Tuesday, 24 March 2015

New South Wales Rush

Contrary to how I would have wished, we have only a few days until the NSW state election and I have yet to post anything substantial on the subject. So this is going to be another dense, intense and something-else-ense post.

Firstly the baseline. Original margins courtesy of Anthony Green, as per usual and adjusted for recent electoral boundary shifts etc. The swing is calculated at 8% away from the Coalition based on the 2PP result in 2011 (64.2:35.8 favouring the Coalition) and the latest polling from Roy Morgan (up to March 23, 56:44 favouring the Coalition).

This gives the Coalition 55 seats (LIB 38, NAT 17) and the ALP 34 with four tossups. Since there are 93 seats, the maximum tossups allowed on this blog are (93/20 = 4.65, rounded down to) four.
Now, after our previous success beating the benchmark based on past voting history, we’ll be playing around with those methods a little this time.

Firstly, because it’s not called the Infographinomicon for nothing, here’s a chart of the historical incumbency of all current seats (including past incarnations) since the abolition of multi-member seats in 1927:

And here’s a chart of the results for all elections since that date:

In both cases there are some strong supporters of the Liberal (e.g. Vaucluse), Labor (e.g. Lakemba) and National parties (e.g. Upper Hunter), which is a promising start for a system of prediction based on seat consistency.

Firstly, let’s try to quantify how consistently each seat supports a given party. In the below chart, the parties are simplified to Labor or Coalition (Liberal + National). Each time the seat voted for the currently ruling Coalition parties, we add 1 to their total. Each time they vote Labor we subtract 1. Independents and minor parties contribute nothing positive or negative to the total. This produces a very primitive profile in the total column where 0 is neutral (voting for the Coalition as often as the ALP), positive numbers indicate the strength of support for the Coalition and negative numbers indicate the strength of support for the Labor party.

However, since 1927 the Labor party has son more elections that the Coalition. This means any seat that has a 50-50 2PP history is actually slightly Coalition leaning by comparison with the state average. And, since polling is averaged across the state, this needs to be taken into account. For this reason the STATE column contains the value the seat would have held if it had voted with the rest of the state at each election during its period(s) of existence. The difference, in the DIF column, is simple calculated by subtracting the STATE benchmark from the seat’s TOTAL, giving the seat’s position relative to the general public.

This number can also be reached by doubling the INDEX. The INDEX is calculated in the following way: starting at 0 add one each time the seat voted for the Coalition against the trend and subtracting one when it voted for the ALP against the trend. The DIF is double the INDEX because voting, for example, for the Liberal Party when the state was predominantly Labor actually increases the DIF margin by two – the seat’s total increases by one while the state’s total decreases by one.

However, these values don’t take the length of time the seat has existed into account. Epping and Monaro both have an index of 3, having voted for the Coalition on three occasions more than for Labor when voting against the trend. However, because Epping has only contested four elections, this equates to a history of entirely Coalition victories, while Monaro has a much more evenly split history dating back to 1927.

To combat this, the next table provides the index, the maximum (or minimum, for Labor leaning seats) value the index could have reached, and the percentage of times the seat favoured a party against the trend (INDEX/MAX x 100%).

A PERCENT rating of 100% means the seat is a bastion that always votes for its preferred party, regardless of any other factor. This value can be diminished by voting for the “other” party, and this diminishing effect will be more significant if voting contrary to the general population. Thus a rating of 0% means either the seat always voted in line with the public (bellwether) or voted against the trend for one party as much as the other.

Also, although this is already captured in the PERCENT, the CONTRA rating is the number of times a seat voted against its “preferred” party contrary to the trend.

The PERCENT rating allows us to recognise what percentage of elections a seat has historically supported their “preferred party” in. If a seat has a 90% rating in favour of the Coalition parties, for example, it supports the coalition in all but the 10% most extreme of Labor-favouring elections. So now we need to calculate whether this coming election is in that 10%.

The easiest way would be to look at all the 2PP results of Labor-won elections and see if the polling suggests the ALP: Coalition ratio is in the top 10%. Easiest mathematically, that is. Unfortunately pre-90’s 2PP data is scarce. Primary vote data, however, is abundant and given the inaccuracy of using decades old data to interpret current attitudes will hopefully suffice. Besides which, as Anthony Green points out [http://blogs.abc.net.au/antonygreen/2015/03/why-the-baird-government-is-vulnerable.html] the optional preferential voting system artificially inflates the 2PP vote. In only three elections since 1927 (1935, 1971 and 1995) has the winning party (and presumably 2PP leader) not been the leader of the primary vote (after combining all coalition votes and factoring Lang Labor into the ALP/NSW Labor vote).

Dividing these into Coalition- and Labor-led races, and ranking them from smallest to largest primary vote we can evenly divide the races into brackets of equal percentages:

Alternatively we can scale them by the leading party’s share of the Coalition/ALP primary votes, ignoring the others all together:

The latest polling (Roy Morgan 20-23 March) has the current primary vote at Coalition 45.5%, ALP 32.5%. A primary vote of 45.5 by the first scale is 30-40%. On the second scale, a 45.5/78.0 is 80-90%. So it’s not a particularly outstanding primary vote, but exceptional primary vote lead. Given that the majority of the “other” vote will eventually drain back to the Coalition or the ALP or exhaust, I’m more inclined to take the latter value. Besides which, ranking the elections by one party is somewhat lopsided. One might as well rank them by non-leading party’s primary vote:

Which puts this election in the 90-100% range.

