Wikiversity talk:Candidates for Custodianship/Diego Grez

Crat chat
I'm unclear on the policy here, but since Diego Grez has failed to reach the "75% supermajority" that is generally used to determined consensus, I'm not sure what to do. Am I supposed to request removal of his tools immediately, or give him 48 hours to find another mentor? --SB_Johnny talk 21:27, 17 August 2010 (UTC)


 * Can you explain how you came up with 66% support? I saw either 70% or 71%, counting his mentor's support. I would say that if 70% or 71% isn't enough of a consensus that the tools should be removed immediately as he had his full probationary period and there wasn't enough support. -- dark lama  21:45, 17 August 2010 (UTC)
 * Ah true, Jt's vote adds one (which makes it 70% exactly). If this were wikipedia, 70% would mean "'crat's discretion", but even in that case I would still call it the same way, as the oppose votes seem to to be in good faith and made out of genuine concern.
 * I don't see any hurry about running off to meta, so let's talk for a day. --SB_Johnny talk 21:50, 17 August 2010 (UTC)
 * The only thing I can find is "If you are not approved, you are free to request another mentorship or withdraw your request." custodianship. The mentorship had already been extended once, though, so I'm not sure that there would be much gained in this case. --mikeu talk 22:19, 17 August 2010 (UTC)


 * Comment: Abigor's comment was obviously made in bad-faith, and was canvassed (as can be seen on here). I strongly ask you to not consider it. Diego Grez 22:03, 17 August 2010 (UTC)
 * Maybe, but we've never discouraged canvassing in the past. --SB_Johnny talk 22:13, 17 August 2010 (UTC)
 * I agree. However Mattwj2002 hadn't contributed in 3 years before supporting you as well, so if comments are to be discounted, Mattwj2002's support should be as well. That would make 75% support by my count. -- dark lama  22:09, 17 August 2010 (UTC)
 * Yup, it's messy. I'm just going with the cards on the table. --SB_Johnny talk 22:13, 17 August 2010 (UTC)

(edit conflict with above)
 * First of all, 75% is not a rigid boundary. You can consider the nature of the !votes, and deprecate some.
 * Consensus (guideline) Consensus is not established just by counting votes. Wikiversity is guided by ideas that are in harmony with the education-oriented mission of the project. In judging consensus, it is the responsibility of all community members to give the most weight to rational arguments that support positions and points of view that are in harmony with the Wikiversity mission. Community members should assign less weight to discussion comments or votes that provide no rational argument to justify a point of view or that disrupt the project.
 * Custodianship (policy) Five days after the request has been listed, a bureaucrat will make the final decision based on the arguments provided in the discussion. (my emphasis)
 * Bureaucratship Bureaucrats act as the final interpreter of consensus with respect to candidacies for user group changes. Bureaucrats are charged with the responsibility of declaring at an appropriate time, whether a custodian, bureaucrat, or bot candidate is granted a user group change or otherwise. Bureaucrats should respect the Wikiversity community's decision on these particular matters.
 * The 75% supermajority is a Wikipedia standard and is, in my view, part of the problem there. It should be both easier to become an administrator there, and easier to have the tools suspended or removed.
 * So, how to interpret the policies? First of all, what I say is simply advice. As a bureaucrat, making a decision, you are responsible to the community, in the end, within the discretion that the community has allowed, and that, in fact, the community needs. The following, then, is how I'd think about this were I in your shoes.
 * The policies clearly refer to arguments, not to numbers.
 * A 'crat may clearly not act against consensus, but this must mean that there is a consensus against the action. If a majority have supported the action, after considering possible problems from participation bias, canvassing, etc, then, clearly, the 'crat is not barred from the action. Below majority support, the action of the 'crat becomes more and more questionable.
 * What is a "community's decision"? If there is "no consensus," there is no decision, it could be argued. Yet if two-thirds of a community has taken a position, it's hard to argue with a straight face that this isn't some kind of decision. It's merely a weaker one than one with a stronger supermajority.
 * Hence, a 'crat is free and, indeed, responsible, to review the poll on a candidate and make a decision based on arguments, as the policy requires. The 'crat may and should, for example, consider if some !votes may be biased or otherwise not representing the community as a whole. In addition, the decision should not normally be against the decision of a majority.
 * This is the problem: if you reject the candidacy, you are going against the decision of a majority, that's quite problematic, in fact. (This has generally been overlooked, but the particular trap here is one that has afflicted every consensus-based organization I've participated in, a supermajority requirement generally leads to a kind of minority rule, where the status quo favors a minority.)
 * I see two reasonable courses here that don't reject the decision of a majority.
 * Promote, based on majority and arguments. Ultimate consideration: welfare of the community.
 * Allow restart of mentorship, with a change of mentor or the same mentor. This could be with or without a new custodianship candidacy. It is clearly allowed with a new candidacy.
 * Policy does not provide for extended probationary custodianship, beyond a permanency decision. But the existing discussion could be considered "no decision." If the candidate is willing and a mentor is willing, the existing !vote could be closed as "no consensus, continue probationary custodianship."
 * Which would I choose? There are only three oppose votes. The first was from a non-participant at Wikiversity who was canvassed, I've seen the IRC log. I asked Abigor to disclose whether he'd been canvassed or not, he declined, but I have no reason at all to suspect that the IRC log I've seen was fraudulent, and the timing showed that the !vote was canvassed. I will show this to a 'crat on request. But it shouldn't be necessary, because the arguments presented by Abigor were, by tradition here, moot. Behavior elsewhere is almost completely moot as to participation and privileges here, that's obvious. The second was from Adambro, who wheel-warred with Diego, on matters where the community generally supports Diego's position. Further, Adambro based his argument, again, largely, on off-wiki behavior as well as, again, issues where Adambro is taking a largely rejected approach, as to community consensus when the matters come to consideration. That is, of course, controversial in itself, but the point is that Adambro is hardly neutral, making a decision purely on the basis of the welfare of the wiki, except as to thinking that his own opinion is identical to that welfare. That leaves JWS, who asked a farrago of questions in many locations, and Diego did attempt to answer some of them. Nobody has answered JWS's questions, ever, in recent times, as far as I've seen, sufficiently to satisfy him. Therefore his arguments must, again, be deprecated.
 * The votes in favor of promotion were six. To shift the result to 75%, it is only necessary to reduce the negative votes in weight by 1/3. That is not much! I would promote, avoiding further disruptive process. Diego's "crime" was that he became active, so that what it would be like to have him as a custodian became visible. As is well-known, any active custodian is likely to make enemies. If we insist on a pure 75% standard, never considering the arguments, as policy actually requires, we will only promote custodians who conceal their true positions, who give us a month of gnoming so that nobody will oppose. And that will be our loss.
 * And thanks for asking. My congratulations. --Abd 22:28, 17 August 2010 (UTC)
 * I did not count the !vote of the mentor. That, then, makes the full vote ratio be 70%, which already takes us within discretionary range with no deprecation at all. However, please remember, the policy requires, not a vote level, but a standard based on the arguments. In the interest of neutrality, perhaps some of the positive votes might be depreciable as well. Good luck. Were I a custodian, I'd be happy to mentor Diego, he was always responsive, and being responsive is even more important than getting everything right the first time. He was responsive to me even though I was just another probationary custodian.... --Abd 22:34, 17 August 2010 (UTC)
 * Just responding to what I saw whilst skimming over the above thesis: (1) if this were a poll in a Quaker business meeting and I were the clerk, this would be no consensus to promote, hands down and an easy call. (2) This isn't a big deal: he can seek another mentor and perhaps convince one or two people given another month to prove himself. (3) The WP standard is the only one I have to go by, unless you think a real consensus is possible in this environment (hint: it isn't). --SB_Johnny talk 23:22, 17 August 2010 (UTC)
 * Whilst I'm disappointed by Abd's apparent eagerness to dismiss my comments, I agree with the idea that promotion wouldn't be unreasonable based upon the votes. Abd encourages us to consider arguments rather than simply counting votes but two of the supporter only voted, presenting no argument in favour which could be assessed and discussed and of those two, Mattwj2002 isn't really active here. Simply saying "support" isn't really a convincing argument so if we based the decision "not a vote level, but a standard based on the arguments" they would carry little weight. It is reasonable for Abd to raise concerns about some of those who've opposed the promotion but he shouldn't overlook issues about some who supported just because it would be convenient to do so. I don't think it would be appropriate to simply extend the probationary period, the choice here is between granting permanent custodianship or not. Adambro 22:47, 17 August 2010 (UTC)
 * Thanks, Adambro, sorry you are disappointed. You have overlooked that I already noted that I'd only presented arguments for deprecating negative votes, and that some arguments might apply to deprecating positive ones. But once we are looking at "arguments," the vote counts are not important. The difference between "extending the probationary period" and starting up a new candidacy is purely a technical difference. The latter is enshrined in policy as a right, so far. There is precedent for extending a period to decide an issue, in recent decisions. I'm gratified to see that Adambro considers promotion not "unreasonable." That is the kind of commentary we need, that steps outside of personal interest, and I've seen Adambro do this many times, and he should be commended. As to the role of vote counts, there should be some sense of support from the community, and the positive votes do show that; negative votes should also be considered in that way. Considering vote counts to determine consensus is not simple in marginal cases. What I see, though, is a system that depends on 'crat discretion, and I'm confident that whatever SBJ decides, it will be within his allowed discretion. --Abd 23:28, 17 August 2010 (UTC)


