Wikimedia Foundation elections 2015/Post mortem

From Meta, a Wikimedia project coordination wiki
Info The election ended 31 May 2015. No more votes will be accepted.
The results were announced on 5 June 2015. Please consider submitting any feedback regarding the 2015 election on the election's post mortem page.

Add your feedback to this page[edit]


A few thoughts on the elections[edit]

Thanks to the Election Committee for the organization of the election! As a candidate for the Board of Trustees, I was thinking of the elections a fair deal, and have written down my thoughts here. Sorry for point to another page instead of simply writing it here, but I thought it was too long to simply drop it here. Feel free to change that if you want, though. --denny (talk) 01:39, 1 June 2015 (UTC)[reply]

Questions to candidates[edit]

Three points here:

  • The longstanding process of requiring that any question be put forward to all candidates has become increasingly unwieldy as time has gone on. There are too many questions and too many candidates, and the quality of most of the questions is poor. Many are irrelevant to the role they are seeking (the only advantage of them is that readers who do understand the role might identify candidates who are out of their depth), failure to answer even irrelevant questions can be viewed as the candidate not being "serious" enough to answer all questions, and frankly it is poor use of the time of voters to read a lot of this.
  • The longstanding restrictions on asking questions of specific candidates is also proving itself to be less than helpful. I could think of several questions that were appropriate only to individual candidates, but irrelevant to the majority of candidates (Denny has pointed some of them out in his own retrospective), and the need to make the questions generic enough to apply to all meant that the community did not get the information it deserved. (I am intentionally writing this well in advance of the results being generated.)
  • I saw some red flags in the questions that some candidates put forward for response by their colleagues, as those questions seemed to be deliberately phrased to support their own platforms. At the same time some of those questions were on-topic and brought forward information that would potentially be useful to voters.

Suggestion for future elections[edit]

  • That questions to individual candidates that are specific to their own situation be permitted. The process should probably require some advance vetting, and will require some changes in the formulation of the question pages.
  • That questions to all candidates be vetted prior to being posted to ensure that they pertain directly to something that is within the scope of the role. This should apply to all of the positions that are subject to the election; there were several questions directed to FDC/FDC ombudsman candidates that were outside of the scope of those roles as well.
  • That consideration be given to developing some mandatory questions for Board of Trustees candidates that is specific to the role, similar to what is done for FDC/FDC ombudsman candidates.
  • The practice of candidates essentially advancing their own platforms by questioning their colleagues raised some serious issues; however, the basis of some of those questions were indeed quite legitimate and would have come across differently if raised by neutral parties or worded in a more neutral manner. This may work more effectively if there is a neutral body or group ensuring that questions are within scope and are neutrally worded.

For consideration, Risker (talk) 04:40, 1 June 2015 (UTC)[reply]

I agree with what Risker wrote above, and would like to add a few remarks. Having all the questions (without the answers) in a single page (or a few pages) is OK. But showing the answers of all candidates in the same page was not very helpful because readers who are only interested in a few questions cannot easily jump to the corresponding answers for a given candidate. Even though I was only interested in the questions related to the resolution of conflicts in WP it took me a long time to read the answers and find which candidates had understood why so many editors give up editing. It would have been much easier if each question was assigned a number and if the presentations had numbered links to the answers.
BTW, regarding the topic that interests me, my feeling is that when a novice editor has a conflict with some experienced editors, the administrators and bureaucrats that handle it do not take the time to listen to the novice editor and tend to believe the experienced editors. IMHO spending enough time in understanding the conflicts is the key to keeping new editors active. Ekkt0r (talk) 12:40, 1 June 2015 (UTC)[reply]
I disagree with the above "in scope" suggestion, at least with respect to the WMF board (I don't know enough about the FDC). The role of the WMF board is an oversight one - it may be called upon, or may independently seek, to oversee *any* aspect of the WMF's work or the work of the communities (the latter may seem surprising to some, but the topic came up in the questions in this election, and produced some interesting answers). It is imperative that WMF staffers and Wikipedia administrators/buereaucrats not be involved in "vetting" questions that are submitted for candidates, because their opinion of what is "in scope" might not align with the actual electorate's opinion. Even when legal or time issues prevent a board member becoming too deeply involved in a topic, the board may still be able to give general, broad input on that topic. Essentially, there is no-one who is sufficiently independent to vet the questions, and the questions should not be vetted in any case.--Greenrd (talk) 11:19, 7 June 2015 (UTC)[reply]

Translation[edit]

For the first time this election, funds were made available to the Election Committee to obtain timely professional translations of selected key messages (including the Board of Trustee's letter encouraging a more diverse candidate pool, banners, questions and responses, and ballots) into a large number of languages. The results are evident: the 2015 candidates came from a much wider range of projects, geographic locations and native languages, and the voter participation from projects outside of the main language groups is significantly higher than ever before. Editors from over 180 projects participated. This was not without glitches:

  • There were some odd translations of some information
  • Some editors mentioned that translations to certain languages were more formal and/or did not take into account gender differences in the usual "wikimedia" way
  • It was not financially possible to obtain professional translations into all Wikimedia languages
  • It does reinforce that sometimes our industrious and valued volunteer translators are sometimes not always able to meet the challenges for major organizational communication

It would be useful to see the perspectives of participants who do not speak English as a first language to more accurately gauge the value of the timely translation of election information; however, at this point it appears that ensuring these translations were available in such a timely way has had a significantly positive effect on both candidate and voter participation. Risker (talk) 05:48, 1 June 2015 (UTC)[reply]

From what I have seen myself and what I heard of others' feedbacks I'd say the professional translators were incompetent for translating wikimedia related texts (though things like messing up with candidates' gender is to general incompetence rather than wikimedia specific one). The thing is that virtually the only people competent for translating wikimedia related texts are wikimedians themselves. We do know what we actually mean by these words "common" people will understand another way. We are in this cauldron for years and even when we don't know something ourselves we know where to find or whom to ask. I think it's crucial next time to focus on translation by volunteers. Before wasting donors' money (there's little of adding to free knowledge by candidates' "speeches") there must be used all means to involve translators from community. There were some letters to translators' mailing list and stuff but I think it's not enough actions when on other scale of weights is wasting money. Another thing was weird way of approving translations. First of all while we already have Extension:Translate on wikis and TWN, there was created a new thing for some reason. Secondly the translations of interface got shown just after some approval (come on we all have access to vandalise main pages of large wikis if we'd like to go to the dark side but we're not to be allowed to translate a candidate's speech or interface and have it approved automatically?) The approval took days if not longer and it's not what encourages to work. AFAIK people finally had to track people with acc on votewiki directly (e.g. via IRC) in order to have them approve a translation. And what's all this for? Anyway the votewiki accholders don't know all the languages in order to check the translation text. It's not in AGF mood but it looks like it all was just to make life of translators worse. To summarise it all I'd say that while it's obvious that translations are important it's crucial to focus on involvement of volunteers and of making comfortable work conditions for them, not for some outside guys who'd get money for badly translating a few messages (I wager they got more than e.g. volunteers who participate in translation rallies on TWN though the latter's work is more important and useful). --Base (talk) 17:58, 3 June 2015 (UTC)[reply]
I don't think it's about wikimedianness necessarily. The Italian translations often misinterpreted English grammar. --Nemo 18:03, 3 June 2015 (UTC)[reply]
If just the original texts were written considering that they have to be translated into languages which sometimes miss equivalent concepts (not only single words) for their expressions, we'd all be quicker in translating and certainly we'd make less mistakes, there would be plenty of volunteer translators and we wouldn't need professionals. We can make mistakes on our own, we always proved to be more than able, and for free :-) Legal systems can be very different, and this is a more specific issue, and legal jargons too, so anything slightly recalling a sort of legalese, actually becomes divisive - matter of fact - instead of working for a deeper unity. And cultural biases can go farther than that. Just consider that - to make a silly real time example - I simply cannot translate the "post mortem" locution (Latin for "after the death") because in Italian culture it can have a meaning only if we are talking about autopsies, which I do believe has nothing to do with how WMF handles the lucky guys who were elected :-)
I know that sometimes one just can't be simple in choosing the words that would be the most proper ones for a complex concept, sometimes you have no choice, and that's ok; but if you start writing without bearing in mind the KISS principle (Keep It Simple Stupid), on my side I receive texts to translate in which not only the unavoidable complexities are complex, and if I'm not in a good mood at that moment, maybe I can find something else to do in another Project, who knows if I'll ever go back to that page; and how, in case.
The wider the audience, the lower the syntactic level should be in order to be correctly received and rendered, because it cannot be rendered in a complex form in all the languages: there will be languages in which only half of the concepts will be close to a word-for-word translation. Keep it simple, and no one needs to be stupid for that, keep it essential, in this case we could rely on a much higher potential number of "wikimedian" translators, who really know what we are talking about, and they know it for free ;-) --g (talk) 15:13, 4 June 2015 (UTC)[reply]

