tally for each candidate represents the BTL (it’s impossible to vote BTL without giving a 1st preference for one of the 20+ candidates). For example in Eastern Metro, the (rounded) figures at this stage of the count are 364,000+ formal votes, just under 347,000 ATL, just under 18,000 BTL which is a little less than 5%.
You can do these calculations for each of the regions.
I advocated limiting the number of preferences a party could put on its lodged ticket. This was to encourage a party to put like minded parties on its ticket and not engage in strategic preference swaps.
I’d be happy to abolish ticket voting altogether, but not if voters then had to fill in preferences for every party, even if that was above the line preferences. You would increase the informal vote, and I still see no reason why voters, after voting for the candidates they know and want, must be corralled into expressing preferences for further candidates they neither know nor care for.
However, i’m not getting into a discussion on that again. All I’ll say is that if the DLP win, it is exactly the same as the victory for Family First, a construct of group ticket voting and compulsory preferential voting.
> Can someone explain why in Ferntree Gully they have the Libs as winning here…
“The VEC declared the Liberal candidate Nick Wakeling as winner of Ferntree Gully district by 17 votes after the final close of counting at 5pm on Monday afternoon. The losing ALP candidate Anne Eckstein has however asked for a second and final recount.”
“The VEC also declared Hastings and Kilsyth as Liberal gains with Gembrook, Forest Hill and Mount Waverley retained by the ALP.”
I don’t believe there is a need to cover the ground again because I believe that we both concluded that OPV would bias a result more to a minor party that achieved more than half a quota than one that achieves less than half, compared to CPV.
In this election I would be interested to see how many BTL votes exhausted, and what impact this has had in the determination of the couple close calls in WVIC and SMET.
Many would argue that an exhausted vote is a bad as an informal vote, and in the interest of democratic purity we would hope to minimise these.
I would be interested in your view on the take up (or not) of the JSCEM recommendations.
South metro – if you were a betting man. Taking into account
1 ATL votes
2 BTL expected votes
3 lack of knowledge about which postals and absentees have been counted
4 preference flows
5 Anything else of relevance
Who would you tip to gt the final 2 spots. I am going with the Greens and Lib’s in that order – anyone else care to make a tip?
The JSCEM recommendation not taken up was the abolition of group ticket voting. There were two reasons. The AEC were horrified about what would happen to informal voting. The political parties were horrified at what it would do to the size and design of their how to vote cards.
You are verballing me on optional preferential voting. I can’t think of an electoral system which isn’t biased in favour of parties that get more of the vote than parties that get fewer votes. If you get more votes, you are more likely to be elected. Ummmm, what’s wrong with that?
And in terms of ‘democratic purity’ some would view an exhausted ballot is as bad as an informal vote?!?!?! Give me a break. I’ve got better things to do than correct such silly statements.
Just to raise the question of the Myers vote again….:-(
In my simulation, I have factored him in as though he registered a conservative-leaning ticket with FFV>DLP>NAT>LIB>ALP>PPV>GRN. That seems to be the way most of these UNG chappies’ voters express preferences. If we swapped ALP and LIB it could make a difference. Ditto if many exhaust.
I suppose nobody has yet found out whether that is remotely true in SMET?
I would think there is a case for making the complete BTL “images” available for public scrutiny- at least after the election- to enable us simulators to get a better handle on some of the paths which BTLs can take. The usual cut-up print-out is only useful in this regard for showing where second preferences go. After that they become hidden inside other parcels.
What’s wrong with that !
An electoral system should have no bias at all !
One vote, one value and every vote counted.
We should do everything we can to maximise those principles.
Geoff,
When the “line” system was first introduced (for the 1984 Senate election, iirc) the rule was that all ticket preferences were displayed in the polling booth. I suppose that this was only of interest to a tiny minority of voters- like those of us who spend time on the poll bludger and similar sites, lol – it did at least mandate transparency. I expect that the procedure was dropped because it mattered only to the few, and those who might have been misled by the ALP-FF deal in 2004 for example would not have bothered to consult the display anyway.
