Wikipedia talk:Banning policy

Should the minimum duration of 24 hours for site ban discussions be increased? — Wug·a·po·des​ 02:00, 9 July 2020 (UTC)[reply]

• A
  WP:CBAN to reflect that. Personally, I think 24 hours is too short, and that we should have them run at least 3 days. I thought about a minimum of 1 week, but I think forcing discussions to be that long will lead to unnecessary stress for little gain. — Wug·a·po·des​ 02:00, 9 July 2020 (UTC)[reply
  ]
• It’s worth pointing out that many of these aren’t controversial, and that extending it to 72 hours minimum in some cases would be pretty unfair to the editor involved. See the Zawl/FWTH ban discussion. There’s no reason that unanimous consensus needed to last 72 hours—much less a week. Tough cases make bad law: there’s currently frustration because a high profile editor was banned (and yes, I proposed that ban), but I had been thinking about proposing this myself. I keep coming back to the Zawl example, and there are likely others more recent, where there was a clear consensus after 24 hours and extending it wouldn’t have helped either the community or the editor in question.
  So yeah, I guess I oppose this, but I would suggest adding an alternative such as Administrators are expected to ensure the community and any impacted editors have had sufficient time to comment before closing, even if the minimum time has been met. TonyBallioni (talk) 02:09, 9 July 2020 (UTC)[reply]
• I suggest that all CBAN discussions, and, indeed, all discussions of any kind on en.wiki be mandated to run for a minimum of 90 days, just to make sure that every Wikipedian has their say. I offer this suggestion in the spirit of the pleasure and enjoyment that all Wikipedia discussions impart to all the participants, and in the desire that everyone should get every last scintilla of joy that these discussions offer.
  How about 48 hours? Two days seems more than sufficient in the majority of cases; perhaps with the caveat that if the subject requests it, an additional 24 hours can be added? Beyond My Ken (talk) 02:37, 9 July 2020 (UTC)[reply]
  • I think 48 hours plus a line similar to what I said above would work well. I don’t want it to be a “they can request and additional 24 hours” in cases where there’s no controversy and someone is just flailing. Put something in the policy saying not just to do the minimum if it seems unfair, and an extension to 48 hours to account for stuff like being filed on a Saturday night, etc. TonyBallioni (talk) 02:46, 9 July 2020 (UTC)[reply]
    • These both sound reasonable, and I especially like the idea of making it clear that the minimum is not the same as the recommended length. I'm not actually that tied to 3 days, but thought it a reasonable place to start discussion from. — Wug·a·po·des​ 02:57, 9 July 2020 (UTC)[reply]
• I would vote that seven to ten days is a happy medium. We have longer waits for much more trivial things. PackMecEng (talk) 04:30, 9 July 2020 (UTC)[reply]
  • There is no functional difference between 7-10 days and thr 90 days I sarcastically suggested above. It's much too much time. Beyond My Ken (talk) 22:29, 9 July 2020 (UTC)[reply]
    • Seems kind of subjective, I think 24-48 is too short. I would be okay with 72 if there are easy options for reasonable extensions. Shit happens and sometimes you just need time to go through everything and make a case. I would favor 7 in general, 10 might be a little long the more I think about it. PackMecEng (talk) 05:17, 12 July 2020 (UTC)[reply]
• I think the trick is that SNOW closes should be allowed after 24 hours. But otherwise 7 days seems reasonable. Hobit (talk) 04:34, 9 July 2020 (UTC)[reply]
  • That would effectively make this policy historical as threads get archived by the bot before then and many of them actually do die out. The other issue is that this would basically encourage admins just to indefinitely block without seeking consensus. A seven day high drama event to do something that’s functionally the same as a block and declined appeal to AN isn’t very attractive. In reality, getting an indefinite block overturned is very difficult, and if the admin insists that it not be lifted without consideration at AN, you have a site ban. You incentive shooting first by creating procedural requirements that in most cases aren’t needed. TonyBallioni (talk) 04:36, 9 July 2020 (UTC)[reply]
    • I am not sure that has anything to do with the length of time though. That scenario plays out right now with the 24 hour version. PackMecEng (talk) 04:51, 9 July 2020 (UTC)[reply]
      • Not really. If you want to use the current example: I could have easily indefinitely blocked Sashi, and requested that the reviewing admin take it to AN for review given the history of the block log, which I think would have been justified. That probably would have been granted. It would have been unlikely that there would have been a consensus to overturn the block, which would have resulted in a community site ban under this policy. I didn't, because I thought community input on such a matter would be important. You're less likely to see people do things like that if it's a guaranteed week of drama even when a clear consensus exists within 36-48 hours.
        The danger in making behavioural policies too burdensome to do is that policy, correctly, gives a lot of discretion on how to block. When you make it more difficult to avail to community based options, what that actually does is shift more power to individual admins to act without consensus, since people are less likely to go through a bureaucratic process when there's an easily available alternative that would usually result in the same thing in practice. The great irony of bureaucracies is that by increasing rules, they also increase decentralization as the centralized functions are unable to handle the growing ruleset, which in turn means that the increasing rules become dead-letter. TonyBallioni (talk) 05:04, 9 July 2020 (UTC)[reply]
        • Do you have an example of that or something similar? PackMecEng (talk) 05:06, 9 July 2020 (UTC)[reply]
          • I just gave a hypothetical for this process, but if you want an actual example, check the
            WP:AC/DS prohibits blocks under DS for more than a year. The current consensus is that you can still indefinitely block someone for an AE violation, but that only the first part has the special AE appeal restrictions. It's a word game: there's a policy that's burdensome in practice, so people find a way to make it work within the existing structures. When you try to write overly prescriptive rules for things like this, it usually backfires. TonyBallioni (talk) 05:16, 9 July 2020 (UTC)[reply
            ]
            • I am not sure I agree with that. Perhaps we have just had different experiences, but thanks for explaining your view on it. PackMecEng (talk) 05:21, 9 July 2020 (UTC)[reply]
        • <ec>That's fair. And I don't like changing rules just to address a one-off problem. But if that closure is upheld because it technically didn't violate any rules, then we probably need to fix the rules. Hobit (talk) 05:11, 9 July 2020 (UTC)[reply]
          • (edit conflict) You can do that by emphasizing that the minimum time isn't mandatory/sufficient and fairness in a way that leaves discretion to individual admins, but gives a policy basis for reversing closures if there are concerns that not enough people have had time to comment. One that is open ended, and is subject to review by either the community or ArbCom. By doing that, you provide an avenue for review while also avoiding the issues raised by having a prescriptive time. TonyBallioni (talk) 05:16, 9 July 2020 (UTC)[reply]
            • Given the current language wasn't enough to stop a case that IMO shouldn't have happened, I think we need more prescriptive language. I don't like it, but I can't support language that lets what happened happen. I get the "bad cases make for bad caselaw" argument, and I agree with that in general. But abuses need to be addressed somehow. And given how rare ban discussions are, I just don't see it hurting that much. Hobit (talk) 06:06, 11 July 2020 (UTC)[reply]

              I find Hard cases make bad law#Hard cases make good law to be 100% applicable to this situation. Levivich ^{[dubious – discuss]} 06:19, 11 July 2020 (UTC)[reply]

• I agree with emphasizing that the usual norms for letting a conversation take its course apply. Closers should not be in a rush to evaluate consensus while it is still in the process of forming, whether or not any minimum standard for discussion has elapsed. I do think that 48 hours is probably a better minimum length to ensure opportunity for the global community to comment. isaacl (talk) 05:24, 9 July 2020 (UTC)[reply]

  Regarding 24 hours being enough to establish clear consensus: many discussions have a consensus among those who comment in the first 24 hours which then gets altered with subsequent discussion. Also, if we are truly treating discussions as not being votes, it is good to provide some time for responses, to allow for commenters to actually discuss. I do agree elaborate rules for governing the minimum time are unnecessary: just follow the usual procedure to let people talk about the situation for as long as it is productive. isaacl (talk) 17:04, 9 July 2020 (UTC)[reply]

• I'd suggest 72 hours, possibly with the potential for SNOW if: 1) the usual SNOW conditions are met; 2) 24 hours have passed; 3) the accused has posted in the discussion or showed substantial activity elsewhere in that time span. A week is too long for a default, but I think we need to have a much stronger "no need for a quick close, even if conversation ceased for a few hours" - a close discussion that's getting a couple more participants each day wouldn't struggle from an extra 3 days activity. Nosebagbear (talk) 09:35, 9 July 2020 (UTC)[reply]
• (Summoned by bot)I tend toward the TonyBallioni position that too many rules actually diminishes effectiveness - especially in a WP environment which is largely based around goodwill and reasonable judgement most of the time from editors and admins. However the 48 hours + some convention for extension + safeguards that this isn't simply a 'delay' tactic, seems a reasonable way forward. I don't know how that could work in practice though. Pincrete (talk) 09:38, 9 July 2020 (UTC)[reply]
• Quorum Per
  WP:CENT
  , which is where I saw notification of this discussion that we're having here.

Of course, if there's some pressing need for a sanction then any admin can impose one immediately on their own authority. The point is that they shouldn't claim community consensus without substantial community involvement. Andrew🐉(talk) 09:51, 9 July 2020 (UTC)[reply]

Almost no decisions receive 100 participants and many of them are still taken on behalf of the community, including almost all RfCs etc and implemented policies. If we got RfAs at the rate we got CBAN requests we'd have nowhere near 200 standard participants. We also get many RfA participants who just participate on basis of knowing the individual. We don't want that for a CBAN, so it's still not analogous. If a CBAN had 30 people all saying ban (and most are near-unianimous), why bother piling-on? This is just straight OTT. Nosebagbear (talk) 13:00, 9 July 2020 (UTC)[reply]

There have been

no big deal, the bans merit greater attention. Andrew🐉(talk) 15:49, 9 July 2020 (UTC)[reply

]

• Absolutely - minimum 3 full days if participation is at least 50 iVotes, up to 5 full days if not. ^Atsme _Talk 📧 13:29, 9 July 2020 (UTC)[reply]

  Adding another important thought that just hit me...aside from the obvious, like vandalism and smelly socks...a minimum number must be reached as a good faith show of "the community". In other words, less than 20 participants with 12 approving a site ban is not representative of our community. This may be a case where the iVote should count in order to eliminate any chance of POV creep when closing and making the final determination. For example, the vote must reach 90% support for a siteban out of a minimum of 50 participants; if not, it goes to ArbCom. ^Atsme _Talk 📧 13:48, 9 July 2020 (UTC)[reply]

• 48 hours should be sufficient. A week or a minimum number of votes is too much. --Guerillero | Parlez Moi 14:18, 9 July 2020 (UTC)[reply]
• Support 72 hours x 168 hours. The iprtant thing is to give everyone the opportunity to have a say; but it's also the most important discussion that an editor can ever face here (excepting, perhaps, the C-ban unblock request!) and I see nothing whatsoever damaging to the project in allowing a mandated period. However, WP:SNOW should definitely be noted: there's no point in a process
  for its own sake, although it should be followed to the letter and not just of the "well, lots of people think this, so I'll do it" vintage. ——Serial ^# 14:50, 9 July 2020 (UTC)[reply
  ]

  @Serial Number 54129: Either I'm losing my aptitude for humour or that really calculates to 21 days². 🤔 --qedk (t 愛 c) 15:06, 9 July 2020 (UTC)[reply]

• 48 hours at most, there is no evidence that longer discussions relate to more accurate consensus - if anything, the drama is proportional to the period something is open (for e.g. see
  WP:SNOW closes after 24 hours should also be considered as valid. --qedk (t 愛 c) 15:00, 9 July 2020 (UTC)[reply
  ]
• Oppose - 24 hours is long enough. From experience, if there's not clear consensus in favour of a site ban after 24 hours then it isn't happening: cooler heads prevail, discussion demonstrates the incidents leading to the proposal weren't that bad, or lesser sanctions are proposed and eventually adopted. The result of bureaucratically forcing such a discussion to go on longer is only unnecessary drama, it doesn't help anyone or improve anything, it just creates another way for disruptive users to wikilawyer their way out of sanctions. Ivanvector (^Talk/_Edits) 16:20, 9 July 2020 (UTC)[reply]
  • @Ivanvector: the case that prompted this is certainly an exception. Hobit (talk) 23:13, 9 July 2020 (UTC)[reply]
• Oppose -
  the encyclopedia is a workshop floor, not a bureaucracy, this is a solution in search of a problem, sometimes a speedy closure is necessary. We should generally rely on uninvolved admins' judgment; in the rare event of a seriously erroneous decision, there are ample avenues for appeal. I especially oppose any complex matrix of rules that would involve a sliding scale based on time open/number of votes/SNOW. Neutrality^talk 16:39, 9 July 2020 (UTC)[reply
  ]
• Support - 7 days, it shouldn't be less to delete an editor than it is to delete a page. Of course I'd also support any extension shorter than 7 days (48 hrs, 3 days, 5 days, all good). I also agree with a quorum requirement. Also, site bans should be advertised at CENT, or else done on a specific siteban-only board. One of the problems is that site bans are decided by the people who watch AN and ANI (myself included), and the editor-at-issue's TPWs, and not by the wider community. That means you get the editor's friends and enemies, plus the ANI regulars, decided site bans. We should do something so that site ban discussions attract more uninvolved editors who are not ANI regulars. Advertising at CENT would do that. Site ban discussions are really important; they're just as important as any other discussion we have. We should treat it as such. Levivich ^{[dubious – discuss]} 17:44, 9 July 2020 (UTC)[reply]

  From my experience, most site-ban discussions have a clear consensus, SashiRolls' was one in a long time with a less-clear consensus - that's hardly the threshold of evidence necessary to prove your point that such clear-cut discussions need so much bureaucracy. It's a waste of community time and nothing more. And please, for the sake of all that is holy, don't make me see siteban discussions on CENT, they don't require the community's viewpoint "definitely" which is the requirement of RfCs that have wide-reaching effects on the community, a single most probably disruptive editor getting banned indefinitely is hardly a justifiable cause to rally. --qedk (t 愛 c) 18:37, 9 July 2020 (UTC)r[reply]

  It's a fallacy that keeping a discussion open longer wastes time. Whether a discussion is open for one day or seven, the amount of time required from volunteers doesn't change. Keeping a discussion open just increases the opportunity for volunteering; it doesn't increase the time required of anyone. Keeping a discussion open longer costs absolutely nothing. And before anyone says, "but if it's open longer, more people will post!" Well, yeah, that's the point. Levivich ^{[dubious – discuss]} 19:53, 9 July 2020 (UTC)[reply]

  What's fallacious is you labelling something a fallacy every time someone says something against your viewpoint. I was quite clear about my opinion arising out of information that is evidentiary in nature, the same does not apply to your opinion as far as I can tell. Keeping discussions open for longer "evidently" leads to more drama in most cases (per the law of entropy, Godwin's Law, qedk's ANI tips and tricks signed edition, et al.) and even in cases where it doesn't, there's no evidence that a shorter discussion would've been less effective. Since you are the one proposing a change, the onus is on you to provide evidence, and yes, Sashi's case is still an outlier which you'd have to prove to be the norm to actually count as a valid opinion. --qedk (t 愛 c) 08:31, 10 July 2020 (UTC)[reply]

  The amount of time you (or I) spend in this thread will not change whether the thread is closed today, tomorrow, or next week. It's not like if this discussion were closed next week, more of your time is required, but if it's closed tomorrow, then less of your time will be required. If a !voter reads a thread and !votes, the amount of time that person spent would be the same regardless of how long after their !vote the thread is closed. Thus, keeping a thread open does not "waste time" because it doesn't require any additional time from anyone. It's about logic, not evidence. And as for evidence, look at the current case. If MastCell has waited until after the weekend to close it, much LESS time would have been spent than has been spent so far on this. By closing it early, MastCell thought he was saving time, but actually ended up like quintupling the amount of time. To the point we're considering a policy change to stop anyone else from doing what MastCell just did. Keeping site ban threads open (or any thread open) only provides opportunity for more to join the conversation. It doesn't waste time (but bad closes do). Levivich ^{[dubious – discuss]} 13:30, 10 July 2020 (UTC)[reply]

  I do agree that many contentious conversations that are closed early based on an ignore all rules/

  outcome is obvious rationale end up costing more of the community's time in the end. But the amount of time spent on a thread does depend on how long it is open, as participants will continue to read new comments, incorporate them into their thought processes, and judge if they wish to make additional comments. If the discussion becomes acrimonious, then experienced community members will spend time and effort trying to redirect the conversation to more productive paths. There are costs to keep a discussion open longer, which must be balanced against the effectiveness of new replies. isaacl (talk) 16:33, 10 July 2020 (UTC)[reply

  ]

  I consider additional participation in a discussion to be a benefit not a cost. Take a hypothetical discussion that gets 5 comments in the first day. If everyone has said all there is to say, then no one else will post in that discussion whether it's open for one day, or two, or twenty. The closer will still need to read the 5 comments regardless of the duration that the discussion is open. If, on the other hand, that discussion stays open and on the second day, 5 more comments are made, those additional 5 comments are a good thing (benefit), not a bad thing (cost). Those 5 comments are to be encouraged and valued. One has to take a pessimistic view of their colleagues to think that, chances are, the 5 comments during the second day in this hypothetical discussion are not valuable. If you look at the site ban re-opening proposal at AN now, you'll see editors refer to additional discussion using terms like "waste of time", "drama", "pointless". That's a very pessimistic view of the value of discussion. If a site ban discussion is open for, say, 72 hours instead of 24, and no one comments beyond the 24 (i.e., the extra time is unnecessary), then it will have cost nothing because the volume of discussion will be the same regardless of whether it's closed at 24 hours or at 72 hours. If, on the other hand, more comments come in on the second and third day, then that's still not an increased cost, that's a good thing, because it means we're discussing. It's not drama, it's not pointless, and it's not wasted time. Levivich ^{[dubious – discuss]} 16:45, 10 July 2020 (UTC)[reply]

