Re: Voting system R&D (Re: 2017 update to the SPI voting algorithm for Board elections)

On 3 March 2017 at 22:05, Barak A. Pearlmutter <barak(at)pearlmutter(dot)net> wrote:
> Dimitri,
>
> Of course we're discussing multiwinner systems: systems that elect
> k-of-n people to a board, where k>1.
>
> As you note, when k=n the election is uncontested so there is no need
> for any voting system.
>
> In the Burlington election discussed, k=1 and n=3. That's about the
> simplest situation you can have, so any credible multiwinner system
> should perform flawlessly in that degenerate case, one would hope!
>
> For a bit of background, all the systems we're considering have the
> following basic character. (I'm going to simplify a bit.) They use
> some "underlying single winner" system to elect a candidate, then they
> adjust the ballots in some way to account for that, and iterate this
> process until k candidates have been elected.
>

STV uses more than just underlying single winner system, in the
multiwinner case.
Moreover STV is only defined for multi-winner case since it reduces to
IRV in single-winner case (as there is never a second winner to
transfer the surpus of the first winners' votes over to)

> The multiwinner system called STV uses IRV for the underlying
> single-winner system.
>

STV is a superset of IRV.

STV uses surplus-transfer + eliminate looser;

IRV uses just eliminate looser.

In some publications STV and IRV terms are used interchangeably, which
imho adds to the confusion.

STV and IRV are two distinct systems, both of which can be used for
multiwinner elections.

IRV is not proposed to be used for the SPI elections.

FairVote in the USA promotes ranked choice voting, which is more
commonly referred to as IRV.

> The multiwinner system called RRV uses RV (range voting) for the
> underlying single-winner system.
>
> STV does the "adjust the ballots" step by basically removing a certain
> number of ballots that supported the candidate that just won.
>

STV does not remove or discard ballots:
1) if winner reached a quota - they are deemed elected
1.1) if they have surpus of votes those are transfered (not
removed/discarded) to next preference
2) Repeat above until there are no candidates left reaching the quota
3) If not enough winners are determined yet, eliminate the looser and
transfer their votes to the next preference.

This works around and produces different results than just IRV, such
that e.g. a candidate who has no first preference votes; but has 100%
of the second preference votes will probably get a seat.

> RRV does the "adjust the ballots" step by putting a numeric weight on
> each ballot, and lowering the weights of ballots that supported the
> candidate that just won.
>

Sidenote, RRV most famously for me used in determining nominations for
Visual Effects award at the Oscars. I cannot find any examples of
anybody credible using RRV for multi-winner elections. It is worrying.
Lack of an article on reweighted range vote on wikipedia also makes me
think it is niche. Range Voting alone is what is most commonly used in
all the serveys where there are no winners at all, and one is
assembling essentially statistics and opinions. There is an advocacy
websites for Range Voting which reads biased =)

I'm failing to find comparisons of RRV with STV (note _not_ comparison
of RRV with IRV) for multi winner elections.

> IRV successively discards the lowest-number-of-first-position
> candidates until one candidate has a majority of first-rank positions
> across ballots, with struck-out candidates not considered.
>
> RV chooses the candidate with the highest average (weighted, in this
> case) score.
>
> Let me note a few things.
>
> FIRST, these systems are not really so different when looked at from
> such a high level.
>
> SECOND, any problems with the underlying single-winner system will
> necessarily cause problems with the multiwinner system.
>

False.

STV uses a combination of two transfer methods. Therefore IRV
single-winner system deficiencies do not spoil the STV in the
multi-winner system case.

> THIRD, these systems were designed at very different times, with very
> different goals and constraints. IRV and STV were designed to be easy
> to perform by tired humans with trays of ballots. STV (and also IRV)
> was designed to be easy to explain to a Victorian voter using a
> natural intuition (namely that each person should get to vote for
> exactly one candidate, but if the person they really want is knocked
> out of the running they should get to switch that vote to someone
> else); and to be easy to process by hand using trays and sheets of
> paper. RRV was designed to be very easy to implement on a computer, to
> avoid pathologies as much as possible, to be amenable to mathematical
> analysis, and to perform well in simulated elections.
>

Somewhat False.

An age of algorithm tells little about its properties. The attack
against age and use of Victorian time reference comes across as ageist
discrimination / an ad hominem attack, rather than a logic statement /
conclusion.

STV is not easy to calculate by hand, due to requirements to
constantly track updated quota requirements after each surplus
transfer. And hence it is usually computerised, as it is defined in
the Scottish STV rules.

IRV is not being proposed for voting in SPI.

The assertion that "if the person they really want is knocked out of
the running they should get to switch that vote to someone else" is
true for IRV and incomplete for STV. If your top preference results in
being the first winner to reach quota, your ballot may be used to
determine the subsequent winners too. And there more chances for your
prefered candidate to receive surpus votes, because knock outs happen
much later for such candidates with STV system.

STV minimises the amount of wasted ballots, and gives more control and
influence for each ballot to contribute towards decision not only for
the top / most popular winner, but for the entire board.

--
Regards,

Dimitri.