BBS Thread/Topic trigger warnings

Really? After almost 3 years, we can expect the volume of flags to suddenly, magically increase?

I’d say that’s … unlikely.

I do think some further restrictions on new users in certain “highly active” (controversial) topics may make sense if we can nail down specifics.

3 Likes

Well, I’m doing my part! :smile:

Actually I used to feel guilty when I flagged, especially if it’s someone I engage with regularly who’s just having a douchebag moment. But I’m over it now.

6 Likes

I flagged this!

11 Likes

Great! That is roughly what I was trying to get confirmation on with my post :slightly_smiling:

I understand the problem with the system, as it is now, is mostly a lack of users flagging posts, I did not know or expect that. But I guess I can see how that would happen.

Flagging is something that I take seriously. Like, I reserve it for really awful stuff, or even only for illegal stuff. I think a lot of people “feel” like that. I’m not saying that is what the button is for obviously it can be for anything we/you decide. That is just the feeling I get from a flag button and I expect others have more or less the same feeling.

I sparsely use the heart button, when I’m on Reddit I hardly never up- or downvote anything. So I don’t think changing the look/name of the button will really help. And I understand a lack of participation among 90% of the visitors is actually a normal function of webcommunities so maybe this cannot at all be improved.

I would be curious to the effects of a request placed directly at the readers. When you get a huge thread, with one of those summaries at the top, maybe add a graphic or a text in the vein of “we need you to flag down unwanted comments”. Kind of like Smokey the Bear or Uncle Sam. I wonder if that would spur people into action.

That said, that rate comment feature looks really sexy!

I really enjoy the fact that you keep on improving this software and taking input from the users. I appreciate that this is a hard process and it is impossible to please everybody and preemptively fix every situation. So I get you have a hard job. But if you only manage to slightly improve the way people interact online that is a great achievement!

5 Likes

Not to keep harping on about the article tags… but I will.

Help me with my reasoning here, I’m full of the flu and everything has gone all echoey. :mask:

So, first there needs to be a way of nailing down those specifics, of (pre-)determining which topics have the potential for controversy / for eliciting bad behaviour from new and extant accounts (I’ve also noticed such bad behaviour coming from ‘sleeper’ accounts).

So, discourse has access to all this myriad historical data. I suppose that includes topics which @falcor has had to close early and reprimand multiple users within. I go on to suppose that this could be cross-referenced against the tags that those articles were labelled with in order to determine the type of topic that this behaviour is likely to emerge within.

I see that the tags themselves seem only to appear at the bottom of the boingboing.net articles and not within the summary pull-down within the bbs. But I imagine it would not be difficult to automatically pull down said tags from the articles associated with the bbs topics. This might be an misassumption, perhaps involving far more work, coding and error checking than I am imagining.

(Perhaps there are other applicable metrics than the tags, but that’s the easiest way for me to imagine this so please assume that any better idea you can think of take the place of the tags.)

The main problem or issue, whatever, seems to be the low volume of flag data to draw quick determinations from as the comments to a topic come flooding in.
The current model seems to rely on letting the topic gestate and then reacting to it with an auto-temporary-closure once a certain number of flags have been applied, then awaiting mod attention to read over, tidy-up and perhaps then re-open the topic once personal warnings have been issued or even accounts suspended or banned.

Is there a place for predictive models? If historical data could be added to the pool of a model, perhaps a prediction could be made about which topics are likely to generate controversy. and then compared to actual behaviour in order to sharpen said predictive model?

I see that there are already messages from the system, such as the ‘you’ve posted a bunch of times already, take a deep breath’ and the suggestions about previously posted, similar topics made available to users. Could similar messages be issued to users as they start to post in a topic, along the lines of ‘Remember how you were reprimanded with a personal message and had multiple comments deleted by our very busy mods in a topic very much like this a few weeks ago, please take the time to compose your thoughts’ or something?

You’d also have a heads-up for the mods about topics that were likely to generate controversy before it had even attracted the trollsome comments. If they already have to spend the time going through temporarily closed topics, cleaning and gutting them for palatability, perhaps such a heads up would lighten the load before such a backlog could develop? Again, I’ve no experience modding so not sure if this would mean more of a burden or not…

Or how about for new accounts "Welcome to the bbs, we’ve noticed that threads concerning this topic are prone to attract many negative comments from new users, as such your flagging threshold has been lowered for this topic. Contributions to other, less controversial topics will maintain your flag threshold at the ordinary level.’

