Misplaced Pages

This is an old revision of this page, as edited by Lowercase sigmabot III (talk | contribs) at 00:31, 24 December 2014 (Archiving 1 discussion(s) from Misplaced Pages talk:Requests for adminship) (bot). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

This is an archive of past discussions on Misplaced Pages:Requests for adminship. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Desysopping proposal

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Well, this is yet another proposal for a desysopping system. I've recently been studying RFA2011 and RFA2013, and it seems quite obvious that reform will never work unless we have a way to remove admins (aside from ArbCom, which is very well known for being slow). Here's the proposal:

1. Before a case request is filed, the issues must have been extensively (struck "extensively" because it is a rather unclear word; "discussed" should be sufficient) discussed at other venues, such as the admin's talk page and ANI. The issues also cannot be minor. For a case to proceed, the admin must have displayed repeated poor judgement, or have committed a particularly serious violation (in these cases, however, ArbCom is likely to deal with it).
2. If all the requirements are met, the concerned user may file a case request.
3. After the case request is filed, two (possibly three; please specify your preference when you comment) completely uninvolved experienced (struck "experienced", as we'll never agree on what "an experienced user" is; any uninvolved user in good standing could probably be trusted to do this) users (possibly admins?) who are in good standing (we'll never agree on what "good standing" is, either) will research the matter, and they will certify the case if they feel that the issues are serious enough to warrant a desysopping case. If the case is not appropriately certified in one week, the case will be closed as stale. If the case is properly certified, the case will proceed.
4. The discussion will run for two weeks. If a supermajority (67%) (changing to a simple majority (51%) per suggestions in discussion; if you prefer the original number, or something different, please mention that in your comment) support desysopping, the admin will be desysopped by a bureaucrat or a steward, although the former admin may file an RfA at any time.

I expect that this proposal will run into the "excessive bureaucracy" problem, seeing that it involves certifications. However, I'm including the certification because I do not want to see admins desysopped over petty issues, such as an isolated bad deletion or block. Also, in RFCU, the cases were certified by very involved editors, which is not fair, in my opinion. This proposal specifies that the case can only be certified by two experienced, completely uninvolved users. Remember that this is only a very rough draft and is open to major modifications, so if you can figure out a way to make this less bureaucratic while still preserving the "mob protection", please say so. Thanks, --Biblioworm 21:29, 2 December 2014 (UTC)

Reforms will never work unless there is a reason for the reform. I understand that providing examples of bad admins who should be desysopped is a bit sensitive, but that's what has to happen. If there is a problem, please link to a discussion with an outline of the claims concerning an admin so others can evaluate whether a reform is needed, or whether there should be swifter retribution for people who try to grind down those who defend the encyclopedia. Johnuniq (talk) 21:49, 2 December 2014 (UTC)

The idea behind the introduction of a desysopping system is that the community will not be as afraid to make new admins, because they can take some comfort in the fact that the admin can be desysopped if s/he turns out to be worse than they thought. The lack of such a system is probably one of the main reasons why only near-perfect candidates pass RfA. Besides, a community-based desysopping system works for some other Wikipedias, so why can't it work for us? I sometimes think that we just make things hard for ourselves. By the way, click here if you want to see a current example of a questionable admin. The issue probably could have been resolved at a much faster pace if there had been a community desysopping system available. --Biblioworm 22:12, 2 December 2014 (UTC)

That's a well-known case which is the kind of thing that the community would have trouble deciding, and for exactly the same reasons that Arbcom has not yet taken any action. However, Arbcom has the issue well in hand and when the admin returns everything will proceed as quickly as is reasonable. Another case is here and it was quickly resolved by the admin effectively retiring. That's fine—we don't need heads on spikes. Johnuniq (talk) 22:37, 2 December 2014 (UTC)

Here's the question—the only problem explicitly identified by Bioblioworm with desysopping through ArbCom is that it is "slow". Why is it necessarily to develop an entirely new process in order to jam through desysoppings rapidly? (And even the "slow" assertion is questionable. ArbCom cases that involve multiple issues, complicated situations, and – particularly – misconduct by multiple parties certainly take rather a long time to proceed to completion. But those complex situations are ones where we shouldn't be extracting one party to rush one sanction.)

In truth, we already have Step 1 and Step 2 of the above scheme, where Step 2 is filed with ArbCom. In situations where there is clearly-described misconduct by an administrator, the ArbCom has repeatedly demonstrated an ability to act swiftly. The ArbCom will desysop by motion. The ArbCom will desysop on the basis of clearly-established, evidence-driven consensus for that action at WP:AN. The ArbCom will desysop in absentia if an admin doesn't respond to proceedings. Simply filing a clearly-stated arbitration case against an administrator is often sufficient to trigger an under-a-cloud resignation (which has the same effect as desysopping).

A major problem is that few people can be bothered to sit down and do Step 1 properly. Gathering evidence and diffs takes time and effort. Editing evidence into a coherent narrative takes time and effort. That effort won't magically disappear with this new process (or ones like it). At least, I hope that it won't—having mob-driven desysoppings for no clearly-elucidated reason seems unlikely to increase the pool of willing adminship candidates.

Finally, making a good-faith attempt to consider alternative, less-drastic remedies takes time and effort. This last step is important but often neglected. Not every error or instance of misconduct actually requires or deserves desysopping. All too often a vocal minority demands desysopping. When they fail to get it, they blame the process. Sometimes the problem lies instead in their own expectations being out of sync with the community. TenOfAllTrades(talk) 22:42, 2 December 2014 (UTC)

• I find myself in sympathy with this proposal, although I would rather have a noticeboard where editors could bring concerns about administrators' conduct. One reason is that that would allow for issues to be raised with one part of the admins' actions without it being necessary to imply they were a net negative; another is that it would by being more ad hoc, more easily allow for resolutions short of threatening the admin with desysopping, and encourage broader participation by not being a formal vote. But I'm inclined to prefer non-bureaucratic and informal processes anyway, and that relates to the points about Arbcom. I share Biblioworm's concern about Arbcom being slow, and I don't share the confidence in Arbcom indicated in some responses above. I find Arbcom's procedures impenetrable and see it making decisions that in whole or in part I do not find helpful. I don't have much confidence in it at all at this point. And I believe Biblioworm is spot on: one of the problems with RfA is that editors fear sysopping is almost irrevocable. I recognize that his/her proposal is designed to minimize the risk of the mob attacking a diligent admin for being diligent, but I'd still rather see a more flexible process. Yngvadottir (talk) 23:27, 2 December 2014 (UTC)

Support, all of the folks comments here were TL, so I DR skimmed over most of it, but I just wanted to say that this could work. Sometimes, certified users are needed for things, like desysopping. (Whether community driven or ArbCom-like, we just need a better desysopping process.) --AmaryllisGardener 23:33, 2 December 2014 (UTC)

And a big thank you to AmaryllisGardener for illustrating so succinctly what would be wrong with this proposed process. TenOfAllTrades(talk) 23:57, 2 December 2014 (UTC)

And a big thank you to TenOfAllTrades for illustrating so succinctly what is wrong with RfA's community. --AmaryllisGardener 00:00, 3 December 2014 (UTC)

That is not a reasonable response. You are supporting a scheme where you dismiss a few paragraphs as TLDR—that perfectly illustrates why a community-based system of attacking admins is not needed. Someone would post a wall of diffs with assertions, and passers-by would support action against the admin without considering whether the diffs actually support the assertions, and without considering the rebuttals that have been made. Johnuniq (talk) 01:11, 3 December 2014 (UTC)

Let me say that "skimmed" would be a more appropriate word, because I read all of it, I just read through it quickly. I don't know why I used TLDR. Anyway, thanks for explaining. --AmaryllisGardener 01:18, 3 December 2014 (UTC)

