Wikipedia talk:Removing administrator rights

kick off

I think this is sensible. Privatemusings (talk) 01:47, 26 August 2008 (UTC)

btw - it's a lot like Wikipedia:Removal_of_adminship, but with a clean slate. The other proposals at Wikipedia:Requests_for_de-adminship#Proposed_processes are well worth studying, but seem to me to have a degree of 'baggage' - I'd like to try and mine the good stuff, without getting bogged down! Privatemusings (talk) 02:26, 26 August 2008 (UTC)

(ec)I agree the community needs some method of removing admins in whom they have lost confidence, that is community based rather than Arbcom based. The problem is in ensuring that the normal work of an admin does not make them have to go through the process every week. Currently RFC is a more generalized process that may have many outcomes, and I think a desysop process (RFDS) should only have 2 outcomes: the admin receives support, and keeps the bit, or they lose it. So as I see it there would need to be some safeguards, first, to ensure that there are not frivolous or unwarranted requests, and secondly that discussion/votes are solely based on admin actions, not on general behavior unrelated to their admin work. So to deal with an admin who is chronically incivil you would use RFC or some other process because the desired outcome is civility, not a desysop.

The first safeguard could be achieved by having the process start only when a certain number of other admins or bureaucrats agree that it is warranted. This is not to prevent non-admins from having their say, just to have users who the community have given their support regarding admin issues make the decision. Once the process is started anyone should be able to have their say.

The second will be harder, as it is a more subjective area. Basically I would advocate striking out discussion that is not relevant to the use of admin tools, or is not at least tangentially related. So if AdminJoe has handled a content issue poorly, but not used the tools in that issue, then that should not be a point of discussion. I'm not sure how this could be fairly enforced though. Kevin (talk) 02:28, 26 August 2008 (UTC)

good points, Kevin... I'm optimistic that the wiki can handle a process like this, without necessarily overly focusing on too many rules at this stage.... I think best practice evolves rather than is written ahead of time (it's a corollary of the 'wikipedia only works in practive, never in theory' thing, for me) - so I agree the issue you mention could be problematic, but I think the community is functional enough to handle it! cheers, Privatemusings (talk) 02:36, 26 August 2008 (UTC)

Our bureaucrats can make sure that the process is fair, and infrequent. This is why I suggest a two stage process: 1/ RFC to show that there is doubt about an administrator, and then 2/ reconfirmation RFA. This allows time and helps make sure there is broad awareness. RFCs are good for generating evidence and hearing multiple views. The typical RFC does not get as many comments as a well attended RFA. Jehochman ^Talk 02:40, 26 August 2008 (UTC)

I would like to see some discussion of the rationale for this. I keep asking and getting no response: what is the problem to which this is a proposed solution? Do we have rogue admins misusing their tools? Are current processes (such as warnings, blocks, Requests for Comment, and Request for Arbitration) insufficient or incapable of dealing with admins who do misuse their tools? Is it especially difficult to desysop problem admins? There seems to be an exceptional amount of interest in this topic recently, much of which seems to be directly attributable to one current Request for Arbitration. All this fuss seems to be creating a mountain out of what amounts to a molehill. Exploding Boy (talk) 02:44, 26 August 2008 (UTC)

Yes, there is a problem that the community has the power to give (and take away) the sysop bit, but there is no clear process for taking away. This has spawned may controversies and arbitration cases that were not necessary. A desysopping process has been a perennial proposal, because people want it to exist. Lack of a community desysop process creates the appearance that adminship is forever, and that once you're in, if you stay on ArbComm's good side, you can abuse as many editors as you like and there will be no consequences. We need to change that. Jehochman ^Talk 02:51, 26 August 2008 (UTC)

ArbCom, as an entity, has far too much potential for corruption to be trusted with something like this, in my opinion. They are very close to the administrators, especially recently and often work in their favor to the detriment of the community and its ideal. They can choose to ignore consensus and act of their own accord. Any alternative process that reallocates power from them to the community itself is good. However, this in and of itself is necessary for other reasons; where the existing power structures have failed to be reactive toward the situation, it is clear that a new system is needed. Since the initiative to take action exists within the community, and we are a project that is supposed to be governed by the community, it seems only logical that it should be the community, and not some specialized entity rife with potential for abuse, to be given the power to take action against administrators who no longer have the trust of the community. This is, of course, assuming that the Committee allows this to occur by not passing some kind of "enforcement provision" to contravene previous policy decided by the community, as they have before. If that were to happen, in the current state of affairs, I'm not really sure how that would turn out ... Celarnor ^{Talk to me} 04:04, 26 August 2008 (UTC)

(to PM)I guess that bureaucrats can keep the discussion part under control, but I'm not convinced that RFC is the best way to initiate the removal of rights. I can't quite see the community agreeing to this type of remedy without causing much drama. I'm in favour of a definite line in the sand. Kevin (talk) 23:49, 26 August 2008 (UTC)

the comments above and below make me think that you've hit the nail on the head, Kevin.... The central issue seems to be a cost / benefit analysis of proposals, likely weighing 'drama' against productive resolution - there's a baby and some bathwater here. I tend to think that there's some implementation of this worth pursuing, and believe that not all the unintended consequences would be bad... Privatemusings (talk) 00:21, 27 August 2008 (UTC)

One of the things that is likely to happen if some process for de-adminning is accepted is that everyones shit list is going to be dragged out, and if there are not clearly defined conditions or guidelines for initiating the process then we will have 50 admins up for removal, and much of the meta-community arguing about the merits of each. We would have to ensure that the net effect is positive. Kevin (talk) 03:53, 27 August 2008 (UTC)

We are not proposing that at all. What you suggest could just as easily happen if people take their grudges to RFC and then to RFAR. But there are controls in place to prevent that. The proposal also includes the some existing controls and some new ones: RFCs need to be certified, and broad community input will be solicited so that no small group of editors will be able to game the system. Our bureaucrats are smart enough to spot that sort of gaming. By the way, the proposal mimics my existing recall criteria, and nobody has attempted to game my recall yet, though there are many who probably would like to do so. Jehochman ^Talk 04:00, 27 August 2008 (UTC)

I understand what you are proposing, that is if a community consensus determines that an admin should no longer be one, a bureaucrat will be able to request a desysop from a steward. Perhaps my thoughts are a little cynical re the community's ability to limit drama, which why I would give more support to a more focussed process. Kevin (talk) 06:09, 27 August 2008 (UTC)

rationale

further to the above - I thought I'd also explain one aspect that attracts me to this proposal. I kinda like the idea of adminship being 'no big deal', and I think this process would help make it less of a big deal than it is at the moment ('easy go' usually follows 'easy come'!) - I hope good things could come of this, and maybe the culture of wiki-administration would benefit :-) Privatemusings (talk) 06:23, 26 August 2008 (UTC)

Exactly. If, as we are constantly bombarded with information supporting such a statement, that adminship is nothing more than a mop that can reasonably be given to anyone with policy knowledge and a decent track record in policy venue and article work, then it should stand to reason that the mop should be able to be revoked if and when it becomes apparent that they're using it contrary to the collective desire of those that gave it to them. It doesn't make sense that something so relatively freely given by a consensus shouldn't be able to revoked by a similar system accessible to the community. Not providing the community with the tool to remove power as well as give it--and no, the Arbitration Committee, as an entity that frequently skirts the community and is seemingly no longer bound to consensus, does not provide that--suggests that there's more to it than that. Celarnor ^{Talk to me} 07:42, 26 August 2008 (UTC)

• Page renamed from Wikipedia:DeSysop, to Wikipedia:Removing administrator rights: probably a more appropriate title. Anthøny ✉ 20:34, 26 August 2008 (UTC)

Some thoughts

It's been said somewhere else by someone (not me) but I can't find it: if proposal 2 is adopted, I'm going to create a shortcut from WP:LYNCH. The process is simply too easy to initiate, and once started, it will gather unwarranted support from people swept up in the heat. Although the probability of the discussion getting so out of hand as to prompt a 'miscarriage' and a desysop where one isn't actually warranted, is remote, the damage caused is much more pervasive: it's going to be a horrible experience for the potentially largely innocent admin involved, having everything they've ever done picked apart and put under the microscope. Proposal 1, by contrast, has much more potential. Starting with an RfC is already the standard practice, and I can see no reason to change it. Given the looser focus on 'desysopping' it's likely to be a friendlier exercise. It's how to proceed from there that is the stone we're currently missing, and I can't see any serious issues with the bureaucrat-determines-consensus-for-confirmation-rfa proposal. If the candidate fails their RfA - judged using the existing criteria - then it's entirely appropriate for the stewards to act as 'technical support' for the process in the same way that 'crats act for BAG in flagging bots. All in all, this proposal serves to link existing chunks of process (RfC, confirmation RfAs, stewards desysopping in response to requests) together in a more usable and coherent fashion. I personally have no problem with the bureaucrats acting as the 'glue' to bind these processes together. Needs considerable revision for clarity, but support the principles of proposal 1. Happy‑melon 14:00, 26 August 2008 (UTC)

I'm not so sure it's remote at all. Look at the difficulty people are having agreeing on a reasonable rationale for endorsing or opposing nominees for adminship. And the five or six concurrent discussions of this proposal going on across Wikipedia suggest no one really has any idea what a fair way of doing this would be either. Exploding Boy (talk) 21:20, 26 August 2008 (UTC)

Arbcom is sufficient

As far as I can tell:

• If there is firm evidence of sockpuppetry or other severe abuse, admins are desysoped almost immediately.
• The arbitration committee regularly desysops admins for less severe, but still inappropriate, use of admin tools.

Given those impressions, I don't see any need for an additional process for desysoping. If there truly were strong community consensus that someone should be desysoped, that would be apparent to the arbitration committee, and they would weigh it appropriately. It seems to me that a desysoping process that leaves out arbcom would only short-circuit the dispute resolution process, and be prone to rash usage.

So I can't support a new system for desysoping by "community consensus" at this time. My reservations are heightened by the lack of any justification for a new desysoping policy on the proposal page.

I would support arbcom if they began (with due consideration) to use remedies requiring specific parties in arbcom cases to undertake reconfirmation RFAs. — Carl (CBM · talk) 15:24, 26 August 2008 (UTC)

There are some cases where arbcom took months to desysop, after it was obvious that it was the right thing to do. I consider them mainly useful for the "smoking gun" type cases. Friday (talk) 15:27, 26 August 2008 (UTC)

I don't think that the complaint "arbcom is too slow" is properly resolved by replacing them with another forum. As I said, there isn't any justification on the proposal page, which makes it hard to give a firm argument against. However, I can say that I don't feel any of these would be a sound justification for a new process:

• Arbcom is too slow. As I said, they move quicky in cases of severe abuse. If there is no need for an immediate desysop, a deliberative process is better.
• Arbcom didn't desysop these people who I really want to be desysoped. Self-explanatory.
• Arbcom desysoped these people who I know shouldn't have been desysoped. Self-explanatory.

— Carl (CBM · talk) 15:33, 26 August 2008 (UTC)

I think Arbcom is sufficient. All of the proposals listed which require a crat to judge and act can as well be judged and handled by Arbcom and there is then no need for changes. Maybe, as CBM says, they should consider re-confirmation as a sanction when de-sysopping is too harsh. SoWhy 20:48, 26 August 2008 (UTC)

• We want to move away from a model where a small group of users, alebeit trusted editors who we appoint, take actions. Wikipedia runs on an open-community, consensus-building model where and when we ever can: steadily reducing our reliance on crude measures such as the Committee, until the point where we return to the "true" Wikipedia model, is what we're ultimately aiming for. I do, of course, largely speak from principle; in practice, the ArbCom performs an essential and crucial role, and we're a long way off yet from no longer needing them. But every little counts, and so long as we think things through (wait until after the RfA review for feedback?), we can do this correctly. From building a successful model for allowing ourselves to implement our own consensus on administrators retaining / losing their status, good things will follow. Anthøny ✉ 22:24, 26 August 2008 (UTC)
  • I'm not sure who "we" is, but I'm not included. — Carl (CBM · talk) 23:28, 26 August 2008 (UTC)
    • I'm guessing "we" is folk whose head fits AGK's cap? I think I'm one! Weeeee! :-) Privatemusings (talk) 23:43, 26 August 2008 (UTC)

Implementation 0 is the way to go. We should keep the focus on actions rather than people. Every active sysop has made enough enemies to form a mob. Judgments like this are better made by arbitrators after other efforts at dispute resolution have failed. Tom Harrison ^Talk 23:06, 26 August 2008 (UTC)

have all admins really made "enough enemies to form a mob"? - that sounds really bad, and may reflect a deeper problem? I think we need to be really careful to avoid squishing conversations here with 'FUD' (being willing to talk, think and examine stuff has got to be a good thing - p'raps better than being certain immediately?) Privatemusings (talk) 23:23, 26 August 2008 (UTC)there's a nice crunchy biscuit in it for anyone who can come up with a suitable joke involving a 'female urinary device' after my comment :-)

If there are any admins who have enough enemies among established users who are assuming good faith, then they probably should have their posession of the tools reexamined. That's kind of the point. Celarnor ^{Talk to me} 05:58, 27 August 2008 (UTC)

Not all admins, certainly, but many, and most of those willing/able to work in more controversial areas. But yes, it is bad and it does reflect deeper problems. It's not a view I've arrived at hastily. The community as it is today is ill suited to making this kind of judgment with careful deliberation. Collaborative editing is a great way to write an encyclopedia; to manage the project, not so much. Leave the process with the arbcom. It's a small group focused on actions instead of personalities. That's more likely to give us results that help the project, and less likely to discourage people from enforcing copyright and blp, making difficult blocks, or doing what's right instead of what's popular. Tom Harrison ^Talk 23:47, 26 August 2008 (UTC)

hmmm.... I get a bit concerned that we're in danger of papering over the cracks... it doesn't seem sustainable to me to say that the community is ill suited to making this kind of judgment - I'm not sure where the boundaries for acceptable community judgment would be? We've seen a sort of similar thing recently in the BLP area - where the community may be equally unable to act appropriately? (perhaps hence the arbcom creation of 'special provisions'?) - without prejudice as to discussion of the competencies of 'the community' as a whole, and the quality of 'community judgment', this just seems like a not-so-healthy (and not-so-wiki) direction to be taking.. it's actually part of my rationale that 'community should be king' - exceptions seem to me to cause more problems than they solve... Privatemusings (talk) 00:00, 27 August 2008 (UTC)

Well, we can leave for another time the underlying problems of our particular community. I'm still not convinced that having a large group make a decision about a man is better than having an elected group make a decision about the man's actions. That seems likely to encourage group think, and to work against the open inclusiveness we both support. I'll continue to follow the discussion and see what people think. Tom Harrison ^Talk 00:19, 27 August 2008 (UTC)

I don't think that arbcom is the solution to a community based problem. The community decides who the admins are, so the decision to remove should also vest in the community. We are talking about a limited subset of user behavior here - actual or potential abuse of admin tools. What is needed are safeguards to ensure that those working in unpopular areas are not unfairly targeted. Kevin (talk) 04:03, 27 August 2008 (UTC)

It doesn't seem sustainable to me to have a smaller group of people make these kinds of decisions. I don't think that the Committee, which frequently delegates its powers to administrators and interacts with him on a basis more regularly than they do normal editors is quite detached enough to handle the issue; while it may work in an ideal world, this is not an ideal world; a power of the community that is that important should remain with the community, not delegated to some group that the community can not directly control, only elect periodically. As the consensus of the community is not binding on the committee's connection, there's no way for us to enforce our will that an individual admin have their tool stripped.

Everything about that idea just seems to fly in the face of the consensus that we use to operate on pretty much everything else, including bringing those people to power. Celarnor ^{Talk to me} 05:55, 27 August 2008 (UTC)

Perennial proposal

At various times, proposals have been made to vote admins off the island by mob rule. All of these have fallen down as stupid. How is this one different to past ones? - David Gerard (talk) 16:47, 26 August 2008 (UTC)

Did you read this? I don't consider the crats a mob, do you? I thought they were generally reasonable people, chosen for their good judgement. Friday (talk) 16:55, 26 August 2008 (UTC)

Crats are chosen for their ability to stick strictly to one-moron-one-vote in RFAs. Look at any RFB in the past few years - David Gerard (talk) 17:25, 26 August 2008 (UTC)

david, can I ask you nicely to sort of tone it down a bit? Whilst some celebrities may want out of here, I don't think there's really any problem with talking about things a little less histrionically (is that even a word?!). I wouldn't describe admins as elected by mob rule, nor is it very nice to describe past proposals as stupid - play nice! Privatemusings (talk) 23:19, 26 August 2008 (UTC)

We need something like WP:DESYSOP more than ever, so that we don't have to read condescending "neener neener, you can't touch me" tripe like this all the time. SashaNein (talk) 14:14, 29 August 2008 (UTC)

Steep slippery slope on fire

Very bad idea. Bad. Jerry talk ¤ count/logs 02:31, 27 August 2008 (UTC)

Thank you. Could you say specifically what is wrong with the proposal? Jehochman ^Talk 02:44, 27 August 2008 (UTC)

Probably that, like other administrators, he could be held accountable for his actions.. :P Celarnor ^{Talk to me} 06:16, 27 August 2008 (UTC)

More that careless editors sometimes sling around accusations and assumptions without assuming any good faith whatsoever. Read directly above for an example! Jerry talk ¤ count/logs 00:18, 28 August 2008 (UTC)

I read Celarnor's comment as rather tongue in cheek, Jerry (hence the :P bit!) - I don't think admin.s need be fearful of the community - and I don't think that fear is healthy.. I hope a proposal like this can help on that level too :-) Privatemusings (talk) 00:27, 28 August 2008 (UTC)

I was being sarcastic. Always assume the assumption of good faith. Naturally, good faith is important, but it isn't the be all and end all of everything that happens here. For example, I'm going to have a hard time believing that a redirect of the main page to goatse is in good faith. Admittedly, that would probably be egregious and obvious enough that the Committee would take action (or, more realistically, a steward or bureaucrat with the later support of the committee). My problem is with the administrators who abuse their power to less obvious extents; those that have a long history of closing AfDs contrary to consensus and getting continually overturned at DRV have no business closing AfDs. However, posting something like that at RFAR would just get you ignored, and rightly so; the committee has for more important and pressing matters to deal with, especially when the actions of the administrator in question can simply be overturned at DRV. These kinds of things are easily performed and easily overlooked actions that need more analysis than the Committee with its finite amount of time and more important concerns can provide to confirm and judge on.

However, such problems still exist, and at the moment, there's no realistic way to go about solving them. While I trust the committee less than I do the general populace as a whole, my reasons for doing so are intrinsic to the Committee's function and power structure; it isn't a personal grievance, it is simply an acknowledgment that isn't really a workable system. In any case, the way it is now, they simply can't deal with complex issues that require a lot of in-depth analysis of contributions and logs unless its a big enough problem to warrant months of investigation. I think the community is in a much better position to offer these kinds of solutions.

Practical considerations aside, if there accusations, they should be dealt with unless they're obviously out of the blue and due to a grudge. If the tools are something to be taken lightly and given to reasonably anyone who doesn't seem like they're going to misuse them, then taking them away for abuse should be just as easy. If they aren't something to be taken lightly, then its all the more important that the community be able to keep a close eye on those who have them (which they can already), and to be able to collectively come to a decision about whether it is a good idea for them to be having them.

Whether that system is arrived at by annual/semi-annual/whatever re-confirmations (my personal preference, as it assumes a positive outcome) where people collectively look at contributions and decide if an administrator is a good candidate to keep using the tools or not, a system where people can go to request a reconfirmation of an administrator, or an outright "This person shouldn't be using the tools because x RFC", the community needs to be able to have some kind of direct say in an administrator's power after they become administrators. Simply because they work in controversial areas should not be a "get out of jail free" card. If an accusation is spurious, I have faith that the community is capable of seeing it as such by judging the relevant contributions. Celarnor ^{Talk to me} 02:11, 28 August 2008 (UTC)

Well stated position, Celarnor, and certainly more appreciated than the tongue-in-cheek poke at me of your previous post. I do disagree, however. Under a pure wiki system, the people are empowered. Since all rules are subject to being ignored if they get in the way of improving the project, then the people can really do anything they want. Let's say, that some new admin really let the position go right to their head, and they started closing XfD's based only on their own opinion, and left a closing comment that included "despite the opinions of the army of numbskulls who contributed here": what would happen today? Each such AfD would go to DRV and each such AfD would quickly become overturned. Who's time would be wasted? One editor would have to submit the DRV, and post a talkpage template for the offending admin. A few (maybe like 5 or 6) editors would leave unanimous comments on the DRV, and some admin would speedy close as overturn. A whole whopping 8 people would waste a whole whopping 2 minutes each. That's 16 person-minutes. And the whole situation would be fixed.

Ah, but only temporarily, you may say. What if the bad admin just keeps on doing it? Well, the community would just keep on DRV'ing them, and they would keep on getting fixed. In this rediculous fictional scenario, the admin would undoubtedly get tired of wasting his own time, and just stop. If not, then he'd get edit-interrupted byt he "you've got new messages" banner all the time as other editotrs and admins keep admonishing him. In the end, he would just stop. If not, then the committee would surely hear the case and put an end to it.

Now, how would a venue where a complaint gets opened, evidence gets posted, people all discuss it, some with completely tangent motives, and some crat having to sift through it all compare? Much more wasted time. And the potential exists for this new venue or process to be abused, such that trivial differences or misunderstandings or poor assumptions (like the misunderstanding between you and I earlier in this thread) get trumped-up to a witch hunt.

Our anti-bureaucratic system simply does not need a new process or layer to handle such things. a pure wiki can withstand most anything, given there are enough interested editors to maintain it.

Thus I say fie on this proposal for more bureaucracy and more process wonkery. Jerry talk ¤ count/logs 15:56, 28 August 2008 (UTC)

Play Nice

Would editors please state specific objections and try to work towards consensus by making constructive criticism. I know a lot of proposals have been rejected before, but that's no reason to stop trying to find a correct formulation. Thanks! Jehochman ^Talk 04:02, 27 August 2008 (UTC)

It's very difficult to state specific objections because no specific justifications have been made on the proposed policy page. Several comments on this talk page seem to assume that there is some intrinsic benefit in a "community based" method, but I don't agree with that. According to m:Foundation issues, consensus is the decision mechanism for content. There's no reason to expect that it should be used for all aspects of the site administration as well. Another possible justification is that currently, if arbcom declines to desysop someone, for "there's no way for us to enforce our will that an individual admin have their tool stripped." [1]. We elect arbcom precisely because enforcing the will of the gathered onlookers is a poor way to resolve disputes. — Carl (CBM · talk) 14:32, 27 August 2008 (UTC)

My specific objections are purely ideological in nature. The thought that the community is not empowered to take away the tools that they are empowered to give really irks me; outside of a slow, deliberative, and uncooperative committee, the community doesn't have any power to act on abuses and improper use of the tools. The committee itself isn't enough unless they can be spurred to action by the community, and that simply doesn't seem to be the case, especially at this point when they not only skirt the community, but also act contrary to it. Celarnor ^{Talk to me} 16:02, 27 August 2008 (UTC)

null proposal reword

I took out the rather poetic bit of the null proposal because I'd like to keep the page as bland and neutral as possible.. plus the less shifting winds the better I reckon.... can someone open a window? :-) Privatemusings (talk) 11:46, 27 August 2008 (UTC)

Time to mark as historical?