Ruling all seats as Coalition unless retained by the ALP in > 80% of cases, treating the 80-90% range as tossups and 90%+ as ALP gives this unlikely prediction:

This gives the ALP 20 seats, the same as last election despite all polls indicating a significant recovery from. The difficulty in determining the bracket a current election falls in is, it seems, a major problem with this method.

What I will do, though, is look at the results after Saturday and work out where this election rates as compared to other elections and see if there is any obvious way to use this method in the future.
In the meantime, this chart compares the Coalition 2PP result for each seat with the state 2PP result for the last four elections. This is represented both as an absolute difference in percentage points (“Dif”) and as a percentage calculated as (Seat/State)-1*100%, where a positive value favours the Coalition and a negative value favours the ALP. Both scores are averaged over the four years, or as many years as data is included for.

Values are not recorded where the seat was not in existence (Balmain 2003, 1999; Castle Hill 2003, 1999; Cootamundra all years; Goulburn 2003, 1999; Holsworthy all years; Newtown all years; Oatley 2003, 1999; Prospect all years; Seven Hills all years; Shellharbour 1003, 1999; Summer Hill all years; Sydney 2003, 1999; Terrigal 2003, 1999; Wollondilly 2003, 1999), where the two party preferred count includes an independent (Barwon 2007; Cabramatta 1999; Charleston 2007; Dubbo all years; Goulburn 2007; Hawkesbury 2007; Hornsby 2011; Keira 1999; Lake Macquarie 2011, 2007; Londonderry 2003; Maitland 2007; Manly 2007, 2003, 1999; Newcastle 2007, Northern Tablelands 2011, 2007, 2003; Orange 2007; Pittwater 2007; Port Macquarie 2011; 2007; 2003; Shellharbour 2007; Sydney 2011, 2007; Tamworth all years; Willoughby 2007, 2003; Wollongong 2011, 2003) or minor party (Greens: Balmain 2011, 2007; Keira 2003; North Shore 2007; Vaucluse 2007)(One Nation: Cessnock 2003, 1999).

The idea of these two values is that they represent the lean from the state standard towards either party. Applying the average Dif and percentage to the baseline of 56% 2PP from the 20-23 March Roy Morgan poll gives the following:

By and large this correlates to the earlier prediction (shown in the COMPARISON column). Bathurst, Coogee, Londonderry and Riverstone are split on this method, favouring the Coalition based on the +DIF value and Labor on the +% score. These have been ruled as tossups.

Drummoyne, Heathcote, Mulgoa, Penrith and Port Stephens prefer the coalition under this method, rather than the ALP. Conversely, Strathfield favours the Labor party. The COMPARISON value in these seats never exceeds 50%, and should be ignored. Tossups under this method result from a lack of data. Where this arises from Independents or the seat not existing the baseline from the polling will be adopted. This leaves Balmain and Newton, where the COMPARISON is 100% ALP but that ignores the influence of the currently incumbent Greens, and Sydney which has been consistently Independent. In the last election (the 2012 by-election) Independent Alex Greenwich received almost 47.3% of the primary vote. This was, however, without ALP contest and may subsequently see Mr Greenwich’s primary vote eroded to the point that he does not make it onto the 2PP list. If he does, however, he should comfortable defeat the Liberal candidate. Similarly the Greens primary vote, based on the Roy Morgan polling, has increased from 10.3% to 12.0%. It is probable that the Greens will retain both seats.

The other consideration is Independents who may retain their seats. Normally these can be passed off as tossups, but there are too many this time: Lake Macquarie, Londonderry, Port Stephens, Swansea, Sydney, Terrigal, The Enterance and Wyong.

As stated, Sydney has a solid IND history. Lake Macquarie has gone to Greg Piper at the last two elections, and a third would be likely based on his 2011 primary vote of 43.7%. The other six independents were Liberal until a series of ICAC investigations led to them all resigning from the Liberal Party. These seats can be called based on the standard 2PP methodology:

Voila! LIB 35; NAT 16; ALP 34; GRN 2; IND 2; TOSSUP 4. Coalition Majority.

Oh, and here's an obligatory "State of the LegCo" chart. Coalition Majority expected there too. One day I'll find a reliable means of predicting the numbers for that...

Coalition ALP Greens CDP SFP

Monday, 16 February 2015

Summary (Davneport and Queensland State)

I’ll keep this quick and simple, because there isn’t a whole lot to analyse. We correctly predicted a Liberal victory in the Davenport by-election. So that’s nice, but doesn’t allow for much introspection.

In Queensland we made predictions for 85 of the 89 seats and got 75 correct. That’s 88.2%, which is a little lower than I’d usually like but it was an odd election, which was a correction for an even odder result previously.

Moreover, the standard polling got 74 of their 84, which is only 88.1%. So on the comparative scale we’re marginally ahead.

I was hoping to see which of the lengthy list of methods from the prediction bested the standard poll-and-pendulum approach and which didn’t, but in the end there were only 5 seats where one method beat the other.

Where we did well: Gladstone, Nicklin and Noosa. These are the three seats we did better than the standard method. Slightly unsatisfyingly, these were all tossups for the pendulum due to the existing 2PP margin not being an ALP v LNQ contest.

The three pendulum tossups we called, we called correctly. By contrast, the two of our tossups the pendulum called (Kallangur and Pumicestone) they got wrong. So we’re better at choosing our tossups. That’s something. Right?

So, how did we choose who to back in these tossups? Historical data.