I guess you don’t even really need to send such a message to new accounts when they are entering into such controversial topics, just make a note of it somewhere in the terms and conditions…

*rasping cough

2 Likes

One thing that would help is not so much a focus on individuals as a couple of specific behaviors that keep coming up a whole lot when people get riled up.

In particular, once things get firey we have people replying to people without reading/bothering to digest the post they’re replying to, and instead simply grabbing a sentence to reply to and regurgitating the same point they started with. This ends up creating huge cycles of essentially useless conversation and frustrating everyone involved.

A fresh example of it is here: Freedom: a #blacklivesmatter anthem

It happens WAY before people get frustrated and is really obvious when it happens, but it’s not necessarily crossing a line so we can’t really do anything about it.

If we could get a public warning when a topic was potentially/was getting ‘heated’ that it was time to raise the bar that we could all refer back to then a lot of these issues wouldn’t escalate and we wouldn’t have so many otherwise interesting conversations getting derailed and/or ending prematurely.

It’d require a second tier of rules, but that shouldn’t be too hard to put together and there are examples of the same put together in debate forums and such online.

3 Likes

How do you automatically determine such behaviour is taking place though?

Relying on mods reading through everything or getting such info through a different flag won’t fly because of the extra work for the mods and the lack of flagging here at bbs.

In theory you can detect it as it’s happening when there’s a large chain of responses between any two users in a thread…that’d just be a query, right? It wouldn’t be a proof (people can go back and forth for a lot of reasons) but it’d bound the box and indicate the convo is worth a second look.

But you shouldn’t count on just that, there’d have to be three other components

  1. A set of forum rules that specifically dictate the standards to be followed when the conversation has been flagged as ‘heated’ or ‘debatey’ or whatever
  2. The use of the ‘Something else’ flag if a new one can’t be created by users in the community
  3. A Mod would have to post in the forum when the conversation had reached the point where we were at the second ‘tier’ of standards. (This would replace abruptly closing the thread and hopefully happen a lot earlier in the process)

So: Part of it could in theory be automated, but really we’re talking about replacing one mod action (closing a thread and ending conversation) with another (notification of ‘time to behave like adults’ with a list of the rules cut and pasted in for easy reference)

Once it’s been done a few times we should be largely trained as a community, and the the primary point of action isn’t the automated component (which would be neat if it worked, but it would by definition add work because there’s no way to perfectly automate it) but is instead the manual component that’s already happening anyway.

1 Like

If what you’re talking about is kinda like a pre-mod action, some reconditioning of the thread in some way before serious mod action need take place, perhaps some TL4 promotions could take care of that.

Or maybe a TL3.5 condition that leverages this pre-mod activity from voted-in users from a TL3 poll.

I know who I’d vote for.

The abilities don’t need to approach proper mod powers but rather could be used to warn or cool down offending users and maybe pass information on to the mods when srs action need be taken?

There are for sure some cool heads in the Regular pool who comment a lot and take the time to make salient arguments without losing their temper and I think we all know who they are… for the most part.

6 Likes

Yeah, and there’d be some flexibility… the notice could be right there in the OP if we thought it’d get heated right off or else it could be added later when things did.

The users in question shouldn’t be that relevant…it’s rarely one user or one conversation, more often it’s a general escalation that gobbles up a lot of people on a big analog scale. It’s the specific behaviors that are the issue after all, and it only applies to specific threads. (not always predictably).

Plus by focusing on the thread we avoid the ‘race to the bottom’ syndrome.

It’d help if it was the same thing either way. You know "This thread has been flagged as ‘X’ and it’s time to adult. In this thread these additional rules are in place (always read the post you reply to, refer back to ‘thou shalt not commit logical fallacies’, no personal attacks, follow rules of civility, etc.)

And like you said, there are quite a few people here who are plenty level headed and probably could be given the additional ability to formalize the decision to make a thread one we have to behave in.

(Not me of course, I like being feisty! I’m gleefuly unqualified and have to waste too many neurons on that stuff IRL) :wink:

2 Likes

I should probably make it clear that I don’t think I come close to being right for that kind of job.

Also, I think there would be some extra pressure to not participate with full, natural ease if there was the weight of some extra responsibility for mod-like conscientiousness.

3 Likes

Or me. But I have gazed into the abyss of tech support too long, like honestly surprised humanity has not destroyed itself yet on a regular basis. I have such low expectations of my fellow human.

