Photo finishes: the Senate

A review of the Senate election results, to be updates as late counting progresses.

Progressive updates of the long and laborious Senate count.

Still in play

Western Australia

Monday 23/9. The ABC computer projection today flipped to show the last two seats going to Louise Pratt and Zhenya Wang at the expense of Scott Ludlam and Wayne Dropulich, the decisive change being that Australian Christians have fallen behind Shooters and Fishers at “Count 21”. This deprives Dropulich of the Shooters and Fishers preferences he needs to survive the subsequent counts.

Tuesday 17/9. The most excellent Senate modelling of PB regular Truth Seeker illustrates the delicate balance of the count here, and the stars that need to remain aligned if Wayne Dropulich of the Australian Sports Party is indeed to find his way to the Senate off 0.2% of the vote. Key to the outcome is Dropulich remaining ahead of the Rise Up Australia party after distribution of preferences from Australian Voice, after which his snowball builds all the way to a quota. This might yet be undone by a gentle trend towards RUA on late counting, together with the unknown quantity of below-the-line votes. Should Dropulich fall short, not only will his own seat instead go to Zhenya Wong of the Palmer United Party, but the complexion of the race for the final seat between Scott Ludlam of the Greens and Labor’s Louise Pratt will change. This is because the comfortable win presently projected for Ludlam is achieved off Palmer preferences, which won’t be available to him if the votes are used to elect Wong. Truth Seeker’s projection is that Pratt will “almost certainly” defeat Ludlam on a scenario in which Wong is elected.

Monday. Together with the Australian Motoring Enthusiast Party in Victoria, the other freak result being projected is that someone called Wayne Dropulich from something called the Australian Sports Party is projected to win off 0.22% of the primary vote. However, there are two points in the projected count where Dropulich narrowly escapes exclusion after finishing slightly ahead of the No Carbon Tax and Rise Up Australia parties. Those hurdles cleared, he harvests almost the entirety of the micro-party vote along with the Liberal Party surplus. If he drops out, it looks like another seat would be in the bag for the Palmer United Party, whose candidate is the little-known Zhenya Wang. The other point at issue is whether the second “left” seat goes to Labor’s number two candidate, incumbent Louise Pratt, or Greens Senator Scott Ludlam. If Zhenya Wang drops out, Ludlam looks the certain winner as he stands to receive the Palmer party’s preferences. But if the Palmer candidate is elected and has no preferences to spare, the result between Ludlam and Pratt at the final count becomes very close, though with Ludlam still appearing better placed.

Election night. One of a number of freakish outcomes currently projected by the ABC computer is that something called the Australian Sports Party wins a seat off 0.22% of the vote. However, there’s a very good chance that they will not in fact make it through the early rounds of the count. It does appear though that a micro-party seat is up for grabs to join the three for the Liberals and the two for Labor and/or Scott Ludlam. As best as I can tell, the only danger to Ludlam is that a share of the Palmer United Party vote might get used to elect one of the micro-parties candidates who are above the Greens on their preference order, which include the Liberal Democrats, the Australian Christians and Family First, perhaps depriving him of the preferences he needs to defeat Louise Pratt at the final count. It is also problematic for him that the PUP, being largely a phenomenon of the late campaign, may fade as pre-poll and postal votes are added.


Tuesday 24/9. The AEC yesterday announced that the computerised preference distribution will be conducted tomorrow, and the result declared on Thursday.

Tuesday 16/9. With two seats each for Labor and Liberal and one for the Greens assured, there are three scenarios for the final seat which could variously see it go to third Liberal candidate Sally Chandler, Jacqui Lambie of the Palmer United Party, or Canberra resident Robbie Swan of the Sex Party. The ABC computer is presently intriguing journalists by giving it to Swan, based on him finishing ahead of Labor at Count 21 by the grand total of 14,275 to 14,274. That sees Swan soak up the Labor/Greens surplus to finish ahead of the Liberal Democrats, whose preferences then put him ahead of Lambie and on to victory with Lambie’s preferences. But if the situation at Count 21 was just one vote different, Swan would be excluded and his preferences distributed in such a way as to leave Lambie about 1000 votes behind the Liberal Democrats, and thus be excluded. Lambie’s preferences would then flow to the Liberals and deliver the seat to Sally Chandler. A win for Lambie thus looks the least likely of the three possible scenarios, although the high rate of below-the-line voting in Tasmania is such that I don’t think it should be entirely ruled out.

Thursday. The ABC projection in Tasmania no longer has Jacqui Lambie of the Palmer United Party winning a seat, as she now finishes behind the Liberal Democrats (28,114 to 27,234) after the Greens surplus is distributed at Count 24. I earlier presumed that this scenario would deliver the seat to the Liberal Democrats, but Palmer preferences are in fact going to the Liberals and, unless Lambie can recover, will deliver the final seat to the third Liberal candidate, Sally Chandler.

