# Talk:Affiliate-selected Board seats/Resolution 2019

## Proposed change to point 2

Hello. I would like to propose a change to point 2, as follows:

There shall be at least three Election Facilitators who are jointly responsible for overseeing and carrying out the election. One facilitator will be the liaison for the Board of Trustees. The elected liaison will also have a deputy to assist in case of need. Election Facilitators shall not be candidates in the election and shall not endorse any candidate nor shall they participate in any public discussions of candidates' merits. Facilitators can steps down from their role and nominate themselves no later than April 6, but no person who has played a substantial role as facilitator in a selection process shall be eligible to be a candidate in that election. Election facilitators are expected to recuse themselves from any decision about how an affiliate organization's vote is cast. Any complaints about the Facilitators' conduct should be submitted to the Trust and Safety team at Wikimedia Foundation: ca@wikimedia.org. If, after an investigation, there is clear evidence of misconduct, the Facilitator will be removed from their role.

My rationale is as follows:

Being an election facilitator is a form of power and is a platform. As a facilitator you send communications about the election, you hold the floor at the Wikimedia Summit to introduce people to the election process, you encourage people to participate. That platform is not given to candidates in the election. There is no session for anyone considering standing to get up on stage and ask how they can help chapters and user groups participate. Of course, it's fine for candidates to talk to affiliates / user groups about how they are casting their vote, but that communication is heard differently if it comes from a neutral facilitator, than if it comes from someone who is standing.

Any candidate who starts off acting as an election facilitator and then changes their mind halfway through therefore gets an unfair advantage in the election process. This undermines the validity of the process. Thanks! Chris Keating (The Land) (talk) 19:05, 3 April 2019 (UTC)

It is important to note that this change was transparently discussed with facilitators and approved by them due to the extraordinary circumstances of this election, the almost impossible timeline they faced and the lack of volunteers. It was then discussed in the Berlin Summit and approved by the majority of approximately 120 participants, as can be seen in the etherpad from the session. The majority approved April 6 as a final date, understanding the unique circumstances the facilitators faced and the unorthodox nature of this elections. Please consider that ignoring these decisions reads like an attempt to undermine the success of a truly inclusive process. Best, Shani Evenstein. 19:21, 3 April 2019 (UTC)
Clarifying what constitutes a conflict of interest has unfortunately become necessary, so I second the proposal. I don't believe this can apply retro-actively anyway and I doubt anyone will adopt this change before April 6. Philip Kopetzky (talk) 20:11, 3 April 2019 (UTC)
I think others have responded below to most of what you said already - I just wanted to point out that having a majority does not make the Conflict of Interest go away. I'm quite disappointed that the separation of duties is not understood by everyone, it is in my view one of the cornerstones when aiming for equality across the movement. Philip Kopetzky (talk) 11:33, 4 April 2019 (UTC)
Hi Shani, To be honest, I am raising this issue because I would like the remaining coordinators, and the affiliates generally, to reflect on the matter for more than five minutes in passing in the middle of a meeting. I am also a bit unsure whether you are speaking here as an ex-facilitator, or a possibly-candidate, or both - which is exactly why the two roles need to be kept distinct in my view. Chris Keating (The Land) (talk) 20:42, 3 April 2019 (UTC)
At risk that my comment here will also be used as a proof for a "chapter cabale", I still want to point out that for me a good or strong potential board member should demonstrate leadership by building bridges between stakeholders in the movement and not by reframing valid concerns about the process (not the goal of inclusivity) to further fuel a conflict in the movement and increase bad faith among affiliates. I think there is a lot of common ground to build on, and we should also not forget that many leaders in user groups, particularly those who will be eligible to vote - including you Shani - are or have been part of chapters. Let's focus on what unites us and not on what divides us. Thanks! P.S: Many of us, including Philip, Chris and me and other working group members were in other meetings and not in the room that day, because we trusted our colleagues from ALL affiliates to make the right decisions and because were were not aware of a potential conflict of interest with the facilitiators, as we assumed the standards of the past, as Chris descrribed above, would also apply this time. And a lack of resources (people etc) and time imho is not a good reason to have a flawed process. --CDG (WMAT staff) (talk) 08:38, 4 April 2019 (UTC)
Hi Shani - thanks for your thoughtful response - I still do not really agree, but I think all the points I would make have been made by other people, so I won't attempt to prolong the conversation even further. Chris Keating (The Land) (talk) 17:09, 6 April 2019 (UTC)

Hi all, some good points made. I think Claudia is right to point out that due to the format of the summit, the ASBS session was missing the working group members and I think there was some important input lost as a result with half the expertise of the conference in other rooms.

I was in the ASBS session and I am unsure how conclusive it was. In terms of the votes, often half the people in the room didn't put their hands up to support or oppose resolutions, showing that many were not engaged in the process. As someone with the movement for many years now and with a fair understanding of governance, I certainly found it confusing.

To be blunt it wasn't planned well or facilitated well. A member of the audience was drafted into assisting with presenting, and Christophe was required to establish what the WMF board actually did. Indicating that this might not be the best forum to dive straight into detailed analysis of board election processes and decisions on them. It overran and certainly needed more time and cooler heads.

I think more discussion would be a good supplement to the session, and I largely agree with Chris's point that there is a COI if a facilitator can run; it's politics, and just being recognised and known gives a huge advantage regardless of qualifications or track record. Anyone running and set on winning would be mad not to take the role of facilitator. I can't remember exactly how the votes went, but this was certainly one of the more heated topics of the session.

As such, I think flagging this up to all attendees to discuss this (or any other resolutions that stand out) here is a good idea. This is obviously urgent given timelines, and keeping it short is key. But if the decisions were clear then, they will remain so. Ironically, I am not sure how best to facilitate this :-). Battleofalma (talk) 09:40, 4 April 2019 (UTC)

I was present on the meeting in Berlin that day and I think I was able to understand what this election means to us as user groups participants. I think it was not only a heated discussion but also a sign that most people want the best for our movement, and a fair treatment to all of those who feel themselves a part of what Wikimedia stands for, and also a sign that there could, of course, be flaws in the process and that no one of us want to let go through because it would mean that all that we want to achieve together could go wrong in the least expected moment. In my humble opinion, there is no time for us to sit and point fingers to potential participants that could raise their voice from their own understanding of what a being a user group is or what being a chapter is. The process is of course perfectible and there will be chances to learn from this not only in the way at this time, but also along the way in the future, but right now we need not to lose focus about what is important just because we have near sighted views on what could be fair. I think it is fair that having agreed upon on one date for anyone to rethink their position as a facilitator, they can have a chance on being part of the selection. I wouldn't vote for anyone that's been in that position just because of mere exposition or because their name sounds familiar to me: My community (user group) and me will have to take the time to discuss internally and think about who we need to support if we want our opinion to be heard and to make it count on the BoT, and that has nothing to do with sending mass messages to mailing lists or standing in front of people once or twice. I think if we do need someone who has a strong view on what being in the BoT means, and also who understands what being part of the movement is, having faced the same struggles user groups and chapters go through, and is smart enough to find solutions from their seat, my community and I will vote for that person because they will have the ability to understand what we need and will do their best to serve not only their own community or mine, but the movement as a whole. That's why I think nobody should be excluded from the election if the rules are already set and agreed upon. --Edjoerv (talk) 13:29, 4 April 2019 (UTC)
I wonder when the above recommendations will be included into the draft document. It is close to April 6 and I understood this is the date that this document will be send to the BoT for final approval. Grijz (talk) 19:33, 5 April 2019 (UTC)

### Final resolution

So I saw that the facilitators posted this message on Wikimedia-l announcing the finalisation of the resolution, which included this comment on this discussion:

"On the talk page of the resolution one issue was raised. The issue looks like to be about a possible candidate. Affiliates will have ample time to discuss the merits of candidates during nomination time, screening time, and they can cast their votes on candidates. The Election Facilitators didn't see the necessity for this change, and left the resolution on this point unchanged"

I just wanted to post that here in case anyone reads this page and wonders what the outcome was, as that will probably be the final word on the subject. I am glad the facilitators considered the matter, and though personally I still have some concerns about it, clearly the important thing now is to make the election as high-participation as possible with as strong a set of candidates as possible. Chris Keating (The Land) (talk) 17:09, 6 April 2019 (UTC)

## Facilitators

Shani Evenstein is removed as facilitator because she stepped down on April 1. Here name is also removed from this page. Shani played an important role during the first phases of this process and I think her name should be retained in the draft, to be followed by the date of her resignation. Also please add he start date to the names of the new facilitators. Grijz (talk) 19:33, 5 April 2019 (UTC)

## STV counting method

The Election Facilitators kept the wording of the STV method the same as used in the previous resolution. They intend to use the same script to count votes as was used in the 2016 election. As a matter of transparency we are looking to publish that script as open source somewhere. Ad Huikeshoven (talk) 10:24, 6 April 2019 (UTC)

Obviously too late to put in the resolution, but it might be worth the facilitators considering how to resolve ties. Ties can potentially happen at many stages of an STV vote, particularly if there are many candidates and relatively few voters. The "usual" method is to break ties in favour of candidates who received the most first preferences, and if there is still a tie then to randomly resolve it. Another option is to resolve the count both ways, and then see if the tie makes any practical difference (often, it will not). Worth considering how this might be approached. When facilitating this process before, I always had in mind that if there was a tie that turned out to be decisive between 2nd and 3rd places then I would propose reporting the tied outcome to the BoT and having them decide whether e.g. one candidate met their criteria better or whether to proceed by drawing lots. However, this was never agreed, I just offer it as a thought. Chris Keating (The Land) (talk) 17:18, 6 April 2019 (UTC)

