So you know how currently a flag is immediately seen by the entire 10k+ rep users community and all the mods in every room. That's not right.

What I'm proposing:

  • When a flag is raised, at first only the room owners can see it and review it, in the spirit of self-moderation.

  • If the room owners are away from keyboard or they choose to not make a decision after 2-3 minutes, or if there are no room owners in a room, the flag goes up to the moderators of the site to which the room is related.

  • If these local mods aren't there to handle the flag in 2-3 minutes, it goes a tier higher to all moderators. And if they're too busy, they may choose to forward it to all 10k users as it's currently working.

  • Flags raised by room owners go to local mods, since if they placed it, they are not able to handle it themselves. (or if the granular room owner privileges/abilities are implemented, to other room owners who have the privileges required to handle chat flags)

At any point the flag reviewer may hit the "I'm reviewing this" button which will extend the timeout after which the flag is taken up a tier (maybe by extra 5 minutes), so that if nobody is reviewing it, it goes up faster. (I don't know if that's going to be necessary, but just in case I decided to include this idea)

All raised flags and their review results are visible to moderators in the chat room's flag history (we're gonna need a page for that). Mods can review them and if some room owners are abusing the flag review system by snuffing them out when it's not appropriate (from what I've read, this happens and needs to be considered), these users may lose their chat flag review privileges across all chat rooms. So if a flag is raised in a room where there are only "untrustworthy" flag reviewers, it goes straight past them to local mods.

(I already posted this as an answer and in comments to "We're more aggressively enforcing self-moderation in chat" discussion, but I think it deserves its own place now that I've fleshed it out enough.)

  • We've already read about messages standing on their own out of context, and if the room owners are found (by mods) deliberately abusing the system by flagging messages which only appear harmful out of context, their flagging privilege may be removed. May 5, 2018 at 2:27
  • Then if a user racks up too many declined chat flags in a row (or a short period of time), their flags may be marked as "not very trusted". After a bit of thinking, removing flagging privilege seems a bit too far, but sorting flags by "trustworthiness" rating seems like it would make the jobs of all flag reviewers easier. They'll know what to expect from a user who flagged 10 posts in the last hour and 10 of them were declined. This might be interesting enough idea for its own proposal post. May 5, 2018 at 2:31
  • Very nearly a duplicate of meta.stackexchange.com/questions/112756/…
    – Shog9
    May 5, 2018 at 2:55
  • Essentially, but with addition of a few further thoughts. That one is pretty basic. I just hope this doesn't get shelved forever like everything else usually is :/ May 5, 2018 at 3:02
  • 4
    Here's a thought experiment that might help. Try to identify a number of different scenarios in which flags are commonly raised, disputed, protested, validated, escalated, etc. Describe an ideal outcome for each of the flagger, the room, and the network. Then theorize how each would be handled in both the current and proposed system. Essentially, treat a flag as a collection of functions that just happen to be accessed via a single interface at present.
    – Shog9
    May 5, 2018 at 3:11
  • 1
    @Shog9 I don't have the experience, intricate knowledge, inside information or mental capacity to do the entire thing, so if you could dissect it for me, as you usually do, I'd like to read your point of view. May 9, 2018 at 11:48
  • You don't give yourself enough credit; I suspect you spend a lot more time using chat than I do. That said, I'll try & write something up if I can find the time.
    – Shog9
    May 10, 2018 at 0:55

1 Answer 1


I'd like to open by saying that I was a proponent of this approach for quite a while, though I do realize that there are three different contexts where this approach may or may not work better.

With recent events though, I guess the big problem is that it makes it too easy to kick an issue under the carpet. If a mod or RO is active, they're there and can see what's going on. Presumably, if they realize that it's a problem, they can deal with it. Done incorrectly, it takes away independent oversight of a chat room. Stuff that shouldn't happen could, and it relies on the ROs and presumably mods to do the right thing. That doesn't always happen.

I'd also like to say, when something is flagged, it's an incident. It's something that needs looking at - and many chat rooms actually have an understanding with other mods. I think folks know that if something goes wrong when I'm asleep, I will follow up with the folks involved.

The one situation where I'd love this is for non-English sites, because flags in Portuguese and Russian are just inscrutible for many of us.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .