I’m creating this post after realizing that the current setup of the moderation system is causing real, serious misconceptions and harming the community in very subtle ways that were not obvious to me before. I’ll be bringing these issues up to the Discourse development team. Still, I wanted to explain how this particular issue happens and hopefully correct some misconceptions about the system as a whole.
Firstly, when a post is flagged, users carry “weight” based on their trust level (essentially how long they’ve been here and how often they’ve been flagged), and whether or not their flags have been agreed with in the past. Enough flags from users (or even one flag from our highest-rated leaders) can result in posts being hidden before a moderator has even seen the flags at all.
This is not clear, and it’s not clear in the automated messages that are sent. This can be even more confusing because no message is sent when a flag is “agreed” with (though we do now send a message if a flag is reversed and a post restored).
This has weird implications when a reply to a bad comment is flagged. Depending on who flagged the original bad comment and who flagged the reply, it can give the impression that the bad comment was ok (because it wasn’t immediately hidden), while reply was not (because it was immediately hidden). This sucks, because in both cases, no moderator has reviewed these posts yet, but messages have gone out saying “The community has flagged your posts!”, and from all appearances, a bad post has been allowed to stand unopposed.
This situation can be corrected once a moderator has had an opportunity to review the problem, but in the time in-between, this is a huge mess, and I’ve now learned in at least one occasion, convinced a longtime member of our community to leave because they believed their response to an insult was deleted, but the insult allowed to stand.
I’m not entirely sure what to do here about this:
- On one hand. we could disable posts from being hidden by the community, but then bad actors could walk all over the forum until a moderator showed up to deal with it, and we don’t have enough moderator hours to cover the blog 24/7.
- On the other hand, if we make any flag hide posts, it creates the ability for bad actors themselves to hide posts and sow confusion by sending automated “your posts were hidden by the community” messages on otherwise innocuous posts - again, until a moderator can show up and undo things.
- We could also raise the threshhold for when flags hide posts further, but we don’t get a lot of flags, and I fear that would create the same issue as option 1 above.
For now, what I’m going to do is rewrite the automated messages to make it clear that posts that are hidden may not have been reviewed and may be restored, and I think I’m going to suggest that a “your post was reviewed” message be sent in the same way a message is sent to the flagger for “something else” flags when they leave a note. But I’m mainly writing this down so that this issue is out in the open - just because your post was hidden doesn’t mean it was reviewed by moderators, and if a post you’ve flagged isn’t hidden, it doesn’t mean that it won’t be once a moderator reviews the situation.
Thanks, and I genunely apologize that this is so confusing!