Okay, I see the STV explanation. It's similar to an IRV process
(especially in that it iterates tabulation rounds by eliminating one
candidate at a time) except it stops when you narrow down to the number
of seats open, so it works for multiple-seat races, and need not invoke
parties (i.e., its not necessarily "proportional representation" per se
-- NCSG does not involve proportional representation -- all office
holders represent the entire SG, not some subset of the SG, and I would
suggest not changing that).
Dan
On 9/7/16 11:40 AM, Neal McBurnett wrote:
> STV (Single transferable vote) is designed for proportional representation, and is not the same as IRV.
> But, indeed, both of them use ranked-choice ballots, and I'm sorry to say that there is much confusion in terminology around the many methods that use ranked-choice ballots.
>
> https://en.wikipedia.org/wiki/Single_transferable_vote
>
> I agree that good ballot design and a well designed user interface for voting is critical when using any voting system. Unfortunately that it is hard, and ICANN may well not offer a good option there.
>
> I also agree that limiting the number of allowed rankings (done to simplify a paper ballot design) is not a good idea, and defeats some of the properties that ranked-choice methods offer.
>
> These are some of the reasons I brought up Reweighted Range Voting, which is easier to implement and perhaps simpler to explain.
>
> http://rangevoting.org/RRV.html
>
> But again, there are a number of hurdles, including the need to amend the charter, which itself requires more participation that we've gotten in recent elections.
>
> Neal McBurnett http://neal.mcburnett.org/
>
> On Wed, Sep 07, 2016 at 11:20:56AM -0700, Dan Krimm wrote:
>> When a new WG is formed, this discussion should move there. But for
>> now it remains here on the general list.
>>
>> One issue about STV (also known as IRV in the US -- instant runoff
>> vote, which is one way to tabulate such ballots but not the only
>> one) is that it is designed for single-seat races. Most of the
>> questions about the recent election had to do with the multiple-seat
>> election and the role of NotA.
>>
>> Just one point about IRV: in the San Francisco Bay Area this has
>> been implemented for a variety of local/municipal elections, but I
>> have great reservations about the local method because it limits the
>> vote to three candidates per ballot even if there are more than four
>> candidates running for the single seat (with four candidates, the
>> one not voted for becomes an implicit 4th choice). Thus, it
>> potentially disenfranchises many valid ballot choices (if none of
>> your three chosen candidates ends up in the final-round head-to-head
>> runoff contest, your vote is effectively irrelevant -- *even though
>> you showed up to vote and cast a ballot*). If there is any talk at
>> all of STV, it *must* be implemented with a full rank-order
>> preference on all candidates running for the office, or else it
>> undermines the whole purpose of that voting system (to allow
>> everyone a voice on the final match without being subject to the
>> split-vote effect ... usually ... ).
>>
>> To Paul's point about voter confusion with STV, the best way to
>> avoid that is with a firm UI that prevents misvotes upon input (and
>> explains errors when necessary -- a learning/teaching tool as well
>> as an input-cleaning tool), rather than a simple form to be filled
>> out like a piece of paper where all sorts of things can go wrong. I
>> doubt that ICANN would provide such a voting system for us -- we'd
>> have to build it ourselves.
>>
>> But honestly, I'm not sure if there is a way to design STV
>> tabulation for multiple-seat races -- never heard of such a thing.
>> Proportional system is more likely in that case, but that entails a
>> party-based system and we don't have "parties" in NCSG -- there are
>> only "independents" in our elections. (I would firmly resist the
>> idea of making the constituencies into "parties" in this context.
>> Better to push back against tribalism in our own ranks, rather than
>> systematically encourage it.)
>>
>> Range voting is an attractive notion, which also avoids the split
>> vote problem and some other issues as well (IRV has some potential
>> special cases that become counterintuitive). It's basically how
>> Olympics are scored with multiple judges per competition. And I
>> believe it could be applied easily with voter weights. Not sure
>> about multiple-seat races, though -- top-N winners?
>>
>> Would have to think more carefully if it accomplishes the mission of NotA...
>>
>> Dan
>>
>>
>> On 9/7/16 10:08 AM, Paul Rosenzweig wrote:
>>> I'd be curious Neal for your response to what I understand is the major
>>> complaint against RRV and SPV methods -- namely that voters often are
>>> confused by them. Perhaps this electorate is sufficiently attuned that it
>>> would not suffer that problem ... but am I correct that it can be a problem,
>>> I think, in other settings.
>>>
>>> Paul
>>>
>>> Paul Rosenzweig
>>> [log in to unmask]
>>> O: +1 (202) 547-0660
>>> M: +1 (202) 329-9650
>>> VOIP: +1 (202) 738-1739
>>> www.redbranchconsulting.com
>>> My PGP Key: http://redbranchconsulting.com/who-we-are/public-pgp-key/
>>>
>>> -----Original Message-----
>>> From: NCSG-Discuss [mailto:[log in to unmask]] On Behalf Of Neal
>>> McBurnett
>>> Sent: Wednesday, September 7, 2016 11:02 AM
>>> To: [log in to unmask]
>>> Subject: Re: voting/tabulation process for future elections
>>>
>>> I'd also like to be on the list, if it is created.
>>>
>>> Joonas, I have made the case for a Proportional Representation (PR) method
>>> to be used, and STV (a PR method) would be an improvement over the current
>>> approach, I think. Reweighted Range Voting (RRV) is another worthy option.
>>>
>>> Neal McBurnett http://neal.mcburnett.org/
>>>
>>> On Wed, Sep 07, 2016 at 12:09:32PM +0300, Joonas Mäkinen wrote:
>>>> I'd like to join the list too if such is created. Has there already
>>>> been a reasoning for/against Single Transferable Vote (SVT)? It gives
>>> great voter satisfaction and discourages tactical voting.
>>>> maanantai 5. syyskuuta 2016 Dan Krimm <[log in to unmask]> kirjoitti:
>>>>
>>>> +1
>>>>
>>>> Now is the time to resolve the election issues, while the topic is
>>> bright in our minds. Let's not put it on the back burner,
>>>> but instead push through and find the consensus. We've already had
>>> several suggestions about how to fix the process, let's
>>>> continue exploring.
>>>>
>>>> One suggestion that was made was to have a No vote for each candidate
>>> in multiple-winner races. There was multiple support for
>>>> that, but also a suggestion that that was not sufficient even so.
>>> Let's continue the discussion.
>>>> Dan
>>>>
>>>>
>>>> On 9/5/16 12:32 AM, dorothy g wrote:
>>>>
>>>> Congrats to all! and finally we can get to work on fixing our
>>> election regulations so that we can have peaceful and
>>>> transparent elections next time around
>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Joonas "JoonasD6" Mäkinen
>>>> www.joonasmakinen.com
>>>>
>>>> Vice Chairperson (international affairs), Pirate Youth of Finland,
>>>> www.piraattinuoret.fi Vice Chairperson, Alternative Party,
>>>> www.altparty.org
>>>>
>>>> Faculty of Medicine +
>>>> Department of Mathematics and Statistics, Faculty of Science
>>>> University of Helsinki
>>>>
>>>> mobile +358 40 700 5190
>>>> Facebook, Twitter, G+, Skype, IRC, Steam: JoonasD6
|
