3
$\begingroup$

This question came up in the review queue, and I was a bit surprised since the question was asked only 17 hours ago, which might not have been enough time for the user to respond to the comment asking for more clarity, yet the question is already up for closure.

Close votes can be discouraging and unwelcoming to new users (this user only joined QCSE 9 days ago!), and if it happens too much unnecessarily (imagine for example that a user decides to close vote their maximum 24 votes/day, every day) then it will become irritating to the volunteers dealing with the review queue.

For quite some time now I've been surprised by a lot of the questions that end up in the close vote queue, and when I posted this: A thought about our close voting practices, not many people have offered to tell us why they are doing it.

Now I want to get more specific:

  • If a user asks a question that needs more details for clarity, shall we close vote immediately or should we give them X hours to respond to comments first, then close vote if they ignore requests for more clarity?
  • If you don't think we should give the user a chance to respond to comments first, why is it so important to give them the "close votes (1)" signal under their question, and to add more workload for review volunteers, so urgently that we can't wait maybe 3 days first?
  • If you think such close votes should occur with urgency (without allowing the user some time to first respond to comments or edit their question), what possible negative consequences might there be, for example do you think it's possible that when a user isn't too desperate for an answer, they might see the close votes and feel discouraged from participating altogether? I'll say that I felt this way when my first ever question on this site got closed at 22:49 on the same day as my question was asked (I have to show a screenshot since the question was later deleted by a moderator, so only I can see it):

enter image description here

enter image description here

At the time, I didn't know much about Stack Exchange at all, and did not know what "closed" meant, so I thought that it meant my entire day's worth of work that went into trying to ask a good question, was now going to be wasted. At the time, I had about ~7 years of experience with randomly asking the occasional question on StackOverflow, SuperUser, TeX SE, English SE, Physics SE, Theoretical Computer Science SE, Travel SE, Biology SE and maybe others, and it seems I had never once had a question "closed" before, and I hadn't been a regular enough user anywhere to have seen what "closed" means.

$\endgroup$
2
$\begingroup$

As far as I can tell, the system is designed so that ideally questions should be closed immediately. Clearly, that's rarely the case, as the voting system needs to do its thing, but I don't think there should be any further unnecessary delay. For example, mods and users that have a sufficiently high score in a tag can close questions unilaterally. Why would the system allow for this if we wanted to introduce delays?

The idea of closing questions is to tell the user: "this question must be edited to meet the site's standards". The ideal cycle of a "not-great" question should therefore be (1) question is asked (2) question is closed as unclear or whatever, optionally with comments saying what should be improved (3) after the question is edited (if that happens), assuming the edits fix the issues, the question is reopened.

Could this discourage people from asking questions? Of course, it often does, and it's never nice to see a question you ask be ill-received or closed. Nonetheless, the way I see it, this quality-control system is a significant part of what makes stackexchange vastly superior to other Q&A sites, so I stand by its effectiveness.

Regardless, even if we did want to give some time to a user to give them the opportunity to edit a question, that would never work. It would imply seeing a question, commenting, and then remembering to go back to check whether the question was improved after X time and if not voting to close. That's cumbersome to say the least.