4 Likes

I used to, but now I also use it when someone is just ruining conversation, demeaning anyone who disagrees, and generally dropping turds in the punchbowl. Also for new users who come in swinging.

As a long time user here, this HAS been improved.

though unlike reddit, flags are not downvotes. Flags are a sign that something has got beyond the communities ability to moderate.

seconded heartily.

7 Likes

Spent a couple of years on the front lines myself. I have seen the dark heart of humanity… and it needs a triple bypass.

Also, I enjoy being a snarkosaurus too much.

3 Likes

Problem is the same as with any other set of guidelines. When someone decides that no, they took a vote and one to zero, the vote says someone is RIGHT and someone else is WRONG.

not much will be done by software to allieviate that users impact on conversation.

Who know what that person needs, but the community needs to put them on a time out like when they were 5, becayse they are behaving like a 5 year old with a good vocab who needs a timeout.

I believe a single flag from a TL3 user (you and I at least) already shuts down a new user comment until mod review.

The only suggestion I could come up with would maybe be a halo or something around a brand new users first couple comments. Like this, effectively:

2 Likes

Guilty!

but “it was a dark and stormy night” as an opening is a great reason to put a book down, isn’t it?

Rather a lot to ask for someone to seek out the nuance in sentence three when sentences one and two were dismissive and querulous. If a user made that wine, that user can drink that cup.

Should I really engage with the nuanced parts and de facto validate the foundation of cards?

I agree. Many people are profoundly unaware that they are covertly aggressing against others in conversation. It’s the ‘thank you, but’ the “I support this, except…” and the "i hear you said you feel but really don’t you feel [Y]?

How we encourage individuals NOT to disrupt a conversation by being conversationally demeaning (less than overtly) when it is pretty clear there is no actual communication going on. When saying ‘hey, that is rude’ yields you a 'how rude you are for singling me out".

I’m a rude fuck sometimes, but not to people arguing in good faith. Not to those who validate but disagree. It’s the invalidation of the people being diagreed with that is getting reactions.

Well said.

5 Likes

correct!

2 Likes

I think this only happens with spam flags.[quote=“AcerPlatanoides, post:115, topic:67711”]
Problem is the same as with any other set of guidelines. When someone decides that no, they took a vote and one to zero, the vote says someone is RIGHT and someone else is WRONG.

not much will be done by software to allieviate that users impact on conversation.
[/quote]

Do you mean that their opinion of their rightness will not change? How about, instead of a warning, a warning and the lowering of the flag threshold to 1 flag from a TL3 user?

One shots from TL3 and above could hide the comment for mod appraisal in those hot topics. Of course, the writers or mods would need to pre-tag the topic as potentially ‘hot’ in order that such conditioning could be applied.

4 Likes

Someone very smart recently mentioned to me in a PM:

Say “I wish the clothes were better” vs. “the clothes are not to my person taste” - one is a value judgement about the object being discussed. The other is a statement of personal opinion.

almost like that sort of thing matters to some people very much, while others are… seemingly mindblind to it and incapable of validating others.

Oh well, it’s when one is above and beyond reproach that the behavioral issue comes in. Be mindblind, just stop being rude about when someione says “hay you, stahp”

I’m really tired of reading arguements where someone argues about what is and is not hurtful to another person.

That sort of demeaning shit should be a 24 hour timeout.

Don’t tell anyone they’re not offended or wrong to be so. DO tell anyone they’re explained their offendedness enough and have cross the boundaries of another repeatedly in a querulous driving trollies ego trip.

Basically, if one cannot IMMEDIATELY say “maybe you are correct” when someone else says “you are behaving badly” then one probably need stop put the keyboard down and have some alone time.

4 Likes

See your point. I mean that when someone is RIGHT, their behavior will not change.

Behavior is the issue, not opinion,. Holding a different opinion without dememaning the other is an adult skill.

I happen to not disagree much in opinion from a lot of people whose styles I find downright repellant. One can be correct factually and still be quite wrong socially. .

But when the facts up for debate are the thoughts and feelings, motivaltions ofone of the parties… that’s no longer a conversation about a topic. When it is about behavior it still is.

Too often people drift into ad hominem responses as though reducing someones credibility, showing them disrespect, somehow discredits their ideas.

I show people disrespect openly and not covertly, for behaving disrespectfully, even when I agree with them.

I’d rather go to the barricades with gentlemen and ladies, thank you very much. :slight_smile:

4 Likes