Wikipedia talk:Admin Accountability Alliance

Essay?
I've provisionally tagged this as an essay because it seems to be given over largely to the expression of personal opinion. --Tony Sidaway 18:43, 13 December 2007 (UTC)

Community enforced administrator recall
I wonder if the above proposal could be merged in with this? It's a formal procedure of admin recall, but allows it to be voluntary in nature. The recall procedure is fair, it needs 6 people to endorse a recall request, followed by a community vote, with 75% in favour of a recall before the admin loses his bit. The numbers can be adjusted slightly, but I think the two proposals could work well together.  Ry an P os tl et hw ai te  20:21, 13 December 2007 (UTC)
 * I think I missed how this is a voluntary procedure. Could you clarify that? Thanks. ++Lar: t/c 21:23, 13 December 2007 (UTC)
 * It isn't at present, but I was suggesting it be merged in with this proposal. It's voluntary, because admins opt into it and if they agree to it and the procedure begins, then they must carry out the recall procedure. If you don't opt in, you aren't subject to it.  Ry an P os tl et hw ai te  00:26, 14 December 2007 (UTC)
 * I'm not thrilled with that proposal. Any admin who hasn't made at least six enemies probably isn't doing his (her) job. But given that it's voluntary, if someone wants to sign on why not. Raymond Arritt (talk) 04:19, 14 December 2007 (UTC)

OK thanks for the clarification. Ryan, why don't you try editing this proposal to incorporate the features of that one that you think have merit. Who knows, maybe your changes will stick. ++Lar: t/c 11:57, 14 December 2007 (UTC)

The history of recalls suggests that a majority of the requests have been turned down in this fashion
That's an assertion that I do not at all agree with, and I suspect that on close examination many others will not either. I've analysed the 11 recalls I am aware of already to specifically refute it, here:. I would strongly suggest that it is a divisive statement, casting bad faith aspersions on wide swaths of people. Such a statement ought to be struck from any policy proposal. Don't take that as opposition for developing this process further but if you want my support you are going to need to remove all of the aspersions you cast in the runup to the policy. Try positioning this as an enhancement to a policy that works, and works well but could stand tweaking instead. ++Lar: t/c 20:58, 13 December 2007 (UTC)
 * We don't have a policy that works. That's the point. We have a nebulous process where everybody can set their own standards, and people can (and do!) simply remove themselves from the process when it gets inconvenient for them (which, for some reason, is called an "unqualified success" for the process...)  &gt; R a d i a n t &lt;  00:25, 14 December 2007 (UTC)
 * The view that the current process (not policy) doesn't work is not universally held. I think a revert like this of good faith changes made to try to gain the idea wider acceptance is not really a good approach and I'm a bit disappointed that you wholesale reverted me. You have the advantage of me now, because I don't revert things back...  ++Lar: t/c 00:41, 14 December 2007 (UTC)


 * I've made some edits in that direction. In general, I just gave the entire proposal a going over to remove assertions that are opinion and to acknowledge there are differences of opinion without changing the proposal itself or the mechanics. My concern about asking stewards to do things remains. ++Lar: t/c 21:21, 13 December 2007 (UTC)
 * No, what you did is reworded everything into political correctness to make it entirely meaningless. Okay, it's clear, you don't like the idea, so you're free to not join. I suspect that in fact very few admins will dare to put their "mop" in the hands of anybody other than themselves and the arbcom, but feel free to prove me wrong on that point.  &gt; R a d i a n t &lt;  00:24, 14 December 2007 (UTC)
 * I don't think it's a valid conclusion that I don't like it and I have no idea how you arrived at that conclusion. If I didn't like it I would not be trying to work with you to try to make the proposal better. ++Lar: t/c 00:43, 14 December 2007 (UTC)

Steward discretion needed
Also, I am struck by how similar to the existing system this is. But the differences are ones that I'm not sure are workable ones. The existing system relies on the force of public opinion, which, in the final analysis, is how policy gets done here, not by fiat. So it may be a distinction without a difference. This policy would also need buyin from a lot of people to make it as official a process as deadminship by ArbCom before stewards would honor it (stewards act on consensus, never on their own... a campaign slogan). The existing process only requires (via the force of public opinion, no other enforcement teeth are needed) that the admin voluntarily ask for de-bitting themselves, which stewards always honor. Worst I fear that you would be asking others to judge how a process went and decide if the admin adhered "closely enough". If I am elected I will never ever do that. It is completely outside the Steward's remit to do that. ++Lar: t/c 20:39, 13 December 2007 (UTC)
 * This is never intended as an "official process". It can't be that, since it's voluntary. The existing recall system relies only on the admin himself, which means it by definition cannot hold anybody accountable for anything. So to create accountability, you need to go for the arbcom, or think of something new.  &gt; R a d i a n t &lt;  00:27, 14 December 2007 (UTC)
 * Stewards cannot be forced to enforce anything that is not official. You may want to consult with a few (more) stewards on that. Therefore if the process is unofficial, a different enforcement mechanism is needed. ++Lar: t/c 00:45, 14 December 2007 (UTC)
 * Being picky I would suggest that it would be hard to create any form of enforcement for something unofficial? You cannot enforce something that is voluntary - I join something voluntarily, I leave it in the same way, maybe it's just me -- Herby  talk thyme 12:13, 14 December 2007 (UTC)
 * Exactly, thanks. If ArbCom embraced stare decisis then a test case in which some admin that said they'd abide by this, and then did not after it went against them, and ArbCom said "we view this as the 'will of the community', you signed up for this process, so you lose your bit because that was the process outcome", might give it the officialness needed. However, ArbCom explicitly does not embrace stare decisis and I believe instead has indicated it mirrors perceived policy the community arrived at, rather than sets it. So some more clever thinking is needed to find a way to enforce a voluntary thing, if the court of public opinion is insufficiently toothy. I remain convinced sufficient cleverness is out there. ++Lar: t/c 15:15, 14 December 2007 (UTC)
 * Stewards are not there to make decisions for the big projects like enwiki. Bad idea. If you need an enforcement mechanism, have ArbCom make a simple factual finding that the requirements necessary for desysopping have occured, and let them make the request to the stewards to alter the user's rights.  NoSeptember  21:23, 14 December 2007 (UTC)
 * This is a good idea. It requires ArbCom to hold that the outcome of this process represents the desire of the participants. That's different than forcing the process to be "official". If ArbCom can be persuaded to do this, it would be ideal as it neatly circumvents requiring stewards to make decisions, which, as I have been pointing out, they, by policy, do not do. ++Lar: t/c 21:35, 14 December 2007 (UTC)

The editors that are "eligible" to make such a request are the other people on the list.
That one I very much support the notion of. See user:Lar/Accountability where I already do this for admins. But it leaves non admins out of the process which is not fair to non admins. So I have (in my criteria given there) made provisions for that as well. Mainspace edits will qualify non admins. That seems fairer than exclusion entirely. ++Lar: t/c 20:39, 13 December 2007 (UTC)