Monday. Labor and the Greens have won over three quotas between them, securing two seats for Labor plus the re-election of Greens Senator Peter Whish-Wilson, despite a headlong plunge in the Greens’ statewide vote from the historic high of 2010. On the right, the Liberals have won their obligatory two seats, but the last is a close-run thing between the third Liberal, Palmer United Party candidate Jacqui Lambie (who unsuccessfully contested Liberal preselection for Braddon) and Clinton Mead of the Liberal Democrats. All three are between 9% and 10% at the second last exclusion, Mead having preference-harvested off a base of 2.29%, Lambie having received the Labor and Greens surplus, and the Australian Christians and Rise Up Australia feeding preferences to third Liberal candidate Sally Chandler. The high rate of below-the-line voting makes this particularly hard to pick.

Election night. Neither Labor nor Liberal appears to have the firepower to get a third member up, leaving Labor preferences to re-elect Peter Whish-Wilson. The last seat, I believe, is a toss-up between the Palmer United Party, currently projected to win the seat, and those pesky Liberal Democrats.


New South Wales

Monday. David Leyonhjelm of the Liberal Democratic Party is currently on 8.9% of the statewide vote, which is almost certainly an accidental consequence of his party being the first listed on the huge ballot paper, and hence the first most voters encountered with the word “Liberal” in its name. The party also receives the preferences of the Democratic Labour Party, which is on a not insubstantial 1.5%. The DLP too did handily on the ballot paper draw, securing the third position out of 44 groups, and similarly owes some of its vote to those who thought they were voting for the other Labor party. A further 0.4% was funnelled to the party by the snappily named Stop the Greens, Smokers Rights and Australian Republicans, whose links to the LDP have been reviewed by Crikey’s Andrew Crook . Other micro-parties feeding Leyonhjelm preferences, either due to their general hostility to larger parties or because they hoped to be the ultimate beneficiary of the preference network, include Katter’s Australian Party, Shooters & Fishers, the Fishing & Lifestyle Party, the Christian Democatic Party, One Nation, the Sex Party, Wikileaks, the Animal Justice Party, HEMP and the Drug Law Reform Party and the Stable Population Party. Not too many of these parties’ supporters would have cast their vote with the intention of electing a party that trades in Ron Paul-style low-tax libertarianism. When combined, Leyonhjelm emerges with a 14.3% quota with at least 4% to spare, and no need for any surplus from the major parties.

Election night. Notwithstanding a sadly typical flurry of excitement based on a meaningless early projection, Pauline Hanson’s chances have been negated by the phenomenon of an accidental 8.9% vote for the Liberal Democrats. This has probably secured a result of three Coalition, two Labor and one Liberal Democrats, although the Greens would be a chance of nabbing one of the Coalition seats if they or Labor improved in late counting for some reason.


Monday. Despite picking up a swing in Victoria, the Coalition has for the second election running failed to achieve a Senate vote sufficient to ensure a third seat. Last time this caused them to lose out on a third seat which instead went to the Democratic Labor Party. However, the DLP’s vote in Victoria was well down this time, mostly no doubt due to the proliferation of competition. Based on Antony Green’s calculator, Ricky Muir of the Australian Motoring Enthusiast Party stands to win the final spot, simply because the Coalition doesn’t have three quotas in its own right and very few parties are favouring it ahead of Muir, who fortuitously outperformed other micro-party candidates due to the way preferences were allocated. As far as I can see, the most likely scenario to thwart Muir involves him falling behind Australian Christians and the Australian Fishing and Lifestyle Party at a point where one of three must be excluded, though this would be easier to envision if Muir had to fall below two candidates rather than just one. If he does fall short, it appears that the last micro-party standing would be Family First, who would not match Muir’s preference firepower owing to the Sex Party (polling close to 2%) putting the Christian parties last. The upshot would be that Liberal incumbent Helen Kroger could get up after all, leaving the Abbott government with one less cross-bencher to worry about.

Election night. The ABC calculator currently projects a win for the Australian Motoring Enthusiasts Party off 0.52% of the vote together with two Liberal, two Labor and one Greens, but I wouldn’t stake the mortage on that. However, the Coalition has fallen short of a third quota and doesn’t get many preferences, helping to explain why the preselection stoush between second placed Scott Ryan and third placed Helen Kroger was so willing. In Kroger’s stead, the final seat would go to some or other right-wing micro-party. Unlikely to be in the hunt are the “DLP Democratic Labour Party”, whose vote from from 2.33% to (on current numbers) 0.69%. This was no doubt partly due to the greater competition for the micro-party vote, but I also suggest it was a bad idea to subtly rename themselves from “Democratic Labor Party (DLP) of Australia”, which I suggest was easier to confuse with the Australian Labor Party.


Monday. Here at least the result is both straightforward, with the Liberal National Party winning three seats, Labor two and the other going to Glenn Lazarus of the Palmer United Party, and inoffensive on a democratic level, since a) Lazarus’s 10.3% of the vote is near enough to a full quota in his own right, and b) people were clearly voting for him on purpose.

Election night. I think you can lock in the currently projected result of three Liberal National Party, two Labor and one Glenn Lazarus of the Palmer United Party looks locked in.

South Australia

Monday. Another of the many extraordinary results was that Nick Xenophon outpolled the Labor Party in South Australia, by 25.88% to 22.78%, and finished only slightly behind the Liberal Party on 26.69%. That looks certain to limit Labor to one seat, Don Farrell’s act of altruism in conceding the top position on the ticket to Penny Wong appearing more consequential than he probably realised at the time. That leaves a decisive Labor surplus to pass on to Sarah Hanson-Young, who had wrongly been written off by many who failed to consider the possibility of Xenophon sucking up enough votes to reduce Labor to one seat. Xenophon could probably have won a seat for his running mate if he had been more ready to engage in preference deals, but very few of the minor players have favoured his running mate over Liberal incumbent Simon Birmingham. The remaining seat looks set to go to Family First, whose candidate is housing tycoon and one-time Liberal candidate Bob Day. Day polled a strong 3.77%, and the only potential roadblock on his path to victory is that he finishes only slightly ahead of the Liberal Democrats at the point where they are excluded. South Australia’s six Senate seats thus look set to be divided between five different parties.

Election night. I may need a fresh pair of eyes on this one this morning, but I think the present projection of two Liberal and one each for Labor, Nick Xenophon, the Greens and Family First is the likely outcome. The one potential disturbance I see is Count 27, where the Liberal Democrats are excluded on 3.67% only just behind 4.03% for Family First. A reversal would, I presume, give the Liberal Democrats the seat instead. Despite Xenophon’s brilliant success in scoring 1.8 quotas, I don’t believe he can secure the remaining 0.2 quotas he needs to elect his running mate. Labor’s failure to secure a second seat, as looks to have happened in Western Australia in results without historical precedent, looks likely re-elect Sarah Hanson-Young. Now recall that the Labor Party initially proposed to give Don Farrell the only winning spot on the ticket at the expense of Penny Wong.


Monday. The territories have never failed to deliver one seat each to the major parties, but Greens candidate Simon Sheikh has at the very least come extremely close to knocking off Liberal candidate Zed Seselja, the former ACT Opposition Leader making a bid for federal parliament after knocking off incumbent Gary Humphries for preselection. In the Northern Territory, there was talk that Nova Peris’s path to the Senate might be blocked by the tightness of preference flows to the indigenous rights party First Nations, but their vote was too low to put them in contention.

Election night. The Greens are making their best fist yet of winning a Senate seat in the ACT, potentially thwarting Zed Seselja’s bid to move from territory to federal politics. However, the ABC projects Zeselja holding by 34.05% to 32.62% at the final count, and my instinct is that that’s unlikely to be overturned on late counting. In the Northern Territory, the First Nations party doesn’t look like it’s going to survive the early stages of the count, thwarting a potential threat to Nova Peris’s election, which probably would have been averted anyway by a near-quota vote for Labor.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

191 comments on “Photo finishes: the Senate”

Comments Page 2 of 4
1 2 3 4
  1. Correction on count 27 in SA. If the Liberal Democrats get ahead of Family First, the FF votes go to No Carbon Tax, and the Lib Dem votes follow soon after. At that point NCT gets elected ahead of X’s number two by Green (!!!) preferences.

  2. Socrates @8 said “I think the AEC also deserves some criticism. They are permitting parties to register with confusing names, and misleading names.”

    It is not the AEC’s fault – they have to apply the law. The issue of whether or not a party having ‘Liberal’ in their name is confusing as resolved back in 2001 when the Adminstrative Appeals Tribunal overturned a decision by the AEC not to register a party called ‘Liberals for Forests’.

    Similarly, an appeal to the AAT against a decision by the AEC to register a party called the ‘Fishing & Lifestyle Party’, despite there being an existing party called the Fishing Party.

    Until the Electoral Act is changed, the AEC will keep registering such parties.

    I’m not specifically singling out the Liberal Democrats. There are two parties with Uniting in their name, there is still the old Australian Democrats as well as the Democratic Labor Party & the Christian Democratic Party. There is a party called Stop the Greens as well as the Greens. Not surprisingly there are over 10 parties with the word Australian in their name.

    If the law was changed, it will be a challenge to decide where to draw the line. But with so many registered partues, the resultant very long Senate ballot papers is likely to mean greater confusion.

    I think paying more attention to the rules enabling a party (or in effect just a party name, as many of these new ‘parties’ are little more than a name in reality) to be registered, and changing the Senate voting system to that now used in NSW in their Upper House is the better way to go.