In Nicklin, the Independent had been in since 1998 and had held off the Liberal National surge of 2012. Gladstone’s Independent was retiring, and had only ever been ALP prior to that.

Noosa was a little harder. But then Noosa was odd – a Coalition v Greens 2PP vote in 2012 but’s the seat high on my list of oddballs. The seat had had 6 Liberal/LibNat victories and 2 Labor wins. That’s a bias but not a conclusive one, especially with the anti-LNQ sentiment of recent Queensland politics.

Although we did discuss the problems with this approach – a strong ALP history for a seat created during the long line of recent ALP victories might be stalwart or bellwether – the polling did actually land slightly more LNQ-leaning than the historical trend, so if anything the seats would be more LNQ friendly than in the past.

Where we did badly: Chatsworth and Everton. Polling had both of these as LNQ wins, but we backed the ALP. Why? Historical data.

Both seats had only one ALP loss since 1977 (i.e. the 2012 abberation). However, throughout that time 1986 was the only election where Labor was not either:
gaining seats, or
the party that went on to govern Queensland.

Other seats in this position include: Brisbane Central, Bulimba, Bundaberg, Cairns, Cook, Ipswitch West, Lytton, Mackay, Murrumba, Nudgee, Rockhampton, Sandgate, South Brisbane and Woodridge. It’s tempting to say that the methodology was sound since we correctly picked all of those. Tempting, and wrong.

First, remove those with even longer-term ALP support (support when the ALP was not in favour). That leaves: Brisbane Central, Murrumba and Woodridge (and maybe, just maybe, Lytton and Rockhampton). And Brisbane Central was only introduced in 1977, so it’s hard to say whether or not this one would have been removed had it had a longer history.

Then scrap Woodridge, because that stayed with the ALP in 2012. That leaves us with Chatsworth and Everton wrong and Brisbane Central and Murrumba correct. 50:50. Compare that witth the poll-and-pendulum which got all four right.

So, while electoral history is very useful – in picking tossups at least – it needs to be tempered with a broader understanding of the state’s political slant. This was something we tried around this time way back in 2014 for the SA state election, and something we will return to for the NSW state election on March 28.

Hopefully I’ll have updated this blog’s allegedly running tally of correct predictions by then.

Sunday, 25 January 2015

Double or Nothing

Hey Paisanos! It's the Super Mar that double election prediction you've all be waiting for!

I'm going to get to Queensland in just a bit, but first a look at...

The Davenport By-Election

...the Davenport by-election. There are five candidates, all aligned to political parties (if you include the Democrats (which the electoral commission doesn't)).

The Candidates

Sam Duluk (LIB): Sam Duluk is an accountant who unsuccessfully challenged the late Bob Such in last election but did not contest the Fisher by-election last month. The Liberal party's "six priorities for Davenport" are:
  • Better local roads and transport
  • Fighting to lower our cost of living
  • More jobs, stronger economy
  • Fixing our schools and hospitals
  • Keeping our community safe
  • A greener, brighter environment
More information on how exactly they plan to achieve this can be found at www.samduluk.com.au/priorities. This is, of course, assuming the Labor majority government allows the Liberals to do these things in the first place.

Natasha Edmonds (FFP): A previous Family First candidate in this seat, Ms Edmonds she received 5.1% of the vote among a similar line-up of LIB, ALP and GRN candidates.

The Family First party is a former Christian Right party either responsibly distancing itself from it's religious origins or hiding them, depending on who you ask. Their pro-life and anti-same-sex-marriage positions are both well known and as one would expect. A few of their other policies, found here, are:

Family, which unsurprisingly for Family First, is a big priority, is supported through support for adult education, vocational training, support for stay-at-home parents and support for carers.

Climate Change (I want to quote directly here so I don't get accused of putting words in mouths where feet should be).
has been occurring since time immemorial. Many hundreds of eminent scientists have strongly criticised both the 'climate change doctrine' and the predictions made by the International Panel on Climate Change. Claims that 'there is a scientific consensus' and 'the science is settled' are not true. ... Carbon dioxide has had no discernible influence on the world's climate in the past and there is no reason to believe it should in the future.
I know, I know, it's a federal issue, but Family First put it on their state website and this does have potential impacts for the South Australian renewable energy sector, environmental concerns and future development plans.
Property Rights (I want to quote directly here, too, because I'm lazy and we still have three parties to get through).
provide the foundation on which many other rights and privileges are exercised. The progressive erosion of the rights of property owners through legislation, heritage listing, water restrictions, native vegetation, rising sea levels, zoning and court decisions is significantly diminishing the opportunities for self reliance.

Jeanie Walker (DEM): Ms Walker contested the Fisher by-election receiving 0.9% of the primary vote. The Democrats have struggled to get much support after their federal GST backflip in the Howard era robbed them of their campaign basis - honesty and integrity - and saw the Greens become the default third party instead (especially among left-wing voters, with FFP or PUP on the right). Their support has been so low they are no longer listed as a party by the federal or state electoral commissions.

I couldn't find a state-based source of Democrat policies, so you get more federal ones. I've tried to pick ones that reflect issues of high priority in the state's political dialogue:

Taxation, under the Democrats, would be a system "broadly based, progressive and based on capacity to pay". (Yeah, yeah, I know I'm quoting directly again. Live with it.) They oppose negative gearing, and see tax reform as a means to other political ends, e.g. promoting rural and regional based enterprises, promote the transition to more environmentally benign products and energy sources are encourage the use of public transport.

