Wikipedia talk:Arbitration Committee Elections December 2013/Archive 2

Congratulations!
Congratulations to the top vote-getters and new committee members! Georgewilliamherbert (talk) 02:05, 16 December 2013 (UTC)


 * I am surprised and humbled by this result. I feel kind of lucky actually, I can try it on for a year and if I hate it it'll just be over. Beeblebrox (talk) 03:02, 16 December 2013 (UTC)


 * Thanks GWH, and thanks too to all the scrutineers, members of the Electoral Commission and other folks who put their time and effort into this. Your efforts are very much appreciated! 28bytes (talk) 03:07, 16 December 2013 (UTC)


 * Thanks, GWH. I'll try not to disappoint everyone (at least, not all at once). --Floquenbeam (talk) 03:49, 16 December 2013 (UTC)

I would like to thank the community for giving me a chance for the second year in a row and I would like to congratulate the group that got elected. -- Guerillero &#124;  My Talk  04:22, 16 December 2013 (UTC)
 * Has someone thanked the stewards? I put flowers on their talk pages when I was a coordinator. The steward policy suggests that this election-related duty is not at all the norm (to put the most liberal interpretation on it), so showing gratitude is important. Tony   (talk)  05:08, 16 December 2013 (UTC)


 * Thank you, and thank you to everyone who put so much time into this! – GorillaWarfare (talk) 05:30, 16 December 2013 (UTC)
 * Thanks, GWH, and many thanks to the coordinators and scrutineers who made sure this all ran smoothly. Seraphimblade Talk to me 03:40, 17 December 2013 (UTC)

Why is the withdrawn candidate cluttering the table?
This should be removed: it was a nuisance for voters on SecurePoll, and the ternary voting breakdown for this person is utterly meaningless. Let's be practical? Tony  (talk)  02:23, 16 December 2013 (UTC)
 * I agree, it seems a bit silly to have that in there. I've commented it out; anyone feel free to revert if you think for some reason it should be displayed. 28bytes (talk) 04:51, 16 December 2013 (UTC)
 * I think it should be displayed; as I said in my edit summary, it may indicate that a certain number of people are registering a protest vote or something. It may also serve as a caution; if a sufficiently large number of people had not marked "oppose" for this non-option (rather than just ignoring it, as most voters apparently did), then the withdrawn candidate would have scored a large percentage and could claim to have been elected. W. P. Uzer (talk) 09:35, 16 December 2013 (UTC)
 * I reckon the 785 "abstain" votes just includes everyone who voted after the candidate withdrew and didn't even see his name on the ballot, so those figures are meaningless. All we know is that by the time he withdrew, the voting was 43 support, 95 oppose, and an unknown number of abstains - that's a 31% support rate. -- Boing! said Zebedee (talk) 09:40, 16 December 2013 (UTC)
 * And having just counted a few, it looks like they all add up to 923. So those who didn't see him on the ballot after he had withdrawn have indeed been included in the "abstain" count - they did not choose to abstain, and did not register a "protest" vote. -- Boing! said Zebedee (talk) 09:43, 16 December 2013 (UTC)
 * OK, sorry, I may have misremembered - I remember seeing this "withdrawn candidate" on the ballot (unnamed), but possibly the voting options had been disabled. Still, the figures are not meaningless - they indicate the degree of support that candidate had at that point (which may be of interest to that candidate or others). W. P. Uzer (talk) 10:42, 16 December 2013 (UTC)
 * I voted after the candidate had withdrawn and did not abstain with regards to that candidate. It is therefore incorrect to assume anything about the level of support prior to withdrawing. Thryduulf (talk) 10:55, 16 December 2013 (UTC)
 * Not from the number of abstains, no; but from the number of supports and opposes (which is also what gives the percentage). I'm happy to let the ex-candidate themself decide whether to display this information or not. W. P. Uzer (talk) 11:01, 16 December 2013 (UTC)
 * Eh? The candidate received support, abstain and oppose votes after withdrawing. Without knowing when each vote for each option was cast (which would require identifying the voters) it is not possible to infer anything. Thryduulf (talk) 12:10, 16 December 2013 (UTC)
 * Hmm, I wonder - I don't remember any withdrawn candidate on the ballot, but I have a vague recollection that they might have appeared in the first version of the ballot that was withdrawn and replaced? -- Boing! said Zebedee (talk) 10:57, 16 December 2013 (UTC)
 * I distinctly remember such a line appearing on the ballot, but now you mention it, the voting options probably were disabled. W. P. Uzer (talk) 11:05, 16 December 2013 (UTC)

The withdrawn candidate withdrew before the election started but after the poll had been made. Since it would have been difficult to remove his line entirely, those in charge elected to simply remove his name. The supports and opposes are all from people who chose to support or oppose "(withdrawn candidate)" on the ballot. I know I opposed the withdrawn candidate purely on the principle of preventing any withdrawn candidates from somehow achieving a passing percentage. Reaper Eternal (talk) 11:33, 16 December 2013 (UTC)
 * Thanks, I thought my memory was failing me. I clearly remember such a line appearing on the ballot, and I certainly did not abstain.  Theopolisme ( talk )  11:48, 16 December 2013 (UTC)

Creative interpretation of the withdrawn candidate data

 * The withdrawn candidate data, a possible interpretation: The withdrawn candidate supports and opposes measure the background count rate, providing data for background subtraction.  Votes for and against the withdrawn candidate represent indiscriminate voting.  To correct, subtract 43 and 95 from all support and oppose counts.  Doing this would cause Richwales to have recorded a net support.  Congratulations, Richwales.  --SmokeyJoe (talk) 12:01, 16 December 2013 (UTC)
 * Nice idea, but it doesn't stand up to scrutiny. My vote was not indiscriminate, nor was Reaper Eternal's (based on their comment above). Thryduulf (talk) 13:18, 16 December 2013 (UTC)
 * Neither was mine, not that it matters.  Theopolisme ( talk )  13:44, 16 December 2013 (UTC)
 * Give me a break—the guy's name wasn't even on the ballot paper. Why has this meaningless rubbish been returned to the results table??? Tony   (talk)  13:49, 16 December 2013 (UTC)
 * - it shows completeness of the results. GiantSnowman 13:50, 16 December 2013 (UTC)
 * He withdrew before the start of polling. End of story. If he'd withdrawn after polling, or even during polling, it might be a different matter. IMO the ballot paper should have been cleansed beforehand. The results are irrelevant and should not clutter the table. Tony   (talk)  13:53, 16 December 2013 (UTC)
 * IIRC the candidate withdrrew in the 'limbo' between the poll being set up and it actually going live. GiantSnowman 13:55, 16 December 2013 (UTC)
 * Yes, he withdrew before the start of polling. Tony   (talk)  14:07, 16 December 2013 (UTC)
 * But after the number of candidates had been finalised. In traditional terms he withdrew after the ballot papers had been printed but before the poll, so his name would have been struck out on each ballot. In neither that scenario nor ours was there anything stopping any voter casting a ballot for that candidate, for whatever reason they chose. It would be absolutely incorrect for those ballots to have been removed from the results. Thryduulf (talk) 14:16, 16 December 2013 (UTC)
 * There is no case for changing the final scrutinised results for ANY reason. They are the absolute record of FACT. Air brushing creates a false interpretation of the final, scrutinised outcome. Leaky  Caldron  14:24, 16 December 2013 (UTC)

"No vote" changed to "Abstain"
Congratulations and best wishes to all the winners.

I changed the "No vote" column heading (and the associated descriptive text) to "Abstain", to reflect the way this year's options were presented to the voters. — Rich wales (no relation to Jimbo) 03:04, 16 December 2013 (UTC)
 * Sure—but no matter what it's called, voting neutral/abstain/no vote is actually a conscious decision on the part of a voter. It is a choice of one of the three options in this ternary-choice system. Tony   (talk)  03:46, 16 December 2013 (UTC)

Agree. "No vote" sounds too much like "negative vote". "Abstain" is more concise. —Neotarf (talk) 06:53, 16 December 2013 (UTC)

Actually "not voting" would appear to be the least confusing term here ... and should not require a third column, if a person votes neither "yes" nor "no" then the software should be able to figure out that the person did not vote on that issue. Collect (talk) 13:15, 17 December 2013 (UTC)

Description of ranking in results
Arbitration Committee Elections December 2013 says "The tally ranks candidates by level of support, defined as the number of votes cast in support of the candidate divided by the total number of votes cast for the candidate ("Abstain" preferences are not counted)." "Level of support" is vague, but it is then defined clearly; but then nowhere in this prose is it mapped to column in the table called "Percentage". And the denominator of that calculation is poorly explained as well: "votes cast for the candidate" (my emphasis) means "support" in lay language (synonymous with the numerator), not "support and oppose". Clearer wording might be "level of support among those who expressed an opinion about the candidate (the percentage of 'for' votes out of the 'for' and 'against' total)". DMacks (talk) 05:10, 16 December 2013 (UTC)
 * Well yes, I'd picked this up immediately, but held back because I've already made a pig of myself in a thread above on this issue. I also don't think the values in that column should be appended with a % sign. Percentage of what? I suppose it's S as a percentage of S+O, which disregards the real level of support for each candidate among the voters—this is a much lower figure, but is a true representation if someone is thinking of ascribing "mandate". Only one candidate—28bytes—gained the support of more than half of the voters. That doesn't bother me, but let's not fool ourselves by ascribing bloated "levels of support" from that formula, expressed as a simple percentage. Again, people voted neutral and oppose for various reasons, some of them purely strategic to advantage the supports they'd chosen (a rational method, IMO). Tony   (talk)  05:24, 16 December 2013 (UTC)
 * Re: "Percentage of what?" - it says so immediately below the table, in note 3. And of course it should have a % sign, because it *is* a percentage. -- Boing! said Zebedee (talk) 09:51, 16 December 2013 (UTC)
 * Pretty abstract for most punters to understand; it's just a mathematical construct. We could capture the "don't care" factor too—the neutrals—because who wants arbs people don't actually feel like supporting, even if they don't oppose? What about multiplying by the square-root of the neutrals, then? Tony   (talk)  14:06, 16 December 2013 (UTC)
 * Not at all abstract, because as is explicitly explained everywhere the winning candidates are those with the highest percentage of support votes out of the total votes supporting and opposing. Despite your repeated attempts to obfuscate the issue, the abstain votes have no relevance to who gets elected - there were 23 individual votes taking place here, each of which you could support or oppose. If you didn't want to take part in one vote you didn't have to. David Gerard was not elected on a turnout of 646, LFaraone was elected on a turnout of 400. If you want abstains to have any effect you need gain consensus to change the electoral system next year. Thryduulf (talk) 14:26, 16 December 2013 (UTC)
 * It's precisely-defined arithmetic - the very opposite of "abstract" -- Boing! said Zebedee (talk) 14:30, 16 December 2013 (UTC)
 * I don't see Tony as arguing against the legitimacy of the S/(S+O) formula as the metric by which the results of the election are calculated; rather, he is attacking the relevance of that metric as the best measure of the electoral mandate of the candidates (presumably in an attempt to have the metric changed in future years). It's important that we recognise, both in discussion and in the wording of the results page, that that metric stands for this election regardless of what individuals or consensus feels about it in the future.  As such I've rewritten the disputed intro somewhat; are people satisfied with that new version?  Happy‑melon 14:42, 16 December 2013 (UTC)
 * If that is what Tony is arguing (and I think you might be right) then he's doing so in the wrong place as it comes across to me at least as disputing the S/S+O metric's relevance to this election. The place to discuss voting systems for next year is next year's RFC. In the meanwhile there is this year's feedback page, which should (at least I intend to, and Carcharoth seems to intend to) be referenced at the start of planning for next year. Perhaps an essay page where he can clearly express his views and preferred alternatives and would allow the arguments to be refined between now and then would be beneficial too? I've tweaked your wording slightly to link to the RFC where this metric was chosen for this year as "established metric" seems a little weak imho. Thryduulf (talk) 15:08, 16 December 2013 (UTC)
 * I think that it would be pretty hard to change the system. Our current method is not only used for ArbCom elections. It is also used for RfAs, RfBs, et cétera all across the English Wikipedia and most WIkimedia projects. However, if we want, we could use the proper electoral system and evaluate this approach: you can only select as much candidates as seats are available, and those with the highest number of votes get elected. — ΛΧΣ  21  15:43, 16 December 2013 (UTC)
 * The system described by Hahc21 above is also known as the General ticket, and was in use until the middle of the 19th century to elect US congressman, and in New York State to elect state senators and assemblymen. Apparently some people found fault with it, and it has been replaced with other systems in the English-speaking part of the world... Kraxler (talk) 17:04, 16 December 2013 (UTC)
 * Your statement isn't exactly correct. The system Hahc referred to is known as plurality at large voting and is still used to this day in at large elections. The general ticket system you described is no longer used because Congress mandated that all U.S. House districts be single member districts, instead of at large districts. (FWIW I would support moving to such a system because I prefer using a system used in the real world instead of the support/oppose system, which makes more sense in a process like RFX than it does with private balloting.)   Hot Stop   19:11, 16 December 2013 (UTC)
 * Exactly. The system I described is used in every single country that has elections to this day. I've never heard of a presidential election were you could vote for more than one candidate; or if you have a council election, you cannot cast more votes that seats available. I believe that we could test this model and see if it works for next year's elections. Like said, we can have an RfC about this by the end of February to make sure we can make the needed modifications to SecurePoll or have the time enough to create a new extension (or voting interface, which could be easier than touching SP) in the case this (or any other suggestion, for that matter) gets approved. —  ΛΧΣ  21  19:40, 16 December 2013 (UTC)
 * Well, I'm going to point out that there are very, very few elections in "democratic" countries that are deliberately intended to result in multiple winners who will have equal responsibility and authority. The voting methodology that comes closest to those systems is approval voting, multiple winner variation - something that this community has explicitly turned down because it offers no opportunity to oppose candidates the voter believes could be harmful. Having worked a lot with SecurePoll over the past few years, I'm hardly a fan of it; however, any replacement system that is developed requires extensive testing, a great deal of care to ensure that it interacts appropriately with MediaWiki and is accepted by the Engineering Department as they will need to pull the voter lists for it. It would also need to be able to work for multi-project elections such as the Board elections. This isn't an easy fix, it will require devoted and focused developer time, plus the assistance of mathematical specialists and programmers to test the voting software. Risker (talk) 19:48, 16 December 2013 (UTC)
 * I think "used in every single country that has elections" is a bit strong, unless you mean the effectively meaningless "someone, somewhere, elects some group"; but it's certainly widely used. SecurePoll has an "approval" ballot type which allows checkbox voting; it would require a little modification to permit a limit on the number of allowable checks. I'm not passing any judgement on the merits of such a voting system; only that it would be easily technically possible. Happy‑melon 19:53, 16 December 2013 (UTC)

Since tactical voting is in vogue, such a change would leave most seats on arbcom unfilled. -- Guerillero &#124;  My Talk  21:15, 16 December 2013 (UTC)


 * Any change or replacement of the existing ArbCom election method is going to have to make it reasonably clear that the chosen winners have community support. I believe this means voters must be given the option of explicitly opposing candidates; otherwise, there will be no real way of determining how many (if any at all) of the candidates really do have the backing of the community (as opposed to merely getting the most votes).


 * I do think, though, that the current S/(S+O) ranking has serious problems.  makes a reasonable point when he questions how meaningful these "support percentages" really are.  And we run a risk here of electing someone who has very little support but also virtually no opposition (i.e., a very low S, but an even smaller O).


 * I would propose a new system where people could vote just as they do now, but instead of ranking candidates per S/(S+O), the winners would be the candidates with the most "support" votes — and with the additional rule that anyone with more "oppose" than "support" votes (i.e., a negative "net" score) would be eliminated and skipped over. This would make the system much simpler — easy to understand and (presumably) to accept — while still allowing voters to register meaningful opposition to candidates they might consider to be really bad news.


 * This modified system would not require any changes to SecurePoll; the only difference would be in the way the winners are selected. To see what I mean, take the existing results and re-sort the candidates in descending order by "support" votes.  The hypothetical results in my proposed modification would be almost the same as the official results — except that Guerillero would have been elected instead of LFaraone, and the one-year seat would have gone to Guerillero instead of Beeblebrox.  (I would still not have won a seat under this modified scheme. )  Again, anyone with a negative "net" score would be eliminated from consideration regardless of their level of support; if more seats needed filling in this hypothetical scenario, the next winners in order would be LFaraone, RegentsPark, Ks0stm (skipping David Gerard and me because of our negative "net" scores), and Gamaliel.


 * I also agree that we should start planning seriously for any needed or desired changes sooner, rather than later, so that there will be enough time to reimplement the voting software if required. —  Rich wales (no relation to Jimbo) 00:13, 17 December 2013 (UTC)
 * I am not sure what problem that would be solving. I went back and looked at how the four previous elections would have been affected by your scenario.  As with this year, it would result at most one (and in one year, zero) arbitrators being elected who were not elected under the current system.  In NONE of the years (including this year) would anyone have been "skipped" over due to getting more opposes than supports.  Leaving aside order-switching and switching between one and two-year terms, here is what difference it would have made:  In 2009, AGK gets elected instead of KnightLago.  (Though interestingly enough, if there had been one more seat, FredBauder and Cla68 would have been skipped over with less than 50%, with Wehwalt getting the last seat.)  In 2010, Sandstein gets elected instead of Jclemens.  In 2011, there is no difference (except for a swap of one-year vs. two-year between Jclemens and SilkTork.)  In 2012, Elen of the Roads gets elected instead of David Fuchs.  What does all this mean?  Maybe it means your scenario gives a slight advantage to someone who is "controversial", but not so controversial that their support percentage falls below 50%. Some people might see that as a good thing, others not.  But the main point is that it would not really make a very big difference.  Neutron (talk) 00:53, 17 December 2013 (UTC)


 * (ec) Unless I am mistaken, "approval voting" (with secret balloting) was used in the first two ArbCom elections, in 2004. (I was not around then, but I have read about it.)  After a series of discussions in 2005, this system was replaced by the support/oppose system with open voting for the January 2006 election.  (The secret ballot was adopted for the December 2009 election, but still with support/oppose voting.)  Before we go too far down the road of "approval voting" (or the more real-world-like system where you vote for a number of candidates equal to or less than the number of available seats, and those with the most votes win), we ought to explore why that system was abandoned.  I think it probably had something to do (as I think Risker is suggesting) with people wanting successful candidates to have "community support", independent of how a candidate stacks up against the other candidates.  I am a little surprised that the idea of approval voting is coming back up; in the past few RfC's (though not this year's) the most popular alternative to the current system was some form of "preference voting", though it didn't go very far, partly because its proponents never really identified a specific system.  As I have said in the past, I would not mind seeing an experiment with Single transferable vote, but I doubt it would get a consensus once people understood that it is designed to elect some candidates who would not get majority support under the current system (meaning, they would get more "opposes" than "supports".)  Neutron (talk) 00:24, 17 December 2013 (UTC)
 * Your proposal is certainly much better than actually using the formula as the sole determinant, and I'd support it if there were no other likely prospect of change. But the question remains as to why voters tick oppose—for multiple reasons, not all to do with their opposition per se to a candidature. By now we're all familiar with "strategic oppose voting" as an approach to maximise the power of one's supports. Why are you so set on building in this negativism, even the barrier of 50? Support is support, I say. If you don't support a candidate, just don't tick support. Binary, not ternary. If more voters oppose than support a candidate, who's to say that many of those voters weren't just advantaging their favoured few (as I and—anecdotally—many others did)? Neutron, you hypothesise that back in the rather crude days of 2005, the simple, widely used system was dispensed with because (some) people argued that "successful candidates [should have] have "community support", independent of how a candidate stacks up against the other candidates [in support votes]". So we got a system where community "support" is defined by how a candidate stacks up against the other candidates in terms of support and oppose votes. It's just another, more complex and flawed way of defining "community support", oddly embracing "support" in negative terms. And let's not forget that one fatal flaw is that the voting intensity is brought into play for each candidate: if a candidate is widely disregarded with neutrals, in particular, 20 supports, 15 opposes, and 888 neutrals gives them a formula ranking of 57.1%. What kind of community support is that? It's an extreme example, but potentially plays out at higher numbers, too. So concerned about this scenario were those who ran the WMF Board of Trustees (including Risker) that they introduced a special "floor" to prevent it. Tony   (talk)  03:01, 17 December 2013 (UTC)
 * I agree with Neutron that if we are to change voting systems (which I'm not sure is necessary) STV is the best option. The major problem I see with it, however, is its complexity. In the first place, I'm not sure whether SecurePoll is equipped to provide a ballot that allows ranking all the candidates in order. And even if it is, it will quite likely be unable to perform the calculations required to determine the quota and distribute surplus and excluded preferences. Plus there isn't just one version of STV. We would have to decide whether to use the Hare quota, the Droop quota, the Imperiali quota or the reinforced Imperiali quota. We'd also have to choose between the random subset, Hare, Cincinnati, Wright, Hare-Clark and Gregory methods for distributing surplus votes, plus the Meek, Warren and Wright methods for dealing with secondary preferences for prior winners. Not to mention whether to use bulk exclusions and whether the single transaction, segmented transaction or iterative count approaches to distribution of excluded candidates' preferences should be used. In short, STV is a good system, but really complicated. Neljack (talk) 06:31, 17 December 2013 (UTC)


 * To quote Einstein, I think the ArbCom election system should be "as simple as possible, but no simpler". Choosing the N candidates who get the most votes is really simple, and I think we should aim for this approach except to the extent (if any) that there are compelling reasons not to.


 * At the same time, there do appear to be good reasons why we need some indication that arbs have a tangible measure of community support (as opposed to just being the N candidates who got the most votes). For example, when the WMF was asked whether non-admin arbs (if any were elected) would be granted access to private data, they agreed to do this because they accepted victory in an ArbCom election as indicating a degree of community support comparable to passing at RfA — and it's reasonable to speculate (though I do admit it is only speculation) that a major factor in this decision was the existence of both "support" and "oppose" options in the ArbCom voting process.  Additionally, I understand Jimbo has said he will not appoint arbs who don't get a net "support" in an election — another reason why we probably do need to continue offering voters the opportunity to explicitly oppose as well as support candidates.


 * I understand, as well, that some people have expressed concern about the possibility of ArbCom getting filled with unqualified, harmful candidates if the election process were solely a matter of who could get the most support. I'm not fully convinced that the existence of an "oppose" option really addresses this concern — a mob of voters supporting a "bad" candidate could, after all, vote to oppose all the "good" candidates just as easily as supporters of "good" candidates could vote to oppose "bad" candidates — but it is an issue that probably does need to be addressed somehow, hopefully in conjunction with an analysis of the various reasons why people do vote to oppose candidates as opposed to simply abstaining.


 * In my opinion, single-transferrable-vote (or any other complex preferential voting system) would unduly complicate the ArbCom election process to no good end. —  Rich wales (no relation to Jimbo) 19:21, 17 December 2013 (UTC)
 * I think STV would probably produce a "better" ArbCom - or at least, one more representative of the community - but I agree that the benefits probably would be outweighed by the complexity of the system. That includes all the wrangling that would have to go into deciding the "detail issues" mentioned by Neljack, only a fraction of which I understand.  And that would be another problem with it - because only a very small number of people really understand all that stuff, the decisions that would affect who gets elected would be made by the same tiny fraction of the community.  And then when the election was over, very few people would understand why Candidate A was elected instead of Candidate B.  So it's not going anywhere - but I don't agree that it would be "to no good end."  It would be good, just not good enough to counterbalance the disadvantages.  Neutron (talk) 23:41, 17 December 2013 (UTC)

Does electing half of ArbCom at each election cause too much disruption?
Firstly, congratulations to the winners of the election. Secondly, this election made me wonder whether we're currently going about things the right way. Is it right that we currently put half of ArbCom up for election every year? For understandable reasons, the rate of arbitrators running for re-election is low, which means at every election, we get a bunch of new ones and almost half of ArbCom is made up of new arbitrators. I have little experience of ArbCom myself, but it strikes me that that must be pretty disruptive to its running every year, with so many new arbs having to learn the ropes at once.

