Suspension, Ban or Hellban?


Note that there’s been a relevant discussion started on Meta StackOverflow here:


all users are equal, apart from some users who are more equal than other ones.


The Great Brain - great reference!


I’m one of the relics from the BBS era, and it wasn’t just Citadel that did the progressive levels of bans. 8/10 boards that you would call - especially those that accepted money for better access had something like that in place.

I made a change, early on to WwiV that would randomly inject high bit ASCII into a user’s editor any time they tried to post a message or reply. There was a function key that allowed the SysOp to do that manually, but nothing automatically based on an access restriction. I used that for people who posted mostly flames or noise, but really helped drive the on line games and file transfers. Usually, those users would get sick of the ‘noise’ and just go play games or upload files.

File leeches were a major problem. If you set your system up so that users had to participate in messaging before downloading stuff, they’d give you crap that you’d end up deleting and go get what they wanted. For these special people, I’d wait until their download started then pick up the phone and start dialing “mary had a little lamb” until even Zmodem couldn’t make sense out of it. They’d give up and go away, instead of trying to make a bunch of new accounts that they saw as a likely waste of time.

I had an ‘invisible’ bit that I could set, but I can’t ever remember finding a good reason to use it. Since we networked message boards via store and forward, it was really easy for the user to figure out what happened, get mad and become even more annoying.

It wasn’t just a hobby for me, not long after launching I went massively multi line and started seeing a profit after Ma Bell and the electric company got their share. I had to keep the place relatively restriction free without allowing a few people to make enough noise to turn away paying customers.

When you come down to it, the real art here is to gently nudge people into doing the things that they are good at while keeping out the people who just have nothing positive to offer.

Sorry for the book, you made me feel a bit nostalgic.


I’ve been a victim (and I find that wording correct) of hellbanning before. It was known to me pretty soon simply because I had a friend on the board, which on that very same day eventually lead to me to find my posts weren’t being seen by him.

The reason I got hellbanned on this particular news site was because I was somewhat critic of the journalist integrity of one of the staff members. I wasn’t abusive in any way. Neither I was motivated by some kind of trolling behavior. Just, if you will, one of those difficult members of the community who may sometimes become inconvenient.

Now, what this technique revealed to me is that this puts a lot of power in the hands of whatever moderators/admins are in charge of a community. The fact this is unknown to anybody allows for all types of abuse. Including quieting down inconvenient voices. That one may guarantee they will never abuse this type of ban, serves very little purpose. Motivations play a significant role in thwarting initial good intentions into seemingly correct actions that are no more than abuses of power. Our brain is very good in entering defensive mode when we are performing bad actions and finding justifications. It can be ultimately said that bad people don’t necessarily feel they are doing bad things. Good people find it even easier.

For any community praising itself of following a democratic role, this presents another problem. Democracy isn’t simply a set of values around the principle of equality. It comprises too principles of Justice. Not giving someone even the right of being warned that an action was taken against them is not democratic. It’s dictatorial. I can understand that some limits may be imposed (like not giving the person a chance to defend themselves) due to time, personnel or technical circumstances. But there’s something fundamentally wrong about a community that includes a mechanism that punishes bad behavior without informing the punished of this decision.

I could never support hellbanning. I find it vicious and, since we are discussing this in terms of a community that is being spoken of as democratic, immoral.


You say that you try to uphold democratic values but I see no evidence of the recognition of anyone’s rights, or any kind of due process that you or any other moderator employs before punishing your “citizens” with your various and cruel forms of ostracism.

So maybe what you are apologizing for in your essay is that you fail to run a republic based on laws which all its citizens must follow, and which no one is above.

I suspect that what you actually run is a pure democracy where the tyranny of the herded majority is completely unrestrained.

I myself have engaged in online communities since the mid 80s, on a topic that remains one of the nastiest on the Internet: Scientology. I’ve been kicked off of many, and they all share the same self-satisfied smugness - either Pro-Scientology or Con - that they are protecting the rest of the community from the “disruptive ones”.

The fact is that democracies can not stand the true individual who willfully stands apart from the herd. Look what they did to Socrates in the democracy of Athens 2500 years ago.

And look - People like you still are still doing it today.

For an online community to uphold all the ideals you seem to wish to forward, you must have due process - an objective, fair method of justice, where one’s accusers must present their specific accusations to those accused, and for objective rules of evidence to be fairly applied to each, among other rudimentary system of laws which are built into republics.

But that’s too expensive in terms of time, energy, and tolerance for views which board owners and moderators would rather wipe with toilet paper from the typical message board.

So don’t kid yourself, or anyone else. You do not uphold democratic ideals, or recognize and uphold anyone’s rights to the freedom of thought, or to the freedom of speech. You run a tyranny, a cruel and stupid tyranny of the majority, just like the Athens which destroyed Socrates.

And you do this only because you can’t be bothered to run anything else.

So if you want to know what you are doing to people, and why people react so strongly to your stunted hubris, read this article:



The problem with slow-banning or error-banning is that the person may take to the social media streets bitching about how slow/crappy your site is, and while you might be laughing inside, you really don’t want that do you?

I’m a big fan of hell banning.



All of these seem like permanent solutions to hide the bad community members.
Since they, in theory, don’t know what is going on, they won’t learn a lesson

That depends on what your goal is. If it’s to act as a user etiquette education site then you’ve failed, but if it’s to provide an enjoyable experience for participants then you’ve succeeded. In any case I suspect that the sort of people who need to be hellbanned aren’t educatable with anything less than a 2x4.


Hellbanning, slowbanning, etc. are counterproductive. Yes, they can be very effective at frustrating and isolating problem users, but at a cost:

  • Legitimate users expend energy and may become frustrated trying to figure out whether they are hell/slow/error-banned or experiencing technical difficulties.
  • Users accidentally caught in these types of restrictions feel duped and cheated, because they couldn't be immediately sure of the problem and address it in an immediately obvious way.
  • You can't legitimately penalize ban-jumpers (even intentional ones) because the system is designed so that the feedback to banned users is ambiguous -- they (theoretically) aren't supposed to know whether or not they are banned.
  • Someone (legitimately or otherwise) restricted in this way and his/her friends and sock puppets will often take the action to be a particularly manipulative form of censorship, which will lead to ban jumping in an attempt to circumvent it (it's easy to counter, but not before it starts drama among other users) or lots of meta-discussion in other places that nonetheless raises drama in your community.

None of these problems are insurmountable or intolerable. However, they do seem to generate more noise for community members, and more work for moderators than straightforward, transparent bans do when implemented properly.

“When implemented properly” is the key phrase here, of course. It’s hard for moderators not to have the last word – we naturally want to be seen as fair, want people (especially the one being banned) to understand the actions we take, and so the natural reaction is to feed the drama beast. Minimizing ban-related drama involves:

  • Having a clear, subjective moderation policy (objective ones are too easily gamed) that is easy to reference and to refer users to when you explain the reason for moderation actions on their user account and/or content.
  • Knowing when to disengage from the meta-conversation about bans, etc.
  • Including not just *procedures* in the moderation policy, but *goals* as well. When the motives behind moderation actions are well understood, there's less fuss and rules lawyering from the productive users. (The problem users will always generate noise -- that's how they came to be called "problem users" to begin with.)
  • Blowing people off (on the occasions that blowing them off is called for) with an action they can take, e.g. "This is how this forum is run, if you are looking for something different, I encourage you to found your own according to your own standards." or " is a democracy; if you disapprove of moderation policies here you are encouraged to nominate yourself in our next moderator election." Either they choose to do nothing (most do) and it's not your fault, or they choose to take action and their forum/nomination flops (also not your fault).
  • Maintaining a policy of never discussing discipline of any user with any other user.

These tips won’t end the ban-related drama – nothing does that – but they can help mitigate it. Rather than avoiding the issue with hellbans and the like, dealing with it directly and moving on quickly won’t let things fester. More importantly, the direct approach prevents legitimate users from interpreting every lag or lack of response as a possible reprimand or an attempt to manipulate the voice of the community.


Remember that these are common techniques used in Chinese censorship, too.


Don’t be sly. If someone is banned, they’re banned. They need to know it. Let the person know they’re being banned. What you don’t need to do is let everyone else know. This isn’t the middle ages where we flail people in the public square for entertainment purposes. If someone needs to be banned for a while, the whole world doesn’t have to know.

Another possibility is putting people’s posts on moderation. They can still post, but nothing shows up unless a moderator approves their post. Have all the people who’ve earned “trusted user” have the power to approve moderated posts. You can even set it up, so they can’t reject a post: Either approve a post, or pass on it. Maybe another moderator will approve it.

If you’re worried about arbitrary abuse of powers, create an appeal process where the user can plead their case before other moderators.

By the way, instead of banned, why not call it a “timeout”?


This sort of reminds me how they ban people on 4chan. One time I used it on a cell phone. And they banned me. The next time I browsed the site they gave me a message saying “Oh noes, I can’t browse 4chan on my cell phone :o!!!” and also a message saying “you are banned”. They obviously didnt like people browsing their site with cell phones. I dont comment very much because I’m afraid I might make a comment that’s not of high quality. Though I love your site, I linked to you from my blog.


Welcome to the wonderful walled garden with the hidden fence.


I once implemented a full justice system on a forum. Mods (police) would give tickets to the troublemakers citing a reason and link to offending post. Then the user was prompted to plead guilty or innocent. If guilty they would be fined (currency/reputation), and if innocent a forum post would be created in Court where they could plead their case. Mods would then vote for majority guilty/innocent or an admin/Judge could make a final ruling. If a user can’t afford their fine, they are jailed, where they can only post in the Prison forum until their time is up (time and fine increased with each infraction). It actually worked really well and was a fun experiment. Only problem was that some people intentionally got in trouble to talk to inmates, so I had to implement a limited use “Just Visiting” card.


Although I confess a quiver of delight at the ‘hella hellban’ (where hellbanned users live in a world composed entirely of hellbanned users), it just seems too Prohibition-y for me; apart from exploit/hack attempts, the only time I ban anyone is for spamming iff the spam is obvious.

If you were running a ‘walled garden’ forum - for Apple users, religious nutjobs, .mil/law enforcement or some other ‘correct line’ ideological straitjacket - maybe in those retarded cases the barely-adequate psychic defences of the readership need to be protected from the cognitive dissonance that can be generated by the idea-marketplace. But apart from that, I favour ‘open slather’ - where good ideas win and bad ideas get ignored and go die in a corner after writing a final batshit-insane 1500-word single-para allcaps post.

Otherwise you’re saying it’s OK to put Jews in the ghetto. And much as I deplore anybody who thinks that they’re superior to me on the basis of a foreskin-swap first organised by an Iron Age goatherd… well, ghettoes are bad.

As Chomsky wrote - if we don’t believe in freedom of expression for those whose ideas we despise, we don’t really believe in it at all.

Ideas, /b/rothers… it’s all about the ideas. The ban paradigm stems from the same idea-sump as the worst aspects of political correctness: it implicitly asserts that the poor frail reader is a delicate flower who is incapable of overlooking the horror of a few words on a webpage. That’s TEH ghey, frankly.


All these comparisons with real life democracy are all too childish. Yet so often triggered !

If you look just a little bit at it, the real life game and the online community game has so few in common, in terms of incentive, punishments and behaviors !
In real life actions have heavy and very real consequences.
Real life laws have been crafted for centuries of all types of experiences entangled all together. And it’s far from perfect. Sometimes don’t you wish real life societies had the efficiency of some online communities ?

So before one goes Rousseau, crying out loud, calling the admin a tyran and demanding a grand jury and full transparency for every single case, maybe one could think and comes to realise why democracy is not just a cure it all patch that can be mindlessly and equaly applied with success to any group of people. (Not likely to happen)

Is the benevolent moderator a judge, a highly trained professional well paid by your taxes money , just to behave in perfect equity ?
Is the user a citizen, paying taxes, whose misbehavior could bring him to real jail, someone so involved that the website owes him explanations for every move ?

I 'd stop citizening and keep using or not, but in peace !


This sounds like an extension of the methods that appear in Phil and Alex’s Guide to Web Publishing.


If you want it to be democratic, how about going to the home of democracy - ancient Greece - and reintroducing ostracism:

I mean, we can vote on most things, and communities would pretty rapidly exclude troublemakers - thus reducing the burden of the ‘who has been banned and why’ conversation.

Just a thought.


Just as an aside, why the snidey comment about clueless programmers? When I read the first paragraph about hellbanning I thought what a cool idea. And if someone independently reinvents it then good for them. We are not all so completely immersed in social software.

But thinking about all these forms of censorship some more I just don’t like it. It all seems to stem from the basic assumption that “we”, the community are all calm, rational people and that “they” are just troublemakers and trollers. As someone above mentioned though what if “we” are the scientologists? Or what if “we” are some protest group with an inclination to extremism and “they” are the small minority saying hang on, let’s just calm down a bit.

There is such a danger of a community becoming ever more self-defining and rejecting minority views.

What’s good for the online world should apply to the real world too. What about freedom of speech? Should the majority of the population be able to silence a minority because they don’t agree with their views? Perhaps protest marches should be banned if they risk upsetting the majority?

Dunno, there’s something smug and elitist about it.

What I’d most like is for everyone to be able to rate a comment either positive or negative (or maybe even with a tag). Then display both the positive and negative counts and let the user sort on either. Because a high negative count may be interesting - what did that person say to create so much feeling?


Since accounts are free, banning a user is suboptimal. As soon as he notices, he got banned, he can just create a new account and stir up trouble again. Thus a hidden ban is definitely preferable to one, where the user knows he got banned. However, all the bans you mentioned above will sooner or later cause the user to notice he got banned, so all these bans are suboptimal, too. I would personally go with an gradual hellban system instead.

Every user gets a trouble counter. People with enough reputation (or possible only moderators) can vote to ban a user; if a user gets enough votes, his trouble counter is increased by one. On the other hand, you said you are against bans for life, so you could have an automated unban system: Every 7 days the trouble counter of a user is decreased by one, so if a user behaves okay for long enough, his trouble counter will eventually reach zero again.

The trouble counter is used to calculate a partial hellban: A trouble counter of 1 means 50%, a trouble counter of 2 means 75%, a trouble counter of 3 means 87.5% and so on. 50% means that 50% of everything a user does (asking a question, answering a question, posting a comment) is hellbanned or IOW, only every second question/answer/commend of this user is ever visible to other users, the other half is visible to the user himself (and maybe moderators), but not to any other community members. That way a user can stir up much less trouble than he could before. If he still stirs up too much trouble, his trouble counter will increase and even more of his contributions are hellbanned. However, since not all is hellbanned immediately, he will still get feedback to some contributions and thus won’t immediately notice he got banned, as he cannot tell if he got no feedback to other contributions because of a ban or just because people keep ignoring him.

This form of soft hellban is much more democratic than a full hellban, since you are not taking any right away of this user, you only “limit” his rights temporarily. It is like putting someone in jail, which is not supposed to take his right of freedom away, but to temporarily limit it (even a prisoner has a certain degree of freedom left in most countries; a lot less than he used to of course). In your case, you are limiting this user’s right of free speech, not by taking this right away, but only by making him produce less noise within a certain community for a certain amount of time.