Transport - while we're on the topic - would be improved by upgrading the rail network, building high-capacity monorails, and making the cost of vehicle registration reflect the environmental footprint of the vehicles, all with the intention of encouraging public transport use and reducing green-house emissions.

Education, under the Democrats' policies, would include free and universally available child care for 3-4 year olds, free (eventually and/or ideally) university education and a requirement to pass year 12 before being eligible for a full drivers licence to increase student retention rates.

Mark Ward (ALP): voting for an ALP candidate has the added advantage that they will hold more sway in getting things done, securing more funding for local projects considering that they are in power, but the disadvantage that they are less likely to fight against government policies that negatively affect the area. Mr Ward is a member of the Mitcham City Council, where he had the benefit of rhyme when encouraging people to vote "Mark Ward for Park Ward".

The South Australian Labor Party website is down at the time of writing, so you are spared another list of policies. Until it's fixed, treat a vote for the ALP as a vote for the status quo.

EDIT: ALP website online on 29/01/2015

Not to be outdone by the Liberal Party's "six priorities for Davenport", the Labor Party has "Seven priorities for South Australia’s future". For more details on each follow the link, but the seven priorities are:
  1. Realising the benefits of the mining boom for all
  2. Every chance for every child
  3. Growing advanced manufacturing
  4. An affordable place to live
  5. Safe communities, healthy neighbourhoods
  6. Premium food and wine from our clean environment
  7. Creating a vibrant city

Jody Moate (GRN): We're having similar issues with the Greens website. I expect it's a server overload problem as the state checks in en masse prior to the by-election. So, you know, just ease up. I've got a blog to write and you election fiends are crashing the sites.

Conveniently, through the magic of Facebook, the Infographinomicon can bring you the Greens policies to the tune of Advance Australia Fair. (Policies in the second verse, around 0:40).


Last election the Liberal's won 51% of the primary vote, taking the seat before preferences were distributed. While there is no direct polling for the Davenport by-election available to me, I sense there is little appreciable movement. Nationally the trend is heading away from the Liberals, but a series of factors in SA - most notably the Gillman land deal - has cast a shadow over the House of Labor too.

That said, the minor parties don't look likely to do very well. The Fisher by-election may not be a very good source of public opinion given the unusual circumstances of Mr Such's legacy and the Woodyatt candidacy, so the Democrats may do better than 0.9%. That said, they'll fall well short of any real challenge and the FFP won't be far behind; it is ard to see them making a particularly massive improvement on last election's 5.1%.

If there is a shift away from the Libs it'll mostly go to Labor, not the Greens, so this'll be your typical ALP vs LIB race. That said, I expect Davenport has enough rusted-on Liberal voters to secure the seat comfortably for the opposition. The seat's electoral history is indicative of this:

Founded 1970
1970-1973 - Liberal and Country League
1973-Preset Liberal (but Independent Liberal briefly in 1985).

Predicted result - Liberal.

The Queensland State Election

Or, as the old (circa 2012) joke runs:

What's the Difference between a Mini-Bus and the Queensland Labor Party?

Queensland is always interesting electorally. For one thing, it doesn't have an upper house, which make the politics strange and the voting simple. For another, the Liberal and National parties are united as the liberal National Party (LNQ). But last year was particularly odd. Of the 89 seats, the Liberal National party won 76. That's over 85% of the seats. Odder still, Labor's share was only part of the remaining 13 - they won 9 of the seats (around 10%), with Katter's Australian party taking two and the last two going to Independents. In 15 of the seats held by the LibNats, the ALP was not even the runner up. Katter's Australian Party (KAP) came second in Beaudesert, Burdekin, Callide, Condamine, Gympie, Hinchinbrook, Lockyer, Nanango, Southern Downs and Thuringowa. Independents came second in Burnett, Maryborough, Nicklin and Warrego. The Greens came second in Noosa.

Normally when the greens come second it's in a left-wing seat where the coalition parties don't get a look in and it's just a question of how left-wing the seat ends up. in Noosa, the Liberal National party was still a serious contender, as evidenced by the fact it won, and the Greens still outperformed the ALP.

Such a thorough trouncing of a major party is almost unheard of in Australian politics. So what's the difference between a mini-bus and the Queensland Labor Party? The mini-bus has more seats.

However, the Liberal premier quickly squandered his popularity the moment he got into parliament - and I do mean into parliament, not into power, because he had not even been elected prior to the election where he lead the Libs to power.

Polls Apart

Drastic cost cutting and various other unpopular moves has the polling very close. The latest polling data has the Liberal party at 52% on two-party preferred counts, which is within the margin of error of being 50-50.

The March 2012 election had the 2PP vote split 62.8% LIB, 37.2% ALP. If the latest polling data is correct, we're looking at a 10.8% swing away from the Liberals. If this were uniform across the board, it would turn the pendulum from this:

Italics indicate retiring MPs
To this:

With LNQ 34, ALP 37 and 18 tossups. Normally this would be our baseline to compare predictions against. However, because my predictions are limited to (Number of Seats) x 5% (= 4.45 for Queensland, rounded to 4) I won't let the baseline carry so many wildcards.
The KAP is responsible for the bulk of cases where the 2PP vote is not LNQ v ALP. Although the movement is against the Liberals, the KAP primary vote has dropped almost two thirds based on the latest relevant polling (3.5% between Jan 16-18 compared to 11.5% in the 2012 election). Any seats the KAP failed to take last time will be harder than before, and any seat where the KAP was in second place, the ALP could not have been within 10.8% of taking the seat. So all LNQ wins against KAP will be called LNQ for the purpose of the baseline. Polling for independents is obviously difficult to come by, but it's hard to believe any second-place INDs would drastically increase their vote, especially with a more competitive ALP this time around, so these have been given to the LNQ too. The Greens vote has improved slightly, retaining Noosa as a tossup.

This leaves five tossups on the baseline, but I'll allow that because any further changes to the baseline begin to get very complicated and it will no longer reflect the "first glance" data we are trying to quantify and, ideally, beat.
I will, however, add a final note of caution about relying on polling. The phrasing of the questions, the format and the polling methodology can very drastically affect the outcome. Compare the "Better Premier" polling for this election for the last two years. (Source: Wikipedia)
At no point has Roy Morgan polling put Campbell Newman above 50% or Annastacia Palaszczuk below 50%. By contrast, Newspoll has never put Ms Palaszczuk over 35% or Mr Newman below 39%.
While most of the Newspoll data is earlier than the Roy Morgan polls, the consistency of the Oct-Dec Newspoll with all previous data, in the midst of equally consistent Roy Morgan polling, is far more likely to indicate methodological differences than actual trends in opinion.


Here's a diagram of each seat's history since the abolition of multiple-member seats before the 1912 election.
Red = ALP; Blue = LNQ (or predecessors); Pink = KAP; Grey = IND; Green (2008-2009) = GRN; Orange (1998-2002) = One Nation; Yellow (2012-2013) = PUP; Black (1944-1963) = North Queensland Labor (Later DLP)
And perhaps more relevantly, here are the results of each general election:
This allows us to work out which major party a seat normally voted for, and the strength of this support, calculated by (F-A)/T where F is the number of times they voted for that party, A is the number of times they voted for the 2PP opposition and T is the number of elections the seat was contested in. By-elections are ignored for simplicity. This gives a fidelity value between 0 (50-50 split of past elections) and 1 (always supports the same party). This differs from the simple win:loss ratio in that a Liberal seat that occasionally votes for an independent will have a higher fidelity score than one that occasionally votes for Labor.
The ALP holds two more seats than the LNQ in this count, and 12 more if we ignore the data from last election as an aberration. We can also weight the results by making each decade count for twice as much as the previous decade, and dividing by the number of elections weighted in the same way. This system ignores last election as it is the only election of the current decade AND is highly atypical. This is potentially catastrophic for any weighted system, giving the atypical result twice the weight of any other data.
So, I know we've come a long way round and I'd normally use a data dump for all of these tables, but they're crucial to the flow of this post, but here is the resulting data, with columns for seat fidelity with and without the 2012 data (the latter only being used for resolving tossups in the former data), the weighted fidelity, and a general summary of the seat's normal political leanings.
I only have 2PP results for the last four elections, but the average LNQ:ALP 2PP ratio is 50.35:49.65. Without 2012 the average ratio is 46.20:53.80. It is difficult to work out what a "typical" polling result would look like, but the 52:48 appears, if anything, to be more Liberal National leaning than normal. Let us then call anything on the above summary marked LNQ as safe, While we know anything that didn't leave ALP hands in 2012 is probably not going to any time soon and particularly not this election. Let's also keep the predictions as they are wherever the pendulum/baseline and above summary agree.
The predicted swing is in the 20% range, so let's keep any of the remaining spots where the margin is over 15% as a seat retained.
When comparing these predictions so far with the fidelity scores, we find we've contradicted a .72 unweighted fidelity for Bundaberg and a weighted fidelity of 1 in Mudgeeraba. Having looked closely at these instances, and in particular the Mudgeeraba issue, and concluded that the fidelity scores are not necessarily helpful; Mudgeeraba was ALP for four elections and LNQ in 2012 which gives it an unweighted (4-1)/5 = 0.6, but weighted value of 1 due to the exclusion of 2012 data. The margin, which had us call it LNQ, is significant (25.93% margin meaning a 75.93% share of the 2PP vote). The fidelity, however, only reflects a recent pattern of ALP favour in an era of strong ALP government.
The remaining gaps are filled first by using the pendulum where the fidelity scores give a tossup value, then considering each remaining seat in detail. Additionally, seats held by parties other than the LNQ and ALP are reconsidered (marked !!! in the above table).
Broadwater: Held by a margin of 11.29, close to the predicted swing, and with a mixed history but many Liberal victories during Labor's strong victories in the 90s, called LNQ.
Burleigh: Identical electoral history to Broadwater, but with a margin of 11.05, called LNQ.
Chatsworth: This seat's 13.94 margin belies a strong recent ALP history - 2012 being the first Liberal victory since the 70s. Called ALP.
Dalrymple: Held by KAP by a margin of 15.22 is hard to peg down. It is a respectable margin for any party, especially a minor one. I'm tempted to call this for KAP because the candidate won a by-election then held out against the LNQ, suggesting this was not just a 2012 aberration. However, there is no data on how consistent the decline in KAP support is. Tossup.
Everton: Like Chatsworth, the margin of 13.15 hides a AP legacy dating back to the 70s. Called ALP.
Gladstone: Held by an Independent who is retiring. Won by a margin over ALP despite the oddities of 2012, and the only election not won by an independent was ALP, so that's what I'm calling it.
Kallangur: 12.43 margin but historically following the general trend of ALP in the 90s flipping to LNQ in 2012. Probably LNQ again but calling this a tossup.
Mount Isa: KAP held by 10.04 since 2012, by a margin over the LNQ despite the recent ALP history of the seat. Could go any of three ways. Tossup.
Nicklin: Independent, the margin is only 4.88 over the LNQ but the candidate has maintained an independent tradition started in 1998, and the margin may have been narrowed by the LNQ support last election which will have waned since. Calling this IND.
Pumicestone: ALP history in this seat is a result of the party's dominance in the 20s and 2000s. The margin of 12.07 is pretty shaky though. Tossup.

