15
$\begingroup$

Note 2: †

This is NOT a duplicate of the linked question: I am NOT asking at all "Under what circumstances is it appropriate to delete a question that has received a good answer?" Again, this post is NOT about when or whether we should delete a question, NO. It is a discussion about the merits of deleting answered questions in the main site.

Added for responding to a close vote.


Note 1:

This is NOT a duplicate of the suggested link (Requests for Reopen & Undeletion Votes (volume 07/2018 - today)) at all. While I disagree with the linked questions below being deleted, this post is NOT about voting to undeleting or reopening questions at all, NO!

Rather, I am seeking a general explanation/discussion of why such deletion is (or could be) of value to the main site of MSE.


[Original question.]

I have noticed recently that quite a lot of answered questions, some of which are still "on-hold", are being deleted by mostly a small group of users in the chat room CRUDE:

I highly doubt the justification of such deletion since it makes others' work (mostly) disappear. A (very) high-rep user (now no longer active) had a similar concern for such action before:

Is there an MSE-like site that is more pleasant to work in? Or are there other solutions?

Question:

What are the merits, if there is any, of deleting (correctly) answered questions in the main site of MSE?

$\endgroup$
28
  • 11
  • 9
    $\begingroup$ You should probably be a clearer about what you want to achieve by poking this bear. Are you in favour of any answered question, no matter how ill-posed or against the 'good question' guidelines being kept? Are you in favour of a certain subset of them (please describe how that subset should be determined)? Are you complaining about what you perceive as organised behaviour (a cartel, perhaps?)? The merits, if they are such, are easily found as YuiTo Cheng has pointed out. $\endgroup$
    – postmortes
    Mar 17, 2019 at 16:56
  • 3
    $\begingroup$ I also notice that you've recently edited the 'how to ask a good question' thread. Are the changes you've made there relevant to this question, perhaps? $\endgroup$
    – postmortes
    Mar 17, 2019 at 16:59
  • 5
    $\begingroup$ @postmortes: "The merits, if they are such, are easily found as YuiTo Cheng has pointed out." Then I suggest you put them into an answer. I do not find it easy to find. $\endgroup$
    – user9464
    Mar 17, 2019 at 17:22
  • 12
    $\begingroup$ The main merit, as I see it, is that it discourages users from answering do-my-homework posts. It's easy to understand how a new user comes here, sees thousands upon thousands of low-quality copy-paste questions answered, and thinks that they can follow along - deleting these posts is a token effort towards correcting that assumption. The only harm that I see is to the pride of the answerers… and to their reputation scores. $\endgroup$
    – user296602
    Mar 17, 2019 at 21:51
  • 6
    $\begingroup$ "The main merit, as I see it, is that it discourages users from answering do-my-homework posts." -- I wonder if there's any sort of statistics that could be somehow obtained that verify how useful closing/deleting such questions is. As frequently as they pop up, one would argue it doesn't help, or at least doesn't do enough to stem the flow of such question. Of course, I haven't been around long enough to have witnessed a period before such a policy on PSQs/etc., so maybe it was just WAY worse back then (if there was a "back then"). Makes me wonder if there's anything else we could do though. $\endgroup$ Mar 17, 2019 at 22:36
  • 5
    $\begingroup$ I find very unfair to delete answered questions (no matter what kind of question) I know that do-my-homework questions or psq questions are not allowed because of the rules of the site, but in the end the answer it's valuable knowledge. It's also not right that if you put a bounty on a do-my-homework post, then no one argues about it. I mean, even one mod said once to not to flag a do-my-homework post or psq question if this one has a bounty... $\endgroup$
    – user486983
    Mar 18, 2019 at 6:26
  • 3
    $\begingroup$ @Isa the value of knowledge lies in its applicability: knowing the maximum percentage of oxygen permissable in air before oxygen-toxicity becomes a problem is valuable knowledge, but arguably not on this site :) Your comment about bountying a do-my-homework post probably deserves its own question. $\endgroup$
    – postmortes
    Mar 18, 2019 at 9:11
  • 5
    $\begingroup$ @Saad I think the OP doesn't just want to undelete specific questions mentioned in his post; Rather, he wants a general explanation of why deletion is of value to this site. $\endgroup$ Mar 18, 2019 at 11:41
  • 7
    $\begingroup$ After I posted this question, two of my old (more than one year) well-received questions are downvoted and one being voted to be closed. As an experienced user who has asked and answered hundreds of questions, I certainly know what such behavior of sudden labeling of other's work as a so-called PSQ means. $\endgroup$
    – user9464
    Mar 18, 2019 at 12:37
  • 3
    $\begingroup$ @YuiToCheng: While they may be relevant (thanks for the links), I have explicitly asked a general question that none of those posts has such scope. $\endgroup$
    – user9464
    Mar 18, 2019 at 12:45
  • 4
    $\begingroup$ And I don't think the answer is that clear. As mentioned as a comment to quid's answer, the OPs of these closed answered questions are probably not coming back anyway, so it won't affect them. Newcomers won't see the deleted questions, so won't learn from their closure/erasure. The only argument is that it will deter the answerers... but again, that's for stuff they answered 5 months or 5 years ago, so... really? Instead of changing their future behavior, that might just piss them off and drive them away (and thus also lead to fewer good answers on good questions) $\endgroup$
    – Clement C.
    Mar 18, 2019 at 16:57
  • 7
    $\begingroup$ I am adding one more upvote - not necessarily to imply that I agree with the post, but simply to get it to the score $+3$, which is the cut-off for being shown in the community bulletin. This might increase that likelihood that more users notice this discussion and contribute to it. $\endgroup$ Mar 19, 2019 at 13:20
  • 4
    $\begingroup$ And... The post has been deleted, marked as duplicate in spite of disclaimers explaining why and how it differed from suggested "duplicates." Well, it's nice to have conversations while they (are allowed to) last, I gather? $\endgroup$
    – Clement C.
    Mar 20, 2019 at 14:38
  • 4
    $\begingroup$ Seems like that's been rectified by now. ... Not sure why people were voting to close despite the clear indications that this was not a duplicate but whatever. @ClementC. $\endgroup$ Mar 20, 2019 at 23:12

4 Answers 4

16
$\begingroup$

Deletion sends a rather clear message that the content is not welcome. To send this message is a merit of deletion. By contrast to put on hold/close yet not delete does not achieve much especially for answered questions.

Of course, one might think that the content is or should be welcome, but that would be a separate concern. If one wishes to discuss this, then one should start a step earlier namely question the closures.

$\endgroup$
43
  • 9
    $\begingroup$ Logically one might welcome an Answer's content while objecting to a Question's content, so the OP may have intentionally broached the issue at the step of deletion. But +1 for addressing the purpose of deleting content. $\endgroup$
    – hardmath
    Mar 18, 2019 at 0:36
  • 9
    $\begingroup$ The main point of closing is to prevent answers from being given. It seems not quite coherent to me to be in favor of preventing answers to a question yet at the same time being in favor of preserving the answers that in one's opinion should not have been given in the first place. At least as a general approach. (There can be corner-cases.) $\endgroup$
    – quid Mod
    Mar 18, 2019 at 2:58
  • 4
    $\begingroup$ But.... that's not going to be effective, right? Recent "previously-closed but now-deleted" questions (at least the latest two I have seen) are by new users, who didn't know the rules before asking, got an answer, and are now probably never coming back. So deleting the question entirely erases it for all future newcomers (they can't see it), and thus... they will repeat the same pattern? $\endgroup$
    – Clement C.
    Mar 18, 2019 at 16:42
  • 4
    $\begingroup$ @ClementC. to a considerable extent the message is intended for those that answered the question. Note that OP insists on answered questions. $\endgroup$
    – quid Mod
    Mar 19, 2019 at 1:22
  • 6
    $\begingroup$ -1: this answer overlooks the fact the mentioned content is "not welcome" by an active minority of users. Unless one tracks their activities the majority of users do not get to see the deletion of answered questions, often years after the fact. As we see here in meta, when such a case is presented openly, it is very common for the deletion decision to be reversed by the majority. So we do have an issue, in that a minority is destroying content, technically in the open, but in practice under wraps. $\endgroup$ Mar 23, 2019 at 2:18
  • 10
    $\begingroup$ @MartinArgerami "it is very common for the deletion decision to be reversed by the majority" those that vote to undelete on any regular basis to me seem to be just as much a minority. Do you have any basis to claim that this is a majority? Majority of what even? In actual fact the vast majority of users does not seem to care much either way. $\endgroup$
    – quid Mod
    Mar 23, 2019 at 8:44
  • 4
    $\begingroup$ @quid: what message? It often happens that not even the person whose answer is deleted learns about the deletion. It's plain destruction of content, with no message. $\endgroup$ Mar 23, 2019 at 15:26
  • 7
    $\begingroup$ Finally, given the constant stream of complaints almost exclusively by those that created the content, clearly it's noted and not appreciated. Whence the message works as intended. That might sound cynical, but that's how it is. $\endgroup$
    – quid Mod
    Mar 23, 2019 at 17:48
  • 3
    $\begingroup$ @JyrkiLahtonen You wrote "from what I have seen, the active users of C.R.U.D.E. are also among the most active reopen/undelete voters". I highly doubt that. I closely monitor CRUDE and it is quite rare that anything is reopened or undeleted there (except in rare cases when non-crude-members make strong objections, and even then it takes tremendous effort to have a small chance of success). $\endgroup$ Apr 6, 2019 at 21:44
  • 5
    $\begingroup$ @quid Thanks for sharing your opinion, but I strongly disagree. I prefer to only review posts on topic where I have much expertise. I recommend that everyone do so. $\endgroup$ Apr 7, 2019 at 20:58
  • 3
    $\begingroup$ @quid Again, thanks for sharing your personal opinions. $\endgroup$ Apr 7, 2019 at 21:02
  • 3
    $\begingroup$ @quid & Jyrki Please don't ping me any more with such nonsense. The amount of false presumptions made above is utterly astounding. How to lie with statistics on steroids! Those numbers are completely meaningless without any understanding whatsoever of the context. $\endgroup$ Apr 7, 2019 at 21:08
  • 3
    $\begingroup$ @quid Do you seriously think that the SE designers want users to vote on posts (far outside) their expertise (or interests)? If so, then the platform is even more poorly designed than I ever imagined. In any case, I rarely use the queues so it is a moot point. $\endgroup$ Apr 7, 2019 at 21:19
  • 3
    $\begingroup$ @Jyrki Once again you are way off the mark in your wild speculations on the motivations of your political opponents. As you surely know from my many remarks on such, I value answers far more than questions, so the review queues are not much use to me since I need to see the answers too in order to properly evaluate the entire thread. On the rare occasions that I peruse the review queues it is usually for other purposes, e.g. seeking statistical data such as interactions with CRUDE, or who was robo-voting very poor choices of dupe targets, etc. $\endgroup$ Apr 8, 2019 at 16:31
  • 3
    $\begingroup$ @Jyrki During such perusals I may notice some threads that I know well and (quickly) vote on them but that is very rare (e.g. 12 total reviews in the past 4 months). If you and your crystal ball truly believe that you can infer something meaningful about my motivations from that infinitesimal part of my time spent here then I recommend that you switch careers from math to astrology. $\endgroup$ Apr 8, 2019 at 16:31
10
$\begingroup$

Promoting some of the comments to an answer so that users can vote on them more freely. I got the impression that at least the OP wanted this to happen. As this is somewhat independent from my longer answer, I want to give the voters the chance to agree/disagree with some parts of what I wanted to say while being indifferent to, or even feeling the opposite way about the other post.


I agree with quid's point in that the merit of deletions is to send a loud and clear message that this content is not welcome. I want to emphasize and clarify that, at least for me, the recipients of that message are the answerers:

  • Answering a duplicate question does not add new content to the site, but it does add more entropy. This is a net loss, and I want to discourage the practice. I am on record (in meta) for being militantly anti-dupe, so I don't want to repeat it all.
  • Answering a copy/pasted homework problem invites more of the same, and gives the site a bad reputation as a homework mill. Also, the same assignment was used in the same course a year earlier, so it is often also a duplicate. I want to discourage both asking and answering this type of questions.