As I understand it, this information has not been available to voters without their chasing it for 15-20 years. During that time, I only saw it published on the web either by the Electoral Commission or as it was during the recent election by Antony Green.
However, on 25 November, there was a notice on the enquiries table at my booth (so I assume that it was a standard instruction toelection-day staff) that the group tickets were available for inspection.
It was a prominent notice on the enquiries table, but would not have attracted the attention of most voters who were shepherded to the tables where the clerks were armed with he electoral roll and the ballot papers.
I think that above the line voting opens up the pandora’s box of murky preference deals – which favour parties who are prepared to deal with anyone (such at Liberal, Labor, DLP, People Power etc). There is zero public input to such deals, and often close to zero party input too – the deals are the provice of apparatchiks. Talk to any Liberal or Labor candidate – they will tell you that it is “out of their hands”. Now we see the DLP getting elected in a very small primary vote as a result, with about 95% of the electorate not even knowing who they are, let alone voting for them.
I don’t think this is good for democracy.
The ATL votes should all exhaust with the party selected – with no preference flows – as the voters are NOT expressing THEIR preference. Voters who wish to express their preference could still go BTL to do so.
Has anyone called the VEC re the situation in Ferntree Gully? Their website gives the seat to both Labor and Liberal at different places on the same page. The figures giving the seat to Labor seem to be more advanced, but who knows?
Geoff, from my research on past ballot papers, for which I was allowed access to the ballots, you can bet some of those ungrouped votes will leak to the nearest group on the ballot paper as preferences. Once they’ve pulled that end of the ballot paper to the middle of the voting booth, it seems voters are a bit inclined to stay at that end of the ballot paper for initial preferences. That is very evident from the (admittedly giant) ballot papers at the 1999 and 2003 NSW elections.
Peter Fuller says When the “line†system was first introduced (for the 1984 Senate election, iirc) the rule was that all ticket preferences were displayed in the polling booth.
No, I want to see every BTL ballot paper cast, not the Tickets. At the moment, the only people who see what has been written on these papers are the counters and the scrutineers….. and Antony*, whose observation above is fascinating. There was already a donkey and reverse donkey effect seen in the tablecloth ballot because of the way that voters draped it over the desk. Evidently it extends to the preferences they choose too (they had 283 to choose from)
Anyway….
New SMET primary figures have come in and guru Green has been further elevated one step in the peerage, because his suggestion that things could alter has been well and truly vindicated. The Greens now get a sufficiently high primary that, when they are elected, they push the ALP over the line, and FFV never get to be cut-up. Counting is now 87% complete here..
(*I did see some of these at Rosebery, when I was shown how the hail-storm damaged so many of them that they feared a new election might be called for. I was only allowed a peek.)
I too would like to see the GVTs abolished, but the cost to democracy would be an unacceptable rise in informality.
I just did a little experiment using my own home brewed calculator. Applying Antony’s suggestion of limiting preferences to say the first 5 from the GVTs in Western VIC, the DLP still get elected. So if the intention is to eliminate “perverse” results like this, then this will not work.
One must question whether there is anything intrinsically wrong with investing my vote with my party of preference, so as they can maximise the chances of getting their candidates elected.
Preferred 3 teir CPV system:
1. Vote 1 ATL exercises GVT; or
2. Number all groups ATL – preferences flow to candidates within group in order of list;
3. BTL – for those who want to order the candidates to their hearts desire.
-> Adam C re: FTG
The first VEC count gave the win to the ALP by 30 votes, they then did a recheck/recount which ended up with the Liberals ahead by 27
The Liberal is notionally the winner according to the VEC as no more votes will be counted – the ALP candidate has however asked for recount
Yes all credit to Antony, but let the record show that I picked the Thornley-on-Green-surplus scenario two days ago.
