Wikipedia talk:Banning policy

From Wikipedia, the free encyclopedia

Shortcut:
WT:BAN
Archive
Archives
Archive 1 Archive 2


Contents

[edit] Wikipedia:Requests for remedies - possible solution to dispute resolution scaling problems

Please review and tweak: Wikipedia:Requests for remedies. A very simple three-step system that can make trusted, final decisions on very tricky or complex matters, based on evaluations from trusted, uninvolved users on a given case in the dispute resolution process. It does add new process, but not many layers, or particularly complex layers by any stretch of the imagination. It's built entirely around consensus and the idea of certification, and is the opposite of Votes For Banning. Please weigh in at Wikipedia talk:Requests for remedies. The community needs a way to move forward in a trusted, fair manner on high-end, complex problems that are either unworkable for normal WP:AN, WP:ANI, or WP:RFC to handle, or that the Arbitration Committee can't take on, or that the Arbitration Committee relegates back to the Community. Lawrence § t/e 22:15, 14 March 2008 (UTC)

[edit] Overly bureaucratic

My understanding was that a community ban was simply an indefinite block that is not going to be overturned. The "banning mechanism" need be no more complicated than adding {{banned}} to someone's userpage once they've been blocked indefinitely ([1], [2]). Or am I doing something wrong? Moreschi (talk) 10:52, 3 April 2008 (UTC)

I'm not an admin, I'm just an observer but as I read the {{banned}} documentation, adding that template will automatically add the user to the Category Banned Wikipedia Users. AFAIK it will not add the user to the Wikipedia:List of banned users, so that should probably be done manually. Once there's consensus, I think those two steps would do it. --SteveMcCluskey (talk) 13:11, 3 May 2008 (UTC)

[edit] Confusion

Hi, I've been reviewing several dispute resolution procedures lately, as part of my involvement with the working group, but I've gotta be honest here, this policy still confuses me. I've read it several times, I've looked at discussions of banned users, I've reviewed the template, and I still don't get it. It's not that I disagree with it, it just seems that "banning" is something that's done by those admins who understand what it is, and the rest kind of ignore it (maybe because they're confused too?). I'd help rewrite this page, except like I said, I don't understand it, so I'm not clear how to even improve it.

I'm perfectly willing to continue ignoring this page, but just in case anyone would be interested in further clarifying things, I would recommend:

  • Go into more detail on the difference between a ban and a block
  • Explain how to go about banning someone. What steps need to occur first? What are common problems to avoid?
  • When is a ban more appropriate than a block?
  • When is a block more appropriate than a ban?
  • Describe different types of bans. Some seem to be blocks, some seem to be "topic bans" or "article bans"? What minimum authority is required to implement such a thing? Do all bans require a "community consensus" discussion, or can some admins just go in and implement them? What are some typical examples?

Thanks, Elonka 06:00, 22 April 2008 (UTC)

Elonka, this policy exists for those admins who view the project not as building an encyclopedia, but as a MMORPG. You can easily identify these admins by looking at their contributions list and seeing mostly edits to project space and block logs several pages long. WP:BANNED includes scores of users who have "de facto" community bans not discussed anywhere on-wiki. 198.203.177.177 (talk) 13:25, 22 April 2008 (UTC)
Elonka raises serious questions that deserve a serious answer, not sarcasm and ad hominem attacks. I agree the banning policy has significant gaps. She outlines most of them quite well, although not quite with the priority I'd give them: a policy page should be understandable by everyone, not merely by the administrators who implement it. DurovaCharge! 14:41, 22 April 2008 (UTC)
Thanks Durova. As a case study, I could bring up an editor that both you and I are familiar with, PHG (talk · contribs), whose behavior led to Wikipedia:Requests for arbitration/Franco-Mongol alliance. Even with a unanimous decision by the arbs, it's still been an ongoing situation for several months now. Would a ban have been an appropriate option at some point? Or if not, why not? And if it was, which "precursor" steps should have been done, that were missed? Is a ban still an option now, or does this come off the table after the arbs are involved? Sorry if these appear to be dumb questions, but I'm still trying to understand this policy. Thanks, Elonka 10:55, 23 April 2008 (UTC)

[edit] Edits by banned users

Proposed text to add: "This includes reverting to bad spellings, bad grammar, and punctuation errors. Stopping a banned editor is more important than the quality of the encyclopedia." It's important to clarify this, that even if a banned editor comes back and reverts vandalism, it's better to revert back to the vandalism than to allow the evil banned user to edit. 198.203.177.177 (talk) 13:15, 22 April 2008 (UTC)

Indeed. Seriously, the anon has a point. The current wording in Wikipedia:Banning_policy#Enforcement_by_reverting_edits section is contrary to our goal of building an encyclopedia, as well as to related policies that tell us to concentrate on edits, not editors. If an established editor makes a mistake, he can be reverted; and if a banned editor makes a good edit he should not be reverted just because he is banned. Rather, we should assume good faith and even consider whether his ban is best for the project. Some time ago a banned editor asked me to correct a spelling error in an article, should I have just ignored him? Recently I saw a bunch of innocent and useful edits (for example, interlinks) reverted just because they were done by a sockpuppet of a banned editor, and this policy was cited as an excuse. As far as I am concerned such an approach is wrong and harmful to our goal of building an encyclopedia, and I suggest rewriting the relevant part of the policy. --Piotr Konieczny aka Prokonsul Piotrus| talk 12:03, 27 May 2008 (UTC)

I believe that the idea behind "revert banned editors on sight" is to simply make it clear and final that they are banned from the project. They are not just banned from making bad edits, they are banned, period. Letting banned users make "good" edits leads to all sorts of questions about what edits are good, what edits are controversial, and so on. It also opens the door for them to gradually evade the ban, after a month of good edits their behavior becomes gradually more controversial, and when someone wants to put the foot down and block the sockpuppet, there will be editors exclaiming "but s/he has done good work here!", and step up to defend the person even if they're banned or been engaging in other clearly abusive practices (we saw that in the Jack Merridew case. [3] [4])There have been enough cases of using a sockpuppet to gradually make more and more controversial edits as a backdoor into Wikipedia, that just saying no to their edits might be the best.

Onto practical consequences, first are you really sure that the "good edit" by the banned user is really good, and not simply well-disguised propaganda, POV-pushing, harassment, or similar? Second, there is no rule which says that you cannot make the edit. If you find an edit from a banned user, it should be reverted, but if you think the edit was good, you can then reinstate it, but you will be taking responsibility for that edit as well, so only do this if you are sure this edit is good. The fact that you reverted and reinstated has the additional benefit of showing to everyone that the user's edit has been dealt with. Sjakkalle (Check!) 13:53, 27 May 2008 (UTC)

I agree we should block socks, ips and any way the banned users try to circumvine their ban. I also have no problem with 'revert and reinstate' good edits to certify they are good, but I have a problem with 'revert good edits just because they were made by a banned user'. Remember: our goal is to build an encyclopedia, not to police the users. The second is necessary, as in any community, but the tool should not become - or overshadow - the goal itself. --Piotr Konieczny aka Prokonsul Piotrus| talk 14:57, 27 May 2008 (UTC)
Um, does that mean we agree? Because if you have no problem with "revert and reinstate", and I too have a problem with "revert good edits just because they were made by a banned user", then I think we are in agreement... :-) Sjakkalle (Check!) 15:01, 27 May 2008 (UTC)
I hope so :) Check my sandbox rewritten version below, and see if you'd like to change anything.--Piotr Konieczny aka Prokonsul Piotrus| talk 16:49, 27 May 2008 (UTC)
I think I support the general sentiments here. I added the bit about core policies and BLP myself last November after a flurry of discussion over on ANI about this policy. While I understand the need to keep banned users banned, a strict "no edits from banned users" policy just opens a new avenue for abuse using reverse psychology. It also discourages second chances. I feel this section and the idea of "meatpuppets" is far too broad and prone to abuse. I support refining this policy to better isolate abuse and disruption, over the mere sight of a name. I made a few modifications below. I'd like to go further and suggest striking the sentance about meatpupptery entirely, and finish on "Users who reinstate edits by banned editors take complete responsibility for the content by so doing." --InkSplotch (talk) 16:03, 27 May 2008 (UTC)
Neither version seems obviously bad to me, and I agree that the wording on meatpuppetry is more scary than it should or needs to be. Nonetheless, I feel the proposed version goes a bit too far on the side of caution — we really do not have the time or ability to review each and every edit made by every sockpuppet of a banned user in order to determine whether their net value is positive or negative (which is what the proposed version could be interpreted as requiring). The point of this policy is that it should be OK to revert edits made by banned users wholesale: they weren't supposed to have been made in the first place, and if any turn out to have been useful after all, someone else will redo them.
Anyway, I've added my own suggestion below; perhaps we can work towards a consensus version combining the best aspects of each proposal. I've left out the paragraph starting with "If a banned user makes helpful edits" from my version, since I think it may be a bit misguided: while we certainly might want to do that in specific cases, we don't really want to encourage the idea that registering a sockpuppet and making helpful edits with it is a recommended way to get unbanned. In its stead, I've proposed a somewhat longer paragraph (which really would belong in the "Reincarnations" section) hopefully more or less describing current practice in those rare situations where a banned user manages to return as a productive editor. —Ilmari Karonen (talk) 18:57, 27 May 2008 (UTC)
I would support simply removing that language from the policy or anything that makes it less stupid. Helpful edits are helpful edits. It shouldn't matter who makes them. Blanket reverting is detrimental to our goal and reverting and reinstating is just making 2 pointless edits, though that's still better than just reverting. Remember, we're here to build an encyclopedia, not score points in an online game. Mr.Z-man 20:20, 27 May 2008 (UTC)
The problems with this are twofold: First, most edits of any significant value are not obviously useful. Sure, one can usually tell at a glance whether a word is spelled correctly or an interwiki link leads to the correct target, but those are trivial edits anyway; if they get reverted, some bot or AWB user will come along and redo them. But if someone has inserted a seemingly innocuous paragraph with a valid-looking citation to an obscure book, how long will it take you to ensure that it definitely doesn't violate NPOV, UNDUE or COATRACK and that the reference it cites is a reliable source and actually supports the statement? Now how about when the sock has made a hundred such edits? Second, we don't ban users just for the heck of it; banned users are banned for a reason, and that reason, whatever it may be in each case, generally involves a judgment that their edits, on the whole, are not worth sorting through to separate the wheat from the chaff — because if they were, we'd be doing that instead of having banned them. If you think a banned user is making edits that are valuable enough to justify that effort, you should be proposing that they be unbanned and promising to take the time to monitor their contributions instead. —Ilmari Karonen (talk) 21:45, 27 May 2008 (UTC)
If we can't tell if its a helpful edit and the user is banned, there's no good faith assumption and we assume the edit may be bad - that is fine, that is practical. Reverting obviously helpful edits for the sole reason that the user who made them is banned is entirely unhelpful. Mr.Z-man 23:10, 27 May 2008 (UTC)
I don't think we, or anyone else in this discussion, disagree on that (possibly modulo the degree to which bad faith may need to be assumed when it comes to some banned users). That said, people will disagree on what constitutes an obviously helpful edit, and will try to bring grief upon those who revert edits they think are helpful "just because they were made by a banned user". There really needs to be something in this section that protects users who revert edits made by banned users because they found them suspicious or just didn't feel like spending half an hour checking to see if they're useful or not. If we ask users to exercise discretion when reverting edits made by banned users, we need to make sure they don't catch grief for actually exercising said discretion. —Ilmari Karonen (talk) 18:57, 28 May 2008 (UTC)
Based partly on some of what you wrote, I've added a new proposed version to the sandbox section below. Comments? It's shorter, too. —Ilmari Karonen (talk) 19:42, 28 May 2008 (UTC)
Best version yet! I like short and concise. --InkSplotch (talk) 02:37, 29 May 2008 (UTC)
Yes, that looks much better than what we currently have. Mr.Z-man 03:09, 29 May 2008 (UTC)
This sounds fine; since we seem to have a consensus I am updating the policy to include the wording of "Second proposed version by Ilmari Karonen" from sandbox below.--Piotr Konieczny aka Prokonsul Piotrus| talk 16:16, 5 June 2008 (UTC)
  • Um. Every time this discussion comes up, people fail to distinguish between addition of content and removal of content. The anon IP who started this discussion came closest. The point is that blanket reversals run the risk of undoing removal of content, and possibly restoring bad content. This is not a hypothetical situation. Let's say that among the edits of the banned user there are some edits where the banned user removed BLP violations or reverted vandalism ranging from obvious to subtle. Those edits will, of course, not be challenged in the normal course of events, but will those watching the pages in question take the time to look further when they see a bot or AWB-assisted editor roll by with an edit summmary of "reverting edits by banned user"? If they just let that go, without checking further, then the BLP vio or vandalism, unwittingly restored to the article by the person blindly applying rollback or undo, may stay in the article for longer than need be, and may even end up causing harm. The point is that it is normally (though not always) safe to revert the addition of content, but it is far more risky to revert the removal of content, and the reason is simple: when you are reverting the removal of content, you are in fact ADDING content to the article. Content should never be added without a manual review. Thus all processes for blanket reversals should be able to assess whether they are reverting the addition or removal of content. Problems arise though with complex edits that involve both removal and addition of content. Still, any guideline on this should at least make the above points about distinguishing between addition and removal of content. Carcharoth (talk) 19:52, 30 May 2008 (UTC)
A good point. Still, the principle that "content should never be added without a manual review" is somewhat problematic. For example, our article on George W. Bush is 124kb long and has 153 references. If a vandal (banned or not) comes along and replaces it all with "He's an idiot.", it's obvious that the vandalism should be removed. What's less obvious is whether it really makes sense to insist that the article should thereafter stand blank until and unless someone is willing to review each reference and personally confirm that each and every claim made in the article satisfies all applicable policies and laws. I suppose the restoration could be done piecemeal, starting with the lede (which is hopefully easily verifiable), but that's still a lot of work just to revert one vandal. Or we could just agree that reversion is a purely technical operation that simply serves to restore the status quo ante, and involves no editorial judgment beyond asserting that the version being reverted to appears to be better than the current one. —Ilmari Karonen (talk) 22:27, 30 May 2008 (UTC)
What should be done, and many people forget to do (including me) is check for recent vandalism just prior to the edit being reverted, since that other vandalism might not have been reverted. Simply reverting the latest edit without looking back a bit through the history is just asking for trouble. Carcharoth (talk) 03:26, 31 May 2008 (UTC)

[edit] Sandbox

Current paragraph:

Any edits made in defiance of a ban may be reverted to enforce the ban, regardless of the merits of the edits themselves. As the banned user is not authorized to make those edits, there is no need to discuss them prior to reversion. When reverting edits, care should be taken not to reinstate material that may be in violation of core policies such as Neutrality, Verifiability, and Biographies of Living Persons.