I was asked by Radiant to comment
I don't feel overly strongly because however it's worded I probably wouldn't sign up personally, but I do worry that it may make admins who are called upon to make tough choices or get involved in insoluable disputes to pussy-foot around rather than make tougher choices, if they know that they are bound to step down if N editors on one or the other side of the dispute are sufficiently displeased with the admin's action or inaction that they recall the admin. Now it could be said that there will be three types of admins: those not open to recall; those open to recall but not open to acting in a particularly contentious arena; those admins soon to be on their way out. We keep referring to ourselves as janitors with mops, but our functions are more than that. Some editors view us as police with batons, but our functions are more than that. Some seek out our help in doing janitorial tasks (tagging crap for speedy deletion; making moves over redirects, reverting lots of vandalism quickly with rollback instead of undo, etc.). Others seek out our help in doing policing tasks (WP:AIV, WP:RPP, WP:ANI, blocking, etc.). I really don't have much knowledge on how one dismisses a janitor or maid - I assume one just gets displeased and sends them packing. However, most places with police forces have a review board so that civilian complaints can be raised and assessed before firing a cop. Why? Because an accusation is cheap and easy and a false accusation has little to no consequence to the accuser. Here, we have such a review board: ArbCom. Now, it isn't bound by all the formalities of litigation and procedure and rights, and it isn't perfect, but it does a decent job of de-sysopping bad admins. Indeed, many admins resign after the ArbCom decides to take a case to look in to whether the admin's conduct was egregious enough to be de-sysopped. Sorry for the wordiness of my comments, butI worry about all new admins being cajoled in the RFA process to accede to this new regime of easy-out making them too timid to be fully useful. All admins make mistakes and a single mistake that makes N people mad earns a de-sysop from anyone who volunteers to that standard will neither make the admin a more careful one nor a better one without simultaneously making them a meek one and a play-it-safe one. And one last comment on the details of the plan (if I have sufferage to tweak a plan I doubt I'll buy into): make sure that it isn't N editors, make it N screw-ups; alternatively, make the N editors all within a short period so no one can game the system by seeing which admins have N-1 editors on their case just to exploit that: akin to basketball teams knowing who is one foul away from fouling out or soccer where you know who has been given a yellow card. Carlossuarez46 (talk) 01:45, 14 December 2007 (UTC)
 * Your points are well taken. I do tough stuff ( or so I would like to think :) ) and under the current CAT:AOTR I am comfortable that I would not be vulnerable to recall by a small group of editors who are annoyed with me for POV reasons, or a small group of trolling editors coming in from one of those sites or suchlike. That's because after a recall were certified (and I have pretty tight criteria for who can petition, the articlespace requirements are designed to keep out trolls) the next step doesn't have to be an RfA. This proposal requires it. For that reason I'd prefer that the number of people who need to certify either be raised significantly (5 is too low if there isn't another step) or be allowed to be variable based on the admin's discretion (set in advance and then adhered to, to be sure, no changing it at the first sign of heavy weather). As I've said elsewhere, properly set up, I could see people wanting to be in both categories/lists/thingies/whatevers. The problem with N screwups is that now becomes an evaluation thing... what is a screwup? What is very nice about this proposal is that for the most part it removes evaluation/wiggle room, addressing the concerns of those who feel there is too much of that in the other scheme. (it also disenfranchises non admins as currently written but that's a side effect that maybe has to be lived with, or maybe not) ++Lar: t/c 11:44, 14 December 2007 (UTC)
 * I still question my sufferage given that I won't be on it, but I would hate to see the admins who do join up get tripped up and deadmined, thus depriving the community of some of its most valuable contributors. The other thing is that there is no qualitative weighing of the community, it's just the number of petitioners, as though 5 (or some N) can trump consensus by banding together to do so. No community can long survive a regime like that without becoming homogeneous over time - and given the track record that most admins who are at risk of desysopping leave the project altogether, we would lose many more people and where we strive to not bite newbies, why should the regulars be bitten - even voluntarily? I am all for freedom of personal decision - want to play blindfolded football on the freeway at night, go ahead - but just know what you may be getting into if this really gets enforced on those who volunteer in good faith and get ousted by others in less than good faith just because some N of them say so. Carlossuarez46 (talk) 02:11, 15 December 2007 (UTC)

I share Carlossuarez46's concerns with this proposal. If you're an admin here long enough, sooner or later you're going to step into a messy dispute. That's part of an admin's job - to be the "older, wiser head" and bring some perspective (and a lot of policy knowledge) to the dispute. To fulfill your admin responsibilities and to do what's right for the project, you're going to take some action that will set off the editors involved. Much as we think of ourselves as janitors, at times we are mediators. Any policeman will tell you that of all the dangerous things they do, they are most likely to get hurt on a domestic disturbance call. It's a thankless, miserable part of the job and is more likely to get all the participants angry at you even as you do what's right for the people involved.