Today:
ALP 1.86872 + 0.08042 (PP) + 0.02214 (Rita) = 1.97128
Lib 2.79566 +0.12732 (FF) + 0.06622 (DLP) = 2.98920
Green 0.93386 + 0.10032 (AD) = 1.03418
ALP + Green surplus = 2.10546
Ray said
Ray: “One must question whether there is anything intrinsically wrong with investing my vote with my party of preference, so as they can maximise the chances of getting their candidates elected.”
Problem is, what YOU want out of preferences and what THEY want may be different. And what they want may be to influence who else gets elected. The cynics say that WVIC is a prime case in point, with the ALP preferencing the DLP in the shrewd knowledge that it might just come to fruition. The 1999 NSW LC election saw this funnelling take place on a monumental scale and was, of course, why their system was changed.
At any rate, the only way a party’s registered preference ticket can, by itself, enhance a party’s own chances is by creating a certain impression among the voters. This can be a strong effect and many voters vote according to their perceptions about “who they are giving their preferences to”, even in single-seat contests, where there is no such a thing as a Ticket.
On FTG….. some sort of a recount or recheck must have been going on. The booth totals have vanished from the XML and even some of the candidate totals have been taken away. I suspect the XML was posted right in the middle of the FTG people re-keying their data.
Correction to my table above:
Today:
ALP 1.86872 + 0.08042 (PP) + 0.02214 (Rita) = 1.97128
Lib 2.79566 +0.12732 (FF) + 0.06622 (DLP) = 2.98920
Green 0.93386 + 0.10032 (AD) = 1.03418
ALP + Green surplus = 2.00546
I am advised that there are 5,500 votes still to count. If they are absent votes, Thornley is alive. If they are postals, he is probably dead – but they are more likely to be absents.
Geoff,
I understand that it can be used for the purpose of determining who else the party might want to work with in parliament, and if you don’t like that choice you can go below the line (or preferably ATL per my option 2 if it were permitted).
If West VIC goes to the DLP, assuming ET gets up in SMET, then if the ALP want to legislate to the right, they have the DLP option. If they want to legislate to the left, they team with the Greens. Sounds like a smart choice, and much better than being a slave of the left with the Greens holding the BoP.
However, the party can maximise their candidates’ chances by preference swaps, which is what others seem to find so objectionable. The only way to eliminate this is to remove GVTs (Option 1) which may not lend itself to much more increased informality if the number of groups is similar to that in this election. Option 2 would be no more onerous than many lower house contest.
Removing GVTs will of itself limit the number of groups, as many are in it hopeful of winning a harvest lottery.
As I have shown by my experiment above, merely limiting the number of preferences a party can lodge on its ticket, may not change the result and exhaust a whole heap of votes.
Ray, you’re thought experiment implies you haven’t the faintest idea of what I’m talking about. You change the preferences rules to try and change party behaviour. So you engage in a thought experiment that is based on ceteris parabis. Get real.
You’re right. I haven’t the faintest idea of what you are talking about, especially when you speak in Latin.
Nor do I hope to hold a candle to your expertise in this area. But I would expect that you would not engage in the occasional put down comment to us mere mortals.
What my experiment does show however is that the parties would not have needed to change their behaviour, if the number of ticket preferences allowed where equivalent to the number of seats available. The result is unchanged under this scenario.
In fact I have extended the experiment to progressively restrict the number of ticket preferences. Only when we limit the number of allowable preferences to 2 does the result change, and all the votes for the Nationals and the DLP exhaust.
Is that a “bridge too far”?
When I look at the preference arrangements, I don’t detect any that were markedly across ideological lines.
Not only will the DLP and National vote exhaust, but also People Power, Country Aliance, Liberals and Family First ie. 47% of the votes will exhaust.
Yes. A bridge too far.