Users are generally expected to refrain from reinstating edits made by banned users and such edits may be viewed as meatpuppetry. Users who reinstate such edits take complete responsibility for the content by so doing. It is not possible to revert newly created pages, as there is nothing to revert to. Such pages may be speedily deleted. Any user can put a {{db-banned}} to mark such a page.

Proposed new version by Piotrus (slight modifications by InkSplotch):

Any edits made in defiance of a ban may be reverted to enforce the ban, care should be however taken not to revert helpful, uncontroversial edits (ex. fixing typos or adding interlinks). Also, when reverting edits, care should be taken not to reinstate material that may be in violation of core policies such as Neutrality, Verifiability, and Biographies of Living Persons. Improving the quality of Wikipedia takes precedence over ban enforcement.

If a banned user makes helpful edits, he should be advised to ask for review of his case and to be unblocked (with or without restrictions, based on previous disruption). If a banned user continues to make harmful edits, his ban should be extended.

Users are expected to refrain from reinstating unhelpful edits made by banned users and such edits may be viewed as meatpuppetry. In all cases, users who reinstate edits by banned editors take complete responsibility for the content by so doing.

It is not possible to revert newly created pages, as there is nothing to revert to. Such pages may be speedily deleted. Any user can put a {{db-banned}} to mark such a page.

Proposed version by Ilmari Karonen:

(I think I'll remove this one soon unless anyone feels it's better than my second proposal below.)

Any edits made in defiance of a ban may be reverted wholesale to enforce the ban, regardless of the merits of the edits themselves. By banning a user, we've decided that we'd rather do without their contributions, and the fact that they've managed to circumvent the technical measures preventing them from editing does not change this. Reverting simply serves to retroactively apply that decision.

This does not mean that obviously beneficial edits (such as fixing typos or undoing vandalism) must be reverted merely because they were made by a banned user, but there is no obligation to retain them either, and the presumption in ambiguous cases, or where the volume of edits precludes individual review, should be in favor of reverting. If the edits were truly useful, someone else will eventually make them again. However, when reverting edits, care should be taken not to reinstate, even temporarily, any material that may be in violation of core policies such as neutrality, verifiability, and biographies of living persons.

Users who reinstate edits made by banned users take complete personal responsibility for the content by so doing. Editing on behalf of a banned user is strongly discouraged and may in some cases be viewed as meatpuppetry, especially if the edits are similar to those that led to the ban in the first place. If in doubt, think twice and consult others more familiar with situation first. If you genuinely believe that an edit made or suggested by a banned user would be an improvement to the encyclopedia, making an equivalent improvement in your own words may be preferable to reinstating the edit verbatim.

It is not possible to revert newly created pages, as there is nothing to revert to. Such pages may be speedily deleted. Any user can put a {{db-banned}} to mark such a page.

(Note: This paragraph really belongs in the Reincarnations section.) Banned users who return to edit constructively under a completely new identity, and manage to refrain from the activities that got them banned as well as anything else that would connect them with their old identity, will generally not suffer any sanctions for the simple reason that they cannot be identified as a banned user. Should such a connection subsequently come to light, and if the user's contributions under the new identity show them to be a productive editor with no sign of the behavior leading to their former ban, this may be considered a reason for rescinding the ban or at least not enforcing it against the new identity. However, this depends strongly on the reasons for the ban; some banned users may simply not be welcome under any circumstances, and in some cases there may be reasons for the ban that may not be apparent from the edit history alone.

Second proposed version by Ilmari Karonen:

Anyone is free to revert any edits made in defiance of a ban. By banning a user, the community has decided that their edits are prima facie unwanted and may be reverted without a further reason. This does not mean that obviously helpful edits (such as fixing typos or undoing vandalism) must be reverted just because they were made by a banned user, but the presumption in ambiguous cases should be to revert. When reverting edits, care should be taken not to reinstate material that may be in violation of core policies such as neutrality, verifiability, and biographies of living persons.

It is not possible to revert newly created pages, as there is nothing to revert to. Such pages may be speedily deleted. Any user can put a {{db-banned}} to mark such a page.

Users may reinstate edits that were reverted due to being made in defiance of a ban, if they sincerely believe the edits are beneficial to the encyclopedia and compliant with policy. Users who reinstate such edits take complete personal responsibility for the content by so doing. Note that editing on behalf of a banned user is strongly discouraged, and may in some cases be viewed as meatpuppetry, especially if the edits in question are similar to those that led to the ban in the first place. If in doubt, think twice and consult others more familiar with the situation first.

[edit] Banned users and Wikipedia namespace pages and discussion pages

Could we have some comments on how to handle Wikipedia pages created by banned users? Effectively, these pages (usually policy pages) embody some idea or concept or guideline that may not be accepted by the community, but may end up generating discussion. Sometimes, the fallout from the banning of the user (or the discovery that the creator was the sock of a banned user) sees people assiduously hunt down such pages and delete, or attempt to delete, them, even when the idea has already taken root and been adopted and community discussion (in good faith) has taken place. Some examples of this can be seen at the following deletion discussions:

There should probably also be something that says that even if a banned user creates a page, subsequent good faith discussion by others on the talk page of that page should be handled sensitively (saying "you've been trolled" is not always the best way to handle such a situation, and in some cases saying that, especially if there was no trolling, can poison a productive discussion). By all means announce to those discussing things on the talk page that the creator of the page is a now banned user, but let those at the talk page come to their own, independent conclusions as to what to do, and don't steam in wielding G5 like the blunt instrument it is. Carcharoth (talk) 20:12, 30 May 2008 (UTC)

[edit] Lacks procedures for Topic Bans

This important article needs a discussion of the procedures for implementing and enforcing topic / article bans. The notion that a ban can be enforced by a block doesn't really apply to a topic ban, since the editor (as I understand it) isn't blocked but is free to edit on all other areas of Wikipedia.

There needs to be some guidance on how and where the editor should be notified. Admins who have experience in implementing bans would probably be the most helpful people for drafting such a section.

A good method for notification would be through a standardized template. There is presently a template for general bans Template:Banned user to go on the User page and for topic or article bans to go on the article or article talk pages Template:User page ban, Template:User article ban, and Template:User article&talk ban, but I don't know of any templates for topic bans to go on the user page. Perhaps something along that line needs to be created.

Finally, enforcement of topic bans, which I presume is done by elevating the topic ban to a general block. Once a topic ban is established, am I right that any admin can initiate a general block once the ban is violated?.

--SteveMcCluskey (talk) 19:24, 25 April 2008 (UTC)

I had been topic banned by the arbcom though the arbcom explicitly stated that I had not edited inappropriately, so I do not accept such a template. Andries (talk) 19:37, 25 April 2008 (UTC)
At the moment, the page says that enforcement is done by reverting edits. That is somewhat strange in cases where there is nothing wrong with the edits themselves. But I find the entire policy very confusing on topic bans, while the concept is really not all that complicated. I've just received an article ban myself and I have no clue, and I get the impression that neither does the banning admin. I'm also with Andries here; since no evidence has been presented for what I am supposed to have done wrong, I would not accept such a template either. Guido den Broeder (talk) 19:43, 25 April 2008 (UTC)
This doesn't seem to be the place to discuss the merits of particular cases. Does anyone have opinions on defining procedures to eliminate the apparent confusion reported by Guido den Broeder? I've seen similar uncertainty among admins on AN/I about how to close and log a topic ban, so there does seem to be a problem. --SteveMcCluskey (talk) 19:37, 27 April 2008 (UTC)

[edit] Proposal on community bans

See WP:AN#Community ban discussion. FT2 (Talk | email) 00:15, 3 May 2008 (UTC)

Copied from AN where these things get lost a bit:

Currently a community ban is de facto created, any time a user is indef-blocked, and no admin is prepared to unblock them.

Two problems:

  1. Most community bans are in practice consensus driven anyway,
  2. With over 1500 admins there is always an admin who will unblock even a block most others agree with. Our consensus model clashes heavily with an "any admin acting unilaterally" model here. It can cause problems, since most other decisions, wide consensus trumps most things.

Question - is it time we moved to a consensus-based view of a community ban?

The proposal would be:

  • A community ban arises where there is a consensus that a user should be banned from the wiki. A ban may either be created by discussion leading to consensus, or by an initial indef-block that a consensus then agrees should be considered a ban. Once created, a community ban may be removed by consensus, or by appeal to the Arbitration Committee. The consensus in each case is of uninvolved admins.

Thoughts? FT2 (Talk | email) 23:53, 2 May 2008 (UTC)

One thing we'll still all disagree on is when "consensus" is reached on a ban. Is 51% consensus, 70%, 85%? Who determines it? Maybe its finally time to get the crats, our consensus determining experts, involved in more things. MBisanz talk 02:49, 7 June 2008 (UTC)
So far, disputed discussions have gone to ArbCom where the arbs decide it, since they're the appeal route for a disputed community ban anyway. This is a solid rational update that shouldn't get derailed over minor points. Anyone who watched the CAMERA fiasco knows it's just too gameable to keep a system where a single corrupt administrator out of 1500 holds such power: too much of a temptation to plant a few operatives who'll act politically. DurovaCharge! 02:58, 7 June 2008 (UTC)
That is something I totally agree with, the idea that 1 or 2 admins can overturn a community ban simply by unblocking is something we really need to work to eliminate from the policy. I'd give up by vagueness objective to get what FT2 is proposing into policy. MBisanz talk 03:20, 7 June 2008 (UTC)
Most of Wikipedia's consensus discussions reach consensus in the neighborhood of 2/3 to 3/4, with weight considered for the quality of the argument. As long as people are satisfied with that general standard we probably don't need to redefine it here, since the Committee and the Foundation exist as appeals routes. DurovaCharge! 03:45, 7 June 2008 (UTC)