Perspective of a non-English speaker[edit]

As a native speaker of Russian and Ukrainian, I could both evaluate professional translations and translate as a volunteer, thus I will make feedback on both.

Professional translations
  • Pros
    • Very helpful for non-English speakers, much better than nothing
    • All statements were translated at the same time, thus we avoid situations when only some candidates have their statements translated.
    • Some translations are of high quality and represent ideas really well (some of statements written in rather poor English were almost poetic in Russian)
    • Translations were published on time owing to huge volunteer involvement by Elections Committee (but honestly speaking, I would prefer seeing these translations published by WMF staff or even by translators — why not to ask them to edit directly on wiki, Translation extension is quite intuitive)
  • Cons
    • Unclear criteria for choise of languages. The criteria should be number of eligible voters who are first- or second-language speakers, not number of speakers. I do not understand how Swahili with 70 active users (thus at most a few dozens eligible voters) got a professional translation, while Polish with 4,000 active users and several hundreds (if not a thousand) eligible voters did not get a translation. I do understand that Swahili is a Global South language, but next time priority should be given to languages with most voters should be chosen.
    • Really anti-wiki spirit: professional translations cannot be edited at all. They are simply not on Meta, and the only way to edit them is making request on the talk page. It would be great to have some easier mechanism of editing them (at least on Translatewiki if one needs to make them editable not by everyone).
    • Volunteer translations cannot overwrite professional ones. Volunteers can add their improved translations on Meta, but this does not replace professional translations, which decreases motivation for improving translations. Same as above: there should be a possibility to replace professional translation by a better one.
    • All translations were done out of context. It was quite common to see strange things like "Wikimedia in French" or "Wikipedia France", the most badly translated word was probably "affiliate". Professional translators should be given a small glossary of wiki-terms (like an explanation that affiliates are organisations, APG grants are to organisations and IEG grants are to individuals) or adviced to look on Meta, as terms like AffCom or FDC are translated there.
    • Even worse, translators were not told that we have female candidates. Almost all languages have some form of gender. It was a ridiculous mistake just a few days after gender gap campaign, and reading that "Maria is a man who founded LGBT wiki project" or "Phoebe is a mail librarian" is just unacceptable. Statements by female candidates should be flagged for translations.
    • And even worse, some translations were really poor and worse than volunteer ones at previous elections. In my view, 1/2 of Russian translations were done by a really good translator, and another half was done by a really bad translator (which was quite unequal). The problem was not only with wiki terminology, but even with common sentences (e.g. in French Maria Sefidari founded English Wikipedia...)
Volunteer translations
  • Pros
    • Sorry, but none this year, except Translate extension
  • Cons
    • It was not clear which languages were chosen for professional translation. It should have been announced well in advance which languages will have professional translations and that the remaining ones can be translated only by volunteers (it was possibly done on some mailing list but it should be done on Meta and perhaps in local village pumps).
    • Not enough promotion for translations into local languages. While translating CentralNotice was promoted via CentralNotice, translating voting interface was not.
    • Volunteer translations were not published automatically, one had to find a person with a VoteWiki account and make them publish their translations. Same as above: should be moved on Meta or Translatewiki next time.
    • Messages that were displayed on VoteWiki were not grouped together. The group of high priority messages included all question pages and quickly became unusable. Statement pages included some random staff like signature of Francis Kaswahili and did not make any distinction between long (display on Meta) and short (display on VoteWiki) versions. As only 100% translations were published, I had to translate twice as many messages to make sure I have not missed anything. It would be really great to create a group for core messages displayed on VoteWiki
    • Once again, female candidates were not flagged. It was possible to find them manually, however, which slightly improved the situation.

Overall, great idea to add professional translations, but it was a kind of superprotect: professional translations could be edited by very few people, while volunteer translations could be edited by everyone but required approval to be viewed by everyone. That's quite anti-wiki and has to be changed next time. Sorry for the long message and thanks to the Election Committee for their job — NickK (talk) 22:17, 4 June 2015 (UTC)[reply]

Thoughts from an English speaker[edit]

  • Remind candidates to use translation-friendly language
  • Encourage short (50-100 word) statements and responses. Provide a space for links to longer responses for those who have extensive thoughts.
  • Highlight the shortest statements: e.g., start with the shorter version of the overall statement, then the longer one (then perhaps a link to full details or other essays)
  • Our 'professional' translations have always been spotty, outside of a couple core languages. They are fine for a first draft, but the final editing and proofing should be done by community members.
  • Have all translators, pro or otherwise, work directly via the Translation extension.
  • Build our community of translators: every opportunity this arises, post to some of the global translation sites welcoming volunteers to help. [This is also a fine way to introduce bilingual editors to our communities and the fact that they can edit wikis.]

SJ talk  19:14, 7 June 2015 (UTC)[reply]

Nearly missed board elections[edit]

Because elections for the board were held in continuation of elections for the FDC, I nearly missed the fact that there was an election for the board (I haven't participated previously). After I had checked out the banners for the FDC election shown on Wikipedia, once the banners for the board election came up, I initially did not perceive the banners had even changed. If the elections had been held at different times of the year, so any banners and announcements would not have been broadcast consecutively, I would have been more likely to have noticed both.--Anders Feder (talk) 23:34, 1 June 2015 (UTC)[reply]

  • It may be reasonable to hold Board elections in odd year as now, but FDC elections in even years. In such an arrangement elections would be held every year, which would be good as it would avoid loosing election committee expertise. Ruslik (talk) 14:20, 13 June 2015 (UTC)[reply]

Signpost coverage[edit]

Hello, I think that the Signpost was very helpful by posting these articles on the candidates. The format with several simple questions was very helpful to find candidates that I disagreed with (which wuere more than I expected).

Rather than posting complex questions with long answers, I think that a simpler questionnaire is more helpful. --NaBUru38 (talk) 01:56, 2 June 2015 (UTC)[reply]

Engaging local community[edit]

You have probably heard that Ukrainian Wikipedia has the highest percentage of eligible voters who participated (about 25%), and that's how we did it.

  1. Translation of statements into local language (on Meta). Reading 1200 * 20 = 24 thousand characters in English is too much for non-native speakers, if you want your community to read the statements, translate them.
  2. Local page with information about candidates in local language. We tried to summarise the most important information that was spread between candidate information (languages and countries were not translatable), statements and questions, for those who don't want to read a lot: uk:Вікіпедія:Вибори Вікімедіа-2015
  3. Announcement in the village pump with a quite lively discussion. It is important to make it more appealing than quite generic MassMessages (which, in addition, were sent in English). It was quite good that people started to write about candidates they know and recommend and candidates they do not recommend. I don't know whether it's good or bad, but it likely influenced other voters: Permalink.
  4. Generating list of eligible voters (we could not get a global one, so we generated it locally)
  5. Sending invitations to eligible voters via MassMessage. Once again, we made them more personal explaining why their vote matters, and we also added link to our village pump to let people read opinions and add their own. It is much better to send them on talk pages and not by email, as we are talking about active users. Example message.