• Support. I was under the impression that this is pretty much how things already worked. We decide issues by community consensus; there is no reason that a policy-backed community consensus to de-admin would not be honored. bd2412 T 01:22, 3 December 2014 (UTC)
• Support, although I think a 50%+1 majority should be sufficient to desysop. Everyking (talk) 02:11, 3 December 2014 (UTC)
• Support, though I think that the specific percentage should be loosened a little, similar to how it is for an RfA. Such a discussion in effect would be a required reconfirmation RfA. Something like approximately 50% to 70% should require 'crats to weigh arguments and/or depending on the circumstances have a bureaucrat discussion to close. With percentages outside those ranges pretty much a rubberstamp keep or desysop case. Admins will make enemies and we want to avoid loosing admins that are willing to make difficult closes that might make enemies. At the same time we should be able to deal with admins that have taken too much of a partisan view and abusing there admin powers to strength one side even if they have a bunch of like minded supporters. PaleAqua (talk) 02:47, 3 December 2014 (UTC)

  I also think that an admin voluntarily resigning during the course of such a discussion should be allowed to bring it to a close immediately, though of course such resignation would be seen as "under a cloud". PaleAqua (talk) 02:52, 3 December 2014 (UTC)

• Oppose Arbcom desysops people plenty when it is really needed. Admins simply doing their job accumulate people who resent them. Hold a vote and everyone they ever blocked will come out with pitchforks and thinly veiled alternative reasons. An admin could not do their job under a popular vote system because enforcing the consensus of the community is an unpopular job.

  Who are these admins that need to be desysopped but the present system is failing? Where is the problem that this is trying to solve? I took a 3 year break recently and when I left there were 4 admins that I thought should not be. When I came back 3 were desysopped and 1 had cleaned up her act. I say the current system works.

  I would support an administrative conduct noticeboard where issues can be reported and discussed, the discussion there could be used to indicate to arbcom if the community wants the matter pursued. Also there is no rule against admins being blocked, topic banned, interaction banned or just plain banned by the consensus of the community. Chillum 04:01, 3 December 2014 (UTC)

• Support - The community needs a vehicle by which the community, not Arbcom, may desysop administrators in whom the community no longer has confidence. Furthermore, I strongly endorse the percentage !votes proposed by PaleAqua above: a simple 50% + 1 should be sufficient to desysop any admin. An admin who cannot maintain at least 50% +1 support has clearly lost the confidence of the community. Requiring a super majority of 67% (or 66 and 2/3 %?) should not be necessary -- that would imply that only 33% of the community still has confidence in the admin, a ridiculously low percentage. A 67% super majority to desysop turns the idea of a minimum of 70% support to be promoted to admin on its head. Dirtlawyer1 (talk) 04:19, 3 December 2014 (UTC)
• Support the community elects admins and the community should be allowed to remove admins as well. Even if this proposal fails, I would recommend the creation of a page where users can recommend that an admin be given a non–binding vote of no confidence. Mellowed Fillmore (talk) 04:42, 3 December 2014 (UTC)
• I think the proposal is too early to support or oppose but I would need to know much more about what we determine to be "uninvolved and experienced editors". Admins but themselves into conflict all the time. If everyone followed the rules and got along and there were no disagreements to be had, then there would be a significantly reduced number of admins. Having any 3 people on the same page, whether they're involved or indirectly is way too low a threshold to trigger a two week community de-sysopping process. Rather than uninvolved and experienced editors, why don't we have bureaucrats review cases filed and they would determine if it should go to a community lead process. Mkdw 06:58, 3 December 2014 (UTC)
• I would generally support something like this, although I'd prefer a 70-75% rate for desysop (i.e., a mirror RFA). WilyD 11:03, 3 December 2014 (UTC)

• @WilyD: As I noted above, a 70–75% affirmative vote to desysop translates as residual community confidence in the admin of only 25–30%. We require minimum community support of 70% to promote an editor to administrator. Do we really want to keep an admin in whom community confidence has fallen to 30% or even less? Dirtlawyer1 (talk) 14:51, 3 December 2014 (UTC)

