19

We still have trouble with unanswered questions, as has been discussed previously. At the moment, the total is 3257 unanswered out of 27476 total questions, or nearly 12%. Regardless of what we do, this situation does not seem to have changed very much with time: always around 10% are unanswered.

I think that part of the problem is that some questions are not obviously unclear or irrelevant to our topic, but nonetheless embody subtle misconceptions that make it impossible to answer productively. A direct answer to this type of question requires acceptance of those misconceptions at face value, while an attempt to guide the querent in a more productive direction would distinctly not be an answer, and in fact would render the question obsolete.

Let us take this one as an example. Here, OP asks for something that certainly has a direct answer, although not in a very compelling or well-defined way (for example, benchmark code should have been given, otherwise the exercise is arbitrary and it will not be meaningful to compare the results). But whatever answers one may give certainly will not be of any use in actually doing GPU-based computations in Mathematica, because the idea behind the question is overly simplistic.

Such questions arise, I think, because people will often ask a question based on an early understanding of their problem. In this sense they are highly localized, but because the OP's assumptions are normally implicit, it is difficult to express that clearly enough in relation to the question content to argue for closure. If the misconceptions related to Mathematica itself, we would have no hesitation or difficulty in closing the question as a simple mistake. The situation is objectively no different if the misconception lies in another field, but since such a question might be perfectly well defined in its relationship to Mathematica, to express this may seem improper.

I have decided to try an experiment and used an unusual argument for closure: that to answer the question would not be productive. Obviously, this opinion is highly subjective, and I would not advocate its unrestrained use. But for long-standing unanswered questions that one strongly feels are ill-posed or misguided, I cannot think of a better close reason, even though it is clearly pointless to have such questions remain open and unanswered forever.

I stop short of a call to action based on this idea, but I would certainly like to hear others' opinions about it. Then, if we decide to use it as an actual close reason, we will be able to link back to this thread to help explain the rationale behind it.

| |
  • 2
    Isn't downvoting the right path if "one strongly feels [the questions] are ill-posed or misguided"? – Dr. belisarius Sep 25 '15 at 14:47
  • 6
    @belisarius downvoting can also be interpreted as an admonishment to the querent, which is not always called for in these situations, and there is the additional obligation to justify it in comments. It also does little to address the problem of closing these questions, especially if someone upvotes. I am not saying one shouldn't downvote, but it doesn't really help, in my opinion. – Oleksandr R. Sep 25 '15 at 15:15
  • 4
    "Querent". What a lovely word; something I would have expected on a different SE site. – bobthechemist Sep 27 '15 at 0:48
  • @bobthechemist in fact, I got the idea of using this word instead of the more plain-spoken "questioner", "asker" or "inquirer" from English.SE. It is nominally obsolete, but I thought it was worth reviving. After all, some SE questions are very much like complaints or inquiries into the occult! – Oleksandr R. Sep 27 '15 at 1:01
  • @OleksandrR. Fair enough – Dr. belisarius Sep 30 '15 at 4:59
7

Since there are now 11 votes in favor and none against, I take it that this approach is going to be reasonably well received. So, I will begin working through the unanswered list when I have some time, and I would be grateful if others will try to do the same.

Note: I don't know what up/down votes on this answer would be intended to express, so please leave a comment if you feel inclined to vote.

Addendum: I have used up all of my close votes for today and didn't have to use this close reason once. So perhaps these questions are not as common as I thought, although at least we have something of a consensus about closing them now.

| |
4

Although I agree that Mathematica.SE has far too many unanswered questions, and that many of them should be closed, I am concerned that too many were proposed for closure today, leading to 20 - 25 questions in the close queue many times during the day. With so many questions to be considered, otherwise diligent reviewers may be tempted to give only a superficial review before voting to close. Thus, I recommend that we not nominate old questions for closure, when the number of questions in the close queue exceeds, say, 10. Further, I recommend that reviewers not succumb to the temptation to vote to close simply because the questions are old.

| |
  • 1
    In my view this would make the process of reducing the unanswered count too slow. I think if I can personally nominate 50 questions for closure in a day with a somewhat reasonable close reason given for each one, that it should not be too challenging for people to decide whether they agree or not. I don't plan to do 50 every day because it is too much like hard work, so the queue will not keep its contents for long. We are fortunate to have many people doing review tasks, so if you get tired of reviewing, just stop for a while and don't feel obliged to do the entire queue. – Oleksandr R. Sep 27 '15 at 4:00
  • 1
    I think it is worth adding that I regularly notice what seem like too-hasty close decisions by others in the queue history and would like to join your call for close votes to be cast only after careful thought. In particular, if a question has answers, it is only rarely useful to close it. I would say this should only be done if the question and its answer(s) are basically not useful to anyone apart from the OP. I saw one question with close votes that admitted a useful answer (which I gave), but several "leave open" votes still did not counteract the accumulated closes. – Oleksandr R. Sep 27 '15 at 4:04
  • 1
    People need to be aware that close votes strongly outweigh "leave open" votes. So one should be certain that one really wants a question closed before voting to close, because the odds are stacked against it staying open after it has close votes. I feel that my own votes are justified but I would not want for people to click "close" just because they trust my judgment. It is important that everyone decides for themselves. – Oleksandr R. Sep 27 '15 at 4:12
  • By the way, it may be better to raise this as a new topic, rather than posting it as an answer here. Even though it is related to the task of reducing the unanswered question count, it does not really have much to do with this proposed close reason, which I haven't actually needed to use so far. – Oleksandr R. Sep 27 '15 at 4:39
  • 3
    153 unanswered questions have been resolved (answered or closed) since I started this thread. This seems to be definite progress and the review queue is almost empty at the moment, so the problem is being dealt with fairly efficiently as it would seem. I have no votes left for another 9 hours, but will flag some more (but maybe not all 24) when I get votes again. – Oleksandr R. Sep 27 '15 at 15:00
3

It's worth pointing out, that questions stay in the unanswered category, even if there are answers, as long as none of the answers is upvoted or accepted. Therefore we should consider giving answers of unanswered questions an upvote, as long as they answer the question sufficiently, even if we don't consider them helpful personally.
Some of the unanswered questions where asked by people no longer active on this site, therefore one can not expect that any answer will be upvoted or accepted by the querent. Moreover, due to the lack of quality of some of these questions, one can not expect high quality answers to them.

| |
  • Yes, it is a good point. I have seen a few questions gaining close votes recently that could quite reasonably be answered or have an answer upvoted as another way to get them off the unanswered list, so I have been doing the same. I would encourage people to prefer that approach where it is in fact realistic to expect an answer, unlike the case I address in this thread, which relates to the situation whereby no answer could be meaningfully be given. – Oleksandr R. Oct 25 '15 at 16:55
  • I made this answer community wiki so others might add similar answers they gave, but that stayed unappreciated, to the end of my answer. That should bump this topic up. – Karsten 7. Oct 25 '15 at 16:55
  • Alternatively, a notice on chat should be OK too. – Karsten 7. Oct 25 '15 at 16:56

You must log in to answer this question.

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