[edit] Prior discussion at WP:AN

See Wikipedia:Administrators' noticeboard/Archive143

Agree with point 1, but is there any evidence of point 2 being an issue? I haven't seen it, and of the 1537 admins, I doubt more than a third are currently active. --Rodhullandemu 23:58, 2 May 2008 (UTC)

The consensus model is seen as important. Also, actual cases probably do exist where an argument has broken out whether sole admin X can unban someone whom a consensus favors keeping blocked (placing consensus vs. unilateral models at loggerheads), or some such. Just seems that this would be a nice way to handle it that effectively codifies how we do it nowadays, where actual community bans mostly are consensus backed. FT2 (Talk | email) 00:05, 3 May 2008 (UTC)
Todays midnight unblock of Jack Merridew supports point 2.--Cube lurker (talk) 00:02, 3 May 2008 (UTC)
Forgive me if I missed that; I've, er, been writing an article, bizarrely enough.--Rodhullandemu 00:07, 3 May 2008 (UTC)
No, it does not support point 2. The editor who unblocked was the one who blocked him in the first place. Very different situation from what you're stating. -- Kesh (talk) 13:17, 4 May 2008 (UTC)
Agree, it would certainly seem inappropriate for community consensus to be overturned by a single admin, and in that sense, it does seem that the current system lacks logic. --Rodhullandemu 00:11, 3 May 2008 (UTC)

I dont agree with the assesment that any indef block is a defacto ban... Remember the differences: accounts are blocked, persons are banned. Many, if not most, indef blocks are a result of an account being used to vandalize or cause any other type of disruption, after which the person can re-register and start over productively. You're not proposing we need to get consensus first for these blocks, I hope? — Edokter • Talk • 00:01, 3 May 2008 (UTC)

I was the user who wrote WP:INDEF so yes, I do remember it :-) More seriously, you might have split the sentence when it wasn't intended you should. Note the wording after the comma, "and no admin is prepared to unblock them". Or else you're discussing clean starts, which are a bit different. We surely don't expect anyone indef-blocked to be able to just start a new account and carry on unchanged. FT2 (Talk | email) 00:08, 3 May 2008 (UTC)

FT2, that sounds like a reasonable concept. Here is a pair of options for implementing it:

  1. Require 2 admins to unblock a community-banned user; they must commit to keeping tabs on the user until the community comes to a consensus that this is no longer necessary (or X months, whichever comes first).
  2. Any user subject to community ban who is unblocked will automatically be reblocked after X edits. At that time, a different admin must review the user's edits and decide to finally unblock. Both admins will thereafter be charged with keeping tabs on the community-banned user until there is community consensus otherwise (or X months pass, whichever comes first).

FT2's vision is cleaner than either of these mechanisms for implementing it, but what I'm trying to do here is offer concrete suggestions that would at least make the community ban more difficult to overturn than it is now. Antelantalk 00:04, 3 May 2008 (UTC)

Specifics of X admins may not be needed. Good consensus is more to the point (see FCYTravis' point below). An addition that "banned users are usually parolled back into the community" will suffice. FT2 (Talk | email) 00:12, 3 May 2008 (UTC)
Yep, I like this approach. If it fails, we can consider other options (such as the X admins watching for X months approach that I've suggested here). I, for one, hope it doesn't come to that. Antelantalk 00:19, 3 May 2008 (UTC)

This needs to be studied very closely to avoid becoming "Votes for Banning." FCYTravis (talk) 00:05, 3 May 2008 (UTC)

Yes. Same applies to any consensus discussion though. CSN was merged back into ANI exactly to ensure many more eyeballs to prevent that. Yes it's important, but the issue here is more "in principle". FT2 (Talk | email) 00:12, 3 May 2008 (UTC)
I support this. A consensus-based approach makes it much more likely that people are going to seriously review these matters in substantial numbers, reducing the probability of inappropriate bans. Furthermore, individual admins should not have veto power; we need to encourage more collaborative decision-making among admins. Everyking (talk) 00:06, 3 May 2008 (UTC)
I have reservations in supporting this, though I am not closed to the concept -- indeed, it is to a certain extent current practice. Votes for banning must absolutely be avoided, and the barrier must be pretty high. The aim of the debate should be "are there any reasonable objections to putting this ban into place?" rather than "does anyone oppose this ban?". This is, of course, difficult to achieve. It might require a decision that whoever says "Support ban ~~~~" has such a comment removed from the discussion. I have great scepticism about Antelan's ideas in particular -- that kind of concrete system is incredibly arbitrary and inflexible -- two principles that must be as far as possible from any consideration of banning. In response to Everyking, admins should not have veto power, but a well-reasoned objection should. Sam Korn (smoddy) 00:17, 3 May 2008 (UTC)
With your permission, if this went ahead, I'd put the "aim of the debate" point as a footnote to the final version, if folks agree. Good call. FT2 (Talk | email) 00:43, 3 May 2008 (UTC)
  • Yes, I have advocated something like this proposal for a long time. We need to reduce drama. If 80% of admins want to ban somebody, they should be banned. Jehochman Talk 00:18, 3 May 2008 (UTC)
    • I think that rather depends what the other 20% think. If they, in general, think "I don't think they need to be banned yet, but it's well on the road", then I think a ban would be reasonable. If they in general think "No, absolutely not, this user has done nothing warranting banning and here's why ...", I think a ban is misplaced. If consensus banning is to be introduced, it has to be real consensus, not an arbitrarily defined figure. Sam Korn (smoddy) 00:27, 3 May 2008 (UTC)
      • But then Admins are supposed to be able to assess consensus on AfDs, and Bureaucrats on RfAs. Since this would be an Admin-input process, would Bureaucrats therefore be the best to assess consensus, or should it be left to Admins? --Rodhullandemu 01:02, 3 May 2008 (UTC)
  • Support in principle, with an important caveat: people who are partisans to a particular conflict should be expected to recuse themselves from the actual consensus and/or disclose their history of involvement. It wouldn't be right for a team of people to gang together and force the outcome of something as important as a ban. DurovaCharge! 00:20, 3 May 2008 (UTC)
  • Absolutely agreed regarding disclosure of involvement. Once concern, though, is that this phrasing makes it sound like this is going to be a vote. Certainly involved admins shouldn't participate in an administrative way in these matters, but involved parties should be welcome to give their input, so long as they disclose their involvement. Antelantalk 00:23, 3 May 2008 (UTC)
  • The amended version looks great - a long overdue update. Thank you very much, FT2, for proposing this. Should reduce the occurrence of periodic drama-fests. DurovaCharge! 00:39, 3 May 2008 (UTC)
  • Support, in principle - but I would like it made clear that the later lack of consensus for a ban resulting from an indef block which was originally supported is not consensus to unblock. The erosion of consensus in banning an editor still means that the block is in place, but that an admin is prepared to argue for the lifting of the block; there is still the need to establish consensus that the block can be lifted. This comment comes directly from the actions performed regarding Jack Merridew a few sections above (at the time of writing). LessHeard vanU (talk) 00:29, 3 May 2008 (UTC)
No consensus to escalate an indef block is not the same as consensus to remove it. I think that's your point. If so, concur. FT2 (Talk | email) 00:44, 3 May 2008 (UTC)
Yes, but also consensus to de-escalating a ban does not mean there is consensus to unblock, only that there is the potential to unblock (first we agree we can discuss unblocking a banned editor, and only then do we talk about lifting the block.) LessHeard vanU (talk) 12:50, 3 May 2008 (UTC)
  • Support the proposal. Based on what happened today the current process is fatally flawed.--Cube lurker (talk) 00:35, 3 May 2008 (UTC)
  • This sits fine with me. At present, the whole "if one admin feels an unblock is good, then it's all fine" concept really doesn't work in situations where the greater community may have something to say about it. In the above situation, I would be surprised if the majority of the community even knows it happened. We're a consensus-driven project, and that should be extended to community bans. Tony Fox (arf!) 01:52, 3 May 2008 (UTC)
  • I support this as well. Wizardman 01:58, 3 May 2008 (UTC)
  • You have my support. - Peregrine Fisher (talk) 05:20, 3 May 2008 (UTC)
This discussion assumes that community bans are normally enforced by a block that can be undone by the action of a single admin. In many cases community bans are topic bans and are, therefore, not enforced by a block but by reverting the edits of the banned user. Any policy revision should keep this in mind.
That being said, I agree with the concept that one admin should not be able to overrule consensus by undoing a community ban. --SteveMcCluskey (talk) 12:41, 3 May 2008 (UTC)
  • Support in the hope that a previously supported block will henceforth not be reversed due to an agreement among a few individuals (like it was yesterday) --PeaceNT (talk) 14:25, 3 May 2008 (UTC)
  • I must oppose this. Banning is a very serious issue, and to be quite frank, the Wikipedia community acts like a mob sometimes. The current rules for a community ban, which allow that if any admin is willing to unblock, the user is not banned, are needed as a safeguard. This is simply holding bans to an extremely high standard of consensus. Already, we see editors being AFD'd from time to time by those who treat community bans as votes for banning. This change will simply empower that. Heimstern Läufer (talk) 17:15, 3 May 2008 (UTC)
    • Another question: How are we to judge when consensus is reached? We have constant difficulties with this at RFA, and yet we're trying to implement a system like that here? This is a bad idea, especially with an issue so serious as banning. I take extreme issue with Jehochman's statement that if 80% of admins think a user should be banned, they should banned. This is the definition of votes for banning, and is something we should be running from, not moving toward. Heimstern Läufer (talk) 07:10, 4 May 2008 (UTC)