  Whether or not I make another comment, I have to incur an additional cost to read the subsequent comments, incorporate them into my thought processes, and decide if I want to make another comment. If I end up deciding I do not need to comment, I have incurred a cost without adding any benefit to the discussion. This gets multiplied by every single participant. The point is not that continuing a conversation has benefits, but that the amount of time I spend on a thread does depend on whether or not it stays open for longer, contrary to what you asserted. Of course there are benefits, but there are also costs, and once the costs outweigh the benefits, then discussions should be brought to a conclusion. isaacl (talk) 17:23, 10 July 2020 (UTC)[reply]

• Support 7 days at a minimum. Levi makes some excellent points. A community ban is a drastic measure that should be based upon the input of a broad-range of community members rather than just the noticeboard regulars and editors who have a vested connection to the editor facing a ban.
  LEPRICAVARK (talk) 18:10, 9 July 2020 (UTC)[reply
  ]
• Oppose As said above, bad cases make bad law and making a new policy based on the SashiRolls CBAN is a definite example of that. There is no need to drag out the process for the vast majority of CBANS, most of which are incredibly obvious and occur after many previous discussions and attempts to modify the person's behavior. Putting hard requirements into place does not put actual protections into place and draws out the process unnecessarily. There have been two main reasons for a hard limit advanced so far: fairness and participation. Isaacl seems to advocate that letting the person proposed for the CBAN have their say is basic fairness and I understand that impulse. In practice, however, any serious (i.e., non-troll, non-battleground) CBAN comes after previous discussions at multiple venues and the person has had ample opportunity to be heard and their opposition to the proposed ban can be readily assumed. Setting a hard minimum length does not actually ensure they proposed banee gets their say. If the problem is, as Levivich and Atsme seem to state above, one of participation and notice, then again a hard minimum does not address the issue. Adding CBANs to notification mechanisms such as CENT might satisfy that concern but simple time limits do not. In both cases, a hard time limit assumes that the desired fairness/participation will automatically appear by the time of the close. This is optimistic, to say the least. Closing admins assessing consensus on serious issues can be expected to take both points into account without time limits. They are also allowed to have judgement on discussions, including bans. If an admin performed an obviously egregious close, it can be reversed and reopened but a hard time limit does not actually address the concerns addressed. At very best, it provides an opening for those concerns and at very worst it provides an opening for further abuse. Eggishorn (talk) (contrib) 19:23, 9 July 2020 (UTC)[reply]

  I didn't use the word "fairness" or mention the person who is under scrutiny. I believe determining the consensus view of the community requires patience, and so time must be allowed for the community to participate in discussion. isaacl (talk) 19:59, 9 July 2020 (UTC)[reply]

  I apologize for the misrepresentation of your position. Eggishorn (talk) (contrib) 20:04, 9 July 2020 (UTC)[reply]

  I think a very important part of the process that needs closer consideration is the time required to find diffs, read them in context, and then deliberate those findings. We should not file a case against an editor one day, and then the next day make a decision to site ban him/her without looking carefully at the evidence presented by both the prosecution and the defense. Giving it

  Systemic bias in Wikipedia that already exists. How fair is that? ^Atsme _Talk 📧 21:35, 9 July 2020 (UTC)[reply

  ]

  And how does three days, or five days, or 90 days, or whatever fixed standard anyone wants to pick, mean that such a careful consideration will happen? On the other hand, why go through such careful consideration on obvious cases? The number of !voters that comment in CBAN discussions who actually "...find diffs, read them in context, and then deliberate those findings" is already very likely a small minority and this won't be helped merely by a longer time limit. The point is: no issue with the process is a time problem. They are participation problems. Using time solutions to participation problems is a mismatch that only increases

  WP:BUREAU. Eggishorn (talk) (contrib) 23:05, 9 July 2020 (UTC)[reply

  ]

  The more time, the more participation, because the more opportunity for participation. Not everybody can comment in the first 24 or 48 hours; but some folks that might be busy on day 1 or 2 might free up on day 3, or 4, or 5. Like, for example, if they have to work, but then on the weekend, they have free time, but the weekend is three days away, they won't "get to it" until day 4. So if you close it on day 3, you don't get that person's participation. But if you leave it open for 5 days, then you will get that person's participation. Levivich ^{[dubious – discuss]} 23:20, 9 July 2020 (UTC)[reply]

  WP:POV railroad comes to mind. What constitutes an obvious case? Are you referring to sock activity? If so, then absolutely - but if it involves site banning a veteran editor, I want to see solid proof, and that editor should be given a chance to defend him/herself - just like at arbcom but not as formal. Sitebanning is a pretty serious step, so if we're not seeing blatantly obvious, repeated incivility issues and non-stop disruption, it's possible that the targeted editor simply has a different POV which has pissed off the opposition who happens to be the majority at that time. Any editor who participates in a site banning and makes an allegation against another editor they need to support such a claim with diffs or they're casting aspersions. Site banning should be closer to a judicial process not a kangaroo court in a banana republic. That's a scary proposition. Of course, I'm just one voice among many but that's what I believe. Consensus rules. ^Atsme _Talk 📧 01:36, 10 July 2020 (UTC)[reply]

  I hear what you're both saying but there's still this false equivalence between time and participation. More to the point, however, there's still unsupported assumptions about how people may or may not budget their time and how a specific limit is necessary to accommodate those assumptions. In SashiRolls's case, in the 2 days, 18 hours, and 1 minute it was open a full 69 editors managed to carve out the time to chime in, including Sashi himself 9 times. So where's the rush to judgement? That's as full and fair a hearing as any editor is likely to get for anything outside an RfA so it can't honestly be considered "too short". And even if we accept that there needs to be a longer time limit, why is 72 hours a magic number? Because I guarantee that, within a year of setting it at three days, there's going to be a potential banee that says: "this discussion was unfair because it took place over Labor Day weekend (or a bank holiday) and I was away with my family and there was a rush to judgement". And so we will have another discussion to make it four days, or five, or a week. and then within a year of that extension the same arguments will again be repeated. Let's stop the silly escalation at the start, let the closing admins exercise restrained judgement on when enough is enough. Avoid adding a rule that doesn't do what it claims and wasn't needed in the inciting incident and wouldn't have made any difference anyway. Eggishorn (talk) (contrib) 02:58, 10 July 2020 (UTC)[reply]

  That's [2 days 18 hrs] as full and fair a hearing as any editor is likely to get for anything outside an RfA so it can't honestly be considered "too short" - I disagree. It's actually the shortest time for a site ban in the last two years for a veteran editor (except for the admin who was already blocked by arbcom for socking). QuackGuru got 7 days and Farmbrough got 4 IIRC, both opposed by wider margins than Sashi's. Also, if we give 7 days for an RFA, then why can't we give 7 days for a site ban discussion? Losing a veteran editor is more important than giving someone the bit, at least IMO. 72 hours isn't really a magic number; what's magic about 7 days is that it covers an entire week and all, well, 7 days of the week, so no matter what someone's work schedule, most people will have a day off somewhere in a week and thus have time to participate. For me this is about allowing non-daily editors to participate in these discussions as much as it's about giving "the accused" time to defend themselves. This rule -- keeping it open longer -- will increase participation. It's not an opinion, it's a fact. Look at any discussion and you'll see that the longer the discussion stays open, the more editors participate. At some point it slows down or stops, but until that point, the longer it stays open, the more people post. It's true of every discussion. Including this one. If we closed this thread 6 hours ago, there'd be fewer comments in this thread. Obviously, right? Levivich ^{[dubious – discuss]} 03:36, 10 July 2020 (UTC)[reply]

  Normal standards for letting an ongoing discussion proceed remain in effect. Under most circumstances, a closer wouldn't close a discussion without hearing from the person under scrutiny, and this would continue to be the case, regardless if the minimum discussion period remains at 24 hours or is extended. isaacl (talk) 03:48, 10 July 2020 (UTC)[reply]

  TMK, the only time in two years that this assumption (more time would be granted when necessary/requested) has been tested, it's been shown false. That's why we're here, right? Levivich ^{[dubious – discuss]} 17:25, 10 July 2020 (UTC)[reply]

  Discussions are often held open beyond theoretical minimum standards when there is ongoing productive conversation. In this particular case, the editor in question was heard from, numerous times. It can definitely be argued that the conversation should have been left open for longer, but there is never a guarantee that discussions will be left to continue as long as the main subject wishes. isaacl (talk) 17:34, 10 July 2020 (UTC)[reply]

  Levivich, by eliding a significant clause in your quotation of me you've (I assume inadvertently) misrepresented what I meant. The time of the AN/I thread was not the determinant of the fair hearing, the contributions was. That thread in fact belies the assumption that time=participation and a hard limit is necessary to create participation. 69 editors is a significant enough cross-section of the active editing population for an admin to judge consensus. I've closed many RfC's, often over less clear-cut proposals, with far fewer participants and any admin worth their salt has done the same. The idea that 72 hours, or any other fixed amount of time, would have resulted in a different result in that case is pure supposition and it is equally supposition that any other CBAN proposal needs a fixed amount of time to achieve "fairness".

Losing experienced content contributors is not a reason to give troublesome editors yet more space and time to bedevil other content contributors whose time should be valued equally. We lose content contributors literally every single day. That's life. We should be interested in promoting an environment where a new editor is encouraged to stay and become an experienced content contributor and value them just as highly. The latter is not helped by giving a pass to problematic editors who wave the "productive mainspace contributor" card. Eggishorn (talk) (contrib) 19:09, 10 July 2020 (UTC)[reply]

• Support 24 hours is a 365
  WP:NOTCOMPULSORY. Yet you still should have a right to fair treatment. People have jobs and some mostly edit during the weekend etc. Someone who files the AN/I thread can collect diffs for a week if he wishes to, yet the accusee is supposed to write a defence during the day (assuming he has to sleep, 24 hrs is never full 24 hrs)? That is unfair and adds to the stress involved in the dramaboard environment. --Pudeo (talk) 20:52, 9 July 2020 (UTC)[reply
  ]
• Support a nuanced version: Objectors are right that when there's unanimity or near-unanimity to site-ban, after 24 hours, there is no need to keep it open longer. But if there's significant debate, or a low turn-out that isn't really sufficient to assess consensus, then keep it open longer. I don't care if it's 48 or 72 hours, but longer than 72 would be excessive.  — SMcCandlish ☏ ¢ 😼  22:47, 9 July 2020 (UTC)[reply]
  • We do have some editors that can really only edit on weekends, (or, for example, nurses that work 3 12s in a row on a regular basis). It's seems unfair to ban someone without giving them time to actually respond. Hobit (talk) 23:11, 9 July 2020 (UTC)[reply]
    • Most closers wouldn't close a discussion without hearing from the person under scrutiny, barring exceptional circumstances. This would continue to be the community-expected norm. isaacl (talk) 03:48, 10 July 2020 (UTC)[reply]
• Support 7 days per Levivich with one caveat: a panel of at least 3 non-self-appointed, community-approved editors and/or admins to close. The community at large is best suited to decide whether a fellow editor deserves a site ban, and Lev correctly suggests 7 days with proper notification to give us ample time to investigate and weigh in. "Wikipediholics" describes a subset of the editing pool, but many of us have busy schedules and can volunteer only a few hours per week (even that is a squeeze). There is no reason our voices aren't as valid as the 'holics'. And there is no reason that a site ban discussion should not last as long as one for article deletion. Consensus cannot be formed in 48 hours; as has been noted, those who show up right away are friends and foes, with the noticeboard regulars thrown in. A consensus from those subsets does not represent community consensus, and is actually much more likely to be swayed by less than honorable motives. By the time I found out Sashi was being discussed, the thread was already closed. I disagree that "longer than 72 hours would be excessive" when a site ban is on the table, without giving the community at large a chance to weigh in, let alone whilst denying the accused an extension to the weekend (as was the case with Sashi, who was forced to continue replying, and ended up getting sick). petrarchan47คุก 23:19, 9 July 2020 (UTC)[reply]

  I like the panel idea, Petra - that's what they're doing now to close the RfC for Fox News - a 3 person panel, and I think all 3 are admins, but TonyBallioni or someone else may know for sure - the discussion was at AN. ^Atsme _Talk 📧 01:42, 10 July 2020 (UTC)[reply]

  FWIW while we're spitballing ideas, I support a panel for a non-snow close. That is, let a single admin make a snow close if it's that clear. Otherwise, a panel. There are very few of these per year, it's not a huge burden, and editors are our most valuable resource, we should be as careful as we can be before we ban one in a non-obvious (e.g. non-snow) case. Also, since such closes will be stressful and contentious, it'll take the heat off any single admin. Levivich ^{[dubious – discuss]} 03:40, 10 July 2020 (UTC)[reply]

  I do not like mandating a multi-person close; I don't think it is a good one-size-fits-all solution. If by non-self-appointed, community-approved closer you mean one the community suggests and approves, I fear that will just lead to a very small number of suitable candidates. Since no one can volunteer to close, the community is likely to only think of a few candidates who are able to gain consensus support. It also means that a whole additional discussion has to be held to do the selection. If the potential closer candidates at the time feel like more people will be of value, they should have the freedom to choose this option, but I don't think it should be required for all cases. isaacl (talk) 04:14, 10 July 2020 (UTC)[reply]

  What isaacl said. It's a terrible idea, it's difficult to find closers as-is for discussions and mandating a panel when most of the cases have an obvious close is a waste of community time (in attempting to find a panel and otherwise). --qedk (t 愛 c) 08:31, 10 July 2020 (UTC)[reply]

• Truly an awful idea, requiring a panel, pure BURO for the sake of BURO. The vast majority of RfCs are closed by a single closer, only particularly difficult ones get panels, so using the occasional panel as a model fro CBANS in unwarranted. The peoeple supporting long or very long time periods seem to be assuming that everyone has a right to edit here. That's not the case. While anyone has the opportunity to edit here, no one has the right to do so, but the community has the right to protect itself and the encyclopedia from disruptive editors. All time spent spent discussing is time not spent editing articles or otherwise improving the encyclopedia, which is what we're here for. Seven days is appropriate for AfDs because it takes that long to get enough opinions to make a determination of consensus (and sometimes not even then). That is not the case with CBANs, which attract editors quickly.
  We're currently at a mandated 24 hours, if the consensus is for more time, 48 hours is more than enough, and if you add in Tony's "sufficient time" caveat, all bases are covered. We need to focus energy where it's need, on the encyclopedia, and not on bureeaucratic redtape. Beyond My Ken (talk) 04:30, 11 July 2020 (UTC)[reply]

• Since we cannot guarantee admins won't make bad calls, a panel close is a really good idea for site bans (except, as Lev notes, in the case of a Snow Close, determined only after 7 days).

Per Levivich

"It's not >2/3. Therefore, it's not inappropriate to close this some other way, such as "no consensus", under the closer's own logic. It's not me who's doing the counting, it's the closer. These are the only specific grounds given for the close. He makes no reference to any specific argument made by anyone, or to any specific incident, or to any specific diff, or to any specific policy. He just talks the numbers, and he gets the numbers wrong."

Per Springee

If this was a close based on the weight of the arguments then the closing shouldn't have said >2/3rds at all. It should have instead discussed why the arguments for were stronger than the arguments against... the closing admin specifically justified "consensus" based on claiming over 2/3rds were in favor. That wasn't true. I get that many feel "consensus" might start at 60% (assuming sound arguments on both sides) but when talking about a CBAN why wouldn't we want to have a higher standard? Would we accept the same weak consensusish result if this were a discussion to strip an admin of their authority? (emphasis mine)

Per Darouet, "what is lost by waiting a week to ban someone?" The goal of attaining community consensus is harmed by limiting discussion to <7 days. It is a fallacy that the majority of editors can donate time to examining a slew of diffs at any time during the week. petrarchan47คุก 19:02, 13 July 2020 (UTC)[reply]

Even the most contentious site-ban sanction discussion close of all time could only garner a "no consensus" in its review after all opinions involved. While ...we cannot guarantee admins won't make bad calls is a technically correct statement, it's a baseless point in practicality. Anyone making bad closes is always subject to community review and panel closes have historically been a waste of time and only really required when it would be a further waste of time later on, in case of very contentious discussions and the point that you clearly don't realize is that most site-ban discussions are relatively uncontentious, so your idea is just

WP:BURO for the sake of BURO. --qedk (t 愛 c) 08:30, 1 August 2020 (UTC)[reply

]

WP does not have a robust way to deal with admin misbehavior. And when it results in a serious outcome like site ban, the best idea from a content creator's perpective is to prevent it in the first place. A team makes it far less likely that admins will ignore policy.

It’s nice you have the (ample) opportunity to attempt to poke holes in my argument. Since consensus is gleened from discussion, a longer period allowing for nuanced, deeper arguments with some back-and-forth from a wider group, ensures a better result. Two or three days is long enough to survey the views of a much smaller group, though not much beyond initial thoughts. Sashi’s was a bad close. It was abruptly cut short during an uptick in activity, with the most recent !votes leaning toward opposing the ban, but that’s not what makes it unequivocally bad, it’s this:

i: Policy is non negotiable; this is especially true for mop holders, for obvious reasons

ii: Policy states per

WP:!VOTE

: It is not the vote that matters, but the reasoning behind the vote

iii: In the Sashi case,

policy

was ignored

iv: Community consensus cannot override policy per

WP:CONLIMITED

; community consensus would only be relevant if used to modify policy

We tell new editors consensus (agreement) is gauged based on the merits of the arguments, not by counting votes.

From

WP:NHC

:

[C]losers are expected and required to exercise their judgment to ensure the decision complies with the spirit of Wikipedia policy and with the project goal. A good closer will transparently explain how the decision was reached. Consensus is not determined by counting heads, but neither is it determined by the closer's own views about what is the most appropriate policy. The closer is there to judge the consensus of the community, after discarding irrelevant arguments: those that flatly contradict established policy, those based on personal opinion only, those that are logically fallacious, and those that show no understanding of the matter of issue.

