I wish to make a modest proposal on closing/reopening questions. Let me begin by noting that by far the most frequent reviews are Close votes -- almost 37,000, and looking quickly most of the close votes (apart from skipping the question) are to Close a question (about 80% of the time) and only rarely to leave it open. In contrast the review queue to reopen questions only has about 6000 reviews, and again most of these (maybe 80% again) are to leave the question closed, and only a small proportion of the votes are to reopen. My own experience with these queues has been very similar to the overall numbers.

To me this suggests that overall people have been doing a good job of removing unsuitable questions, and only relatively rarely are there borderline questions that are closed. Even when these borderline questions are reexamined by other users, only rarely do they get reopened. The fact that questions get reopened suggests also that some users really like or wish to see the question on this site.

Would it make sense to have a feature whereby a question that has been closed and then reopened is no longer considered for closing again? That is, I am proposing a no "double jeopardy" rule. Added Stefan Kohl in comments has proposed the alternative of requiring more votes in each new close/reopen cycle.

The pros: It seems friendlier to do this, and respect the opinions of others who wish to have a question open. (In my view, it is not symmetrical to want a question closed versus wanting it open.) It will save us the occasional close/open war, which ends up getting heated. Similar arguments came up in What shall we do with stone soup? but nothing concrete seems to have been resolved there.

The cons: Every once in a while one might get irritated by a question so bad that it's hard to understand why five others voted to reopen! I find that there is a question of this sort that I voted to close the second time around; I do feel that I would now prefer to be personally irritated, but let the question stand on general principle. Another con would be that the proposed rule is irreversible (and I am generally wary of irreversible steps) -- this could be mitigated by either having moderators deal with extremely rare situations, or resetting the close/reopen possibilities if a question is edited. Added Stefan Kohl's suggestion to increase the number of close/reopen votes in each cycle, I think, gives a good way to guard against potential abuses.

In any case, I would be curious to hear what people think of this. For my part, I do not intend anymore to vote to close a question if others have reopened it.

  • $\begingroup$ In brief you want to let five 3k+ users dictate what is admissible. $\endgroup$ – user9072 Mar 23 '16 at 1:10
  • 6
    $\begingroup$ I think the idea behind your feature request makes sense, but implementing it in the suggested form would come along with the risk that a small group of people extends the scope of the site in a way with which the vast majority of the community does not agree. What one could do though is to increase the number of required votes in each close/reopen cycle. $\endgroup$ – Stefan Kohl Mar 23 '16 at 10:09
  • $\begingroup$ @StefanKohl: Excellent idea! $\endgroup$ – Lucia Mar 23 '16 at 12:23
  • 2
    $\begingroup$ I don't think I'm particularly in favor of such devices. I think what frustrates me more are cases where people state their concerns about a question, a meta is set up to discuss them, but there is no response or acknowledgment from the OP about such concerns, and so people wind up talking as if to themselves. (It's equally bad when closers don't express what is wrong, but that's not what I'm talking about here.) I don't think "Caudine Forks" is a very helpful way to frame it -- closers are not usually actively setting out to humiliate, but often want to express legitimate concerns. $\endgroup$ – Todd Trimble Mar 23 '16 at 13:14
  • 3
    $\begingroup$ In short, I would call for more communication between the parties concerned. The recent meta post by Jon Bannon (on Intuition Pumps, here: meta.mathoverflow.net/questions/2802/…) is a good example of a right approach to take in these matters. $\endgroup$ – Todd Trimble Mar 23 '16 at 13:16
  • $\begingroup$ What do you actually propose now? It makes little sense to have a feature request and to propose two completely different things. (It is fine for a discussion.) Please focus on one proposal or adapt the tags. $\endgroup$ – user9072 Mar 23 '16 at 15:02
  • 1
    $\begingroup$ @quid: Both proposals are based around the same idea of being more inclusive, and giving greater consideration to people who want to see some questions open. I'm interested to see what people think about it. $\endgroup$ – Lucia Mar 23 '16 at 15:05
  • $\begingroup$ Fine, so it seems you want to discuss about the subject and not propose one specific new feature. So I retagged the question accordingly. $\endgroup$ – user9072 Mar 23 '16 at 15:07
  • 1
    $\begingroup$ Actually, a milder version of this was already proposed years ago by Gil Kalai; it should be somewhere on the second page of the linked discussion. For Stefan's proposal: It is somewhat reminiscent of the idea of cancelling votes to close proposed by Kaveh, especially if one would take the "biased" version by Daniel Moskovich it seems the number would even match. $\endgroup$ – user9072 Mar 24 '16 at 2:13
  • 1
    $\begingroup$ @quid: Thanks! Those are useful pointers, and all those ideas you link to seem good to me. $\endgroup$ – Lucia Mar 24 '16 at 2:19

I agree with part of the intent of this proposal. Namely, I agree that it makes sense to take into account approval of a question already expressed by others when deciding to vote to close it. This goes for up-votes, reopen-votes, and reasoned arguments (in ascending order of relevance, as far as I am concerned).

[I also agree that the situation is not symmetric, but I do not see it as a one-way-road either, and in fact the precise details how to weigh things should depend on they type of question and concerns voiced. But this is tangential here.]

However, I feel the specific (original) proposal is not a good idea. (I am more or less indifferent towards Stefan Kohl's proposal.)

There are various things one could say about details, but I do not, at least not initially, since I do not want to distract from my main concern regarding this proposal.

Making this change would change the dynamic of the situation drastically; I think it would change the dynamic to the worse, for the following reason.

  • The current process (with all its faults and negative side-effects) is geared towards establishing some kind of consensus or at least modus vivendi. There is an incentive to find some common ground, in order to get out of a close/reopen cycle that is a nuisance for everybody.

  • The proposed process makes it so that basically all that counts (when faced with a closure) is finding enough support, the five votes to reopen. Then one wins, and all concerns voiced can simply be ignored; locally it is also not relevant if in the process of finding this support one alienated even more users. In brief, this process seems geared towards confrontation.

I do not believe this new process would make the site friendlier or avoid heat, in fact quite the contrary.

  • $\begingroup$ Thank you for the answer. I understand your view, although I don't agree with it. In the last month about 400 questions were closed, of which 9 have been reopened. Taking a more tolerant stance towards the reopened questions I don't think will make the site any worse. I also don't quite know why you say that the proposal will make the site less friendly -- you might well be right, but I just don't see your reasoning. $\endgroup$ – Lucia Mar 25 '16 at 0:25
  • $\begingroup$ The statistics over the last three months is also similar -- about 1150 questions closed, of which 20 were reopened. The proportion of questions reopened is even lower than what I would have guessed. $\endgroup$ – Lucia Mar 25 '16 at 0:53
  • $\begingroup$ Indeed the number of questions that get reopened rarely exceed 10 per month: 8,10,6,7,11,5,3,3,8 is the respective count for the last couple month. What seems more relevant for your proposal though, is the number of question that got closed more than once. This is over the history of MO significantly below 100 to be precise 63 is what the query gives, but it may missed some deleted ones, not sure (about 20 of those ended up open in the end). So, we are talking about a one in thousand event. The problem you want addressed is so rare that it is best handled on a case by case basis. $\endgroup$ – user9072 Mar 25 '16 at 1:39
  • $\begingroup$ Indeed it is extremely rare, and yet these are the questions leading to endless discussions. My take is -- it is so rare that accepting other people's interest in the question would not change the site at all. $\endgroup$ – Lucia Mar 25 '16 at 1:42
  • $\begingroup$ Yes, I meant to come to this in a second comment. There are two orthogonal things. First the more abstract point: the data is the one that came about given a certain context. This context hardly ever creates the problem you want solved. Still you want to change it. Yet, in the other direction, you propose to change the context completely and somehow assume the outcome will stay about the same except for treating better those one in thousand event. This assumption seems based on nothing but wishful thinking. $\endgroup$ – user9072 Mar 25 '16 at 1:50
  • $\begingroup$ Second, right some question lead to discussion. But imo the plausible explanation for this is that the question is in a certain way controversial, and not the technical fact users can vote to close again after reopen. I just don't believe that all user seeing a reopened q they find problematic yet not being able to vtc anymore will all just think "oh well, nothing to be done, moving on" but at least some will complain in cmnt or rant on meta or do still something else (like giving a "fun" answer to mock the question, post parodies of the question etc [to be sure, all this already happened]) $\endgroup$ – user9072 Mar 25 '16 at 1:57
  • $\begingroup$ None of which seems more friendly than a simple vote-to-close. There was a recent example, the "Go" question. You found the meta thread useless, I know, but I feel confirmed in that it was not,. It gave a venue to discuss, actually not that much discussion happened, there was a bit of voting and the thing fizzled out without any significant problem. What mainly frustrates many users is if they feel they cannot do anything (but rant, and then they do this). This is very visible in the type of complaints one sees by new and low-point users under closed questions. $\endgroup$ – user9072 Mar 25 '16 at 2:06
  • $\begingroup$ Finally, let me specifically come back to "it is so rare that accepting other people's interest in the question would not change the site at all" while I already addressed it indirectly. I am very convinced that certain types of questions, of the form that are prone to lead to such problems, would be a lot more frequent if community-moderation would not have pushed back on them rather consistently over the years. $\endgroup$ – user9072 Mar 25 '16 at 2:12
  • $\begingroup$ Quid -- I've voted to close more than my fair share (usually when I feel sure of my view), and I certainly don't have anything against the community moderation. However, I really don't see how changing the community attitude to accept that other people might find a question valuable even if one doesn't care for it oneself, will really lead to the catastrophes that you predict. $\endgroup$ – Lucia Mar 25 '16 at 2:17
  • $\begingroup$ Voting to close or reopen should have nothing to do with "care for [the question] oneself", these votes are not to be used to express personal preferences (instead to express those one uses up and down votes). The decision should depend on whether one considers the question as suitable for the site or not (given the current form of the q, and the current scope of the site); everything else is a misuse. In view of this it also not necessarily true that for a qu that gets reopened some "really like or wish to see the question on this site;" it is not true for most of my reopen votes. $\endgroup$ – user9072 Mar 25 '16 at 2:32
  • $\begingroup$ I also made an effort not to predict any catastrophes, in any case not short term. However, the change you propose over time could have a significant negative effect. It is also somewhat ironic that you seem to have an aversion against discussions, yet a main type of question that would benefit from the change you propose are questions closed because they are likely to create too much discussion. There are plenty of controversial subjects somebody might want to discuss on MO, and i am pretty sure that for some of those there are a couple 3k+ users that want to do so too. $\endgroup$ – user9072 Mar 25 '16 at 2:39
  • 3
    $\begingroup$ I also just do not think that five 3k+ users are always right in their judgment. There is a somewhat recent case of a "historical" question, reopened, of which it was clear that the entire context was pure fiction and the thing just posted for the fun in it. I really do not see why we should host such things even if a handful of 3k+ users think so. $\endgroup$ – user9072 Mar 25 '16 at 2:44
  • 1
    $\begingroup$ Ok, I give up! You're right that I don't care for these discussions. Nor do I particularly like soft questions or "discussiony" ones, which as you say are the ones likely to benefit from a tolerant attitude. However, it is also very easy for me to ignore questions that I don't like, and I don't really need reasons for why somebody else might like them. $\endgroup$ – Lucia Mar 25 '16 at 2:57
  • 1
    $\begingroup$ Although I don't agree with some of your views, I accept this answer. Some points that you make are surely valid, and I think perhaps are also reflected in the overall lack of enthusiasm for my proposal. Thanks also for pointing out the previous discussions; I found still earlier versions on tea especially regarding the exchange of votes to keep open and close, which indeed were hard to implement well. $\endgroup$ – Lucia Mar 30 '16 at 0:18

I think you are proposing a technical solution (or change to a technical solution) to a problem that is more cultural (open/close tug-of-war). It seems to me that a cultural change should be made instead.

I don't vote to close, but I do suggest why things should be closed or migrated or, more importantly, how they should be changed to fit (my idea of) what is appropriate/acceptable for a MathOverflow post. If some of us maintained the habit of thinking "How could I make this a better question for MathOverflow" and giving a brief (even stock) comment toward that, it might lead the way toward making better questions, rather than fighting over how/whether the original poster should make any change. In particular, the following experiment could be tried: for a given closed question which holds promise, those who vote to reopen edit a section above the original question, giving the essence of what the proper version of the question should be, and leave the original for comparison.

Gerhard "Yes, It's Not As Easy" Paseman, 2016.03.23.

  • $\begingroup$ Thank you for the answer. In some ways you are right that a cultural shift would be preferable. However, the discussion (see especially Gil Kalai's answer) in the linked stone soup question suggests such a cultural shift, but I don't think this has happened. People forget, and new people enter the system. It seems to me that a technical solution may in fact pave the way for a lasting change of attitude. $\endgroup$ – Lucia Mar 23 '16 at 21:33
  • 1
    $\begingroup$ @Lucia the top-voted answer there asks for edits to such questions by those that value them. Requests for edits to the questions were explicitly made (by a moderator) on the last two question that started a considerable meta activity ("LIGO" and "Go"). No edits happened. It seems a feature needed to get the culture in line with what was discussed there could be a question cannot be reopened unless it is edited. [Added: I do not actually mean to propose this feature it is intended as an illustration for what is in my opinion part of the problem.] $\endgroup$ – user9072 Mar 23 '16 at 21:51
  • $\begingroup$ And I see the opposite: the more lasting (and in my view, worthwhile) effects are cultural and not technological. New people come in and do not understand, so the community members help them with training. There are a lot of technical innovations that came with 2.0 that I ignore or am slowly coming to appreciate like this symbol thingy, @Lucia. While I see your list of pros and cons, I am not seeing much benefit or cure in your proposed change as I do in mine. Gerhard "You Want Change? Be It." Paseman, 2016.03.23. $\endgroup$ – Gerhard Paseman Mar 23 '16 at 23:51

You must log in to answer this question.

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