Wikipedia talk:WikiProject Administrator/Five Problems with a Single Solution
![]() | This project page does not require a rating on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||
|
This is great!
[edit]Trust me, it is; I am, after all, an infallible admin! LessHeard vanU (talk) 18:50, 31 October 2009 (UTC)
- I agree heartily! You really have done a wonderful job of articulating lucidly what a lot of us have been thinking. Thanks! --Tryptofish (talk) 19:23, 31 October 2009 (UTC)
Not a trial? No stigma? Good luck with that
[edit]Wikipedia is edited by human beings. If you think we can have an process for yanking someone's userrights based on a discussion* without it feeling like a drama-filled trial and having the those who lose their bit stigmatized, you're only kidding yourselves.
*On the other hand, if you yanked userrights based on a mechanical criteria, such as a period of inactivity, you might do that without a trial, stigma, or drama. davidwr/(talk)/(contribs)/(e-mail) 00:13, 23 November 2009 (UTC)
- Perhaps that needs clarification. Some people think that the only grounds for deciding who should and shouldn't be an admin is whether they are likely to misuse the tools. Another point of view is that all wikipedia actions by a user may be considered when making the decision.
- At RFA, I think we can say the "all actions may be relevant" view has come to be consensus. There is still discussion on the flip-side of that-- you'll occasionally bump into people who argue that incivility or edit-warring by an admin is not admissible in the consideration of their admin status since admin tools weren't used. It seems to me, no opinion polls conducted, just my impression-- but I think point of view seems to be on the way out also.
- Anyway, the point behind "not a trial" and "no stigma" is that the reforms proposed here are like RFA-- the community uses whatever standards it sees fit. Because of this, it's not a trial that produces a finding of wrongdoing, it's a !vote that produces a decision about role assignment. And as a consequence of that, it isn't an official stigma in the same way that a forced arbcom desysop would be-- something that is automatically implies guilt or wrongdoing.
- When I say "not a stigma", I mean that I hope it will be viewed more like "losing election for arbcom" or "having a failed RFA". Certainly, the original reasons people opposed someone being an admin/arb may remain, but the fact that you didn't win an election isn't, itself, a reason to oppose. Whatever actions you've done may be a stigma, but the actual result itself isn't a stigma. Again, in contrast to a forcible desysop, where your actions were so bad that other wikipedians and Arbcom spent weeks or months of its time deciding that you couldn't be trusted with the bit.
- Would people feel "on trial" or "stigmatized"? I hope not, but I admit they probably would-- humans are humans. My hope, however, is that as it gets easier to desysop, it will get easier to re-sysop too. I can off the top of my head think of several people who resigned their adminship who I would would fanatically support for a return to adminship if admin status was easier to revisit. Under our current RFA/arbcom-deadmin system, they are forever locked out of adminship because of the effort required to revisit the issue-- but with a working DRFA, that would change.
- So, basically, humans are humans, but policy is policy. "not a trial" and "not a stigma" refer more to the policy than the social perception. --Alecmconroy (talk) 12:32, 23 November 2009 (UTC)
- If this is a mirror of RFA, then it would take 70-80% no confidence to remove the bit. The odds of such a person gaining 70-80% in a future RfA would be bleak. As for being on trial or being stigmatized, losing a vote of no confidence is like a judge losing a retention election - it's your constituency saying "your fired," which is inherently stigmatizing. davidwr/(talk)/(contribs)/(e-mail) 13:05, 23 November 2009 (UTC)
- Which is why a constituency that feels strongly enough to actually say "you're fired" ought to be heard. No one has a "right" to be an administrator. --Tryptofish (talk) 14:40, 23 November 2009 (UTC)
- Yes, but we do have a right not to be told "this process isn't a trial" or "this process doesn't impose stigma" when such is obviously not the case. Let's have a little truth in advertising: If you are nominated for losing your bit, 10+ people think you are a crappy administrator or have personal reasons for wanting to cause you grief. If you lose the discussion, then 50+ people fall into that category and they outnumber people who think you deserve to keep the bit and/or have personal reasons to support you by better than 7-3. That's stigma and, for all practical purposes, a trial, so we shouldn't pretend it's not. davidwr/(talk)/(contribs)/(e-mail) 16:23, 23 November 2009 (UTC)
- By the way, for all but the most non-controversial candidates, RFA is a trial/crucible and, except for those who lose due to lack of experience or "almost, try again soon" reasons, not getting the bit can be stigmatizing. If DRFA is to mirror RFA, it will be similar. davidwr/(talk)/(contribs)/(e-mail) 16:25, 23 November 2009 (UTC)
- Yes, but we do have a right not to be told "this process isn't a trial" or "this process doesn't impose stigma" when such is obviously not the case. Let's have a little truth in advertising: If you are nominated for losing your bit, 10+ people think you are a crappy administrator or have personal reasons for wanting to cause you grief. If you lose the discussion, then 50+ people fall into that category and they outnumber people who think you deserve to keep the bit and/or have personal reasons to support you by better than 7-3. That's stigma and, for all practical purposes, a trial, so we shouldn't pretend it's not. davidwr/(talk)/(contribs)/(e-mail) 16:23, 23 November 2009 (UTC)
- Which is why a constituency that feels strongly enough to actually say "you're fired" ought to be heard. No one has a "right" to be an administrator. --Tryptofish (talk) 14:40, 23 November 2009 (UTC)
- If this is a mirror of RFA, then it would take 70-80% no confidence to remove the bit. The odds of such a person gaining 70-80% in a future RfA would be bleak. As for being on trial or being stigmatized, losing a vote of no confidence is like a judge losing a retention election - it's your constituency saying "your fired," which is inherently stigmatizing. davidwr/(talk)/(contribs)/(e-mail) 13:05, 23 November 2009 (UTC)
- You know, after I was pointed here by Tryptofish (and Gigs) from the latest de-adminship proposal, I was inclined to come over to this talk page and make almost exactly the same post as Davidwr. While this document is long on bromides and platitudes, it's rather short on specifics — and unfortunately 'The Fix' is self-contradictory to boot. Consider the system's purported virtues. (I've italicized the text, but otherwise reproduced the text as it appears on the project page.)
- High barrier to RfDA-- RfDA must be difficult to initiate.
- Resistant to mere retribution--- RfDA must recognize that even the most level-headed and consistently civil admin will still attract a certain amount of controversy.
- These first two points sound quite reasonable — but also seem to be features that are present quite thoroughly in the existing system (Arbitration). If the proposed process is supposed to have a high barrier (and quite legitimately so), then what can be improved about the present process? Arbitration already requires brief statements of complaint by the involved parties, presentation of at least a prima facie case through diffs and other suitable evidence, and a preliminary screening of the evidence by community-selected experts: the Arbs. Further, the process is (to some extent) monitored for adherence to proper process and for proper behaviour by the Arbitration clerks.
- Community-driven-- The community retains the ability to assign users to the various roles.
- It could be argued – and not facetiously or speciously – that the community has done exactly that in selecting (indeed, electing) certain of its members to act in our stead as Arbitrators.
- Not a trial-- While misuse of tools is one potential problem, there is also widespread agreement that a good admin also exhibits other traits like civility, impartiality, and cool-headedness.
- That's quibbling about semantics, really. Any process, whether this one or Arbitration, is evaluating the actions of an admin to determine whether or not they should retain their admin bit. Evidence will be presented, weighed, and discussed, and ultimately a final, binding decision will be rendered. While it appears that the intent of this passage is to affirm that all edits (and not just those involving the tools) will be considered, that doesn't make it any less of a trial.[1]
- No Stigma--- Unlike a forced desysopping, RfDA must not imply any wrongdoing, abuse, or any other "guilt".
- That's bollocks, really. The purpose of this process is to cause an individual to be desysopped, presumably without requiring their expressed desire or consent for that outcome. Call it whatever your like, but it's certainly a 'forced' desysopping. And a 'successful' RfDA absolutely does imply a failing of some sort on the part of the individual concerned. They're unsuitable for adminship, and their bit is being taken away. Whether it's for misuse of the tools or because they're just too much of a jerk, they're being fired for some sort of (at best) incompetence[1] and a stigma does attach to that. Further, even a 'failed' RfDA would bear a certain stigma. If the first two points (a high barrier, and an apparent resistance to vexatious or vengeful litigants) are met, then the successful start of an RfDA implies that any complaints enumerated within have at least some credibility on their face. the 'failed' RfDA will be presented as evidence that an admin is on notice, on a short leash, or is just good at recruiting friends to the discussion.
- Appeals to Arbcom-- As a final safety net, the Arbitration Committee must be able to over-rule decisions made by RfDA
- That's a reasonable safety net, though I do worry that there may be a bit of 'then-I'll-just-ask-the-other-parent'. The ArbCom will be very reluctant to overturn even a poorly-though-out successful RfDA for fear of provoking endless wikidrama. And if an RfDA fails, should there be yet another bite at the apple for the complainant? See also my further comments below, about how this process would unnaturally separate desysopping from all other aspects of dispute resolution.
- [1]Reading between the lines a bit, it seems that this proposal is trying to hint at some changes to the standards (or perceived standards) for desysopping, with the implicit assumption that creating a whole new desysopping process is the only way to implement those standards. I note that in the past, the ArbCom has held that administrators (like other editors) are expected to be civil in their dealings with others, and to meet reasonable requests for explanation of their actions. Further, while admins are expected to be impartial in the use of their tools, we cannot compel them to lack opinions or biases in all aspects of their on-wiki conduct. I would strongly advise that instead of constructing an entirely new regime for desysopping, it might be wiser simply to ask the community to endorse suitable changes to the adminship policy with regards to expected standards of conduct. If those changes were based on a suitably thorough and credible community discussion, the ArbCom should reasonably be expected to follow them.
- I'm concerned and confused that this process attempts to simply extract a portion of the Arbitration process to deal solely with desysopping. While I'm not prepared to troll the Arbitration case logs, I would be shocked to discover that more than a very tiny fraction of cases (if any) involving desysopping also involved no other case findings, cautions, or remedies. Going ahead with a process that forcibly separates the question of desysopping from all of the other surrounding (mis)conduct by other actors seems to be a dangerous notion. By the time a case reaches Arbitration, it's almost always true that there is more than one party who is behaving badly. (This is also at odds with the 'resistant to retribution' principle. I have seen disputants at RfArb who have tried – unsuccessfully – to insist that the scope of the cases they raise should be narrowed to not consider their own behaviour; they just want to see the other guy get punished. While the ArbCom is generally quite sensible about smacking down such unreasonable and self-serving limitations of scope, it seems that this process would explicitly enable and enforce such gaming.)
- Finally, on the current proposal — this process would seem to put a very blunt tool into the hands of an angry mob. The entire process, as proposed, can be run through in less than ten days (not more than three days for the nomination, plus seven days of voting). God help you if someone makes a spurious claim which you can't refute for a couple of days because you're working, or visiting family, or just have better things to do than guard your RfDA for the full week. This process will have the same problem as RfA — that is, everyone will feel that everything has to be said, and analyzed, and responded to, and re-responded to, and decided, in a narrow time window. No opportunity for sober reflection, no easy way to take a step back from the process and pause if things are going off the rails. The most emotionally-invested individuals will dominate the discussion; there will be bickering threads a dozen colons deep. At the end of it all, a coarse, binary output: yes, you keep your bit; or no, don't let the door hit you on the way out. (It's possible for a small ArbCom with a self-regulated pace to weigh multiple outcomes: permanent desysopping, temporary desysop, probation, topic area restrictions, all of various duration. It's not credible to suggest that a mob of fifty or a hundred drop-in editors working under a one-week deadline will be able to do the same.) TenOfAllTrades(talk) 17:25, 5 December 2009 (UTC)
Discussion! Hurrah!
[edit]A pamphlet, not a bill-- bear in mind, this page is an attempt to show that what look like separate problems are, in fact, entangled and related. Over at RFA, they have been debating improving that procedure-- to make it less stressful, learn what standards are good ones, etc. Over at arbcom, they've been working on caseloads and wait-issues. Lots of places have talked about concerns of drama, factionalism, incivility, etc.
The take home point of this document is:
- All these seemingly separate problems are fundamentally connected.
- Altering our system to make de-admin easier would have a lot of good effects on all these separate problems.
So, above it's noted that this page "is long on bromides and platitudes, it's rather short on specifics"-- which I think is actually quite charitable. It's almost completely devoid of specifics because it's not a specific proposal at all. There are a lot of different specific proposal already written-- this was just sort of a general framework to think about the problems, along with a vague sketch of what I think the solution will look like.
Arbcom as RFDA - Tenofalltrades suggests Arbcom, under altered policy constraints, could perform the needed role. This is, in fact, one theoretically possible way to implement RFDA. Expanded seats, sub-panels, lower courts. Of the people who accept the existence of problems, most people who have helped out so far have generally tended to think a community-based process as the solution, but that doesn't mean we should dogmatically accept that. Indeed, it's possible to say this year's arbcom has taken initial baby-steps in the direction.
I personally don't know how make an arbcom-level solution come to pass. Essentially, I'm saying "In general, I think it would be good if you were substantially more likely to desysop admins who are causing trouble". They already have this power, they already do this to an extent-- I think the limiting factor is just the amount of time needed for a small, overworked group of people to review the behavior of a much large population. But if anyone agreed that there are problems that need fixing can find a way to fix things through arbcom, that would definitely be excellent, and I'd whole-heartedly support that effort.
Appeals to Arbcom-- definitely don't want this to become an automatic "asking the other parent". Basically, when I said Arbcom has the final say, I was think of it as a way to make sure that people aren't desysopped if arbcom thinks the process was bad. If arbcom blindly accepted every appeal, there'd be an automatic "asking the other parent" in place, but I don't think they'll do that.
Another way to view-- RFC that can be directed acted upon.-- another way to implement an RFDA would be to make user-conduct RFCs not automatically appealed to arbcom. As it is, user-conduct RFC for admins is basically nonexistent, since any RFC consensus about role assignment or user trust can't, currently, be acted upon. One way to implement an RFDA would be to use the usual conduct-RFC process, with modification that consensus at that level can be certified and acted upon, rather than automatically triggering an RFAr.
