6 ms·

> "It affects only a few comments which are unlikely to be particularly valuable anyway." More importantly, it affects the occasional person who makes mostly g

> "It affects only a few comments which are unlikely to be particularly valuable anyway." More importantly, it affects the occasional person who makes mostly good comments but got hellbanned 3 years ago for something that probably could have been handled with a simple downvote and a "please don't do this, here's what the guidelines say" instead. I regularly see high-quality comments, sometimes the highest quality in an entire discussion, that are marked dead because the account is banned.

hueving 11y ago
hellbanning is easily the worst thing about HN. It's the embodiment of a childish idea that nobody can change, ever. It's not even effective against trolls because a troll will just change accounts immediately. It just silences a section of the population that happened to touch a nerve back in the day. A disgusting black mark on an otherwise reasonable site.
- brudgers 11y ago
  What are the better alternatives to hellbanning?
  - PhasmaFelis 11y ago
    Regular banning, of course. You tell the user that they've been banned and why. Optionally, offer tempbans (with a stated duration), and/or have an appeal mechanism; but even with neither of those, it's less sadistic than a hellban.
    - brudgers 11y ago
      I've had showdead=yes for a while and haven't seen anything that made me suspect live banning was employed for pleasure by Hacker News moderators. Admittedly, I'm not inside their heads and have only paid attention to a small fraction of dead posts. One difference between regular banning and live banning is that the banned person is kicked out of the community with a regular ban, but not with a live ban. A live ban mitigates the effects of problematic behaviors on the community while still allowing the person behaving problematically to retain their identity and participate in ways that aren't problematic: i.e. people who are live banned can remain part of the community. Because the live banned person remains within the community, there is an opportunity for the community to recognize unwarranted live bans based on actual exhibited behavior within the context where it is relevant. Anecdotally, I've seen unwarranted live bans lifted in real time. I've also seen members of the Hacker News community who have benefited from keeping their identity within the community over an extended period of time while continuing to exhibit problematic behaviors. People may look at that situation and see something different, I see live banning as a highly compassionate way to solve the corner cases that need high levels of compassion. From a practical standpoint doing regular banning "right" with formal notification, explanation, and appeals processes requires a non-trivial moderator time and energy. If most bans are justified, that means all that energy is wasted on accounts that the owner doesn't value and accounts that the owner values solely as a conduit for argument and/or insult. Even in cases where the person values their account as an identity within the community, formal processes are an escalation. The most likely proximate cause for needing to limit that type of account is the manner in which disagreement is expressed. Creating a context that threatens identity is unlikely to suddenly produce better behavior in regard to disagreement over a sanction already enforced. Regular banning is confrontational. One of the ways Hacker News differentiates itself from other sites where people type into boxes is by discouraging that very behavior. Lunch at Cafe Hellban comes at some cost to the community, but I believe it's lower than the alternatives.
      - krapp 11y ago
        >i.e. people who are live banned can remain part of the community But the purpose of "live banning" is to fool a user into believing they're part of the community, while trying to keep them from actually participating through subterfuge. If live banning were simply a way to allow "problematic" users to participate, then the subterfuge wouldn't be necessary, and showdead wouldn't be opt-in, it would be opt-out. Banned users are meant to be ignored and forgotten about by default. >I've also seen members of the Hacker News community who have benefited from keeping their identity within the community over an extended period of time while continuing to exhibit problematic behaviors. If you're talking about TempleOS - I would argue that actual banning would be healthier for everyone in his case. He is the ur-example of why "live banning" doesn't really work as intended, or why, at the very least, actual banning should also be an option alongside live banning. Encouraging him to stay and post while banned is what has turned him into a museum exhibit and mascot for this site, and I think the effect on the community of having to justify this as being, somehow, the best of all possible worlds is toxic. >From a practical standpoint doing regular banning "right" with formal notification, explanation, and appeals processes requires a non-trivial moderator time and energy. You want moderators to apply more than the minimum possible amount of time and energy to their duties. But given that there is already an informal appeals process for bans involving someone publicly calling out a banned user and having them email the site, applying that process to a form doesn't seem like much greater effort. Sites with far more traffic and far greater hostility manage with entirely volunteer moderators. >that means all that energy is wasted on accounts that the owner doesn't value and accounts that the owner values solely as a conduit for argument and/or insult Of course, that assumption only holds true for some banned accounts, not all. Those obviously deserving accounts could be shown a generic banned message, mentioning it can't be appealed. Or show the generic message and a contact email in all cases. For other banned accounts, it would be unethical to consider appeals to be a waste of effort because it involves work for the moderators. There is no reason the process has to be unreasonably complicated in any case, as long as a banned user knows they've been banned and has some way of contacting staff to discuss this. >Regular banning is confrontational. "live" banning can be confrontational as well, a passive-aggressive, 'slowly poisoning someone's food with arsenic kind' of aggression. Dang now warns people repeatedly before banning them, which is confrontational, but also demonstrably more humane than simply dropping them into the oubliette. Although, of course, new users may not know what "live-banning" is or when it occurs, they may expect a normal ban. But a user who has been warned repeatedly and then live-banned seems more likely to deserve it than one who's banned without warning. >Creating a context that threatens identity is unlikely to suddenly produce better behavior in regard to disagreement over a sanction already enforced Except with live bans, if the sanction is enforced in such a way that the user is unaware of it, the reasons behind it, or the necessity for self-correction, obviously this will not result in a change in behavior. To expect someone's reaction to a ban that they're aware of to be the same as their reaction to one they're unaware of is, I think, unreasonable except in extreme cases.
    - DanBC 11y ago
      > You tell the user that they've been banned and why. Optionally, offer tempbans (with a stated duration), and/or have an appeal mechanism You know this sometimes happens on HN, right? Some people are given many warnings, and then they're banned and told that they're welcome to come back if they stop that behaviour?
  - PhasmaFelis 11y ago
    Regular banning, of course. You tell the user that they've been banned and why. Optionally, offer tempbans (with a stated duration), and/or have an appeal mechanism; but even with neither of those, it's less sadistic than a hellban.
- lotharbot 11y ago
  I think hellbanning has its place. There are people who join websites with no intent of contributing -- who want to make trouble, to troll, to harass, to create link-juice for other sites, or otherwise benefit themselves at the expense of others. An ordinary ban simply invites them to return under a new name and keep it up. A hellban both stops their negative influence from propagating, and stops them from realizing that it's time to make a new account. The problem, of course, is that sometimes people actually do intend to contribute but are caught in a hellban because they made a single mistake, or possibly even because they said something that bothered a moderator but was completely reasonable for them to have said. Allowing users to "vouch" is a nice counter to this -- if we see a genuine contributor who has been mistakenly flagged as a troll, we the community can fix it. This is a fantastic change.
  - hueving 11y ago
    Did you miss the part where a troll/spammer can just create a new account? It's trivial to see if you're hellbanned (incognito window).
    - DanBC 11y ago
      But spammers don't tend to just make new accounts. And it's fine that trolls create new accounts - their new accounts get banned quickly if they troll in their new account. And we don't know if there are methods that detect account hopping. I suspect that there are.
    - lotharbot 11y ago
      I'm well aware of that step. But not every troll or spammer is. How many trolls or spammers actually research the target forum well enough to know that they use such a moderation style? It seems to be a very small number, given how many trolls and spammers get hellbanned and keep on posting, compared to how few seem to create new accounts immediately.
DanBC 11y ago
Do you have any examples? I know that it does happen that good comments are banned, but it's not that common. And this announcement gives us tools to fix it.
- lotharbot 11y ago
  Not off the top of my head. I don't mean "regularly" like once every few minutes, but more like a couple times a month.