In the Sashi case, the above was ignored entirely and core policy was flipped on its head (read closing statement here).

law is the backbone of the encyclopedia. petrarchan47คุก 20:16, 4 August 2020 (UTC)[reply

]

I'm poking holes at your argument because it's fundamentally wrong, you keep throwing statements like ensures a better result with no evidence for the same. The only credible background for this entire RfC is literally one siteban discussion and that itself, was a disputed bad close, not a clear-cut bad close at all - so quit citing that over and over? Similarly, I don't have much sympathy for your perspective on administrators, I myself have not been comfortable with the status accorded to admins (even before I was one) and have raised it as a problem on multiple occassions, including when supporting unbundling of a tools. It's quite apparent that it's the community itself which set up this structure, making a very hard barrier for entry and easy barrier to exit, eventually making it harder and harder for less "exceptional" or "non-controversial" editors from becoming administrators. That is a completely separate issue and you seem to be conflating the two. --qedk (t 愛 c) 13:06, 8 August 2020 (UTC)[reply]

• I'd suggest a minimum of 72 hours, any more than that would seem excessive. There should be a couple caveats on that so that if the subject of the ban request has commented and SNOW conditions are met then it can be closed after 24 hours. As an alternative, I support Tony's suggestion that a sentence be added along the line of Administrators are expected to ensure the community and any impacted editors have had sufficient time to comment before closing, even if the minimum time has been met. I oppose a panel closure for the same reasons that Tony gives above as creating an untenable procedural requirement. Generally, ban discussions aren't too complicated to close, much less so than complicated RfCs in any case. Callanecc (talk • contribs • logs) 02:59, 10 July 2020 (UTC)[reply]

  I agree that "ban discussions aren't too complicated to close" primarily because it's status quo under the adopted process but I don't agree that's how it should be done for the following reasons: 2 sides, each with a different POV + a single admin closing a discussion = potential for POV creep, intentional or otherwise. It is not an easy task to leave one's biases at login, and that makes it a bit more difficult to AGF when we're dealing with anonymity. Veteran editors in a particular topic area can easily recognize admins with strong biases in that topic area, whereas admitting to one's bias and recusing is a different story. Admins with strong biases or who may be prejudiced against certain editors because of their POV should recuse from closing a community siteban discussion, and that includes any other unilateral action at sole discretion against that editor. Another important circumstance to consider is that the editor who filed the case had plenty of time to gather both support and diffs against their opposition, and that includes the time necessary to craft a strong argument. On the other hand, the accused was unaware that any of this was taking place, and probably surprised when they received notice. Do you not see that as an ambush? The accused will naturally be in a state of upheaval knowing they are ill-prepared. How would you feel? It's pretty obvious the accused will need at least 12 hrs. to calm down, especially if they believe the filing is a WP:POV railroad attempt. They will need time to read and evaluate the allegations with a clear head, which involves reading diffs in context to refresh one's memory, and then finding diffs of their own to present a strong argument and proper defense. As for participants, uninvolved editors deserve the most weight, and that does not include the few drama board regulars who have consistently demonstrated a strong POV at AN & ANI. Prejudice comes in all forms, including preconceived notions born of one's history/interactions with the accused. The uninvolved editors need time to conduct their own research, and will hopefully steer clear of groupthink and conformity. And that's why I support more time and a panel. ^Atsme _Talk 📧 18:31, 10 July 2020 (UTC)[reply]

• I'd (also) suggest a minimum of 72 hours, and everthing else per Callanecc above. 24 hours is too short for such a serious event. Editors who are up for site banning may have enemies, and they may appear first before other independent/non-involved editors have considered the case. The additional wording suggested by Tony/Callanecc would also place a marker down against closing discussions too early, without making the process to cumbersome. Britishfinance (talk) 11:30, 10 July 2020 (UTC)[reply]
• Oppose I think a 24 minimum is sufficient. May I suggest a process similar to AFD, where borderline situations can be kept open longer as necessary to achieve an appropriate level of community involvement/consensus? -- Dolotta (talk) 13:28, 10 July 2020 (UTC)[reply]
  • Just a note, AFD are kept open seven days. PackMecEng (talk) 16:49, 10 July 2020 (UTC)[reply]
• Oppose This does not only apply to site bans and site ban discussions are rare, and come after months and months of ongoing discussions and issues. 24 hours is fine, as a minimum, it does not prevent the matter from being open longer, and there is nothing about just longer that equates to fairer, certainly not the other numbers above, they are all just arbitrary -- longer does equate to more dragging-out, and likely small group fighting. If the concern is about fairness, create/enhance review process. Alanscottwalker (talk) 14:02, 10 July 2020 (UTC)[reply]
  • Isn't 24 hours just as arbitrary as any suggestion above? Longer also does equate to more time for people to sift though evidence to form an informed decision and for the subject to read through everything, reflect, and make a proper case. Whereas the person purposing the CBAN has had plenty of time to write up a rational and assemble diffs, why should the other person have similar opportunity? PackMecEng (talk) 16:49, 10 July 2020 (UTC)[reply]
    • As I said, flat hours are arbitrary. There is no way to know, let alone enforce, what editors do with their time. And, we have no way to know whether any amount of time will be enough for any one editor. Alanscottwalker (talk) 17:09, 10 July 2020 (UTC)[reply]

      Sure we do. We can ask them, and if an editor says "give me through the weekend", then we know what amount of time will be enough for that editor. Levivich ^{[dubious – discuss]} 17:12, 10 July 2020 (UTC)[reply]

      • Not with the proposal in issue, if you want another proposal about different requirements, including asking, that will need a different RfC. Alanscottwalker (talk) 17:20, 10 July 2020 (UTC)[reply]

        I think this proposal is underpinned by the proposition that 24 hours is not enough time for any editor to respond to a site ban proposal. Levivich ^{[dubious – discuss]} 17:21, 10 July 2020 (UTC)[reply]

        Then it is based on an entirely unsupported assumption about all editors. Alanscottwalker (talk) 17:26, 10 July 2020 (UTC)[reply]

        No it's the other way around. In order to believe that 24 hours is enough time, one must believe that every editor edits every day, and that an editor can respond to a site ban proposal on the same day it's made. That is the bad assumption. We aren't even giving enough time for someone to sleep on it before responding. Levivich ^{[dubious – discuss]} 17:29, 10 July 2020 (UTC)[reply]

        • Entirely false. Nothing about the minimum requires that assumption. Nor does it account for the several different ways these proposals arise, which are often after long discussions. -- Alanscottwalker (talk) 17:33, 10 July 2020 (UTC)[reply]

        You do realize that the same editor (and literally your only example) who claimed to have no time made over 45 comments in the same period... basically a lie. Quit using this argument already. --qedk (t 愛 c) 18:55, 10 July 2020 (UTC)[reply]

        QEDK, I quote SilkTork from the re-open discussion at AN, who put it better than I can: although the user had continued to comment in the thread, these do not appear to be considered posts, but quick emotional responses. I disagree this is "basically a lie". Fundamentally, that Sashi wasn't given until the weekend per his request, is the primary reason I support re-opening that discussion, and the primary reason I support a 7-day minimum. So, sorry, you'll be hearing me (and I think others) continue to "use this argument" ("use" it for what?), or, rather, continue to raise this concern. Levivich ^{[dubious – discuss]} 19:00, 10 July 2020 (UTC)[reply]

        If an editor claims they don't have time, they don't have time, their claims would be rendered valid if they truly made none or a very few in the period, a believable claim, with a possibility for lying - but when the very claim they used to push an argument is rendered invalid by their own actions, that's equivalent to incriminating themselves. You disagree that it's a lie but it still is, the fact that they responded multiple times is equivalent to the fact that they had time to respond, logical, no? (in your own words). The distinction between "posts" and "emotional responses" is not one that can be discerned by anyone other than SR and for that reason, we must accept that all responses are of the same nature. Secondly, I don't mind hearing from you, my statement was simply advice, one that you're free to not accede to. So yes, you're free to raise your concerns always, similar to how I am entitled to respond to those concerns. --qedk (t 愛 c) 19:08, 10 July 2020 (UTC)[reply]

• Support 72 hours for making a site ban permanent. Any admin can feel free to block before the 72 hours, but then it will be considered a normal block which can be undone by any other administrator (of course, an informal discussion on the blocked user's talk page before unblocking would be best practice). Either the behavior is serious enough that it warrants a full 3-day period to allow busy people to respond, or it is run-of-the-mill indeffable behavior that should follow an "easy block, easy unblock" paradigm. -- King of ♥ ♦ ♣ ♠ 21:36, 10 July 2020 (UTC)[reply]
• Support 72 hours for making a site ban permanent per King of Hearts. -- Valjean (talk) 01:00, 11 July 2020 (UTC)[reply]
• But suppose the editor involved is a "weekend only" editor, shouldn't those 72 hours be mandated to contain a weekend day? and shouldn't that weekend day start the 72 hour clock rolling, because that's when the weekend editor would first see it? And what about if the 72 hour period contained a religious holiday which prevents the observant subject editor from being online? Should that day not count? And what if one weekend the editor went to visit their parents int he country, where there's no Internet, and came back to find that they've been banned? Maybe we should require all editors we contemplate having a banning discussion about to fill out a form which lays out their availability, and then we can negotiate the days on which discussions will take place. Or, better yet, let's have ALL Wikipedians fill out such a form, and we'll only CBAN those for whom a 72 hour ban discussion can be arranged for conveniently for all parties. Or maybe we should never ban anyone, and just let the project and the encyclopedia suffer from rampant disruption, that's really the easiest path. Beyond My Ken (talk) 04:41, 11 July 2020 (UTC)[reply]

