Even that would help besides the currently just one setting. Our most severe abuser is TL1, and is indeed contributing interesting topics (despite the mass flag spamming), so downranking to TL0 (with all its other limitations) and only allow flagging for TL1+ is not an option here. So, yes, max flags per day
per TL would be nice and tremendously helpful here. If users progress to higher TLs, they can flag more. Makes perfectly sense to me, and those communities that don’t need those settings can default them.
The threat of being locked at TL0 (with a reminder of what they cannot do) could be added to a warning notice. That may jar them into thinking twice about being so free with their flags.
Thank goodness the only flags we had were for Discobot.
You have three options (well, four)
-
directly intervene with the user and indicate there will be consequences for future negative behavior (casting lots of inappropriate flags)
-
turn down “max flags per day” globally in your site settings
-
lock this user to TL0
-
reduce flag sensitivity in your review dashboard, so it takes a higher cumulative flag score for thresholds to be met
I suggest you try those out before jumping to “discourse must add a new feature because of {problem user}”.
That said I am not opposed to limiting flags per day for low flag reliablity users, it is a sensible direction.
Thanks, @codinghorror, I opt for option 5, patch core in our case as our communtiy seems to be a border case, compared to others.
BTW: “Reduce flag sensitivity in your review dashboard” is not set in the review dashboard, but in Admin > Site Settings > hide post sensitivity
I’m having this problem too. One user abusing the flag system. Kicked up a huge fuss when demoted to TL0. The flagging is the only problem so I already reduced flags to 2 per day.
I wonder if there’s a way we could add a user to a ‘no flag’ group and stop them from flagging completely?
I’d also welcome an algorithmic temporary ban on flagging triggered on a bad flag together with a low flag score.
This seems like a challenging issue, does there seem to be any legitimate reason why they are flagging things or no?
If there are not, that may be grounds to suspend or ban their account for that.
If you ask the person who is making these flag reports about if they have any reasons they can explain for those, don’t see what the problem would be unless their reasons aren’t considered to be valid by the site management + moderation team.
One potential strategy is for there to be some kind of a price for flagging, as in whoever is making these flag reports will be asked to explain their reasons for the flags. In addition potentially they could be asked to help with mediation in talking with the person/account that is being flagged about what is causing them to flag things that other folks don’t agree need flags.
Make flag spamming part of your user code of conduct or terms and then use the Discourse moderation tools - warning, silencing and suspension. Treat flag spamming like any other disruptive forum behavior - set out clear standards of what is acceptable flagging and also consequences when the system is abused.
That is important the terms and conditions + code of conduct and community guidelines, here is quote of the official statements for this support site and the default for new sites unless edited/changed:
If You See a Problem, Flag It
Moderators have special authority; they are responsible for this forum. But so are you. With your help, moderators can be community facilitators, not just janitors or police.
When you see bad behavior, don’t reply. It encourages the bad behavior by acknowledging it, consumes your energy, and wastes everyone’s time. Just flag it. If enough flags accrue, action will be taken, either automatically or by moderator intervention.
In order to maintain our community, moderators reserve the right to remove any content and any user account for any reason at any time. Moderators do not preview new posts in any way; the moderators and site operators take no responsibility for any content posted by the community.
This clearly recommends and encourages people to flag anything and everything considered to be “bad behavior,” which can definitely be a reason why people are flagging things more often than not. This also puts a lot more responsibility on moderators to deal with problems/arguments, as opposed to guest members having more opportunity to work through things amongst themselves.
Could start a new topic about overall moderation + flag philosophy and to develop drafts for alternative community guidelines to this. I spent a lot of time reading through these guidelines and have found difficulty with a lot about how these are written for my own ideals about launching new sites, may be best to start over completely with independent policies for each site.
Yes it does encourage users to flag bad behaviour, but I disagree with you saying that the forums FaQ is the reason behind flag spamming. A reasonable person would know what to flag and what not to flag depending on the context of the post etc.
The behaviour is problematic for moderators as it creates a higher workload. As Lilly said above, you need to create rules around flagging or go down the official warning and suspension route.
That would have been quite a different statement to say those community guidelines are “the” reason for spam flagging, what I wrote was just that they “can be” one of many possible reasons why there are many flags.
Difficult to know what all is happening without more information about context and hearing both sides of a situation. I am very cautious to accept any one person’s testimony as an absolute truth.
I agree with those urging communication with the user before implementing settings changes.
Ultimately, I would guess that you want the behavior to change. Restricting the user’s ability to flag won’t give them that opportunity. I think a really important piece of moderation, especially in political forums, is to lead with communication and not punishment. Be clear about what behavior needs to change and why, with examples, and what will result if the user doesn’t change.
I think it’s good to start with an open-ended question to the user, like: “Hey [user], I noticed you flagged [other user’s] response to your post. Can you help me understand why you flagged it?” Or something similar, so you can gain an understanding of where they’re coming from, and respond to them accordingly.
Culture shifts can’t be legislated or settings-changed into existence. Settings help us design guardrails but it’s communication and enabling users to demonstrate change that will do it, at the end of the day.
I think you may be underestimating the role that the guidelines can have in shaping this behavior, though.
Four years ago, in response to the OP here, @riking offered up a possible warning message regarding flag-gaming:
…But the guidelines actually say the exact opposite of that, and at least heavily imply that flag-gaming may be an effective tool:
(Which basically is exactly how Facebook works — except for the “or by moderator intervention” part, since their first-line seems completely automated at this point — so if mods are going to claim that Discourse doesn’t work like that, it’d be a really good idea to change the guidelines to reflect that claim.)
In addition, @anon36555649 suggested possibly putting the flagger together with their flagee(s) to discuss the issues behind their flagging abuse. I don’t really agree with that suggestion at all as I think it’s a recipe to heighten tensions, not reduce them, but if someone were to go that route it would basically be in direct opposition to the guidelines, which advise users:
Take the really murky, hand-wavey definition of “bad behavior” — a painfully vague term, employed in a situation that calls for being extremely specific and clear — and combine it with a few users who are all too willing to misunderstand, particularly when they think it’ll be to their advantage, and it’s easy to see how these problems can come up.
Not quite. That sentence is meaning if enough flags from different users accrue.
Exactly! That’s precisely how flag gaming works. A few people gang up on a post, or a user’s posts, that they disagree with or want to silence, and together they trigger the automated system to take it down.
That is the textbook definition of flag-gaming. Each individual user only contributes their one set of flags (which is often how they escape consequences, also, because individually they “didn’t do anything wrong”), but it only takes a fairly small group to game the system.
(Those users also don’t have to be colluding or conspiring with each other, they merely have to be like-minded and share the same targets.)
Why not just kick them out, right away without warnings? And making it public to teach every user that shitty behaviour is not accepted any way.
It worked on my Facebook group of close to 30k members on risky topic.
Then you have a moderation problem.
Do group mods/creators have visibility into the identity of flaggers, on Facebook? I know that regular users don’t, even when they’re the ones targeted — Facebook keeps the identity of flaggers completely anonymous, so there’s no way for someone who’s been targeted to know who’s gaming the system against them. (And Facebook’s own human moderation staff is famously absent from participation or contact with regular users, typically.)
This entire conversation started as a question about how to better moderate such activity, so that seems like a given.
You miss now the obvious point, and you must be a coder It has been never about flags. It is about toxic behaving and how to deal with that. Fixing one symptom that way it doesn’t bother moderators is pseudo act. The actual issue is still there.
I’m not underestimating anything. The guidelines set on a site are (or should) be obvious to understand as to avoid any misunderstandings. If on your site users are misinterpreting the guidelines, it means they weren’t written well enough and have to be amended or you have a moderation issue on your site if you’re being swamped by flags and can’t deal with it.
mmh maybe but at the end of the day all the flags will be reviewed by a moderator so they will have the ultimate choice whether to reject / agree with the flag.
Personally, I’d go for what riking said in sending a warning or what Jagster suggested just suspend/remove them if they are going to be an issue.
(Those users also don’t have to be colluding or conspiring with each other, they merely have to be like-minded and share the same targets.)
This seems very close to also being the description for community moderation. But if the same group of people are continually mis-flagging the same type of post/the same user then this pattern will be obvious to spot from the Review Queue info and they can then be dealt with further by a moderator.
There are also some built-in guards to dampen this behaviour as well where the accuracy of someone’s flagging has an effect on how much the auto-tools value their opinion. If someone keeps on casting flags which a moderator rejects then their accuracy score gets lower and lower, and their flags carry much less weight.
Sites can also adjust the thresholds for auto-hiding posts too, or turn it off completely, which can be very useful if they encounter a problem like this.
Ultimately though, if people are using flags to bully or harass other members then they’ve got to be dealt with outside of the auto-tools, as no one wants that behaviour within their community.
In addition, @anon36555649 suggested possibly putting the flagger together with their flagee(s) to discuss the issues behind their flagging abuse. I don’t really agree with that suggestion at all as I think it’s a recipe to heighten tensions, not reduce them, but if someone were to go that route it would basically be in direct opposition to the guidelines, which advise users:
And, for the record, you’re right. That is a terrible idea. I do not recommend that in almost any circumstance.