The core problem with this proposal is that it requires the admin in question to agree to accept the judgment of the people making the complaint without any prior assertion that they will be impartial or fair in their presentation of the complaint or that they be knowledgable about the policies that you may have been enforcing. No matter where you set the bar, it will be possible for the partisans to game it and deliberately abuse the system.

Now, you might be able to mitigate that concern if there were adverse consequences for a frivilous or abusive complaint. But then who decides that a complaint was frivilous? The proposal rightly says that the current process of letting the accused admin determine whether the complaint is reasonable is a conflict of interest. So it has to be someone else. Who? Are you going to establish the equivalent of the police review board that Carlos mentioned above? (And, by the way, once that someone has decided that a complaint was frivilous, how will any sanctions be enforced? What will stop an abusive nominator from simply creating a new account, free from whatever consequences you tried to impose?)  Without answers to those concerns, the "adverse consequences for a frivilous complaint" fail as a mitigating control.

There are two models that I see that might be effective. One is the same one that Carlos mentioned above - some equivalent to a police review board. The ArbCom largely fills that role.

The other is a "jury of your peers". If the policy set a standard where the accused felt that he/she could reasonably trust the judgment of the people making the nomination, voluntary recall might be a viable alternative. Unfortunately, the only threshold of trust that we have at Wikipedia is the granting of adminship. I suspect that you could get more people to sign up if the policy required some quorum of other admins to determine that de-admining was appropriate. The problem with that approach is that it reinforces the impression that adminship is a club. Some (many?) non-admins will be skeptical of the willingness of an admin to comment harshly about another admin. Still, I think with that modification the policy might gain support.

Without something like that, I have grave concerns that even on a voluntary sign-up basis, this will make it harder for us to keep the good admins. Rossami (talk) 02:22, 11 January 2008 (UTC)

You know what, not a bad idea
At first when I read this, I outright dismissed it once i read "they make the solemn promise to step down given sufficient complaints about them, even if they are not convinced themselves", because if an admin isn't being complained about somewhere in the world, they're not doing a good job. But then I read who comprise the eligible editors. While I think the general community should air their grievances (what's weird is that I just finished watching the episode of Seinfeld in which they celebrate Festivus), I like the idea of the final judgment power being given to the others in the category. I think it could work. J- ſtan TalkContribs 02:34, 14 December 2007 (UTC)
 * I too think it has merit. There is one major flaw regarding enforcement (as I outlined) above that needs to have a clever solution, and the wording of the justification ought to read less polemically, but it is a novel proposal that is worth exploring and fleshing out. There is no reason it cannot coexist with CAT:AOTR, with specific admins in neither (list/category/thingie/whatever), one or the other, or both. ++Lar: t/c 04:08, 14 December 2007 (UTC)
 * Should it instead be "Those editors who have been on this list for at least 2 weeks"? Otherwise, if I don't like Lar and only 4 list-members have said he should step down, I could join, demand that Lar step down, and then leave the list. "Drive by deadministrations" as it were. – Quadell (talk) (random) 20:59, 14 December 2007 (UTC)
 * "Drive by deadministrations". I like that. Maybe not two weeks, but maybe four days, as with the autoconfirmation process. But it probably should be changed. J- ſtan TalkContribs 00:24, 15 December 2007 (UTC)
 * Go ahead, Quadell-san... make my day! You have to ask yourself, do I already have 6 admins lined up to pull the trigger on you as soon as you join... or was it only 5? Do you feel lucky? well, do ya?  :) Seriously, I hate to keep trotting out my own criteria in the current system, but there is a time requirement I put in there, you can't just pop in, petition, and leave again. ++Lar: t/c 02:36, 15 December 2007 (UTC)
 * No, it's cool that you have something similar that could be used to tune this proposal. I do like the time limit system. If we really wanted to do this right, have the time limit be as long as a standard recall case. There couldn't be any Drive-bys (or Clints :) ) J- ſtan TalkContribs 02:49, 15 December 2007 (UTC)

RFC
I like this idea; but in addition, if someone were to open an RFC about me and suggests that I step down as an administrator, and if there were consensus on that page (as judged by a bcrat) that I should step down, then I would. Is there a problem with this solution? Should this be mandatory for admins? – Quadell (talk) (random) 20:53, 14 December 2007 (UTC)

Potential problem: the number 5
There are currently around 1000 active admins. Let's imagine this becomes quite popular, and half of all active admins join up. Now let's say there's a Big Divisive Issue, where many admins take a position that is strongly opposed by many other admins. "Allegations of Apartheid" articles; deletion non-free images; ads; blocking based on non-public information; whether NPOV = SPOV; use your imagination. So ten of the hard-core members of Faction One decide to ask for the deadminship of a dozen or so leading members of Faction Two, many of whom had performed questionable acts, but which wouldn't be a huge problem to a non-partisan. These dozen from F2 counter-deadmin the ten from F1, and so do their friends, who are offended at the attack. Repercussions ensue. It's mentioned on AN/I, and some F2 members ask for the deadmining of even those who voiced support of the F1 faction's actions. Uninvolved admins ask for the deadmining of that person, simply because he "abused the AAA system", and some F2 members deadmin them. Normally, these things would die down rather quickly, but if all it takes is 5 more admins, out of 500, there could be a remarkable amount of bloodshed before it ends. 5 is a big enough number when there are only 50 AAA members that it should be manageable, but it doesn't scale.

So should it be 10%? Or "more supporting the deadmining than opposing"? What do you think? – Quadell (talk) (random) 21:17, 14 December 2007 (UTC)


 * I just read the proposal, and came here to post about almost precisely the same problem which Quadell has so described very well.
 * However, I don't think that quadell's nightmare scenario of bloodletting started by big factions is the only possibility or even the most likely. The more likely outcome of the 5-black-balls-and-you're-out rule would probably initially be something more akin to guerilla war: admin A hates admin B, so A emails 5 friends to evict B. 5 of B's friends evict one of A's team, and so we begin a cycle of tit-for-tat killings.
 * Faced with this, a group of uninvolved admins soon decide it's time to protect themselves, and form The Alliance: 100 admins who make an all-for-one-and-one-for-all vow, a "you evict any of us and we'll get all of you" grouping In response 200 other admins form the "Self-Defence league", declaring their hated of the Alliance, and vowing to decapitate ten members of the Alliance for every Self-Defencer evicted. Within the week, Alliance member admin X is de-adminned by his own side as a suspected double-agent, and after a leak inquiry the Self-Defence League decapitates five of its own people whose innocence cannot be proven beyond doubt. Ten uninvolved admins who plead for sanity are themselves de-adminned ... and by the end of the next week there are only 5 admins left.
 * Any such system needs to be much more robust than suggested. Why not simply have something closer to the RFA system, and allow any admin to be recalled by a supermajority of admins in a vote? --BrownHairedGirl (talk) • (contribs) 04:09, 16 December 2007 (UTC)
 * Pretty cool scenario though you forgot the obligatory mention of "For great justice." I like the idea of a supermajority vote for de-adminning; it would probably go a lot faster than arbcom, which can be agonizing. Raymond Arritt (talk) 04:20, 16 December 2007 (UTC)

What counts
Hello, I'm wondering what happens with those who are not eligible to request. Non admins for example will clearly have no say here, as they would not be on the list. This, to me makes the adminship position seem extremely elite. I'd hope never to have to request someone step down, but as I'm not an admin, with this proposal, I'd not even have the choice. I think that it should be "editors in good standing". Of course, admins are preferable, and if the admin being asked to step down disagrees that a particular non admin is not in good standing, then if other editors disagree the "vote" should still count. Still, even better would be a minimum amount of edits/time etc. Thanks.  Red rocket  boy  21:19, 14 December 2007 (UTC)


 * I can see your point that "the community" more broadly should have a say. This has to be balanced against the fact that on a project this size it wouldn't be too hard to assemble 6 (or 10, whatever) people acting as a gang to stir up trouble against admins who were only doing their job. If someone is really screwing up it's probably going to be obvious to at least 6 out of our 1400 admins. Raymond Arritt (talk) 21:24, 14 December 2007 (UTC)


 * Non-admins have nothing to lose. (They can't lose administratorship.) So if non-admins had a say, we would be in the odd position that non-admins would have an advantage over admins. (A non-admin could join the AAA at no risk, but an admin could only join if she was willing to put her chips on the table, as it were.) – Quadell (talk) (random) 21:36, 14 December 2007 (UTC)


 * I think that limiting suffrage to only admins who choose to be in this "alliance" is basically putting the opinions of a small group of admins over that of the community. 1 != 2  21:44, 14 December 2007 (UTC)


 * However, other than the suffrage requirements this seems like one of the better accountability proposals. 1 != 2  21:47, 14 December 2007 (UTC)
 * Hmm, let's see... What about suffrage for any admin, not just the ones signed up here? Raymond Arritt (talk) 02:44, 15 December 2007 (UTC)
 * Agree with Raymond here. I'm a non-admin and I also think 1400 admins are enough to be trusted with this responsibility. Everything else will just create problems. In real life as well there are voting restrictions for everything. Its not like in every vote, any person can participate. There are requirements everywhere. Good standing is hard to determine anyway, no set criteria there. --Matt57 (talk•contribs) 00:16, 15 December 2007 (UTC)

A few issues that aren't addressed
Here are some issues that, in my opinion aren't addressed:


 * How do we ensure that all these discussions have the necessary amount of visibility? Certain examples spring in mind when people propose Yet Another User Conduct Noticeboard that ends up being read by less people than it absolutely needs for all-around fair treatment...
 * Is calling out demonstrable foul play allowed by the admins in question? "Step down!" "Says ten obvious sockpuppets." "No, step down! You can't complain!"... After all, that's one form of stating that you're not convinced =)
 * Are there specific requirements on what exactly is a good consensus in this case? Now it just says that there merely has to be a significant number of users to demand action. There should be some provisions in this that define exactly how to determine the consensus; I'd hate to see this turn into "Votes for Desysopping" or "We, the Undersigned Legion of Meatpuppets, Have a Non-Complainable Demand for the Admin in Question to Waste Some More Time in RfA... Sixth Time This Month."
 * In worst case it has a big potential trouble spot: Why initiate discussion with an admin, if they can't ultimately complain about it? Who is allowed to defend them? If so, doesn't this get a bit confusing? Again?
 * I'm not sure how this solves any problems that other DR discussions/RfARBs already do. At least those processes are tried-and-true and have some sort of actual influence. If there's any level of voluntarity involved (in this case, admins have to volunteer for this thing) I don't know if it will work at all. This policy says that its idea is to combat the "empty promise", yet, it merely trades it for another form of empty promise: Nothing demands the admins to join this thing, nor is here anything that demands the admins to stay in once they've joined. Exactly same thing as before, just with one goal changed. What's the point, if I may ask?

Don't get me wrong, the idea is generally speaking good. I just feel that this idea in current form leaves too many loopholes in all ways, and adds yet another process – this sort of things have not flied all that well in past, so we should learn from the mistakes. We have tons of these processes as is and adding more processes makes things even more murky. We have big processes and I don't know if adding yet another partially redundant process for special purposes does any good. One way I could see this work would be that this thing would be somehow meshed into existing best practices and policies. Currently, it stands out as yet another gnarl in our already twisted and gnarled policy tree. It kind of says to me "If you want to address a problem, don't go to the court if you can just pass leaflets to people on the street, thereby forcing someone else to go to the court and waste their time and money."

