[EM] STAR

[C.Benham]

Toby,

According to one definition of a "clone" is that they all get the same 
rating on a ratings ballot. STAR has ratings ballots that in the final 
stage it interprets
as ranking ballots.  Voters can't make any ranking distinction among 
candidates they score the same.

This would be much less egregious if many many more scores were 
available, say 0- 10000.

The Clone-Loser thing enabling a clone of the score-winner that pairwise 
loses to the score runner-up to displace the score runner-up out of the 
final
is relatively benign.

If we define clones as candidates that are voted together on all ballots 
with the minimum score distinction needed to make ranking distinction 
among them,
then the method fails Clone-Winner.

Frankly I can't see any place for "proportional" anything in 
single-winner method.  The point of  STAR is that it is very simple so 
people can understand how it
works and on-the-surface seems more-or-less benign and "democratic", and 
it isn't IRV which is much much better.

It isn't something that anyone serious looks at with a view to 
"improving" it with some relatively complicated kludge.

Chris

On 21/08/2023 10:20 pm, Toby Pereira wrote:
> Chris
>
> I thought one of your big problems with STAR was the cloning thing. So 
> the fact that score ballots might jeopardise the clone run-off would 
> be a good thing wouldn't it?
>
> Your approval opposition run-off is perhaps similar in outlook to what 
> I suggested previously about using a sequential proportional method to 
> elect two candidates to the run-off. Except that I would deal with 
> clones by allowing a single candidate (or their exact clone depending 
> on how you want to define it) to win both seats in the run-off. In the 
> case where single candidate has so much support that they could win 
> the first two seats in a proportional election, it would be deemed 
> that no run-off is needed.
>
> Toby
>
> On Monday, 21 August 2023 at 13:38:45 BST, C.Benham 
> <cbenham at adam.com.au> wrote:
>
>
> Toby,
>
>> Also a run-off between the most two approved candidates still has 
>> STAR's clone problem.
>
> Some years ago I suggested that 2-round Top-Two Runoff could be 
> improved by using approval ballots in the first round
> and then having a runoff between the most approved candidate (the AW) 
> and the candidate with the most approval opposition
> to the AW (i.e. is most approved on ballots that don't approve the AW).
>
>> If the most approved candidate is cloned, the run-off becomes 
>> irrelevant.
>>
> Spoken like someone who lives in parliamentist country. "Clones" 
> aren't necessarily identical.  There could be slight political differences
> or one may be less corrupt, or one could just have a much better haircut.
>
> Parties being having incentive to each field two candidates (even if 
> they are "clones") is maybe not too bad.  But STAR uses score ballots
> so there is a danger that there being two candidates from the same 
> party might cause voters to not give both of them max score enough
> to stop both of them from making the final.
>
>> However, I just don't think that STAR's failure here can reasonably 
>> be called a monotonicity failure.
>
> I think it is very much like one and it's claiming of bragging rights 
> on that point over IRV is unfair and misleading.
>
> Chris
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20230822/0b8ad234/attachment.htm>