Wikipedia:Arbitration Committee Elections December 2010/Feedback

This page is for feedback on the December 2010 Arbitration Committee elections from all involved. Constructive reviews, comments and suggestions for future elections are most welcome.

Feedback by role

edit

Voters

edit
  • Apparently, the idea of notifying all eligible voters by bot on their user talk pages about the election was shot down. I hope the idea of using bot notifications will be considered again for the next election, however, as I think it would help increase voter participation. Cla68 (talk) 03:49, 6 December 2010 (UTC)[reply]
    That would potentially run to millions of accounts, whereas fewer than a thousand could be bothered under the current ad regime... Skomorokh 03:53, 6 December 2010 (UTC)[reply]
    About 760,000 accounts, I think. Definitely not worth it unless someone can calculate the number of active, not just eligible accounts. Also, we'd need to count out alternate accounts of people and stuff. And most people who don't know it's ACE season probably don't care too much about ArbCom tbh. /ƒETCHCOMMS/ 04:58, 6 December 2010 (UTC)[reply]
    ArbCom is a specialised discussion area of Wikipedia which is appropriately not of concern to the majority of users; as such it wouldn't be appropriate to give all users yellow talkpage alerts. We don't alert people to RfA candidates, even though admin appointments have more impact on the day to day the running of the site. SilkTork *YES! 16:29, 6 December 2010 (UTC)[reply]
    ArbCom decisions, especially in regard to areas where editing is contentious, have the potential to affect many editors. Rather than suggesting that only some sort of "inner circle" has an appropriate concern, would it not be better to make Arbcom more widely known among the community. At the moment, ArbCom is elected by the votes of little more than 1 in 1000 Wikipedians, which is not much of a mandate. And as for RfAs, instead of using the lack of alerts there as justification for none here, perhaps we should introduce them there also. Jimmy Pitt talk 21:53, 6 December 2010 (UTC)[reply]
    I agree that gathering more votes would be nice. What about making it similar (in appearance) to "personal appeals from Jimmy Wales", but shown only to those who's eligible? Shouldn't be that bad implementing, and shouldn't be too bad for those who don't have any interest in it (one click on 'X' button once per year isn't too much IMHO). Ipsign (talk) 16:39, 7 December 2010 (UTC)[reply]
  • Put simply: more time was needed. I ended up voting neutral on most of the candidates, because within the time frame for discussion and voting, I simply didn't have the time to examine the ones I didn't already know. RayTalk 15:16, 6 December 2010 (UTC)[reply]
  • More time to vote would have been nice. The voting period seemed unexpectedly short compared to other years. Voting responsibly takes a fair amount of time and effort. --Alecmconroy (talk) 16:27, 6 December 2010 (UTC)[reply]
  • We got nearly as many votes as last year, with the interference of Thanksgiving in the US. I'm unsure why Thanksgiving, which lasts for a weekend (or even a week, is it?) stops people from voting during a 10-day period that purposely included two weekends. In the view of many editors, including me, a 49-day election hoopla such as applied to ACE2009 is out of balance (that's almost one-seventh of the time). There was a deliberate attempt to wind it back to about a month this year. It is a lot of work even when it lasts a month; please remember that election coordinators have to give up a lot of time to make it work. I do agree that the fallow period between nom and voting was too short—that was my fault. Perhaps five days instead of two; but I would resist a wholesale blow-out of the total duration. Tony (talk) 07:28, 7 December 2010 (UTC)[reply]
  • Agree with needing more time to vote. Two weeks instead of ten days would be my suggestion. It's just another four days. Heimstern Läufer (talk) 09:02, 7 December 2010 (UTC)[reply]
  • Voting time was too short. I went to the polls and found them closed already. :( Suomi Finland 2009 (talk) 16:46, 8 December 2010 (UTC)[reply]
  • I think the voting time was actually more than enough. Ten days, including two weekends, is more than enough time to actually hit the voting page. What may be needed is more time between the closing of nominations, and the start of voting. If we wanted to keep the total time for the election process the same, I would suggest 8 days for nominations, 6 days for discussion, and 8 days for voting. The nomination period could also be shorter as long as it was announced ahead of time, as I don't believe that a serious candidate would be a "weekend only" editor. Jim Miller See me | Touch me 17:02, 8 December 2010 (UTC)[reply]
  • Agree that more time is needed for voting. The November-December months are big vacation periods. Two weeks minimum. Three weeks or a month would be preferable. ScottyBerg (talk) 20:16, 8 December 2010 (UTC)[reply]
  • To be honest the only reason I voted at all was so I would have the right to complain when and if arbcom and I disagree over something. I don't really know any of these guys, and having only a few days to decide didn't help either. Also, why is no one working on the bot angle? Psychologically I would think that being invited to vote would produce better results than a drive by notice. One last thought: it may help to explain what arbcom is and does in relation to the us personally, I suspect that if the registered users felt they were a part of the process they may be a little more incline to vote in the elections. TomStar81 (Talk) 18:34, 10 December 2010 (UTC)[reply]
  • The quality of the candidates was impressive, and the number of candidates was good, although the numbers were worrying to begin with. I wish that the discussion period following nominations, and then the voting period, were much longer. While I wish that I knew why others opposed candidates I thought suitable, I accept that this is a desire, not a need. I like the public logging of voters. I'd be interested in seeing a demographic study of the voters. Congratulations especially to Newyorkbrad for such an impressive vote of confidence. --SmokeyJoe (talk)12:47, 25 December 2010 (UTC)[reply]
  • the length of time seemed fine to me. Making it longer won't solve many problems and may even remove some urgency which motivates marginal voters to participate. We do need to change how the "default" is recorded, with either no radio button selected or "no opinion recorded" as the default option because people persist in applying the norms of AfD/RfA to these elections despite the substantive difference between a volunteered "neutral" opinion and an ambiguous default "neutral" option on a ballot. the number of voters is likely to decrease year over year as the "core" Wikipedia userbase shrinks. Please don't interpret this as cause to make dramatic changes to the format or advertising for the election (though I don't mind a bot message to elegible, active accounts). Eventually if the electorate shrinks enough we can go back to a public discussion style, but I think that won't be for a while now. I can't comment on the questions because I didn't read any of them. Like RfA questions they tend to be unenlightening in the main; some are uniquely helpful but some others (rarely) are largely preening by the questioner. I also don't really care if we link to guides from the "official" template (can't put enough quotes around official!). Elections are a community process and community guides which are minimally helpful should be given some prominence if they help voters choose candidates. Overall the election seemed to be handled with professionalism (with one exception to be noted in a bit). There will always be good natured ribbing about the length of time it takes to scrutineer <1000 votes, but it is not a big deal. The business with jimbo at the end pondering whether or not to ask detailed questions about criminal history was unseemly. If you want to background check arbitrators make it a foundation position out of the hands of the community, pay salaries and post real names. If you want a community process then the community needs to handle it (beyond what is literally necessary for CU status). Jimbo's "cerimonial" (again, can't put enough quotes around that as it is clearly not ceremonial) position being used to add an extra layer of basically useless and intrusive scrutiny doesn't add any value. Protonk (talk) 03:47, 6 January 2011 (UTC)[reply]

Candidates

edit

I'd like to thank the election coordinators, scrutineers, and everyone else who handled (or is now handling) the logistics of the elections, as well as the authors of candidate guides and all the voters. Beyond that, I like the idea of asking for input at this time while any issues with the election are fresh in everyone's mind, but I don't have much to add: apart from the obvious controversies, the election process unfolded reasonably smoothly from my point of view. Newyorkbrad (talk) 02:16, 6 December 2010 (UTC)[reply]

I found the feedback for each candidate pretty quiet on the whole. Not fussed about it really but does make me ponder the election guides, which (I guess) then assume a greater prominence in the relative vacuum compared with non-secret voting pages. Given this, I sorta think the more guides the merrier really so as we get more of a quantitative rather than qualitative emphasis :) Casliber (talk · contribs) 02:30, 6 December 2010 (UTC)[reply]
PS: Yeah ditto/what NYB said.. Casliber (talk · contribs) 02:27, 6 December 2010 (UTC)[reply]
Jclemens' feedback
  • To the Coordinators The nomination tools, candidate instructions, and voting instructions were clear. As a candidate, I thought things were well handled. I appreciated the suppression of one question, even though I didn't think it was out of bounds and I was OK answering it. If there's anything I would change for the next election, I might incorporate more of Lar's questions--they amounted to a de facto set of official questions that all candidates had to answer or face the "wrath" of the guide writers, so why not just incorporate them? A dialogue between the guide writers who intend to return to the role, giving them a reasonable amount of input into the question process, might be appropriate before next election.
  • To the guide writers Half of you engaged with us candidates, accepted feedback, posted questions, changed your minds when merited and otherwise made it clear that you took the responsibility seriously. Others... didn't. Lar gets props for his questions. If anything, I was surprised by the number of guide writers who made up their minds about me without asking me any of their own questions. Jclemens (talk) 03:03, 6 December 2010 (UTC)[reply]
    • As a guide writer, generally my own philosophy is more based on what someone has done, rather than what they are promising to do. So asking a question isn't particularly helpful for me, unless I ask something like, "Please give me links to where you have offered comments in previous arbitration cases". But then again, I can usually search that up on my own, which is why I don't bother asking! --Elonka 03:54, 6 December 2010 (UTC)[reply]
      • Agreed: promises are dust. Conversely, I was surprised by how few candidates engaged with the guide writers William M. Connolley (talk) 21:59, 6 December 2010 (UTC)[reply]
      • Agreed: promises are indeed dust. Which is why questions eliciting promises are less worthwhile than questions eliciting what people think about issues and how they would handle situations. I try not to ask any of those sorts of questions, myself. But having different guide writers analyse using different factors (and explain their analysis) is goodness. I hope we have twice as many, no 5 times as many, guides next year if we are going to continue with secret balloting. ++Lar: t/c 03:33, 7 December 2010 (UTC)[reply]
      • I didn't ask candidates any specific questions of my own this year, as by time I came round to give my view on candidates, the existing questions told me what I needed to know (including one general question based on a question asked by me in the 2009 Election). If any candidate wanted me elaborate on my position, or to clarify any issues I had raised on them, they were welcome to do so on my guide talk page, and some did do this last year. CT Cooper · talk 21:29, 7 December 2010 (UTC)[reply]
Feedback by Sandstein

I agree with what Newyorkbrad and Casliber said: the elections were conducted smoothly and in a calm, professional manner. I appreciated the number of interesting questions that were asked of me. What I appreciated less was several people wanting to copy-paste a huge catalogue of generic questions to each candidate's talk page, even though there was presumably a reason that there was no consensus to include these questions in the community-developed set of general questions (probably because several of these questions dealt with nuances of wikipolitics that are only interesting to a rather limited number of people, and have little relevance to the job of arbitrator). I also found it rather odd that the many "election guides", which reflected the private views of individuals, were prominently linked to from the "official" election template.  Sandstein  07:27, 6 December 2010 (UTC)[reply]

Well I guess I'm disappointed that you didn't want people to know what you thought about BLPs, flagged revisions, stalking, outing, offsite criticism, vested contributors or factions, among other topics I asked about. Most other candidates handled my questions just fine and I suspect I'm not the only voter that opposed you (and perhaps other candidates as well) over your refusal to engage with voters. These questions are not any more "generic" than the standard ones are. They're just better than the standard ones because they go into more detail and ask harder subquestions. Maybe if you can't answer hard questions, arbitration isn't really your cup of tea? Further, I followed up with a number of candidates to clarify or elaborate on their answers when I didn't understand. That in at least one case converted an undecided into a support. You missed out on that. Your loss, not mine. ++Lar: t/c 03:31, 7 December 2010 (UTC)[reply]
Lar, if you want to ask lots of questions, that's fine, and if you vote on the basis of the answers, that's also fine. However, Sandstein has done more arbitration enforcement than anyone else, so in my opinion, he has demonstrated that he possesses the necessary skills to serve on ArbCom. Just because he chose not to answer your questions, you shouldn't suggest he is incapable. PhilKnight (talk) 17:19, 7 December 2010 (UTC)[reply]
How is arbitration enforcement a predictor of the skill set necessary to be a good arb? Or an explicative of how the candidate views important matters affecting the whole wiki. Sandstein is well within his rights not to answer anything but his candidacy suffers as a result, IMHO anyway. I won't buy a pig in a poke... I'm not suggesting he IS incapable, just that we have no way of knowing. I'm also suggesting that if he refuses to answer questions now, that's a predictor that he might not be responsive to questions (about decisions, or about what he meant) while serving. THAT's concerning. At least to me, maybe not to you. ++Lar: t/c 17:27, 7 December 2010 (UTC)[reply]
Elen of the Roads sez

Didn't mind the questions, but did find it a bit difficult having them in two places. This could be because every page starts withWikipedia:Arbitration Committee Elections December 2010, so I found it somewhat difficult to spot 'em apart. I'd also prefer if the questions were formatted into individual sections - I did that on the talkpage but wasn't sure if I was allowed to fiddle with the 'front' page, as I thought it was being transcluded. I don't mind if guides are written wholly on the basis of which candidates the guide writer would like to buy them a beer, but I found it somewhat confusing when a guide writer who seemed to be working on that basis, then said he hadn't read anything I'd written and asked if I wanted him to read it.Elen of the Roads (talk) 16:27, 6 December 2010 (UTC)[reply]

Question for the candidates

There were a few comments here and there alluding to the paperwork candidates had to get through as an impediment to standing. I wonder if any of you would care to comment on this – was there anything superfluous, unnecessarily difficult, or better handled by volunteers other than the candidates? Were the templates, preloaders and whatnot a help or a hindrance in getting your statements and question pages set up correctly? Skomorokh 18:37, 6 December 2010 (UTC)[reply]

I do think anyone who couldn't complete the paperwork would struggle as an arb. As a tiny side note, I didn't find out the statement was only supposed to be 400 words until after I posted it, and then I had to edit about 10 words out to make the limit. <---That was me Elen of the Roads (talk) 11:57, 7 December 2010 (UTC)[reply]
I didn't think that the paperwork was a problem, it was all very well set up.  Sandstein  21:37, 6 December 2010 (UTC)[reply]
A couple of parts of it were confusing to me on the first try, but it was easy to get past that and get it right. Not perfect, but not bad all things considered. It certainly kept things more consistent than I've seen happen before. Georgewilliamherbert (talk) 03:40, 7 December 2010 (UTC)[reply]

Election volunteers

edit
Comments by Fetchcomms
  • Maybe a better way than "anyone is welcome" because there were some concerns raised about the coordinators' judgments this year. Also, not sure coordinators should be asking questions to candidates, though definitely not endorsing/writing "guides". Could be more organized with a clear "who does what" as well, but I don't think there was much issue. /ƒETCHCOMMS/ 05:02, 6 December 2010 (UTC)[reply]
Comments by Sven Manguard

The first thing that I want to do is publicly go on the record and thank fellow coordinators Skomorokh and Tony1. Everyone should follow suit. Most of the coordinators did minor things, template maintenance, rules compliance, and the like. I made questions number properly. That's all well and good, but none of it would have been possible without Skomorokh and Tony1. The two of them ran the show: they planned everything out, recruited the election admins, stewards, and technical coordinator, wrote for the signpost, manned the IRC channel, were a constant presence in answering the questions at the various election pages, and directed the rest of us.