This proposal seems to have lost all interest, and discussion seems to have ended. Unless there are objections or renewed activity, I intend to mark it as historical in a few days. Jerry talk ¤ count/logs 01:28, 31 August 2008 (UTC)

There is an ongoing, related discussion at Wikipedia talk:Requests for adminship##The_concept_of_.22community_confidence.22. Jehochman ^Talk 02:31, 31 August 2008 (UTC)

I know wiki time moves fast - but I'm not sure I'd mark this as 'historical' just yet!! - In fact my reading of this, and other conversations indicates broad support for the concept, with a strong minority dissent. It's not clear really yet whether this could gain community consensus, or be seen to be inappropriate.. but I think matters are moving in the right direction :-) Privatemusings (talk) 03:16, 1 September 2008 (UTC)

Too early to mark as historical. We're in no hurry here. I'm not initially inclined to agree with the proposal, but I've only just found out about it, and would appreciate a little time to think without the body being buried before it's cold. -GTBacchus^(talk) 04:19, 1 September 2008 (UTC)

I don't think closing it at this point is appropriate; there's a lot of discussion regarding recent matters that this would serve as a solution to, both at RfA and elsewhere, and I don't think marking a page only a few weeks old as historical is particularly helpful or useful. Celarnor ^{Talk to me} 04:22, 1 September 2008 (UTC)

Not a bureaucracy

I've been thinking about this and similar proposals, and my instinct is to say, "we don't need to set up another bureaucratic procedure to formalize what already happens by natural processes." I understand however, that people want to have a formal procedure for desysoping, because the perception among many Wikipedians now is that the community is empowered to grant the bit, but not to remove it. We could set up a procedure to put the desysoping power explicitly in the hands of the community, and it would probably make some people feel better.

I still oppose it though, and here's why. If people perceive that the community lacks power to remove admins, and if we therefore set up a procedure for it, then we're agreeing with that perception, and reinforcing the idea that the community needs formal procedures in order to have power.

It would be much better, IMO, to convince the community that we already have that power, without needing to create bureaucracy about it. It is much better to educate editors about the Wiki way of doing things than to give up that struggle entirely, and resign ourselves to letting Wikipedia become more bureaucratic. Setting up a procedure for desysoping would be a surrender, and I don't think there's a need to do that. We would be saying that we're helpless without red tape, and I don't think that's true.

Those who believe that the community lacks power over sysops should not be coddled and enabled in that belief, but rather educated as to how we do have that power. Rather than inventing procedures, why not explore how we can already do what we need to do within existing structures? -GTBacchus^(talk) 15:29, 1 September 2008 (UTC)

We do have that power; or perhaps, more accurately, we did, until we set up all this bureaucracy we have now. We lost it when we gave it to ArbCom, investing it in them with the trust that they would use it when we needed it. Personally, I think we should eliminate the existing bureaucracy and take it back to an RfC-style decision where the community takes part with community concerns, in the original wiki fashion.

Whether you think we still have it or not depends on whether or not you perceive ArbCom as part of, or obliged to serve, the community. While this may have been the case at one point, the data is steadily starting to point in the opposite direction. The committee may have the power the sysops, but until we can convince the Committee that they are here to serve the community, not to spend all their time picking and choosing cases that allow them to go about collecting all decision-making power onto themselves, we don't anything other than the ability to cry pathetically and uselessly in the face of misconduct, which is pretty much what we do now. The Committee has the ability to simply not take the case; they don't even have to come out and say "No, we aren't going to desysop per the consensus of the community"; they can just ignore us without doing anything. Until the community has some viable mechanism to override bad decisions (or non-decisions) like that, the only power we have is the same power had by serfs in medieval fuedalism: beg our lords to hear our pleas for help.

Hyperbole aside, it doesn't need a formal procedure. It doesn't need another committee. I'm fine with what we have now--that is, RfCs--I just think it should be mandatory in such cases that our existing bureaucrats who are there to judge meta-level consensus do what we gave them the bit to do so and flip the switch. Then, if someone has a good idea, it should be taken to the Committee; the default then, i.e, what would happen if they simply ignored it, would be to keep the existing consensus. If there is good evidence to the contrary, then it can be reversed. This way, the Committee continues to have the power to override bad decisions, but the power to make them is returned to the community. Celarnor ^{Talk to me} 17:07, 1 September 2008 (UTC)

I think we do have the power to remove bad sysops because I've seen it happen, repeatedly. What we don't have the power to do is make it happen quickly, on demand by a group of editors. Something not having happened yet is not proof that it cannot happen. Arbcom may turn us away n times, but they eventually are forced to listen by sheer numbers of discontented editors. The viable mechanism is the same one we've got for dealing with content questions: reasoned discourse, combined with consensus-building and patience.

Also, remember who elects the ArbCom. If you take the long view, this problem is much less of a problem that I'm seeing people make it out to be, I think. -GTBacchus^(talk) 18:00, 1 September 2008 (UTC)

You've seen it happen? How? Outside of the committee, which isn't bound to consensus anyway, under the current system, I'm not aware of anything that allows us to desysop abusive admins.