  3. Yes, I mentioned that earlier although I did not work out whether it was NCTCS or Xenophon number 2 Stirling Griff that won.
    However the trend in the vote count last night was for the Lib Dem vote to be dropping and they were getting further behind Family First.

  4. Did I actually read this right? In Antony Green’s Running of the Counts (SA Senate version) on the ABC, after the election of the Greens’ Sarah Hanson-Young by preferences from the ALP’s eliminated Don Farrell, we see the following (bearing in mind that there are now only 3 players left in the game: Liberals, Nick Xenophon’s #2 and Family First):
    9,990 (1.28%) votes (55,097 ballot papers at 0.1813 transfer value) originally from Australian Greens distributed to Family First (Bob DAY) via preference 54.
    32,138 (4.13%) votes (177,246 ballot papers at 0.1813 transfer value) originally from Australian Labor Party distributed to Family First (Bob DAY) via preference 7.

    That’s right, both the Greens and the ALP chose to elect a Family First senator over a 2nd Nick Xenophon rep. Because Family First’s commitment to, e.g., the environment and climate issues is so much more philosophically similar to the Greens and ALP’s, and not influenced by denialist nutbags at all:

    Those views scored the nod ahead of the ticket of a senator who, regardless of whether his particular scheme is loved or loathed, clearly recognises that there is an issue and doesn’t advocate ‘denying and/or doing nothing’ as a policy response:

    So the strategic geniuses in SA Greens and the ALP have very likely taken a huge step to putting in place a senate that will pass Abbott’s laws abolishing both a carbon tax and ETS. Oh, they’re so clever with their preference swaps— they will just harvest, harvest, harvest and never have to give anything back, right?

    ALP and Green reps: don’t whinge about how micro-parties have vote harvested with each other. Just strongly advocate & pursue changing the rules & laws to put power back in the hands of voters (optional preferential ATL & BTL) and out of the hands of backroom dealers. As you should have done when you were put on notice of this issue around a decade ago. I’m sure change won’t be too fast coming, giving you plenty of time to reflect on how you handed power to minor- and micro-parties who are trenchantly opposed to your philosophies.

  5. SimonH,
    Labor put Nick Xenophon at preference 10 while they put Stirling Griff his running mate at preference 49, with Family First and Liberals in between the two of them. So it seems the Labor party has something against Stirling Griff personally.

  6. I really don’t understand the intricate niceties of the Senate voting sysytem and given that I would never get that part of my life back I’m not going to spend much time studying it.

    However, perhaps someone with better understanding than me might be able to answer this:

    If the parties actually put other candidates in the order of their real political preference for them (ie how closely their policies matched), rather than trying to play “strategic” preferences, would much of this drama be eliminated?

  7. With reference to Joe Bullock and Lousie Pratt on the WA senate ticket and Paul Howes rumors my much better half noted seconds ago “it is almost like they don’t care who runs the country so long as they get their seat.”

    If they put Shorten into the leadership or Howes into the senate I don’t think anyone could disagree.

  8. A question for anyone with process or past knowledge:

    How many below the line votes are there still to be counted, and what is the chance of these affecting results? For example, Sinodinis has ~1.1 quotas and the Greens ~0.9 in NSW. Any chance of movement here?

  9. Re the WA senate counting it is noted by several posters that there are small differences which could see the Sports Party eliminated early. However small differences between parties that poll only small amounts of votes is significant. “Truthseeker” models possible outcomes and works on variation of 15% of the votes for microparties and 4% for large parties. 10% of 2000 votes is 200 and that is a small difference for the Liberals and Labor which they would easily catch up but it is a very big and probably insurmountable difference between two microparties. The only reason for those sort of changes to occur between microparties is if an error is found on checking ballot papers or if there is a unusual pocket of supporters already counted or yet to be counted.

    So I expect early exclusions to stand and that changes will occur at later stages of the count as postals boost the major parties.

  10. God, the Senate preferencing system is a joke.

    Totally and utterly undemocratic, and leads to absurd results where people voting for one party (e.g. Green) might be the ones who get the party they would least like to see in power (e.g. No Carbon Tax/Family First) over the line.

    The situation in SA is very poor, and reflects poorly on both the Greens and Xenophon who should have found a way to help one another rather than benefitting the LNP and various right wing loonies.

    The whole system needs urgent reform. Above the line preferencing sounds like a good idea.

  11. “Liberal Senate candidate Zed Seselja has a Senate quota in his own right after further counting of votes on Monday.
    The Australian Electoral Commission says it has now counted almost 200,000 of the votes cast for the ACT Senate race, or three quarters of those enrolled.
    On Tuesday afternoon Mr Seselja reached 33.37 per cent of the primary votes counted so far, just above the required one third quota.

    Read more:

  12. Western Australia Preference flow

    There are a lot of pints were the preference flow can be broken as the count progresses.

    AlphaCode, Group Name, GroupPref
    A, Smokers Rights, LDP, SPRT, PUP, ALP, GRN
    B, Liberal Democrats, LDP, SPRT, PUP, ALP, GRN
    C, Australian Christians, SPRT, PUP, ALP, LDP, GRN
    D, Help End Marijuana Prohibition (HEMP) Party, SPRT, LDP, GRN, ALP, PUP
    E, Socialist Equality Party, GRN, SPRT, ALP, LDP, PUP
    E, Socialist Equality Party, ALP, LDP, PUP, GRN, SPRT
    E, Socialist Equality Party, LDP, PUP, GRN, SPRT, ALP
    F, Palmer United Party, PUP, LDP, GRN, SPRT, ALP
    G, Shooters and Fishers, SPRT, LDP, PUP, ALP, GRN
    H, Australian Voice Party, SPRT, PUP, LDP, ALP, GRN
    I, Sex Party, SPRT, LDP, GRN, ALP, PUP
    J, Secular Party of Australia, GRN, LDP, ALP, PUP, SPRT
    K, Australian Independents, SPRT, LDP, PUP, GRN, ALP
    L, The Wikileaks Party, SPRT, GRN, LDP, ALP, PUP
    M, Katter’s Australian Party, PUP, SPRT, LDP, GRN, ALP
    N, Family First Party, SPRT, LDP, PUP, ALP, GRN
    O, No Carbon Tax Climate Sceptics, SPRT, LDP, PUP, ALP, GRN
    P, Stable Population Party, SPRT, PUP, LDP, GRN, ALP
    P, Stable Population Party, SPRT, PUP, LDP, ALP, GRN
    P, Stable Population Party, SPRT, PUP, LDP, GRN, ALP
    Q, Stop The Greens, LDP, SPRT, PUP, ALP, GRN
    R, Australian Democrats, LDP, SPRT, GRN, PUP, ALP
    R, Australian Democrats, LDP, SPRT, GRN, PUP, ALP
    S, The Greens (WA), GRN, SPRT, ALP, PUP, LDP
    T, Animal Justice Party, SPRT, GRN, PUP, LDP, ALP
    U, The Nationals, LDP, PUP, SPRT, ALP, GRN
    V, Australian Fishing and Lifestyle Party, SPRT, LDP, PUP, ALP, GRN
    W, Australian Motoring Enthusiast Party, SPRT, LDP, PUP, ALP, GRN
    W, Australian Motoring Enthusiast Party, SPRT, LDP, PUP, ALP, GRN
    X, Australian Sports Party, SPRT, PUP, LDP, GRN, ALP
    Y, Rise Up Australia Party, SPRT, PUP, ALP, LDP, GRN
    Z, Australian Labor Party, ALP, GRN, SPRT, PUP, LDP
    AA, Liberal, SPRT, PUP, LDP, ALP, GRN

  13. The AEC MUST publish copies of the BTL preference data-entry data-file and provide copies progressively during the count. Without access to this information it is impossible to verify the integrity of the computer count

    It’s akin to shopping at the supermarket and not being given a receipt and running total as goods are scanned.

    There is no technical or legislative reason why this information can not be published on-line and in real time.

  14. the other factors that need review is

    1. the calculation of the surplus transfer value: It must be based on the value of the ballot not the number of ballot papers.

    2. The method of distribution of excluded candidates: The count should be reset and restarted on every exclusion. This would correctly reflect the voters intentions without skipping or jumping candidates. Every full Value vote must be allocated to the next available continuing candidate

    3. remainders should be retained: With a computer count this is possible

    4. Voters should be able to preference above the line

    5. Deposits should be increased to $5,000 per candidate: refunds issued for every 4% of the vote received or candidate is elected.

    6. the Droop quota should be abolished; Why divide the cake by X/(y+ 1) only tho throw away a slice, The quota should be x/y – Pure proportional representation


  15. Lenxyz@60

    Thanks for taking an interest in my blog. I use a higher variance partially to allow for the greater likelihood of such “pockets” being found, but also to allow for the fact that these micros may have a significant % of btl votes.

    I acknowledge it’s not perfect, but it’s also designed to avoid a “false” 100% reading – No chance of getting elected with a 100% outcome by my model when my variance is higher than actual.

    I am currently working on a theoretical model to accurately portray btl votes. Keep watching my blog for progress of this!

    Also, daily around 10-11pm, I will update my blog with most current AEC senate numbers. Tonight I also plan a small piece outlining how the GRN candidate was not as close as everyone thought.

    Follow me on Twitter @AU_Truth_Seeker for notification of my new blog posts.


    300-400 VOTES added to either TCS FFP or RUA and SPRT loses out and Palmer elected.