hueving 11y ago

hellbanning is easily the worst thing about HN. It's the embodiment of a childish idea that nobody can change, ever. It's not even effective against trolls because a troll will just change accounts immediately. It just silences a section of the population that happened to touch a nerve back in the day. A disgusting black mark on an otherwise reasonable site.

brudgers 11y ago
What are the better alternatives to hellbanning?
- PhasmaFelis 11y ago
  Regular banning, of course. You tell the user that they've been banned and why. Optionally, offer tempbans (with a stated duration), and/or have an appeal mechanism; but even with neither of those, it's less sadistic than a hellban.
  - brudgers 11y ago
    I've had showdead=yes for a while and haven't seen anything that made me suspect live banning was employed for pleasure by Hacker News moderators. Admittedly, I'm not inside their heads and have only paid attention to a small fraction of dead posts. One difference between regular banning and live banning is that the banned person is kicked out of the community with a regular ban, but not with a live ban. A live ban mitigates the effects of problematic behaviors on the community while still allowing the person behaving problematically to retain their identity and participate in ways that aren't problematic: i.e. people who are live banned can remain part of the community. Because the live banned person remains within the community, there is an opportunity for the community to recognize unwarranted live bans based on actual exhibited behavior within the context where it is relevant. Anecdotally, I've seen unwarranted live bans lifted in real time. I've also seen members of the Hacker News community who have benefited from keeping their identity within the community over an extended period of time while continuing to exhibit problematic behaviors. People may look at that situation and see something different, I see live banning as a highly compassionate way to solve the corner cases that need high levels of compassion. From a practical standpoint doing regular banning "right" with formal notification, explanation, and appeals processes requires a non-trivial moderator time and energy. If most bans are justified, that means all that energy is wasted on accounts that the owner doesn't value and accounts that the owner values solely as a conduit for argument and/or insult. Even in cases where the person values their account as an identity within the community, formal processes are an escalation. The most likely proximate cause for needing to limit that type of account is the manner in which disagreement is expressed. Creating a context that threatens identity is unlikely to suddenly produce better behavior in regard to disagreement over a sanction already enforced. Regular banning is confrontational. One of the ways Hacker News differentiates itself from other sites where people type into boxes is by discouraging that very behavior. Lunch at Cafe Hellban comes at some cost to the community, but I believe it's lower than the alternatives.
    - krapp 11y ago
      >i.e. people who are live banned can remain part of the community But the purpose of "live banning" is to fool a user into believing they're part of the community, while trying to keep them from actually participating through subterfuge. If live banning were simply a way to allow "problematic" users to participate, then the subterfuge wouldn't be necessary, and showdead wouldn't be opt-in, it would be opt-out. Banned users are meant to be ignored and forgotten about by default. >I've also seen members of the Hacker News community who have benefited from keeping their identity within the community over an extended period of time while continuing to exhibit problematic behaviors. If you're talking about TempleOS - I would argue that actual banning would be healthier for everyone in his case. He is the ur-example of why "live banning" doesn't really work as intended, or why, at the very least, actual banning should also be an option alongside live banning. Encouraging him to stay and post while banned is what has turned him into a museum exhibit and mascot for this site, and I think the effect on the community of having to justify this as being, somehow, the best of all possible worlds is toxic. >From a practical standpoint doing regular banning "right" with formal notification, explanation, and appeals processes requires a non-trivial moderator time and energy. You want moderators to apply more than the minimum possible amount of time and energy to their duties. But given that there is already an informal appeals process for bans involving someone publicly calling out a banned user and having them email the site, applying that process to a form doesn't seem like much greater effort. Sites with far more traffic and far greater hostility manage with entirely volunteer moderators. >that means all that energy is wasted on accounts that the owner doesn't value and accounts that the owner values solely as a conduit for argument and/or insult Of course, that assumption only holds true for some banned accounts, not all. Those obviously deserving accounts could be shown a generic banned message, mentioning it can't be appealed. Or show the generic message and a contact email in all cases. For other banned accounts, it would be unethical to consider appeals to be a waste of effort because it involves work for the moderators. There is no reason the process has to be unreasonably complicated in any case, as long as a banned user knows they've been banned and has some way of contacting staff to discuss this. >Regular banning is confrontational. "live" banning can be confrontational as well, a passive-aggressive, 'slowly poisoning someone's food with arsenic kind' of aggression. Dang now warns people repeatedly before banning them, which is confrontational, but also demonstrably more humane than simply dropping them into the oubliette. Although, of course, new users may not know what "live-banning" is or when it occurs, they may expect a normal ban. But a user who has been warned repeatedly and then live-banned seems more likely to deserve it than one who's banned without warning. >Creating a context that threatens identity is unlikely to suddenly produce better behavior in regard to disagreement over a sanction already enforced Except with live bans, if the sanction is enforced in such a way that the user is unaware of it, the reasons behind it, or the necessity for self-correction, obviously this will not result in a change in behavior. To expect someone's reaction to a ban that they're aware of to be the same as their reaction to one they're unaware of is, I think, unreasonable except in extreme cases.
  - DanBC 11y ago
    > You tell the user that they've been banned and why. Optionally, offer tempbans (with a stated duration), and/or have an appeal mechanism You know this sometimes happens on HN, right? Some people are given many warnings, and then they're banned and told that they're welcome to come back if they stop that behaviour?
- PhasmaFelis 11y ago
  Regular banning, of course. You tell the user that they've been banned and why. Optionally, offer tempbans (with a stated duration), and/or have an appeal mechanism; but even with neither of those, it's less sadistic than a hellban.
lotharbot 11y ago
I think hellbanning has its place. There are people who join websites with no intent of contributing -- who want to make trouble, to troll, to harass, to create link-juice for other sites, or otherwise benefit themselves at the expense of others. An ordinary ban simply invites them to return under a new name and keep it up. A hellban both stops their negative influence from propagating, and stops them from realizing that it's time to make a new account. The problem, of course, is that sometimes people actually do intend to contribute but are caught in a hellban because they made a single mistake, or possibly even because they said something that bothered a moderator but was completely reasonable for them to have said. Allowing users to "vouch" is a nice counter to this -- if we see a genuine contributor who has been mistakenly flagged as a troll, we the community can fix it. This is a fantastic change.
- hueving 11y ago
  Did you miss the part where a troll/spammer can just create a new account? It's trivial to see if you're hellbanned (incognito window).
  - DanBC 11y ago
    But spammers don't tend to just make new accounts. And it's fine that trolls create new accounts - their new accounts get banned quickly if they troll in their new account. And we don't know if there are methods that detect account hopping. I suspect that there are.
  - lotharbot 11y ago
    I'm well aware of that step. But not every troll or spammer is. How many trolls or spammers actually research the target forum well enough to know that they use such a moderation style? It seems to be a very small number, given how many trolls and spammers get hellbanned and keep on posting, compared to how few seem to create new accounts immediately.

DanBC 11y ago

Do you have any examples? I know that it does happen that good comments are banned, but it's not that common. And this announcement gives us tools to fix it.

lotharbot 11y ago
Not off the top of my head. I don't mean "regularly" like once every few minutes, but more like a couple times a month.