Community-driven-- Ten says: "It could be argued [...] that the community has done exactly that in selecting [...] Arbitrators." Absolutely Arbcom is an community-driven process for resolving disputes. Some people might object, on principle, to Arbcom serving that role, but I certainly don't.
The problem is that for normal disputes Arbcom is the last resort of dispute resolution. But for disputes involving admin behavior, Arbcom isn't just the last resort, it's the only resort. For Wikipedia, Arbitrator time is our most precious and limited resource we have. Imagine if we had every single RFA go through arbcom instead of letting the community directly select admins.
So, I guess the big question to the view that RFDA should be handled by Arbcom alone but still think the current system has problems: 1) how can we help it do so more effectively and efficiently, and 2) how and whether Arbcom consider community confidence during those re-evaluations. --Alecmconroy (talk) 00:48, 6 December 2009 (UTC)
more later. --Alecmconroy (talk) 22:11, 5 December 2009 (UTC)
- To clarify, my commments above are largely a response to a couple of editor at the latest RfDA proposal (Wikipedia talk:Community de-adminship/Draft RfC#A pause to look at the numbers) have suggested that this document will answer my concerns about the proposed process there. In brief, it does not, and it has some flaws (noted above) on its own, as well. Some of my criticisms are more specific to the current RfDA proposal, others are more general; I'm sorry for any confusion that might have occurred.
![](http://upload.wikimedia.org/wikipedia/en/thumb/3/3b/WikiprojectAdmin_TheFix_Updated.png/300px-WikiprojectAdmin_TheFix_Updated.png)
.
- Responding to Alecmconroy's first two bulleted points — I'm willing to give you the first one. At some level, virtually everything on the project is connected, and it's certainly possible to draw plausble-sounding links between these seemingly disparate problems. The second point is based on assumptions which I'm unprepared to grant. Will the threat of a rapid desysopping process actually result in improved civility and reduced drama? Nope. What a rapid desysopping process means is that any controversial action will be met with threats of desysopping. (Don't like a decision? Just whip up a lynch mob.) Civility will decline, not rise — and it's not as though editors are currently particularly restrained in insulting, abusing, and threatening admins. Whether the process successfully de-admins someone or not, most cases would likely be followed on by Arbitration in any case — either to review the decision itself, or to go on and sanction the other bad actors who got a free pass by narrowly scoping the RfDA process.
- There is an implicit assumption that there exists a large pool of unsuitable admins who must be dealt with rapidly in order to save the project. There is a further assumption that existing processes cannot handle these cases. I'm just not seeing the evidence for either point. The figures next to the text are beautiful conceptual diagrams, but I feel you've left out the role of ArbCom. What actually happens is what I've drawn in at right. Parallel RfDA processes would be adding a second upward-pointing funnel to duplicate the RFArb process. TenOfAllTrades(talk) 17:52, 6 December 2009 (UTC)
- I actually made a diagram very similar to the one you created-- less funnel more of a pipe, pretty clogged, to illustrate the limited number of arbcom cases heard. Probably should have put it in. Because RFAr does do this job now to a limited extent. . --Alecmconroy (talk) 19:33, 6 December 2009 (UTC)
CSD hangon
[edit]I would like to contest the CSD nomination. I can certainly understand how the creator of the page would have requested CSD in good faith, but I believe that there will be a future use of this essay by other Wikipedians, potentially including me. Administrator policy and administrator recall remain issues where the community has discussions. This essay contains some useful insights from the time when it was written. It would be fine to mark it as "historical". I would also be quite willing to have it moved into my user space, if that would help. Thanks. --Tryptofish (talk) 18:28, 3 August 2011 (UTC)
- Just noting that a third editor reverted the CSD tag, which is fine with me. My other comments still stand. --Tryptofish (talk) 18:39, 3 August 2011 (UTC)
- Yeah, that was me. I've also requested undeletion of the images that were used on it. The page is very useful in highlighting the serious issues plaguing WP:RFA and adminship in general. I've left a note to that effect on the prime editor's talk page. --Hammersoft (talk) 19:38, 3 August 2011 (UTC)
- Yes, I know it was you. Good, I agree. I also left a request at the talk page of the admin who deleted the images, but your request will probably get done first. Thanks. --Tryptofish (talk) 19:54, 3 August 2011 (UTC)
Steward comparison
[edit]The stewards handle this with almost no drama, despite the hundreds of participants: every two years, each steward goes up for reappointment. This is less intensive than standing in the first place. Noone asks a lot of questions unless there is a problem; the equivalent of a bureaucrat (a grou of stewards serving in that role) closes all of those discussions. – SJ + 21:38, 21 June 2013 (UTC)