That's how we made our community involved in this election and feeling about this election as about some important local community decision. Hope more communities will try to increase participation and share their approaches — NickK (talk) 22:37, 4 June 2015 (UTC)[reply]

I wasn't aware that you had done invitation to vote on the local wiki using MassMessage. That's fantastic! Philippe Beaudette, Wikimedia Foundation (talk) 19:39, 7 June 2015 (UTC)[reply]

P.S. As we send these invitations to eligible voters on talk pages, people could be engaged in discussion easily. Active users. It would be too private via emails, and one does not feel him/herself so engaged with something bigger just reading a letter in his/her inbox, without opportunity to ask more questions --アンタナナ 08:25, 6 June 2015 (UTC)[reply]

Direct user talk page notifications about elections[edit]

It is my understanding that two Wikipedia projects, Italian Wikipedia and Ukrainian Wikipedia, directly notified qualified voters about one or both elections by mass-messaging user talk pages. This seemed to have a positive effect on the number of editors from the two projects who voted or otherwise participated in the election process. I would be very interested in knowing:

  • if there was any on-wiki discussion before the project decided to proceed with this process
  • if there was any significant pushback or on-wiki discussion that expressed displeasure about these direct notifications
  • what processes the two projects used to identify the qualified voters, draft the message(s) and distribute the messages.

If there was generally positive response within the two communities to this process, and it is possible to scale it, this may be a worthwhile task to consider for future elections, except on a much larger scale. On the other hand, if there was significant disgruntlement or unhappiness with the process (despite the effects on participation), it may be worthwhile to consider other options in the future (such as how to improve banners/email distribution/etc). I hope that some individuals who participated in this "get out the vote" activity would provide some feedback. Risker (talk) 19:27, 5 June 2015 (UTC)[reply]

I would also be curious about using Echo (that notification thing next to your username) to do it. Perhaps less intrusive? --Varnent (talk)(COI) 20:41, 5 June 2015 (UTC)[reply]
This discussion is exactly about #2. --Ricordisamoa 23:49, 5 June 2015 (UTC)[reply]
Echo notifications would be lovely; maybe we can look into this for future board (and steward??) elections. Ajraddatz (talk) 19:00, 6 June 2015 (UTC)[reply]
talking about it.wp, we posted 4 notices at our Village Pump, both for FDC and board elections (1, 2, 3, 4) - for transparency, I wrote the last one, I support these "manoeuvres". This wasn't the first time we posted at our Village Pump to inform about ongoing elections, we started in 2007. We also added a line in our "Gazette", and sitenotice was frequently reset so to repeatedly show to anyone. Following a personal inspiration (which most of us agreed upon, though), in the meanwhile a few users started messaging user talk pages. This wasn't a mechanical mass-posting, even if it was a copy-paste of always the same text, page after page (one was written in esperanto, who knows why, but I was told it wasn't different :-). Mass-messaging wouldn't be perfectly coherent with our habits, since we had a few episodes of bad-faith campaigning in the past, so they are now explicitly forbidden for quite anything, in particular for sysops' elections and RfDs. Posts at the Bar (village Pump) and in user talk pages invited generically to vote, and they didn't indicate or suggest any particular candidate.
All was done without prior discussion, the ones who felt comfortable with these invitations, just wrote what they wanted to write autonomously and freely. At our Bar we received some unhappy comments, indeed, and they are important because a notification at the Bar always has something of bureaucratic in itself, so it might look like an official call and this could prevent some users from getting into a more open discussion. The point is that if you see an unhappy message, you have to consider there might be some other silent users who would criticize if it just didn't seem to be an "official" call. In talk pages, instead, many users thanked those who wrote them, and took the time to vote. But here too, I believe there have been negative reactions which weren't written because - perhaps - of the sympathetic personality of those who mass-posted and in respect of their (deserved) reputation. I am not aware that the ping or the echo notifications were used: posts at the Bar were impersonal and talk pages (usually) don't require any other tool.
Here (last post at the Bar), a serious and experienced user criticized the mass-messaging in talk pages mostly because of the contrast with the prohibition of campaigning; also, he underlines, this was a selective messaging, because not all the users received the message (I didn't, for instance :-), so it was not a generic list of users that was targeted, but a selection of them only, by a criterion which remained unknown.
My personal view is that the sum of all these invitations worked, actually the objections were not many and focused on formal issues, and I would use again these procedure to bring more users to the ballots. I also would try to include somewhere in our Project the real-time translation of the Q&A and questionnaires we unfortunately read too late in the Signpost (precious work, wow :-). --g (talk) 01:17, 7 June 2015 (UTC)[reply]
James and I have been begging for Echo notifications for elections for some time. We've got a request in front of engineering. We'll keep pushing for it. Philippe Beaudette, Wikimedia Foundation (talk) 19:40, 7 June 2015 (UTC)[reply]

Standing Election Committee[edit]

In my personal post-mortem of the 2013 elections, I recommended that the Board give serious consideration to a Standing Election Committee that would have the responsibility of reviewing different voting methodologies, as well as setting other election-related criteria well in advance of an election. It is clear that, given the very short time between the call for Election Committee (EC) members and the appointment of the EC, as well as the extremely short time between their appointment and the need to have the fundamental structure of the elections in place (less than 2 weeks), there was no reasonable opportunity for the EC to evaluate a range of voting methodologies, to have the software built and tested if a different methodology was selected, or even to do much more than fine-tune the processes from the 2013 election. The 2015 EC used that very short time very effectively. Nonetheless, there is feedback from several directions that it's time to revisit at least the voting methodology, and probably several other aspects of the election process. This takes time, and would benefit from having community comment as well as a period devoted to reviewing reference information. Thus, I urge the Board again to consider appointing a community-based standing election committee which would be appointed once the current committee completes its work (including post mortem assessment), probably by mid-July. This standing committee may be only a handful of members, to be supplemented by some additional members for the election itself. Risker (talk) 21:34, 5 June 2015 (UTC)[reply]

I totally agree. There is a lot of conceptual work to be done, and also many things before the elections themselves should not be done hastily. Pundit (talk) 08:15, 6 June 2015 (UTC)[reply]
Definitely. SJ talk  18:35, 7 June 2015 (UTC)[reply]
I believe this is my sixth Wikimedia election... that probably makes me the cranky old guy in the corner... and I couldn't possibly agree more. A standing committee is an idea whose time has come. Philippe Beaudette, Wikimedia Foundation (talk) 19:42, 7 June 2015 (UTC)[reply]

Electoral system[edit]

  • Many acknowledge that tactical opposing is intrinsic in the system. After the fact, some additional considerations are possible.
  • Results display a tight margin between "truly opposed" (serious) candidates and background noise: only ~200 votes, 300 at most.
  • Conversely, while there is higher variance in support votes, few dozens supports would suffice to alter the result and 2-300 supports would significantly alter the
  • Considering that converting a voter to a support vote also avoids a tactical oppose, and vice versa, the numbers needed are actually lower.
  • It's rather easy to move 100-200 votes from larger communities without leaving public traces.(signature missing)
as for background noise, you have good indication of this in the votes for the two with least support. To give Pete support is a bit hard to un~derstand, when he had withdrawn, but there were just 100 votes =(2%) that then could be seen as noisevotes. Also the last one had 1500 oppose which for me is a sign those had read through the candidates page. So I am personally very impressed by the low niose level, and find the voters to have taken the voting very seriously. For the reasons for oppose, I do not agree with you, I consider a fair standpoint is that all these were geiven with careful consideration and worth full respect.Anders Wennersten (talk) 10:42, 6 June 2015 (UTC)[reply]
One problem that we should be able to analyze from the data the EC has may be that different cultures have different approaches to opposing votes. In some cultures it may be very rare to use them - and as a result, different communities may want to express the same kind of support, but with the current mechanics, will cast entirely different votes. Pundit (talk) 10:56, 6 June 2015 (UTC)[reply]
It may be tricky to define which regions voted in which way; I'm not even sure if SecurePoll would be able to spit out that data because of its vote security. I'm not convinced there's any reason to believe there is correlation between geographic location and voter feelings about opposing candidates; more importantly, they're secret ballots that probably should not be technically capable of being correlated that way, in order to protect the privacy of the votes and the voters. (Yes, I know that in "real world" elections there are breakdowns even to individual polling stations, but each polling station tends to have more voters than we have in this election, so it remains much more anonymized.) I don't think I'd want to participate in an election that analyzed the votes of all 30-50 participants from my country to this point.

