Dealing with HNQ drama before it happens - how do we deal with questions that can lead to trouble?

Question

Sometimes stuff creeps into HNQ that doesn't belong. While I'm a moderator of one site that doesn't get into the HNQ, and the worst I'd face on the other is a question that's decidedly meh, having a rough idea what to do when I see a potential hot button issue on another site in the HNQ list would be nice.

We already have means of auditing what got there and kicking something out and means to handle a community emergency in progress.

That said, we can check when and if a question is HNQed, but not what's HNQed as I understand it, so extra visibility is nice, so our vision's not 20/20.

How best do we assist community moderators in seeing potential issues? While we have standard flags if a post is rude or abusive - should we be custom flagging them to let a moderator know it's HNQed and potentially problematic? If we're interacting with them on chat - would mentioning such a post be a useful escalation method?

Answer 1

If you see something in the HNQ that you don't think should be there, you have several options, everything from editing to closing to flagging for moderator attention.

If you think the question is on topic but unfixable, please raise a flag. If you also happen to see one of the mods for that site in chat, feel free to let them know about it but please, also flag. Chat messages are not something that is easy to track after the fact to know why something was removed and we don't have a way to let moderators indicate why they removed a question at the moment. Using a custom moderator flag addresses all of this. A generic "remove from HNQ flag" would not do this.

I've been wanting to compile a list of solutions for various situations. There are a couple of things that we've already implemented to address problematic content in the HNQ and there's some things that are in my list of stuff to consider as future upgrades to the HNQ list.

Stuff we already have:

Per-site and network-wide title blacklists

If there are words showing up in titles that should be prevented from showing up in the HNQ list on either a per-site or network-wide basis, we can add those words to the blacklist to prevent the questions from ever showing up. This is a pretty blunt tool so we need to be sure that the questions will be considered pretty universally bad for the network. There are likely many NSFW words we could add to this list. If there's something specific you think needs to be added to this list, please make a Feature Request about it on the relevant meta site.

Question editing - particularly the title

If the question is a good fit for the site but the title itself contains problematic content, editing the title and the body to clean up that content is a good start.

Question closure

Users on the site can close a question to get it off the HNQ list. Now, this requires that the question be a bad fit for the site for more than just the content - too broad, unclear, or off topic - but it's definitely an option that should be considered.

Moderator question removal

Flagging for moderator question removal is the best way of getting attention to a problematic question to have it reviewed for exclusion. Do remember, though, that moderators are not always on the site, particularly on weekends and especially on holidays. The options above this are preventative or something most users on the site can do to address the issue. They're preferred both because they don't require moderator intervention and they can be undone. Even a question flagged for moderator attention may only be edited or closed if the moderator feels that's the correct solution.

Stuff we don't have that might be useful:

Delay before a question can be added to the HNQ list

Right now we're working on adding a delay before a question can be eligible for the HNQ list. This means that a question will need to be at least 6-8 hours old before it can appear on the list, even if it qualifies otherwise. This is designed to give the users on the site the opportunity to vote on, edit or close the question. If a question is problematic, you can vote it down as that will prevent it from showing up on the list and since it can't show up in the first few hours, the question (and answers) will get community votes rather than network-wide votes that tend towards upvotes.

Give moderators (and other users) a heads-up when a question appears on the HNQ list

There are many ways we could implement this. I personally favor a HNQ hub of some sort - similar to the bounties tab - that would only be visible if there were at least one HNQ list question on the site. This would give a quick overview of what questions might need some attention. In conjunction with the previous tool, we could include new questions that will be added to the HNQ list once they meet the minimum age requirement. This allows users to vet the questions more closely before they are added to the list.

I'd also want to make this list a bit more sticky than the regular HNQ list. The HNQ list updates every 15 minutes, which means questions at the bottom of the list may come and go several times. I'd probably want this to capture all questions that have appeared on the HNQ list within the last hour or two unless they've aged out completely or been removed by a moderator.

There's a suggestion on the announcement post (and in a comment on the question here) to make this light up the diamond indicator like the Meta posts do. I understand that but it's not my first choice for solution. I don't personally see a reason to limit this list to moderators only. Such a suggestion also has a risk of being noisy on sites with many HNQs, particularly if that site rarely has controversial ones.

Let mods indicate a reason for removing a question from the HNQ list

This would help us see connections better - why was a question removed, is there something about the question that we could have filtered for to prevent it from showing up? Without it, we're having to guess. Though, in the interim, moderators wishing to explain removing a question from the HNQ list can flag the question themselves, explain the reason, and then handle the flag.

---

There may be other stuff I'm missing (it's early and I'm half-awake) but these are things I think will help a lot. If you have ideas, feel free to add them to my announcement post of the HNQ update.

I'm hoping that the minimum age update will be out later this week or next, so that should be good to see what impact it has.

Answer 2

in need of moderator intervention

A problem not listed above that requires action by a moderator. Be specific and detailed!

Since ♦ moderators have the ability to directly remove a question from the Hot Network Questions list, and the only* feasible option for regular users is to close the question (which is usually not the desired action, and requires 5 users with 3k reputation to do so), I'd say this is a valid use case for a custom moderator flag.