Final Prediction:

That's LNQ 44, ALP 40, IND 1, TOSSUP 4. If the ALP gets the four tossups it's still a hung parliament, with both parties relying on IND support for government. In all probability the split could go either way, so a LNQ government is likely.

Sunday, 18 January 2015

Double Take

We’re just doing a quick retrospective this week to look at our predictions for Victoria and Fisher and see us tow very wrong we were.

First, Fisher, because that’s a simple right or, in this case, wrong.


This was a really close election. As in, recount close. For simplicity we’re just going to look at the recount results, not the pre-recount ones.

Prediction: D. Woodyatt (IND)
Result: N. Cook (ALP)
My prediction had placed Woodyatt, Labor and Liberal in the final three, and guessed that Woodyatt would out-poll Cook, receive the bulk of preference flows and soar ahead of the Libs. I was correct in my guess of final three – not that that is worth anything, even bragging rights, in this case.

However, Labor out polled Woodyatt 6325 to 6106. 219 votes is not much out of 20,559, but it’s not super close either. The independent’s votes split more to Labor than Liberals, suggesting he claimed more pragmatic Labor supporters than disenfranchised Libs, but I doubt the relationship is symmetric.

While Woodyatt’s preferences split 2147:3959 (or about 1:2) against the Libs, I’d suggest that if Nat Cook had dropped out there would have been enough anti-Liberal sentiment in the ALP ranks that the preference flow would have been far more extreme in it’s anti-Coalition skew.


Predictions did well here, with a 93.2% success. Unfortunately we’ve yet again failed to meet the baseline of 94.0% set by applying the polling data to existing margins.

However, let’s not feel too bad. We called every right answer that the polling did. Both systems failed to predict a Liberal win in Forrest Hill, Ripon or South Barwon. Neither expected the Greens to take Prahran, even given the Federal electoral landscape. And no practical method could even have expected the Independent victory in Shepparton.

The only seats we got wrong polling didn’t were the ones the polling couldn’t deal with – the tossups in Brunswick, Melbourne, Northcote and Richmond where 2PP polling didn’t work because the race was ALP v Greens. Our method backed Labor in all four, but Melbourne ended up with the Greens.

So while we got a lower percentage than the polling, we actually got more right (82 instead of 79) as a result of taking a punt at contests the baseline was too chicken to deal with.

Sunday, 11 January 2015

2015 Schedule

It's going to be a reasonably relaxed year after the New South Wales state election on March 28. Before then, however, we have both the Queensland state election and Davenport by-election on January 31.

At last, the methodology for the Victorian predictions has been outlined as of this weekend, along with this schedule.

Next week (17th or 18th of January) I hope to analyse both the Victorian state election and Fisher by-election predictions, and maybe even update the so-called running tally of correct predictions to the right.

The weekend after that (24th or 25th of January), if all goes to plan, we'll have the predictions for both the Queensland and Davenport elections.

The following weekend is the actual vote in both of those elections. We may have results by then, or counting may be ongoing, but I'd expect to have both elections finished with be mid February, when we'll gear up for NSW.

I may throw in a few special elections if and as they arise, and I'll probably take a second shot at Eurovision at some point to, but we'll see how that goes closer to the date.

Finally, a thankyou to me select and discerning readers for their patience and continued support. Without you the Infographinomicon would be... pretty much the same, actually, except the task would feel even more lonely and futile than it does.

Have a statistically great year,

The Psephology Kid.

Friday, 5 December 2014

Gone Fishing

T’was the night before a by-election in the seat of Fisher, and a prediction was thus due. The problem with by-elections are that you get the result 100% right or 100% wrong, which in a close race either massively exaggerates or conceals the actual value of the predictions. For this reason I won’t be comparing my prediction to any baseline, although the vote itself will still count towards the blog’s tally whenever I get around to updating it.

Also due to the all-in or all-out nature of the prediction, a lot of sources that typically toe the “it’ll be a close race, let’s wait and see what happens” line have flipped off their psephological perches.

It’s been one of those elections…
ABC state news has tonight tentatively backed Liberal candidate Heidi Harris to win. The Advertiser seems to be leaning that way too. Crikey.com is backing first-time Independent Dan Woddyatt which would normally be quite a gutsy bet, but is probably pretty sound. And everyone keeps saying that Labor will put up a strong contest, although no one will touch the idea of an ALP victory with a 10-foot poll. (Hahaha. Poll.)

There is some dubious polling available for the seat. The Advertiser sampled 400 voters to produce this little gem:
Source: http://resources2.news.com.au/images/2014/12/01/1227141/408650-ef267956-793d-11e4-a0f0-f6e11a4d80ae.jpg

which is particularly uninformative despite what The Advertiser assures is a sufficient sample size. The lack of detailed survey methodology, demographic breakdown or any of the basic standard requirements for a meaningful poll aside, it demonstrates only that preferences are going to play a huge role and provides absolutely no data on preferences.

Polls apart

So, on a first-past-the-post count the Liberals would win this. Their primary vote is the highest, but not enough to win outright, so it comes down to preferences. One bad way to estimate preference flows is to assume that voting preferences are a series of independent probabilities; in other words if a three-party race splits the vote 4:3:2, the 2/9ths of voters that go to second preference will split 4:3 in accordance with all the other votes. In such a scenario there is no real difference between preferential voting and first-past-the-post and the Libs would win Fisher. Obviously, though, some parties (and Independents) are more closely aligned than others, so this rarely works.

We could look at how preferences flowed last election and approximate something similar, but the comparison is a poor one. The issues are different. The campaign is different. Television channels are not plastered with grainy black-and-white scare campaigns. The last election saw long-term Independent Bob Such outpoll the Liberals on their primary vote; now the Libs lead the field against three independents and several minor party reps. Malwina Wyra, the Greens candidate, is the only candidate from the March election also running this time.

The other option is one of broad generalisation. One major theme of the campaign has been the value of an Independent representative traded off against the power of a candidate with party backing when they come to the negotiating table. The Independent vs Major Party issue was covered in one of the four questions polled by The Advertiser. 50.5% of respondents thought an Independent would do a better job representing the seat of Fisher than a major party (although the Labor and the Liberal candidates polled a combined 54.75% of the primary vote).

68% say we spend too much on foreign aid. 59% want foreign aid cut.
I’m going to assume that, based on the reporting so far and possibly because of it, people who vote 1 for an Independent will probably preference other Independents pretty highly too.

The Stop Population Growth Party and Democrats are predicted to drop out first, but their preferences are unpredictable. Even if they go entirely to Labor (which is highly unlikely to feature anywhere nearly that high in the preferences of either voter base) who then also scoop up the Greens vote, Dan Golding and Rob de Jonge will keep Dan Woodyatt in a safe second. So long as Woddyatt out-polls the ALP he’ll do very well from the anti-Liberal vote and will be in a very good position to win. This is my gut instinct on how things will actually play out, and if Woddyatt only absorbs Labors primary vote he still gets over 50%.

Of course in the real world, he won’t get all of Labor’s vote. Still, Woddyatt’ll do well from ALP support along with preferences from other Independants, a share of the Dems and the Stop Population Growth Party, and a fair chunk of the Greens. This is despite reports that “strategists in both major parties say they still expect Labor’s more polished ground game to lift their actual vote at the weekend into second place behind the Liberals”.

Dramatis Personae

After Bob Such passed away, a by-election (which would normally draw more contenders than usual anyhow) seemed like a good shot for any aspiring Independent. It also attracted yet another round of attempts from the Stop Population Growth Party, saw the Democrats (now the Independent Australian Democrats) enter once more into the breach, and also has a Greens candidate because the Greens are now making a point of contesting every seat every time.

My gut instinct that Dan Woddyatt will ride a wave of preferences until Labor pushes him over the line is partly based on Bob Such’s early successes as an Independent in Fisher. It is also partly based on the campaign material I’ve seen.

Dan Woodyatt (IND) has primarily campaigned on continuing Bob Such’s legacy. His “party” on the ticket will be listed as Independent Continue Such’s Legacy, and he has the added bonuses of Bob Such’s widow’s endorsement and (as a result of the ensuing news coverage) is the highest profile of the Independents.

Heidi Harris (LIB) has campaigned on two fronts. She has taken the somewhat radical approach of actually listing things she supports or intends to do in her document Heidi Harris’s plan for Fisher to get 50 things done in the first 50 days: Includes 10 benefits for the people of Aberfoyle Park.
This includes:

18. I will help eligible multicultural organisations to apply for a Land Tax Relief Grant
24. I will establish a ‘footpath register’ so people can inform me of footpaths in their neighbourhood that need maintenance
38. I will work with Trees For Life to revegetate local parks

Now I don’t normally comment on policy, but I will talk about campaigning and advertising so I should at least point out that I have chosen these examples because I find them amusing in that they can be done without winning a seat in parliament (and, in the case of the footpaths, are actually a Local Government issue). There are more traditional promises with regards to the Emergency Services Levy (point 1), Payroll tax (13) and road upgrades/bus services (3, 5, 6, 7, 8, 12, 14, 16, 20, 29, 32, 37, 39 and 41). Back on the less predictable side there are also plans for: talking with the fire brigade to make sure they’re ready in case of a fire (17) and let them know when they need more money (27), conducting fundraisers for causes yet to be determined (19), handing out stickers (21), talking about graffiti (26), making a website (28), networking young professionals (34), reading to students at every local primary school (43), helping volunteers to volunteer for stuff (45), taking guns off the streets (47) and planting vegetables in schools (49).

The second prong is an attack campaign (no surprises there) against Labor (predictably) and to a greater extent Dan Woodyatt (okay, maybe there is a small surprise), including a letter from Bob Such’s Electorate Officer suggesting that Ms Harris is the real Bob Such 2.0, and (separately) criticisms of Woodyatt include that he used to be an ALP member, referred to Jay Weatherill as a “Statesman” and lives at Bellevue Heights just beyond the electorate boundary.

This is particularly ironic given that Nat Cook’s (ALP) campaign includes a comparison with Ms Harris, pointing out that Ms Harris also lives outside the electorate. The remainder of Ms Cook’s leaflets that I have seen do not outline her plans or ideals (which is probably unnecessary since they’ll be in accordance with publically available Labor policy information), but emphasises her 25 years as a nurse and creation of the Sammy D foundation after her son was murdered.

Possibly in a completely unrelated move (but also totally not unrelated, given that it is apparently unique to the Fisher electorate) Labor has been delivering its own attack campaign documents, criticising the Abbott (i.e. Federal) government over proposed outsourcing of submarine production, university deregulation, cuts to health and education, the GP tax and petrol levy, backed up by an impersonalised letter from Penny Wong. The fact that none of these federal issues really come into play in state politics is not really mentioned anywhere.
That covers all three candidates polling above 5% of the primary vote, but for completeness Rob de Jonge (IND) is baking on 8 years experience in council and supports fairer fines, unspecified health and education improvements and two-weekly green waste collection. Dan Golding (IND) has apparently posted some leaflets but I have not found any, and conducted an online campaign through low-resolution policy images

Jeanie Walker (DEM) and Malwina Wyra (GRN) are both running campaigns from facebook like de Jonge and Golding, which in my opinion just looks like a lack of party backing, although at least Ms Wyra links to the SA Greens website so you can get some idea of the policies involved. Bob Couch (SPG), by contrast, has a manifesto listed by the ABC as his primary campaign website, where he blames unemployemt, housing affordability, utility costs, health care quality, traffic, high density urbanisation and loss of biodiversity on population growth, and provides his policies ranging from tripling penalties for drug-related crime, legalising euthanasia and preventing foreign ownership of Australian real estate.

While I have not found a lot of additional information on any but the three leading candidates (Harris, Woodyatt and Cook), I feel confident that there are unlikely to be any surprises form the back field. The Libs seem keen to keep Woodyatt down, presumably because if he falls behind Nat Cook and dros out the Libs get a lot more preferences than Labor. Labor, in turn, is prepared to forego any obvious policy campaigning to keep the Libs back in striking range of Woodyatt.

Bonus Prediction

Obviously Labor will try to win, but Fisher is so heavily pro-Liberal I doubt there will be too much ALP disappointment if Woodyatt wins off Labor preferences and sides with the government.

My additional prediction is that Woodyatt will be reasonably happy to work with the ALP given:

A)     They have the power
B)      He used to be an ALP member and is therefore presumably more closely aligned with Labor policies than Liberal
C)      The Liberals have targeted him specifically in this campaign


Woodyatt wins on preferences, and sides with ALP for matters of confidence and supply but will make enough noise to retain his Independent for Fisher status next election.

Friday, 28 November 2014

Victoria 2014

This post was supposed to be coming to you from the great southern continent of Antarctica. That, in itself, posed a few challenges. However after my plans were interrupted by mechanical failure, a bout of food poisoning and an entire Friday wasted as a result of an absent minded mistake this post became even harder to create, despite coming to you from our regular base of operations.

I had hoped to get some party profiles done it time, but didn't. However, a rough guide to most of the parties can be gauged by the last federal election's senate ticket for Victoria, analysed here.

More information on how the predictions below were reached will be forthcoming, however for now I will simply assert that I expect the following results in the lower house:

Albert Park: Labor
Altona: Labor
Bass: Liberal
Bayswater: Liberal
Bellarine: Labor
Benambra: Liberal
Bendigo East: Labor
Bendigo West: Labor
Bentleigh: Labor
Box Hill: Liberal
Brighton: Liberal
Broadmeadows: Labor
Brunswick: Labor
Bulleen: Liberal
Bundoora: Labor
Buninyong: Labor
Burwood: Liberal
Carrum: Labor
Caulfield: Liberal
Clarinda: Labor
Cranbourne: Labor
Croydon: Liberal
Dandenong: Labor
Eildon: Liberal
Eltham: Labor
Essendon: Labor
Euroa: Liberal
Evelyn: Liberal
Ferntree Gully: Liberal
Footscray: Labor
Forest Hill: Labor
Frankston: Labor
Geelong: Labor
Gembrook: Liberal
Gippsland East: National
Gippsland South: National
Hastings: Liberal
Hawthorn: Liberal
Ivanhoe: Labor
Kew: Liberal
Keysborough: Labor
Kororoit: Labor
Lara: Labor
Lowan: National
Macedon: Labor
Malvern: Liberal
Melbourne: Labor
Melton: Labor
Mildura: National
Mill Park: Labor
Monbulk: Labor
Mordialloc: Labor
Mornington: Liberal
Morwell: National
Mount Waverley: Liberal
Mulgrave: Labor
Murray Plains: National
Narracan: Liberal
Narre Warren North: Labor
Narre Warren South: Labor
Nepean: Liberal
Niddrie: Labor
Northcote: Labor
Oakleigh: Labor
Ovens Valley: National
Pascoe Vale: Labor
Polwarth: Liberal
Prahran: Labor
Preston: Labor
Richmond: Labor
Ringwood: Liberal
Ripon: Labor
Rowville: Liberal
Sandringham: Liberal
Shepparton: National
South Barwon: Labor
South-West Coast: Liberal
St Albans: Labor
Sunbury: Labor
Sydenham: Labor
Tarneit: Labor
Thomastown: Labor
Warrandyte: Liberal
Wendouree: Labor
Werribee: Labor
Williamstown: Labor
Yan Yean: Labor
Yuroke: Labor

Gembrook and Mount Waverly were both a bit dubious, but I did not use any of the 4 tossups available to me. All of these results do, coincidentally, coincide with the swing compared to the adjusted polling from the 2010 election, but makes definite claims in Brunswick, Melbourne, Northcote and Richmond, which would all be tossups under the swing-based system as the 2010 2PP results were ALP vs Greens, and thus the Labor-Liberal swing would be irrelevant.
No predictions have been made for the upper house.