The Schulze method that was being used is the one that is specifically intended to give only one winner; probably most people don't know that Schulze also created a separate system that was intended to give multiple winners. It is a very confusing system and many people unintentionally gave support to candidates they did not believe should have a chance.
One of the things that really becomes obvious using the S/N/O system is the number of *non-votes* or neutral votes: almost all of the candidates had more neutral votes than support and oppose votes combined. The effect of not requiring voters to decide how to classify each candidate (in Schulze, to rank the candidate; in S/N/O, to support or oppose) has radically different effects in the two systems. In S/N/O, the neutral votes have no effect at all on the outcome. In the Schulze system, not ranking a candidate is the equivalent of an oppose vote; every candidate who is ranked (even if they are ranked at a level well below the number of candidates) is ranked higher than a candidate who is not ranked at all. This is counter-intuitive and gives no effective way for people to differentiate between candidates that they really really do not think should be on the board and candidates about whom they have not formulated an opinion, or even candidates about whom they are indifferent. It is a serious weakness in the Schulze system. Nonetheless, the S/N/O system has significant weaknesses as well, as others have pointed out.
There are other systems that allow only as many supports as there are seats open, which might be worth considering. There are systems that only allow support votes and no opposition. There are not very many systems, though, that are specifically designed to give multiple winners when one of the conditions is that they *not* be running on a shared ticket.
We did not have enough time in 2013 (nor, to be honest, the interest amongst Election Committee members) to do a thorough review of multiple-winner voting systems. That year, we had to develop all of the processes for electing FDC members and FDC ombuds, which was a lot of work. This year, the committee barely had enough time to do the tasks that were absolutely required just to make the election happen, and in order to incorporate the specific instructions of the board with respect to outreach, seeking of diverse candidates, and increasing voter participation (all of which proved very worthwhile), they didn't have time to fine-tune a lot of the processes that were already developed. I would have loved to see changes in the way that questions are handled, and a rethinking of the voting methodology, for example. But there simply was not time to come up with a well-considered *better* way.
So...yes, I agree with Milos and many others that a Standing Election Committee is needed to re-examine the way that Board candidates are elected, and to re-examine the entire framework on which the elections are based - indeed, I recommended it after the 2013 election.
I find it interesting that nobody seems all that worried about the FDC election (where 5 of 11 candidates got seats) or the FDC Ombud election (where both candidates came forward in the last 24 hours before nominations closed). These two elections suggest some pretty big underlying problems as well. Nobody seems all that upset that fewer than 10% of all the candidates for the 2015 elections were women - one of the lowest percentages ever - and that not a single woman was elected to any role for the first time in any election where more than one candidate was being elected. On the whole, despite having a fair number of candidates outside of the US and areas represented by large national chapters, not a single non-white, non-male candidate, not a single Asian, African or Latin American candidate was elected. We're pretty good at talking about diversity, but very poor at implementing it.
Risker/Anne
On 6 June 2015 at 13:55, MF-Warburg mfwarburg@googlemail.com wrote:
I still think it was a big mistake (of the electcom? I don't remember, but /someone/ pushed it through without discussions) in the 2013 election to abolish the Schulze method. Am 06.06.2015 19:16 schrieb "Milos Rancic" millosh@gmail.com:
Moving this discussion into a separate thread, to leave the main one for best wishes and similar :)
Before I start talking about the voting system itself, I have to say that, from my personal perspective, I wouldn't imagine better outcome: a Polish steward (my favorite Wikimedian group :) ), a Croat founder of Wikidata (whom I consider as a friend) and a very prominent English Wikipedian, with significant record of working with smaller languages (BTW, I didn't know that he's a candidate till I saw the results; I didn't vote, as I still don't think I am able to make informed decision; useful note: one year out of movement requires more than one year to be able to fully participate again).
When I read the results for the first time, I thought that it's about structural changes. However, it was not. Present Board members were just punished as present board members (some people will always object your work) with negative votes, as well as Sj was punished with lack of positive votes because of his laziness :P
The problem is obviously the voting system. And it's one more reason why standing committee should be created. With more time, they would know why it's perfect for stewards and why it isn't for any kind of democratic representatives (including English Wikipedia ArbCom; as far as I remember, this is exactly the method how en.wp ArbCom is elected).
Stewards have to be trusted all over the projects and 80% threshold follows that idea. However, stewards are not reelected, they have to show to that they are doing good job and there is the space for those who are doing important, but not visible job. Bottom line is that stewards themselves decide if somebody would stay a steward or not. (If there were objections from the community.) And stewards are doing that job perfectly.
It should be also noted that stewards are elected managers, not democratic representatives, which Board members and en.wp ArbCom members are.
This system is bad because of two main reasons: (1) it isn't suitable for electing democratic representatives; and (2) it's very vulnerable to abuse, which could easily create negative culture.
Applying this to the democratic elections consistently means one of two things: we want to have conformists in the Board or we want to change Board members every two years.
I hope the first is not our idea. The second could be, but two years in office is too short period of time for a Board member to do anything substantially. So, this method would be a valid one if the term of a Board member would be, let's say, four years.
The output of the elections is not democratic, as well. It's obvious that Maria got the most support and it's 5% more than the first one, as well as Phoebe had more support than the second one.
While I think that opposing votes are important, they shouldn't be *that* important. Successful candidate had to gather 3 supporting votes for every opposing one. If the supporting and opposing votes have the same weight, it would be more fair.
With the formula S-O, the results would be:
- Dariusz: 2028-556=1472
- Maria: 2184-775=1409
- Phoebe: 1995-714=1281
- James: 1857-578=1279
- Denny: 1628-544=1084
And the results would be much more according to the expressed will of the community: Dariusz is well respected steward and community has given him a lot of support, and as he is a new candidate he didn't do anything which would annoy a part of the community. Maria had significant opposition, but also the biggest number of supporters, which has to be acknowledged. Phoebe and James would have been very close, while Denny wouldn't reach support threshold.
If one opposing vote has weight of three supporting votes, this could easily change the strategy of the groups interested to see one of their candidates as Board members. Instead of "vote for", we'd get "vote against" attitude. That's not just abusive toward the system, but also creates negative atmosphere, where candidates and supporting groups could start looking into each other as enemies, not as fellow Wikimedians.
So, while the current voting system has given refreshing results, it would be bad to keep it as it's now. To be honest, I would avoid negative votes at all, as I am sure that even more fair system would be implemented, if it contains negative votes next time, we'll get much more negative votes than this time, with negative consequences for our culture.
On Sat, Jun 6, 2015 at 2:16 PM, Gregory Varnum <gregory.varnum@gmail.com
wrote:
I have a lot of personal opinions on the method, questions process,
etc.
Many of them will be shared in the committee's post mortem (others I
will
be discarding as I now process the last several weeks).
Also, we are beginning to post some statistics that folks may find
helpful:
https://meta.wikimedia.org/wiki/Wikimedia_Foundation_elections_2015/Stats
We will be posting more on the blog next week about what all goes into running the elections, and I am open to feedback on what additional information we can share that would be helpful to the community. Our
group
made an early commitment to transparency, and I hope that has come
across
in our posting of major meeting minutes, posting of these stats, open dialogue on Meta and email, a post mortem from the committee, and the upcoming blog post.
Finally, I want to give a big thank you to my colleagues on the
Elections
Committee. I was, by the nature of my tasks, a bit more visible - but please know that everyone worked very hard, did a great job, and
deserves
equal gratitude. Thank you Adrian, Anders, Daniel, Katie, Mardetanha, Ruslan, Savh, and Trijnstel - as well as Risker, James, Alice,
Philippe,
Geoff, Stephen, Sylvia, Heather, Tim, and a few others I'm sure I'm forgetting.
-greg (User:Varnent)
On Sat, Jun 6, 2015 at 6:19 AM, Chris Keating <
chriskeatingwiki@gmail.com>
wrote:
Congratulations to the new Board members - I am sure you will do a
great
job. And commiserations to those who will be leaving the Board - thank
you
for all your hard work over many years.
Also it is good to see a much higher turnout in this year's elections
than
in 2013 - well done to those involved :)
On the subject of voting systems, though...
On Sat, Jun 6, 2015 at 9:08 AM, Anders Wennersten < mail@anderswennersten.se> wrote:
David Cuenca Tudela skrev den 2015-06-06 09:01:
However I must say that the results of this election are hilarious.
The
person with the most support votes doesn't win because of oppose
votes
:D
Why hilarious? We had a full consensus in the election Committee
to
go
for S/N/O voting, it is a kind of standard procedure in the
Wikimedia
world.
Many people looked at voting systems before the Wikimedia movement
existed
and virtually none of them settled on the system we ended up with.
Perhaps
this should tell us something!
To my mind the key problems with the present system are:
- Oppose votes have greater weight than support votes. In this case,
Maria
would have needed 136 additional support votes to win, or 46 fewer
oppose
votes. In effect an Oppose vote was worth 2.96 times as much as a
support
vote for her. As a result, being non-opposed is much more important
than
being supported. The penalty for doing anything controversial is significant.
- There is nothing in the process to produce any diversity in the
result.
Say that there was a 2/3 to 1/3 split in the electorate on some
important
issue. The right answer would surely be that you elect 2 people with
one
view and 1 with the other. However, in this voting system you would
likely
end up electing 3 people from the majority point of view. Because the Wikimedia movement is much more complex than this it is difficult to conclude that there was any particular issue like this that would have affected the result, but still, the point applies. The voting system
builds
in homogeneity not diversity.
Regards,
Chris _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
,
mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe
Wikimedia-l mailing list, guidelines at:
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe