# Under what circumstances is it appropriate to delete a question that has received a good answer?

I would like to have a discussion about the following question:

When is it appropriate to delete a question which has been closed as "missing context or details," but which has generated mathematically good and upvoted answers? What considerations should be made when voting?

The recent spate of deletions via downvotes and closure votes has been triggered mostly through the actions of a very small group of users, and one user in particular (who claims a role in deleting 3500 posts in about three months). There has been a significant increase in deletions recently.

This has come up in numerous threads throughout meta recently, although the boxed question has not been discussed specifically in its own question. The level of deletion has caused great frustration among some users who view this as causing the removal of a significant volume of answers and teaching effort - this is severe to the point that some users are considering leaving the site. Other users believe it is necessary in order to protect the quality standards on MSE, and (possibly decreasing) quality standards have cause a lot of frustration as well.

$\scriptsize{\text{(I will reserve my own opinion for the comments.)}}$

• Deletion always seems a little harsh to me, and especially so in the particular circumstance you describe. It removes someone's significant input. I think it is wrong unless there are clear extenuating circumstances. (Contests are a good example, I answered a question earlier, followed the OPs link to their other questions, then realised that I had probably answered a contest question. Of course, enforcing this uniformly must be problematic.) – copper.hat Aug 2 '14 at 3:48
• Is the question about delete votes or down and close votes? – Scott H. Aug 2 '14 at 5:22
• Good to have this discussion (+1). I certainly go by instinct in trying to draw the line between "vote to close"-only and "vote to delete". It is very likely that this thread will bring some new angles to the surface allowing me to reconsider/fine-tune my approach. – Jyrki Lahtonen Aug 2 '14 at 7:46
• I think that the primary goal on the site is to answer the person who asked. The "permanent record" aspect is secondary, but surely if we do want to keep a "permanent record" of some questions, it should be the best ones - not hastily composed PSQs about routine exercises from unengaged askers. – Carl Mummert Aug 2 '14 at 19:11
• ... because deleting an answered question once in a blue moon won't do it. I think closed questions should be kept sparingly. Maybe the best 10 percent can be kept. When you vote to close a question, you are saying it should not have an answer; if so, why would you keep the answer? – user147263 Aug 4 '14 at 6:28
• @900, the reason to keep the answer is blindingly obvious; the answer may contain helpful mathematical content, even if the question was (deservedly or otherwise) closed. I think I asked you elsewhere whether you checked to see whether answers had helpful mathematical content before you downvote them; I don't think you answered that question. – Gerry Myerson Aug 4 '14 at 6:50
• @GerryMyerson I thought I did, but perhaps not explicitly. I will spell it out: how I arrive at my voting decisions is none of your business. – user147263 Aug 4 '14 at 7:03
• I'll take that as an admission that you have no qualms about designating terrific answers for deletion. – Gerry Myerson Aug 4 '14 at 7:04
• @GerryMyerson Among $\le 1$ scoring answers to closed questions, the percentage of terrific ones is rather low. I see them once in a long while; then I'll usually try to edit question into shape and get it reopened. I do mean terrific, which does not include nice-but-easily-reproduced explanations. From the meta rhetoric one would think I'm burning unread Ramanujan's manuscripts... it's more like a few copies of a mass-produced solution manual. – user147263 Aug 4 '14 at 18:08
• @900 Hallelujah! It's refreshing to hear that you found some answers terrific by your standard. But if they were so terrific then why didn't you upvote them? (currently you have zero upvotes and, alas, $4423$ downvotes). What will it take to earn your upvote, an elementary proof of FLT? If your answer standard is so high that rarely does an answer meet it, then it seems almost all answers are subject to your mass deletion campaign. – Bill Dubuque Aug 4 '14 at 20:41
• @BillDubuque I think the site needs downvotes more than it needs upvotes. Maximizing utility function $Au+Bd$ with $A<B$, subject to $u+d\le M$ leads to $u=0$, $d=M$. ... As for the last sentence, see: "Among $\le 1$ scoring answers to closed questions..." just above. – user147263 Aug 4 '14 at 20:44
• You asked for my opinion, @900. I think your actions here are harmful and disruptive; the exploits to the system that you've found, and the scale on which you've willing to use them to shape the site after your desired quality standards are ridiculous. Cheers. – user61527 Aug 4 '14 at 21:23
• @900 Of course I know the formula. But some of the best answers on the site have few upvotes (e.g. if they were not quick, or if they are in very specialized topics, etc). Pick your favorite teachers and browse the tail end of their answers and surely you will find many gems with few votes. It greatly worries me that these excellent answers will meet the mechanical standards applied by the answer grim reaper. – Bill Dubuque Aug 4 '14 at 22:09
• @T.Bongers But that's just about the little ol' me. I thought you proposed a discussion about the site, not about a user. – user147263 Aug 4 '14 at 22:25
• @T.Bongers I would suggest that while it might be appropriate to delete questions with no answers, common courtesy would indicate that anyone considering deleting a post (other than the OP) that has been answered at least inform the answerer. Alternatively, perhaps MSE could implement a feature where the post could only be deleted with the consent of the people whose "work" would be deleted. – user137481 Aug 14 '14 at 1:22

One aspect of not deleting a bad question is that it sends a message to the reader: "MSE accepts these sorts of questions". Also, "This is the kind of way you should ask questions."

Often times, the answer to a question assigning an exercise will contain a solution, in which case we are telling the internet "Come to MSE and post your exercises and you will receive solutions."

Any merits that a particular answer to a question may have must be balanced against this point.

Deletion is the easiest fix for this demerit. If one believes a particular answer should be salvaged, one could have your cake and eat it too by one of the following means.

• Edit the question to make it a good one, or at least a passable one.
• Create a good question, and suggesting the answerer move his answer to the good question. And if he doesn't, copy it yourself (and give attribution to the original answerer)
• Same as the last answer, but by finding a good abstract duplicate rather than writing a new question.
• The biggest problem I see with the "edit to save question" philosophy is, how can I edit a question to show the OP's work? Given that no-effort questions can lead to some very informative answers, how can I edit the "no effort" question to save it? – apnorton Aug 2 '14 at 14:27
• @anorton You can't add OPs work. But you can add context from an impersonal point of view... Context: This integral is similar to ... which can be solved with ... But here we have ... so that approach does not work. The closing reason refers to absence of any thought, it does not mandate posting a half-done solution. – user147263 Aug 2 '14 at 16:06
• @anorton: I generally view the work as a means, rather than an end: the point of showing the work is so a knowledgeable reader can extract the precise good question that the OP needs to ask, but ultimately for the purposes of being a searchable mathematical reference, it's only the good question that matters.... – Hurkyl Aug 4 '14 at 19:45
• ... but now I think I finally understand that people are viewing MSE as a forum to engage in tutoring activities, in which case having the OP show work is an end, as it makes them part of a dialog.... – Hurkyl Aug 4 '14 at 19:46
• ... but that said, I don't think MSE is a good fit for having interactive dialog, (and I believe such is contrary to the general stackexchange 'mission' anyways), and as many have noticed, the gamification actively encourages people to preempt dialog. My leanings have been increasingly towards the position that such activities are really off-topic for MSE.... – Hurkyl Aug 4 '14 at 19:47
• ... and in particular, the poor fit is by far the most pronounced when it comes to the 'hopelessly confused', who are by far the most in need of having interactive dialog, and least helped by a "brief question" - "brief paragraph on the topic" interaction. (or worse, a "brief question" "brief solution" exchange) – Hurkyl Aug 4 '14 at 19:51
• I suppose it's a bit ironic that I hadn't understood that, because when I first came to MSE, I did look at it as a place to engage in more tutoring activities. (although I have since come to think as I describe above) – Hurkyl Aug 4 '14 at 20:03
• @anorton After some time (for example, >1 year) has passed, the "showing work" as a defense against homework has no potency. The actual OP is purely incidental, a number associated to a post and nothing more. – user1729 Aug 6 '14 at 8:20
• @user1729 That is correct. But, we still have a "close as lacking context." For some reason, I made that equivalent to "close as not showing work," which 900 reminded me is not the same thing. – apnorton Aug 6 '14 at 13:54
• @anorton A recent example: I edited this question which was closed for "no context", and it got reopened. Used OP's comment for "source", and added "observations" of my own (not very deep ones, but something that may help appreciate the question, and set better example for others). – user147263 Aug 6 '14 at 16:11

I hesitate to give an answer as I am not (yet) involved in this aspect of the moderation (on this site). Yet, I try to present an argument for a certain way of seeing things, and I would be curious to know what the specific points of disagreement should be in case there are some. The thing below applies mainly to closures of recent questions.

There are two main points for putting on hold/closing (in the current context at least):

1. To signal that the question is (in its current form) not appropriate for the site.

2. To prevent answers being given to that inappropriate question.

Thus, if the question that was put on hold is not improved the decision that the question is not appropriate for the site stands and the natural next step is deletion.

Whether or not the question got answer(s) is not that important in general (but see below). Typically, there was no doubt that somebody could or would give an answer. Indeed, almost the entire point of putting on hold was to prevent that expectation becoming a reality.

Now, that the process failed since somebody gave an answer typically does not change the initial judgment that the question should not be on the site and/or should not be answered. There seems no reason not to delete it. Indeed, answering sabotaged community moderation (often this will be accidental but sometimes it seems even deliberate). Why should this be honored by preserving the answer. In fact, not to delete it now makes the putting on hold somewhat moot.

There can be situations were an answer is really a game-changer for a question, since it finds a surprising aspect in the question or is truly great (just good does not suffice, it should really be surprisingly excellent, a good answer being forthcoming was expected when voting to close already). In these cases one could make an exception.

To sum it up: closed questions can and should be deleted even if the have good but standard answers; only truly great answers should rather not be deleted.

• By the way, you should totally get those missing 1365 points. On Math.SE you get 50 close votes a day, versus measly 25 on MathOverflow. – user147263 Aug 4 '14 at 17:57
• Okay, okay. Done 260 today. But 900 is too much for me. – quid Aug 6 '14 at 0:03
• This. So much this. – Jyrki Lahtonen Aug 6 '14 at 9:20

Fragmentation of discussion hurts the site

Some (many? most?) of the questions that should get deleted are simply exercises, without any specific requests for exposition on a particular point.

As such, there is much duplication as the same abstract exercise (and sometimes even the same concrete exercise) gets posted over and over again. Without cleanup, this creates the unfortunate situation that the answers to a particular question are spread across multiple postings, and sometimes there are even duplicate answers!

A good answer that is difficult to find isn't much better than no answer at all.

Furthermore, the visibility of the answers gets reduced, and vote counts watered down, so we also diminish the functionality of the site to rate answers.

If there is an answer worth salvaging, then it should be salvaged, rather than preserving the question to let it rot in place. Some approaches to salvaging are suggested in my other answer.

• +1 because I like the sentiment, but I want to clarify that I believe the solution to this to be very difficult to implement. (It's hard to consolidate questions because of the difficulty of searching mathematics.) – apnorton Aug 2 '14 at 16:08

Testing the waters with:

### I propose that if the answer(s) can be reproduced with a call to Wolfram Alpha, then the Q&A can be deleted without any qualms.

So for example standard tricks to reproduce an integral or a solution to a basic congruence or a (system of) equation(s) are at risk.

This would leave plenty of room for teaching methods/concepts.

The same test can be applied to a question. If it can be answered by WA, it could be declared off-topic, and all such answers are not helpful. Thus the question must be about understanding a step as opposed to how to turn the crank.

This would get a rid of lot of crap. Not all of it, granted. But it would steer the site towards explanations, concepts, ideas, problem solving et cetera.

This probably needs a lot of fine-tuning to be palatable to a significant fraction of the user base.

• That has high risk of throwing out the baby with the bath water, since many teachers use such elementary problems to highlight deeper ideas in simple contexts. For example, I use a nonlinear generalization of Heaviside's rule as one way to teach modular computation. Of course WA can compute such partial fractions by brute-force, but without the conceptual structure I impose on the problem - which serves carefully-crafted pedagogical purposes (which could easily be overlooked by trigger happy reviewers). – Bill Dubuque Aug 3 '14 at 18:12
• @BillDubuque There is no risk to that answer, simply because the question is not closed (nor is it likely to be closed in the future). Users cannot vote to delete a non-closed question. If you have similarly valuable answers that are at risk, see my answer here. – user147263 Aug 3 '14 at 21:32
• @900 My point was to give one example of certain types of answers which may be at risk if the above proposal were implemented (not to imply that said answer was at risk as placed in its current thread). – Bill Dubuque Aug 3 '14 at 21:39
• Ugh. Again with the "teachers". Many of my former teachers, which are excellent teachers beyond anyone's dispute, would have told me to stop wasting my time online. Even more so since I seem to spend so much of it on meta arguments. – Asaf Karagila Aug 3 '14 at 22:54
• @Asaf I prefer to write "teacher" vs. "answerer". – Bill Dubuque Aug 3 '14 at 23:53
• I would like to go one step further and say "If a question can be typed into Wolfram Alpha and solved, then it should be off topic to begin with and closed as soon as possible." Yes, many of my early answers would be in violation of this rule of thumb, but I think it would be better for the site in the long run. – apnorton Aug 6 '14 at 13:57
• I disagree simply because Wolfram Alpha is a useless pedagogical tool. See i.stack.imgur.com/bxUZ2.png for example. – user21820 Apr 3 '17 at 16:28
• @user21820 That much is clear. My feelings about this vary very much. Lately I have been somewhat cynical, so I advertised this. After all, we have very many answers that do not offer anything pedagogically either. The key question IMHO is whether getting rid of a bunch of useless answers outweighs the loss of a few teaching opportunities. Another thing is that some relatively high rep users repeat the same tune day in day out - without adding anything new (doesn't apply to Bill who will often turn a few more stones looking for a new angle). – Jyrki Lahtonen Apr 3 '17 at 17:35
• Yes I totally agree that many answers do not offer any pedagogical value, and sometimes even are misleading or downright wrong or logically invalid and yet are highly upvoted. That last part irks me the most. I believe however that deleting bad questions (according to the usual criteria) will get rid of much of these useless answers anyway. =) – user21820 Apr 3 '17 at 17:51

What considerations should be made when voting?

The premise of the question is flawed in that it proposes there could be some consensus about how people vote. Voting is not by consensus, it is everyone's expression of their own opinion.

When your opinion becomes a crusade affecting hundreds of questions and answers, then it's a bit more than an opinion.

No, it is still just an opinion of one user, who happens to think that those Q&As should be deleted. Perhaps I hold this opinion a bit stronger than others hold theirs, but I am not responsible for other people's weaknesses. I see such Q&As as broken windows, through which new users enter (via Google search, typically), and conclude that the site is of low standards. I don't want their first sight of the house to be an overturned trash container.

If others' opinions lean toward inclusionism over deletionism, they can express that by their votes, too: upvote, reopen, undelete. The fittest will survive.

I don't see mathematical correctness as a sufficient condition for preservation. Wolfram Alpha generates mathematically correct answers (with more details in the Pro version), but it would not be useful to have its input/output pairs to be logged on this site as Q&A pairs. Which isn't much different from what is going on.

Asker: How solve $\int \frac{\sin(2/x)}{x^2}\,dx$? HALP! (Question titled "Calculus problem help")

Answerer: Hint: Try the substitution $u=\color{blue}{\dfrac{2}{x}}$, so that $du=\color{blue}{-\dfrac{ 2}{x^2}dx}$ (Makes no edits to the question)

Others: Gosh, that's a mathematically good answer, and now the whole thing is sacrosanct content destined to stay forever on the site!

### View from the sidelines

It's sometimes interesting to find how the site is seen by someone who does not participate in it, but pays attention to what goes on around the network.

I recommend this post by MichaelT to your attention.

• Mathematical correctness is clearly not a sufficient condition, as your example shows. But it seems a pity to delete significant contributions because the question is deemed bad. – copper.hat Aug 2 '14 at 18:17
• Aah! I hate when people call those sorts of things "hints". It's not a hint if you do the only challenging part of a problem for the person. I even think calling it a hint is actively harmful, as it can leave the person with the impression that if they can then do all of the mechanical work themselves, then they mostly understood how to do the problem. Sorry for the rant. – Hurkyl Aug 2 '14 at 21:23
• @Hurkyl: When I write homework questions, I sometimes includes a small hint, which is something like "Look at ..." and sometimes a larger hint which includes several (if not all) the checkpoints one would need to pass for a correct solution, like "Use this, then do that, then use the previous question". For the former I use the Hebrew word "hint" and for the latter "guidance". Here I might use "hint" for both, because I'm not sure what is the correct term for such an outline. That been said, I do feel that such outline for the solution can be a valid and helpful [...] – Asaf Karagila Aug 2 '14 at 21:28
• [...] answer, since it leaves the task of writing properly to the students. Of course, it is always better to know that what the OP knows and did (or was supposed to do) in those cases. But I wouldn't consider them necessarily actively harmful. – Asaf Karagila Aug 2 '14 at 21:29
• "No, it is still just an opinion of one user, who happens to think that those Q&As should be deleted.... If others' opinions lean toward inclusionism over deletionism, they can express that by their votes, too: upvote, reopen, undelete." This strikes me as highly disingenuous, as you know bloody well that you have found a way to amplify the power of the opinion of one user to an extent and purpose that appalls a great many other users. And it is hard for those other users to upvote, reopen, and undelete when there is no way to know what answers you have targeted. – Gerry Myerson Aug 2 '14 at 23:45
• Moreover, it is not clear to me that you even read the answers you target, to make sure they are of the type you parody above, before you vote them down. So, here's my suggestion: before you vote an answer down, post to meta to announce your intentions. Let 24 hours go by. If no one has come to the rescue of the targeted answer, go ahead, carry out your plan. – Gerry Myerson Aug 2 '14 at 23:48
• @Ayman: There is a problem with that page, though. The list of "recently deleted" is not detailed enough. It only keeps a maximal number of entries, regardless to the timespan you are asking for. This means that if there has been, say, 60 deletions at the same time, then nothing before that (and possibly some of those deletions) are not going to show on the list. It's nearly impossible to monitor when the only tool for monitoring doesn't give full information. I have already raised a feature request for that, which 900 tried to raise awareness to on meta.SE to no avail. – Asaf Karagila Aug 3 '14 at 0:35
• @T.Bongers That was culture shock, amplified by the provocative tone of my answer. I do hope for a few more DVs so that I can have the most downvoted answer ever... I haven't changed my position, and haven't been silent about it either; the votes on my subsequent posts were largely positive. – user147263 Aug 3 '14 at 0:45
• @GerryMyerson "before you vote an answer down, post to meta to announce your intentions. Let 24 hours go by." That proposal seems infeasible. It seems quite likely some will vote-up just to make a point (and some other to counter this) causing huge friction. And, there is a thread for undeletion requests. If somebody cares about some lost post they can post there, if noone cares or even notices than not that much seems lost. – quid Aug 3 '14 at 10:04
• @quid, there is already huge friction on this issue. And Asaf points out a few comments up it's not that easy to know what's being deleted. – Gerry Myerson Aug 3 '14 at 12:25
• @GerryMyerson To be clear the suggestion that one should come to meta and make a thread if one even contemplates downvoting a question is serious? – Scott H. Aug 3 '14 at 16:22
• @Scott, the suggestion is that if you contemplate voting down an answer with the goal of having the Community user delete the question (along with all answers posted to that question) then you first post to meta to see whether the community agrees that answer should go. One need not make a new thread for each contemplated downvote; it could be a new post on a downclose-devoted thread, like the thread we already have for reopenings and undeletions. – Gerry Myerson Aug 4 '14 at 1:49
• @quid, when an old answer by another user is downvoted in order to cause the Community user to delete the question and all answers to it, this is not something I would notice "while following my usual activity". – Gerry Myerson Aug 4 '14 at 1:54
• @quid, the questioner may be long gone. The answerer may notice that she has acquired a downvote, but what can she do? She can't upvote her own answer. All she can do, so far as I can see, is keep an eye out to see whether the question gets deleted and, if it does, vote to undelete it. Not, I think, a satisfcatory state of affairs, when a determined user is on a crusade to delete questions regardless of whether or not there are useful answers. – Gerry Myerson Aug 4 '14 at 13:01
• @GerryMyerson since the deletion of answers that are worth keeping in this way (dv + autodel) should be rather rare the way you sketch seems quite alright to me. Give a list of examples illustrating the problem and I might change my mind. Until now it seems noone could produce anything remotely supporting the alarmist claims. – quid Aug 4 '14 at 14:51

(This is my first active contribution to meta (as far as I remember), although I've been watching for a while. I'm not sure if this should be a comment, but I think it's too long. I don't think it directly addresses the question, but as I understand it this is meant to be a discussion).

My instinct would be to take something of a different approach (which I haven't seen mentioned anywhere, although I may have just missed it). I don't know what the practical aspects would be like.

It seems to me that there are two main things going on. We want to reduce the number of questions basically asking for us to solve homework problems or similar. But we also want to keep the answers that are good teaching (and not really doing what we suspect the OP wants us to do).

How about actively creating more questions focused on teaching the more basic material? Set up questions on the more general questions we know are there (eg how do I tackle this type of integral?), and then rather than posting new answers to new questions, direct the OP to the already-provided teaching material on the subject. This might create a clearer space to work for those who want to focus on teaching, make it easier to shut bad questions as duplicates, and maybe also make it easier for students to find what they actually need (and so hopefully stop them asking the questions so often).

• Good idea, but not entirely new. Close to the abstract duplicate suggestion. I also approve of the alternative name minor variant suggested in that thread. The main obstacles to this are the answerers and the askers. It is difficult to make either group of posters comply. Admittedly it may also be the case that we haven't tried hard enough. – Jyrki Lahtonen Aug 14 '14 at 12:12

I'll go ahead and throw out this opinion (not necessarily my own, but one that people might like to discuss):

Answers are more valuable than questions. As a result, we should focus on preserving good answers, and simply make poor questions better.

If we don't like a question, we should close it before it's answered. Much of our problem regarding deleting questions (and answers) via autodelete stems from the division we have over closing low-effort questions. If we "closed ranks" and decided to close all low-effort questions and not answer them, then this problem would go away.

Let's not close questions as "no effort shown" that are already completely answered (unless it's been shown to be a certain instance of cheating), because this puts the answer at risk of being autodeleted.

In short, to answer the title question: "It is never acceptable, unless 1) all answers to the question are unhelpful or 2) it is certainly an attempt at cheating."

• close it before it's answered is unrealistic. It takes under a minute to bang out a FGITW answer, perhaps followed by some polishing within the 5-minute grace period. 3K users are not refreshing the Close Review queue every minute... if we decided to [...] not answer them -- sure, and if people around the world decided to live in peace and harmony, many problems would go away. – user147263 Aug 2 '14 at 4:48
• Not voting on this yet. I like the second paragraph, and the theory of the third (but the FGITWs are a reality - even in the case of a probable duplicate). The last sentence I have strong reservations about, because "helpful" is very vague, and does not give any guarantees of quality. – Jyrki Lahtonen Aug 2 '14 at 7:35
• Also, at least to a veteran, the presence of a couple of votes to put a question on hold should act as a warning sign (this effort may later go to waste) as opposed to a red flag (better answer this quick while I still can). I have more sympathy for an unsuspecting noob. – Jyrki Lahtonen Aug 2 '14 at 7:43
• @JyrkiLahtonen Sorry for being unclear with "helpful"--I meant "unhelpful" as in the "not an answer" flag option. (I wasn't looking at the options at the time, and thought that flag option was "This answer is unhelpful," not "This is not an answer.") – apnorton Aug 2 '14 at 14:15
• Though I upvoted this, I strongly disagree with the third (off-topic) paragraph. If you don't like the question, then instead of preventing teachers from answering it, do something constructive and move on to a question that you do like. – Bill Dubuque Aug 2 '14 at 15:19
• @Bill Dubuque: I have hesitated to write this, but I hope you can recognize that regardless of the merit of your argument you make it weaker by claiming so often that some users are "teachers" (while others are presumably something else). Surely opinions about what constitutes excellent "teaching" vary widely among users here. – Carl Mummert Aug 2 '14 at 19:15
• @anorton: I agree one solution would be for everyone to delay answering low quality questions. I try to do my part. But (1) it is often mathematically satisfying to write an answer; (2) easy questions are good low-hanging fruit for new answerers to build up rep to meet the key thresholds; and (3) the rep system is designed to encourage users to answer questions (much like a good slot machine is designed to encourage people to keep playing). So I think it is, unfortunately, a losing battle to close bad questions before they are answered. But I don't have any solution, yet, that I can suggest. – Carl Mummert Aug 2 '14 at 19:19
• @CarlMummert As usual, I have no interest in in discussing yet another of your strawmen. – Bill Dubuque Aug 2 '14 at 19:28

I believe that the "missing context or details" close reason should be reserved for "PSQ" - questions which simply state a question and lack context. Otherwise, "unclear what you are asking" is more appropriate. Basically, this close reason is for homework questions, and is saying "we are not really in the business of doing your homework for you, sorry!". Therefore, if such a question has generated mathematically good and upvoted answers, and a sufficient amount of time has passed (60 days, say), then I believe the solution is simple:

Edit if needed, and re-open.

The OP will have forgotten about the question, and their homework will be been handed in already. So we are no longer doing the OPs homework. So I see no problem here...

• Old questions and answers set examples for new users; they are the first thing a new user sees when arriving to the site via search. Nobody's got time for FAQs. So your message to them would be: "this is not a hw-on-demand site! But if you demand your hw to be done, we'll do it anyway." – user147263 Aug 4 '14 at 12:14
• Yes, they set examples. So edit them to be good! Also, I would like to point out that the standards of the site have progressed: A perfectly fine question from 3 years ago might be closed today. I do not think that perfectly fine answers should be deleted simply because of a shift in policy. – user1729 Aug 4 '14 at 12:28
• Sure, and I edited some. But them's the breaks: if I am the one picking up toys after the kids had their fun, I get to decide whether a toy goes back on a shelf or to the dumpster. Those who want to make sure their toys avoid the dumpster should pick them up themselves. ... Incidentally, I happen to have Copy Editor badge and you don't. So you telling me to "edit them" sounds a little odd... – user147263 Aug 4 '14 at 12:38
• @900 Two things. Firstly, my answer was addressing the abstract question, and essentially ignoring the bits addressing yourself. But because you commented I will say - you are very fast. Sure, if you are the one picking up the toys then it is, perhaps, true that you should get to decide what stays and what goes...but you still have to give others a chance! Slow down a bit - whats the rush? Second thing: My editing is irrelevant, in the sense that I am not trawling through ancient posts trying to improve this site. If I was, I would be editing them, and only deleting when necessary. As it is... – user1729 Aug 4 '14 at 13:02
• ...I have real life to contend with. – user1729 Aug 4 '14 at 13:02
• I figure others had their chance to improve posts from 2013. When I close them, they complain that close queue is flooded; when I edit, they complain that the front page is flooded and it takes One Whole Click to get the new questions. Should I quote my idol again? I will... – user147263 Aug 4 '14 at 13:14
• "Guess what: if your site is full of crappy questions, your site sucks - even if they're not highly-ranked by your own users, folks are finding them via Google, and that's where the vast majority of your readers are coming from. You can work to fix that - as painful as that process is - or you can bury your head in the sand and blame it on all of those stupid people from elsewhere." -- Shog9 – user147263 Aug 4 '14 at 13:15
• @900 I would rather you flooded the front page. As complaints go, this is less serious than loosing a good piece of work (although I am willing to admit I am not a fan of posterity, I do feel that loosing work when it could be saved is wasteful). – user1729 Aug 4 '14 at 13:18
• @900 I feel your quote is miss-placed. I do not believe that it is applicable to Math.SE in the same way as to other sites. See, for example, this question about increasing site traffic. – user1729 Aug 4 '14 at 13:19
• @user1729 IMO the post linked is further evidence of the problem and the applicability of the quote: since the quality of the questions and organisation of the content is lacking, people outside do not find what they are searching for here. I think the math site(s) really could learn some things in this regard from tex.SE for example. – quid Aug 4 '14 at 15:01
• @quid Hmm, I understand what you are saying, but I do not think I agree. If you are wanting help with a specific integral, then I do not know how searching would help. In other parts of the site, searching is already useful. But, perhaps, there is a 3-way correlation between quality, level of question (school vs PG), and possibility of using words in a searchable way. – user1729 Aug 4 '14 at 15:32
• All these "good" answers by some of the "best teachers" full of nothing but formulas, even implication arrows all around(!), on top of that pointless formatting and still on top of that done abusing MathJax for it. Obviously things get tricky search wise with posts full of nonsense like ${}{}{}{}{}$ and without any words.(Okay, exactly that one might not be in a post but a comment tyically, but you will know what I mean I guess.) – quid Aug 4 '14 at 16:42
• @BillDubuque you might try to read again what I wrote.; it might make sense then. But is not that important either. We might discuss it when you explain the unique specialness of a math site to me. – quid Aug 4 '14 at 22:42
• @BillDubuque This is really pretty tangential here, but okay here is one. There are better examples but for instance math.stackexchange.com/a/129826 – quid Aug 4 '14 at 22:55
• I agree that objecting to the implication arrows is a little odd -- my issue with some of the posts in the style you refer to are when a key piece of information is pulled out of thin air, so that the implications are mainly useful for showing how to verify a solution and not how to find the solution, or how to carry out a line of attack and not how to discover the attack. (of course, more verbose style can also suffer from these problems) – Hurkyl Aug 5 '14 at 23:54