Time limits for unfocused discourse threads?

As commented by @mbauman in How do we Julians win big when the situation is so unfair? - #75 by mbauman “open-ended” threads (ala “how can Julia be more popular” or “what don’t you like about Julia”) rarely seem to go anywhere productive, require continual attention by moderators to split subtopics off from the ever-branching discussion, and by construction can never be resolved satisfactorily.

In that recent such thread, @mbauman took an alternative of imposing a 24-hour auto-close clock. This allowed people to get a last few thoughts in, and didn’t single out any particular poster, while preventing the discussion from cycling endlessly.

Can we do this more consistently? e.g. have a new flag for “unfocused topic”, where if it’s approved by moderators it results in a time limit on the thread (e.g. 48 hours).

37 Likes

I’m actually okay with it. Maybe I see the thread reply goes up and what I believe is that I created lots of engagement. Maybe put some good metrics on what constitutes a good thread than just bringing people in?

2 Likes

I think that is a good tactic once the discussion stops moving anywhere interesting, but I wouldn’t want to get in the habit of imposing that restriction right away on new topics. That would be overly strict IMO (like StackOverflow). Precisely focused topics with directly implementable solutions are often better on Github anyway.

Plus, the “unfocused” topics often garner much more engagement which has some benefits:

  • community pulse on the current state
  • diverse opinions
  • byproduct PRs
1 Like

“Engagement” is a good metric for companies looking to raise ad revenue, but you don’t have to look very far on the Internet to find unpleasant forums with lots of engagement.

Many unending discussions can simply suck people’s energy, foster acrimony, and make this feel like a less productive and less enjoyable space. It’s hard to simply ignore threads that devolve into gripe-fests, but at the same time more engagement does not and cannot lead to such an open-ended thread being “resolved” … and generally has the opposite effect: an explosion of topics under the umbrella of a single thread. Even if the discussion contains some useful comments, they are difficult for later readers to find if they are buried in a single long thread, so they also make the forum less useful as an archive.

26 Likes

I also sought some informal feedback from the slack #community channel last Wednesday — folks there were largely supportive but noted that the messaging is important.

For example, we could allow for some open discussions as long as it’s productive (and not too circular). The gradual descent into the abyss is often visible from the edge. At which point the message is simply:

This discussion seems to have run its course; let’s plan on bringing things to a close in 24 hours.

16 Likes

I see this as a benefit if they are able to be broken out properly. Natural discussion in a meandering thread can lead to topics (and PRs) that wouldn’t have been brought up otherwise. Maybe there is some Discourse tool that could nest or link to a new topic from an existing thread bringing the readership with it? Some middle ground between starting an empty new thread from scratch with no followers and continuing to dogpile on the same mega-thread?

1 Like

Moderators can and do split off posts into new topics (which link to the old one), but this requires the mods to continually monitor the thread (as I mentioned in my first post above). This cannot always be done cleanly since individual posts within such a thread often intermingle multiple topics. Moreover, the split-off threads are often sub-optimal, e.g. lacking a clear statement of the new topic, because they were originally written to be read in another context.

I think the forum is a much more pleasant environment when people are encouraged to start new threads when they want to raise new topics, rather than tacking them onto old threads and relying on moderators to disentangle them.

(Personally, I dislike following long-running mega-threads just in case someone happens to say something of interest, but they are also hard to ignore, especially if they are going downhill. Whereas a clearly labeled, focused topic can safely be skipped if it’s not in an area of interest to me.)

4 Likes

I agree totally. I mean a new option for users (not mods): maybe this new button would start a “related” thread that links and alerts readers of the source thread.

That’s already possible! It’s just not very discoverable. Click that little image button when formulating a reply to a specific post and you’ll see a menu with these options:

image

8 Likes

I did not know about that! Does it only alert the one person you reply to though? I’d like to make sure it also migrates people who are lurking. Or maybe it leaves a link in the original post that people would see?

This post actually ended up being a good example.

  1. The original topic is thread time limits.
  2. The discussion prompted me to think about alternative methods to discourage mega-threads.
  3. I replied with one such idea (that would not have occurred to me without participating in this thread).
  4. We are now on a tangent about how a similar existing feature works.

Was this an okay reply in this thread or should it have been a new linked thread?

It is a good example — which post of yours would make for a good first post of a topic? None really make sense on their own.

This is precisely why topic-splits aren’t always an effective moderation tool.

My mind goes to gardening metaphors — topics are a bit like a plant pot. Ideally, the OP would spend some time doing research, prepping the soil, labeling the plant, and carefully planting the seed of an idea they hope to grow. Unfocused discussions lead to many others bringing their own seeds and prompting other off-shoots from the main theme. Those secondary saplings and cuttings often aren’t started with enough intentionality to survive a transplant.