$\endgroup$
12
  • $\begingroup$ This is more or less right, although sadly things are a bit more complicated than this. I hope people forgive me for getting into the politics of SE a bit here but it helps explain where these kinds of problems come from. When I became a mod, got to know other mods, looked around SE meta etc. it seemed obvious to me that closing questions quickly is part of how SE works. It's literally built into the design of the system. However, in recent years, this isn't how SE is sold to its newer users. Ultimately, this issue of telling different people different things causes conflict $\endgroup$
    – Mithrandir24601 Mod
    Feb 8 at 9:53
  • $\begingroup$ @Mithrandir24601 what do you mean exactly with SE not being "sold" this way? Are you referring to "be-nice policy" et al. or something else? $\endgroup$
    – glS Mod
    Feb 8 at 10:01
  • $\begingroup$ It's considerably better now (to be fair), but I'm referring to specific terminology used when a user first logs onto the site and sees (well, saw) things like 'get answers to your questions, fast' - it wasn't about helping, just 'you will get an answer'. But anyway, as far as I can tell, these sorts of issues are fixed now, it's just that there was a while when this did cause issues and confusion $\endgroup$
    – Mithrandir24601 Mod
    Feb 8 at 10:09
  • $\begingroup$ @Mithrandir24601 I see. Still, I don't think we can change our policies to follow the whims of the coders in charge of writing those welcoming messages etc, now can we? Tbh I see as a more significant issue in this regard the fact that the messages shown to low-rep users when their questions are closed now don't spell the exact close reason (at least I think that recently became the case... unless it changed again). $\endgroup$
    – glS Mod
    Feb 8 at 10:14
  • $\begingroup$ but I guess the underlying problem behind these discussions is the neverending struggle between wanting stackexchange to be more welcoming and nice to new users, and wanting it to maintain the high quality of posts. These two requirements, imo, are intrinsically incompatible to a large degree, and thus we are stuck constantly looking for the optimal compromise between niceness and quality-control. Not that this is necessarily a problem: as long as the needle doesn't violently pull one direction or the other things will work well enough I think. $\endgroup$
    – glS Mod
    Feb 8 at 10:19
  • $\begingroup$ "For example, mods and users that have a sufficiently high score in a tag can close questions unilaterally. Why would the system allow for this if we wanted to introduce delays?" Those two things have little to do with each other! If we want to give users a chance for about 3 days to respond to comments before closing the question, the fact that mods and high-rep users can unilaterally close the question is not at all incompatible with that! Mods can already unilaterally close questions, why is it that it almost NEVER happens? having gold-badge users on a tag is just like having more mods. $\endgroup$ Feb 8 at 19:26
  • $\begingroup$ "the way I see it, this quality-control system is a significant part of what makes SE vastly superior to other Q&A sites, so I stand by its effectiveness" - what evidence do you have that this is true? The reason I like using SE is that I get answers to my questions quickly (whether I've googled something and an SE answer came up first in the search result, or whether I posted the question myself). As a question asker my experience with the 1st question I asked on QCSE made it seem vastly inferior to other Q/A sites because the question got "closed" so it felt like a big waste of time. $\endgroup$ Feb 8 at 19:33
  • $\begingroup$ "That's cumbersome to say the least" ~ What's cumbersome is going through the close vote queue and seeing questions up for closure for needing "details or clarity" when OP hasn't even had the opportunity to respond to the questions asking for clarity. If it's true that the system is designed for questions to be closed immediately then why do we need 5 close votes to close a question? Why not one? Why does SOCVR exist? Why is the bar for hammer closing so extremely high? Why isn't the alert that close-reviews are pending, not right in people's face when they log in? $\endgroup$ Feb 8 at 19:42
  • $\begingroup$ "It would imply seeing a question, commenting, and then remembering to go back to check whether the question was improved after X time", well when you close-vote a question for which you've asked the OP to make some edits, you better remember to look back and check if they've made the edits so that you can retract your close-vote, because otherwise you still end up with questions like this: quantumcomputing.stackexchange.com/q/15803/2293 in the review queue, which have 3 close-votes that weren't yet retracted when OP made all requested edits. $\endgroup$ Feb 8 at 20:16
  • $\begingroup$ "that would never work" -- It would NEVER work? $\endgroup$ Feb 8 at 20:16
  • $\begingroup$ reopening doesn't happen often probably because many questions that get closed are asked by people that don't bother editing, or just find it easier to ask a new one; I didn't give evidence that the closing system is a crucial component of quality control on SE, I stated it as my opinion - providing such evidence would probably require a lot of data I don't have; of course, different people have different experiences with how the system works, that's normal I'd say; regarding the high closing threshold, I guess for democratic reasons? (cont) $\endgroup$
    – glS Mod
    Feb 8 at 22:20
  • $\begingroup$ I agree that 5 is high, but that's the number that was chosen, for whatever reason. This number has been debated before; I don't know what SOCVR means; I don't understand the comment about the alert not being in people's faces; regarding remembering to go back and check: no I don't need to, because I (usually) get notified if the user replies to a comment, and regardless, if the question gets closed and then edited it ends up in reopen queue again; retracting close votes almost never happens, the idea is to reopen when due not to prevent closing $\endgroup$
    – glS Mod
    Feb 8 at 22:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .