11
$\begingroup$

This question is about users who stubbornly only provide Maple code and their output as answers, even if the question does not ask for it.

Yesterday I decided to review some low-quality posts (there were 3 in the queue). While doing it, something completely unusual happened: I was flooded with posts flagged as "low-quality" (by whom? by users or by the MSE software?), all of them belonging to a particular user. They all followed the same pattern: they provided Maple code solving the OP's problem.

I had a hard time deciding what to do: the answers were properly formatted, clear, intelligent and correct. When a result (as opposed to an explanation) was required by the OP, the answers provided it. It is also obvious that this user knows Maple quite well. On the other hand, these answers never explained anything on the conceptual level, not even when the OP explicitly requested it. Plus, it is my strong belief that blind computing does not equal understanding or insight, even though it sometimes leads to it.

In the end, I decided to vote against deletion, but I also downvoted all of them (I think they were 10 of them, much more than the queue was initially showing). To my surprise, less than 24 hours later they have all been deleted, probably some other reviewers were stricter than me.

Should a moderator have a talk with such users, or is this attitude considered acceptable?

$\endgroup$
  • 1
    $\begingroup$ @quid: Indeed, the question that you refer above is motivated by the behaviour of the same user. This makes me wonder: is he a salesperson for Maple? Shouldn't he disclose this in his profile? Please also note that the other question hasn't obtained an acceptable answer, therefore I cannot click on "That solved my problem!". $\endgroup$ – Alex M. Aug 30 '15 at 11:09
  • 3
    $\begingroup$ 'Please also note that the other question hasn't obtained an acceptable answer, therefore I cannot click on "That solved my problem!".' I am not sure what you mean to say. Generally, I feel way too much is made of this. The site is full of answer that just provide solutions where an explanation is requested or would be desirable. (Some of them presumably also obtained by a CAS while not even this is disclosed.) If you think an answer is not useful: downvote and move on. $\endgroup$ – quid♦ Aug 30 '15 at 11:22
  • $\begingroup$ @quid: No, this is a funny misunderstanding. Take a look at my post above: what I see is a clickable button on which it is written: "That solved my problem!". Do you see it too, or is it shown only to me as the OP? I do not know what happens if I click it. $\endgroup$ – Alex M. Aug 30 '15 at 11:25
  • $\begingroup$ No I do not see it. But I knew you see it and I know that when you click it it would close your question as duplicate. What I do not know is what aspect of your question is not address by answers to the question I link to. There is an answer by a moderators saying they are not concerned by this. $\endgroup$ – quid♦ Aug 30 '15 at 11:27
  • 1
    $\begingroup$ @quid: I see. Well, apparently I am not satisfied by any of the answers there, in the same way that Bill Dubuque hasn't been either (or else he would have accepted one of them). My problem is that I suspect this user to be an undisclosed Maple salesperson. Without swift moderator action, this runs the chance of turning MSE into an advertising and marketing battleground, which neither of us probably wants. Think of what Wikipedia is undergoing with biased edits - something similar in spirit (but of lesser intensity, agreed) could happen here in the near future. $\endgroup$ – Alex M. Aug 30 '15 at 11:32
  • 2
    $\begingroup$ The question you ask got answered there. That you do not like or agree with the answer is irrelevant for deciding if it is a duplicate. $\endgroup$ – quid♦ Aug 30 '15 at 11:36
  • 1
    $\begingroup$ @ArthurFischer: Please note that your edit of my question completely changes its meaning in a way that I cannot condone anymore. It would have been nicer to suggest those edits in a comment, rather than do this kind of surgery on it which, from my point of view, is close to vandalism. Let us stop being protective of users' sensibilities: my question was specifically about one precise user, having almost 300 "borderline" answers. How many such weird cases are there on MSE in order to warrant the formulation "such users" instead of "this specific user"? $\endgroup$ – Alex M. Aug 30 '15 at 11:48
  • 3
    $\begingroup$ I would continue to downvote consistently and hope others follow. If the poster of these low-quality answers starts to whine and complain, then raise a flag. (The poster likely has anyway.) @AlexM.: I have had similar suspicions that the poster is in Maple sales. If that is true, then that is a very very poor reflection on Maple that they would target this site for pollution. $\endgroup$ – Ron Gordon Aug 30 '15 at 11:51
  • 4
    $\begingroup$ @Alex: Questions on meta should really be about behaviour and not about individuals. The behaviour appears to be the posting of CAS-only answers. And it shouldn't matter who is posting them, be it user994985, André Nicolas or myself. $\endgroup$ – user642796 Aug 30 '15 at 11:54
  • 1
    $\begingroup$ @Alex: you did the right thing by considering the content and coming to a decision whether to downvote and/or delete. I might have been more firm about deleting ones that give a Maple answer to a conceptual question. But the decision is up to you as a reviewer, in the end. Unfortunately, there are several users who post non-mathematical "answer only" posts, and there are users who vote up such posts, even though the answers are arguably not mathematical. $\endgroup$ – Carl Mummert Aug 30 '15 at 15:19
5
$\begingroup$

Good question. The answer, as almost always, is "it depends".

When the question (clearly) requests a conceptual answer, the Maple code will likely not be very useful. In these cases, I would consider downvotes and/or deletion votes warranted.

If on the other hand, there is both a conceptual and a computational aspect to the question, then the Maple answer can be argued to have some value. One can still decide to downvote on a case-by-case basis, based on personal preference, but deletion seems too much.

Surely, these answers are perhaps less likely to become the accepted answer, but they have value by exposing the capabilities of Maple. This may aid OP and visitors in e.g. finding the value of an integral before asking conceptual questions about techniques for finding the value.

In my opinion maths.SE should allow for different approaches towards a question. CAS's are one possible approach among a broader spectrum. Having a question answered from multiple perspectives can greatly add to the value for future visitors. So if one sees a question with only a CAS answer that one thinks deserves a more thorough theoretical explanation, please consider writing another answer, to exist side by side with the CAS answer.

NB. It might well be that nearly all answers of the user under discussion fall into the first category – each may vote as they want. However, I deemed it more useful to discuss the general situation.

$\endgroup$
  • 3
    $\begingroup$ The problem with a CAS-generated answer is that there is zero insight behind the answer. In most cases, and in Maple and Mathematica certainly, the result is generated with a proprietary algorithm that none of us save employees of these companies will ever know. So we have to take the correctness and of the answer on faith. Maybe I am being too extreme, but here's a question: how do you know that a Maple algorithm works in all possible cases? Where might it have problems? While it is of course a valid approach to use a CAS to perform computations, there is no insight behind it. $\endgroup$ – Ron Gordon Aug 30 '15 at 11:59
  • $\begingroup$ @Ron You certainly have valid points. People should be aware of possible errors in CAS's and should not rely solely on them. In most cases, I would prefer a question to have an insightful answer besides any CAS answers. However, I think that this does not prohibit CAS answers. They might not be the most useful, but I firmly contend that they have a place on MSE, albeit small. (Obviously I'm not talking about "just put Integrate[f, x] into Mathematica" type answers. For those, I agree with you. It's more about "Here's a nontrivial way to use built-in functions to compute what you need".) $\endgroup$ – Lord_Farin Aug 30 '15 at 12:14
  • 6
    $\begingroup$ I think that, unless the question explicitly asks how to compute something in a CAS, a mathematical answer to "How do I compute this" is not "Use Maple". However, the use of Maple might make a reasonable comment below the question, to add context to the question of how to perform the computation. $\endgroup$ – Carl Mummert Aug 30 '15 at 15:20

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