Rich wrote:
> Sam R. wrote:
> > Each booth counts separately, tallies centrally. =A0If booths are
> > working on an expectation that Sue will be eliminated first, but Jane
> > is eliminated first, then all counts prior to that elimination will
> > have to be recounted in the booth. =A0As the count was based on booth
> > preferences flowing from Sue to Jane, but in reality they flow from
> > Jane to Sue...
>
> What a Rube Goldberg mess!
>
> No wonder the count is taking a month.
Tyranny of Distance, mate, and it works fine in general when the
electorate individually and collectively are decisive. Normally
marginal seats take a while to count. Normally the last few marginals
don't matter to the formation of Government.
> ISTM that attempting to process vote
> transfers anywhere but at a single location
> per electoral division is asking for trouble,
> and that counting at a single site would
> go _much_ faster.
>
> Also, no count is possible until all votable
> ballots have been returned. I note that
> Corangamite is still undecided because
> several hundred mail ballots haven't
> come in. The deadline for that should
> have been a week after the election.
Not going to disagree with you there; but, given that paper and this
method has worked fine and produced great results in timely fashions
so far, and, that the electorate is highly suspicious of
constitutional change and electoral change (such as electronic
voting), it is unlikely to result in centralisation.
> What would happen if there were three
> parties A/B/C/D, and the vote was
> distributed like this?
>
> 35 A/D/B/C
> 30 B/D/C/A
> 25 C/D/A/B
> 10 D/C/B/A
>
First round: A 35, B 30, C 25, D 10. D drops, D votes flow to C.
>
Second round: A 35, B 30, C primary 25 from D 10 total 35. B drops, B
votes incapable of flowing to eliminated D, flow to C.
>
Third round: A 35, C 65. C wins.
>
> Except: when B drops, shouldn't those 30
> votes go to D, which is ahead of C on them?
No, because D has been eliminated already.
> In which case D reappears in the third round,
> which is A 35, C 25, D 40 and C drops.
Nope, elimination by least number of primary votes. This is why much
of the election is focused Vote [1] NiceParty. Primary preferences
matter in split vote situations. If B voters really wanted D, they
would have voted D first!
> Since D was the first or second choice of
> all voters, that seems a better result than
> C winning - D was preferred to C 75/25,
> to B 70/30, and to A 65/35.
Yes, this is an illustrative example of how every electoral system has
limits. The Australian one was focused on allowing competing members
of the Country Party to run against each other, so mid-point
preference perversity is not seen as a great issue. It also requires
a relatively even split between 4 parties in a ballot, where as in
Australian reality, the primary preference split is at most between
three potential candidates as in the Division of Melbourne where
Liberal voters preferenced Greens, allowing Greens to defeat the ALP,
despite the ALP having more first preferences than the Greens.
Given that Australia had up until 1990 a 2.5 party system, and
subsequently a 2.75 party system, this doesn't disturb most voters.
As the National Party seems content to merge into the Liberal Party,
and as the Greens seem to be forming a progressive bloc left of Labor,
it may return to a 2.5 party system (with more independents than prior
to 1990).
Them's the costs of having a constitution. Though, in Australia due
to Multi-member quota voting in the Senate, we get a bite of the apple
in another direction to make up for any flaws in the composition of
the Lower House.
(Give me the simplicity of continuously sitting deliberative councils
of the entire working class!)
yours,
Sam R. |