And yes the minor party that takes the cake (the Greens) has more than half a quota.
Do we have any conclusive E met and W vic results yet? If not does anyone know when we will?
Sorry I mean S met not E
No and no.
5pm figures are better for ET
Lib 2.77629 + 0.12782 (FF) + 0.06825 (DLP) = 2.97236
ALP 1.87202 + 0.0815 (PP) + 0.02290 (Rita) = 1.97642
Green 0.94423 + 0.10172 (AD) = 1.045595
ALP + Green surplus = 2.02237
ET’s lead is 1,317 votes
Antony… where are you…. whats the latest in Western Victoria? Is the DLP still in front and what about Northern Metro? Is the ALP in trouble from the DLP?
I think the one thing that may save the Libs in Sth Metro is the below the line votes of PP.
More than two thirds of PP votes are BTL. Mayne’s natural constituancy is small L Liberals. It will be very interesting to see what happens when the VEC press the button on the computer and BING we see where BTL votes go or don’t go in the case of votes that exhaust.
Fascinating to watch. It’s like watching a marathon race where everyone wants first and second but no one wants third.
I thought that the DLP will come up the middle. My analysis which I have broken down more shows it to be a very very close race in Southern Metro. The Greens havinbg topped the ALP surplus has handed the seat to the DLP who in turn had survived the NP and Lib surplus it was a close race.
Three are unlnown issues also with the segmenation and value inflated votes from teh paperbased formula/ I know Antony discounts this issue (Why I fail to understand maybe its as good as his model in predicting Mlebourne being a win for the Greens (The Senente based analysis is still holding its head above water and much more reliable then the 2002 single member profile)/ The key to teh DLP success was a range of factors much of which was first identified by the Speaker in his fortune telling calculator 🙂
MelbCity your posts are hard enough to understand as it is but if you’re to start being a bitch as well I won’t bother even trying to read them.
Sorry, long day of other committments. Western Victoria has no changes, except DLP is a further ahead of the Nationals at the key point. Breaking out the BTL votes, Labor remains about 100 behind the Greens, and my guess would be that the Greens will do better on the drift of BTL votes. So, unless the votes remaining to be counted lift the Labor primary vote compared to the Greens, then the totals are pointing to Labor preferences electing the DLP.
South Metro was a good day for the Greens which is a good day for Labor. Greens might be as many as 1,400 votes over the quota when the Democrat ticket goes out, that’s the first ‘segment’ that MelbCity keeps on about. Libs have slipped to be 2884 short of a quota, Labor 3826, but will then recieve almost all of the Green surplus, putting them around 2450 short. This is based on about 5,300 BTL votes sloshing around.
Adam.
I was scrutineering in Sandringham yesterday. BTl’s are favouring Sue Pennicuik. Some votes exhausting without going to any of the three candidates that matter. But interestingly John Lenders votes are going to Sue Pennicuik as much as thornley. Labor ticket and Liberal ticket (surprisingly) is leaking to the Greens lead candidate. Even DLP and FF BTLs are going to the Greens evenly. In Postals counted yesterday Libs were way ahead of the ALP (double). Greens were very respectable, but I think Thornley will find it difficult. Left to count postals and absentees from Brighton, Bentleigh and Sandringham. Thornley will find it difficult, If the Greens don’t make quota and surplus.
Gosh what will the ALP say about Greens preferences electing another Labor MP. Would they take all that tripe back?
On another note. What will the impact on Labor be if the DLP gets elected on their preferences? Family First at Federal election and DLP in State on Labor prefs. Poor Bronny will get a lashing in the Gay press.
