The short question is: What factors do you take into account when you decide on rules for a community?
I live in a puzzling community for 5 years. It has an entry test. It is simple enough, but it filters people. Therefore, on the community forum we do not have rules at all; we envision that the users are intelligent enough to not spoil the forum.
I would say this worked quite well up to now. Recently we had two problems with it:
A user appeared and provoked others by making strict judgments about their personality. As result, the others would react similarly and the forum started to become ugly.
One user brought up a political topic. The topic would provoke people so that it lead to a similar result as the one in problem 1.
Both problems were solved in strict manner. The first problem was solved by banning the user, the second by closing the topic and issuing temporary bans for some users, who offended the others.
I do not like this situation, because the bans were not based on rules, but on the subjective feeling of the community moderators, that they "have to do this". This, in my opinion, can result in an unsafe feeling for users, who are not able to predict the behavior of the moderators. Worse still, the moderators could provoke a banned person to feel offended and deal more damage to the community.
I tried to figure out some rules, which would prevent such situations, but two things are quite hard for me:
How can one make rules objective enough? If you say, "respect the others", then what does respect mean? Everyone can understand it differently. As result, the same unsafety for users is present, as moderators and users share differing opinions.
How can one avoid situation when users find loopholes in the objective rules? When you formulate rules very specifically and objectively, there are always loopholes in them and members will use the rules against the moderators. For example, when I gave a user a ban saying, "You are banned for offensive judgments about another person's personality (Ad hominem)", he replied, "Ok, I see, so I can't say that he is xxxx, but I can say that his mother is xxxx?".
On the other hand, another user said, "So I can't say anything bad about others explicitly, but I can mention this implicitly, putting a hidden layer of meaning in my words", which is also definitely not what we wanted for the community - it is even harder to rid the community from implicit, offensive behavior of users.
In addition, moderators enforcing such rules looks like tyrants: nothing is allowed. So, how one can formulate rules in such a way that moderators can ban unwanted issues falling under these rules, rather than using their own opinions to the extent that community members rebel against moderators?
Are implicit rules bad at all, or should the rules be explicit?