Anyone can do an analysis of the home wikis of the voters since that is public information; one must keep in mind, however, that the geographic location of many voters will not be revealed by the home wiki. For example, there are plenty of voters from internationalized wikis (Meta, Commons, Wikidata, etc); English Wikipedia is the "home" wiki for people from all over the world, and not even just English-speaking countries; and several other language versions of projects (Spanish, French, Arabic, Hebrew, Chinese dialects to name but a few) have editors from multiple countries. Risker (talk) 05:16, 8 June 2015 (UTC)[reply]

Invalid votes, recast[edit]

All of the 6,26% of the votes that were invalid were "recast"? Quite a high percentage. What does recast mean precisely? --Sargoth (talk) 11:28, 6 June 2015 (UTC)[reply]

Aspiriniks found the answer and posted it on deWP. It seems, that a lot of users used a feature that was a bug. Shouldn't happen next time, in my eyes. --Sargoth (talk) 11:36, 6 June 2015 (UTC)[reply]
Recast means that they voted multiple times, and only their final ballot was accepted. This is pretty common each election that SecurePoll has been used. People change their minds, forgot someone, etc. It is not related to the bugs, or an indication of much - from what we can tell anyway. --Varnent (talk)(COI) 11:47, 6 June 2015 (UTC)[reply]
So if they recast and change their votes, their first votes are invalid only? Strange. But ok. --Sargoth (talk) 13:34, 6 June 2015 (UTC)[reply]
It means you can change your mind and only your last response counts. Doc James (talk · contribs · email) 18:05, 6 June 2015 (UTC)[reply]
I suspect keeping but invalidating the results is actually preferred to deleteing or removing the votes. It's probably useful to have a record of all voting, in case there is concern of fraud. Nil Einne (talk) 07:36, 7 June 2015 (UTC)[reply]
We do keep a record of all voting, and the ballots themselves are not deleted, it's just that when the tallying script runs it doesn't count the 'old' ballots. It counts the most recent ballot from each person. Jalexander--WMF 08:11, 7 June 2015 (UTC)[reply]

Has the SecurePoll extension been audited?[edit]

Has the SecurePoll extension, either the version used in the election or any version, ever been professionally audited (for security, correctness, etc.)? Also has consideration been given to rewriting (parts of) it in a more formally-specified programming language than PHP?

I realise that some of the concerns that have been raised about electronic voting in national elections (such as coercion by standing behind someone while they vote and forcing them to revote until they vote "correctly") are less likely to be relevant here, but I think some of them are still valid for these smaller types of elections.--Greenrd (talk) 11:32, 7 June 2015 (UTC)[reply]

Greenrd, can I ask you which concerns, precisely, would be relevant in our case? (in my country there are no national elections with electronic voting, so I don't have any experience about that) --g (talk) 15:11, 7 June 2015 (UTC)[reply]

Rewriting Secure Poll extrension[edit]

I think WMF should invest some efforts into updating this extension to make setting up polls easier without need to manually edit xml files. Some other features should also be improved such as creation of voter lists, eligibility determination and translations. The general look&feel of polls is also quite outdated and is in urgent need of an update. Ruslik (talk) 17:13, 11 June 2015 (UTC)[reply]

Possibility to submit spoilt vote[edit]

