Given Jeff's recent blog post about hellbanning users, I must say I think the idea (making that user's content visible to only them) is a good one.

However, the people in whose hands that power will be placed terrify me.

Jeff is saying that he doesn't like that bans can degenerate into meta-discussions about the fairness of banning, and in that respect I agree with him. But the meta-discussions do accomplish one thing -- they get more than one person to look at the ban and weigh in on what they think.

This isn't that I don't trust moderators (or, indeed, diamond moderators), but I don't trust any one single human being with an irrevocable and complete banhammer, where there's no potential to dispute things. (And if the user doesn't know they're banned, there's no potential for dispute)

Basically, I'd sleep a lot more soundly at night knowing that if one of these "enhanced bans" are put into place, that at least two people have looked at the "problem child", and only after more than one person has signed off on such a ban, does that go into effect. I don't care exactly what that process is, but I don't want to have to worry about pissing off a moderator and getting banned as a result.

We all get pissed off here -- it's part of any community. But "the nuclear option(s)" need(s) to be decided upon by someone with a cool head, not someone who's pissed off.

  • 3
    "I think everyone has a right to know what sort of jail their community uses, even these secret, invisible ones." Sounds a lot like installing the servers in Guantanamo Bay... not the path I would like to see for SE.
    – Aleadam
    Commented Jun 6, 2011 at 7:25
  • 1
    @Aleadam: At least the "users" know they've been detained, and in Gitmo the majority of the time everyone else knows they've been detained. Commented Jun 6, 2011 at 9:23
  • 1
    It might be worth mentioning that hellbanning, slowbanning, errorbanning and randombanning are widely used on reddit and hacker news, among other sites. Usually it is kept secret and any posts complaining about it are deleted as soon as possible. I fear that this may be the case for SE too. One thing is the official stance, and another thing is what's actually there.
    – siamii
    Commented Jan 18, 2012 at 14:07

7 Answers 7


Hellbanning, slowbanning, errorbanning, and randombanning are all things we have never experimented with and do not intend to.

It is very hard to make the argument that using extreme evil to battle evil makes the internet better.

Our current measures of banning and stopping the huge flow of extreme low quality posts seem to be doing fine. In fact, I think we are doing way better than we were 3 months ago.

We also seem to be fine dealing with people complaining about "why kan i nt aks Qestion URGNT?".

I think we can stop the flow of crap by improving our existing systems, while being transparent about what it is we are doing.

When we stop people from asking questions, we stop them from asking questions; we do not put them in a virtual reality sandbox.

  • 8
    Err.. then why is Jeff proposing that on his blog? Commented Jun 5, 2011 at 7:40
  • 9
    @Billy, maybe because he's wondering aloud? "This might naturally lead one to wonder – what if we could suspend problematic users without anyone knowing they had been suspended?" - and is at the same time being transparent about what the SE sites do (timed suspensions)
    – clairesuzy
    Commented Jun 5, 2011 at 7:55
  • 17
    @clairesuzy: Often though, Jeff's "wonderings aloud" become official SE policy. Commented Jun 5, 2011 at 9:11
  • 1
    @Billy as @clairesuzy says, my guess is that he is just conducting some sort of thought experiment out loud.
    – waffles
    Commented Jun 5, 2011 at 22:39

Short Version: I have deep reservations about putting in a system that attempts to promote good behavior by behaving badly.

Let me start by pointing out that, to some degree, we already take out the trash without you knowing about it, and without getting into endless discussions on meta every time it happens. This is possible because, as a community, we have come to a consensus over time about

  1. What the vision of StackOverflow is, and
  2. What behaviors are consistent with that vision.

So while not everyone agrees with, for example, moderators unilaterally closing (and in some cases) deleting questions without debate, we do it anyway, because the community (for the most part) has already decided, via a democratic process that we want to live in a nice place, not one overrun with broken windows.

Recently, SE has begun implementing certain control mechanisms that are not only invisible to most users, but are also automatic. This includes the "low quality" IP ban that you may have seen new users complain on Meta about. When this was first implemented, there was a debate on meta, not about the relative merits of such a system, but about the error message that was being shown to these users:

Oops! Your question couldn't be submitted because:
Sorry, we can't accept this question.

The error message was eventually changed to this:

Oops! Your question couldn't be submitted because:
It does not meet our quality standards.

If this had happened on a non-programmers site, the error message might have never gotten changed. But programmers are a pedantic lot, and since the error message didn't actually say anything, the community rightly challenged it.

Hellbanning takes this innocent bit of obfuscation, and raises it to a whole new level of dishonesty.

How can you expect your community members to act with integrity, when you can't do so yourselves?

Yes, representative democracy is messy. But would you rather have this??

  • 3
    I confess I'm having a hard time understanding what you are arguing in the last part of your question. Are you arguing that hellbanning is positive or negative? Are you arguing that automatic control mechanisms are good or bad? I see the examples, but what lessons are you drawing from them, and why? I apologize if I'm being dense.
    – D.W.
    Commented Jun 4, 2011 at 18:53
  • 12
    @D.W: I am arguing that hellbanning is patently dishonest, and you shouldn't expect your community members to have good behavior if you can't be well-behaved yourself. See Jeff's blog entry for a detailed description of hellbanning.
    – user102937
    Commented Jun 4, 2011 at 18:57
  • 1
    Perhaps the first question I should have asked is "Should we implement hellbanning? :P" I thought it was a good idea but your answer has convinced me otherwise. Commented Jun 4, 2011 at 19:07
  • 1
    I can't say I'm a fan of it either, but sometimes you can't choose between a good solution and a bad solution to a problem, just the one that you think sucks the least.
    – Rob Moir
    Commented Jun 6, 2011 at 8:53

Hellbanning shows a real lack of integrity.

Everything should be open and transparent. Users should be given the chance to change their behaviour, and if they're hellbanned they won't even know they need to change. If they're bad enough to hellban, then they're bad enough to ban. We just need exact, objective reasons and then if they don't change after a few suspensions, then really ban them, don't fake them out.

All hellbanning does is keep those users from complaining elsewhere, and maybe we need to make sure they can complain elsewhere, so that banning doesn't become just a matter of course.

Every ban should be carefully thought out on a case-by-case basis, based on objective standards.

  • 28
    I think it's fairly obvious that such a process would be reserved for those disruptive users for whom warnings and suspensions don't work. You say "if they don't change after a few suspensions, then really ban them", but that doesn't work either because they can just create a new identity and come back. If they don't know they've been "banned", then they'll eventually just get bored and leave.
    – Aarobot
    Commented Jun 4, 2011 at 19:13
  • You could require real names and photos... Doh!
    – user291305
    Commented Dec 15, 2016 at 19:15

I definitely agree that there should be a process to follow for this special kind of ban. My guess is that hellbanning will be a last resort, after several rounds of increasingly lengthy regular timed suspensions have already been tried.

As for checks and balances, there is a list of currently suspended users that all moderators can see, and any one of us can lift a suspension that another has placed. We also all get a system message when a user is placed in timed suspension, so it would be difficult for any one of us to suspend someone without cause.