 * What is your "common law" interpretation of Wikiversity talk:Consensus where 3 people for most of the discussion there disagreed on whether a policy was needed, what consensus should or shouldn't be, and than suddenly the page is made a guideline without further discussion or resolution of the disagreement, but two people supported the change to a guideline? -- dark lama  22:51, 17 August 2010 (UTC)
 * I looked at the history of the guideline page and it was stable, unedited since almost a year ago. That creates a presumption of consensus on it. The issue of "consensus" is thorny, in fact, and I have extensive experience with consensus organizations. Briefly, my conclusion: majority rule with defined participation rights, and appropriate deliberative structure is, long-term, far cleaner and less disruptive. Consensus, however, meaning, now, approval greater than a majority, is highly desirable, so, to me, the conclusion is that a process which seeks maximized consensus -- always -- but which allows ad-hoc or temporary decisions to be made by majority rule is the best idea.
 * Wiki common law? There is a technical legal term for the state of wiki common law on this issue: Okay, I won't say it, but it's a word in SNAFU. There are a few wikis, now, that may be starting to grasp the problem. See, when the community is small, consensus works as a standard, but it starts to break down somewhere around thirty active participants or even less. Long-term, I've seen, even thirty will burn out on it, if they stay active. Wikis try to make do without addressing the problem of participation bias; long-term, participation becomes warped in favor of those who will sit through long, boring discussions. Or who will simply ignore the discussions or make quick, sound-bite comments with no depth, and no attempt to actually seek consensus. There are classic solutions and wikis generally reject them because .... they are classic solutions, and we don't need no stinkin' classic solutions. Our Way is better, and you Outsiders should first learn Our way before you make suggestions, and ... I'm watchin' ya now, one false move and you are outa here, troll!
 * Ahem! The classic solutions were developed over centuries, and they are the classic solutions because they work. However, what I really propose is synthetic, incorporating the classic solutions but adapting them to the ad hoc nature of wikis, with bottom-up hierarchical structure that, when it's proposed, you might like. Those who understand it often get quite excited! Meanwhile, we gotta muddle through. There is usually a way. Thanks for asking. --Abd 23:19, 17 August 2010 (UTC)


 * I would suggest allowing Diego about 48 to 72 hours to find a new mentor, and then if that fails, have the tools removed on meta. This would minimize the requirement for steward intervention, and prevent the redundancy of having to turn them back on if he finds a mentor. Geoff Plourde 00:53, 18 August 2010 (UTC)

