Rule creep happens.

There's that one guy who enjoys watching the ballgame while sitting in the bathroom stall, so you end up getting a departmental mail sent to the entire company explaining that it is unprofessional to sit in the bathroom watching baseball during working hours. The same thing happens in forums. Someone violates the spirit of the rules to go all internet-lawyer on the community, so the community decides it's just easier to make a new rule to shut them up.

Eventually you end up with a mess of rules that nobody can really explain rationally, much like the monkeys gathered around the ladder saying, "I don't know why, that's just how we've always done things here."

What can be done to prevent (or at least minimize) this sort of rule creep while still maintaining the ability of the community to handle troubled users?

4 Answers 4


If you make rules to address every situation that comes up, you'll be making rules forever (because users will keep coming up with new ideas) and then you'll end up with a large, unmanageable pile of not-necessarily-self-consistent rules, and an obligation to administer them. Yuck. That's no fun.

Instead, you want fewer rules and broader license. The CS department at my alma mater (Carnegie Mellon) had a pretty strong community when I was there -- there was an active lounge (students, staff, and faculty), there were departmental activities, and there was even a chore rotation for everything from cleaning the lounge fridge to maintaining the software to manage the monthly cheese buy (I am not making this up). This community had one rule, the Reasonable Person Principle: strive to be reasonable, grant that everyone else is doing the same thing, and don't take disagreement personally.

That all sounds fine and dandy, but what if definitions of "reasonable" collide? The RPP states:

Not all people share the same model of reasonableness, so disagreements inevitably occur. Under the reasonable person principle, the first thing to do is work it out privately (perhaps in person, since e-mail is known to amplify feelings). Indeed, many people would find it unreasonable to bring in third parties before trying personal discussion.

This is easier to do in a small, local community where you can sit down with the other person over coffee, but the idea still works even if the other person is halfway across the world. If you can at all manage to have a private synchronous conversation (phone, IM, whatever), instead of bouncing email back and forth, I've found that works better. If someone has hours to stew over, and work up a head of steam in replying to, a piece of email, things degenerate more quickly.

The CS department didn't have moderators (people with the authority to enforce rules against people), but if it had had moderators we would still have encouraged the community to address these problems where possible. You don't like that Joe cooks smelly curry in the microwave, or that user666 posts snarky comments on every post? Take some ownership and try to work it out amongst yourselves before asking a moderator to step in; it's what reasonable people do. If a problem comes to you as a moderator (or you, as a mod, are the person who objects to the behavior), then follow the RPP like anybody else. A conversation between people who have different perspectives on a community they care deeply about will accomplish way more than citing rule 42(a)(iii)(paragraph 17) (and adding paragraph 18, because who would have thought to regulate washroom viewings of the World Cup?).

  • 1
    Excellent Answer
    – Malachi
    Commented Aug 11, 2014 at 15:39

This list of techniques was tested in a huge amount of communities, they all work very well and prevent a lot of moderation workload. I am mainly referring to my experience in communities related to Persistent World, a modification to Mount&Blade: Warband game. Due to the focus of the mod (role-playing) being very different from the focus of online play in the vanilla game (only fighting), the lack of software solutions to make things peaceful (don't even think they are possible), rules were needed to keep peace. Because of a huge amount of ways to disrupt role-playing, some servers ended up with a huge wall of text on their rules page that noone wanted to read. During my times of admining, though, the following techniques were used to avoid it.

  • Implement a "common sense rule" -- while listing some particular types of inappropriate behavior that happen frequently, tell your users to behave reasonable, as Monica Cellio pointed. This prevents users from intentionally abusing loopholes in your rules, also giving you an official ground to punish those who anyways do. Persistent World servers often have (had, better to say: I didn't play it in the last couple of years, nor did I admin), something like "abusing the loopholes in the rules shall only make your punishment worse".
  • Make rules address not only one particular action (watching ballgames in WC during working hours), but rather a group of them (making something not related to your work during working hours). It is hard to provide a Persistent World example, I cannot come up with one: pretty much all of the rules we used were more or less broad, or it would be even bigger walls of texts.
  • Many situations can be just addressed once by talking to a particular problem user and deleting particular problem content, without making a special rule for the case. Only make rules for offenses that happen relatively frequently, or that are "popular" among new users. For example, most servers didn't have special rules against offensive character and faction names, we just asked players with nicknames like "Adolf_Hitler666" to change the name and kicked them. If some offense is particularly harsh, you can sometimes even just ban a player, possibly giving an opportunity to appeal it (like if that player comes again under the same name, or under another offensive one). Most of the time it was enough to warn and ask not to do it again. On the other hand, a Minecraft role-playing community I was in for some kinda long time had special rules against non-letter characters in names, because for some unknown reasons a lot of people wanted their playing characters named "john99", "yourmom2005", etc.
  • Go through the list of your rules from time to time and rework/delete rules that are no longer important/related. If you have a long list of them, it might be possible, for example, that some sentences contain artifacts of previous versions: most of the long Persistent World rules had a lot of internal references. It is also a common issue on old forums with long lists of rules that change frequently.
  • Understand that new users don't always follow the rules. Even if you make the list very short, there will still be people who didn't find the rules, didn't bother to read them, or simply did not understand the principles you imply after reading your rules for the first time. As an admin I always asked a new user when I got a complain about a typical, but not-so-harsh rules violation: "Did you do X?". If he admitted doing it, I asked: "Do you know that it is against the rules?", the answer being "no" 99% of the time. Then I said "You should read the rules of our server, it is kinda special and different from vanilla Mount&Blade experience. You shouldn't do X because Y." After making sure that the user understands this rule now and knows where to find the rule set, I said something like "Ok. Good that you get it now. Unfortunately, since you broke the rules, I am banning you for one hour. Feel free to come back and play with us when you read the rules.". If the offence was too strong to be treated with just a 1-hour-long ban, and the player was clearly new to the thing, I tried to settle it down by offering a compensation to the complaining side. This is a very, very rough description of how was it actually done, the real way the conversation goes depending on many other things. Of course, if the user totally denied the action he has done (not the offence), there was something like "I know for sure that you have done that, I am applying [a harsh punishment], you are strongly suggested not to neither repeat your main violation nor lie to an admin again, and reread the rules".

  • Make a "short" and a "long" versions of the rules. Short version basically being a "be reasonable" article with just a few guidelines on how to behave, while the "long" version being a long list of inappropriate action that have already been punished in your community. This I have seen tested in LARP (Live-action role-play), where full rules may sometimes take as many as 50 A4 pages. Short version works as a summary of the rules which is more useful to the people who are already familiar with how things go in LARP in my area, while the long version is for new players. At the same time, it was advised that everyone reads the long version too because it is actually the thing being used. I have to note that while community rules, as well as Persistent World rules, are more like a code of conduct, behavior standards, rules of LARP are more like game mechanics on paper, but I also know some Russian communities that implemented short version of the rules, and it was pretty much successful. I just didn't see that much of it in action, so I am referring to the LARP experience more.

  • Do not list some offenses that do not need to be listed, such as spamming, or trying to apply a DDoS attack on the community. People who come to spam are not people, they are bots, they don't read the rules, and actual users don't do spamming (well, unless your definition of spam includes, say, posting any kinds of URLs; I have seen that). It is common sense that DDoSing is not OK, and if someone tries it, just take some actions to protect your community and ban the user. Anyone who spams or DDoSs is not a good fit for you. For example, in the aforementioned Minecraft community there was a schoolboy who threatened the admin with a DDoS attack after the admin game him a warning -- the boy was banned, but there is still no rule against DDoS. At the same time, most Minecraft communities that I have seen implement special rules against doing harm to the server in some game-specific and way more common ways.

  • 1
    Welcome to Community Building! You said these suggestions were tested in a lot of communities; could you say more about that? Where did you see or learn about these approaches? (There's a lot of good common sense here; I'm just wondering about the testing you alluded to.) Thanks. Commented Aug 20, 2017 at 18:47
  • 1
    @MonicaCellio I added more info about the experience I had with those techniques in Persistent World (role-playing mod for Mount&Blade: Warband), Minecraft role-playing community I have been into for a relatively long time, and in LARP (Live-action role-play). Not sure if my answer actually became better from that edit. Commented Aug 20, 2017 at 19:28

Simple is better. Manage expectations from as early on as possible that moderators exist for avoiding rule creep, not simply being rule enforcement robots. Set guidelines for atmosphere. Things like "don't be disruptive" are general enough that moderators can take actions to deal with a problem if it is a problem.

The flip side of this is that you need to make sure to be careful with moderator selection. Don't pick people who are looking for power, but rather looking to make the community better. Ensure there is open communication between moderators and that there is a relatively consistent understanding of where the lines are.

If you keep things general enough, then moderators are free to take whatever actions are necessary to remove problems from the community. If a user is being obnoxious, you don't need a new rule not to do that particular obnoxious thing, you just need moderators who can see that the user is annoying the community, ask the user to stop and remove the user if he is unwilling to behave in a way that fits within the community.

Moderators should also be sure to take action promptly to avoid a bad example being set and to make sure it is clear to the community as a whole what is expected. It is difficult to get to a point where it is unmanageable for moderators if you are sure to deal with issues early on in a consistent manner. This also requires clear communication between moderators.

That said, rules do sometimes need to be added, but they should be added to address a pattern of behavior, not an individual action. If you see recurring problems by multiple users, then this speaks to an additional rule being needed to clarify what the community expectations are. You should also make sure that the new rule clearly identifies the type of behavior that isn't acceptable and why so that other similar behavior will still be caught under the umbrella of the guideline.

  • So yes, I get that this is how things should work, but in my 20+ years of online communities I've never actually seen things work out without rules being added. Nobody knows 100% which rules are required at the start, and after a year nobody knows 100% which rules were added because they were needed, and which rules were added because they were more convenient in moderating. This strikes me as idealistic, not realistic, and it requires moderators to actively engage every single problem user which doesn't scale well when you have a lot of problem users.
    – jmac
    Commented Jul 30, 2014 at 16:11
  • @Jmac - I'd counter that if you are having lots of problem users, then mods were not active enough previously in eliminating problem users quickly. That said, I'm not saying rules don't need to be added sometimes, but if you are adding a rule for a specific case rather than for an observed pattern of issues, then you are doing it wrong.
    – AJ Henderson
    Commented Jul 30, 2014 at 16:20
  • So pretend for a second that they were all rule changes for 'observed patterns of issues'. It is still rule creep. How do you prevent that? Are we really doomed to creating layers upon layers of rules to handle exceptions? Is there really nothing that can be done to mitigate the harm of rule creep making community standards/new user integration far more complex than they should be?
    – jmac
    Commented Jul 30, 2014 at 16:41
  • I don't expect such a rule would have to be added regularly, or even ever provided a community is working well. If you have situations that require another rule because the community is not understanding how to integrate, then another guideline is helpful, if the guidelines are well defined, then their shouldn't be an issue. The best way to prevent rule creep is to only create rules when absolutely necessary (expected behavior is broadly not being communicated correctly) and make them such that you have the least chance of needing more (moderators are able to handle the individual cases.)
    – AJ Henderson
    Commented Jul 30, 2014 at 16:48

In this case, I believe the most important thing is to approach the offending user privately first before attempting to "squash" the unwanted behavior with a blanket rule to ban the behavior. Often, the user may not understand the community all that well yet and just needs a little reminder about what's appropriate (so in the case of the guy watching the game for 40 minutes in the stall, remind him that it's not appropriate).

Rules should be made with the community's thoughts at heart. After all, the users are what makes up the community and keeps people coming back, so driving people away with frivolous rules will definitely be counter-productive.

So the best way to stop it: don't do it and approach the users individually instead. If the users become argumentative saying that it's "not against the rules", it's then that you go back to the community at large and address the issue in a conversational way rather than ruling with an iron fist.

  • So yes, I get that this is how things should work, but in my 20+ years of online communities I've never actually seen things work out without rules being added. Nobody knows 100% which rules are required at the start, and after a year nobody knows 100% which rules were added because they were needed, and which rules were added because they were more convenient in moderating. This strikes me as idealistic, not realistic, and it requires moderators to actively engage every single problem user which doesn't scale well when you have a lot of problem users.
    – jmac
    Commented Jul 30, 2014 at 16:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.