  BMK, you articulate well the reasons why I support 7 days: it covers almost all of those situations: holiday, travel, weekend, whatever. Levivich ^{[dubious – discuss]} 05:14, 11 July 2020 (UTC)[reply]

  WP:SARCASM --qedk (t 愛 c) 06:29, 11 July 2020 (UTC)[reply

  ]

  WP:SARCASM :-) Hobit (talk) 15:25, 11 July 2020 (UTC)[reply

  ]

  WP:SARCASM --qedk (t 愛 c) 15:56, 11 July 2020 (UTC)[reply

  ]

  You win! Tabularize sarcasm ;-) Hobit (talk) 17:41, 11 July 2020 (UTC)[reply]

• Oppose as unnecessary, per Tony et al. Most CBAN discussions are more of a formality than anything, so adding more BURO doesn't really help. Administrators should be encouraged to leave discussions open longer when necessary, but I see no reason to prohibit SNOW closures after 24h. --AntiCompositeNumber (talk) 05:26, 11 July 2020 (UTC)[reply]

  Even if the initial !votes are

  WP:SNOWing, it is possible that someone who doesn't visit Wikipedia every 24 hours has a strong argument to make. We don't want to miss out on their voices, and when the the evidence in favor of the user "on trial" is not so easy to find, closing administrators would not know what "when necessary" even means; in fact, it is explicitly not their job to dig up new evidence. -- King of ♥ ♦ ♣ ♠ 17:59, 11 July 2020 (UTC)[reply

  ]

• Of course, what remains undecided is how long a discussion about how long a CBAN discussion should remain open should remain open. I think that infinity days is the only reasonable course, that way we'll get the views of Wikipedians yet to be born, which is important because we know that different generations think differently, and we wouldn't want to decide how long a CBAN discussion should last without hearing the views of those who are currently fetuses, infants, or pre-cognitive, let alone our vast reservoir of sperm and eggs. This discussion is an attempt to maximize fairness, after all, and notions of what is fair are constantly changing, so it would be wrong to decide how long this discussion should last without allowing all future Wikipedian generations to have their say about it. Beyond My Ken (talk) 05:06, 12 July 2020 (UTC)[reply]

  Sorry, I oppose your discussion length proposal. It fails to consider editors who have left the site for good in the distant past, those who have never edited, and those who were active previously but are deceased. As a compromise I propose that all future ban discussions are required to have started on January 1, 1970. Ivanvector (^Talk/_Edits) 15:15, 13 July 2020 (UTC)[reply]

• 24 hours are sufficient as a hard limit. The burden not to close an exceptionally contentious discussion too early is on the closing administrator. ~ ToBeFree (talk) 12:08, 12 July 2020 (UTC)[reply]
• Support Should be equal to, or longer than an AFD, assuming that whatever triggered the discussion isn't blatantly ongoing. Personally, when I'm not furloughed (4 months of Covid lockdown here now ...), I can easily go the entire work-week without even logging in, if it's busy ... and then there's family stuff on the weekends. Trying to rush a defence is probably going to fall flat, or come across as being defensive. I don't think many appreciate how stressful it is for a long-term editor to be in the middle of one of these processes, with real-life going on. (sigh ... looking forward to real-life again ...) Nfitz (talk) 00:12, 13 July 2020 (UTC)[reply]
• 72 hours minimum. Wikipedia editors shouldn't feel that they can't take a couple of days off without something being proposed, decided, and closed while they were out having fun. --Guy Macon (talk) 03:30, 13 July 2020 (UTC)[reply]
• 72 hours minimum with a strong preference for a week. We give articles a week minimum for discussion; it is absurd that such a serious measure as permanently removing someone from the project is considered as requiring less discussion. See also Wikipedia:Miscellany for deletion/Wikipedia:Community sanction noticeboard (second nomination). Stifle (talk) 13:59, 13 July 2020 (UTC)[reply]

  Stifle, Except putting a person (not an article) in the dock for extended criticism seems quite a bit crueler. -- Alanscottwalker (talk) 14:47, 13 July 2020 (UTC)[reply]

  Alanscottwalker A ban proposal that is clearly failing can be closed early; 72 hours would only be required to find in favor of a ban. If the user under discussion no longer wants to take the criticism, they can self-request a ban. Otherwise, I can't imagine anyone who would rather be banned than continue to receive criticism for a few days. -- King of ♥ ♦ ♣ ♠ 15:01, 13 July 2020 (UTC)[reply]

  King of Hearts, Well, I was addressing the 7 day afd comparison, but even so under the 72, it seems certain that for 48 hours relentless criticism of the person won't be closed. -- Alanscottwalker (talk) 15:13, 13 July 2020 (UTC)[reply]

  Again, regardless of length, a user under discussion can shut it down early if they consent to being banned. IMO "think of the defendant!" is not a good argument, because most people are willing to do almost anything to avoid being banned. -- King of ♥ ♦ ♣ ♠ 15:17, 13 July 2020 (UTC)[reply]

  King of Hearts, beating them into submission is the dynamic you describe. -- Alanscottwalker (talk) 15:27, 13 July 2020 (UTC)[reply]

  (edit conflict) Why would the length requirement be only to find in favour of a ban? We're saying it's likely a person's detractors will top-load the discussion early, but what's to say the opposite does not occur, that someone proposes a user be banned and all that user's allies top-load the discussion? If there's going to be a limit it should be for all ban discussions, not just the ones that are successful. Ivanvector (^Talk/_Edits) 15:19, 13 July 2020 (UTC)[reply]

  Ivanvector, Even votes that decide not to ban regularly have extended even biting criticism of the person. -- Alanscottwalker (talk) 15:22, 13 July 2020 (UTC)[reply]

  So let the discussions be snow-closed, then, in both directions. I'm not really sure I understand your point: I get that we don't want to leave subjects of ban discussions to be

  unblockables, users who should be banned but all their friends always show up in defense to shut down any proposal. Ivanvector (^Talk/_Edits) 15:32, 13 July 2020 (UTC)[reply

  ]

  Ivanvector, My comments do not promote a one-way snow close, that was King-of-Hearts. But friends can show-up at any hour, so no time limit addresses that. Alanscottwalker (talk) 15:47, 13 July 2020 (UTC)[reply]

  Because then anyone can, in bad faith, start a discussion on a user in good standing, and the discussion would have to remain open for 72 hours / a week simply because of that rule. There is inherent asymmetry in a ban discussion; banning a user requires a consensus to ban, but not banning them does not require a consensus not to ban. An additional asymmetry is that the ban proposer has all the time in the word to present the opening statement, but the user (or their defenders) may need more than 24 hours to prepare a defense.

  It is already current practice to speedy close (before 24 hours) a ban discussion that obviously will not succeed. To clarify: I believe that a ban discussion with less than ~20% support (i.e. a

  WP:ANI that even the most unblockable user cannot prevent an initial ~20% of people from agreeing with a ban when deserved. -- King of ♥ ♦ ♣ ♠ 15:45, 13 July 2020 (UTC)[reply

  ]

  King of Hearts, First, all site-ban proposals are exceedingly rare. Second, no real ban proposal comes out-of-the-blue (if some hapless person makes such a magic proposal, they'll get their heads handed to them). Alanscottwalker (talk) 15:55, 13 July 2020 (UTC)[reply]

  (edit conflict) Well, if there are enough level-headed people monitoring ANI, and presumably being level-headed includes recognizing situations which compel a discussion to be kept open and/or closed early (as generally we do expect experienced closers to be able to sense, and among which "the accused user has not edited since before this discussion started" ought to be included) then why have a minimum duration at all? Ivanvector (^Talk/_Edits) 16:01, 13 July 2020 (UTC)[reply]

  (after edit conflict) Perhaps I should suggest, if we are going to have a lengthy minimum duration to consider a ban proposal successful, that early closes be permitted only in the case of proposals which an experienced closer deems clearly frivolous, rather than just "not likely to succeed"? And also to ensure admins preserve the discretion to block users involved in the discussion, including the target, for unnecessarily disruptive behaviour (personal attacks in particular). I need to think about it but I would be more likely to support if we could move in that direction. Ivanvector (^Talk/_Edits) 16:01, 13 July 2020 (UTC)[reply]

  Yes, I would agree with that. So a support ratio of ~20% of regular contributors is prima facie evidence that the request is not frivolous, but below that threshold an admin may use discretion to determine whether a request is frivolous. -- King of ♥ ♦ ♣ ♠ 16:39, 13 July 2020 (UTC)[reply]