    If the ALP increases it’s percentage to above 27-28% then the Greens lose out

    It only takes a misplaced bundle of votes allocated to the wrong pile or postal.Absentee votes to shift the balance and the order of election changes in WA

    If both ALP and LNP Votes increase then this could also tip the Greens out of the last position

    All parties will be wanting to monitor the WA preference count and review/recount the above the line votes and examine informals

  17. Andrew Bartlett 53

    Thanks for the clarification and in that case, yes I should have blamed the Admin Affairs Tribunal, or the legislation itself, and not the AEC.

    In that case I agree with your point about the need for taking more care in the registration of parties as parties. There is an obvious advantage in getting above the line on Senate voting papers with a column of your own, compared to ungrouped independents. There need to be specific criteria as to what constitutes a political party, rather than just a group of like minded individuals.

  18. Michael Kroger needs to be corrected, he blasted the AEC on QandA tonight about the Liberal Democrats name confusion and he was totally incorrect. He also said that the 8 percent in NSW was against 3% in other states, but in other states they also benefited by name confusion as well I am sure. In SA it probably has meant we have a FF senator instead of NCTCS senator.

  19. I have updated my Monte Carlo analysis – with very few votes counted there’s very little change in likely elections. Just a small drop in PUP vote in TAS and hence a small drop in PUP likelihood of election.

    However, it may all be academic because as Kevin Bonham points out LDP receive more prefs than others after GRN election so more exposed to BTL leakage anyway, more than likely locking in PUP’s Jackie Lambie.

    Details on blog

  20. Truth your model is way out.

    The problem with the last seat is the Droop quota. There is no justification to maintain it under a proportional system.

    We should have a pure proportional system.

    Most likely the ALP and the LNP will agree on implementing a representation threshold

    The Threshold will have to apply to groups not individual candidates, And If implemented I would suggest that it be 4%. What most people forget is that the LNP and ALP 3rd candidate has less than 1000 primary votes and no one complains about them getting elected.

    Tye distortion in the way the vote is counted also is a factor. If you look at the count the main parties vote increased disproportionally in value as a result of the formula used to calculate the surplus transfer value.

    You have 50,000 votes at a value of one and 500000 votes at a fraction value all being transferred at the same calculated value of say 0,25. teh 500,000 vptes juts went up in value

    Add to this the distortion in the distribution of excluded vote preferences where a vote skips candidates already elected earlier in the count the preferential system again produces errors in the count. In Queensland 2007 this prevented the Greens from being elected. The easiest way to test this is to recount the Queensland 2007 Senate excluding all candidates except the last seven standing 3 ALP, 3 LNP and 1 GRN.

    IN order to give reflection to the voters preference the cont should be reset and restarted ion every exclusion. This way a full value vote would always be allocated to a continuing candidate as if the excluded candidate(s) had not stood.

    A reiterative count also would provide allow the quota to be adjusted on each iteration taking into account any votes that exhaust should an Optional Preferential voting be adopted. Optional preferential voting has unintended consequences as we saw in Western Metropolitan where the Holy Trinity elected the Greens candidate by default.

    If we are to adopt a representation threshold then a reiterative counting becomes even more so important as the votes for minor candidates would need to be distributed before any surpluses were calculated. If you apply one rule to one class of votes then you should apply the same to all other votes

    See Wright System for an explanation of a reiterative count. It is a weighted proportional system that resets and restarts the count in each exclusion. reiterative counts are not new and were previously used in Victorian State Local Governments. The main reason they were not considered is that in a manual count it was time prohibitive. With a computerised counting system this is no longer a limitation

    we need to also consider more the use of the Droop Quota (x/(y+1)) This locks up votes in the count and denies up to 14% of constituents a say in who should represent them., There is no overriding justification to use the droop quota. Again in a manual count it allowed for the result to be obtained at an earlier stage without having to distribute all preferences this is not an issue with a computerised electronic counting system

    Preferential voting above the line

    Yes I like this opt6ion and had previously advocated it m my submissions to the JSEM

    Above the line voting was not just implements to give expression to party group voting tickets but to easy t6eh manual counting system. Unless we adopt a online voting system above the line preferential voting would take a long time to data-enter in the results. if we adopt a online voter data-entry process then this would not be an issue but additional problems begin to rise in relation to electronic vote security

    As it stand the AEC refuses to provide copies of the Below-the-Line Preference data-file. Without access to this there is n o way a scrutineer can monitor or verify the integrity of the count. There are errors in the data-entry transcription process that are not picked up. They are small in number bit in a close election can make a difference. It is also possible that a person who has access to this data-file count manipulate it in such a way that it would not be detected. By publishing copies of the data-entry file in real time this allows for the data stream to be monitored for any unauthorised changes.

    It also allows scrutineers to highlight potential tipping points and allows more attention and review of those votes that could change the outcome of the election and aa such subject to greater scrutiny during the data-entry process.

    The AEC in not providing access to copies of the data-file as the count progresses undermines confidence in the election process. There is no justification or technical reason why this data can not be published in real time as is the case with the lower house count.

    The AEC has a legal and moral obligation to provide copies of this file to scrutineers on request. But they don’t.

  21. Truthseeker,

    Your model shows a 7% chance the Democrats can win the last spot in NSW. This number has improved since yesterday. Are they the only party which can overtake the Libs for the last spot?

  22. Benji @72, that made my eyes pop out of my head too. That said, as an ex-Dem who is embarrassed by what the party has become, I would not know whether to laugh or smash something if they pulled off the upset. But I feel sure they won’t (although Sinodinos was 110 on my BTL ballot, which I expect to end up with Cate Faehrmann).

  23. edward o, yes the Dems are riven by infighting. Still cant be any worse than the Motoring Party or the Sports Party! I suppose it depends if they can get ahead of the Animal Justice Party at Count 34. They should then harvest prefs off that Party plus Motorists etc. They are only 35 votes behind at this stage.

  24. My model generates a sample of 1000 Monte Carlo simulations. Accordingly, minute change from one sample to the next is normal and expected – the same as polling for PUP might be 7% one day, 5% the next, 8% the next.

    What the model shows is that where the vote of the parties is modified slightly, in 7% of such modifications DEM wins. In practice, my model has two degrees of inaccuracies:
    1. No recognition of BTL votes. This means my model would overestimate the likelihood of a party that relies on harvesting.
    2. My model variation of each party’s vote is wider than actual variation. This is deliberate and partially offsets the #1 issue.

    7% however is not much chance. I’m working on how to model BTL votes and will try to get this model up and running this week (no guarantees though! 🙂

  25. I was thinking that I couldn’t imagine anyone being both switched on enough to vote BTL but switched off enough to actually vote Democrat in 2013, but I in fact know one or two people who did.

  26. I know, I was just being a bit brief. I don’t know anyone that gave them a [1], but I do know some people who proudly showed me their BTL ballot from or BelowTheLine or whatnot which had them in the top ten behind, say, Secular Party or Drug Law Reform.

  27. Pancho@59

    A question for anyone with process or past knowledge:

    How many below the line votes are there still to be counted, and what is the chance of these affecting results? For example, Sinodinis has ~1.1 quotas and the Greens ~0.9 in NSW. Any chance of movement here?

    There is a common confusion here. BTLs and ATLs are all included in the original count together as unallocated ticket votes. It is not the case, as many imagine, that ATLs are counted first then BTLs added. So a party that is good at getting BTLs does not see its primary vote rise in late counting.

    When considering the impact of BTLs it is necessary to look at who is most exposed to BTLs. Typically a party that has to snowball up from a very low vote total is more exposed, especially if its preference sources are parties with normally high BTL voting rates. A party that has most of a quota to begin with is much less exposed to BTLs and more likely to benefit from them.

    In NSW I would think there were so few BTLs that this wouldn’t affect a 0.9 vs 1.1 contest. But what it does sometimes do is change the exclusion order and produce a different contest altogether.

  28. Make that 2 votes now. Does anyone know what could happen if the Democrats overtake that AJP at that stage of the count? Or is it just too much speculation to bother?

  29. PUP are now very close to notionally dropping below the Lib Dems in Tasmania on the calculator. This is an artificial result because the Lib Dems are snowballing upwards from a low vote and will be hurt by the c.11% BTL Tas vote which tends to be higher among their feeder micros. But if the post-count sees further softening in PUP’s position they might not win.

  30. The AEC should publish the BTL preference data file. The process of distribution and the calculation of the Surplus Transfer file can and does effect the outcome of the election. In Victoria 2007 the value of distortion in the calculation of the Surplus Transfer Value represents 70,000 votes. The segmentation distribution also distorts the outcome in Queensland 2007 this elected the wrong Senator

    The AEC is still in the process of distributing.sorting the number of ballot papers from above the line to below the line papers. It is unclear if they will presort the votes into primary allocations. At present they are in groups. IN most cases the BTL votes follow teh Above the line tickets. How ever for the micro parties this might not be the case. The Holly Trinityy is the key to the microparty stack climb At present they are 250 to 300 votes behind Sport Party in WA., If the number of BTL votes or prepoll/Postals bridges that gap then the results in WA will change as Sports party will be excluded before they benefit for the Holly Tinity alliance distribution.

    The media should be asking why the Holy Trinity entered into such an unholly alliance with other microparties preference swap

  31. Good evening fellow bludgers,

    I have done my daily after-work run of my Monte Carlo simulators and have two interesting results:
    1. NSW: Sinodinos (LNP3) now only winning 80% of scenarios, down from 85% yesterday
    2. Tas: PUP win likelihood decreasing daily, only picking up 3% of new votes counted. I consider PUP much less likely to win now and this low-new-vote trend appears set to continue.

