Question on PM4E-Australia

[matvail2002]

Could somebody explains to me how does the ''second choice vote'' works and affects the results?

[gopprogressive]

Basically it becomes "2-party preferred" which you can select upon in the game for polling data. However, it only applies if a candidate does not secure an absolute majority. For example, if Candidate A gets 52% of the vote, Candidate B gets 36%, Candidate C gets 7% and Candidate D gets 5%; then Candidate A wins by default.

However, if Candidate A got only 45% of the vote while B's votes remain at 36% but C and D's votes increase- then C and D's preferences will decide who wins. If C and D's voters overwhelmingly support B, then B will win; if they split 50-50 then A will still win.

Does this help? It's pretty much how the game works.

[matvail2002]

Basically it becomes "2-party preferred" which you can select upon in the game for polling data. However, it only applies if a candidate does not secure an absolute majority. For example, if Candidate A gets 52% of the vote, Candidate B gets 36%, Candidate C gets 7% and Candidate D gets 5%; then Candidate A wins by default.

However, if Candidate A got only 45% of the vote while B's votes remain at 36% but C and D's votes increase- then C and D's preferences will decide who wins. If C and D's voters overwhelmingly support B, then B will win; if they split 50-50 then A will still win.

Does this help? It's pretty much how the game works.

So, does the results indicate how many ''second choice'' votes wants to the winner?

[Smid]

So, does the results indicate how many ''second choice'' votes wants to the winner?

As an Australian, I can confirm that what GOP Progressive says is entirely correct.

Additionally, in his/her first example of:

A - 52%,

B - 36%,

C - 7%

D - 5%

They would still typically allocate everyone's preferences to result in a 2PP result. Indeed, those numbers are remarkably similar to the results in both Cook and better yet, Wannon, in the 2008 Federal Election - http://results.aec.gov.au/13745/Website/HouseFirstPrefsTppByDivision-13745-NAT.htm (if you click on the name of the electorate, you can get more detailed information, including number of votes for each candidate at particular polling booths, etc).

Cook:

Liberal - 52.4%

Labor - 35.91%

Green - 6.2%

Other (Including the Christian Democrats, One Nation and two separate independents) - 4.29%

Wannon:

Liberal - 52.56%

Labor - 36.17%

Green - 6.98%

Family First - 4.29%

You'll note that in both cases, although a party received more than 50% of the primary vote, the electoral commission still distributed preferences, resulting in a two party preferred result of 56.57% Lib vs 43.43% ALP in Cook and 57.47% Lib vs 42.53% ALP in Wannon.

In the game, you can switch between primary vote and two party preferred - which distributes preferences in all seats.

[Alfonzo]

Basically it becomes "2-party preferred" which you can select upon in the game for polling data. However, it only applies if a candidate does not secure an absolute majority. For example, if Candidate A gets 52% of the vote, Candidate B gets 36%, Candidate C gets 7% and Candidate D gets 5%; then Candidate A wins by default.

However, if Candidate A got only 45% of the vote while B's votes remain at 36% but C and D's votes increase- then C and D's preferences will decide who wins. If C and D's voters overwhelmingly support B, then B will win; if they split 50-50 then A will still win.

Does this help? It's pretty much how the game works.

That's not entirely how it works.

It A gets 35.8% of the vote, B gets 22.6% of the vote, C gets 21.3% of the vote, D gets 19.0% of the vote, in instant-runoff elections, it would reduce to a 2PP between A and B.

In Australian elections however, D gets eliminated first, and D's votes distributed accordingly. Let's say D's preferences flow 56.7% to C. This is enough to put C ahead of B, which means that B is the next to be eliminated. From there, we have a 2PP between A and C. Now let's say B's preferences flow 81.5% to C. This puts candidate C on 51.2%. Causing a candidate who came third in first preference votes to be elected.

This is precisely what happened in the seat of Denison in the 2010 Australian federal election.

I'm not sure if preferencing calculations in PM4E takes this into account. For PM4E, it should work by successively eliminating the lowest placing candidate and distributing their votes. But I remember reading somewhere that the game works off instant-runoff by eliminating all but the two highest placing candidates, and distributes preferences of the eliminated candidates, in one calculation.

Ideally, we would have a second option when turning preferences on in the PM4E engine. 0 would be no preferences, 1 would be instant-runoff preferences, 2 would be successively eliminating the lowest placing candidates.