I'm sorry to paint bleak pictures, but I feel that possible bad sides of the proposals have to be brought to light =( --wwwwolf (barks/growls) 12:55, 16 December 2007 (UTC)

Contradiction
I do not know how to resolve this, except having people be more willing to confront troublesome admins. In practice complaints on AN/I usually do get a fair response. There is also a side channel: most admins will if they think appropriate comment privately to each other. Almost everyone wants to stay in the good opinion of their colleagues. DGG (talk) 04:53, 18 December 2007 (UTC)
 * 1) Experience shows that security of tenure is necessary for freedom from political influence. Any admin active in controversial areas will always have at least 5 enemies. The better they are at it, the more.
 * 2) There are many poor quality admins keeping just below the radar who ought to be removed or reconfirmed. The existing arb com structure can only deal with those who make drastic violations of accepted procedure.

Wikilawyering
OK, so I'm an admin. It is, however, regardless of this fact that I don't like this idea - not insomuch as it is part of the controversial history of trying to make an effective process for removing adminship, but rather in that it ignores the principal method of resolving disputes on Wikipedia. It has been my experience that the most accepted method of questioning some edit, action, or message is to approach the person for comment, and to seek outside mediation only once such an approach has proved fruitless. I know that as an admin, if someone comes to me with a comment about what I've done, I will try to address that to the best of my ability because it is my responsibility to be accountable for my actions. If I have some reason to suggest that a matter needs no further action or that I will not take such actions, I will say why. It is not so much that admins are necessarily evil or authoritarian, but that sometimes admins appear to not be accountable for actions that they have took, and refuse to address it adequately.

I would propose that any policy or guideline or even essay like this must take into account that a process like this should be created in such a way that it is much more difficult to try to desysop a user in the case where they are open to discussion about an action they have made. In particular, I think that it's unreasonable to take one controversial action as a reason for desysopping - if it isn't clear that what something someone did was unacceptable, then it is for the community to decide whether the action, not the user, was problematic. Granted, there will be cases where users who are administrators are acting improperly using the mop and are actually abusing it (the Alkivar ArbCom case, as closed, comes to mind) - but this should not be a reason for us to potentially alienate good users who might have made a couple of bad decisions but are willing to realize that they could have done better and make an honest effort to improve, or who are being slammed for actions that are controversial in a case where they might just have been slightly bold.

I like the idea of admins being accountable for their actions (I certainly hope people find I am :) ), but all too often proposals don't leave adequate room in for avoiding drama. Nihiltres { t .l } 17:30, 18 December 2007 (UTC)


 * In fact, I find it hard to imagine a proposal that would go further to increase it. I move it be marked rejected. DGG (talk) 04:35, 19 December 2007 (UTC)

Stewards
You want the stewards to enforce something, but "this is not going to be policy". WTF? &mdash;Ashley Y 05:08, 21 December 2007 (UTC)


 * I doubt the stewards would ever respect a process where only members can vote, and the rest of the community can not. So in the end this is just as unenforceable as the current recall process. It relies on the admin to choose to step down, like any other admin.  1 != 2  14:53, 21 December 2007 (UTC)
 * Well, yeah, but I think the point is that the admin will step down out of his own decision, and nothing will have to be enforced. J- ſtan ContribsUser page 17:15, 22 December 2007 (UTC)
 * That's what we have now then... any admin who wanted to adopt the spirit of this idea could thus join CAT:AOTR and list their recall petitioner criteria, recall process and resignation/next step process to work the way this proposal outlines. (and then hopefully list their criteria at Administrators open to recall/Admin criteria so we could all learn from them. By the way, as a new steward I would be happy to ask my fellow stewards whether what I was asserting before (about whether stewards would consider this binding or not) was correct or not... ++Lar: t/c 01:35, 25 December 2007 (UTC)


 * The relevant case is when some admin commits to some criteria and then vigorously disputes that those criteria have been met and/or attempts to revoke the commitment. Furthermore, there will be no policy and no arb decision supporting revocation. Will the stewards nevertheless revoke the admin right of someone who will contest their action? &mdash;Ashley Y 06:43, 25 December 2007 (UTC)
 * Nope. At least that's my interpretation of policy, and as I offered, I could raise this with the other stewards to confirm it. The point is that this new proposal is no different than CAT:AOTR as regards enforcement, unless, as I suggested above, some way to make this official, yet unofficial... voluntary, yet binding... could be developed. (I was responding to J-stan's suggesting it would work voluntarily...) ++Lar: t/c 16:23, 25 December 2007 (UTC)

That's what I thought. I've edited the page to mention this. &mdash;Ashley Y 00:38, 26 December 2007 (UTC)

Community decision
The promise comes into play if at least five eligible editors take sufficient issue with said admin that they request him or her to step down, on said talk page. At this point, the admin must either (1) seek immediate reconfirmation at requests for adminship, with the understanding that adminship will be revoked should the RFA fail; or (2) take up to one week to discuss the issue, clear things up, and make amends as desired; if, after one week, there are still five eligible editors that want the admin to resign, his or her adminship will be revoked by one of the Stewards. In either case, the admin may reapply at requests for adminship at any later time. 

I do not think this is in accordance with Steward policies on meta (which, regarding the stewards, overrule enwiki policies, and most surely unofficial ones, imho). ("Their task is to implement valid community decisions." (...) "The only exceptions are in emergency cases where no local user with that right is available, or for projects that demonstrably have no community.")

A request by five people (unless we are speaking of a committee, which is a totally different matter) does not seem to me like a community decision at all, and only the vote part (if allowed by local policies) would be in accordance imho. So even if this would become an official policy, I would have serious doubts about implementing it (being a steward myself).

I see a possibility with a community vote, but that would require probably (if I understand enwiki's forest of policies well enough) a change of policies, so that also the community would have the ability to desysop admins. Which is, by the way, a quite regular option in other wiki's. However, it would require careful caution with setting up those policies, as they are sensitive for trolling. effeietsanders 09:50, 27 December 2007 (UTC)


 * Shall we mark this "historical", then? &mdash;Ashley Y 10:23, 1 January 2008 (UTC)

A Way This Could Work
Currently this probosal is too similar to admins open to recall, and other similar ideas which havn't really changed much. The Problem as this proposal notes is that admins are only responsible to themselves. As such convincing an admin to step down is often futile. If however, the admins who volintarally join this idea, select another admin who would have to be convinved of a need to step down instead of themselves, this would work a lot better, as a third party (even if not nesicarally impartial) would be making the decision. --T-rex 04:53, 8 January 2008 (UTC)
 * You forget a main issue: What you propose might not be compatible with the Stewards policy, which tells stewards what is a valid decision. I actually doubt that a decision like you propose it would be compatible enough in the sense that a steward would act without a community consensus nor arbcom decision. (And if the local policies do not allow a vote on this subject, I doubt that this vote could be seen as a valid community decision. However, I have to say that I am for reconfirmations and stuff, but I think the only way to go is either totally voluntarily, either just plain obligatory, and keep it for every one the same. if you do the latter, you should make damn sure the procedure is troll-proof btw. effeietsanders 22:33, 8 January 2008 (UTC)
 * It doesn't matter what stewards policy is or anything. It's about admins agreeing to have somebody else make a rational choice for them. No need to even bother the stewards in this instance --T-rex 23:16, 8 January 2008 (UTC)
 * No, I guess you'll remove the rights yourself hae? ;-) effeietsanders 00:08, 12 January 2008 (UTC)

Historical
I've marked this as historical, because it seems to have died. &mdash;Ashley Y 07:04, 16 January 2008 (UTC)
 * The proposal author was gone for over a month but was back recently. Perhaps he could be nudged. But it's not enforceable by stewards in its present form which makes it not workable, in my view. ++Lar: t/c 07:11, 20 January 2008 (UTC)