Thanks...
Thanks to everyone who voted. Sincerely I expected that the vote was going to be better, but it was not. I requested the removal of my rights at meta, and it feels more free! :) Will try to complete my work at Mapudungun, and hoping positive feedback through all this process, I reiterate my acknowledgements to all of you. Diego Grez 03:57, 19 August 2010 (UTC)
 * Just give it some time, make standard patrolling edits, keep your nose clean, and then reapply for a mentor later. That should be no problem. You did have seven supports, which is quite a lot, so don't feel bad. Ottava Rima (talk) 04:06, 19 August 2010 (UTC)
 * Thanks, Diego. That's really the best way to go about it, IMO. Tomorrow is another day :-). --SB_Johnny talk 13:05, 19 August 2010 (UTC)
 * Shall we have a photo op of everyone hugging and smiling, or is that too soon?
 * Okay, one mystery here. We have more than two-thirds of the community voting, even if we include an apparent non-participant (It's 88% without that), yet no promotion, and more to the point, no volunteer custodian to continue the mentorship. It may merely be a formatting and structure problem. The actual vote shows community support, the policy requires consideration of arguments, not votes, which are a means of assessing level of consensus, and only a rough estimate because of participation bias. Diego certainly had adequate "minimum community support" in terms of numbers supporting, and the opposition from within the community was only the kind of opposition that would normally be expected from someone who has been active as a custodian, instead of as a pretend custodian, i.e., someone who only does wikignoming. I conclude that something is off, and I intend to suggest policy changes. Two-thirds is adequate to change policy, but we'll see how the community feels about it. The actual problem, right now, might or might not be serious, but this is the bottom line: a community could become locked in a state where a number of existing active custodians are at variance with consensus as to choices they make in use of tools. If a new candidate comes up for !vote, and is seen as a threat to the power of this group, one custodian blocking plus a canvassed vote from outside, and some custodians have extensive contacts outside and could make sure that something like this comes in, can prevent promotion. This is a general problem with consensus process which should be discussed elsewhere, but here we can see a possible example where strict interpretation of vote levels can prevent "substantial consensus" from acting to improve balance.
 * At the same time as some are arguing that we can't support recusal policy because we have too few custodians, the same people are rejecting -- or not supporting -- a new custodian who seems quite adequately qualified. SBJ, you have mentioned Quaker consensus. I can't write with authority on specific Quaker practices, but I would not call the situation here "consensus" within, say, the rules of Alcoholics Anonymous, as to considering a matter to be completely settled. However, even AA will consider a two-thirds vote at the AA World Service Conference to be "consensus" for the purpose of making decisions. The actual controlling body over property, the Board of AA World Services, Inc., is advised by Conference consensus, but itself makes decisions as is normal for boards (and as is probably legally required), by majority vote. With that low a level, two-thirds, the Conference will allow reconsideration, and they have been known to unanimously reject a motion that had passed with only a handful of oppositions, after reconsideration was requested and new arguments made. Here, what is being done is to consider a "motion" dead because it failed to reach a higher supermajority than two-thirds, when, in fact, the policy requires, not a supermajority, but preponderance of arguments (and vote counts are an implicit argument, only one factor, though an important one).
 * If Diego has such high support, why has no replacement mentor volunteered? It may simply be a process problem having to do with visibility. Diego, if you feel ready, please start up a new candidacy, requesting a mentor. Jtneill may or may not be willing to continue, and if you like, you could ask him privately.
 * Most worrisome would be that in the secret and semi-secret blizzard of off-wiki communications that I've been finding play a large role here, some deal may have been cut to keep Diego out, even with the presence of such a high level of support. I'm advising Diego not to get involved in this, except to simply ask for a mentor again. If, with such a record and such support as was shown, he can't find a mentor, something is very, very off. Addressing it may take time. --Abd 13:41, 19 August 2010 (UTC)