Simply closing the questions does not discourage any of the above practices. All the closed posts can still be voted on. The same applies to duplicates. Deletion is the only sufficient deterrent at my disposal. And also hitting the answerer is exactly my goal.


But, let's not forget that

  • Duplicates are sometimes hard to find. I really should do a better job not making this message personal. A biblical parable involving a beam and a mote springs to mind also.
  • I want to commend our moderators, and confess that once this fight against dupes/low quality content became too personal for me, and they had to send me a private reminder/warning not to do that.
  • I do believe that not all the answerers of LQ questions are in it for the sake of internet points. Yet, when the same names recur as answerers to questions in a close/delete queue I have a hard time A) remembering this, B) believing their excuses. We can blame the gamification aspects of the site for many a bad thing. But, SE wants us to play, so it is not unnatural to ask for a level playing field.
$\endgroup$
4
  • 3
    $\begingroup$ Unfortunately there are too many different points in one answer - I want to upvote some and downvote others, so the post won't be very useful as an opinion poll. Maybe you should break it into separate answers. $\endgroup$ Apr 6, 2019 at 21:19
  • $\begingroup$ I see @Bill. I was hoping to prevent that from happening, but I got too riled up, and couldn't keep the lid on. (Scratches head) $\endgroup$ Apr 6, 2019 at 21:20
  • 2
    $\begingroup$ There at least $7$ independent opinions posted in the answer and I wouldn't be surprised if all $2^7 = 128$ possible boolean combinations of such do occur in the community. Good luck interpreting the votes. $\endgroup$ Apr 6, 2019 at 21:31
  • $\begingroup$ That is unfortunate :-( Yet, this is a complicated problem, making contact with several other site policies. And, in spite of the pleas, there are no simple answers. $\endgroup$ Apr 8, 2019 at 6:45
9
$\begingroup$

The site norms about this have gradually changed over the years, so it may be helpful to take a bit of "historical" look at the problem. I think it sheds some additional light. Also, I try to explain how the scale, more precisely, the growth of the site also plays a role.

StackOverflow is a couple of years older than us, and an order of magnitude bigger. So it may be a good idea to link to a few posts I find relevant. Both the posts I want to encourage you to read were written by Shog9 - one of the community managers working for StackExchange. He has extensive understanding of the dynamics of these sites (it's his job). He is also one of the CMs who regularly interacts with Math.SE moderators, and, when needed, will post here.

You are welcome to think that I am horribly misrepresenting his points. I probably am. You see, he is using metaphors necessarily open to interpretation. See for yourselves. His older post and a more recent one. I probably shouldn't try my hand using an agricultural metaphor, but I will do it anyway :-) You can have a field day (bad pun intended) breaking my metaphor. It will do just that. Probably faster than I expect it to.


In its infancy this site did not have this problem. There was room for everything. Nearly all the content was new. The users were not unlike happy gardeners in a paradise. Some wanted to nurture rare orchids, some wanted to concentrate on farming food. Yet, everything was small scale. If a weed showed up, we had the resources to study it carefully. Painstakingly isolate it from the wheat and the orchids, try and mutate it to something more interesting. Failing that we would uproot it, by hand, so as not to disturb the plants near by. Some did not want to participate in that kind of gardening, but that was tolerated, because dealing with the weeds was not too time consuming.

When the site started growing, problems started appearing. Some would have liked to continue doing what they had always done. Others pointed out the benefits of crop rotation, pointing out that monocropping depletes the soil of some nutrients, making it less fertile for seasons to come, and also making it easier for pests to appear. Fights broke out. In the end some kind of a truce was reached. Even though there was some collateral damage, we still had enough resources to protect some of the useful plants that happened to be close to a weed.

But the site kept growing. Needing to feed a lot of people, we, almost without noticing exactly when the change happened, switched to intensive farming. We still want to do crop rotation. But, the scale has gone up. We no longer have the resources to carefully examine each and every plant by hand. Says Wikipedia: It takes very few wild oat plants to cause a significant reduction in the yield of a wheat or cultivated oat field, even though the seeds are a type of oat. So when we spot a few wild oat plants, severe collateral damage will be necessary. The farmer happily sacrifices many good stalks to make sure all the wild oats are destroyed. Whatever it takes to save the main crop.


So that's why many deletions in the current stage of the development of the site are IMHO necessary.

  • The questions may be monocropping duplicates, or pests of attempts to outsource homework.
  • Both of those must be stopped, for they will invite their "friends" to crash the party. If the price we have to pay is the destruction of a few correct answers, so be it.

But, that's not the end of it. What lies in the future? Reread Shog9's later post. Water. If we don't accommodate for it, it will destroy the farm. And. Water. Does. Not. Care. What/who is water here? Many interpretations are possible. His most important message may be (my interpretation) that we all need to brace for a future when we ALL need to compromise on some of our dearest principles. Linking to the infamous post describing four types of SE users

  1. The librarians may be forced to accept that they cannot have full control of the content. Some may leave the site in disgust, but do observe that the librarians will get reinforcements from the ranks of former prolific answerers.
  2. The noobs will also need to learn the site norms. But, the continued existence of the site needs new users. Many a noob will inadvertently start out as one of this type. But we need many of them to stay.
  3. The most knowledgable of the noobs will become prolific answerers. If left to themselves usually not adding new content. So they will need to learn to compromise. And, in the end, they will be converted to librarians (seen this happen a few times already)...
  4. ... or possibly to an apathic shaking their head on the sidelines (this fate awaits a few librarians also).

But, while all this played out, there was the hermit, happily tending to a few orchids. Not worrying about the ego points and politics. Just beautiful math. Very Zen (unless somebody steps on his orchids).

$\endgroup$
3
$\begingroup$

My opinion, lifted from some comments I wrote above. It is not solving the problem, but arguing that the current approach to said problem is misguided.

I will develop my argumentation by building on two comments left by @postmortes (comments I disagree with), who wrote

"[I think] that's what the debate is about: whether this really is a site to get your homework done for you"

and

"If the site isn't about getting your homework done then it will be hard for anyone to argue that PSQs shouldn't be deleted."

Indeed, we are talking about past, already answered PSQs, not new ones. The debate is rather

Is deleting those past, answered, closed questions an efficient way to prevent new ones (PSQs, HW, etc) appearing?

And I don't think the answer is that clear. Namely:

  • The OPs of those closed, answered questions were usually first-time posters; were either discouraged by the response (closure) or got what they want, and are probably not coming back anyway: so it won't affect them.

  • Newcomers won't see the deleted questions, so won't learn from their closure/erasure.

So it doesn't deter the posters. The only possible argument is then that it will deter the answerers from answering those poor-quality undesirable questions in the future. Right?

But that's penalizing said answerers for things they answered 5 months or even 5 years ago! I really doubt this will have the desired effect. Instead of changing their future behavior, that might just (understandably, if you ask me) piss them off and drive them away — and thus also lead to fewer good answers on good questions.

So you end up with less valuable content (which in many case was good quality, even though the questions were not), and fewer of the previous answerers willing to contribute to good content in the future. Maybe I'm pessimistic, but that does not seem like a sound policy to implement.

$\endgroup$
31
  • 5
    $\begingroup$ This is probably quite difficult to quantify. Some user might leave because they see some of their content deleted (or content provided by others). Some users might leave because they see that Mathematics Stack Exchange resembles a homework mill. So the same actions which might drive some part of users away might be helpful in keeping other users on the site. $\endgroup$ Mar 18, 2019 at 21:08
  • 3
    $\begingroup$ @MartinSleziak Yes, and I did add these caveats (it's my opinion, and the conclusion in the middle is "I don't think the answer is that clear."). I believe it's therefore important to have these conversations, instead of simply having a few (or even many) people going to implement these new half-baked policies on their own without actually consulting first (which, and maybe I'm wrong, looks like what's currently happening). However, given the (at the time of writing) 3 "close" votes on the OP's question, some people may not agree this conversation is worth having. $\endgroup$
    – Clement C.
    Mar 18, 2019 at 21:10
  • 4
    $\begingroup$ The examples in the question mostly seem to be rather recent questions. Thus the argument on old stuff is somewhat besides the point. There is also a built in mechanism. Namely points for well-received old content are not lost after deletion. $\endgroup$
    – quid Mod
    Mar 19, 2019 at 1:26
  • 2
    $\begingroup$ The thing about reputation which quid mentioned can be found here: Do you lose the reputation (points) you acquired from a question if/when it is deleted? and Reputation tab seems to be insensitive to deletion of upvoted answers. (However, I would hope that this site has enough users for who the reputation points are not really that important.) $\endgroup$ Mar 19, 2019 at 5:56
  • 3
    $\begingroup$ I get your point. However, I am not sure that characterizing this as new and done without consulting is accurate. This topic (or at least very close to this) was discussed on meta several times, for example, here is a post from 2014: Under what circumstances is it appropriate to delete a question that has received a good answer? $\endgroup$ Mar 19, 2019 at 6:01
  • 2
    $\begingroup$ @Isa well, and then they did not leave but continued to answer, until then they stopped to be active for in all likelihood altogether different reasons. $\endgroup$
    – quid Mod
    Mar 19, 2019 at 6:07
  • 2
    $\begingroup$ Re: Newcomers won't see the deleted questions. That's true. Still they see the ones that are not deleted and can use them as examples. There were many posts on meta of the form: "Why was my question closed? Look - this one is similar and it wasn't." (Of course, this is argument more in favor of closure than in favor of deleting. Even if the question is kept, it should be clear that it does not have sufficient quality. Here is a relatively recent related discussion: Historical lock for preserving old, low quality questions.) $\endgroup$ Mar 19, 2019 at 6:07
  • 7
    $\begingroup$ Clement, thanks for coming here. If a representative fraction of users opposed to deletions would come to a round table discussion, then we might actually talk it over and come up with a compromise. Alas, I'm not very optimistic about that happening, but I appreciate your effort. $\endgroup$ Mar 19, 2019 at 6:11
  • 3
    $\begingroup$ OTOH, I disagree with your conclusion in the sense that deterring answers to bad questions is exactly the goal here. If it works as a deterrent, fine, mission accomplished. But it does not look like it works. We have plenty of users making an industry out of answering PSQs. $\endgroup$ Mar 19, 2019 at 6:19
  • 2
    $\begingroup$ @ClementC Like what I suggested here? Not impossible. The objections raised were: A) it does not deter the answerers not motivated by rep points, and B) it does not deter the answerers motivated by rep points, because many upvotes come quickly, and a moderator on-duty cannot convert the thread to CW fast enough unless they learn about it very quickly. $\endgroup$ Mar 19, 2019 at 7:36
  • 4
    $\begingroup$ A downvote? It is not a deterrent to an answerer motivated by rep. The reason is that the difference a downvote makes is the difference between +50 and +48 rep points. OTOH, a deletion (if done within the 60 day window) is the difference between +50 and +0. You see why only one of the alternative actions is actually a deterrent. $\endgroup$ Mar 19, 2019 at 7:39
  • 7
    $\begingroup$ @Isa and others want to leave because of the volume of content they do not appreciate, including answers to poor questions. Yet they stay because they have a good heart and despite all the frustration keep trying to maintain the quality of the resources. There is no monopoly for goodheartedness with those that answer a lot. $\endgroup$
    – quid Mod
    Mar 19, 2019 at 18:20
  • 5
    $\begingroup$ Considering the close votes to this question, most of which are from the very users who are taking the close-deletion action mentioned in the post, I now have more doubt on the justification of coordinated deletion in the CRUDE room. They do not even bother to post an answer stating clearly the merits they think exist. $\endgroup$
    – user9464
    Mar 20, 2019 at 11:49
  • 11
    $\begingroup$ @Jack Not to be presumptuous, but maybe they agree with my answer, no need for duplicating answers. Voting suffices. Further, while you insist that this is not a dupe, arguably the same discussion was had numerous times already. Right, if you want you can have your new thread, but you cannot expect that everybody engages in a near identical discussion yet again. $\endgroup$
    – quid Mod
    Mar 20, 2019 at 17:53
  • 2
    $\begingroup$ @ClementC. I only skimmed them for now. I don't think such escalating rhetoric is useful. If you think something is over the line please flag it. $\endgroup$
    – quid Mod
    Mar 28, 2019 at 9:13

You must log in to answer this question.