[EM] Proportional election method needed for the Czech Green party - Council elections

[Peter Zbornik]

Dear Raph Frank,

Thanks, for sorting things out and for the example.

Based on your comments, I'll try to explain what I meant by the
"unified" method, even though you basically said it all in your
previous email.

Thus, as you pointed out, the "unified" Schulze method is equivalent
to Schulze STV, if it is modified to always include already elected
members. You mention, that the Schulze proportional ranking thus is a
special case of this method which always elects only one member.

The "unified" Schulze method is also equivalent to Schulze's
proportional ranking, if it is modified to elect groups of hopefuls.
A special case of this method is Schulze STV - where the group size is
the same as the number of seats.

I guess, that we can say, that Schulze proportional ranking and
Schulze STV are special cases of an underlying Schulze method.

Thanks for pointing these things out.
I really have to take a closer look at the paper, and at Schulze STV.

Best regards
Peter Zborník

2010/5/7, Raph Frank <raphfrk at gmail.com>:
> On Fri, May 7, 2010 at 4:27 PM, Peter Zbornik <pzbornik at gmail.com> wrote:
>> The proportional ranking needed is not P>VPa>VPb>Ma>Mb>Mc>Md,
>> but P>[VPa, VPb]>[Ma, Mb, Mc, Md].
>> Let us call this required ranking for "boundary conditions".
>
> Schulze's method can do that too.
>
> Step 1: Elect the Schulze single seat method winner as President
> Step 2: Elect a 3 person council using Schulze-STV, but the President
> must be a member
> Step 3: Elect a 7 person council using Schulze-STV, but the President
> + VPs must be members.
>
> I think this is what you meant by your unified method?
>
> Schulze rankings is just Schulze-STV, except you elect councils that
> increase in size by one each iteration, and members elected in
> previous iterations must be members of subsequent councils.
>
>> Example (from an email by Schulze):
>> "40 ABC
>> 25 BAC
>> 35 CBA
>> The Schulze proportional ranking is BAC.
>> However, for two seats, Droop proportionality, requires that A and C are
>> elected."
>>
>> The "unified" method for two seats without boundary conditions would
>> select
>> BA (i.e.Schulze STV)
>
> Schulze-STV meets the Droop criterion, so would elect A and C in a 2 seat
> race.
>
> Schulze-rankings elects B and then A as you say.
>
> There are 2 steps:
>
> *** Work out A's score vs C:  ***
>
> We split the voters in 2 groups
>
> Voters who prefer B to A: 60
> Voters who prefer C to A: 35
>
> There are no options in which group each voter can be placed, as no
> voter is eligible for both groups.
>
> The smallest group has 35 voters so, A better than than C gets 35 votes
>
> *** Work out C's score vs A ***
>
> Again we split into 2 groups
>
> Voters who prefer only B to C: 0
> Voters who prefer only A to C: 0
> Voters who prefer both to C: 65
>
> Thus we split the third group into 2 parts, as they can be placed in
> either group.
>
> Voters who prefer B to C: 32.5
> Voters who prefer A to C: 32.5
>
> The smallest group has 32.5 voters so, C better than A gets 32.5 votes
>
> Thus the result is
>
> A gets 35 votes and C get 32.5 votes, so A wins the 2nd seat.
>
> Anyway, I think the rankings method can be generalised to allow groups
> of candidates to be elected at once, rather than electing them one at
> a time.
>

-- 
Odesláno z mobilního zařízení