Since there is also a time component involved (this is not a flag which can wait for a day), pinging them in chat is IMHO a valid option to convey the urgency of your message. If one of the moderators indicates they don't want to be pinged for this, you should of course respect that, but I can't imagine them getting mad at you for trying this course of action.

*_{: Protecting the question can be done by a single user with 15k reputation, but is almost never sufficient to stop any potential drama. Downvoting question & answers helps a little but only delays the inevitable and might not be objectively warranted either.}

Answer 3

Let's get this straight. A question from SE Workplace entitled

Caught masturbating at work

entered the HNQ over Easter, and remained on that list even after its title was changed by a mod. It took another mod to pull it off the list yesterday. The question with its 5,000 visits is now closed.

In the name of everything good, how the heck does that word pass through the nets? Aren't sirens supposed to be wailing whenever words of a sexual nature are in titles?

Apparently not, because I recently flagged a question on EL&U, whose title was:

Origin of “cooter” meaning “vagina”

My message:

A perfectly legitimate English language question, and its title was perfectly acceptable too. How would you edit that title without harming or misrepresenting the OP? But in the light of recent events, it was an inappropriate question for HNQ. There was another question which I also flagged but that was a less egregious example, although that too was pulled out of HNQ, so flagging does help.

My point being, those type of questions shouldn't even enter the HNQ in the first place. Sanitizing the Workplace title helped tremendously (How can I regain a professional atmosphere with someone who may have seen me misbehave?) but all SE needed was a female developer to tweet the title and ask their followers, "look at the latest offering from Stack Overflow/Stack Exchange", and now The Workplace or EL&U would be facing temporary suspension from HNQ.

Or wouldn't they?

Answer 4

How best do we assist community moderators in seeing potential issues?

I think the best possible way to do this would be via flags. The problem then is, by the time a custom flag gets handled, the question may have already spent a day in HNQ or so.

Not all moderators use chat, and requiring them to would be a bad idea in my opinion.

I'm personally a big fan of Catija's proposal to allow sites to decide on a "lag" time before a question becomes eligible for HNQ, because it would nearly completly solve the issue of custom flags being slow, while at the same time allowing more space for the community to course-correct.

Answer 5

TL;DR This concern is rather tangential to main interest of active site users and moderators and if we want it addressed we better make changes to the system to make it easier for them to deal with it.

Let's not forget that primary motivation of site users and moderators is to ask, answer and maintain on-topic questions and it's unrealistic to expect them invest a lot of efforts in something barely related to that. One can argue that HNQ impact how site is perceived from other parts of the SE network and possibly attract some new users from the network (to rather limited extent) but this seems too weakly related to site primary mission ("questions, answers, no distractions") to make people do a lot of work improving things.

---

Thing that feels most troublesome to me is how it is supposed to work with off-topic questions? A normal way for an active site user with these questions is to vote / flag to close and move on.

I can see how people may invest effort in editing if they see the way to salvage and make it on-topic, but expecting them to edit with the purpose to just make safe entertainment for HNQ passers-by seems awkward... softly speaking. A more reliable way in cases like that would be if system penalized / ignored close-worthy questions (those with many votes and flags to close) when selecting HNQ.

Another troublesome thing about how system works is questions can get to hot list too soon for community to do any meaningful cleanup. This is especially noticeable at smaller sites.

Expecting community (especially at smaller sites) to be able to immediately improve the questions assumes people will carry sort of 24/7 watch, things are simply not supposed to work that way, nor is it realistic to expect that it ever will work that way. Not to mention how terribly unfair it would be to expect stuff like that of volunteers focused on on-topic questions and answers. If we want to give users a reasonable chance to cleanup things, system would better have some sensible time lag prior to pushing questions to HNQ.

Yet another difficulty seems to be a matter of scale. System allows up to five questions from the same site in the hot list.

This can be too much for small site to handle if you consider that getting to the list tends to trigger "explosions" of answers and comments from newcomers unfamiliar with site quality norms - which in turn involves a fairly heavy moderation load (which can easily turn out more important from site perspective than polishing question titles). If we want site regulars to focus more on improving their hot questions we better do something to ensure reasonable amount of other moderation load carried by these questions. A simple way to achieve this would be to decrease the limit from five to 2-3 questions.

The last, but not the least - flags. Currently flags on hot questions are buried among those on regular posts.

It would take quite a lot of effort of flaggers and moderators to find and communicate about these - and, maybe most importantly, from site perspective there is no good reason for them to invest much effort into that. If we want moderators to handle issues in hot questions in more timely manner we better make system aid them in finding flags cast on these.

---

Summing up, system would better make it easier for site users and moderators to maintain troublesome hot questions. Currently it is rather difficult to them. Once again, it looks unrealistic to expect site community invest a lot of efforts in something barely related to their primary goal (of asking and answering on-topic questions).

_{(For the sake of completeness, some may be motivated to maintain hot questions for fear of excluding their site from hot list like it happened to IPS. But I honestly doubt that it could have solid long-term effect. Not to mention that fear may be bad for morale.)}