• Well, until it actually happens a few times, it's a bit tough to guess what the right choice is, of course. Requests for de-adminship is unlikely to be a random sample of the community (much like Requests for adminship), which makes it hard to guess. If you think of it as yes/no, perhaps 50/50 makes sense. But in a more Misplaced Pages-esque decision space with "no consensus" in the middle, there's also space for "no consensus" to de-admin. It's just a guess, and would be worth tinkering with if people are unhappy with the results. WilyD 16:03, 3 December 2014 (UTC)

• Wily, I agree that it's a guess, but the historical pattern at RfA has been a high percentage of participation by existing admins, and I would expect that pattern to continue in a vastly more controversial community desysop RfC. If a majority, or even a substantial plurality of admins !vote to remove, the subject admin is a goner; likewise, if a strong majority of admins continue to support, removal is unlikely. Recent re-confirmation RfAs have shown that even controversial admins usually retain community confidence of 60+%. I would expect that attaining super majorities of 67 to 75% for desysopping would be be virtually impossible to obtain short of the admin's conduct being outrageously egregious and/or a complete meltdown, in which case the admin is more likely to resign than wait for the community's axe to fall. Bottom line: getting a 50% +1 majority to agree to remove an admin will not be an easy result to obtain, and getting a super majority of 67 to 75% would be virtually impossible. At the end of the day, this is about accountability, and if an admin no longer has support of 50% of the community, he or she needs to be held accountable. Dirtlawyer1 (talk) 16:35, 3 December 2014 (UTC)

• For your information, I've made a few changes to the proposal, which are written in small print. --Biblioworm 16:38, 3 December 2014 (UTC)
• Well, both of us are merely guessing; neither of us knows how likely people are to show up and support or oppose a desysoping, or with what standards. I think it's wiser to take whatever guess, and make it clear that it's subject to change if the choice is wrong, so people aren't needlessly intransegient in the future. As long as the choice is reasonable, it shouldn't impact whether the proposal is endorsed. WilyD 17:53, 3 December 2014 (UTC)

• I don't support the proposal as stands, but I would really like to see a community de-sysopping procedure. There are two significant issues with the current proposal - 1) It's a highly negative process. RfA is difficult for individuals when people start to oppose - this is asking for multiple burnout retirements, whether or not the final decision is for a desysop. Perhaps removing comments from any such vote might improve that, but we should be taking into account the real person behind the keyboard in any discussions here. Even better would be securePoll, if one could be set up. 2) The "starting criteria" are excessively wooly. How long should discussions discussions be to be considered extensive? What time period should they be over? How many events should they cover? How many people should have been involved?

  That said, I do like the idea of the "gatekeeper" being two uninvolved users. I'd suggest three, because a "hung jury" is less likely. There's a balance to be found, and I think this proposal is a step in the right direction. Worm(talk) 11:28, 3 December 2014 (UTC)

• Definitely would support a due process. Like any other action on WP, if there is a consensus to do something then this is no different. VegasCasinoKid (talk) 12:12, 3 December 2014 (UTC)
• Support in general, but I like the idea of 50%+1 rather than supermajority to prove the admin has lost the community's confidence. --SarekOfVulcan (talk) 15:02, 3 December 2014 (UTC)

  Also, 3 would be better than two, yes. Not more than that. --SarekOfVulcan (talk) 16:52, 3 December 2014 (UTC)

• Support with caveats; I think you need more than two "experienced" users to certify (three? five?). Also, I'd suggest a minimum edit-count for those voting, in a similar way to ArbCom elections (although perhaps not as onerous as that - min. one month & 50 edits?) otherwise I can predict votes being overrun with socks which is just a waste of everyone's time. I'd also agree with 67% - at least - admins make enemies, even when they're doing their jobs correctly. This is especially true of anyone working in controversial area (I/P, fringe medicine, etc etc). Black Kite (talk) 15:13, 3 December 2014 (UTC)

• Black Kite, I agree that the certification panel should include three (or five) members not two. I regularly draft closely-held corporate and partnership documents, and it is never a good idea to have a decision-making board with an even number of votes: it invites the inevitable deadlock. I also agree that only registered users should be permitted to participate, and with the further caveat that participants must have been auto-confirmed before the case was initiated. On the other hand, I disagree that a 67 to 75% super majority should be required for removal; that translates into continued community support of only 25 to 33% -- and that's a ringing vote of "no confidence" under any circumstances. Dirtlawyer1 (talk) 18:58, 3 December 2014 (UTC)

• My point was that 67/75 might be required (instead of 50) to offset the "this nasty admin once blocked me / refused my request / wouldn't block the idiot I was edit-warring with / protected an article in the wrong version / etc. etc., despite the fact that it was completely warranted" type of oppose. Even the best admins will, by the very nature of their work, have gathered a list of editors that don't like them. Black Kite (talk) 20:10, 3 December 2014 (UTC)

• Support, at this point I'd support pretty much any suggestion for a community desysop process, no matter what the details. --Floquenbeam (talk) 15:16, 3 December 2014 (UTC)
• Support It may need some work, but this is at least a good starting point. Intothatdarkness 16:39, 3 December 2014 (UTC)
• Question what is the desysop procedure for other (larger) wikipedias? Dusti 18:20, 3 December 2014 (UTC)

Proposal is off topic for this page

This is the talk page for our process for making new admins. A proposal that brings about a new process and an entirely new class of users need to be discussed in a more public venue like the village pump. This is the talk page for RfA, I don't see it as a relevant place to product a new process to remove admin access.

This venue does not provide the wide attention of the community needed to create a consensus significant enough to enact a change of this magnitude. Chillum 04:27, 3 December 2014 (UTC)

I'm just getting a general opinion right now, as I'm not quite ready to put this at the proposal page. (There's still some things that need to be worked out, such as the percentage that would be required to desysop.) When everything is smoothed out, I'll take this to the village pump. However, I could create a subpage for this in my userspace and have the discussion continue there. --Biblioworm 04:39, 3 December 2014 (UTC)

Fair enough, it is at least tangentially related to RfA. Chillum 04:58, 3 December 2014 (UTC)

• Yes, WT:RfA is definitely the wrong venue, but I support the proposal in principle because I think it's high time some major RfCs were started on the topic of community desysoping which I broadly support anyway. If the proposal gets laughed out (as one of mine did recently), no harm done - come up with another. It would be good to keep the community on its toes for a while on such an important issue. Sooner or later one idea will gain traction and after a bit of the inevitable tweaking, the community will agree on something. --Kudpung กุดผึ้ง (talk) 14:22, 3 December 2014 (UTC)

What is an "experienced user in good standing"?

Since this wording may appear to be vague in the proposal, I want to obtain some opinions on what constitutes an "experienced user in good standing". I would define an "experienced user" to be someone with 3-6 months of active editing, and about 1,500-2,000 edits. A user in good standing would be someone who is not currently under any sanctions and has not been under any for at least 3 months. --Biblioworm 16:43, 3 December 2014 (UTC)

I would say an "experienced user" is a user who's been here 6+ months with 4,000+ edits. just my 2¢. --AmaryllisGardener 16:46, 3 December 2014 (UTC)

Biblioworm, under your criteria, I wouldn't be an experienced user in good standing, so I'd have to do a NIMBY oppose. :-) Any suggestions on how to loosen it without going to a straight edit-count/time criteria? --SarekOfVulcan (talk) 16:48, 3 December 2014 (UTC)