OK, maybe that’s a little strained, but the point is that the effort folks put into replies is typically well below what’s required to start a good topic… and it’s just not possible to split coherently in a way that leaves both threads sensible.

What I think might be ideal is if participants themselves note an aside starting to grow and then preemptively create a dedicated topic for it (if it’s an aside they care about). Continuing to reply to the aside (as we’re doing here) only makes the original thread harder to follow.

Please do try starting a new thread about topic splits if it’s something you care about!

6 Likes

Back on track, one thing I — as a moderator — really like about timed closes is that it avoids any perception of me (or any other moderator) “getting the last word in.”

I think the only question I have is if we should flag unfocused topics from the get-go (and give them something like 7 days by default) or if we should just wait for things to develop (and then give 24 hours once we think an intervention might eventually be needed).

7 Likes

I think that’s a judgement call that would have to be made on a case-by-case basis.

For example, with topics that simply echo many previous threads of this type (ala “how can Julia be better”), I suspect that putting a timelimit on quickly might be appropriate, maybe accompanied by a link to previous threads. For more novel topics, excluding flamebait, I think the moderators would want to wait to see how it develops, but I would still lean towards timelimits if it turns out to be branching rapidly.

3 Likes

I’m going to start an aside again here because I think it is related enough and I don’t think it warrants a grand new topic …

That is a good point. If I am in a hot thread, see a tangent starting to form, and decide to start a new topic, then it is going to take me a while to write a solid first post. Hopefully by then there haven’t been ten other quick replies in the original post that have already moved past that point.

The slow Discourse mode may be effective at encouraging smaller, focused threads by removing the race to get your ideas out before the discussion moves on. Time limits may have a similar effect though since you wouldn’t want to start a new aside that you want to bloom in a dying thread.

Making it easier for users to seamlessly transition a discussion to a new thread topic and encouraging that behavior should hopefully help avoid long, unfocused threads from forming. Maybe there should be some way to mark in an existing thread: “writing a new related topic on X, so get your ideas ready to post at the new location not here”.

Yes, the above paragraph is ironic to post here, but I don’t have a good feeling for where the line is! If every post is very strictly on topic, then the discussion thread just amounts to agree because … / disagree because … / one solution … Discussion and Discourse comes from wandering slightly.

I think that in the majority of cases it is possible to tell that a topic will be unfocused from the very first post. My impression is that the majority of users do not star topics like this. It is a few (<10) individuals who perpetrated these in the past year or so.

I have been limiting my participation in this forum (compared to the past) partly because these topics make this forum less pleasant than it used to be.

I would propose that we emphasize that certain kinds of posts are not welcome here, and then after someone violates this, require that a moderator approves their posts for a time duration (eg 1 month). The OP whose topic gets a lot of “unfocused topic” flags should have the opportunity to close / delete the topic for a short grace period without any sanctions.

10 Likes

I agree with this—you can spot these threads almost from the title alone most of the time. The trouble with “waiting for the discussion to stop moving anywhere interesting” is that interesting is subjective and a discussion can still be interesting in some valid sense yet be a general emotional drain for anyone who makes the mistake of getting involved or even reading, and ultimately be unconstructive in the sense that no positive actions will actually stem from it. Such a judgement also requires moderators to be following along, which, let me tell you is a huge burden in both time and emotional energy. I completely ignore these threads, because I’ve been doing this too long and I do not have the time or energy to get sucked into yet another vague unactionable thread.

So I would favor a timed close approach: a moderator spots one of these threads and as soon as they do, they put a close timer on it; when it’s closed, it’s closed. 24 hours may be a little short, we could do 48 instead. If there’s anything interesting in the thread, people are free to spin off more specific threads.

I would also add that we already tried to implement a “no overly broad” discussion topics policy, which this is just an iteration of really. This is allowing them for a while instead of closing them immediately which we haven’t been doing, I suspect because it feels a little too strict and may seem defensive.

16 Likes

I guess I have some sort of psychological trait that makes me easily pulled into these sorts of escalating discussions. I try my best to behave myself and be a non-escalating influence, but often fail at that. Even if I say something non-inflammatory, just posting in the thread can cause it to get refreshed to the top of the recent posts screen.

So I often appreciate (at least in retrospect) when cooler heads come along and close discussions that are going in bad directions.

This makes me think though that maybe an additional lever available to the moderation team would just be to just block a certain thread from being bumped to the Discourse front page each time a user posts in the thread? This could give a more organic way for threads to die down.

8 Likes

Please go make an account on Meta and :heart: that exact feature request here:

We can completely unlist a thread, but then it’s completely hidden and only folks that have notifications or a browser history can find it.


I often think about not just what I write, but also how folks will likely be wont to reply. And I do actively fight the urge to “get the last word in” — if I’ve said my piece, I’ve said it and continuing to reply just isn’t valuable.

5 Likes

Done.

2 Likes