• Support 72 hours. I think 24 hours is too short, but I'm also not convinced a full week is necessary. A week is appropriate for some discussions -- like XFD and requested moves, where some editors might wish to wait several days before commenting -- but most AN and ANI discussions seem to taper off after three days or so. Additionally, I do think we should codify somewhere that the "defendant" should be given the right to respond, but perhaps that is a discussion for another time.
  Calidum 16:35, 13 July 2020 (UTC)[reply
  ]
  • Edit: I have to oppose the suggestion made by several editors that a three-admin panel should be required to close all ban discussions. While there may be some instances where a three-person close is appropriate, I do not feel it is necessary in general.
    Calidum 17:57, 13 July 2020 (UTC)[reply
    ]
• 72 hours minimum looks like a number which is gathering some support and it is, really a bare minimum. We are not professional editors, we do not have to be here every day. So if we get involved (the more so if unwillingly) in a discussion we should be granted a bare minimum time to participate. Closes should not be by a single admin. Obviously, no hard rule, a clear case may be close in minutes, a difficult kept over a week. Actually, this should be a *general rule* for *everything*. (it would need to be software aided to be practical, I guess) A system where a few people, even if trusted by the community, are self appointed judge, jury and executioner, is a society which lends itself to massive abuse of power. (Nothing against admins in general, I am one of them. Or of us. :-) - Nabla (talk) 17:49, 13 July 2020 (UTC)[reply]
• Support but open to options. If the concern is an active editor will cause issues during the discussion I'm OK with blocking them assuming they are allowed to reply to the discussion. I think 3 days is probably the minimum we should require. I'm unsure about cases of SNOW since the SashiRolls case started as a SNOW and ended up in a dispute over the no-consensus vs consensus. It should also be understood that longer than minimum should be the default. If the discussion is remotely active then it shouldn't be closed. Springee (talk) 19:15, 13 July 2020 (UTC)[reply]
• Support 72 hours per discussion above, with the caveats that admins have discretion to speedy close clearly frivolous ban proposals, and preserve discretion to sanction users being unreasonably disruptive in the discussion. Strongly oppose requiring multiple closers or a panel to close ban discussions: per
  beating a dead horse). Ivanvector (^Talk/_Edits) 20:15, 13 July 2020 (UTC)[reply
  ]

  Just so's I understand: Because of vociferous complaints on one editor's ban, we are now going to remove administrative discretion (the very discretion that every admin candidate gets entrusted with by passing RfA) and put into place multiple requirements and standards that....enshrine the actual practices already in place? And this is supposed to avoid

  WP:BURO? A discussion just short of 72 hours which was somehow so quick that a rigid 72 hour cutoff is needed but, wait, that's too rigid so we're going to put another policy on top of it for the exact standard of SNOW closes. I must be missing something. Instead of trying to exactly split that hair, why not just, you know, let admins do what they were entrusted in the first place to do? Eggishorn (talk) (contrib) 21:23, 13 July 2020 (UTC)[reply

  ]

  SNOW closes in favor of a ban are already prohibited, within 24 hours. This proposal simply raises the limit to 72 hours. -- King of ♥ ♦ ♣ ♠ 22:34, 13 July 2020 (UTC)[reply]

  @King of Hearts:, I understand that. What I don't understand is how it is not ludicrous to start adding rule upon rule upon rule to the already redundant process to bend over backwards so that every appearance of some unquantifiable notion of "fairness" is extended to individuals who already have a documented track record of not wanting to work together with the rest of the community. The amount of pixels spilt over SR's CBAN has not changed anything and these new rules won't prevent the same thing from happening the next time around. Adding these rules and pretending that they will proactively satisfy in the future is extremely optimistic. Far better to retain the current administrative discretion that has already been extended to users such as yourself, Ivanvector, MastCell, etc. I can understand the appeal of having a policy in place to point to but that policy already exists and by adding finer and finer gradations we only increase, not decrease, the potential for wikilawyering after the next CBAN. Bureaucracy for bureaucracy's sake is bad enough but when it is not going to provide anything more than wishy-washy hoped-for benefits against a concrete and definite cost, it's not worth it. Eggishorn (talk) (contrib) 04:00, 14 July 2020 (UTC)[reply]

  The only change here (other than everything I wrote as a tack-on proposal, nobody has yet commented on it) is changing the already-established minimum duration of 24 hours to 72 hours. I wrote as much more as I did ("caveats") just to express that extending this time limit should not also come with weakening admin discretion to pacify unruly participants, which does occasionally happen when things like this are proposed, unintended consequences and such. I don't want to see ban discussions turn into a sanctioned three-day free-for-all where all of a person's enemies come out of the woodwork to say mean things with no consequences. We have RfA for that. Ivanvector (^Talk/_Edits) 13:23, 14 July 2020 (UTC)[reply]

• Support 127 hours I'd give my right arm for that to happen. Lugnuts ^{Fire Walk with Me} 16:54, 15 July 2020 (UTC)[reply]

  Relieved to know it's just an arm you'd be giving up. ^Atsme _Talk 📧 16:57, 15 July 2020 (UTC)[reply]

  Allusion -- Alanscottwalker (talk) 18:25, 15 July 2020 (UTC)[reply]

• Support minimum 72 hours (from initial talk page notice for the accused) we need to bear in mind that not everyone has the time to respond immediately. Accusers have pretty much unlimited time to collect evidence, improve wording on initial comments etc. The accused only has 24 hours to respond in cases which are closed quickly. Real life may effect how much time they can spend responding to the allegations / problems, and not giving the accused a chance to properly defend themselves for something as major as a CBAN is not right. CBANs are a big thing and shouldn't be rushed.

I'm unsure about SNOW closes being exempt from the 72 hour minimum, but perhaps a lower time minimum for SNOW closes seems appropriate (like 30 hours or 24 hours). Enough time for a user to provide an initial response and enough time to ensure that the discussion has been seen by many editors. Dreamy Jazz ^{talk to me | my contributions} 22:18, 15 July 2020 (UTC)[reply]

• Support minimum 72 hours - I recently participated in an un-topicban discussion at ARCA that took over three months to finally unban the editor in question. If it takes three months to unban, it should take at least three days to ban or topicban. Jusdafax (talk) 10:58, 16 July 2020 (UTC)[reply]
• Support 72 hour minimum except in cases of copyvios, spamming, vandalism, socking, LTA etc., handled by one admin. But, the banned user (fired person) has a 7 day window to appeal (not 90 days or 3 years) IFF they did not have time to present their side of the story (to address the weekend warrior argument) OR if they can show the evidence does not agree with the consensus (i.e., the close was improper). Abusive use of the appeal should be shut down ASAP. This appeal would be handled by a 3 person panel, preferably with one of them being either a crat or arb (who had higher standards for election). Would prefer the panel present some sort of findings of fact and come to their own conclusion (not used as a supervote), then weigh that against the consensus. If the consensus (ban/no consen./not ban) agrees with their findings of the evidence presented and the accused's response, then the close is good. If their conclusion differs from the consensus, then the consensus must be properly weighted in their conclusion, based on policy/guidelines. (need to flush this out more, this is meant to prevent a whole discussion being over run by friends or foes.) I don't know, too much red tape and bureaucracy I guess, but firing someone is heavy. Rgrds. --
  talk) 19:38, 17 July 2020 (UTC)[reply
  ]
  • The exceptions you mention would all be handled as well by indef blocks (or other tools), to a first approximation. All the best: Rich Farmbrough 14:55, 18 July 2020 (UTC).[reply]
    
• Support 72 hour minimum per the good points made by Levivich and Andrew, the "minimum" part because of the issue raised that people don't have time to deal with this stuff in the week. Bear in mind that the "subject" can be blocked while the discussion takes place. Other important points made are the idea of some sort of quora, as discussion with minimal numbers of editors involved are not really community decisions, or if they are to be treated as such should be considered fragile and easy to overturn with a !re-vote. All the best: Rich Farmbrough 14:51, 18 July 2020 (UTC).[reply]
  
• For indef or long (6 month plus) blocks and bans, I think the least courtesy that should be offered is keep the discussion open for a reasonable duration. 72 hours at a bare minimum, to allow various editors to participate. Perhaps 7 days is appropriate, to accommodate for editors who only edit on certain days, weekends, etc. Yes, a small number of editors' votes probably won't change the result, but their comments could influence it. ProcrastinatingReader (talk) 09:40, 21 July 2020 (UTC)[reply]
• Unlike blocks, which can be enacted quickly to prevent ongoing damage or stop immediate behavior, bans can afford to be deliberative (i.e. we can discuss banning a user who is already blocked), and as such, I have no problem with support 72 hours as a reasonable minimum limit. --Jayron32 14:30, 21 July 2020 (UTC)[reply]
• Support 72 hour minimum to offset first-mover advantage (per Atsme, Pudeo) and wider publication for more severe penalties (per, e.g., Levivich. Several mentioned issues re evidence. Related to that, Levivich and I been working a (still unfinished) {{diff table}} aka {{dtable}} to facilitate and encourage more complete collection and organized presentation of evidence. Suggestions welcome. Humanengr (talk) 04:48, 23 July 2020 (UTC)[reply]
• Support 72 hours as the minimum: per Humanengr, Atsme, et al., with a strong preference for a week — I haven't been active recently, and that's due to work. I am certain that such circumstances also apply to others. The standard indefinite blocks aside, I cannot foresee any reasonable need to close a CBAN after only 24 hours. — Javert2113 (Siarad.|¤) 23:23, 23 July 2020 (UTC)[reply]
• Comment: Regarding 72 hours vs. 1 week, I would say this: First off, a clearly frivolous request can be closed immediately. 72 hours is minimum for closing a (non-frivolous) ban discussion which is obviously succeeding or failing. If more time is needed to reach a consensus, then we can allow as much as a week. -- King of ♥ ♦ ♣ ♠ 23:41, 23 July 2020 (UTC)[reply]
• Oppose
  WP:SNOW applies, and the result is obvious. No need to make the process even more complicated or arbitrary or drawn out. Also, the fact that this RfC stems from the SashiRolls case sours my stomach. Wikipedia is not SashiRoll's, or anyone's, personal appellate court. Attempts to make Wikipedia more court like only takes us down the road towards inefficiency and bureaucracy. CaptainEek ^{Edits Ho Cap'n!}⚓ 07:42, 27 July 2020 (UTC)[reply
  ]
• Oppose - 24 hours is simply a minimum window to allow for a fair review and ample participation, so that bans cannot be implemented by only a few users within the span of only a few hours. It is a sufficient timeframe to allow clear consensuses to be closed without excessive drama, and when a consensus is unclear, discussions can and do go on for as long as is necessary. It's a bit silly to hold up SashiRolls' ban as an example a failure of the 24 hour rule, because it was not closed even near to the 24 hour window, but after 66 hours, with a fairly strong consensus that is unlikely to have changed had the minimum rule been 72 hours instead, and an unusually high level of participation. Given that the discussion was more in line with with the OP's proposed 72 hours, and the fact that the 24 hour rule was not even observed, much less unfairly exploited, I fail to see how this is anything but a solution in search of a problem. ~Swarm~ ^{sting} 16:43, 30 July 2020 (UTC)[reply]
  • @Swarm and CaptainEek: To be clear, I'm not using the SashiRolls case as an example of where the 24 hour issue failed. The point of mentioning the case is that the community opinions there suggested consensus had changed regarding a 24 hour minimum, and substantial discussion about what the minimum duration ought to be had occured there. Part of me regrets mentioning it, because I think this is a useful discussion regardless of the outcome regarding SashiRolls. — Wug·a·po·des​ 00:44, 7 August 2020 (UTC)[reply]
• Comment Why not optionally allow a provisional close if consensus is clear after 24 hours but require the discussions to run a full 7 days before formerly closing. For simple cases the final close could just affirm the provisional one. This would give discussions that need time to have it, without slowing down the process for routine cases. Blocking of course is still an option during the discussion, and especially if a provisional close of ban is reached, the subject should normally be indef banned until the final outcome. PaleAqua (talk) 17:10, 30 July 2020 (UTC)[reply]
• Observation: It appears to me that there's a change afoot. No less than 3 things contribute to my views. 1. Recent RfA comments and !votes. 2. Reducing the amount of time an admin may be inactive. and 3. This proposal. It seems to me that wiki is trying to do away with "administrator's discretion" and place more emphasis on "administrator activity". It seems we're trying to exchange "trust" for "activity". Just a recent/current thought experiment I've been engaged in, but perhaps after more thought I'll think differently - one never knows. — Ched (talk) 00:42, 31 July 2020 (UTC)[reply]
• Support, for a CBAN, seven days minimum. Seven days minimum is the standard for any discussion that claims to hold the participation of the community. Shorter periods exclude a large part of the calmer section of the community that doesn’t watch and respond to the drama boards on a daily basis. Seven days for a CBAN in no way limited any admin from imposing a WP:BLOCK, and in fact, CBAN discussions should really presuppose that the user has just recently been blocked. One clear sign of over hasty closes is trial of highly related discussions that follow. A good timely close is evidenced by the respect for the finality of the closed discussion. —SmokeyJoe (talk) 09:20, 1 August 2020 (UTC)[reply]
• It's disappointing that nobody so far mentioned time zones. It is not reasonable to expect an accused editor to even notice a case in progress in less than 12 hours. That leaves little time to mount a defence even if the editor has nothing else going on in their lives. I support a minimum of 72 hours before a ban can be imposed. A frivolous case can be dismissed earlier. Zero^talk 09:47, 1 August 2020 (UTC)[reply]

  I referred to this: I do think that 48 hours is probably a better minimum length to ensure opportunity for the global community to comment. isaacl (talk) 17:18, 1 August 2020 (UTC)[reply]

• Support 48 hours at a minimum. A week would not be unreasonable. Carrite (talk) 16:50, 1 August 2020 (UTC)[reply]
• Weak support 2 or 3 days for the reasons given in the previous discussions at least for site bans which is clearly a serious matter, temporary injunctions can be imposed pending discussion if needed. Crouch, Swale (talk) 21:23, 1 August 2020 (UTC)[reply]
• Oppose There is no evidence that more bureaucracy means better outcomes. In fact, the opposite seems true. There is nothing wrong with a ban discussion taking a few days, nor anything wrong with it taking 24 hours. Each case is different. Hamstringing the community, and admin, will not guarantee "better" outcomes. This ONE case is the basis for the RFC and as usual, basing a change due to ONE case is a bad idea. Dennis Brown - 2¢ 01:04, 3 August 2020 (UTC)[reply]
• Support 72 hours minimum. I am entirely persuaded that yes, we shouldnt be able to remove a person in a day when it often takes a week to get rid of articles. Only in death does duty end (talk) 08:00, 3 August 2020 (UTC)[reply]
• Oppose 72 hours minimum - as Dennis Brown and others have pointed out, this is based on one case and that's not a good reason for such a major change, and would hamstring the community and Admins unnecessarily. I can't get my head around the argument relating to AfDs, a non-notable argument is no where near the problem a disruptive editor is. And if somehow a mistake is made, then surely it can and will be fixed on appeal. I'm happy with a 24 hour minimum with the caveat that that can be shortened in egregious cases. Doug Weller talk 14:04, 8 August 2020 (UTC)[reply]

  @Doug Weller: If an editor is really causing acute disruption, such that 48 hours are going to make a difference, then an admin should simply block them of their own accord. In general these kinds of ban discussions are for people who cause chronic disruption, and 48 more hours of them isn't the end of the world. -- King of ♥ ♦ ♣ ♠ 17:01, 8 August 2020 (UTC)[reply]

  @King of Hearts: I suspect that any Admin who did that would be accused by someone of pre-judging the issue. Doug Weller talk 17:36, 8 August 2020 (UTC)[reply]

  We currently do hold ban discussions on users who are already blocked. So the typical sequence would be to block the user to stop the immediate disruption, and then start up a ban discussion. Increasing the length from 24 to 72 hours wouldn't change that. -- King of ♥ ♦ ♣ ♠ 17:38, 8 August 2020 (UTC)[reply]

• I support a 72 hour minimum, eith a normal defualt of 7 days. The point about allowing editors whyo only can or normally dom edit on cetian daysn of the week a reasonable chance to participate in a potentially vital discussion has meerit. So does the point that the normal 7 dday minimum for AfD is comperable. Moreover, and rather more than in an AfD, back-and-foprth discussion, in which editors attempt to rebut or respond to points made by others is of value, and tht takes time. So does proper reading and analysis of multiple diffs in any complicted case. A 24 hour discussion is biased toward those editors who are active on a daily basis, and they are not a represenative subset. In caases that are at all contentious a panel close would be a good idea, but I don't see a need to mandate it. In cases of urgent disruption, a block need notr wait on a ban discussion, as mentioned by King of Hearts and others above. 24 hours is just as arbitrary a tiem persiod as 72 hours or 7 days, and changing from one arbitrary minimumn mto a different one is in no way an increase of BURO. There is still a limit. Note that in the SashiRolls case the early response looked significantly different than the discussion just befre it was closed. This i8s evidence that an early apparent consensus may not truly mean as wide a consensus as it seems. DES ^(talk)_{DESiegel Contribs} 15:54, 10 August 2020 (UTC)[reply]
• Oppose. 24 hours is more than enough time for the vast majority of cases. As for the few exceptions, hard cases make bad law. Jayjg ^(talk) 19:28, 10 August 2020 (UTC)[reply]
• More time is needed, also improvements to the process if it is to be regarded as legitimate. If it's urgent, a block can be used to prevent disruption.
  CIR", incorrectly provides that as the only basis for a ban, and seems to have been structured so RTG would be expected to reply to everything and other editors to just support (or oppose) without reading properly. As expected, the replies by RTG are also long, and there is support for a ban just because people think it's a waste of their time having to read it. Because of this lack of structure, discussions are dominated by people with strong opinions for or against banning, who are involved in disputes with the editor, or to show support for the editor who is proposing the ban. Requests such as this should be rejected as they would be by the Arbitration Committee, instead they should be summarised, a valid reason should be provided and, it there is any opposition to a ban, alternatives proposed. If there is still no consensus (as in the SashiRolls discussion) then an arbitration case would be necessary. Peter James (talk) 22:23, 11 August 2020 (UTC)[reply
  ]

  Your insinuation that people voting did not read the arguments and just picked a side is very misleading and more of an accusation than the claim of

  WP:CIR (not a personal attack, to clarify) which was very apparent from the thread and prior behaviour and the fact that you can't see it seems to be a problem on your part and not on the community's. That thread is probably the finest reasoning for why added bureaucracy is bad. --qedk (t 愛 c) 08:24, 13 August 2020 (UTC)[reply

  ]