    It’s all outlined in quantitative glory on my blog, which I’m sure you’ve all got bookmarked now. 🙂

  32. I have analysed Tasmania in great detail here:

    Although PUP’s position is weakening and they will soon cease showing as the winner on the calculator, I find that the two scenarios that threaten them are both very likely to be undermined by BTLs, even at the end point of a full projection of their current slide.

    On that basis I think that Lambie will soon lose the lead on the calculator, but win anyway.

  33. The one thing I have not really considered is whether BTLs causing a different order of exclusion of micro parties can make any difference.

    I am thinking that it would have shown up as a non-zero chance on your Monte Carlos if there was anything in it.

  34. My hunch would be to agree. As of Monday, I slightly lowered my vote variations. But even the higher variations I used on Sunday showed SXP, SFP, PUP, FF, LIB as the only likely electees… the “3-party preferred” at the moment has only:
    as the micros that do not support LDP.

    In order to change the penultimate elimination, you’d need to find an even greater concentration of micros to overtake PUP at this point. Alternatively, you’d need to find a candidate on the ALP GVT between PUP and GRN. Such tickets are (in order)
    The first three of which polled reasonably… But still unlikely.

  35. Yes I got a screenshot when that first happened.

    PUP got a nasty whack on the last trickle of ordinaries so I’ve downgraded them from “very likely” to “likely”, out of caution.

    FF were not that far off winning. Today at one stage I mistakenly entered NCTCS as having .58% instead of .058% on my projection. The calculator “elected” Family First based on that error.

  36. I think it’s time for a little critical analysis of the LDP vote in NSW and the combined influence of name confusion + ballot paper position.

    Ignoring Vic (no GVT submitted:), let’s consider the relative ticket positions of the LDP and LIB or LIB/NAT:

    NSW: Strong advantage to LDP – first in ticket, LIB/NAT buried in middle, tiny fonts.
    QLD: Strong disadvantage to LDP, almost opposite to NSW, Lib/Nat second and LDP buried in middle
    WA and SA: Moderate advantage to LDP (2nd or 5th), maybe offset by LIB being right at other end
    TAS: Moderate advantage to LDP (2nd) over LIB (5th), perhaps reduced by lower number of groups / larger font size.

    If we accept the above as reasonable… then consider this:

    NSW (massive LDP visual advantage):
    LDP at 8.9%, LIB + NAT House vote = 47.2%, LIB/NAT Senate vote 34.6%, incredible 12.6% difference between House and Senate. (I needed to add the NAT house vote as they run a joint ticket in NSW).

    QLD (massive LDP visual disadvantage):
    LDP at 0.7%, LNP House 45.3%, LNP Senate 40.7%, House – Senate 4.6%.

    WA, SA, TAS (moderate LDP visual advantage):
    LDP at 3.4, 3.5 and 2.3% respectively.

    Best comparison is QLD vs NSW.

    a) The LDP had a fantastic campaign in NSW and achieved a much higher vote than in neighbouring QLD due partly to the massive campaigning efforts of party officer David Leyjonhelm
    b) I’m a purple hippopotamus.

  37. Labor’s taken a whack on ordinary votes of about 500 votes in Tasmania. Not sure where it came from. I’ve also found an error in my modelling worth about 600 – this is a trap for beginners; don’t try to model out-of-electorate prepolls off within-electorate PPVCs. Both of these disadvantage Labor at the point where the Sex Party could eliminate them and snowball to victory. They also both disadvantage PUP at the point where PUP can be excluded by the Lib Dems. Combined these issues make the last seat in Tasmania still very unclear between at least three parties.

  38. I’ve just posted and tweeted a correction to my above comment. Sports Party Dropulich now has an estimated lead of just 90 to 170 votes. This is approximately 0.006% of the ~1.5m West Australians who will be voting.

  39. I’ve updated my Tas article to include the remote scenario of Family First still getting up. It showed up as 0.3% on truth seeker’s simulation today; I think that given Sex Party is dropping on postals and ALP will drop on prepolls, it’s higher than that. Still quite remote. They’re 770 shy on the calculator of jumping Sex Party, probably closer in reality, and will get closer still. If they manage that, they might jump ALP as well (but probably won’t). If they get through both they’re probably home.

    Absolute mess of a post-count to model. Yesterday my article was giving people who read it a headache and now it’s giving me one!

  40. Am I reading it wrong, or is Helen Kroger likely to get up on primary vote alone?

    There are over 500,000 votes still to count, and the Libs are doing very well as postal and pre-polling come in.

Comments are closed.

Comments Page 2 of 4
1 2 3 4