This is based on a misconception, in my opinion. If someone is indefinitely blocked, and an admin unblocks, then there arises the possibility that the admin will re-block, possibly/hopefully following discussion. If then the unblocking admin leaves the block then it's a ban, since they have reached the conclusion that their action has no support. If it continues to be lifted by that admin or others, then there is no ban for now — if a lone warrior persists in the face of vast opposition, then they will be stopped eventually by ordinary means, perhaps including an emergency arbitration. In the end, perhaps that collection of unblocking admins will as a whole come to the ban view, and leave it in place, in which case it's a ban by consensus. If they do not, then it cannot be viewed as tenable that the individual is likely to remain blocked, and thus no ban can possibly be in place.

None of these facts change if you happen to legislate some other "consensus" model, as is being mooted here – it matters not if there is a so-called consensus of a handful of passing admins to ban since if someone feels strongly enough that the person should be unblocked they will still do so. You cannot possibly stop unblocks by attempting to legislate in this way; a nice demonstration of why policy is descriptive and not legislative. This new piece of legislation tries to stop something that simply cannot be stopped while any admin retains technical abilities to unblock since feelings will eventually run high just as they do now. Finally, I dispute FT2's #2 as demonstrably untrue in the numerously many cases where indefinite blocks have stuck in the past. If we accept FT2's #2, however, then the proposal is doomed to failure as a matter of definition for the reasons I have just described. Also, of course, this will in practise function identically to the disastrous Community Sanction Noticeboard, and it is hopeless to suppose otherwise. Splash - tk 17:19, 3 May 2008 (UTC)

  • I oppose this proposal per Heimstern Laufer, but I also find Sam Korn's argument persuasive. I don't think every admin should have veto power, but I also don't think that a simple majority is sufficient to convict a user to spending the rest of their life on activities more worthwhile than Wikipedia. :) What the numbers game fails to consider is that a single user who wishes to unblock a user who most of the community wishes to ban will have a good reason, and in some cases will have a strong conviction that the ban is just wrong. I think this sort of reasoning also explains why RFA requires a supermajority: people who oppose RFAs almost universally feel more strongly about the issue than supporters. So if one admin wants to unblock and nobody agrees with him, maybe that admin is just loony. If two or three or four other admins agree with him, even against a large majority, my instinct is to let the admins who wish to unban take responsibility for the situation.
  • I'm also concerned about any process that ends with an appeal to the Arbitration Committee. ArbCom is overworked and doesn't need to be consulted every time an indef-blocked user wishes to receive a second chance. I don't think a majority is needed to overturn a ban: a significant minority is enough for me. Shalom (Hello • Peace) 03:56, 5 May 2008 (UTC)

Call it WP:CREEP if you want, but we need some sort of guideline or process for this. A significant minority or even a majority can't be the deciding factor as shown by what happened recently (non canvassing emails?). I can see WP:ILIKETHEM and WP:OTHERBANNEDUSERsEXIST type of essays helping in some sort of discussion. - Peregrine Fisher (talk) 04:09, 5 May 2008 (UTC)

  • Disagree with the unban/unblock rule. There needs to be parity in blocking and keeping blocked, in banning and keeping banned. But this proposal turns that parity upside down. If it requires a community consensus to ban someone, then it continues to require a community consensus to keep that editor banned. An unban discussion reaching no consnesus is proof that there is no longer consensus to ban that person, and thus they are not community banned by consensus. Any other rule will result in votes for banning again. GRBerry 14:26, 5 May 2008 (UTC)
Comment: I think the #Jack Merridew thread on this page demonstrates the potential of a serious problem. I also want to point out why the community sanction board was abolished. This initiative should avoid such pitfalls. -- Cat chi? 18:49, 5 May 2008 (UTC)
Something to remember

We get caught up in our discussion of the way in which Wikipedia operates that sometimes we forget that it operates at all. We are consumer driven, millions more readers than the thousands of editors (hell, I'm not even an "editor" as much as I am a reader). I've been reading Ed Fitzgerald's userpage, which got me thinking about this thread.

The point is that we should all remember that Community Bans don't happen that often in the grand scheme of having the office open 24/7/366this year for business. At best a dozen a year.

These sort of community ban reviews are even more rare. I can only think of a couple times this has happened in recent years.

Sometimes smoke is caused by char and ashes. There is not always fire.

I'm just sayin. Keegantalk 04:34, 3 May 2008 (UTC)

[edit] Pretty diagram

http://img166.imageshack.us/img166/1953/banningza1.jpg

Thoughts? Daniel (talk) 02:52, 7 June 2008 (UTC)