I understand the previous system elected one third of ArbCom every three years, but this was changed because three-year terms were felt to be too long. What about if we instead kept the terms at two years, but elected one-quarter of arbitrators every six months? That would mean holding elections twice as often, but only one-quarter of ArbCom would change at each one, creating greater continuity in the committee. Or would more frequent elections be more disruptive overall?

I'd be particularly interested to hear from current and former arbitrators, and others who have experience dealing with ArbCom, on whether more frequent elections would be a good or bad idea. Robofish (talk) 01:25, 17 December 2013 (UTC)
 * Having recently looked through the history of ArbCom elections, the turnouts and the history of the candidates who were elected, I already feel like the pool of potential arbitrators is already diluted (no offense to those recently elected). Having more frequent elections is only going to result in us electing someone unfit to be an arbitrator and more members of ArbCom to be ousted because of outcomes to tough cases. Regards, — Moe   Epsilon  01:37, 17 December 2013 (UTC)


 * I never thought I would say this, but I think the 3-year terms actually worked better. We still (re)elected about half the committee each time, but there was a little bit more depth in the ability to carry out knowledge transfer. I'm currently in the process of "on-boarding" new arbitrators, and it's a lot more work than meets the eye: I have a spreadsheet to track who needs access to what and it has 20 columns (including notes, confirmation of email, confirmation that ID was submitted). That doesn't count updating pages and preparing knowledge transfer documents all over the place. One election a year is enough for everyone.  Risker (talk) 02:08, 17 December 2013 (UTC)
 * Elections are stressful. I don't think elections every six months is a solution. Also, while three year terms might have meant lower turnover, that is quite a commitment to make and look at arbitrators who can't serve out two years, much less three. Looking over the elected new arbitrators, I know almost all of their names which means that they already have quite a presence on Wikipedia. There are not newbies getting elected here, most of these folks have been on WP for years. Yes, they need to familiarize themselves with the ways of ARBCOM but I'm sure current arbitrators and ARBCOM clerks help orient the new arbitrators.
 * But above all, this is best a question that should be asked of past arbitrators...were they prepared? Was it hard to get up to speed? What helped? What was hard? Without information from editors who have actually been through the experience of coming aboard ARBCOM, it's all just speculation on what we think they are going through. Liz  Read! Talk! 02:30, 17 December 2013 (UTC)
 * Liz, in brief, I'd say that almost every year something very big and potentially divisive hits in the first 6 weeks of a new committee's term. Some years, this helps the new team to come together and gel, and it results in a very productive committee. Other years, it results in a splintering of the committee that it can never quite overcome. And each year, new arbitrators have more and more past history to try to absorb and understand, as they're asked to amend or clarify cases from the past, or recurrent situations rooted in precedent. There's a lot to learn. Risker (talk) 06:03, 17 December 2013 (UTC)
 * As an outside observer, a big and divisive thing not occasionally strikes in the sour spot of the transition. Each transition seems to be (from an outsiders) perspective, an opportunity for profound mischief, so increasing the number of transitions can make that far worse. --Tznkai (talk) 19:52, 18 December 2013 (UTC)

I don't think elections every six months would be practical because the turnover would be far too frequent, among other reasons. Risker's point about possibly going back to three-year terms is one that probably be considered heading into the next election. I'm somewhat surprised it hasn't come up since the terms were changed in 2009.  Hot Stop   03:17, 17 December 2013 (UTC)
 * As an off-the-top-of-my-head suggestion, would a longer transfer period help? i.e. instead of electing people for 2 year terms every 12 months, we elect people for 2 year 3 month terms every 12 months. This would mean a larger committee between December and March than the rest of the year but would allow a more seamless transfer. Majorities for accepting/declining a case would be calculated ignoring those in the final three months of their term. Thryduulf (talk) 10:08, 17 December 2013 (UTC)
 * There are two things I would suggest. The first is that the core election period should be April/May or May/June and the second is that the voting be designed to end no later than 4 weeks before the end of the previous committee's term. That way, neither the election nor the transition is competing with what, for most westerners, is a very busy social and work/school season. This is, of course, my personal opinion. Risker (talk) 12:51, 17 December 2013 (UTC)

Responding to the original post, the Arbitration Committee elections each year take up an enormous amount of the community's time and effort. I don't think it's justifiable to further increase that commitment of our contributors' time and effort which is, at all times, our most important resource. A couple of years ago, when we were having periodic checkuser/oversight and AUSC member elections on top of the ArbCom elections, it became obvious to me that "elections fatigue" and diminishing returns were setting in. If we were to hold ArbCom elections more than annually, I fear that the same thing would occur here as well. Newyorkbrad (talk) 17:11, 17 December 2013 (UTC)
 * Thanks to those above for your replies. It's clear that there's no support for more frequent ArbCom elections. Other possibilities suggested above, like extending terms or changing the election time/transfer period, are things to possibly think about before the next round of elections. Robofish (talk) 21:17, 17 December 2013 (UTC)


 * Risker, any chance you might write a guide for new arbitrators, extracted out of your spreadsheet and other materials. It could be recycled each year with minimal updating. It took years for us to write a guide for stewards who work on the auditing for the elections, but we finally got there thanks to valiant efforts by a few election officials. Tony   (talk)  09:21, 18 December 2013 (UTC)

7% female arbitrators
In the future, I would like to see the number of women arbitrators meet or exceed the number of female editors, including by pre-allocation of female-only seats. EllenCT (talk) 05:33, 17 December 2013 (UTC)
 * Well, it's pretty much at that level right now. You need to keep in mind that 50% of the women candidates for this role were elected, which is a much better percentage than can be said of just about any other "identifiable" group of candidates. If there are few women candidates, it's unlikely this proportion will change; it has been years since there were more than two women on the committee at any time, and last year I was the only one. This year, GorillaWarfare is it. Speaking as a woman who is just finishing 5 years on the committee and knows the nature and volume of work involved, I can say honestly that I see absolutely no value in "pre-allocation" of seats to individuals who can't otherwise get themselves elected. If they've got the skill set and they're willing to stand, women candidates generally do quite well, and are actually considerably more likely to be elected than male candidates. Risker (talk) 05:54, 17 December 2013 (UTC)
 * Indeed, if we want more female arbitrators we need more female candidates, so the first question that needs answering here is "why are more female editors not standing for election?". Until we can answer that question, we wont know what needs to be done to increase their representation on the committee. As a starting point for that, I'd encourage any woman reading this to answer the following questions:
 * Why did you not stand for the Arbitration Committee this year?
 * What change(s) to the committee or the arbitration system would make you stand?
 * I don't expect that this will give anywhere near a representative sample or provide the answers. Certainly nothing suggested as an answer to 2 should blindly applied. This is just to get some idea of what we need to start looking into and to get some ideas to evaluate and discuss. Thryduulf (talk) 10:20, 17 December 2013 (UTC)
 * Well, speaking as a female editor who a couple people urged to run this year: 1) I didn't run because I am already harassed quite enough (about my gender, my sexuality, and my topics of interest, as well as a number of other things that have little to do with being female) due to my Wikimedia participation, and painting a bigger target on my back is not something I'm inclined to do. 2) Any system in which arbitrators are "fair game" for abuse and harassment, either on-wiki or off, is going to put off female editors who think like I do (which, mind you, is not all of them). I don't have any ideas for how to cut down on the amount of abuse arbitrators take beyond "maybe start enforcing some level of civility?", but until something happens to limit the abuse/harassment, only a very, very particular personality, in a very, very particular life situation is going to be prepared to step into that role. My sense is that fewer women than men have that particular mindset and are in that particular life situation. A fluffernutter is a sandwich! (talk) 14:59, 17 December 2013 (UTC)
 * The proposition as it stands is of course nonsense; "I would like to see the number of women arbitrators meet or exceed the number of female editors." However, assuming Ellen means a proportional representation, where does it end? How about ethnic minorities and disabled editors being "fairly represented"? I see no good case for affirmative action of any sort in a structure such as Arbcom. Remember, in a slightly different context, WP is not a democracy. Leaky  Caldron  10:42, 17 December 2013 (UTC)
 * So given that we're not aiming to be a democracy, we might perceive that women (or some other group, though probably not the ones you mention) are typically better (than men) at dispute resolution, but typically more reluctant to stand in elections, and thus the project would benefit from some sort of action which encourages them to stand. W. P. Uzer (talk) 10:57, 17 December 2013 (UTC)
 * The ideal arbitration committee would represent the broad range of backgrounds of the community it serves. Geographically I don't think we're doing too badly, but the gender of the committee is disproportionately male. Ethnicity is a harder one, as what constitutes a minority is geographically dependent, as a white British person living in London I'm part of the ethnic majority, but if I were to move to India say I'd instantly become part of a minority. I don't support affirmative action to get anyone onto the committee, but when there is no limit to the number of candidates, encouraging under-represented groups to stand is imho a Good Thing. Before we can do that however we need to understand why they aren't standing at present. Thryduulf (talk) 11:23, 17 December 2013 (UTC)
 * My point would be, would it lead to better Arbcom. decisions? Typically a case reaches Arbcom. due to the failure of Alpha males to agree amongst each other and after other forms of dispute resolution have failed. How would injecting Alpha females at the Arbcom. level lead to "better" decisions? What is a "better" Arbcom. decision anyway? Most of the cases are straightforward, despite the ludicrously lengthy processes and pompous, posturing tone of their deliberations. Maybe alpha females generally would be less pompous and posturing. If so I would vote for them.  Leaky  Caldron  11:38, 17 December 2013 (UTC)

I don't view the proposal as "nonsense". I once worked for a private consulting firm who elected its own board. We had several "inside limits" which over-rode the preferences that would otherwise arise from the voting. My recollection is that we required one board member to be younger than 40 years of age, and we had some geographical minimums by continent, to make sure we had geographical representation. I do not automatically rule out consideration of similar limits, although I can see potential for abuse. One mathematical note, I would urge that the limit be set below the proportionate representation, or problems can ensue. For example, if we decided, as a community, that there was value in having female representation, I would support a minimum that was below 7%, but not one that was effectively 7%. As an example of something I would consider supporting, I think a non-admin might be a useful addition to the committee, so would consider an inside limit that elects at least one non-admin, should any be qualified to run (possibly also requiring some minimum support.)-- S Philbrick (Talk)  19:40, 17 December 2013 (UTC)
 * "I would like to see the number of women arbitrators meet or exceed the number of female editors." If not nonsense, it's nonsensical. Leaky  Caldron  20:27, 17 December 2013 (UTC)

"Another victim of the manocentric male-ocracy!" GiantSnowman 20:30, 17 December 2013 (UTC)
 * @GiantI missed this jibe 2 days ago. Is it an attempt at humour, a veiled personal attack or intended to clarify something? Leaky  Caldron  12:59, 19 December 2013 (UTC)
 * - humour, of course. Your lack of good faith is disturbing. GiantSnowman 13:02, 19 December 2013 (UTC)
 * WP:AGF isn't a noose. Your reference to "manocentric male-ocracy" might suggest that I was referring to the OP's suggestion as nonsense because she appears to be a woman, whereas the reason I described it as nonsense is because, as it stands, it is nonsense. Or maybe you haven't read it thoroughly. Leaky  Caldron  13:23, 19 December 2013 (UTC)
 * My comment - a quote from Futurama by the way - was in no way a response to you, hence why I out-dented it. For what it's worth, I fully agree with your feeling that this is nonsense - because it is. GiantSnowman 13:38, 19 December 2013 (UTC)

As I understand it, one of the nine arbitrators elected this year (and one of the eight elected to a full two-year term) is female. That's 11.11...% of the total elected this year and 12.5% of those elected to two-year terms. As of one year ago, two of the 15 arbitrators were women, which was 13.33% of the total. I also believe (though I can't prove) that if more women ran, more women would be elected. (There were only two female candidates this year, I believe.) So if 7% percent is your goal, the voters are capable of achieving that without any "allocation", and have done so. Also, isn't it clear to everybody that the under-representation of females on ArbCom is just a symptom of the larger issue of females choosing to edit Wikipedia at a much lower rate than males? Once that problem fixes itself, the rest should fall into place. Neutron (talk) 23:04, 17 December 2013 (UTC)

If you would like to see more female arbitrators elected, encourage qualified female candidates to run. Quotas are not a good idea. We cannot get more female candidates elected than ones who run. Seraphimblade Talk to me 06:06, 18 December 2013 (UTC)


 * I am female, voted for females, and miss a female I voted for last year, --Gerda Arendt (talk) 09:27, 18 December 2013 (UTC)
 * I shall certainly miss the one who is leaving. But like I said, there can't be more than candidates. This year, we had two female candidates running. One was, well, something of an unusual candidacy that was highly unlikely to go anywhere, and the other one was elected handily. It's not like we have dozens running and only a few getting elected. To get elected, one must run in the first place. Seraphimblade Talk to me 12:43, 18 December 2013 (UTC)


 * You're very kind, Seraphimblade. :-) Realistically, though, there are many, many other things that affect the balance of the committee.  In 2014, with the exception of one lone European, all of the arbitrators will come from the US or the UK; no Australians, Canadians, Africans, Asians, etc. There aren't a lot of members of the committee whose primary focus has been content improvement/development; the few who were once primarily content editors have been away from it for a while. There is a surfeit of members with technical expertise (most of whose skills I had no knowledge of beforehand!). But designating seats for geography, for gender identity or sex, for age, or for onwiki expertise is a mugs game. The encyclopedia risks having to accept someone who's not really the right fit simply because of their preferred categorization, or not getting an outstanding candidate because s/he came in second or third in the preferred categorization.  Risker (talk) 20:05, 18 December 2013 (UTC)
 * Let me see if I have this straight... There were two female candidates and twenty male candidates. Of those one female and eight males got elected.  That is 50% of the female candidates and 40% of the male candidates.  I'm not sure I'm seeing the problem here as stated that not enough females are being elected.  It seems to me that females have a higher percentage of being elected than males.  That being said, I would like to see a few more qualified females maybe run at the election next year, assuming of course that there are more qualified females that are willing to take a the nonsense complaints and agravations that accompanies such a position. Technical 13 (talk) 14:04, 19 December 2013 (UTC)

novel proposal
EllenCT and many others raise interesting points about what the committee ought to be, and how its composition ought be determined. Added:  The list of queries below is not a catenation of desirable or undesirable attributes - it seeks to find which attributes the community finds should be given the greatest weight Collect (talk) 17:04, 17 December 2013 (UTC) 


 * Ought its members be as close to the median of Wikipedia editors as possible (eliminating any who are far from the median) on all Wikipedia issues,
 * ought it be representative of the demographics of the Wikipedia community as nearly as practicable, or
 * ought it be representative of groups within the community on the basis of opinions regarding Wikipedia policies and practices, or
 * ought it reflect those who, regardless of any demographics, are best suited temperamentally and intellectually to make rulings congruent with the scope of the committee set forth in the policy establishing it?

Once the criteria have been determined, then we can possibly effect changes to make the results in accord with those criteria. At this point, folks who wish to be "Ruler of the Queen's Navee" appear to have the edge, but I do not think anyone has actually advanced that as a proper discriminant for members of that committee. And, of course, additions to the postulated four choices are absolutely welcome. Collect (talk) 13:36, 17 December 2013 (UTC)


 * Is the (active) median Wikipedia editor an administrator?  Volunteer Marek   14:23, 17 December 2013 (UTC)
 * I expect that the median Wikipedia editor is a schoolkid adding penis vandalism to his high-school article, unfortunately, so you'll need to adjust the baseline of who you think it should be representative of... :D Happy‑melon</b> 14:28, 17 December 2013 (UTC)e
 * That's why I put the "(active)" in there. "Active" as in survives to make, say, 100 edits (though some define "active" as more than 5 edits, especially when the numbers of "active" editors isn't looking too good).  Volunteer Marek   16:30, 17 December 2013 (UTC)
 * Propose a wording for "choice 5" in that case -- which subcriteria ought to be added to make it more representative of your own opinion? Thanks. Collect (talk) 14:34, 17 December 2013 (UTC)
 * What I hope arbitrators to be is people who look at facts with an open mind free of bias, - answer yourself if this describes the median Wikipedia editor, --Gerda Arendt (talk) 14:36, 17 December 2013 (UTC)
 * Yes, but does that describe the *median administrator*? That's the more relevant question here.  Volunteer Marek   16:31, 17 December 2013 (UTC)
 * "Ellen and many others raise interesting points about what the committee ought to be". Where? The small section above? Leaky  Caldron  14:39, 17 December 2013 (UTC)
 * Going back a ways even before this year -- even one proposal for "proportional representation" in the past, etc. Probably fifty editors or more have discussed such issues -- do you wish me to list them all?  Cheers. Collect (talk) 15:14, 17 December 2013 (UTC)

OK: adding:


 * 5 Ought its members be as close to the median of Wikipedia administrators as possible (eliminating any who are far from the median) on all Wikipedia issues?
 * 6 Ought its members be those who approach all issues free of any bias or prejudgment, in accord with the scope of the committee,  regardless of any editorial opinions they may have?

I think this addresses Happy and Gerda's posts -- if not, please edit them. Thanks. Collect (talk) 14:43, 17 December 2013 (UTC)
 * I am not a fan of the Arbcom and there needs to be reform here but this doesn't make sense even to me. First we have enough trouble getting people to do this, by making it even harder we are guaranteeing that any candidates would be those who have carefully managed their wikicareer. I personally don't want more bureaucrats than we already have on Arbcom. The second problem with this is that the language "eliminating any who are far from the median" is too vague. What does that mean? Too Americanized? Not enough? Must be Catholic? Fond of cats and kids? It doesn't tell us anything and we have enough problems with extremism in this project in regards to politics and religion especially. Why does it need to be an admin, there are plenty of non admins fit for the task of Arbitrator. In fact I would argue that the committee should be part administrator, part regular editor and at least a couple members from the Foundation. Maybe someone like Maggie from the community liaison section. The committee should be a representation of all parties and stakeholders. The next problem is the demographic argument. We don't need to be appointing by statistic. We need the most qualified people (not that we have those now mind you). We should not be appointing based on race, gender, religion, diet or some other equally "arbitrary" cultural subgrouping. 138.162.8.58 (talk) 16:22, 17 December 2013 (UTC)
 * The proposed choices or queries are not catenated -- they are separate issues to see and discuss exactly what we want ArbCom members to be.  It is clearly impossible for anyone to think they meet every possible criterion -- but unless we know which criteria the community deems important, the idea that we can "improve the election process" is risible <g>.  Cheers. Collect (talk) 16:36, 17 December 2013 (UTC)

What specific, identifiable problem with Arbcom. - past, present & future - does this novel proposal intended to address? It sets out a list of desirable attributes, but what is the definition of "the problem", please? Leaky Caldron  16:57, 17 December 2013 (UTC)
 * I commend you to read the whole dang talk page above. And the proposal does not list "desirable attributes." It asks which attributes the community assigns or ought to assign the greatest weights to. Cheers.  Collect (talk) 16:59, 17 December 2013 (UTC)
 * Most of this talk page has nothing to do with the structure of the AC and attributes for being a member of it. Leaky  Caldron  17:07, 17 December 2013 (UTC)
 * I actually agree with both of you. In collect's defense Arbcom and the Arbitration process needs a makeover (along with other areas of Wikipedia), but I also agree with Leaky that before we start coming up with solutions we need to identify what the problems are. Are some of the things mentioned above problems, yes I think we can agree that most are. Not enough women, sure I think we could agree with that but that is a problem throughout the project, not just here. Do we need a wider demographic? I think we will all agree that would be good too, but how do we get there. None of these are problems with the committee or the process though. There are a lot or those so what we need to do is identify the core requirement for the committee. Are we meeting that (not really in my opinion), are we straying to far from that mandate (I believe it is), is it fair (no not really), are the cases managed fairly and appropriately (no not really), does Arbcom followup to ensure the cases are being enforced fairly (no not at all), it takes to long, its heavily biased towards admins over reguler editors, etc. This doesn't even include that Admins have to be brought before arbcom to be desysopped which I think is a waste of time and unnecessarily burdeonsome. So as you can see there are quite a few areas where we need to improve the process. If we can get more women involved and expand the demographic (which I'm not even sure what the demographics are frankly) that's great but I don't think its a critical problem. 138.162.8.58 (talk) 19:24, 17 December 2013 (UTC)
 * The community is made up of many people with their own views on this. I think the arbs we get is already the aggregate opinion of said community members. Resolute 19:49, 17 December 2013 (UTC)
 * Exactly. I would add that this kind of discussion is really one that would make more sense within a political party; what should be the attributes of the candidates we nominate?  In some places, this is done overtly; in some countries where elections are done by "party list", Party X might have a rule that Candidate #2 must be of a different gender than Candidate #1, at least one of the Top 5 must be 35 or under, and at least one must be 65 or older, no more than 6 of the Top 10 can be of the same gender, and so on and so forth.  But Wikipedia does not have a party system (whether there are factions that could be called "parties" is a different story, but the elections are not organized along those lines.)  In the electorate as a whole, however, I don't see where this discussion is going.  It doesn't really mean anything in the abstract, meaning, without candidates to match with the attributes.  And as Resolute suggests, in an actual election, the voters look at the attributes of the candidates, at least so far as those attributes are disclosed, and the aggregate of the voters' individual decisions decides who is elected.  Neutron (talk) 23:23, 17 December 2013 (UTC)

The problem (if there is one) seems to be that some people think there are too many admins on Arbcom. But as long as we are electing Arbcom, most voters can be expected to vote for Admins, because they have the experience and proven willingness to do the kind of job that Arbcom does. If one wanted change it might be to stop electing Arbcom, and instead elect a Parliament that appoints and fires Arbcom members as it sees fit, with members of the Parliament barred from serving on Arbcom. In that case one could elect non-admins to Parliament in droves. Arbcom would still be populated by Admins but these would be accountable to the Parliament on a daily basis, instead of not being accountable to anybody for two years. But I'm not sure that anybody would notice much difference in practice. Tlhslobus (talk) 16:09, 4 January 2014 (UTC)
 * Sure they would. Nobody sensible would be willing to be an arbitrator because they'd be subject to the caprices of a body that...well, has nothing else to do but meddle in their responsibilities and critique every single vote or decision each individual arbitrator makes; frankly, that's even worse than the level of scrutiny applied now. Arbcom isn't about politics. It's not about representation. It's about problem solving. Risker (talk) 18:26, 4 January 2014 (UTC)


 * Thanks, Risker, but I think you're over-dramatizing a bit. What I suggested (or half-suggested, since I haven't actually indicated I think it's a good idea, rather than one that won't make much difference but might keep a few people happy) is roughly the equivalent of democratically elected police authorities. Roughly speaking, Admins are our cops, and Arbcom are our top cops (it's not an exact parallel as Admins and Arbcoms can also be seen as our judges, etc, but it will do). And all the criticisms you make can and probably have been made of democratically elected police authorities, but in practice such authorities don't stop people choosing to become cops or top cops. Indeed it may make life easier for them by leaving them free to ignore criticism from large numbers of ill-informed nutters out there in the community, in favour of merely taking on board some of the criticisms of a much smaller number of Authority members, many of whom are actually well-informed and reasonable people. I suspect that mostly they make very little difference in practice (which is perhaps why there's no huge demand for them in most places, but not much heated opposition to them where they do exist), but in some places (such as Northern Ireland) they arguably help make the police somewhat more acceptable to some dissatisfied sections of the community. Clearly the discussion here suggests some people are not happy with the kind of Arbcom we have, and if that is a serious problem (I'm not at all sure that it is), then such an 'Authority' might offer a partial solution (though obviously there'd have to be plenty of further discussion about the details of who gets to elect the Authority, and precisely what powers it got to have, etc). But I suspect it wouldn't actually make much difference, and it's probably not going to happen, if only because there currently isn't anybody particularly interested in pushing the idea. But I thought I'd mention it, just in case somebody out there does think it's a good idea and wants to try to carry it forward. Tlhslobus (talk) 06:06, 6 January 2014 (UTC)
 * Geez, do you ever misunderstand the purpose of the Arbitration Committee. It sure as heck isn't about policing; they have no opportunity to go out and enforce policies absent a very specific request from the community. You've still not said what else this superbody is supposed to be doing; if it's just picking and choosing who gets to do dispute resolution, I can't understand why you think that's superior to the direct selection that is happening now.  Risker (talk) 03:29, 8 January 2014 (UTC)

Voter demographics
In light of some discussions at Jimbo's page, has anyone done, or is anyone willing to do some analysis of the voters in the election? I'd be interested to know the percentage of admins/active admins/'crats/Active 'crats/active non-admin editors that voted. I'd also be interested to know what proportion of votes were cast by editors who have ever been named in an ArbCom case.-- S Philbrick (Talk)  20:06, 17 December 2013 (UTC)
 * I think i could do this, but it will take a while.... — ΛΧΣ  21  20:26, 17 December 2013 (UTC)
 * I don't see a rush, but if there is any debate about turnout, it might be useful to have some stats. Unlike Political elections, where the ratio of votes to voters is unambiguous, and has meaning, it is more complicated here, so i think categories such as these would be useful.-- S Philbrick (Talk)  20:48, 17 December 2013 (UTC)
 * No rush from me either, but I was wondering if you could also get a time/vote chart to see at what rate the votes were coming in and if maybe the election was run too short of a period of time. Based on what I read of that above linked discussion, there are some that think that 900 votes in two weeks wasn't enough time for a majority to get a chance to vote, and if the rate of votes per hour supports that, perhaps it would be appropriate to offer the election for a longer period next year? Technical 13 (talk) 13:45, 19 December 2013 (UTC)
 * 900 people could have voted in 2 hours if they had wanted. 20,000 could have voted in 2 weeks. Voting is optional, reminders are displayed. People vote if they want, when they want. The entire election process is already too long. Extending the voting period would be akin to Parkinson's Law. Leaky  Caldron  13:55, 19 December 2013 (UTC)
 * Technical 13: I think I could do that. Although, Leaky has a point. I think that we could just follow Risker's advice and move the elections to another period of the year, since a lot of users are on holidays when elections take place. — ΛΧΣ  21  15:16, 19 December 2013 (UTC)
 * I don't see what relevance votes-per-hour would have to whether the voting period was long enough. It's not like people were queuing up at a polling station and got turned away when the time ran out. <b style="color:forestgreen">Happy</b>‑<b style="color:darkorange">melon</b> 15:35, 19 December 2013 (UTC)
 * , they could have all voted in the first two hours, in which case would there be any need to keep the election open for two weeks? They could have all voted in the last two hours, in which case, perhaps a more aggressive form of advertising the election may be appropriate and maybe if there had been two more hours we'd have 1,800 voters instead of 900... Without the data, there are just way too many possibilities to speculate, so I propose we just wait and see what the data  returns with indicates.  Thanks! Technical 13 (talk) 17:32, 19 December 2013 (UTC)
 * You're looking for evidence of a problem that doesn't exist. Thanks! Leaky  Caldron  17:41, 19 December 2013 (UTC)
 * Well that doesn't rely on painstaking analysis of voter demographics, you can get that just from the list of timestamps on the ballot list. Broadly, 10% of the votes are in the first two hours, 33% in the first day, then there's pretty much a perfect exponential decay curve until the final day, when the final 10% of votes come in in the last 12 hours.  Absolutely nothing remarkable, IMO. <b style="color:forestgreen">Happy</b>‑<b style="color:darkorange">melon</b> 17:58, 19 December 2013 (UTC)

Okay, so I got the first piece of data together. According to what I did gather by now, most people vote on the first and last day of the elections. 304 users (or 33%) casted their first vote (only the first timestamp is considered for my studies) on 26 November, whereas 118 (12.78%) did so on 9 December. This represents 422 out of 923 votes (or 45.72%), and it means that almost half of the votes were accomplished in two of the fourteen days available to vote (or 14.28% of the time available to vote). If we add 27 November, which is the third and only day left with more than 100 votes, we'd have 524 votes (or 56.7%). A lot of additional conclusions could be drawn from the graphic at the right. Cheers. — ΛΧΣ  21  Call me Hahc21 22:55, 19 December 2013 (UTC)
 * The 145% increase on 9 December coincides with the [//en.wikipedia.org/w/index.php?title=MediaWiki%3AWatchlist-details&diff=585296601&oldid=584593858 angry red watchlist notice] I added about the elections closing. (I also increased the cookie count, so anybody who had dismissed the notice and forgotten about the elections would be reminded again.) I think the elections may need to be advertised a little more prominently, though I accept the sole use of a watchlist notice is by consensus at the last ACE RFC. AGK  [•] 07:10, 21 December 2013 (UTC)
 * Has it been investigated what percentage of (eligible-to-vote) editors regularly access their watchlist? I know I do all the time, but I could imagine that quite large numbers might not, and therefore might not be aware of the election. W. P. Uzer (talk) 10:07, 21 December 2013 (UTC)
 * Maybe a site notice would be good? I know that some wikis use them to advertise important events and elections. — ΛΧΣ  21  Call me Hahc21 18:35, 21 December 2013 (UTC)
 * I will mention I know of at least three or four helpers on the #wikipedia-en-help IRC channel that were not pleased to find they had missed the election. It is what it is, but I agree that some better advertising may be the responsible thing to do if we want as many eligible voters as possible. Technical 13 (talk) 22:20, 21 December 2013 (UTC)

If there was a quota, should it match editors or readers?
The reaction to my proposal makes me wish that I had proposed that the quota, if there ever is one, would meet the reader demographics if not meet or exceed the editor demographics towards the reader demographics. EllenCT (talk) 00:36, 20 December 2013 (UTC)
 * There should not be a quota. Period. Nobody at all has demonstrated that Arbcom is better or worse depending on any of the demographics that have been batted around. In fact, one could argue that rejection of good candidates because they don't fit into some specific demographic would make Arbcom worse. Arbcom isn't intended to be representative of any group other than those that experienced editors consider capable of managing disputes in relation to Wikipedia.  It is not a governing body. It is not an exercise in democracy, and it's definitely not there for political correctness.  Risker (talk) 01:03, 20 December 2013 (UTC)
 * I agree with Risker's view on this issue, as expressed here and elsewhere on this page. Newyorkbrad (talk) 01:08, 20 December 2013 (UTC)
 * How is it possible to demonstrate that arbcom is better or worse depending on whether its demographics match readers without an arbcom with demographics that match readers? There is only one way to find out. EllenCT (talk) 01:11, 20 December 2013 (UTC)
 * I'll be completely blunt. Quota systems encourage the practice of placing inferior candidates ahead of those more qualified. When you're electing based on gender, skin colour, religion, ethnicity, etc. before capability, then you will not end up with a better committee. Resolute 01:19, 20 December 2013 (UTC)
 * On the contrary, "males are somewhat less likely to reach agreements than females." Testosterone has consequences. EllenCT (talk) 01:32, 20 December 2013 (UTC)
 * That has nothing to do with the work of the Arbitration Committee; appeals to authority, particularly authority that does not speak to the issue at hand, is a false logic. I think you need to understand the work of the committee before you start proposing to change its makeup. Risker (talk) 02:04, 20 December 2013 (UTC)
 * I am very interested in your reasons for saying so. EllenCT (talk) 02:19, 20 December 2013 (UTC)
 * I say so for the same reason that I would have no faith in the ability of someone to build a better internal combustion engine if they had not bothered to figure out what an internal combustion engine is supposed to do. Risker (talk) 02:38, 20 December 2013 (UTC)
 * I am more interested in the reasons for your first statement about whether reaching agreement has anything to do with the work of the arbitration committee and the implication that the authority I cited does not speak to that work. Are you saying that there would be no way to improve a lawnmower engine prior to the invention of the lawnmower? EllenCT (talk) 05:44, 20 December 2013 (UTC)
 * That's a pretty awful analogy, given that there were lawnmowers long before there were lawnmower engines. Have you read the arbitration policy yet? There's nothing wrong with creating a new idea; however, unless you're extremely familiar with the conventions of a community, it's generally a poor idea to appropriate the name of one process and try to apply it to something that does not appear to have any relation to the original concept. Maybe there's a reason to have a group that is statistically representative of the community as a whole, but since you've not defined what that group would do, and you're clearly not familiar with what the Arbitration Committee does, it would be better if you would give your new group some other label.  Risker (talk) 05:03, 21 December 2013 (UTC)
 * Yes. I would rather improve an existing article than start a new one. Is there any part of the committee's work which is easier to improve than reaching agreements over the internet? Should I propose a shadow arbcom, call it the ajucom, which acts in parallel to the arbcom but is at least 50% female and 20% people with statistics degrees? EllenCT (talk) 00:58, 22 December 2013 (UTC)
 * To start with, take it off this page. This is about the recent election. I'm not going to help you, because I think your idea is entirely wrong-headed (not to mention how you seem to be going out of your way to not inform yourself even when people give you links to read), but maybe someone else will help you develop a request for comment. But keep something firmly in mind: negotiation is not at all a role of the Arbitration Committee; decision-making is its role. Don't let the word "arbitration" fool you - like many other English language terms, Wikipedia has taken the word "arbitration" and turned it on its head.  Risker (talk) 19:10, 22 December 2013 (UTC)