This is really unlikely to be a real problem. If multiple candidates have the same sums of votes in a certain step of the process:
1. if they are the least-voted and in this step no candidate reach the quota, they all lose and the process continues.
2. if they are the most-voted and they reach the quota, they all win. Note that by definition of Droop quota the number of winning candidates cannot exceed the available seats.
3. in any other case, the process continues normally.
The process will always return a list of winners, no more than the available seats. This is how the script we used in the past is implemented.
There is only one possible problem I can think of, and it's that it is possible that less candidates are selected than the seats available. This is a problem of the Droop quota; the solution is very easy but requires using a slightly different quota.
To make an example, if we have 120 voters and two seats available, then the Droop quota is 120 / (2 + 1) + 1 = 41.
If 41 organizations vote for candidate A, 40 for candidate B, and 39 for candidate C, then A is elected, but neither B nor C reaches the quota. This is very visible if you have a small number of voters, but with 40 or 120 is highly unlikely to happen and I wouldn't worry about this.
The solution is that instead of saying that a candidate is elected if the sum of its votes is at least: ${\displaystyle \lfloor voters/(seats+1)\rfloor +1}$ we say that a candidate is elected if the sum of its votes is higher than: voters / (seats + 1).
This is simpler, fairer, and avoids stalemates except for the cases in which they are unavoidable (e.g., 40 votes each).
In the past I did not to propose this approach because it is non-standard and would have meant departing from what was written in the resolution, but I think it is technically better from any point of view. - Laurentius (talk) 09:49, 10 April 2019 (UTC)
In case at a step all candidates have the same sum of votes and the number of candidates exceed the still available seats, there is a tie. Common practice is to draw a lot to determine the winners. However, the resolution does not specify what to do in case of a tie. The script of Lorenzo will report "STALEMATE" in case of such a tie. So, the process will *not* always return a list of winners. Ad Huikeshoven (talk) 17:53, 6 June 2019 (UTC)
Correct. As I said above, "avoids stalemates except for the cases in which they are unavoidable". In the unlikely case of perfect parity, the facilitators would have to propose a way to solve the tie. - Laurentius (talk) 18:12, 1 August 2019 (UTC)

## Variables for links to Wikipedia

Hi,

You’re using variables for translations of links to Wikipedia (`\$stv` and `\$droop`), but the target pages have equivalents in other languages (in my case there’s a French article on the single transferable vote). Shouldn’t we be able to put localized links? If yes, you should remove the variables and simply put the links.

Thanks,

Frigory (talk) 17:31, 7 April 2019 (UTC)

Hi Frigory. I think I did like you said, but didn't want to invalidate existing translations. LMK if that works, and thanks for your remark. --Elitre (WMF) (talk) 06:48, 10 April 2019 (UTC)
@Elitre: It works. Thanks! -- Frigory (talk) 15:12, 13 April 2019 (UTC)

## Why not IRV?

--Liuxinyu970226 (talk) 03:51, 9 April 2019 (UTC)

Instant-runoff voting (IRV) is specifically designed to result in a single winner, and this election is for two seats. IRV does not allow for ranking of candidates with an equal preference, as is permitted with the STV system being used. IRV is rather strongly linked to the donkey voting problem, although I do not believe that would really come into play in an election of this nature, where even the "voters" (i.e., individuals who will vote on behalf of their group) are carefully selected and well-informed. Risker (talk) 05:02, 9 April 2019 (UTC)

## Board approval

Hi all, this is to inform everyone that the WMF Board has approved the resolution with no changes and we can proceed with the next phase. (Ping Ad.) Kind regards, Raystorm (talk) 11:57, 12 April 2019 (UTC)

## Closed-source Qualtrics software

It is quite problematic that closed-source software is being used for this. In the future, I think the affiliate organizations should specifically make sure to use open-source software for managing voting. --Yair rand (talk) 19:09, 14 April 2019 (UTC)

Do you have any suggestions about open-source software that would work for this? Chris Keating (The Land) (talk) 20:02, 14 April 2019 (UTC)
@The Land: Any reason Mediawiki's own SecurePoll couldn't work for this? --Yair rand (talk) 04:29, 18 April 2019 (UTC)
The same that are explained on that page and on the talk page - very old, unmaintained code that would have required a lot of dev's time over several weeks. I'll be sharing more info on the topic once this selection is over, and I'll be glad to assist anyone who wants to look into this closely to make sure that a more satisfying solution is found for the next run. --Elitre (WMF) (talk) 09:10, 18 April 2019 (UTC)
In general, I agree that better, open-source software should be used. I spend much of my life working on election integrity. But I'll note that since it is not a secret-ballot election, and the votes will be published along with the results, and voters do get an email indicating how their vote was recorded, many of the security issues are much easier to deal with. In the future, the Helios system would be excellent for this if we adopt a proportional-representation method that doesn't require ranking of ballots. Other options are out there which we can check out after the election. ★NealMcB★ (talk) 17:45, 30 April 2019 (UTC)

## Minor fix

The resolution point number 6: Each nominee must have two (and only two) endorsements from any type of Affiliate in order to be a validate candidate. Shouldn't this be valid candidate.? –Ammarpad (talk) 16:20, 28 April 2019 (UTC)

Done. Translations take some time. Ad Huikeshoven (talk) 15:58, 29 April 2019 (UTC)