  • 1
    How about this: meta.stackexchange.com/questions/93786/…
    – Predator
    Commented Jun 4, 2011 at 18:14
  • 7
    @Gens: Is that a problem that actually needs to be solved? I've been on SO for nearly three years now, as a normal community member and as a moderator, and I have yet to see a moderator perform a "wrong" action. I've seen disputed actions, but they are rare, and are generally reversed if there is a doubt.
    – user102937
    Commented Jun 4, 2011 at 18:24
  • 2
    @Gens: I think every time I rightfully closed a question my PKI score would drop, making it a pretty worthless metric. I prefer to have open discussions about specific moderator behavior. Commented Jun 4, 2011 at 18:26
  • 2
    @Gens: I have no doubt that the majority are rational and smart, but most people aren't even aware of mod actions unless they're on the receiving end of them. Commented Jun 4, 2011 at 18:37
  • 2
    @Robert: It's that "generally reversed if there is a doubt" bit that I'm pointing out here. If nobody knows by the person who enacted the ban, then there's no doubt that can ever be had -- only the moderator knows! Commented Jun 4, 2011 at 18:51
  • 1
    @Bill the Lizard: I see. The existing measures in place do make me sleep a bit more soundly, but I would feel better if two people had to acknowledge before something this drastic happened. I'm thinking along the lines of the "silence does not equal consent" rule. Commented Jun 4, 2011 at 18:54
  • 3
    @jcolebrand: That is sometimes the case (that things are discussed among mods beforehand) -- however at present time that is not enforced. There's nothing stopping a moderator from just suspending someone they don't like. Luckily for us, they generally don't, but I'm a fan of the system enforcing things where it's simple to do. Commented Jun 4, 2011 at 19:06
  • 4
    @jcolebrand: The difference between this scenario and the President of the United States is that everyone knows (or has a means of knowing, if they wish) the decisions the POTUS makes. Nobody even is aware of the decisions under the proposed hellbanning system. Just because you're generally elected by the community, or generally trusted, does not mean that you are infallible. Commented Jun 4, 2011 at 19:10
  • 3
    @jcolebrand: I can't show a case of abuse of a system that has yet to be implemented. (And even if the system were to be implemented, there'd be no way for me to show that abuse, because the system by design hides the fact that it's in use from everyone) Commented Jun 4, 2011 at 19:12
  • 3
    @jcolebrand, he's not bikeshedding, there have been plenty of examples in the past, most of which can't even be shown because they've been deleted off the site (rightly so). Please don't try to derail his point, by asking for specifics, when his point is general. Address the general concept. And of course this is a conversation about proposed future actions, so there can be no examples. Commented Jun 4, 2011 at 19:13
  • 4
    @Billy: I'd be in favor of getting at least one other moderator's consent before hellbanning someone. It's an extreme form of suspension, and I don't think it's too much to ask us to at least confer with other moderators before using it. If Jeff decides to go ahead with hellbanning, I'll make this a feature request. Commented Jun 4, 2011 at 19:26
  • 5
    @Billy @BillT @Gens @Lance @Robery so we're clear, I'm all for hellbanning being something the dev team has to carry on their own, meaning, don't make it a mod tool. The devs have enough to worry about without looking on individual sites for malicious users, so it would be the case of the moderator having to provide evidence to the devs of reason for hellbanning. That requires a valid argument and a strong case. Ergo, I'm not just saying "hell yes let's hellban everyone" but I'm saying ffs quit trying to micromanage the moderators. Go get elected yourself and participate as a mod.
    – jcolebrand
    Commented Jun 4, 2011 at 19:31
  • 2
    @jcolebrand: I don't believe I said I wanted to micromanage mods anywhere. Just that the proposed hellbanning needs to be confirmed by at least two people. Don't lump me in with Gens' moderator points plan thing. I don't agree with that either. Commented Jun 4, 2011 at 19:40
  • 3
    @jcolebrand: I believe the feature has been proposed and devised. That's Jeff's blog post. I'm not saying there's no way for a mod to make a rational decision on their own. I'm saying it's possible for them to make an irrational decision on their own. It's possible if there are two or more too of course, but much less likely. I'm not saying any existing feature or moderator tool should change, merely that if the proposed new feature is implemented, because it is not transparent, it needs stronger checks in place. Commented Jun 4, 2011 at 19:47
  • 5
    @jcolebrand: Okay, whatever. If you want to put words in my mouth, and then judge what I say based not on what I say, but what you said I said, so be it. I'm done. That way lies insanity. Commented Jun 4, 2011 at 19:52

As others have pointed out, I didn't read Jeff's post as a feature announcement but as a general musing.

That said, I share all the reservations expressed here.

On another note, maybe I'm being thick, but don't hellbans only work on people who don't know that they exist? Because an easy way to tell would be to logout and go looking for your questions?

EDIT The other thing which bothers me about hellbanning is 'wasting' all that time implementing what amounts to an exotic Easter egg - for users you don't even want... And all that just to avoid discussions which are mostly fuelled by lack of trust.

It's difficult to balance the conflicting needs for transparency (letting people air their doubts and informing them appropriately) and tranquillity (not being drowned in vitriol and 'meta' noise), but I strongly doubt that hellbanning is the solution.

  • I don't think the type of users who get hellbanned usually know enough to do this. But a professional troll might, I suppose. Commented Jun 6, 2011 at 6:19
  • @Jeff, ok, so I'm not too thick, that's reassuring :)
    – Benjol
    Commented Jun 6, 2011 at 6:48

Hellbanning would not work well as:

  • People access the sites from more then one IP address
  • People can check if their post is on the site via their favourite search engine
  • There is no need to be logged on to access the site.

Therefore it is very likely that sooner or later a person that had been "hellbanned" would find out.

  • 1
    If they find don't find out in the "sooner" timespan, then chances are they have already left out of boredom and don't reach the "later" timespan in which they would find out. Commented Jun 9, 2011 at 7:39

First, I think hellbanning is a useful tool in "private" forums. (I'll define private as "obviously owned by a single person or small group"). In that context, I treat their forum as part of their house - they can choose the rules. So, for example, if someone is really honking Jeff off on the Coding Horror forum... I'd say it's completely within his rights to make the offender hellbanned, if Jeff chooses.

Around here, I don't think the system would work as well - mainly because hellbanning only works if people aren't aware that it's happening. With a larger group of moderators (who would all have to be aware, if not actively agreeing), the required discretion wouldn't be possible. Hellbanning with the offender's knowledge is just read-only access.

So, to answer the question, I'd say that hellbanning isn't the right tool in our toolbox.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .