so you are blaming the tooling for what exactly? what is the “hole”? I’ve read all your posts in this topic and I am really struggling to see what your point to all this is, other than that you are unhappy with the moderation on another forum that happens to use Discourse.
what undesirable outcomes? that you had a post flagged that you didn’t agree with? how is it the “tooling” that is at fault here?
is this what you are asking?
if you have over-zealous flaggers, perhaps you can create a custom group for such users then omit them from this setting (and maybe lock them to TL-1), or trust level rate limits, etc. flags start pms with moderators so you can appeal there.
Seems that the problem you are talking about is the way the system automatically responds to flags when there is not moderator presence to review those and they are just left unresolved? Thousands of members may be too many for this community where you are experiencing this.
Would a solution to this be a change in settings so that there isn’t an automatic response to community flag reports, but those just go to alert a moderator who can make a decision about those? If the mods/admins at the site you are talking about aren’t open to new ideas like that then there may be nothing you can do.
You could indeed adjust the hide post flag threshold. However then you risk having posts visible that maybe should be hidden until properly reviewed.
In my experience be it a Discourse Forum, Reddit, FB etc.. is often ppl are too impatient and expect quick resolutions. Well maybe a complete culture issue as you find this also with purchased products that require after sales support.
Sure if the forum is experiencing rapid growth then the mod team may not have time to scale up accordingly.
Here’s a thought: if the community has become large and too many flags are low-quality, how about removing the ability to flag from TL1 and move it up to TL2? The TL2 users will presumably be among the better behaved and more committed members of the community, and have more to lose if threatened by demotion for misuse of flags.
Censorship is a strong word and may not be correct to use in this context.
While you are a guest at someone else’s site they have a right to edit what you want to publish in public view with their platform.
If a government is prohibiting you from publishing what you want on your own site or with multiple different platforms/publishers then that may be censorship.
Apologies if my last post wasn’t in support of this vision, someone flagged that but mod here elected to not moderate that out. I agree that would be a different talk about the definition of censorship and the history of that is a whole different thing.
If you want to keep this topic going seems like you may be proposing changes to the core flag system for the platform? A feature request topic may be better for that.
Do you have any suggestion on how to better handle this?
It’s a problem that I’ve personally never faced, so I’m curious to hear if you have a recommendation
Until that time, I think @KhoiUSA, @Ed_S and @ondrej nailed it in their initial replies:
Quick note: when looking through the lens of People, Process, Technology; a lot of times, we community managers look at Technology for solutions, while often it’s much better to deal with the People part first
I thank everyone who has read and/or replied here - and endured some confusion.
OP point: TL1 ability to flag content is easily abused.
I presented a live example and requested community discussion. This was not a request for a specific change.
Note from the list of suggested/related topics that this is not a unique concept - others have brought up this topic over the years in one way or another.
This is a summary of many suggestions and points of view on the topic (not all from here) :
Consider TL2.
Define a policy about to acknowledge the situation.
Make mods aware of it and treat abuse of community moderation like any other.
Possible option for if/when Groups replaces trust levels.
Define/trace/limit over-flagging/reputation as number/% of flags by an individual, perhaps against another individual.
Ability to contest/flag/appeal a community flag.
Categories or tags that cannot be community flagged (like mod appeals).
“I don’t understand.”
“I understand but disagree that it’s a problem.”
“;TL/DR;”
This thread was mod-closed, I appealed that decision, the thread was re-opened - thank you very much for the opportunity to appeal. This is the established and proper moderation process - exactly the process which I’ve attempted to profile as being a problem with community moderation.
I don’t feel I have anything more to contribute. We don’t have solid concensus that there is a problem, or among those who recognize the situtation that there can or should be a technical solution. Cool. I was just looking for a discussion among peer administrators and respected platform “users”, not a conclusion. I look forward to other comments that might appear here over time.
I think this maybe isn’t as large an open goal as it first sounds.
Mostly, TL1 is a reasonably brief phase for most people (default 15 visits and some light participation), so the window of opportunity to do this isn’t that large before they become TL2.
But to hide a post it needs to accrue enough ‘points’ to push it over the hide post threshold (as admin/mod, you can see the reviewable score for a particular one by clicking on the question mark icon in the top corner of each reviewable in the queue). For low trust users they get a lower point boost than higher trust levels (only 1 point for TL1) so you’d need more flags from more users to build up enough to hide a post.
There’s also a user’s flag reputation as well, which has a direct effect on whether it adds or subtracts points based on how many previous flags have been agreed/disagreed with. So if the mods are disagreeing with all the flags these TL1 users are submitting then their reputation should tank and future flags will have even less chance of hiding anything.
This topic has some info in about how the user reputations work (and hide sensitivities too if you were looking to change something on your own site):
Also, it’s worth noting that the ‘something else’ flags don’t trigger a hide post.
I think normally when topics crop up about disingenuous flaggers it’s more about them wasting staff time rather than having too much of a public impact. Though the ‘hide post threshold’ is a dynamic number based on some magic under the hood, and I think I have seen at least one example of it being unintentionally super low which did mean it was more of a hair trigger than it was intended to be.
But if everything is working as intended then there are already some reasonably sophisticated safeguards in place.
@JammyDodger와 저도 언급했듯이, 동의/반대 플래그에 기반한 명성(reputation)이 %로 표시되는 시스템이 실제로 존재합니다.
플래깅과 같은 기능을 제한하기 위해 그룹을 생성할 수 있는 아이디어에 동의합니다. 이는 현재 Dynamic Groups 플러그인을 통해 가능할 수 있습니다. 다만 기능 요청이나 PR이 필요할 수도 있습니다.
또한 플래그에 대한 이의를 제기하는 것이 더 쉬워져야 한다는 점에도 동의합니다. 하나의 아이디어로는 템플릿 시스템 게시물에 이의 제기 링크를 포함하는 것입니다. 플래그가 달린 게시물에 대한 링크가 포함되도록 몇 가지 조정이 필요할 수 있습니다.
이는 사이트 팀/카테고리 모더레이터 간의 커뮤니케이션에 해당합니다.
이것은 해당 카테고리나 태그 내에서 플래그 버튼에 대해 display none을 적용하는 커스텀 #customization:theme-component의 CSS를 사용하여 구현할 수 있다고 생각합니다. “플래그 남용자” 그룹에 속한 사용자의 경우에도 플래그 버튼을 숨기는 데 사용할 수 있습니다.
훌륭한 토론이었습니다. 다양한 사이트 설정 또는 오버라이드를 그룹으로 이동하는 것은 분명히 훌륭한 아이디어입니다. 제가 언급한 플러그인은 이러한 것 중 일부를 달성할 수 있을지도 모릅니다. 저자도 Category restrictor 플러그인을 가지고 있는데, 이는 사이트 전체의 침묵 대신 특정 카테고리에서 사용자를 침묵시킬 수 있는 기능을 추가합니다.
이 주제에 대해 추가로 말씀드리자면, 정지(suspension) 시 이의를 제기할 수 있다는 메시지가 표시된다고 기억합니다. 그러나 이 템플릿에는 이메일 링크나 특정 그룹 또는 지정된 사용자에게 DM/PM을 보낼 수 있는 기능이 포함되어야 합니다. 이론적으로는 사이트 연락처 이메일이 충분할 수 있습니다. 하지만 더 쉽게 만들기 위해 플래그/정지 게시물로 연결되는 텍스트를 제공해야 할지도 모릅니다.
There certainly seemed to be at least one problem in the original situation you described, but that is difficult to diagnose correctley with limited information. There didn’t seem to be a clear technical problem that could be identified with the platform which has a fairly sophisticated flag system.
This is an important statement I agree with this. There can be all kinds of problems with leadership and staff policy implementation that are unrelated to the community discussion platform.