Ray
How to understand Anthony Green’s Latin, all other things being equal 🙂
http://en.wikipedia.org/wiki/Ceteris_parabis
smiles
Disasterboy
Adam I can not reply if I do not uinderstand what you don’t understand. presumably it is the sengameatuion and rules of that apply. The system that is in place is not “proportional” and certianly not one vote one value. The method used to calculate te suplus distorts the value of some votes, Some majopr party ticket votes increase in valu e whilst the BTL secondary preferences increase in value. It bnecause the rules that is appolied is based on the number of ballot papers and not the value attached to the Ballot Paper. Whedn a surplus is transfered all votes carry the same “new” value and it does not matter if some votes (say 10000) were worth 0.1 and others (Say 1000) at 1.0 they all are transfered on at the same revised value…. Its basic maths..
The other issue iof teh segmentation rulkes which were orginal designed to minimise the distortion in the paperbased surplus calculation. (All related to the days of a manual count.)
Southern Metro
Based on the information Antony had listed above and I have not had a chance to review the latest data.. It would look like Labor has managed to secure a healthy lead. The main issue “The point I am bitching about” whether you agree or not is that teh VEC has not porvided anywhere near sufficent information in detail and in a timely fashion as to how many votes have been counted and how many and from where they are still to be counted.It is an auditors nightmare and there is a lack of transparency and accountability as a result. We live in te age of the Interent and more information shouldbe readily available. The VEC has an obligation not just to teh medai to meet their requirments but they have a greater obligation and responsibility to te public. This information should be readily available and opublsihed on the interent and that includes teh polling booth breakdwon of the Upeprhouse ballot. The only reason they parted form the practice of the past is that some on the media did not want this information. Thats fine they can ignore it if they want but the VEC’s obligation is beyond that of the media’s perceived requirements. Polling Booth data shoudlbe available and so should the BTL preference data that is data-entrered into a database.
Its basic really and I do not think it is that hard to grasp…
In Western Victoria, Antony, how far ahead do you now put DLP ahead of FF, and thereafter, DLP ahead of Nationals? It’s looking like these 2 cut off points will be critical.
Barry Cassidy on ABC radio in Sydney yesterday remarked on the possibility the DLP might win a seat in the Victorian Upper House.
I’m still supporting my namesake Mr Thornley(us Evans stick together).
Will there be a recount in Ferntree Gully? A 17 vote margin would seem to necessitate this, surely.
I think I predicted 55 seats for Bracks before the election, so I was right, at least for once.
Hello Dinesh
Do you have actual numbers to attach to the BTL observations?… i.e. number of ballots seen, % this, % that, etc. Such numbers for Dems, PPV and ALP would be very useful. On Tickets alone, ALP currently win 5th spot with a bit of room to spare. I currently have the Greens on a surplus of 2700, which is twice what Antony has (6pm figs perhaps?…. or it might be the dreaded segmentation!).
Just to liven things up, I might mention that it seems like The Greens have pulled narrowly ahead in the count for the last place in Western Metropolitan.
According to the most recent figures from the Victorian Electoral Commission, it looks like the lead Green candidate will be narrowly ahead of the third Labor candidate after distribution of Democrat and People Power preferences, whilst the second Liberal candidate will be short of a quota after DLP and Family First preferences. The upshot is that the Green will be elected ahead of the Liberal on Labor preferences. At least that’s the impression this Brisbane resident gets from inspecting the VEC web site.
Can people closer to the action in Melbourne shed any further light on what’s happening in Western Metropolitan?
There are a number of ironies in the situation:
(a) the Greens’ influence in the new parliament will be diminished rather than increased if Sue Pennicuik continues to gain as the count proceeds;
(b) Sue is a former AMWU and ACTU official, so the Greens have broken Labor’s monopoly on the election of union officials to parliament;
(c) if I’m right about Western Metropolitan, this will mean that 2 of the Greens’ 3 MLCs will be from the two most working class regions in the Upper House.
Another irony is that Colleen Hartland is a former cleaner in Parliament House – now that’s a career trajectory! But I have no idea if it really is looking like Labor will be running third when it’s down to Labor Liberal and Green for the 5th seat in Western Metro. On the figures it does look possible though.