• I'm going to oppose. An experienced user in good standing? What, is this a new rights group or is it arbitrary or ...? We already have autoconfirmed users. If you arbitrarily set this, what means will be in place to prevent an ever increasing benchmark of the subjective assessment of "experienced" and "good standing"? Why not apply this metric as a suffrage benchmark for RfA? If not, why not? The further and further we get away from the principle of "the 💕 that anyone can edit" the further down into the pit of hell we go. Either you respect all editors who have not proven themselves to be a malicious entity on the project, or you don't. If you don't, then you are opposed to the very foundation on which Misplaced Pages was founded. --Hammersoft (talk) 17:46, 3 December 2014 (UTC)
  • I've struck "experienced" in the proposal, as we'll never agree on what exactly constitutes an experienced user. Besides, multiple users would have to certify any case, so they would probably keep each other in check. Now, we need to decide what a user in good standing is... --Biblioworm 18:03, 3 December 2014 (UTC)
    • We'll never agree on 'good standing' either. I think good standing, at a maximum, is equivalent to autoconfirmed and not blocked. --Hammersoft (talk) 18:19, 3 December 2014 (UTC)
      • In the interests of simplicity, I've struck both. I've also added a note that admins could take on the role of certifiers. --Biblioworm 23:20, 3 December 2014 (UTC)

A "user in good standing" is either very vague or it is is a new class of user that needs to have a selection process and... a removal process. Sound familiar? I suppose we will have a RfUIGS board to vote them in, and some then people will start to disagree with them and demand a system for their removal. Who chooses which UIGS are used? If one UIGS says it is bunk but 2 UIGS say it is valid is it a question of who gets there first like a race?

Any solution that simply shifts trust around will not work as we will be in the same situation we started in.

I say anything that it worthy of losing your admin bit is worthy of a block. The community just needs to be willing to block admins. Any admin who unblocks themselves will lose their bit. Any admin that reverses a block that was the result of a community consensus will likely lose their bit.

Arbcom can desysop, the community can block and even ban an admin. If we can come to a consensus to desysop then we can come to a consensus to restrain poorly behaved admins. Admins are not special and the community needs to stop making them a special case. Chillum 18:05, 3 December 2014 (UTC)

• Comment neither supporting nor opposing. I support the idea in general, but I cannot support the proposal as now stated. Selection of uninvolved editors will be problematic at best, because even individuals who have had no previous involvement with the admin in question may be perhaps involved in some other way, such as perhaps the admin engaging in dubious conduct with an editor that individual likes on-wiki or off-wiki. Without a fairly clear idea how such people would be selected or drafted, and I might prefer drafting or at least requested to take part from a list of theoretically eligible editors, I can't see how this would have much chance of success. John Carter (talk) 19:07, 3 December 2014 (UTC)
• As stated above, I think having only two random users review a case to trigger a community de-sysopping process is too ambiguous and could too easily be misused or abused. Based upon some of the RFA's I've seen, 3-6 months experience would not be remotely enough time for someone to know whether or not a sysop was abusing their administrative privileges. I believe there should either be an appointed committee to review submitted cases, or it should be given to an existing group that already has community support such as bureaucrats or even other administrators who are appointed by community consensus. Mkdw 21:10, 3 December 2014 (UTC)

This is just a version of WP:RFC/U

For all the good intentions here, this really just works out to a version of WP:RFC/U..with teeth. It would be part of dispute resolution, two or more people need to certify it, and it remains open for two weeks (instead of four). Sound familiar? We're about to decommission WP:RFC/U in a landslide and landmark decision for what was previously an important part of the project. Now we're going to replace it with what is essentially a clone, but this time with teeth? Ummm... --Hammersoft (talk) 20:06, 3 December 2014 (UTC)

Could admins take on the role of certifiers?

I know there will be objection to this idea, but I wonder if uninvolved admins could be the certifiers for a desysopping case. I'm proposing this because "uninvolved users" might be a point of concern when this proposal is formally proposed, as the community might worry that any user, newbies included, could certify a case. To avoid this issue, do you think the certifiers should be admins? (Then again, we'd be defeating the entire purpose of creating this, as this process is supposed to advance the idea that "adminship is no big deal".) --Biblioworm 23:20, 3 December 2014 (UTC)

• Nope nope nope nope. This has been thought of before, and soundly rejected because then the idea that admins are a self enforcing cabal comes right forward. Want to desysop an admin? Well, you've got to get two admins to agree with you before it can even go to a vote. It just doesn't fly, and won't fly. See, the vast majority of people agree there needs to be a desysopping process. The problem is the devil is in the details, and nobody has come up with any system that isn't fraught with serious issues. --Hammersoft (talk) 23:36, 3 December 2014 (UTC)

By the way, I looked at Misplaced Pages:Requests_for_de-adminship#Proposed_processes and found two processes right off that sound very similar to what you proposed above. Both are from 2005...9 years ago:

1. This one, in which ten users needed to certify for it to go to a vote.
2. Misplaced Pages:Requests_for_de-adminship/Proposal_2#Petition, in which ten users plus three admins had to certify for it to go vote (and users had suffrage requirements).

All this has happened before, and will happen again. Everything old is new, everything new is old. We keep returning to the same proposals, and they keep being rejected for fundamental and/or grave structural issues. --Hammersoft (talk) 23:48, 3 December 2014 (UTC)

Refined version

I've refined the proposal a little, in accordance with some suggestions from the users giving opinions on this. Does it look a bit better now?

1. Before a case request is filed, the issues must have been discussed at other venues, such as the admin's talk page and ANI. The issues also cannot be minor. For a case to proceed, the admin must have displayed repeated poor judgement, or have committed a particularly serious violation.
2. If all the requirements are met, the concerned user may file a case request.
3. After the case request is filed, three uninvolved users will research the matter, and they will certify the case if they feel that the issues are serious enough to warrant a desysopping case. If the case is not appropriately certified in one week, the case will be closed as stale. If the case is properly certified, the case will proceed.
4. The discussion will run for two weeks. If a simple majority (51%) support desysopping, the admin will be desysopped by a bureaucrat or a steward, although the former admin may file an RfA at any time. (This part of the proposal is still open to modification; for example, should there be a discretionary range?)

--Biblioworm 23:57, 3 December 2014 (UTC)

There are several things that stand out here as being serious structural issues. There are requirements that

1. Prior to the case being brought, it must have been discussed at other venues. Who decides if it has or not?
2. The issues must not be minor. Who is the arbiter of what is minor and not minor?
3. The admin must have displayed repeated poor judgment. Who decides what is poor judgment and what is not?
4. Once those requirements are met, the case may be filed. Who decides if all the requirements are met?
5. A case can be certified by three users. Are they the ones that make the above decisions?
6. Do we checkuser all the certifiers to ensure there is no sockpuppetry going on?
7. The certifiers must be uninvolved. How do we determine that? The term 'uninvolved' has been embroiled in controversy before. Unfortunately, you're going to have to define 'uninvolved'.
8. The certifiers are expected to do research into the case. What constitutes research?
9. Is there an evidence page where people can submit items as evidence? Who decides what is admissible evidence and what is not?
10. What if three users certify, but three oppose, stating that the case is frivolous?

The whole notion sounds promising...until you look into the details of it. If you're asking me to support the idea that we need a desysop process separate from ArbCom, sure. If you're asking me to support the proposal as worded? The proposal, as is, is empty of significant details to know if this system would work or not. I'm being asked to buy a car without knowing anything about the car, other than it's a car. --Hammersoft (talk) 00:03, 4 December 2014 (UTC)

Please continue discussion at User:Biblioworm/Desysopping proposal

The people who have commented on this thread have brought several issues to light, and therefore, I do not believe that this is quite ready to be formally presented at the village pump. I have copied this proposal to User:Biblioworm/Desysopping proposal, where I will work on it and try to incorporate the suggestions that the commenters have given. Please continue discussion of this proposal on the relevant talk page. (You may want to watchlist the page if you're interested in keeping track of the changes.) Thanks, --Biblioworm 00:21, 4 December 2014 (UTC)