 * Actually -- a slightly false position. The key to reasonable quota systems depends on looking at the universe of qualified people for the task -- and determining what the percentages are within that universe.  Then the number of positions must be sufficient to allow for multiple subgroups -- say a "jury universe" for Wikipedia editors of about 100 would allow for any subgroup chosen in accord with "quotas" to be attainable.  This means, moreover, that if (say) 20 out of the hundred individuals were to be used for a given case, that if one group were underrepresented in the group of 100, that they would then be proportionately more likely to serve in the group of 20.   Would it be an improvement? No one really knows -- but it is a possible route to take. Collect (talk) 01:28, 20 December 2013 (UTC)
 * Precisely. Perhaps there is an advantage to having at least twice as many statisticians on the committee as are present in the readership or editing community. Again, there is only one way to find out. EllenCT (talk) 01:32, 20 December 2013 (UTC)


 * We've been trying to get through to the community for years that case work is only about half of the work of the committee. On what basis do you think that jury pools are going to handle the rest of the stuff? Before arguing for a system like this, it would be far more useful to everyone in the community, including the arbitration committee, to figure out how to return some of those tasks (in particular, block/ban reviews of community or single-admin blocks/bans) to the community. Perhaps more importantly, I don't see any indication that there is a problem this proposal is trying to solve, nor any indication that the solution is solving an actual, demonstrable problem rather than a hypothetical one. Risker (talk) 01:44, 20 December 2013 (UTC)
 * I was addressing a possible concern expressed above by others. Is there only one single perfect way to choose arbitrators - or is the current system now hallowed? The clear solution is to discuss the possible choices and the reasons each one may be good or bad -- not to say "we hare already at perfection" <g>. I have specifically not stated that one specific course is ideal - I seek to facilitate reasoned discussion on the options for the future as best I can. à bientôt Collect (talk) 01:52, 20 December 2013 (UTC)
 * I doubt the current system is hallowed, but I think you need more than "we should do this because we can" before change is likely. But I will reiterate that I will never support a system that makes competence a secondary attribute. Resolute 03:31, 20 December 2013 (UTC)
 * Kindly note that I made no such comment -- my initial comments were with respect to statements over a long period of time from multiple editors concerning criteria for administrators, and I asked which of a whole slew of such criteria were felt to be most important. Then, and only then, concrete discussions as to how to meet the desired criteria would be useful.  Cheers. Collect (talk) 16:09, 21 December 2013 (UTC)
 * Okay, my Adjucom of One having sufficiently deliberated, we propose that the arbcom return block/ban reviews of community or single-admin blocks/bans to the community. If you want to second my proposal, then please set up an RFC on it. EllenCT (talk) 01:23, 22 December 2013 (UTC)
 * That's already true. The community can already remove community sanctions or admin discretion actions by consensus, and it's never not been that way. I've closed more than one such community discussion myself with the result of community actions being scaled back or removed. Seraphimblade Talk to me 02:02, 22 December 2013 (UTC)
 * Fantastic, I was afraid I would make a proposal which would cause more work. The Adjucom of One thanks you for your helpful reminder, admonishes you for use of a double negative, thanks you for your service to the community, and asks you to forgive that the committee was somehow unable to thank you for both of those things in conjunction. EllenCT (talk) 15:46, 22 December 2013 (UTC)

Request
Congratulations to the new Arbitrators. Could I request, please, that each of the winners contact me via e-mail with a snail mail address to which I can send an envelope bulging with untraceable US $100 bills, just in case I ever need to influence their decisions in the future? (Just a joke, obviously, what I really want to send is the business card of my uncle, Benny "Two Eyes" Ken, who's in the import/export business here in NYC, and can get you very good deals on stuff that somehow falls off the truck during shipping. He's a really sweet guy who hasn't been soured by his occasional time upstate.) Beyond My Ken (talk) 04:26, 20 December 2013 (UTC)
 * Don't trust him, I have seen this movie before.:-) Kumioko. 108.45.104.69 (talk) 04:32, 20 December 2013 (UTC)
 * Nonsense, I don't even own a duffel bag! And the one I have has baseball equipment in it. And it's not mine, it belongs to my wife. I mean... Beyond My Ken (talk) 06:09, 20 December 2013 (UTC)

Discarded ballots
I'm not trying to start a controversy here, but I was looking at last year's turnout to see how it to this year's (in relation to the above thread on turnout). Anyway, I noticed that 136 of the 1,039 total ballots cast were discarded (for lack of a better term) this year by the scrutineers, compared to 2012 when 34 of 858 total ballots cast were discarded. The 2010 and 2011 election pages don't list a total number of ballots cast, but judging from their respective log pages that only handful of votes were discarded in those years. I'm just curious as to why the number of discards increased so much.  Hot Stop   00:09, 22 December 2013 (UTC)
 * Just a minor correction: there were 116 discared votes, not 136 (1,039 - 116 = 923), of which almost 100 were ballots of people who voted more than once (I actually saw a user who voted like 10 times, of which, obviously, 9 were discarded.) — ΛΧΣ  21  Call me Hahc21 00:18, 22 December 2013 (UTC)
 * Thanks that makes sense. (And math is hard for me sometimes.)  Hot Stop   00:24, 22 December 2013 (UTC)
 * You're welcome :) — ΛΧΣ  21  Call me Hahc21 01:19, 22 December 2013 (UTC)
 * Why not discard all 10, as a punishment for cheating? W. P. Uzer (talk) 09:43, 22 December 2013 (UTC)
 * Mainly because in most cases they bore identical time stamps indicating a software glitch. Rules allow for "changing votes" by the way, so a second vote is not "cheating" - it is expected if you change your mind.    "Cheating" is where socks are used -- and I trust all socks were caught.  Cheers. Collect (talk) 10:15, 22 December 2013 (UTC)
 * I see, thanks. (And how does it work with "socks" - if you catch someone dishonestly voting twice or more, do you discard all the votes, or all but one? Just out of interest.) W. P. Uzer (talk) 11:48, 22 December 2013 (UTC)
 * I believe they used a "checkuser tool" to see if IP addresses of the registered users and access agents were duplicated, as a minimum, but I do not recall anyone saying of only duplicates were removed in such cases. Collect (talk) 12:26, 22 December 2013 (UTC)
 * In general, sockpuppet votes should be considered the same as general multiple votes from a single user: all but the most recent vote should be struck manually, in the same way the software strikes earlier votes from the same user automatically. If the sockpuppeteer were to be banned (not just blocked) before voting concluded, then there would be a case for striking the final vote as well; although I don't think that situation has ever arisen.  <b style="color:forestgreen">Happy</b>‑<b style="color:darkorange">melon</b> 12:54, 22 December 2013 (UTC)
 * I think it indeed happened a while ago when a candidate was banned mid-elections. I don't recall it all, but I believe that their vote was also striken. — ΛΧΣ  21  Call me Hahc21 19:15, 22 December 2013 (UTC)
 * It would presumably have to be done by the rules, and this year's rules say that an editor who is "blocked at the time of their vote" is ineligible. So I suppose that you can't strike someone's vote retrospectively following a ban, unless they had been blocked at the time they voted. In previous years the rules may have been different. W. P. Uzer (talk) 20:13, 22 December 2013 (UTC)