Sometimes when a question is closed, different people have a different reason for closing it. You can see this while voting: there is a count in the margin saying how many votes of a given type have been recorded.

However, once the question is closed, not all of this information is retained. Often, only one reason to close is stated in the 'Closed' dialogue below the question. Not only is not all the information passed to the OP (or certainly to the world); it's also harder to fix multiple problems at once.

Wouldn't it be better if a question needs 5 close votes of the same type (e.g. 'Too broad' or 'Off topic'), rather than any 5? What are some benefits that the current system has to the system I propose?

If not that, wouldn't it at least be beneficial to show all the reasons for closure, and not just the ones with two votes or more, cf. Todd Trimble's comment below?

  • 4
    $\begingroup$ No. The point is to close the question. If there are five people who think the question should be closed, does it matter why? Do you need five people to tell you that you should eat blueberries for the same reason? Can't it just be good for you and tasty? $\endgroup$ – Asaf Karagila Mod Jul 23 '18 at 9:39
  • 6
    $\begingroup$ Quite often the offered close reasons are only a loose fit, and different voters make different random choices as to which point to tick. Not to mention that some of the close reasons are almost equivalent in the first place (e.g., plain off-topic and off-topic with math.stackexchange.com suggestion). $\endgroup$ – Emil Jeřábek Jul 23 '18 at 11:47
  • 8
    $\begingroup$ That said, I don't see a great reason why all the close reasons couldn't be displayed. There is a bit of discussion on this here meta.stackexchange.com/questions/312658/… but I didn't really buy the reason in the answer. One thing to note: before a question has been closed, the current reasons for closing are displayed when you hit the vote-to-close button (which of course you don't have to follow through on). Also, if at least two people use the same reason, that reason will display after all the close votes are in. $\endgroup$ – Todd Trimble Mod Jul 23 '18 at 12:05
  • $\begingroup$ @AsafKaragila: "The point is to close the question" - do you mean to imply that the main goal of closing questions is to shut them down rather than to improve them? In cases where a question really isn't suited for the site, it should be easy to get 5 "Off Topic" votes. But I have also seen reasonable questions get closed (and sometimes reopened again, sometimes after some mild changes, or even just by disagreement), with no consensus on what exactly is the problem with the question. $\endgroup$ – R. van Dobben de Bruyn Jul 23 '18 at 12:25
  • 4
    $\begingroup$ Yes, closing questions is not "means for improvements", it is simply a way to say "No, this is not content valid for this website". Whether or not the author (or someone else) makes changes that make the content a better fit, or if other people dispute this decision, is a whole other story. The point is to close the question, because in its present form, it does not fit to the site (as best as the users voting could tell) for one of many possible reasons. $\endgroup$ – Asaf Karagila Mod Jul 23 '18 at 12:28
  • $\begingroup$ @AsafKaragila: The On Hold dialogue reads: "If this question can be reworded to fit the rules in the help center, please edit the question." I see this as an invitation to improve the question, and I think this is easier if there is more information about its problems. $\endgroup$ – R. van Dobben de Bruyn Jul 23 '18 at 12:36
  • 1
    $\begingroup$ The information communicated by close reasons is too unspecific to be of any use for improvement of the question. This requires someone to actually post a helpful comment. $\endgroup$ – Emil Jeřábek Jul 23 '18 at 12:46
  • 4
    $\begingroup$ Re what Emil said: in practice this is usually true. Usually the author of a question will not really be able to act on "not research level" in order to improve the question. Some of the others (like "unclear what you're asking" or "primarily opinion-based") do seem more like an invitation to improvement, although usually the author will need a hand-hold to discover what exactly is unclear, etc. But I think the OP has a laudable aim here: to try to give questions more of a chance. Most of the time, closures have a ring of the death knell to them, and it seems he wants to ameliorate that. $\endgroup$ – Todd Trimble Mod Jul 23 '18 at 13:42
  • 1
    $\begingroup$ Rather, wouldn't it be enough to have two close reasons "off-topic" and "no longer relevant" -- where closing a question as "off-topic" deletes existing answers and adds a message like "This question does not appear to be a good fit for this site in its present form. If it can be changed in order to make it suitable for this site, you might find relevant advice in the comments.", and closing a question as "no longer relevant" adds a message like "This question has run its course, and does not accept further answers."? $\endgroup$ – Stefan Kohl Mod Jul 27 '18 at 12:32
  • 1
    $\begingroup$ @StefanKohl: interesting; I didn't think to challenge the available options. There is certainly some ambiguity in them, and they could benefit from a revision. To add to your options, I think migration to math SE or academia SE could both be helpful tools as well (the former currently exists; the latter doesn't). $\endgroup$ – R. van Dobben de Bruyn Jul 27 '18 at 13:18
  • $\begingroup$ @R.vanDobbendeBruyn: As to migration, I think it is easy enough for the OP to paste their question into a question box at the other site themselves -- that way they get a fresh question there which gets better visibility in general. Thus on the side of MO, it would suffice to close the question, and to suggest (a) site(s) where to post it instead -- and leave the choice to the OP. $\endgroup$ – Stefan Kohl Mod Jul 27 '18 at 16:07
  • 1
    $\begingroup$ You may want to upvote this feature request on Meta Stack Exchange: Distinguish close votes by reason $\endgroup$ – Federico Poloni Aug 5 '18 at 11:46

Sometimes questions do not fit the forum. When I comment on these questions, I normally suggest they picked the wrong forum, and sometimes suggest a different forum. If there is an obvious fix that would make the question better, I suggest that. On a recent number theory example I suggested the poster do more research and not expect the community to do it for them. I have voted to close perhaps twice, and usually on questions which I thought would cause harm, not on questions that would be just low quality that others would clean up.

If everyone had the willpower to leave a comment on why a question should be closed, MathOverflow might be seen as a bastion of principled critical negativity, rather than just an unwelcoming place. Similarly if no reason were given and the question were just closed. The present system is a compromise between those two extremes, and accommodates the fact that most people do not like writing comments.

I dimly recall a set of comment templates compiled by others (Martin Sleziak?). If people just use one or two of them when closing a question, that would convey an attitude of encouraging the poster to try again, which is not always the desirable outcome. I think asking for five identical close reasons would sharply reduce the quality of nonclosed questions on the forum.

Gerhard "When It's Not Majorly Broken..." Paseman, 2018.07.24.

  • $\begingroup$ I am not sure, do you refer to the templates in the answers to this question: Comment template for questions more appropriate for MSE? I should point out that several sites have lists of comments templates useful in various situations, for example, Mathematics: List of comment templates. $\endgroup$ – Martin Sleziak Jul 24 '18 at 15:33
  • $\begingroup$ Sorry, but my memory is fuzzy and I don't recall. Both of the links seem appropriate though, and if you want to add them to the answer, that's OK with me. Gerhard "Likes Seeing Others Add Quality" Paseman, 2018.07.24. $\endgroup$ – Gerhard Paseman Jul 24 '18 at 15:43

You must log in to answer this question.

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