And, no, ArbCom is not "forced" to deal with us. Barring extreme misconduct on their part, they're pretty much invulnerable to the community; our only recourse is not to elect them again, which doesn't heal the damage that their inaction causes; it simply keeps them from further maiming what we have now any more, and we get to repeat the cycle and hope that the next set of people decides to listening to us.

The Committee, or maybe an individual steward if enough people bother one, may have the power to remove bad sysops, but we certainly don't. Like the editor below me said, all we can do is stand around with increasingly strong statements to the effect of "That was a bad idea, please don't do that again", "Please stop doing that", "No, seriously, stop doing that!", etc, which is really pointless when we can't back them up with anything. We can't really do anything about it without the Committee, which, apart from only being bound to their own consensus and can completely disregard ours, takes a default stance of non-action; combined, those two factors is entirely unacceptable in my opinion. I'd rather see something to the effect of "A community RFC with widespread support can result in the desysopping of an abusive administrator" now than watch the inevitable unfold later. Celarnor ^{Talk to me} 18:11, 1 September 2008 (UTC)

I've seen it happen, via ArbCom. The pattern is something like this: A problem admin first upsets a small group of people, who complain with no effect. Then a larger group is upset, and there's an RfC that is dismissed by many power-haves as the whinging of a small group of discontents. Then, a significant chunk of the community begins to notice the problem, and there are more RfCs, with increasing support, and the admin's enablers making less and less sense. Then, someone files an ArbCom case, which is rejected. This happens two or three times, until finally the problem admin is involved in a high-profile dispute with other influential editors who are important enough that ArbCom reluctantly takes the case. If they're not desysoped the first time they're taken before ArbCom, then it happens on the second or third iteration, or else they resign, "under a cloud".

Now, that process is slow and frustrating; I agree. I would also like to see it happen more quickly, but I think the only way to make that happen, is, as I argue below, to change the culture. Suppose we enact what you say: "A community RFC with widespread support can result in the desysopping of an abusive administrator". Now, define "widespread support". What if there's also widespread opposition? Must the support outweigh the opposition? By what measure - certainly not a simple nose-count? If we have a culture of coddling and enabling bad admins, then what procedure can work against that? -GTBacchus^(talk) 18:30, 1 September 2008 (UTC)

I never said I have a solution, and I certainly don't have the specifics; I'm just acknowledging that this is a problems. Perhaps a nose count of enough people complaining can force the admin to be put up for reconfirmation with a slightly reduced requirement; perhaps some mechanism to force ArbCom to deliberate on the matter and come to a decision other than "We aren't going to take this case", or perhaps a simple RfC with certain codified participation requirements and a beaurocrat to analyze the resulting discussion and distill some consensus from it. Celarnor ^{Talk to me} 19:01, 1 September 2008 (UTC)

I realize that none of us has the perfect solution, and I have to say, I'm a bit torn. Part of me sees just where you're coming from, but another part of me says that it would be better for people who see it as a problem to address a cultural rather than a policy solution (which is much, much more difficult, as are many worthwhile things).

I mean, those ideas don't sound bad, but... what if we list a reconfirmation RfA, and then the coddlers come out in large numbers and win the day? Then that's not a solution. What if ArbCom, forced to take the case, says, "we see no problem here." Then that's not a solution. What if a B'crat just says: "I see no problem here," when we agree that there is one. Then that's not a solution.

Do you see what I mean, about a cultural problem? Any layer of bureaucracy that's steeped in a culture of coddling and enabling problem behavior isn't going to help curb that behavior. -GTBacchus^(talk) 19:11, 1 September 2008 (UTC)

The community has the power to stand around, point at things, and say "Yep, that's bad- you shouldn't do that." In most cases, this is all that's needed to keep sysops (or any other editor) in line. In a few cases, this doesn't do the job. All talk of revoking sysop rights is for those cases when normal corrective feedback doesn't work. If anyone has ideas on other things to try, let's hear them. But I don't see them above. Friday (talk) 17:26, 1 September 2008 (UTC)

I've seen ArbCom take sysop rights away from abusive admins several times. Are we claiming that this doesn't really happen, or that it doesn't happen often enough, or that it doesn't happen in certain cases where we'd like to see it happen, or... what, precisely are we claiming here?

Of the most abusive admins I recall seeing, most have lost the bit. A few others have changed their ways, and a couple are still active and abusive, but the more people they upset, the more their power erodes, and the ice on which they skate gets thinner and thinner... -GTBacchus^(talk) 18:00, 1 September 2008 (UTC)

I've seen it too. I just think problems like this should be nipped in the bud rather than allowed to fester. Of the bad admins I remember, yes, some have lost the bit via arbcom. Some have stomped off on their own. Some are still with us. And yes, some of them tantrum themselves into irrelevancy. My main complaint is that this is slow, and they continue causing problems during their downward spiral. Why would we not want to solve these problems in a timely manner? There's a huge bloc of editors who say things like "Yes, so-and-so is a problem admin. But, let's hold his hand and buy him a puppy and hope he shapes up." I'd rather pull the bit and then hope he shapes up. Problem admins can shape up on their own time- our goal in dealing with them should be to limit the damage they cause. Wikipedia is not therapy. Friday (talk) 18:14, 1 September 2008 (UTC)

As long as there is a significant bloc of empowered editors who are willing to coddle problem admins, no procedure or administrative trick will solve that problem. If we decide that an RfC needs to clearly indicate problem behavior, then we run into the fact that "clearly indicating problem behavior" is subjective, and enough people (remember, TINC) saying that the admin is not a problem make the determination unclear, leaving us right back where we are now. I love the idea of solving these problems more quickly (although I don't think of it as urgent), but unless we can change the culture of enabling problem behavior, how are we actually going to get that done? In whose hands can we put that power, and then trust that the power will be applied to our liking?

I think it's a cultural problem, and that it therefore cannot be addressed by a political solution. Does that make sense, or what am I missing? -GTBacchus^(talk) 18:31, 1 September 2008 (UTC)

You're probably not missing much. I have no magic solution. My best stab at a fix right now is for the crats to step up and fix this problem. They'd need to be willing to ignore the crowd of coddlers and do what's right. This may or may not work; the main reason I suggest it is that it might work, and if it doesn't.. so what? All that means is that we're back to where we are right now. Friday (talk) 18:38, 1 September 2008 (UTC)

Ok, what would it take for B'crats to do that? Do one simply walk into Mordor and ask them? Will they need some kind of community mandate? Will Jimbo or the Board have to confer those powers on them? I admit to not knowing much about that level of administration; wiki-gnome types like myself have trouble breathing the rarefied air up there... -GTBacchus^(talk) 18:48, 1 September 2008 (UTC)

(ec) I agree that social pressure is a good way to go about reigning in bad sysops, and that we should not create any new bureaucracy, and if possible, we should dismantle existing bureaucracy. However, if a sysop is behaving badly, at some point a bureaucrat should step in and say, "Enough is enough, you (dear sysop) appear to have lost the trust of the community. Get thee to RFA and pass again (at a reduce threshold of 65% or whatever), or else I am going to ask Mr. Steward to take away your bit." We shouldn't have to wait for a ticking drama bomb to explode. Better to defuse it once a situation becomes evident. We are not asking for any extra bureaucracy. All the necessary components already exist: RFC, RFA, and bureaucrats with the ability to judge consensus. The less we rely on ArbCom, the better, because ArbCom is a hellish bureaucracy. Jehochman ^Talk 18:18, 1 September 2008 (UTC)

Ok, why would you trust a Bureaucrat or a Steward more than you trust ArbCom? They're even harder to remove, aren't they? -GTBacchus^(talk) 18:33, 1 September 2008 (UTC)

For my part, I trust a bureaucrat or a steward more than I trust members of the Committee because I haven't seen any bad stewards or crats; the system seems good enough to only produce good crats and stewards. The Committee, not so much. I've seen two generations of irresponsible, non-responsive, community-ignoring ArbCom, and I haven't seen anything to indicate that this is going to improve any. But, the point is, with a system like this, a decision has to be made; whether the decision is "There is consensus for the removal of this administrator" or "There is no consensus for the removal of this administrator", someone has to say something somewhere along the line (barring, of course, the entire crat group deciding to ignore us, which I personally think is unlikely); there is no "We aren't taking this case". Celarnor ^{Talk to me} 19:06, 1 September 2008 (UTC)

How?

Ok, supposing we decide to go with a solution that involves empowering 'Crats in the way you suggest. What's the "how" of it? How does the community get that into effect? -GTBacchus^(talk) 19:13, 1 September 2008 (UTC)

I've wondered if an individual crat could run on the ticket of "I intend to fix any mistakes I make with promotions." This would be tricky- some people might like the candidate but not the campaign promise, and some would be the opposite. If on some crazy chance such a candidate was successful, they'd have a plausible community mandate. It's a longshot, I know. Friday (talk) 19:20, 1 September 2008 (UTC)

We have to encourage the bureaucrats to step in. Wikipedia policy is descriptive, not normative. We don't follow written policy; we make policy by actions based on consensus, which are then recorded in writing. Sooner enough an RFC will appear that shows an administrator has lost community trust. When that happens, if a bureaucrat steps in and asks for a reconfirmation RFA, if the community supports that action, then it starts to become policy. Rinse and repeat a few times, successfully, and it is policy. Jehochman ^Talk 19:23, 1 September 2008 (UTC)

So, in other words, we need a B'crat willing to do it. Who is that going to be? -GTBacchus^(talk) 19:35, 1 September 2008 (UTC)

Hopefully, at least one of them; more once it gets codified as policy, hopefully. This is why I don't like the idea of dividing powers...Celarnor ^{Talk to me} 19:49, 1 September 2008 (UTC)

So, it's just a matter of waiting for a case to come up, and then asking B'crats until one says yes? I could support that; how do we coordinate it? -GTBacchus^(talk) 20:16, 1 September 2008 (UTC)

Hint: Do it somehow differently than User_talk:Kingturtle/Archive14#If_you.27re_into_creating_policy_by_precedent.. and User:Raul654/archive14#Ryulong. Friday (talk) 20:33, 1 September 2008 (UTC)

Perhaps if there is evidence of discussion and consensus, a bureaucrat will be more likely to take a chance. Jehochman ^Talk 20:41, 1 September 2008 (UTC)

Good, but we need to stay far far away from relying on any kind of strength-of-numbers argument. History has shown that even obviously unfit admins often still have large crowds of cheerleaders. Maybe a crat-dictated reconfirmation RFA would help, if the crats were willing to disregard mindless cheerleading in favor of actual sensible arguments. Friday (talk) 21:03, 1 September 2008 (UTC)

From the examples posted by Friday, it appears that at least some 'Crats are unwilling to do something that they feel is beyond the powers they've been granted by the community. Maybe something like this would work:

There's an RfC on some controversial admin, with a wide base of support for the idea that the admin has lost the community's trust. Someone makes a statement in the RfC that it would be appropriate for a Steward to BOLDly remove that admin's bit. A lot of editors, including many admins and other trusted, long-time contributors, sign their support of that statement. The RfC is then brought to the attention of the Stewards, and... we see what happens?

I'm still not convinced that we're better off doing this than we are teaching each other to be patient and persistent (the squeaky wheel, etc), but I'm willing to follow the train of thought for a while and see where it leads. My prediction, based on my admittedly limited familiarity with Wikipedia, is that the Stewards say, "no, you've got to work this out between the community and ArbCom," and then we're back to teaching each other to be patient and persistent. I think it's possible that even the most acrimonious and drawn-out path to desysoping is beneficial to the community, in terms of what we learn. -GTBacchus^(talk) 21:56, 1 September 2008 (UTC)

This sounds worth trying. I'm growing increasingly concerned that some of arbcom are quite out-of-touch. There's a potential case right now that involves allegations of repeated misuse of the block button. A few of the arbitrators have already declined it. If we can't count on arbcom to at least look into allegations of tool misuse, it seem pretty clear we're entirely on our own trying to solve these problems. We better start trying new things- the established channels appear to be failing us. Friday (talk) 14:11, 2 September 2008 (UTC)

No, that won't work. Stewards only do what they are told. There is no judgment of consensus in their job description. Their job is to make sure they have proper authorization before changing user rights. What we need is a bureaucrat to ask the subject of an RFC to go back to RFA. Hopefully we can find a situation where the subject will agree. If that happens a few times, it becomes customary, and we can gently record the process as a policy. This needs to be created and strengthened incrementally over time. Posting a bold new policy is a quick way to get it shot down. Jehochman ^Talk 15:55, 2 September 2008 (UTC)

To what extent is a Steward's job description fixed by higher-ups, and to what extent is it determined by a wiki-like process? Also, would the same argument apply: that it's not in the B'crats job description to ask someone to go back to RfA? Are they more free to try new things than Stewards are?

I don't think that anyone is suggesting posting a bold new policy, which I'm taking to mean writing up a policy page and posting it in the project namespace. If I came across that way, then I didn't present my idea well. For the record, I think that writing up a new policy before it grows organically on its own is a terrible idea. However, if this is going to involve a B'crat doing something new, then unless one of them shows up and volunteers, someone is going to have to say something to some one of them. That's going to be the means by which it grows organically - you plant a seed. An RfC might be a good context in which to do that, or it might not... -GTBacchus^(talk) 17:27, 2 September 2008 (UTC)

I think we have our answer- Wikipedia:Bureaucrats'_noticeboard#What_happens_when_consensus_changes.3F. The crats (well, the ones who've commented so far, anyway) want an explicit mandate from the community. Maybe I'm crazy but I think a simple little proposal like that might stand a chance. Friday (talk) 17:32, 2 September 2008 (UTC)

"Requests for comment"

This is not an avenue for desysoping, it basically says correctly that if the community has expressed a loss in confidence in you that the admin may voluntarily step down. Well that is true without an RFC, and an RFC cannot compel a person to step down

The same issue exists with 1 which refers to admins willing to step down. Being open for recall is not an extra avenue of desysoping. It relies on the admin voluntarily stepping down. An admin need not step down even if open to recall and an admin may step down when not open to recall.

I think this page is misleading as it is. I would suggest that the manner in which admin rights are removed are always one of the following:

1. The user decides to give up his admin bit
2. Arbcom decides to remove the admin bit
3. A steward acting in an emergency situation decides to remove the admin bit.
4. Jimbo decides to remove the admin bit

To give the same information in another way, a user may attempt to initiate a desysoping in the following manners:

1. Convince the admin to step down
2. File for arbcom and convince them of the need to desysop the admin
3. Convince a steward that an emergency situation exists that requires desysoping(such as a compromised account or an admin gone nuts)
4. Hope Jimbo decides to do it(Please don't bug Jimbo to desysop people, he will likely refer you to arbcom)

All else is just muddying the waters. Already at least one person has been confused by this. Chillum 18:42, 2 September 2008 (UTC)

Good comments, thank you. I have made adjustments. Feel free to do more. The idea behind this page is that it should help users understand the various processes for removing sysop access. Jehochman ^Talk 20:03, 2 September 2008 (UTC)

Much better. I have also added a note that in cases of admin recall it is still a voluntary action which may be refused by the admin. Chillum 21:12, 2 September 2008 (UTC)

I made a change or two that I hope are accurate and helpful. I think we cannot emphasize enough that getting ArbCom to remove the bit - although the most common way, and possibly the easiest - is not at all trivial, and that it requires a lot of patience and persistence. There is a challenge in not seeing this fact as a weakness, but instead as a Good Thing.

Such cases keep pushing the community closer to an understanding that certain maturity and "people skills" are necessary in our admins, because look at what a mess it is when immature or antisocial editors are given the bit, and then proceed (unsurprisingly) to behave in an immature or antisocial way, which absolutely damages the project. If this is a lesson we have to learn, as a community, then these drawn-out drama fests seem to be what it takes to learn it. It's years of Internet culture, going back to Usenet, that we're trying to alter. It won't happen in a day. -GTBacchus^(talk) 23:26, 2 September 2008 (UTC)

Redirects

Currently there is considerable similarity between redirects to this page and redirects to the much older page Requests for de-adminship: to this page; to the other page. This should be sorted out somehow--I don't know or necessarily care how, except that I think requests for de-adminship, having been around a lot longer, has a lot more traction in the community as a whole, so I would not support all redirects simply pointing here instead. Possibly the two should be merged anyway, since they cover similar ground. In any case, I've added a dab link. Chick Bowen 00:21, 22 September 2008 (UTC)

Administrators open to recall

This is listed as the second "Established processes for removal of administrator access." Has an admin ever lost his/her rights through open recall? As far as I know, admins have either resigned or been desysopped by ArbCom, Jimbo or similar means. Kingturtle (talk) 19:05, 18 May 2009 (UTC)

I seem to remember this happening before. It was a drama fest that came to a mob mentality result if I remember correct. Chillum 19:10, 18 May 2009 (UTC)

I think it is this list (note: list is not automatically updated and may not be up to date) OhanaUnited^{Talk page} 06:53, 15 August 2012 (UTC)