[EM] A PR-STV calculator

[Richard Lung]

Helllo Kevin,

It is up to your list manager to decide whether my links to Binomial STV 
are safe. For instance Github suffered a massive malware attack in 2023. 
I hope he will be forthcoming on this. I myself have no technical 
expertrise.

With regard to Binomial STV being both an election count and an 
exclusion count (first order binomial stv is just one election count and 
one exclusion count). One votere election preferences are another voters 
exclusion preferences. So there is no difference in principle as to how 
they should be counted. This meets the scientific requirement that there 
is but one truth to aspire to, and only one way to count both election 
and exclusion.

One can see clearly that election and exclusion are on the same footing, 
by looking at the history of two-party systems. A prime motive of the 
big two parties is to exclude third parties as a "wasted" vote. And a 
prime motive of third or other parties is to exclude either of the big 
two parties by "strategic" or "tactical" voting. So, really what we have 
been seeing in two-party states is not so much "elections" as 
exclusions, among dissenters from a two-party system.

The limitation of the two-party system is that some people are only 
allowed to have exclusions rather than elections. Conversely some voters 
only have elections and not exclusions. Binomial STV overcomes this 
limitation, by allowing the voters to have both elections and 
exclusions. The most basic, first order form, which should be sufficient 
for all voting purposes, consists of one election count and one 
exclusion count, which is just an iteration of the election count, but 
to exclude, rather than elect candidates.

If a candidate reaches an election quota, they are elected, in the 
normal way. But if a candidate reaches an exclusion quota, they are 
excluded. If a candidate is both elected and excluded, they are a 
(potentially) famous case of what Forest Simmons calls a "Schrodinger 
candidate" (both "alive" and "dead" to the electorate). In this case a 
"Quotient" of the exclusion count to the election count decides the 
issue. If unity or less, the candidate is elected. (The quotient is the 
square of the geometric mean exclusion and election keep values.)

The Binomial STV computer count is why we have a division of labor. But 
the voters do have to have a certain amount of instruction (as is the 
case, more or less, in all elections). For example, take a vote of 10 
candidates competing for 5 seats. Voters have a maximum of 10 
preferences. Their first five preferences (should they choose to make 
them) will more or less help to elect all the candidates. But their next 
five preferences (should they choose to make them) will less or more 
help to exclude candidates. Thus a tenth preference counts as much 
against a candidate, as a first preference counts for a candidate.

But the voters may leave out any or all preferences. Abstentions go 
towards an abstentions quota, which leaves a seat unfilled. Thus 
Binomial STV meets a first principle of natural science: the 
conservation of (preferential) information. This is not a returning 
officers election, by which the presiding official has to make sure all 
the seats are filled, regardless of whether the voters like the 
candidates. A carte blanche (as in French presidential elections) is 
equivalent to NOTA.

Regards,

Richard Lung.



On 19/03/2024 01:22, Kevin Venzke wrote:
> Hi Richard,
>
> Yes, I did see your post about a contractor who implemented Binomial STV, but from
> the thread it seemed unclear whether you agreed with his implementation.
>
> Normally in a multiwinner proportional method the concept is that each elected
> candidate will represent some section of the voters. It's hard to understand what
> the role of exclusion votes should be in this context. Do you consider Binomial
> STV a proportional method? Is the purpose of exclusion votes that one voter should
> be able to affect which candidates are elected to represent *other* voters?
>
> Kevin
> votingmethods.net
>
>
> Richard Lung <voting at ukscientists.com> a écrit :
>> Hello Kevin Venzke.
>>   
>> I've just hired a programmer of a binomial stv that is the first voting
>> system for the voters to rationally exclude, as well as elect,
>> candidates. Present voting methods are two-truth systems; not just stv,
>> is inconsistent or illogical; all don't have the same rules for electing
>> and excluding, and none of them allow voters to rationally exclude
>> candidates, on the same rational terms as electing them.
>>   
>> I posted the links to the list. But the moderator intercepted them and I
>> have not heard from him since.
>>   
>> Regards,
>>   
>> Richard Lung.
>>   
>>   
>> On 17/03/2024 16:16, Kevin Venzke wrote:
>>> Hi all,
>>>
>>> Not sure if this is of interest but I thought I'd try my hand at a PR-STV calculator:
>>> votingmethods.net/stv
>>>
>>> I included a couple of aids to visualize the outcome, in particular an indication of
>>> which preferences on each ballot contributed to someone's election.
>>>
>>> The Wikipedia STV example computes properly and is included on the page for
>>> convenience. The topic is not my strong suit, but hopefully I've done it justice.
>>> (If not, I'll fix it.)
>>>
>>> Kevin
>>> votingmethods.net