Recovering from a paranoid community

The Discourse FAQ mentions that “This is a Civilized Place for Public Discussion”, “Be Agreeable, Even When You Disagree”, and “Always Be Civil”. We agree 100% and have the same values on our Discourse forum. Unfortunately, a sizable portion of our members don’t value that sort of community. They can’t fathom the concept of being constructive and its benefits.

Because of this lack of understanding, they misinterpret moderation action as disallowing any sort of criticism. If we take down a PA targeted at a company team, the community jumps to the conclusion that they’re not allowed to criticize the company. The actual answer is that instead of posting the below they should post the one that follows:

This update is awful. The responsible team obviously doesn’t care about their jobs.

This update is awful. Here is a list of past updates that have had the same problem – this is a reoccurring problem. Can we maybe look for more community feedback before pushing to production in the future?

but, again, since they can’t fathom a constructive community, this is all Greek to them. I tried to tell the user who made that post that it was a PA, but they kept insisting that it wasn’t – that it couldn’t be. They refused to believe anything other than “criticism of the company isn’t allowed”.

We could just to remove users from the community who can’t understand what it means to be constructive, but a mass exodus like that (there are a lot of users like this) would cause enormous tension within the community. I’m not sure what else we could do though. The longer this is a problem, the more it festers – it’s pretty critical we resolve this soon rather than letting it take its course.

Have any of you experienced similar problems in your own communities before? What did you do to resolve the problem?


This is an interesting subject. We do also have a lot of users who are very direct and not that helpful in their criticisms.

We deal with that with civil responses. In this case, I would not remove their posts but rather respond something like “sorry to hear you had issues with this update, could you give more details about the problem you’ve encountered?”

When you have an unhappy customer, and they care enough to share it with you, the best response it to try to find out what the problem is. If they feel shut down by you removing their posts, they will likely grow even more dissatisfied with your services.


We were more lenient in the past, but the abyss looks back and these kind of posts rubbed off on the community – we started seeing even more. The posts we leave set the tone of the community. By removing the posts we’ve helped decreased the amount of new ones, and the feedback we receive is a lot more valuable. Leaving them up is unfortunately not an option.


I think the key is reiterating that you accept constructive criticism but out going hate is not acceptable - if people have a problem they can take it up privately with you. At the end of the day if they dont like it they are well within their reigns to go somewhere else.


This is tough topic, amongst the one that users also start to PM others forcing solutions and answers. Most of them are well known community members, and they are literally spammed by impolite users.

This becomes hard to analyse when you’re not connected to those members, or they fear to tell you in the first place. This really needs well-written guidelines and actions taken in a reasonable, explaining way. Sometimes you also need to use the hammer, and tell users that they have crossed a line.

On the old Wolflab platform, users could also send emails to others. One who doesn’t have that much experience tries to help, and the other one abuses such. There were also user walls, where the conversation was going on further, hidden from the public audience. This really annoyed community members, and I literally needed to watch and monitor this activity. There was no point to stop this, rather to entire disable this alternative communication way.

Speaking for myself, I use to know many things, and I am not always in the good mood to explain everything to users. I for example expect basic Linux knowledge when someone installs a monitoring software I do support. Some users tend to think that this gets in the personal level, and that I am better than them. This clearly is a cultural thing, I mostly have seen this in the European region.

In the past, I was used to ignore such things. At some point I had users which started to attack other users in a thread. Including myself. The solution, to explain in a civil manner how things work, and stopping others to join my side by locking the topic, wasn’t really satisfying.

The tipping point was one who registered an account to just create a topic and write many things not very civilized. This included many personal attacks. Leaving this thread public for Google and as bad example for others, well, actually I do not want to have such bad karma around.

After I while I’ve deleted locked threads (7 in a year, not that much, but still alarming), and started to issue warnings and bans. That’s the regions I don’t want to go again.

One thing I am actively doing is to write many helpful things and explanations into the FAQ, and also linking that more prominent in the header. I’ve taken the default from Discourse, which already is very civil and good, and added some, let’s say, “triggering emotions” points into it. That way new users get to know that we are all human beings where just one little word can hurt our feelings.

I am also happy to have a small group of staff members who have grown out of this community, and they care a lot, and encourage others. Right on, they get used to the user trust levels, and enforcing forum rules.

The platform was previously a mix of German and English, as it was founded 2003 where it only was German. The Discourse platform breaks with that habit, and “enforces” English as inclusive community language. There are some users who ignore the announcement banner, and the FAQ and still post in German. We’ve agreed on locking these threads with a gentle reply to ask to create one in English.

TL;DR - there really isn’t the best way to handle such aggressive, problematic behaviours. Things to keep in mind

  • always be friendly, use emojis even when the other one is an “asshole”.
  • don’t respond immediately to threats. the other side then recognizes that you don’t care much about them.
  • tell users when they’ve crossed the line. Tell them that they’ve hurt someone’s feelings. This is mostly the point where they get touched from my experience.
  • enforce strict conversation rules. lock topics, warn/ban users, don’t give them your platform
  • listen and encourage the other users involved in this “situation”. Ask them how you can proceed with help, if e.g. their question was locked down.
  • don’t take sides in a discussion. be a civilized moderator and listen to everyone, still pointing out wrong behaviour.
  • some trolls will come again and again. Ignore them, and do nothing except for maybe moving topics into a category which is listed at the bottom (“anything else”).

The golden rule is to always be friendly and don’t let others down. Depending on the culture on your platform, I would encourage other well known community members to step up, and be more regulative. Grant them trust level 3 and explain the problem, ask them to take care. Sometimes users also tend to work against the admin/mod team, just because they have the shield icon.

Kind regards,


So having spent over 10 years as part of, and in leadership in, the XDA community I have seen my fair share of this sort of thing. One thing I have found to be extremely helpful is to recognize that quite often these type of responses come from the passion they have for the community/product/topic.

Passion manifests itself in a lot of different ways for a lot of different people. If you can reach out to them personally, either via PM or getting on a call, and figure out what is driving their responses it can go a long way to changing the tone - and even getting them involved to finding a solution is a good step. Notice I said “can” because sometimes you just can’t change their interaction or their feelings or their behavior/tone. At that point you need to have in place a public process for how you deal with it, and just do it.


I think this is an important thought, and I just read some related feelings about the gaming industry:

Unfortunately there are no clear answers, but this is a problem a lot of people are feeling. The article makes a good point in that unmoderated spaces like Twitter normalized a lot of this abuse as an acceptable way to talk to another person…

The fact that the bulk of the discourse takes place in unmoderated spaces has shaped that discourse. We actually do moderate the Steam forums for our games, and people are always shocked when they get banned for heaping abuse on other forum members (the only thing we regularly ban for).


I do PM users to get a better understanding of their concerns, but some of the tantrums seem intentional at this point. For instance, just recently:

  • User makes minor mistake in thread, and we send feedback
  • Corrects that thread, but then makes a new thread with the exact same content we asked him to correct
  • When we tell him to fix again, conveniently decides to make a controversial thread
  • Says he just wants to help, but more likely he wanted to start drama
  • Banned once we find out he was PMing people with "wink wink nudge nudge <threadlink>"
  • Goes on for hours in DMs insisting he was just trying to help
  • Once banned, 24/7 posting of derogatory memes about forum staff ensues

I’m pretty sure there’s a small group of users who despise some of the rules for the forum (e.g. we delete off-topic chains of memes from serious discussions such as Feature Requests) and intentionally try to cause drama to get back at us. This particular user will frequently post derogatory memes “other people send him” that aren’t true at all, and with as often as he posts, respectable users probably believe the same now.

Users thinking we’re censoring them is only a recent thing, so I’m pretty sure this propaganda is responsible. We can’t really do anything about it though because if we ban this person we just confirm what they’re spreading. It’ll be even more difficult to recover from this paranoid community.


What we did a few times at XDA was to make it quite clear to the offending user why they were being reprimanded/banned/etc. and point to specific instances in our rules/code of conduct that they were in violation. Banning was always a last resort though.

Once someone was banned that was causing a large issue, and especially if they were a big name in the community and had a number of friends who were taking up their cause, we would post a public notice in a general category about what had happened, what our attempts to rectify the situation were, and ultimately our decision. We would make sure to link to the code of conduct/forum rules as well in that. Then we endeavored to make sure we were consistent - but each situation is different.


So much this. You need to model the behaviour that you do want to see. Enlisting power users to help with that will speed things up. Effecting behaviour change is a long game though.

Potentially interesting related reading here:


The best would be not to explain the meaning of the word “constructive” but to demonstrate it.

Saying “sorry, this is not constructive” IS, well, not constructive :wink:

For example, in response to:

This update is awful. The responsible team obviously doesn’t care about their jobs.


So great we have some immediate feedback here! May you please elaborate a bit, what exactly is awful: we wanna keep our users happy, but your feedback needs some more details before our team can discuss it.

Don’t even mention it in your answer that it is not constructive – by saying so you’ll basically state that the writer is incompetent in giving any feedback in written form; to many people it may sound offending (i.e. you are publicly saying that what they wrote is inconsistent or not useful).


My experience is coming at this more from a politics/social-issues perspective. Where problem-solving is obviously a lot more subjective, value-laden, and controversial than with tech issues.

So often, people (including mods) will interpret “constructive” not in terms of personal conduct on the forum, but of a political or activism goal. IE ideas and opinions do not align with their preferred ideological framing, they are dismissed as societally non-constructive.

It happens! I have left two different Discourse-based fora as a consequence of this.