How would you make statistics of power relations?

I have a Discourse instance where people tend to use the forum in quite surprising ways. At first, the forum was organized by groups-categories, where each group (reflecting an actual human group in a shared physical space) would have its own “private” category, and there was a common category where all groups would maintain a single topic to “publish” their own state.

Lately, the organization evolved under peer pressure to allow everyone to see all categories, and (ab)using mentions as a mean to force notifications. Of course in practice this does not really work because one cannot mention more than 10 users or groups in a single post.

I realized that a lot of discussions go through private messages, back-channeling categories and overriding group discussions. I’d like to find a way to expose statistics to demonstrate that this back-channeling is actually influencing the community in wrong ways, to insist on using categories limited to groups instead of pseudo-transparency that defeats the purpose of notifications and self-chosen level of engagement.

So the question is actually double:

  • do you have similar experiences and how do you handle them?
  • did you come up with special requests to gather such statistics in a useful way?
1 Like

This is horrible. I’ve seen it once or twice and it’s just a terrible way to spam the community with noise.

I’m not sure what statistics would even exist for this? Personally with vaguely similar issues I’ve just pulled out a couple of topics which clearly conflicted with the goals of the forum, presented them as case studies and described the effects of the given behaviour. In particular describing the chain of activity: when users do x, y happens and then we end up with z. We don’t want z because…

It sounds like in general the community is suffering from a culture of bad practice and lack of clear direction. Personally I would spend a bit of time looking like what the ideal interaction between parties would look like. From that I would start to look at restricting the actions which conflict with the overall goal – you don’t need to include and use all of the features of Discourse.

For example, I’ve disable private messages for anyone below trust level 4, making them essentially unused – ensuring all content is posted as Topics and not lost inside a private message.