There would not be an election if it were not for those two users. Period. A hearty round of applause is in order from all.

Moving to a less positive note, I would also like to go on the record and register my displeasure in a number of issues that plagued this election. I could write paragraphs and paragraphs of observations, there were several issues, not the least of which were a drop in votes, a less than ideal spread of candidates (both in quality and quantity), and the seemingly endless flow of unnecessary, often uncivil, and certainly unproductive flow of drama.

What I am instead choosing to focus on is the community's poor sense of timing on changing or trying to change the policy and procedure of the election. There was an RfC preceding the election by a few weeks, where changes to the rules were proposed, voted on, and finalized. Participation was decent, about 100 people were involved, and most of the consensuses were clear. However, the RfC forced mainly on the voting method. Once that was decided, and the election began, people began complaining about the voting method, the page layouts, whether or not guides should be included on the ACE2010 template, and confoundingly enough, whether or not we should enforce the preexisting policy on identifying to the foundation.

All of these are valid topics for the community to decide, but doing so in the middle of an election, where it would be confusing and disruptive to actually change anything, seems insane to me. If people have suggestions or problems, these need to be raised before the elections began. Considering that the placement and quantity of questions was an issue in the 2009 election, someone should have thought to raise the issue in an RfC in August or October. Without the voice of the community in an RfC, the coordinators had to rely on the community voice from the various pages in the 2009 election, hence the changes made this year. We can only work with what we know.

I know from asking around that in previous years, RfCs on various facets of the ArbCom election were attempted after elections, and months before elections. Those RfCs, which I have not myself seen, allegedly died from a lack of participation. If January won't work for the community and August won't work for the community, and no one cares to bring these things up in the October RfC, then in my opinion the community has no right to complain about the processes used in the elections. It is clear that there are issues in need of a community decision. It is also clear that there is a poor track record for RfCs not immediately proceeding the elections. If changes are to be made, someone has to either organize an RfC or bring it up in next year's October RfC. Therefore, my proposal is simple. If someone really cares about an issue that they feel was problematic form this election, or wants something to be adopted as policy for 2011, they need to push it. They need to set up the RfC, they need post the question, and they need to fight to ensure that the discussion gets adequate participation. If no one is willing to do that, I propose that no one have the right to complain.

One final note. This election, two editors signed up as coordinators and were subsequently asked to step down (by editors at large, not by election staff or other coordinators.) One did not. One did, and then rejoined after I persuaded him to do so. Asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable. Like all other non-elected positions and tasks on Wikipedia, being a coordinator is an individual editor's personal choice. If you believe that there should be a guideline as to who is eligible to serve as a coordinator, bring it up in an RfC. Barring that, let me repeat, asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable.

TLDR

  • Special thanks to Skomorokh and Tony1, the unsung heroes among the coordinators.
  • I'd love to lambaste the community for various forms of disgusting and uncivil behavior, but other people will cover that, I'm sure.
  • Participation in the RfC on the method of voting was high, however there was no RfC on other issues, such as the templates, the guides, and the questions. If people really care about these issues, they need to craft RfAs. If they can't be bothered to do that, they have no right to complain during the election.
  • Asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable. Let me repeat this, because I was absolutely disgusted by this. Asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable. Period.

Sven Manguard Wha? 05:37, 6 December 2010 (UTC)[reply]

    • "Asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable. Let me repeat this, because I was absolutely disgusted by this. Asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable. Period." I'm afraid that's rubbish. We have the right to request people to stand down if they are not doing their jobs appropriately. That is the community's prerogative, and no full stops from you will ever change that. No opinion on whether the specific demands here were appropriate, but a wholesale assertion that asking someone to step down is unacceptable is just plain nonsense. Heimstern Läufer (talk) 09:06, 7 December 2010 (UTC)[reply]
      • In both of the cases where this happened this year, the people that demanded coordinators step down were not coordinators themselves. The people asked to step down were doing perfectly fine as far as the coordinators were concerned. It's not okay for members of a volunteer project, one that prides itself on an anyone can participate model, to pick and choose who can and cannot volunteer. It doesn't work like that, especially in a role like that of the coordinators which has no advanced rights and is not elected. That is what happened in these cases. Established editors asked coordinators to step down, not because the coordinators did anything wrong, but because the editors didn't like the coordinators. If these people did things that were against the guidelines or the spirit of the position, then the rest of the coordinators might have grounds to step in, but chasing out eager and able hands without cause is an absolutely idiotic idea. All it will lead to is less and less eager hands, and the steady attrition of the able ones. Sven Manguard Wha? 05:17, 15 December 2010 (UTC)[reply]
Comments by Skomorokh

I'll add to this as issues occur to me.

  • First of all, a thank you to my colleagues. a huge thanks are due to Tony1, who doggedly worked behind the scenes to ensure the necessary people and structures were in place in time (not to mention covering the election almost single-handedly as Signpost correspondent), and without whom the elections would have most likely been an unengaging, disorganised, last-minute permanent crisis. Similarly, MBisanz's attentiveness and readiness to assist was very helpful, and Happy-Melon came through in a big way when it was needed. The selflessness of the scrutineers and the volunteering of the coordinators in difficult circumstances is commendable.
  • We needed more coordinators with experience who were prepared to make potentially difficult decisions. While the guidelines for this election were undermined by a lack of community input, once those guidelines are established they need to be enforced. The role of patroller and gatekeeper of editor behaviour around the elections is a necessary one, and with the notable exception of Jehochman there were few editors who were willing to take this on, leaving too much of this to those, like me, who are ill-suited to the role. This situation lead to avoidable drama and inconsistent application of standards. Skomorokh 16:08, 7 December 2010 (UTC)[reply]
Comments by MuZemike

First off, I'll add that this was my first year helping out of sorts with the ArbCom election, and I am amazed of how much work goes into running it.

While I did start Wikipedia:Requests for comment/2010 ArbCom election voting procedure (which was mainly out of community need and that time was rather short), for the most part, I did relegate myself to observing the odds and ends, consistently watching the WT:ACE2010 page. On the CheckUser end, I assisted in two CheckUser investigations, both of which sock puppetry was discovered.

Anyways, here are some of my comments and observations:

  • It seems that the navigation around the WP:ACE2010 pages are rather haphazard and unclear. Perhaps I would recommend something along the lines of fewer subpages to help improve on that end, that's all I can think of right now.
  • We still get lots of commentary on WT:ACE2010 about users making complaints about things that have already been discussed in previous RfCs. Common reasons include not being around during said previous RfCs to comment, not being arsed to comment until the election is underway, or even users who may now have strong opinions due to a recent event which they consider their own personal travesties.
  • Voter guides seem like a decent idea; if somebody wants to publicly disclose who they support or oppose, then they reserve the right to do so.

I look forward to looking at the results and especially seeing where the percentages stand per my RfC statement here. –MuZemike 19:13, 7 December 2010 (UTC)[reply]

Comments by Happy-melon

I would first of all like to echo the thanks that have been given to the other election volunteers, especially Tony1 and Skomorokh. The success of this election is largely thanks to their foresightedness and commitment.

The biggest concern for me was the community's willingness to attempt to change the policy and procedure governing the election in midswing, something which has been frowned upon in recent RfCs as well as being just plain stupid. 90% of the discussion about identification needn't have happened: as soon as Jimbo said the first time that he would not seat an unidentified candidate, that should have been the end of it for this election, because the procedure for this election is that Jimbo takes the results as advisory and reserves the right to seat or not seat candidates as he chooses. All further discussion should have focused on reviewing that procedure for the next election. There were other similar issues of smaller scale, particularly the withdrawals; fortunately it was fairly easy to curtail discussion there by indicating "there are only three users who have the technical means to make the change that you want, and we all agree that it should not be made". But people really do need to get into the habit of accepting that an election is not amenable to being rewired midway through.

The proliferation of fragmented discussions on numerous disparate talkpages was, frankly, a mess. I have sixteen ACE10-related pages on my watchlist, and I'm sure I haven't covered them all. Much more thought needs to be given to which talkpages should stand alone (and what they should be for), and which should be redirected to one of said consolidated discussion fora.

There are still a number of things which need to be done to SecurePoll, the top priority of which is proper logging of tally generation and of edits to messages. I'd say that the community needs to thoroughly review the voting method before next year (by which I mean well before next year). And ArbCom really need to finish off WP:ARBPOLDRAFT; it clears up so many ambiguities which exist in the current ArbCom policy. That's my £0.03. (also)Happymelon 20:19, 10 December 2010 (UTC)[reply]

  • General query: Not including Tony and Skomorokh, what exactly did the other coordinators do (in their role as a coordinator, as opposed to their role as a checkuser)? NW (Talk) 05:38, 15 December 2010 (UTC)[reply]

Feedback by topic

edit

Candidate eligibility

edit

Just to start things off, it seems like we need more clarity next year about exactly what the expectations are regarding identification--how it is done, what happens to the identifying information after the age/identity is confirmed, what is done to ensure the identification actually represents the editor in question, and what happens to candidates who refuse to identify. Now, perhaps we should hold of on such a discussion until this years appointments are determined, but I think that next year we could definitely do with an upfront, unambiguous statement. I don't know how much of that statement would need to be authored by the Foundation and how much could be authored by the community, but the current confusion seems...undesirable. Qwyrxian (talk) 02:24, 6 December 2010 (UTC)[reply]

Agree. The rules should be clear, and should be enforced before a candidate gets on the ballot. It is very undesirable to think a candidate might be elected, and then be rejected. If the successful candidates can trust WMF with their identities, so can all the candidates. Jehochman Talk 03:37, 6 December 2010 (UTC)[reply]

Not quite sure where to ask this, but if there is anyone reading this who considered putting their name forward as a candidate, but decided not to for reasons that could be taken into consideration for next year, would they consider stating those reasons here, or submitting them in confidence to the election co-ordinators? Carcharoth (talk) 03:38, 6 December 2010 (UTC)[reply]

I second this, and if anyone would like me to proxy their comments, in the strictest confidence, please feel free to get in touch via email. (also)Happymelon 20:54, 6 December 2010 (UTC)[reply]

Voter eligibility

edit
  • A quick note on this subject. During the voting period, I asked a user with technical knowledge and access to preform a check to determine the number of eligible voters. The number that he came back with was 170,000. This number does not account for recent activity, and at the time of this posting, I have yet to secure a cross reference between the list of eligible voters and the list of active users. Sven Manguard Wha? 05:43, 6 December 2010 (UTC)[reply]
    • Depends how we want to define active? According to the Wikimedia statistics,[1] in September 2010 there were 3644 editors on the English Wikipedia actively making 100+ edits/month, and 35222 with over 5 edits/month. --Elonka 05:52, 6 December 2010 (UTC)[reply]
  • The eligibility requirements are too high. We went from a requirement of 150 mainspace edits, to a requirement of 1000, in the space of one year! When I joined Wikipedia, 1000 mainspace edits was "why aren't you an admin yet?" material. Great, we have users who are more dedicated and more obsessive now, but why exclude the people with 150-1000 edits who five years ago would have been seen as key parts of the Wikipedia community? This certainly reinforces the popular image of Wikipedia editors as an exclusive club who try hard to make themselves even more exclusive. rspεεr (talk) 09:34, 6 December 2010 (UTC)[reply]
    As far as I can tell, 1000 mainspace edits has been required of candidates since 2006 (with some debate about it each year). Of voters, the requirement was and still is only 150. Rami R 10:41, 6 December 2010 (UTC)[reply]

Candidate questions

edit

The question limitation was useless and unhelpful. It was a nice experiment, but it should be reverted to the previous rules. NW (Talk) 02:16, 6 December 2010 (UTC)[reply]

I think the limitations foundered because they were unsuccessful in encouraging questioners to alter their behaviour. The main effect was to split the questions across multiple locations. Skomorokh 02:19, 6 December 2010 (UTC)[reply]

I didn't find the general questions to be that useful since so many of the answers were similar (especially the ones on principles). To me, the only two that really helped were the one of a case the user agreed and disagreed with and the last question on changes they'd propose. I'd also like to see a few days added to the period after noms end and before voting so actual discussion can take place. DC TC 02:27, 6 December 2010 (UTC)[reply]

I'd agree with Skomorokh - one does feel obliged to answer questions once asked...and to try and answer at face value even if aware of the circumstances around the question. Despite this I suppose, many questions were worthwhile. Casliber (talk · contribs) 02:35, 6 December 2010 (UTC)[reply]
I also think more discussion time is needed. It didn't help this year that some candidates entered very close to the cut-off time for nominations. A voting period of 10 days is about right, but at least that amount of time should also be allowed for pre-voting discussion, if only to allow candidates to withdraw before voting starts if the discussions throw up problems with their candidacy. I felt that I was only beginning to get a feel for the candidates and their suitability for the roles at about the time voting was finishing, though maybe that is because I was coming at the question from a different perspective to most voters, in that I have experience of what the role entails. *shrug* I can't expect other voters to view things from that perspective, of course! But to take RfA for example. Imagine there were 22 RfAs running simultaneously. If there are RfA regulars here that look closely at an RFA candidate's contributions, would they be able to say if they would be able to comment on 22 RfAs within a ten-day period? If not, then how can voters in these elections be expected to do the same? Carcharoth (talk) 02:46, 6 December 2010 (UTC)[reply]
I agree. I'd favor a longer period between the close of nominations and the beginning of voting, where the questions can be asked and guides drafted. I get the fact that everyone can go back and revise their vote later, but I really believe that should be the exception, rather than the rule. Jclemens (talk) 03:09, 6 December 2010 (UTC)[reply]
Also in agreement that it felt like too little time between the end of nominations and the beginning of voting, especially with all the last-minute candidates. --Elonka 03:57, 6 December 2010 (UTC)[reply]
  • I'd like to minimize the number of copy and paste questions. The questions that really have value are the ones unique to each candidate. What sort of interviewer asks standard questions? Perhaps we could ask the candidates to fill out a "job application" consisting of general question to draws out their basic positions on issues of common interest. This should be filled out when the candidacy is announced. Then, voters should have at least 10 days to review the statements and general question answers, and an opportunity to ask one follow up question at a time. This way voters could ask more than one question, but not all at once. That would help prevent one voter from overwhelming a question page so much that other voters are precluded from asking their questions. Jehochman Talk 03:35, 6 December 2010 (UTC)[reply]
    I like this idea. It would be difficult, however, to stop people applying at the last-minute with a half-filled application and some sob-story and promises to fill in the rest of the application form pronto. I do like the idea of limiting questions to successive ones, but that can also be abused by persistent questioning on one topic (though that is something arbs will have to learn how to handle). Carcharoth (talk) 03:40, 6 December 2010 (UTC)[reply]
    I agree with Jehochman that this is an issue. One problem is that a great deal of questions are not questions at all, but rather a technique to bully candidate's with one's opinion. (This is the effect, if not the intent.) The only real purpose of these litmus test-style questions is social coercion to elicit politically correct answers (you will be opposed for answering many questions in a certain way). This becomes corrosive because the more people do this, and the more we tolerate this as a community, the more likely we are to essentially cede control of the committee to whoever asks the windiest questions and most ferociously promotes their litmus testing. That said, I'm not sure what is best to be done about it. --JayHenry (talk) 03:48, 6 December 2010 (UTC)[reply]
    I agree with JH and the "one at a time" idea. The problem with many individual questions, including most of Lar's, was that they diverted attention away from the skills and experience that make a good arbitrator. I am unconvinced that probing every candidate's views on of BLP and other political/policy issues gives voters a better idea of how to vote. Probably only a small proportion of voters got that far. Might be worth conducting a survey of 50 voters randomly selected from the log on a few issues. But we have more important matters to sort out, like the voting system and a revamp of the software. Tony (talk) 14:28, 6 December 2010 (UTC)[reply]
    I think you're confused. You really are having a hard time with the notion that many candidates and many voters find my questions useful, aren't you? You seem to be more focused on form than substance. We need the best arbcommers we can get, not the smoothest election. The election is nothing compared to the torture that awaits the winners. ++Lar: t/c 03:21, 7 December 2010 (UTC)[reply]
    To be fair, I suspect that the truth of the matter is that some people like some of your questions, but not everyone likes all of your questions. I certainly find some of them could do with a bit of tweaking and copyediting and changing (some of your updates don't go far enough, for example), and some could be dumped altogether. Lar, would you be willing to allow the community to edit your questions next year and select the most useful from your questions and dump the rest, or is it an all-or-nothing approach? Rschen7754 also has an individual set of questions (and a metric), and I developed a set of questions this year (but so late that probably no-one noticed). What you have to consider is the advantages and disadvantages of a community-edited set of questions, compared with multiple sets of questions from individuals (see what JayHenry said elsewhere about how questions from individuals can be overly political). If you want an example of what to dump from your questions, it would be the color question. :) Clearly some psychobabble there... :) :) Surely not just deliberate annoyance designed to test the candidates' reactions to smilies? :) :) :) Surely not trying to be or look clever? :) :) :) :) Carcharoth (talk) 03:47, 7 December 2010 (UTC) OK, maybe that question is designed to be funny, but you know what they say about repeating the same jokes year after year.[reply]
    All or nothing? Hardly. I'd be delighted to see some of my questions (including their nuances) edited and included in the standard set, which as I have said, was sadly lacking this year. I offered them up last year, and I offered them up again this year. None were taken by the coordinators. It struck me that it appeared (perhaps not actually true, but the appearance was there) that Tony's got a preconceived notion that they are no good, and that no one likes them, and as an influential coordinator, that meant none were incorporated. If some get incorporated next year, great. If not all got included and I was still convinced some more depth was needed in some areas I'd probably still tack some on somehow. As for the color question, it's designed to be funny, but it's 1 of 11, and it's interesting to see what people do with it. Ref my RfA where someone asked a "funny" about Lir/Lar and was impressed at what I did with it. ++Lar: t/c 15:26, 7 December 2010 (UTC)[reply]
    Speaking as someone who asked no questions but read the answers to every question posed, I found them extremely helpful--both the custom tailored ones and the bulk ones (Lar's especially). Lar's questions demanded a multi-part format (they would not have worked "one at a time"). I don't mind the way it was formatted (one per person on the front page, the rest on the discussion), if only because I can imagine some voters wouldn't have wanted to wade through that many questions and answers. I don't understand how someone can be an Arbcom member, given how many times we were told that it's an insane amount of work, and not be interested or willing to take the time to answer the key questions. I strongly disagree with Tony1--the questions Lar posed, while not definitive, cut to some of the key issues that need to be faced. Furthermore, answering them required the candidates demonstrate the ability to answer nuanced questions seems to me to be a key Arbcom skill. Qwyrxian (talk) 14:47, 6 December 2010 (UTC)[reply]
    Nicely said, Qwyrxian. I agree with those thoughts. --Elonka 16:46, 6 December 2010 (UTC)[reply]
    I didn't mind the questions, I'd prefer they be cut down though. DC TC 20:56, 6 December 2010 (UTC)[reply]
    I actually disagree with Qwyrxian. Spending a great deal of time just telling someone what they want to hear is not a key Arbcom skill. Especially not when it's just pandering to an editor self-absorbed enough to ask every candidate a dozen politically-loaded questions. We're moving in the wrong direction by embracing this notion that it's okay to pointlessly waste the time of arbcom candidates, members and the community like this. One reason the ArbCom has become so inefficient is this: the election process self-selects those who enjoy tedium. --JayHenry (talk)
    Except that {I, probably others} wasn't being told what I wanted to hear. In fact, the answers more often led me to oppose candidates than support them. Questions are key for someone like me: highly interested in the project, following key policy discussions and noticeboards, but active for less than a year. Examining the answers to the questions certainly took me time, but it was possible--there was no way I could somehow look through the editing history of all of the candidates, looking for key points that would help me figure out what their philosophies were in practice. Not having questions seems to me like saying that voters/interest groups shouldn't be allowed to question political candidates, and that people should only make up their mind based on searching through the person's floor speeches and voting records. Ideal, of course, but not in the least bit practical. This is the same reason why the voting guides were valuable--not because they told me how to vote, but by using the guides as a lens, I could see how others perceived the candidates. Without the questions and guides, the election really comes down to just popularity and/or voting by a small, elite few. Qwyrxian (talk) 23:52, 6 December 2010 (UTC)[reply]

Hypothetically speaking, how would those opposed to the mass-questioning propose to put a stop to it, in light of the failure of the measures adopted this year? The point has been made that prohibiting questioners from engaging in this activity on election pages wouldn't stop them from doing so on candidate user talk pages, for example. Skomorokh 22:34, 6 December 2010 (UTC)[reply]

The election volunteers could remove them. DC TC 22:37, 6 December 2010 (UTC)[reply]
Not from user talk pages. My point is that if the only consequence of prohibition is relocation, then we ought to be talking about the which locations are best for the mass questions. They were initially allowed on candidate question pages, and this year only the talkpages of those question pages; was that any better? Skomorokh 22:41, 6 December 2010 (UTC)[reply]
I think it's best to push these sorts of things as far off official pages as possible. The question page was still ridiculously long, but was dramatically improved over recent years. For editors who do care, but don't have dozens of hours to devote to elections, it's best to strive for something manageable. This improves enfranchisement. The current system was good, and could be improved further by reducing the generic form questions. --JayHenry (talk) 22:52, 6 December 2010 (UTC)[reply]

As in past years, my questions are liked by some and disliked by others (I think rather less but I could be wrong). Most (although not all) candidates found them thought provoking, if they took the time to answer them. Some candidates couldn't be bothered. I put to you that those candidates would make poor arbcom members since they don't want us to know how they think on important matters. Yes, my questions were wide ranging. That's because the official questions, with a couple of exceptions, are sadly lacking in depth and nuance. They're softballs, for the most part. Nothing prevents mine from being adopted as official questions, of course. If some scheme is come up with to make it more difficult to ask them next year, in direct contravention of what the majority of people want, I will nevertheless ask them, in a way that adheres to whatever Byzantine apparatus is necessary. As I did this year, what I did was completely within the rules, and it was of substantial benefit to the electorate and the candidates. Despite the best efforts of the self-appointed squelchers. ++Lar: t/c 02:56, 7 December 2010 (UTC)[reply]

Generally, I agree with Lar on this - hardball questions here were generally constructive. As a candidate (and then, as a voter, reading all the other active candidates' answers), what I found problematic was formatting. I had problems with it, and a lot of other candidates had inconsistent style/format as well.
I would prefer to perhaps see if either a more strongly enforced style guide / format guide, or something like a tabular question/answer format.
For large question dumps - Lar's in particular - I'd promote some sort of inclusive system where a community vetting process of the questions includes them in the "official page". We need somewhere for candidate specific questions, but if you're asking the same set of questions to everyone, and they're not off base or abusive, then I'd prefer one bundle to work with.
Georgewilliamherbert (talk) 03:57, 7 December 2010 (UTC)[reply]
Yes, I think centralizing the questions and discussions is the way to go. ScottyBerg (talk) 16:33, 7 December 2010 (UTC)[reply]
Lar and NW seem to misunderstand the skill-base and experience required by an arbitrator, who overwhelmingly needs to be able to judge behaviour, not policy or content. Stalking is relevant, but I found too many of Lar's questions were oriented towards politics and policy. This is dangerous, since it misconstrues ArbCom as a governing body, not the judicial body that it primarily is. In previous elections, the actual skill-base relevant to the role of arbitrators became hopelessly blurred with wider political questions.

I hope Lar joins in the discussion about the general questions next year. Why doesn't he join the team of election coordinators? Then he might see things from a different perspective—one of streamlining the whole process for all concerned.

Unlike Skomorokh, I do not see the guidelines for questioning as having "failed"; no one has presented evidence that the voters wanted to wade through more text, or that discussion with and about candidates was truncated. Tony (talk) 07:40, 7 December 2010 (UTC)[reply]

No, I think you misunderstand the skill base required, as well as the things that matter to voters. I'm not only concerned with how good a judge of behavior someone is, I'm also concerned with their views on important questions. If a candidate, for example, denied that there are factions or that they are a problem, that's (to me) a show stopper... they may be a good judge of behavior but they're lacking in clue about the wiki itself. The BLP question is another example. ArbCom went right to the edge of policy and stared into the chasm in some recent cases. I want to know if the candidate thought that was too far, just right, not far enough? And why. This project (like any endeavor involving more than one person) has political aspects. ArbCommers are politicians, like it or not. So political questions are fair game. You're continuing to be confused about a fair bit, I'm afraid. Give it up, Tony. And yes, the guidelines failed miserably since they made a fair mess of the questions, led to poor formatting, made things harder to find, and so forth. ++Lar: t/c 15:26, 7 December 2010 (UTC)[reply]
I just totally disagree with this. This was the first year that the question page was even close to manageable. When these pages are all 200K it's incredibly disenfranchising to all but the most profoundly wiki-addicted. It's hilarious that we argue about franchise for people with only 150 edits and yet are trying to create a system where an evaluation of the candidate takes dozens of hours? The more we allow everyone with the self-righteous inclination to ask officious reams of pet questions the more we make the process grotesque and off-putting. Streamlining something like this is an extremely important objective for election coordinators, and kudos for their work. --JayHenry (talk) 01:56, 8 December 2010 (UTC)[reply]
Wow, lots of loaded words there. I think your view is far from universally held, and your tone is less than helpful which is not like you... ++Lar: t/c 15:15, 8 December 2010 (UTC)[reply]
I hate to say this, Lar, but you seem to be shooting the messenger here. I agree with what most of what JayHenry says, though I think the phrases "self-righteous inclination" and "officious reams of pet questions" probably did go too far. The rest of what he said was fine, though. BTW, if you want to see a really long set of questions and answers, see my election questions page - it was apparently one of the longest (that was my fault, of course)... :-) What is really needed is to have no-limits discussion areas, and then to have objective journalists to report on and summmarise things to the electorate (just like happens in real elections). But then you would have to accept that some journalists (let's call them "commentators") might describe your questions in ways you might not like. How would you react, for example, to someone writing a guide or report that criticised your questions and pulled them apart one by one and dissected them to within an inch of their proverbial lives? That is what happens in real political campaigns. Do we want to go down that route? Carcharoth (talk) 03:21, 9 December 2010 (UTC)[reply]
I strongly agree with Tony and Jay. If there are no limits on the questions page, this may easily become overwhelming for voters. If you have, say, 30 candidates and 50 questions for each, you need to read 1500 questions and answers, or about 10,000 sentences. In addition, many questions and answers will contain links to ArbCom cases and various other discussions. Is that feasible for an average voter? By doing so, you will reduce the number of people who will make an informed choice to a few dozen. Is that our goal? Also I am amazed that Lar is talking about his questions only, as if quetsions of all other other editors don't matter. I am opposed to using questions of any one editor as a basis for official questions. I am not saying Lar's are bad (his constant theme is that they are good as if this is all that matters). I think they are good if a bit dry. But that's not the point. The point is that there are many other good questions - at least good in the mind of those who ask. And they have as much right to ask numerous questions as lar does. - BorisG (talk) 15:53, 8 December 2010 (UTC)[reply]
Speaking as a fairly ordinary voter, what overwhelmed me was the dizzying array of possible discussion pages, not large numbers of questions. Nor am I worried about the quality of questions. Let's say they're hostile. So what? Arbcom will have to deal with hostility. ScottyBerg (talk) 16:05, 8 December 2010 (UTC)[reply]
It's not about just mine, it's about everyone's, although I focus on mine since some folk have a habit of attacking them specifically. In past years we did not have an overwhelming number of questions. I see no reason why we would in future. I think BorisG is raising a red herring. Limiting questions is wrong. Use collapse boxes or whatever if you must, for the individually raised ones, but keep them all on one page. I am seeing many ordinary voters saying that they found the profusion of pages confusing, more so than "too many questions". (On the contrary, the voices here seem to be mostly saying they liked the questions. There are a few outliers, yes, but that's the majority I see.) One page for questions, of whatever sort, one page for discussion, of whatever sort, per candidate. Reduce the confusion, not reduce the information. ++Lar: t/c 16:52, 8 December 2010 (UTC)[reply]
Just to clarify, I thought your questions were pretty good and not hostile at all. But even if they were "bad," so what? If someone is running for arb, they can take it. ScottyBerg (talk) 16:57, 8 December 2010 (UTC)[reply]
For the record, I agree with Tony and Jay and Boris G, so maybe their position isn't such an outlier as Lar thinks. I think that keeping the profusion of questions under control is important, for the voters, not for the candidates. As far as a profusion of pages go (and to answer ScottyBerg's question) there were a total of four pages for each candidate. A candidate page (used for the statement), a statement talk page (used as the general discussion page that ScottyBerg was asking for), a questions page, and a questions talk page. It was slightly confusing, but I didn't find it a "profusion" of pages. Carcharoth (talk) 03:11, 9 December 2010 (UTC)[reply]
(edit conflict) Tony, instead of saying we are wrong all the time, can we instead stipulate that we disagree? To me, it is clear that how someone thinks about specific policies and their application makes an enormous impact on how they will vote on cases. Do you disagree? NW (Talk) 15:28, 7 December 2010 (UTC)[reply]
Seriously, seconded. Tony, your "my way or the highway" approach to this whole election has been really, really off-putting. You leave no room for anyone to disagree with how you do things; if they do, they're just plain wrong. I don't think that's fair to the community. Heimstern Läufer (talk) 04:20, 8 December 2010 (UTC)[reply]

Limiting questions cannot be done unless we're going to patrol the candidates' talk pages and remove them, which would be patently inappropriate. Candidates may ignore questions as they choose. That is what they should do if they feel overburdened by questions, not try to limit them. And before the candidates complain about too many questions, remember this: You are asking for a vote on binding sanctions. Voters have every right to want to know a lot about your views and if you don't want to talk about them at length, maybe you just shouldn't be on the committee. Heimstern Läufer (talk) 09:13, 7 December 2010 (UTC)[reply]

I agree that limiting questions is a bad idea. There is no such thing as too many questions, and pretty much no such thing as a "bad" question. If the questions are repetitious, the candidate can refer to his/her answer previously. My main concern about the process is that discussion of the candidates was scattered on a bunch of pages. There should be one centralized discussion page for each candidate, where voters could discuss why each particular candidate was good, bad or indifferent. For instance, I had profoundly mixed feelings about candidate X (no point in naming the person now) but I was unclear as to where to voice my misgivings. I think the profusion of pages might have discouraged voters from participating in the process. Yes, I could have established my own "voter guide," but a problem with the guides is that they are not conducive to dialogue with other editors about the candidates. So if I say, "I am troubled by candidate X for this and that reason," there's no simple way for another user to step in to say "yes, but take into consideration y and z." ScottyBerg (talk) 14:30, 7 December 2010 (UTC)[reply]

Exactly. One page for questions and answers, and one page for discussion. At most. The current system is balkanized and is a likely cause of less discussion. ++Lar: t/c 15:30, 7 December 2010 (UTC)[reply]
Fewer pages sounds good. But if there are no limits on the questions page, this may easily become overwhelming for voters. If you have, say, 30 candidates and 50 questions for each, I need to read 1500 questions and answers, or about 10,000 sentences. In addition, many questions and answers will contain links to ArbCom cases and various other discussions. Is that feasible for an average voter? - BorisG (talk) 11:39, 8 December 2010 (UTC)[reply]
I don't see how more information disadvantages a voter. The candidates may find it time consuming, but they will find arbcom to be even more time consuming than answering fifty questions. ScottyBerg (talk) 16:02, 8 December 2010 (UTC)[reply]
More good information does not disadvantage the voter. However, having gems of information hidden amongst noise is a huge disadvantage to the voter. Reams of extra "standard" questions that are not specific to the candidate do create a lot of noise. For example I would not be concerned with the carefully crafted answers on BLP and flagged revisions (possibly designed to win votes or at least not piss off too many voters anyway) of a returning arb rerunning for arbcom. I would be very concerned with what they had actually been doing as an arb for the last year. Tonnes of carefully crafted answers to standard questions add noise and can actually hinder us getting to the point, the real meat on the bones. Polargeo (talk) 16:25, 8 December 2010 (UTC)[reply]
Well you see, that's where I think the central discussion page would be of value to the voter. Sure, there's going to be a lot of data on the question pages. But individual voters, without going to the trouble of creating a voter guide, can just stop in and say "Candidate X was really great in his answer to Question 3a" or "I'm troubled by his answer to Question 4b" or whatever.Or "the questions by User:JohnDoe were really off-point and he handled them well." ScottyBerg (talk) 16:41, 8 December 2010 (UTC)[reply]
Another thing to remember is that questions are just part of what people need to keep in mind in voting. Maybe the candidate has a bad record. That can be discussed on the discussion page, and can result in questions on the question page. Most of these candidates I'd never heard of before, and there was no central place to have a thorough discussion of who they are. I voted for and against a couple of candidates based on only general impressions without really knowing a thing about them. ScottyBerg (talk) 16:47, 8 December 2010 (UTC)[reply]
Good points. Polargeo (talk) 16:54, 8 December 2010 (UTC)[reply]
I don't get this, though. ScottyBerg says above (and elsewhere): "there was no central place to have a thorough discussion of who they are". There was a central place to discuss each candidate. For each candidate there was a link labelled "Discuss this candidate" with a link to the candidate talk page. For example: here. These were all transcluded on this page, and that was linked from the election template, so either ScottyBerg failed to find these discussion pages, or he was looking for something different. Carcharoth (talk) 03:28, 9 December 2010 (UTC)[reply]
The discussion was separated and the supposed central place generally received little attention. This is because much of the discussion took place on the question talkpage because of the question restrictions many questions were asked on the question talkpage to get around this issue, it was a mess. Just calling something a central discussion place does not make it so. Just to highlight this, three of the candidates were not discussed at all on their central discussion pages. What is the point of leaving any message on a page virtually nobody is watching or involved with? Polargeo (talk) 11:08, 13 December 2010 (UTC)[reply]
  • I think that a large wall of questions and answers means that only those who know the "inside baseball" can realistically vote with any degree of clue. This effectively disenfranchises a large number of people. At the same time, I don't see a way to fix it. I think only open voting has a realistic chance of doing so as you can get a sense of the perceived issues and advantages associated with each candidate. But if we can't go back to that, I favor keeping things they way they are. I personally ignored the questions not on the questions page (heck I only read 20% of that...) At least it stopped one set of questions from dominating. Hobit (talk) 07:52, 1 January 2011 (UTC)[reply]

Securepoll interface

edit
  • I had no problems with the SecurePoll interface. I do think voters should be told that voting through the non-secure and secure servers will cause double-counting of votes, and also that their IP addresses will be seen by the scrutineers (i.e. if you vote and then re-vote, best to do it from a single location if you edit from more than one location). Carcharoth (talk) 02:50, 6 December 2010 (UTC)[reply]
  • Could we please have the candidates listed in alphabetical order? NW (Talk) 16:57, 6 December 2010 (UTC)[reply]
The reason they aren't listed as such is because some studies have shown candidates benefit if they're on the top of the ballot, so it's randomized to deal with this. DC TC 18:01, 6 December 2010 (UTC)[reply]
I'm fairly certain that is only the case when voters have to pick one person out of a group, not when people have to select a choice for each person. If there are other confounding variables, it is true that randomization would probably eliminate them. However, you have to weigh the slight chance of that against the fact that it almost made me (and maybe other people) screw up entering my selections. NW (Talk) 18:37, 6 December 2010 (UTC)[reply]
I think one way to fix it would be to make the ballot sortable. And I agree that the name on top probably helps more in an election were you pick a certain number of candidates, not the Support/Neutral/Oppose we used this year. DC TC 18:48, 6 December 2010 (UTC)[reply]
When I voted, I was working from a list. Which was in alpha already! Presenting me with a random list wasted significant time on doublechecking my work. If it must be random at least keep the names in alpha but start at a random point so I can slip my list by the right number of names and wrap. But sortable ballots are the best idea. Also the ability to set all votes to oppose or support rather than having to start from all neutral would be nice. (this election, setting all to oppose and then ticking the supports I wanted would have saved a bit...) that's a nice to have but alpha or sortability is almost a must have. ++Lar: t/c 03:18, 7 December 2010 (UTC)[reply]
The selection bias due to alphabetical name placement has been shown to be as high as 2.5%, which is very significant. However, NW is quite right that the relevant literature is generally considering single- or multi-winner elections, not what is essentially 21 separate support/neutral/oppose questions. Food for thought... (also)Happymelon 21:05, 6 December 2010 (UTC)[reply]
It would be helpful if the software remembered how we voted so that a voter did not have to re-enter all of them if they decided to change a single vote.--Chaser (talk) 06:47, 7 December 2010 (UTC)[reply]
The software needs a revamp so it can be used without requiring serious input from WMF developers. This might be possible if we present a community request early enough in the year. We need to re-examine which method of voting we want before doing this. October is too late. Tony (talk) 07:44, 7 December 2010 (UTC)[reply]
Strong agreement with this and with Chaser's suggestion as well. ++Lar: t/c 15:31, 7 December 2010 (UTC)[reply]
While I understand the point of randomizing candidates on the ballot, would it be possible to have the ballot and the candidate listing match once the ballot order has been determined? It was rather awkward to try to go down the ballot and the candidate statements the way they ended up. I would think that once the nomination period has ended, the ballot position could be determined, and the list of candidates refactored to match their ballot positions. Jim Miller See me | Touch me 16:41, 8 December 2010 (UTC)[reply]
The whole point is that it's random for each voter (indeed for each display of the ballot form), so that the selection bias towards candidates at the top of the list is eventually distributed evenly over all the candidates and hence neutralised. So the ballot order is never "determined", except when you load the ballot page. (also)Happymelon 17:21, 10 December 2010 (UTC)[reply]

Implications for thresholds in other elections

edit

A few months back, the election for Checkusers and Oversighters yielded fewer successful candidates than the number of positions that we had hoped to fill. There was some discussion then about whether the Secure Poll system had led to more voters opposing than public voting had used to do. In the ArbCom vote that we just had, there were some candidates with support at more than 50% and less than 60%. Perhaps there should be some discussion to assess what the community thinks about the proper threshold for success in Secure Poll elections, both for ArbCom and for the other positions, and maybe it would be better to have that discussion soon, rather than putting it off until just before the next elections. --Tryptofish (talk) 21:44, 9 December 2010 (UTC)[reply]

I agree. While I feel pretty good about the successful candidates this time around, it makes me nervous that people are getting in with support in the low 50% range. Gigs (talk) 14:32, 10 December 2010 (UTC)[reply]
Also we had only 18 candidates left in the end for 12 places. Six of these arbcom places will be taken by candidates who received around 1 support for every 3 voters. Polargeo (talk) 14:41, 10 December 2010 (UTC)[reply]
I worded my opening comment in a way that I hoped would not swing too strongly towards expressing approval or disapproval. But I'll say now that my personal opinion is that, actually, it might not be such a bad thing for successful candidates to pass with greater than 50% support, even if it's only a smidge greater. It's a trade-off: making sure the community places its trust in the appointees balanced against the need for people to fill the roles. I think things worked pretty well in this election, but a case can be made that the bar had been set too high for the Checkusers and Oversighters, necessitating appointment of more people after that election. --Tryptofish (talk) 19:32, 10 December 2010 (UTC)[reply]
I played around with the election return numbers, e.g. changing the weight of opposes and recalculating the results, but found almost nothing I did changed the election outcome materially. I would have thought these elections should be more like RFA/RFB where one oppose vote effectively cancels several support votes. Maybe I should re-check my calculations but it seems like it wouldn't have made any difference. I do think too many seats were open at the same time. 67.117.130.143 (talk) 09:32, 13 December 2010 (UTC)[reply]
  • I second the concerns here. I feel okay about our winning candidates this time around (the ones I looked into, anyhow), but there is room for concern when people are getting ArbCom type privileges with support from fewer than 40% of all people voting. RayTalk 23:17, 16 December 2010 (UTC)[reply]

Voter guides

edit

I disliked these last minute changes to the template. A format should be decided upon early and then not changed after that. NW (Talk) 02:22, 6 December 2010 (UTC)[reply]

Rather late in the election, as the number of voter guides had grown to overwhelm the template, the list of guides was placed in a collapse box. For next year, I think a list of all the candidates should be placed inside a similar collapse box, and the voter guides only accessed through a single link. Voter guides are all very well, but the focus should be on what the candidates themselves are saying, not an undue focus on what others are saying about the candidates. Carcharoth (talk) 02:25, 6 December 2010 (UTC)[reply]

I agree with Carcaroth here, it should be possible to find candidates using the templates. DC TC 02:30, 6 December 2010 (UTC)[reply]
....and in the very likely absence of feedback, I think voter guides will be more numerous....Casliber (talk · contribs) 02:36, 6 December 2010 (UTC)[reply]
The Candidates row was too demure? Listing 20+ candidates with individual links runs into the bloat problem again... Skomorokh 02:49, 6 December 2010 (UTC)[reply]
Hence the suggestion to use a collapse box. First row is the main links, including a link to a page listing the voter guides, then a collapse box showing the candidate names. Play around with a copy of this year's template to see what works best. Or set up a dummy template to play around with over the coming year (or rather, over the next few weeks before everyone loses interest again until a few days before next year's election). Carcharoth (talk) 02:53, 6 December 2010 (UTC)[reply]
That's why I rushed to make changes this year. In my experience with web development, nobody looks at the new website until after we post it live. I can put the site on a testing server and ask for feedback, and everybody ignores it. I favor using a collapse box to list the candidates, and then either a single link or a second collapse box to list the guides. Jehochman Talk 03:30, 6 December 2010 (UTC)[reply]
I am in agreement with NuclearWarfare, that it was very unpleasant to have sudden changes to the template towards the end of the election. It was also very disappointing to see personal attacks and accusations of bad faith being levied against guide writers.[2][3][4][5][6] --Elonka 04:04, 6 December 2010 (UTC)[reply]
Elonka, it is unhelpful the way you continuously wikihound me with accusations. Please stop talking about me in all venues. If I say something in a community discussion, you are welcome to disagree, but you are hardly -- hardly -- anywhere near objective when you talk about me. Jehochman Talk 04:19, 6 December 2010 (UTC)[reply]
OK, you've both had your say. In the interests of this feedback page being kept useful for others to give their opinions, I suggest that any disagreements between you two (Elonka and Jehochman) be taken to one or other of your talk pages. Carcharoth (talk) 04:33, 6 December 2010 (UTC)[reply]

Linking voter guides in a template is unnecessary and should be dispensed with next year in favor of a subpage of WP:ACE2011. jps (talk) 03:45, 6 December 2010 (UTC)[reply]

Voting guides were helpful. What wasn't and isn't is that some editors decided that they have the right to come up with unclear inclusion criteria (how do you judge good faith or whether the guide is written responsibly, and why they have to be serious, and satire is not allowed...). The self-appointed censors intervention is not welcome; by calling some guides inappropriate, their actions might have created the same bias that (AGF) they wanted to eliminate. I believe that all guides should be eligible for inclusion, unless there is a consensus from at least five editors other than the guide's author (or 80% of discussion participants, min 6) that it is unhelpful and needs to be removed. (For the record, my guide was uncontroversial and was never removed by anybody, but at least one other guide I enjoyed reading was subject to censors' cuts). In a free society, people should be allowed to express their opinions, as long as the guides don't violate our policies (such as CIV/NPA), censorship of them should not be allowed. --Piotr Konieczny aka Prokonsul Piotrus| talk 04:33, 6 December 2010 (UTC)[reply]

  • I looked at every single voting guide and they did help me decide who I was going to vote for. I appreciate the thought and effort the writers put into their guides. Cla68 (talk) 11:52, 7 December 2010 (UTC)[reply]
  • Just wanted to note here (for the record) that other voter guides did exist, both in userspace and on talk pages. My one was here, but I never really finished it off (or even stated who I supported or opposed). It was more intended to draw together the questions that I asked the candidates. I am aware of one other election guide, which was here. I suspect that one got more views than my one. :-) There may have been other guides as well - a search restricted to the userspace namespaces usually ferrets most of them out. A quick search using "intitle:ACE2010" found at least two more: User:Atmoz/ACE2010 and User:Access Denied/ACE2010. Rather scarily, a userspace search for the terms "arbitration", "election" and "2010" gets 261 hits in userspace alone (and over 3000 if you include user talk space), though lots appear to be false hits. Carcharoth (talk) 00:25, 10 December 2010 (UTC)[reply]
I thought the reason given in User:Atmoz/ACE2010 for supporting Sandstein was interesting. PhilKnight (talk) 02:15, 10 December 2010 (UTC)[reply]

Guide writer power

edit

The following was initially placed on the main discussion page

 
Page traffic statistics for the 2010 Arbcom election guides, during the voting period

I have put together the page traffic stats for the election guide pages (see figure) I am aware of how crude this measurement is. For example, some guides were not even available for the whole of this period. Also guides being worked on extensively by their authors during the election will have a boost to their traffic, although many of the guides were completed before voting began and I have purposefully left out the pre voting period to reduce this anomaly.

I find it interesting but not surprising that three well established guides come out on top in the page traffic, possibly giving them considerable influence. SandyGeorgia (1051), Lar (889) and Elonka (837) were 1, 2 and 3 respectively, these guide writers also published guides linked to by the election templates in 2008 and 2009. Two other guides came close to these established guides in traffic numbers and those were NuclearWarfare's (825) and my own Polargeo (768) at 4 and 5. The rest of the guides were some way behind in the traffic stats and headed up by Wizardman (582).

I invited and engaged in a discussion before the voting regarding election guides being linked to from the template (see Wikipedia talk:Arbitration Committee Elections December 2010/Archive 1#Election_guides) because I was concerned about the potential for excessive influence of established users on the election. Several established guide writers defended their positions at the time. I think it is good to have some post election reflection, rather than just leave it to the mayhem of the run up next year. I also think the guide writers (including myself) should not dominate any discussion as involved users. Obviously I hope to be able to write a guide next year :). Polargeo (talk) 11:54, 7 December 2010 (UTC)[reply]

  • Just a small point, it was mentioned that my arguments had something to do with tall poppy syndrome [7]. Now I can modestly claim to be a tall guide writing poppy myself (see figure) I still present the same argument. More seriously this election is about the candidates, they are the tall poppies and I would not wish them to be cut down for being good, we want tall poppies on arbcom. Polargeo (talk) 11:54, 7 December 2010 (UTC)[reply]

Merely measuring page views doesn't really measure how influential a guide is. VM's guide got a lot of views but I would be very surprised if it had a lot of influence. I think people went to your guide out of novelty, to see what the "bad boy" was up to. Since it (while amusing) was almost entirely content free, I expect your numbers might be somewhat lower next year, unless you actually do more than come up with amusing/insulting pictures of the candidates. Do some analysis and give some reasons for your views on candidates, reasons that other voters can use to help form their own opinions. Sandy's views, NW's views, and mine (as well as the views of other serious writers) differed in many areas, but since we gave some analysis, some background, some examples, voters could use our various lenses to get a better picture of the candidate than if they just read the candidate's (severely length limited and thus mostly worthless) statement and answers to the standard (softball, and thus mostly worthless) questions. That's the point of a guide, which as I said before, I think you entirely missed. ++Lar: t/c 14:54, 7 December 2010 (UTC)[reply]

@Lar, first you state the obvious, then you insult me, then you lecture me in that order. If you must please comment on the issues I raise in a constructive way. Polargeo (talk) 15:27, 7 December 2010 (UTC)[reply]
If that's what you took away from what I said, I don't think you read it carefully enough. ++Lar: t/c 17:19, 7 December 2010 (UTC)[reply]
Interesting data, Polargeo, thanks for pulling it together! Just curious, where did you get the numbers from? --Elonka 15:18, 7 December 2010 (UTC)[reply]
here Polargeo (talk) 15:24, 7 December 2010 (UTC)[reply]
Does somebody have the time to compile similar statistics for the candidates statement and/or question pages? I think it would be interesting to see what materials the voters are reviewing. This would help us understand what should be expanded or given more time. Individual reflections are nice, but scientific evidence is better. Jehochman Talk 15:42, 7 December 2010 (UTC)[reply]
  • The profusion of voter guides tended to lessen the impact of any individual guide. In the aggregate, they were less helpful than a central page to discuss each candidate. This is not to say that we should stop having voter guides. By all means, keep them going, but make it easier for voters to discuss, and read discussions of, each candidate. This will benefit the candidates, as it will give them a chance to show how smart they are, and their willingness/unwillingness to engage with voters. ScottyBerg (talk) 15:42, 7 December 2010 (UTC)[reply]
  • Not sure my guide belongs in this table, as it was not released (deliberately) until after the election ended! The only traffic to that page during the election period was me editing it! Much as I'm delighted to be an outlier, it wasn't a difficult achievement... Geometry guy 22:04, 7 December 2010 (UTC)[reply]
    Mine too probably doesn't belong here. It was in the template for a few days but I removed it before the end of the election because (a) of the (pointless) kerfuffle over the template and (b) because the page history stats showed that no one was looking at it anyway. Regardless of all that, I found writing a guide to be a useful and enjoyable process (for me) and will, assuming I'm still around, do it again next year. (P.S. I liked your guide GGuy!) --RegentsPark (talk) 15:09, 9 December 2010 (UTC)[reply]
    Thanks, but... but... you received thanks from NewYorkBrad, whereas I (sniff, sniff, brushes tear aside) received nothing :( Geometry guy 22:39, 11 December 2010 (UTC) ;)[reply]
  • I think that the argument made two posts above, that the quantity had an impact on how we perceived any individual guide, does not take into account that most people who are participating at this level of the project have the capability to separate the wheat from the chaff, so to speak. I looked at every guide as a coordinator, to make sure everything was above the belt, but when it came time for me to assess candidates, I started out looking only at a tiny number of guides, and even then I only looked at guides after doing my individual and personal assessment of each candidate. I used the guides to ensure that there wasn't some well known issue that I missed, and to see what more experienced users thought. One guide I went to often. Two I visited occasionally. The others I had very easily determined to be wastes of my time. Sorry to be so blunt, but it's true. That being said, the three good guides were worth having access to, even if there were 20 other not good guides. I support for next year linking the guides to the template, but placing them in a collapse box from the start. Sven Manguard Wha? 22:29, 7 December 2010 (UTC)[reply]
I expect nothing less from an election coordinator than thorough assessment of all of the information. Still hovering on 3 guides can mean you end up receiving a large amount of your quality information through a very narrow prism. I would be interested in how many less experienced voters there were (say using a rough <4000 edits) for example. Polargeo (talk) 09:21, 8 December 2010 (UTC)[reply]
I neglected to mention it, as I didn't see it as being on point at the time, but I did spend more than a bit of time gathering opinions and assessments from other people I trust via conversations on the IRC. I did not rely only on the opinions of three people. That being said, when it came time for me to cast my vote, the most influential opinion in the matter was my own. As far as I am aware, my vote did not match exactly any of the guides, or the intended votes of any of the people who gave me their intended votes. Don't confuse a limited number of guides used with being under-informed. Sven Manguard Wha? 04:11, 13 December 2010 (UTC)[reply]
  • Also, don't think that just because a guide is highly trafficked that that meant it necessarily had overpowering influence. I visited many of the guides numerous times, but since I know I disagree with the guide writers, many times I was using them as a lens of "if this is what this guide writer thinks, then maybe this candidate isn't actually as good as I thought..." They still have influence, but not necessarily in the way they may have desired. Qwyrxian (talk) 02:51, 8 December 2010 (UTC)[reply]
  • Question for Polargeo - On the graph, are those units in the vertical scale individual hits or IPs accessing the page or logged in users accessing the page or what? Thanks. Georgewilliamherbert (talk) 03:04, 8 December 2010 (UTC)[reply]
Just hits. So I do understand there are many issues with interpretation of what on earth this means but it does appear to come out with sensible results that give at least some indication of how much viewing each guide received. Going on nothing more than my gut feeling it appears to represent the guides which users turned to. Polargeo (talk) 09:14, 8 December 2010 (UTC)[reply]
My connection slow and I have to jump off soonish - anyone wanna run through and rank the guides WRT the actual results and see whose were most on the money? A guide league table :) Cheers, Casliber (talk · contribs) 00:35, 9 December 2010 (UTC)[reply]
Um. My guide tells how I thought the election should come out, not how it would. It for example opines that you should not have been elected... that's different than predicting that you nevertheless would be anyway. Perhaps if you want such a thing a mechanism for mailing in guesses to somewhere and revealing after the fact could be devised. ++Lar: t/c 01:46, 9 December 2010 (UTC)[reply]
The first person pronoun being the key word there Lar - in which case it gives a guide to readers how the guides actually relate to the community voting, and might be worth noting next year. Casliber (talk · contribs) 02:11, 9 December 2010 (UTC)[reply]
Sure. Such a league table would be ideal for those voters who want to vote the way they think the election's going to come out rather than the way they think things ought to be. Perhaps that's interesting to some folk. In real life I vote Libertarian every year, not because I think they're going to win, but because I think they should. I have some (not very good) friends who pride themselves on claiming they always voted for the winner. Why exactly did they bother to vote at all? ++Lar: t/c 03:38, 9 December 2010 (UTC)[reply]
Yeah, well unlike Cas, I knew who to ask: User talk:Elonka#Voting table. Carcharoth (talk) 03:51, 9 December 2010 (UTC)[reply]
Meanwhile I worked up User:SandyGeorgia/ArbVotes2010/Guides (see the second table)... which tries to see who advocated for/against those who actually won/lost and tots up tallies. But only for those guide writers Sandy thought were "meaningful". Anyone that's interested should check my work I was working in a hurry. A ? or neutral is a "fail" in that metric, you had to go all in for/against. I think that's what Cas wanted. If it's any consolation I wasn't the most right among the guide writers Sandy looked at. ++Lar: t/c 04:15, 9 December 2010 (UTC)[reply]
Lar's right, Cas. You endorse those you wish would win, not those you predict would. Skomorokh 03:43, 9 December 2010 (UTC)[reply]
Yep, you are all right - Elonka reminded me of the correct word I was looking for. I guess it depends how an advocator's ideal model of wikipedia matches up with a consensus model, and how they visualise and navigate the differences. Casliber (talk · contribs) 20:12, 10 December 2010 (UTC)[reply]
They are totals for the period. Polargeo (talk) 15:43, 9 December 2010 (UTC)[reply]
  • So Sandy's guide agreed with 13 out of 19 results voted for by the community. If you count the 2 withdrawals she ended up with 15 out of 22 of the results she wanted. Is this the best return for any guide writer? Her guide was also the most visited by some margin (see my figure above). Question is, does Sandy have her finger on the pulse of the community or does the tail wag the dog? I think I would get top marks from Geometry guy's cliche rating guide for that comment :). Polargeo (talk) (comment originally unsigned at 10:01 :))
Darn foiled again by the deviously clever Lar. Polargeo (talk) 15:21, 9 December 2010 (UTC)[reply]
Here's some analysis, which looks at guides in different ways, depending on how you slice it: If we are going strictly by which guides were best predictors of the voting, and we also rate things very strictly, meaning we draw the cutoff at exactly the 12th candidate, so we give one point for supporting #12, and one point for opposing #13, then the most predictive guides were:
  • Secret (17 correct)
  • RegentsPark (15)
  • SandyGeorgia (15)
  • Elonka (14)
  • Ealdgyth (14)
  • Lar (14)
If we look at the average of voter guide opinions compared to actual vote percentages, there are 3 candidates where the votes really differ from the guides: John Vandenberg and Xeno did much better in the votes than the guides, and Giano (GiacomoReturned) did much worse in the votes than the guides. There's also Georgewilliamherbert, who in position did much worse in votes than guides (16th instead of 9th), but I think that was just because the percentages were so close: In guides he got 46.67%, and in votes he got 44.26%. There were several candidates in that 45% range, so a single point could move things around a lot. In any case, looking at the obvious cases, meaning John Vandenberg, Xeno, and Giano, the guides that were more in line with the community on those three, meaning they supported Xeno and John Vandenberg, and opposed Giano, were:
  • Secret
  • Elonka
  • Aiken Drum
  • WereSpielChequers
If we're looking at guides and asking, "Who got the top 5 right, and the bottom 5 right?" the answer is, "No one". Or at least, no single guide supported the top 5 candidates and opposed the bottom 5. There were a few who had a mix of neutrals in, so managed to avoid opposing a top 5 candidate, or supporting a bottom 5 candidate. Including the neutrals, the guides which most accurately predicted were:
  • Heimstern (3 supports, 3 opposes)
  • PrivateMusings (4 supports, no opposes)
  • AGK (2 supports, 3 opposes)
  • Wizardman (4 supports, 4 opposes)
However, I must strongly agree with Lar here: The guides were not written to predict, but to advocate. I'd hate to think that guide writers next year would be writing their guides based on who they predicted was going to win, rather than who they thought should win! So YMMV. BTW, for more comparison data, check User:Elonka/ACE2010#Comparison between guide data and actual results. --Elonka 15:40, 9 December 2010 (UTC)[reply]

One thing that tends to get underestimated when considering what draws people to a particular guide is the number of times it is linked from elsewhere. By this I mean linked from places outside of the template, or just mentioned offhand during discussions of the election. Some guide writers are very active in discussions and (naturally) care about their guides and mention them a fair amount to others. Other guide writers don't mention their guides. Some guides are also mentioned by others. All this helps to 'promote' some guides more than others. Also, some guides get mentioned/discussed on the user talk page of the author, and this will also drive traffic to those guides, especially if the author's user talk page is heavily watched or followed by active users (note that this is a different value to the number of talk page watchers). Any consideration of which guides are more influential has to take all this into account. Carcharoth (talk) 10:10, 10 December 2010 (UTC)[reply]

I didn't put much work into my guide, I'm surpriced that my guide was the most accurate, also one of the lowest seen guides. Secret account 20:11, 10 December 2010 (UTC)[reply]

I didn't do anything to promote my guide either, but then it was the first year I'd done one. If I do one next year I might well put something on my user page. I think I got a fairly high success rate despite supporting Balloonman and opposing one of the most popular candidates. I'm not sure how I wound up one of the guides that was closer to the community on Xeno, John Vandenburg and Giano, it isn't as if many people read my guide and I don't know if anyone was influenced by it. ϢereSpielChequers 22:29, 11 December 2010 (UTC)[reply]

Guide amalgamation

edit

I am concerned that some of the post election analysis is taking place at User:SandyGeorgia/ArbVotes2010/Guides and not here. I was also far more concerned that during the election the author of perhaps the most popular guide selected her own group of "the most helpful ArbCom election guides." on that page. This selection was then able to draw even more attention to a limited portion of the guides that were endorsed by that user. When a guide is linked to from the template using that guide as a vehicle for the selection of other guides is not something that should be allowed to be repeated in future elections. In a way it semi bypasses the fairness and equal treatment which the template at least tries to achieve. Polargeo (talk) 15:05, 9 December 2010 (UTC)[reply]

Polargeo, this apparent issue you seem to be having with SandyGeorgia is getting ridiculous. Sandy made a particular page with guides which she found useful to her, and an analysis is being conducted of that particular sample. It has to some extent shaped her own voting (and guide) so she acknowledges that in her guide. The analysis hopefully lets her (and others who were looking at that page) see how that sample compared to the actual results. I'm not sure why you have this particular fixation on SandyGeorgia; maybe you were upset that your guide was not part of her selection - I mean, you made the effort to highlight that during the election? Or are you just unhappy that despite the Signpost article unduly focussing on your guide, Sandy's guide continued to attract the most views? I really don't know what it is. Obviously, you could have created a separate page with guides that you found helpful (in fact, they might be ALL of the guides), and you could conduct a post-election analysis. You had every opportunity to link to such a page during the election. But at the end of the day, I suppose that you really don't care about any of that because your pointy "campaign" (as you call it) is with a very different and petty agenda - to make on-wiki guides extinct. Each to their own, I guess, but I do wish you would stop trying to make an issue out of a non-issue. Ncmvocalist (talk) 16:52, 9 December 2010 (UTC)[reply]
Sorry but I do not think it is a non-issue which should be dismissed using the sort of tactics you have just used. Declaring that I have a petty agenda is very insulting and dismissive and I don't quite know how you think you can get away with that sort of language, is it because you think you can dismiss my views? I have no agenda to make guides extinct. I do have an agenda for fairness. Polargeo (talk) 16:56, 9 December 2010 (UTC)[reply]
Really? Tactics? I would not have even been made aware of your campaign/agenda, had you not clearly said "I could easily have written a load of criteria and showed exactly how I applied this criteria to my assessment but I rejected the idea of doing this and went for the eye catching low level journalistic style. I think it paid off in getting my guide noticed, next time I might do it a different way but I will still campaign for these guides to not exist, including my own. Polargeo (talk) 14:41, 6 December 2010 (UTC)" [emphasis added] I assumed that was you who said that, not someone else - am I wrong? Your agenda isn't really for analysing guides or for extracting the helpful points from them, but it's to advocate for their extinction is what I am reading from that. I find that really petty. Now you are saying that your agenda is for fairness - right, but I fail to see how both correlate (let alone for the benefit of the project) otherwise I might have been inclined to support it. I'm sorry you find it insulting and dismissive that I think it is petty, and I'm sure you also find it insulting that people make such guides despite your concerns. But I stand by my view and I've quite clearly provided reasons for coming to that view. Right now, I truly don't understand how you can genuinely think that what you are saying cannot be seen as extreme or extraordinary. Similarly, I also don't understand why you keep on focusing on a particular individual, SandyGeorgia, absent what I've said above. Ncmvocalist (talk) 17:25, 9 December 2010 (UTC)[reply]
When I said not exist what I meant is that I have no problem with users writing guides, just that they should not exist on an individual's userpage linked from the template. I have argued this case strongly and can show several diffs of myself stating just that. In fact I think you are aware of this as you were involved in the initial debate with me on guides prior to the election. At the time I didn't appreciate I was having a debate about not having guides linked from the template with mainly guide writers such as yourself. Polargeo (talk) 18:01, 9 December 2010 (UTC)[reply]
Heh, had it not been for the brief discussion I had with Roger, and the relative shortage of guides at the time (or so I thought), I would be a non-guide writer like I had always been. ;) I don't remember having a discussion with you though; you would naturally have views on the arbitrator-selection-process after being affected by the biggest arb case of the year, so I would have been curious. Still, it would be helpful to be clear about what you mean at the venue where you said it - most people are not up to date about each user's every content position; it's awful when people interpret what you say in a way that you did not intend (as the ultimate outcome is unfortunate). Similarly, it's awful when you deliberately omit key information to support a particular campaign - like the word "my" in your comment below; it makes the entire basis appear as bad-faith. Ncmvocalist (talk) 18:35, 9 December 2010 (UTC)[reply]
It was used to promote certain guides. It directly stated "summary of the most helpful ArbCom election guides" it was self selected and it received 332 hits. Bearing in mind it started 3 days after voting began and was only a subpage of Sandy's election guide this is a huge number of hits. Using the popularity of your own template linked guide to direct people to your preferred subset of guides is not good, all I am saying is please can we make sure this does not happen in the future. Polargeo (talk) 17:45, 9 December 2010 (UTC)[reply]
That sounds like bad faith to me and doesn't seem very fair in itself; you've missed the "My" bit - "My summary of the most helpful ArbCom election guides"; that means "mine" (Sandy's), not everyone's summary of the most helpful. If the most unpopular guide did the same thing or if it was someone other than SandyGeorgia, would you be having this issue? Maybe your issue makes sense to others, but based on what you are saying, it doesn't make sense to me. Personally, what you are advocating for seems to be unfair, if anything. It's not like it was a summary of unhelpful guides with the aim of disparaging others; it was just the most helpful guides to her deliberations - like the ones that raised additional points she considered - that's how I read it (in good faith). I'll await Sandy's clarification about her position and others who might get what you are saying and communicate it to me in a way that I'll at least understand. Ncmvocalist (talk) 17:58, 9 December 2010 (UTC)[reply]
What Polargeo is saying is that if the most watched guide links to a subset of all the guides, then that raises the profile of that smaller subset. Indirectly increasing the influence of that most watched guide on the election process (it leads to a kind of Selection bias). As I've pointed out on Polargeo's talk page, I believe that there is no procedural way to avoid this as long as the guide pointed to by the template is a good faith 'candidate' guide (which it clearly is) and as long as the 'Select' guide is clearly marked as personal opinions (which they are). (I should add that I think that Sandy has built her guide reputation legitimately.) However, the concern is not an invalid one and beyond pointing out the procedural kosherness of all this, I'm not sure if it needs to be so disparaged as being in bad faith. Regards. --RegentsPark (talk) 18:36, 9 December 2010 (UTC)[reply]
"bad faith" hey. Another follow up attack on my motives. There is no point debating with people such as yourself who spend most of their time attacking a user's motives rather than debating the issue. Polargeo (talk) 18:08, 9 December 2010 (UTC)[reply]

ACE2011 RfC

edit

FYI, there is an ongoing discussion about the RfC for ACE2011. If there are suggestions about the types of things that need to be discussed for ACE2011 voter guides, now is the time to bring up the questions. See Wikipedia talk:Arbitration Committee#RfC on WP:ACE2011 in the works. --Elonka 04:54, 12 September 2011 (UTC)[reply]

The RfC for ACE2011 has been opened, and can be found here. --Elonka 01:37, 19 September 2011 (UTC)[reply]

Voting system

edit
  • Not seeing anywhere else to put this (Someone moved it to a new section, thanks!) so this seems the section appropriate to comment on voting systems used. I found myself supporting a large number of candidates (17 of 19). I considered only supporting 11 or 12 to fit the number of seats, but that seemed pointless as I felt strongly that I should only oppose those where I had clear concerns about their ability to either do the job, or grow into the role. Others may have supported less and opposed more, but I would have preferred to rate those I supported relative to each other, rather than define some arbitrary cut-off point (which varies from voter to voter) where I supported all on one side and opposed all on the other side. The alternative of voting for a small number of supports, and a small number of opposes, and a large number of neutrals, I rejected as being unhelpful. Similarly, voting for only a few supports and opposing the rest, I thought would cause strange results overall, so I avoided that. Essentially, there were many different strategies voters could have chosen, but due to voting 'blind', many of the strategies make no sense but people probably followed them anyway. So the likely outcome is a mixture of results from different voting strategies, combining to produce a mixed picture. Ideally, to allow analysis of this, it would be good to have a system that allows limited analysis, such as "How many people voted all supports, or all opposes, or all neutrals, and any combinations in-between". This could be obtained by a survey of voters, but I see no reason why the system shouldn't be configured to allow this sort of analysis of voting patterns. If lots of people were following the "support 12 candidates" strategy, that would point to preferential voting being used. Other patterns might suggest other factors came into play. In short, no problems with the technical aspect of SecurePoll, but I do think a discussion (another one!) of the voting method is needed. Carcharoth (talk) 02:38, 6 December 2010 (UTC) I should note here that if everyone supported between 15 and 17 candidates, you get silly results as well. You do need a fair number of opposes for any sensible result. 03:34, 6 December 2010 (UTC)[reply]
    • I would much prefer a preferential system where you put the candidates in order rather than simply voting for or against. It would enable a much more nuanced vote as there were several candidates who I supported rather more strongly than others, it would also avoid the temptation to only vote for as many candidates as you think there are seats. But it would force us to choose what sort of preferential voting we introduce. One method would be a series of Alternative Vote counts looking at all the votes but ignoring preferences for already elected candidates, this would arguably get us the best qualified arbcom, but at the price of excluding controversial candidates who are liked by a large minority but are the last choice of the majority. Alternatively we could use the Irish system of Single Transferable Vote, this would result in a more diverse Arbcom as it could easily elect controversial candidates who are the first choice of 10% and the last choice of 90%. ϢereSpielChequers 16:03, 6 December 2010 (UTC)[reply]
      • A preferential system allows voters to express 21 degrees of favour/disfavour. The current system allows only 3 degrees and is impossibly clunky. Worse, the current system makes the intention of Oppose votes unclear: they could mean "I don't want this candidate", or "I'm voting oppose strategically so my favoured few will be further advantaged; I'd otherwise vote neutral." Yet people interpret opposes as meaning the former, only. This is why people like Sandy appear not to understand this, with her calling for a minimum of not a 50 but a 60% ranking. It wouldn't mean 60% favour a candidate. The current system builds in the possibility of failure, as we saw with the functionary elections earlier this year and could see in the current ArbCom election if a floor is applied. A preferential system would remove the notion of a floor, simply matching the most preferred candidates with the number of vacancies. Tony (talk) 16:43, 6 December 2010 (UTC)[reply]
      • I don't think STV, or any other proportional method, is right for ArbCom. If the ArbCom were a proportional representation of Wikipedia, then all the battles that Wikipedia users fight would be fought again on ArbCom. And I can see the formation of a Troll Party that would easily get a seat. Unlike political legislation, the basic rules of conduct that ArbCom adjudicates actually are things where it makes sense to impose the majority's will on everyone, so majoritarian methods are preferable. I can see Carcharoth's point that we are approaching the silly strategic zone of approval voting, and I would be in favor of the Schulze method (the same way we already elect the board). rspεεr (talk) 17:29, 6 December 2010 (UTC)[reply]
        • Sorry, why would preferential voting politicise the Committee? Schultze is a form of preferential voting, so I'm confused by your last point. Tony (talk) 07:47, 7 December 2010 (UTC)[reply]
          • Yes, both Schulze and STV are preferential. The word I said was proportional. STV is proportional while Schulze is majoritarian.
            In an idealized proportional election where 60% of the voters cast the same ballot, 60% of the seats go to their top choices. In the same situation in a majoritarian election, all of the seats go to their top choices. Proportional voting sounds fairer, and in elections that are already factionalized they are fairer. But here, a proportional election would invite factionalism where there currently isn't any. I'm all for preferential voting, just make sure it's majoritarian so we get an ArbCom that can possibly agree on things. rspεεr (talk) 04:19, 8 December 2010 (UTC)[reply]
  • I would personally not want to see a preferential system, but that's just because it doesn't reflect the way I approached the election or the candidates. I completely disregarded the number of open seats and simply approached each candidate as to whether or not I was cmofortable with them as an Arb. Yes, no, or indifferent were the only answers I had. Requiring me to rank the candidates in order of preference is more analysis than I really want to do. And I don't want my opposes given less weight due to a ranking. If I was opposed to a candidate, I was absolutely opposed to the candidate. A weighted system would require me to figure out among the ones I oppose, which is most likely to be close so that I can use the strongest possible oppose vote against them. Again, too much analysis. Jim Miller See me | Touch me 17:15, 6 December 2010 (UTC)[reply]
    • That's an argument against positional voting systems like the Borda count, where the sole distinction between different places on your ballot is how many "points" they are worth (and props to you for noticing the problem with positional voting systems so clearly). But you can be absolutely opposed to many candidates in Schulze and many other methods, without reducing the effectiveness of your votes for other people -- you just leave them off your ballot. rspεεr (talk) 17:29, 6 December 2010 (UTC)[reply]

I think there needs to be a well-planned, well-publicized, multi-step process to discuss any potential changes in the voting method. This year, in a sort-of rushed RfC before the election, there was a proposal for the Schulze method and one for the Single Transferable Vote method (by the same editor I believe). There were one or two people who claimed that the Schulze method (which I believe was supported by 11 people) won a consensus, leading to some debate over the results of the RfC. (The portion of the RfC regarding changes to the voting method was closed as "No consensus.) Hopefully a new process can involve many more editors. It would also have to end with a definite result some months before the election so that the WMF software developers (who do a great job by the way) can have enough time to make it happen. Maybe I am just dreaming here, because I have observed past RfC's on complex subjects with multiple possible outcomes (which I think the issue of voting systems is) end in a torrent of arguments about not only what the result should be, but what the result was. Hopefully that can be avoided this time. I also think that, because there are many different voting methods to choose from, the first step should not be an RfC, but sort of a brainstorming process to get all the ideas that have some level of support on the table, and then a discussion to winnow those ideas down to a manageable few for consideration in an RfC.

As for the specific methods themselves, since there seems to be some support for the Schulze method, I just have a few comments about that. I know the Schulze method is used in WMF board elections, though I don't know exactly how that came about. I do not know whether there was a consensus in each project for that, or whether it was decreed top-down by the Foundation. The reason I am curious about these questions is, after reading our article on the Schulze method, I have to wonder how many Wikipedia editors really understand it. I have to say, after reading that article, I don't understand it any better than before I read it (and I have actually read it three or four times now.) If we do get into a full-fledged discussion of the Schulze method, my first question to its supporters will be, can you please explain it in plain English? In other words, when I vote I will indicate on the ballot my preferences for the candidates... and what happens then? And no fair talking about the Schwartz set heuristic, sequential dropping or for all i = 1,...,(n-1): d[C(i),C(i+1)] > d[C(i+1),C(i)]. I think that I The Voter have a right to know how my ballot is counted and what impact it has on the election even though I am not a mathematician. The current support/oppose method has faults, but one of its strengths is that the voter knows exactly how his/her vote impacts on the outcome. In my opinion, although STV is a little murkier in that regard, it is still simple enough that the average voter can understand how his/her ballot contributes to the result. Maybe someone will be able to explain Schulze so that it meets this criterion as well -- but it hasn't happened yet. Neutron (talk) 23:59, 6 December 2010 (UTC)[reply]

Everyone is equally enfranchised, but your vote is divided between the number of candidates based on a ranking system you devise. The more that people rate candidate A above candidate B, the more likely (strengthened) the outcome that candidate A wins and candidate B loses. That's pretty much it. jps (talk) 14:22, 7 December 2010 (UTC)[reply]
  • I don't want any Schulze or STV or preferential. Straight up down and the top n candidates get it. I especially don't want Schulze. All of these more complex systems suffer a flaw in that there is always a combination of circumstances in which everyone's preferences aren't exactly honored. I'm not sure I even see the need for a neutral really. I managed to get to either support or oppose for each candidate and I voted exactly as my guide said I would. (you'll have to take my word for it, but I did) If there IS to be a change, have it agreed to by 6 months before the election, no sooner. No last minute stuff please. ++Lar: t/c 03:14, 7 December 2010 (UTC)[reply]
  • The peril of any fancy voting system is that people have to understand it. With this year's system, if the majority of voters do not oppose those whom they feel they don't know enough about to support, they just leave it neutral, it is possible to end up with candidates with very low levels of support being elected - I'm looking at a spreadsheet I knocked up here with 9 candidates and 10 voters.
these three candidates are well known to the voters
Candidate Support Oppose S/(S+O) S/voters
Alice 5 5 50% 50%
Bob 7 1 88 70
Lewis 1 5 17 10
these three candidates are not well known, but no-one has uncovered a good reason to oppose
Candidate Support Oppose S/(S+O) S/voters
Fred 2 0 100% 20%
Jane 2 2 50 20
Kate 2 1 67 20

We would elect these lower three candidates, even though only 20% of voters voted for them. In fact, by assuming that around half of votes cast were neutral, I managed to get a result where 8 candidates were elected, but only two of them had the support of 50% or more of voters. Elen of the Roads (talk) 13:51, 7 December 2010 (UTC)[reply]

Schulze is an over-determined method for arbcom elections since Wikipedia loves consensus. The problem is that the ranked order of candidates cannot be interpreted properly by those who certify the election unless it is known ahead of time exactly how many seats will be filled. The exact number of seats filled is right now subject to change by post-facto fiat, arbitrary cut-offs can be put into place which would exclude or include candidates that voters would have preferred to not have seen on arbcom. The solution to this, if you're really a Schulze fan, is to make it abundantly clear that it is a requisite for a certain number of seats to be filled but that ensures that people with very little support (in elections like this) will be seated. jps (talk) 14:22, 7 December 2010 (UTC)[reply]
Schulze is only useful when you have only one candidate to pick. If I would like to see Alice Bob and Kate elected, why would I want to rank them. It's a yes/no option.Elen of the Roads (talk) 16:57, 7 December 2010 (UTC)[reply]
That simply isn't true, Elen. If you want to use a Support/Oppose system, Schulze can easily encompass that: just rank the candidates you like as 1 and those you don't as 2 or not at all. It's exceedingly simple; see Schulze_method#Ballot. Skomorokh 17:07, 7 December 2010 (UTC)[reply]
Skomorokh, the first sentence of Wikipedia's article on the Schulze method (the same article that you have cited to a section of) says the method "selects a single winner using votes that express preferences. (My emphasis.) The ArbCom election is a multiple-winner election. Is that statement in the [[[Schulze method]] article incorrect? (And I'm asking that sincerely, I do not know whether it is correct or it isn't, and in my opinion the article is so overly-technical and confusing that I cannot tell whether it is correct or not. I also notice there is a "confusing or unclear" tag on the article, which I did not put there, so apparently I am not alone in my opinion.) Neutron (talk) 19:29, 7 December 2010 (UTC)[reply]
There are variants for single- and mutli-winner elections, yes (e.g. Schulze STV). I wouldn't trust those Wikipedia articles though, I hear anyone can edit them. Skomorokh 19:33, 7 December 2010 (UTC)[reply]
Schulze STV is like using a chainsaw to carve a turkey. There's a far simpler multiple-winner generalization of Schulze. Schulze gives you a ranking as a result, so just take the top N people in the ranking instead of the top 1. This is how we already elect the Board. rspεεr (talk) 04:23, 8 December 2010 (UTC)[reply]

Dare I suggest that my arcane fiddling about with Excel suggests that we would do just as well offering a simple "vote only for the ones you want to get elected" poll, and picking the ones that more than X% of voters supported. That avoids all the uncertainty of "ooh, never heard of Kate, should I oppose or leave it neutral?" which has a significant influence on the outcome of the election. MHO - whether the voters understood the instructions should never be the deciding factor in who wins the election.Elen of the Roads (talk) 17:02, 7 December 2010 (UTC)[reply]

I agree support and if you don't support a candidate it is considered to be exactly that, you don't support them (not that you don't care one way or the other or that you are in active opposition, just that you don't support them). Polargeo (talk) 17:09, 7 December 2010 (UTC)[reply]
Just as a historical note, straight approval voting was used in Wikipedia:Arbitration Committee Elections December 2004 and rejected in Wikipedia:Arbitration Committee Elections January 2006/Straw poll. Looking at the former link, it does seem very much like a popularity contest. Skomorokh 17:19, 7 December 2010 (UTC)[reply]
At present candidates can get through based on voter apathy towards them. And it surely only becomes a true popularity contest if voters can make individual votes for individual candidates. In the current case any candidate not supported by a voter would be exactly that, not supported. Did the election in 2004 go by numbers and not percentages and were people able to turn up and just vote for an individual without their vote being counted as a non-support for the rest? Polargeo (talk) 17:38, 7 December 2010 (UTC)[reply]
Yes, it was by total number of supports (the percentages listed seem to be support votes for that candidate divided by the total number of voters). And you have a point about the current system favouring bland candidates; this is why we need to adopt a preferential system for fairness and balance. Skomorokh 17:43, 7 December 2010 (UTC)[reply]
  • Concur that I absolutely do not want a preferential system, particularly not the Schulze system. People tend not to want to "rank" candidates they do not know well, but rank those they do know either favourably (1, 2, 3, etc.) or negatively (whatever lowest number there is). Unfortunately, preferential systems either make it mandatory to rank all candidates, or (like the Schulze system) place completely unranked candidates *below* a candidate who is ranked as "last place" (e.g., in this election, if I identified a candidate in 21st position, but didn't rank two other candidates, the one I put in "slot 21" would still come out ahead of the two I didn't rank). I am fairly certain that we saw an impact at the last WMF board elections, where a candidate was ranked much, much higher than expected simply because more people ranked him as the least desirable candidate and didn't rank other candidates. I do want to have an option to oppose a candidate straight out, and I don't want to have to do complex mathematical computations to determine the impact of my vote.

    As to "lesser known" candidates getting in with lots of neutral votes, last year's experience didn't bear that out. We'll see what this year's experience tells us, though; I don't think one year's worth of data is sufficient to really assess whether this voting system is the best for our community. Risker (talk) 19:26, 7 December 2010 (UTC)[reply]

Note: The Schulze system is not supposed to rank unranked candidates below all the other candidates. It's supposed to include "no information" for candidates who receive no number. Likely what you saw in the WMF board election was the result of a situation where few people ranked the undesirable candidate at all, but a fair fraction of those that actually did rank the candidate placed him above all other candidates. jps (talk) 23:34, 7 December 2010 (UTC)Ack! That's a modified Schulze system that I described (which is what my department uses in decision-making processes). Proper Schulze does prefer ranked over unranked. That's dumb. jps (talk) 23:40, 7 December 2010 (UTC)[reply]
That's my point about systems that demand that the voter understand how the count works. Mind, I can't talk much. The UK first past the post system looks deceptively easy (just vote for the chap you want to win) but I remember the baffled look on my daughter's face (first time voter this general election) on discovering that a vote for the third party in the race actually only benefited the chap who the bookies had in first place.Elen of the Roads (talk) 21:37, 7 December 2010 (UTC)[reply]

I hope that what everybody takes away from this discussion is that when we reach the final step in the decision-making process about next year's election -- regardless of whether that is in the form of an RfC, a "poll" or something else -- the different options need to spelled out clearly as mutually exclusive choices, with all the "brainstorming" having been done at a previous stage in the process. Since there are more than two choices and some of the choices may have variations, people should be permitted to give rankings of the choices, i.e., First choice, Second choice, Third choice, much as the ArbCom itself does when voting on a proposed decision. I have seen some "community polls" on Wikipedia that worked that way, but others that did not, and I'm hoping that this one does. For example, it may be that someone whose first choice is "Support/Oppose voting with a 60% support requirement" would not favor "Support/Oppose voting with a 50% support requirement" and should therefore have the ability to make one of the other options (i.e. approval voting, Schulze, etc.) their second choice. I want to make it clear I am not arguing in favor of or against any of the options right now (although I guess I have hinted at my reservations about the Schulze method earlier), but rather I am focusing on the process for arriving at a community decision. Neutron (talk) 21:50, 7 December 2010 (UTC)[reply]

Well it looks like Sandstein has lost this race, despite getting more support votes than 5 of the winners. Anyone else think it's time to either use a plurality system (basically what we have now, but without the supports) or adopt a new system. DC TC 04:02, 9 December 2010 (UTC)[reply]
More people voted for Sandstein, but more people voted against him as well. Half of those who expressed a preference didn't want him as an Arb. In a 'first past the post' type system, he would have got in, but would that have been any more representative an outcome? I don't understand what you mean by 'plurality system' - could you explain (preferably with coloured pencils)Elen of the Roads (talk) 18:59, 9 December 2010 (UTC)[reply]
Plurality was probably the wrong term (you're correct that I meant first past the post). DC TC 19:03, 9 December 2010 (UTC)[reply]
No worries. I've seen so many voting systems go by these last few days, most of such impenetrable complexity, that the idea of drawing lots like the Athenians did, or Jimbo going 'eeny meeny miny moe" has developed a certain appeal. Elen of the Roads (talk) 21:32, 9 December 2010 (UTC)[reply]

I like the idea of maybe two or three levels of "Support", although I don't see any real benefit in having different levels of "Neutral" or "Oppose". If we allowed, for instance, "Strong Support," "Support", and "Weak Support," with no more candidates in the higher positions than the number of spaces open, I think it would allow voters to indicate which candidates they really want to see elected, while also indicating those that they would accept if that were how things developed. In this scenario, those with the higher degrees of support would be more likely to be chosen, while those who received weaker support would still be eligible if there weren't enough candidates with more support to fill the positions. John Carter (talk) 19:08, 9 December 2010 (UTC)[reply]

No thank you. If another editor gets to place 1, 2, or 3 support votes on a given candidate, then I wish to preserve my right to cast 1, 2, or 3 oppose votes on that candidate. I noted above that I believe that this is too much analysis to ask of the voters, but there should not be proportional supports without proportional opposes. Oppose votes kept at least one candidate from being elected, and I don't know that we should question the veracity of those opposes (AGF!) by weakening the value of an oppose vote in such a way. Jim Miller See me | Touch me 22:18, 9 December 2010 (UTC)[reply]
I'd strongly favour Schulze as it's adaptable to meet any sort of preference for voting style. If you want to "strongly support" a number of candidates, you can rank them all 1; "strongly oppose", you can rank as 999. The numbers actually don't matter, but it helps the voter feel as though they have expressed their intentions. It also allows as many shades of preference as there are candidates, or as few divisions as desired. I disagree with Lar that "there is always a combination of circumstances in which everyone's preferences aren't exactly honored" - it's clear that Schulze is better at honouring voters' preferences than anything else suggested (see Schulze method#Satisfied criteria). At the risk of over-complication, you can even use it to produce a cut-off for the number of seats elected: simply introduce a dummy candidate called "Cut Off" and let the voters include it in their ranking (the intention is that they wish to see elected those candidates that they ranked higher). The result of the election using Schulze would then place the candidates in rank order, with those ending up below "Cut Off" judged as failing to gain support from the community. I doubt you'll find any other system able to express community consensus so accurately using a single ballot. --RexxS (talk) 22:39, 9 December 2010 (UTC)[reply]
The reason I proposed the fewer tier system in the first place was to, basically, reduce the possibilities of "ties". Something like I proposed would allow a given voter the right to strongly support as few as zero candidates, or as many as the number of positions to be filled. The latter point, which would limit the possibility of everyone getting "strong support" or the equivalent and reducing the possibilities of ties or near-ties, seems to me to be an advantage. John Carter (talk) 18:58, 12 December 2010 (UTC)[reply]

Election personnel

edit

The fact that the developer, scrutineers and election admins were headhunted and selected by myself and the other self-appointed coordinators to operate a system that selects the most powerful group of editors of the project as well as allow officials to see any voter's IP info without having that logged sits very uneasily with me. Coordinators have a great deal of discretion, and the setup is ripe for abuse, cabalism and manipulation given the small number of people involved and the lack of any community mandate. Pre-selection by voting would likely be overly bureaucratic but these issues are definitely worth keeping in mind in the context of the secret ballot system. Skomorokh 03:08, 6 December 2010 (UTC)[reply]

  • It sort of goes without saying, but coordinators should not publicly endorse any candidates, nor should they write guides. As long as coordinators act responsibly, there is no further structure needed. If a coordinator were to get out of line, presumably the other coordinators would ask them to step away from the process. Jehochman Talk 03:26, 6 December 2010 (UTC) <<< Election Coordinator[reply]
    • Ya, I don't completely agree with that view. It depends on what exactly the coordinator is doing. Technical work on template formatting and the like isn't disqualifying from having views on candidates and making them known. But avoidance of even the appearance of impropriety is important, so as soon as you step away from completely routine gnoming, then no... (for example making any sort of decision about whose guide is in or out of the list isn't gnoming). ++Lar: t/c 03:07, 7 December 2010 (UTC) <<< Election Coordinator[reply]
  • Agree that the concerns Skomorokh raises need to be addressed and tackled head-on, though I don't have any clear answers. What would be a good start is for those commenting here to declare whether or not they are election coordinators, and a list provided here of all the election personnel, including co-ordinators (rather than just a link to a list). I should add that I think the election co-ordination has been handled well to date. Any criticism are of the system, not those who volunteered this year. Carcharoth (talk) 04:37, 6 December 2010 (UTC)[reply]

I'm going to repeat this, cut straight from my comments above, as I find this issue to be one that bothered be greatly. This election, two editors signed up as coordinators and were subsequently asked to step down (by editors at large, not by election staff or other coordinators.) One did not. One did, and then rejoined after I persuaded him to do so. Asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable. Like all other non-elected positions and tasks on Wikipedia, being a coordinator is an individual editor's personal choice. If you believe that there should be a guideline as to who is eligible to serve as a coordinator, bring it up in an RfC. Barring that, let me repeat, asking, telling, or demanding people not to be coordinators is completely and utterly unacceptable. Sven Manguard Wha? 05:48, 6 December 2010 (UTC)[reply]

  • Skomorokh says, "the setup is ripe for abuse, cabalism and manipulation given the small number of people involved and the lack of any community mandate"—Many roles have no community mandate. To start with, the featured article and FAR directors; the selectors of some of the main-page material; the managing editor of The Signpost, not to mention the journalists; the ArbCom functionaries; and let's not forget Jimbo.

    Three factors militate against abuse and manipulation by the election coordinators: (1) anyone can self-select, and that goes for editors who do not trust anyone who has already self-selected; (2) the actions of coordinators are largely exposed to scrutiny and community complaint; and (3) those same actions are subject to scrutiny by the independent stewards.

    I am more concerned that running the election smoothly now requires experience and skill that are not readily transferable. Thanks to Happy-melon, we now have a page that sets out the scrutineers' and election admins' roles; we need an equivalent for the election coordinators, so that a larger pool of editors might feel reasonably confident in walking into the role in future years. Skomorokh and I both had the advantage of experience from ACE2009; I don't know how it could have been done without this experience. We need the process to be viable with totally new personnel, if that were to be necessary. Tony (talk) 14:17, 6 December 2010 (UTC)[reply]

    • Cutting away some of the thicket of rules about various things might make the process a bit more repeatable. ++Lar: t/c 03:07, 7 December 2010 (UTC)[reply]
      • Lar, there's an irony on a wiki that a procedure needs quite elaborate rules to restrain bloat; think of it as transferring bloat from the operational side to the process side (i.e., making the grid a little more rule-bound to ensure that what fills the grid in practice is not overwhelming). That was the lesson I learned in designing my failed "AdminReview", which was an interesting experiment. Bloat is the enemy of wikis, and is certainly the enemy of a calm, professionally run election, and reasonably prompt arbitration cases. Are you going to be a coordinator next year? That would be good, IMO. Tony (talk) 07:57, 7 December 2010 (UTC)[reply]
        • I'm not sure I agree with you about what makes a good election. Or a good arbitrator for that matter. As to coordinating, if there is a hard and fast rule that coordinators can't write guides or share their views on candidates during the process, then no. Asking questions and writing guides is more important to me. That said, I was a coordinator this election (although not the way you're thinking, I expect), I did help with various gnomish things here and there. ++Lar: t/c 15:09, 7 December 2010 (UTC)[reply]
  • Having watched this cycle of elections, people then proceeding to sit on the ArbCom, & what happens next for some time, I have found that a term on the ArbCom is more of a thankless job than some plum assignment. Difficult & contentious disputes go there to be decided & be resolved, & no matter how they are resolved the ArbCom is going to make someone upset -- perhaps a valuable contributor who decides at that point to leave Wikipedia. And that's if an ArbCom member does a good job; a bad job will lead to an even more vicious response. Then there are some people, who are elected to it, but after a gradual fall-off in their contributions vanish from Wikipedia. If I were an election coordinator, & were intent on misusing my authority, I'd be tempted more to abuse that power to select someone I didn't like to the ArbCom, not a Wiki-friend. -- llywrch (talk) 22:31, 6 December 2010 (UTC)[reply]
Agree with Llywrch that being a coordinator is a bad position from which to cherry-pick favourites, and I think people who agree to be election staff are on the whole pretty neutral about who's who, although we all, of course, voted. You get drowned in the procedure and nervous about keeping it all hanging together at certain points (like the start of noms and the start of voting); this is inevitable, and I'm sure the clerks have similar experiences.

Being an arb can be stressful, and is a drag on your time. It can involve unpleasant pressure from parties. But it has rewards in the high-profile of the role. My hope is that the procedures of the Committee can be gradually reformed to make the job less time-consuming. Tony (talk) 07:57, 7 December 2010 (UTC)[reply]

Election timing

edit

I liked that the elections were held a week or two earlier than last year, but feel that they could be moved to another time of year entirely. In previous years, the elections interfered with the holidays (Jimbo had to announce the next year's arbitrators on Christmas Eve!), and this year it interfered with Thanksgiving in the U.S. Especially with the short amount of time between the end of nominations (Tuesday before Thanksgiving) and the beginning of voting (day after Thanksgiving), it took some tricky scheduling on my part to make time to do a proper review of the candidates, and I still felt very rushed. Maybe in 2011, we could move the election process to a less stressful time of year? August, October? And then instead of having new arbs start on January 1 (which again is interfering with holidays), have an "arb changeover" date of something more reasonable, such as November 1? --Elonka 04:10, 6 December 2010 (UTC)[reply]

I think the elections should last longer. The voting period should be at least two weeks long. --Piotr Konieczny aka Prokonsul Piotrus| talk 04:35, 6 December 2010 (UTC)[reply]
Here's some data on number of voters per day. This is from the voter log, so each voter is only counted once (though I'm not sure if that means their first or their last poll)
Date Number of voters
26 November 2010 216
27 November 2010 91
28 November 2010 70
29 November 2010 95
30 November 2010 62
1 December 2010 77
2 December 2010 55
3 December 2010 40
4 December 2010 57
5 December 2010 91
Or in other words, yes, I agree that the election should have lasted longer, since we were getting a substantial number of voters each day, and it didn't look like things had trickled off. In 10 days of voting we had about 850 total voters this year, which was substantially fewer than last year, when we had a 2-week voting period, and nearly 1000 voters. --Elonka 05:17, 6 December 2010 (UTC)[reply]
I'm pretty sure voters went in the log by their first date. I revised my votes, yet still show as voting on the first day. Jclemens (talk) 05:22, 6 December 2010 (UTC)[reply]
Thanks for the clarification! --Elonka 05:27, 6 December 2010 (UTC)[reply]
I agree with Piotrus. This is a poor time of year for a lot of people—first you have Thanksgiving weekend in the United States, and then you have a massive uptake in work around December (for me at least), and of course final examinations studying for all of our American undergraduates, which is a substantial portion of our editorbase. Having a much longer election (15 to 21 days of voting, at minimum 7 days of discussion between end of candidacy filing and voting opening) would be helpful. NW (Talk) 13:47, 6 December 2010 (UTC)[reply]
I believe it is out of proportion to spend more than one month of every 12 in election mode. It is a lot of work, and a lot of disruption to many people. ACE2009 lasted 49 days; this one will probably last less than 30 days, because the general questions were organised beforehand, the nom and voting were 10 + 10 rather than 14 + 14, and the fallow period was two days. IMO, the last was the only misjudgement (my fault); I suggest it be four or five days in future. Tony (talk) 13:55, 6 December 2010 (UTC)[reply]
I agree that we want to keep this short if possible. If the problem is that late Nov/early Dec is a busy time for a significant part of our electorate, why not move the whole process to a shorter slot in a different time of the year? Would the first half of October or late Jan/early Feb clash with anything significant? ϢereSpielChequers 16:12, 6 December 2010 (UTC)[reply]
I am going to disagree with Elonka and say that I thought the timing worked out as well as it could. There were extra days off from work where I could take the time to read all of the candidates profiles and answers to their questions. There are only so many hours in a given day to spend on WP, and the election naturally results in a lot to read. This time of year (school breaks, work holidays) allows for people to take the time to inform themselves about the election and the candidates. Moving to another time of year may be more difficult for many people to dedicate the needed time. Jim Miller See me | Touch me 16:43, 6 December 2010 (UTC)[reply]

Agree. Elections should have lasted longer, and should not have ended during the weekend William M. Connolley (talk) 17:51, 6 December 2010 (UTC)[reply]

Setting aside the length of the voting period, isn't starting and ending on weekends the optimum use of time given that English-speaking peoples tend to have their most free time at weekends? Skomorokh 18:39, 6 December 2010 (UTC)[reply]
For those of us whose work squinternet is more reliable than home, during work time is better :-). But the real problem was that I didn't realise the voting period was so short William M. Connolley (talk) 18:43, 6 December 2010 (UTC)[reply]
Well, the dates have been posted during the entire election. Not to sound callous, but if you didn't know that the voting period was "so short" that is entirely your fault. I, for one, found time to read all the statements, questions, comments, and as several guides, as well as prepare my decisions, all within the 12 day period after nominations ended. I think the timing was fine. Sven Manguard Wha? 22:39, 7 December 2010 (UTC)[reply]

I'm not too sussed about the absolute timings with respect to the calendar but I am among those who think we need more time between the end of nominations and the start of voting, to allow voters to more thoroughly question and evaluate candidates, including followup as needed. ++Lar: t/c 15:12, 7 December 2010 (UTC)[reply]

The elections came when I had a lot of coursework to do for my degree course, which rather limited how much time I could spend reviewing candidates, though I still found time to review them all and create a guide. I don't have a problem with the existing ten day voting period, though I wouldn't have a problem with either extending it to fourteen days, or better still, increasing the time between the nominations and the voting to four or five days, to give more time to review last minute candidates, as suggested by Tony1. CT Cooper · talk 23:28, 7 December 2010 (UTC)[reply]

The election timing was good for me, with exams of my students finished. Would be even better if they were during the Christmas holidays since I would have more time, but maybe my idea of holidays is not shared by many :) — Preceding unsigned comment added by BorisG (talkcontribs) 06:42, 8 December 2010


Criteria for election/appointment

edit

A point I raised in the RFC prior to this election, and which continues to raise its head in discussions, is minimal criteria for election/appointment to the Arbitration Committee. The three points that are discussed are:

  1. Minimum percentage of support as calculated by support/(support + oppose). I personally favour no less than 50% support, and I do not believe it is reasonable to raise that above 60% as long as we are using SecurePoll or any other secret balloting system.
  2. All appointed arbitrators must be candidates in the election immediately before appointment, and meet all other minimal criteria. This is so that only those who have received a degree of community scrutiny will be on the Committee; no appointments from the floor or reappointments without a second election.
  3. Only candidates who receive a minimum of 150 support votes may be appointed.
  1. This last point addresses the hypothetical "what if only 4 people support but everyone else is neutral" scenario, which I personally believe is extremely unlikely. Now, I've not done any mathematical studies to determine if this is the "right" number, and I believe the minimum should be lower than what any prior successful candidate has achieved, because we cannot count on the number of voters from election to election, especially as our number of active editors continues to decline.

Thoughts to start off this discussion. Risker (talk) 17:35, 10 December 2010 (UTC)[reply]

A couple of thoughts in response:
  • I agree with you that it must be not less than 50% support, but also that there is no need to go as high as 60%. (As I also commented on in #Implications for thresholds in other elections, above, I'm concerned that we may not be taking sufficient account of the latter point in elections for other functionary positions.)
  • About your point 3, perhaps another way (not mutually exclusive, as we could do both) would be to require a minimum number for the "Net" value (ie, Supports minus Opposes). That protects against the (extremely) hypothetical 4 support, 3 oppose, and 500 neutral (where Net would be 1).
--Tryptofish (talk) 19:45, 10 December 2010 (UTC)[reply]
I would avoid selecting an arbitrary number for the minimum support level; best to keep it as a proportional measure to allow for scaling. Support from 1/5th of the voting electorate would be one such metric. Otherwise agreed on establishing these as sound principles. Skomorokh 15:18, 14 December 2010 (UTC)[reply]
I agree with Skomorokh. Carcharoth (talk) 01:42, 15 December 2010 (UTC)[reply]
Yes, I agree with that too. --Tryptofish (talk) 20:05, 17 December 2010 (UTC)[reply]

Now that Jimbo has just announced the appointments, it occurs to me that one thing we should consider about the 50%/60% issue, reflecting what has just been practiced, is to say 60% is required for a two-year appointment, and 50% is required for a one-year appointment. --Tryptofish (talk) 20:05, 17 December 2010 (UTC)[reply]

That seems reasonably straightforward and sensible to me, as someone who met the 50%, but not 60%, threshold. The other less-supported arbs and I essentially have a year in which to prove ourselves to the community, or face replacement. Jclemens (talk) 21:44, 17 December 2010 (UTC)[reply]

Do we need something in writing?

edit

This election saw a great deal of discussion generated from a lack of clarity or disputed legitimacy of aspects of the process. Contentious issues included the voter guides, the guidelines for questions and the eligibility criteria for candidates. To take a look at the volume of discord over the last issue, here are some of the longer discussions:

Many areas of the election saw a great deal of latitude in what the election volunteers could establish, including the questions guidelines, candidate eligibility criteria (uncharted waters regarding identification and disclosure of alt accounts), and the directions to the scrutineers (making inroads into steward permissions locally).

The problem is that although some of these instances of editorial boldness passed without comment this year, an editor who chose to pick a fight over them at the next election might ignite another firestorm of fractious, ill-tempered and confused discussion. The identification fracas linked above shows, I think, that we as a community are no longer satisfied to accept precedent-as-policy as an acceptable justification for election procedure, and are demanding community- or foundation-derived justification for contentious practices in order to satisfy ourselves as to their legitimacy.

So the question that comes to mind is "do we run the next elections on the fly, or is there a need for policy?" Comments solicited, Skomorokh 15:04, 14 December 2010 (UTC)[reply]

I think it's pretty clear that getting things in writing well in advance of the next election is likely to decrease drama during the election. But it's less clear what, exactly, should be put in writing. Actually, it's not clear to me what the outcome of this page is going to be. Perhaps the coordinators from the just-past election could get together, review this page and the other things you link above, and present some proposals to the community, for the community to react to. Otherwise, I have a feeling that no one is going to act on any of this before the next election. --Tryptofish (talk) 21:28, 14 December 2010 (UTC)[reply]
I think a helpful thing would be to summarise what issues/queries have arisen in functionary elections, and then look at what needs to have a stated procedure. As I said, I am happy to help collate, formulate and word anything that needs to be put before the community, so it gets there in a timeous manner. And it would be helpful if someone could code up AWB to put a notification on the talk page of everyone who voted when we get around to discussing it!! Elen of the Roads (talk) 02:01, 15 December 2010 (UTC)[reply]
You don't need AWB, you need a bot; EdwardsBot and the like do fine, they would simply work off the voter list. Having said that, last year there was similar enthusiasm, which completely dissipated after the election process was over, and it is only out of the kindness of Tony and Skomorokh that we got as much done as we did this year; nobody else showed the least bit of interest in doing the hard slogging. Decisions need to be made with enough lead time to allow significant rewriting of SecurePoll, but we also need to keep in mind that there is likely to be some significant coding change to SecurePoll in the spring due to the WMF Board of Directors election. The developers may not be willing to rewrite the software twice in a year, although that is something entirely within their scope. Risker (talk) 04:37, 15 December 2010 (UTC)[reply]