Moderator Dashboard: Potential Areas of Concern Report

Continuing the discussion from What cool data explorer queries have you come up with?:

I think @fefrei 's “Who has been sending the most messages in the last week?” Data Explorer query is really helpful for identifying abuse of the PM system in a community.

Inspired by that, I would like to propose (perhaps the wildly unrealistic feature) that some of these potential areas of abuse be run automatically and be brought together into a space on the moderator dashboard. It might have a neutral name since the data in itself wouldn’t be determinative.

Some items it might highlight:

  • Users who are a few standard deviations above the average number of PMs
  • Users who are a few standard deviations above the average number of post edits (their own or someone else’s)
  • Users whose posts have received a high number of flags
  • Users who have had a high number of messages to/from moderators
  • Other areas??

This dashboard would helpfully aggregate and clarify where moderators might want to prioritize their time.

I don’t really feel that addresses the actual issue. Are you seeing a lot of users send excess PMs?

Remember that TL0 users can’t send PMs at all. It is possible there should be stricter rate limits on sending PMs for users at TL1.


Hm, can we make " PM " a watched word here on meta?

Our organization has a very peaceful and respectful community. I’m incredibly encouraged by how much kindness and care is taken to discuss even contentious issues. It isn’t perfect, but there’s a lot to celebrate.

Unfortunately, there are a handful of people who have committed to trolling our organization and seem to have considerable free time to do so. My goal is to be proactive and have systems and processes in place to identify different ways they might seek to cause trouble.

1 Like

All the other items you listed are basically covered, the only possible point of “not knowing” is high volume of PMs, particularly for TL1 users, as we know, TL0 users can’t send PMs at all. That’s why I think it might be better to have more cohesive trust level 1 PM limits here, for the benefit of all Discourse sites.