Apologies for not paying enough attention last election, but I was wondering if it was possible in the past elections to submit a blank or spoilt ballot. Also, whether this will be possible in the future. --Tobias talk · contrib 14:33, 8 January 2016 (UTC)[reply]

If we go back to the Schulze method, you will be able to give everyone the same score, which is practically the equivalent of a blank ballot. Nemo 17:26, 8 January 2016 (UTC)[reply]
Alright, but that's not really what I am trying to get at. There should be a possibility for users to express their rejection of the election process on the ballot, by submitting a ballot that will be counted separately as invalid/spoilt. This is both a long-standing tradition in real life elections as well as democratic processes on Wikipedia: for instance on German Wikipedia, you always have the option to say "Yes", "No", to abstain or to "Refuse", where refusal is usually chosen by those who disagree on formal grounds with the vote. If a large number of people "refuse" on a vote, this reduces the legitimacy of the result, which would not be equivalent to what you describe. --Tobias talk · contrib 18:45, 8 January 2016 (UTC)[reply]
Well, the same effect would be achieved, as long as the dump of the ballots happened to be published as used to be done regularly. Sadly the election around WMF elections has been constantly decreasing over the last 5 or so years. :( Nemo 19:41, 8 January 2016 (UTC)[reply]
I think there is a clear difference between selecting every candidate equally and submitting a spoilt ballot. The former is a good option for voters who wish to support the election in general, but can't decide for a particular candidate. The latter is a way of expressing dissatisfaction with the voting process and/or the candidates. Not casting a ballot at all is not a good substitute either, since it indistinguishable from apathy --Tobias talk · contrib 12:22, 9 January 2016 (UTC)[reply]
@Church of emacs: I'm not 100% sure if I understand the question but I think you're asking if you're able to submit a ballot that, in the end, is meant to have no actual "effect" on the election (in has no effect on anyone's success or loss) but has you in the list of voters? If so this is, indeed, possible and likely would be (but depends a bit on how future elections are set up so I don't want to make a blanket statement). In this specific election and ballot submitted with only neutral votes (the default position so just pressing submit after loading the ballot) would have that effect. Jalexander--WMF 01:34, 10 January 2016 (UTC)[reply]
To understand the question, it may be useful to do some basic reading on the matter, e.g. w:Spoilt vote. IATE offers translations in multiple languages: [1] [2]. Nemo 10:46, 10 January 2016 (UTC)[reply]
(edit conflict) Let me put it a different way: electronic and online elections are meant to replace paper elections. They thus need to provide the same options to the voter. In a paper election I am able to write "I don't agree with this vote" on the ballot, thus purposely invalidating the ballot. It will NOT be counted as supporting every candidate equally, instead it will be put on a pile of spoilt votes. The spoilt votes don't have any effect on who wins the election, but they can have a psychological effect: if 2/3 of the voters going to the voting booth submit a purposely spoilt vote, then it is a strong sign that whoever is elected by the remaining 1/3 has a big legitimacy problem (despite being elected by the majority of valid ballots). This is clearly a different outcome from where someone who doesn't agree with the election votes for all candidates equally. It also sends a stronger message than simply not taking part in the vote, which can be interpreted as apathy. If of 100% eligible voters, there's a 30% turnout but 2/3 of the votes are purposely spoilt as a sign of protest, it is much less legitimizing than a vote with 10% turnout without any spoilt votes.
In some elections you also have an option "None of the above", which is similar to what I have in mind. Notice how "None of the above" is very much different from "Every candidate equally", because despite the same outcome (no effect on the election scheme) it captures a very different message. --Tobias talk · contrib 10:56, 10 January 2016 (UTC)[reply]
"no effect on the election" is ambiguous. There's formal effect, which is an impact on the calculation of the outcome of the election given a set of votes, and there is psychological effect, which influences how we think about the election. There's no formal effect that spoilt votes have on the election, but there certainly is a psychological effect. --Tobias talk · contrib 11:01, 10 January 2016 (UTC)[reply]