24
$\begingroup$

Recently, user answered this question Inequality for Distribution of points in space, but his solution was wrong. (I don't know in what detail am I supposed to explain things here since this is my first post in meta, but basically, he swapped $\sin$ with $\cos$ and then miscalculated something. If needed, I can provide more detail)

I read his answer and saw it was wrong, so I commented on that. He almost immediately erased the answer. I think it would have been better if the answer stayed there - it showed one false solution using a wrong counterexample that probably someone other would come up with, and then I commented on why it was wrong both intuitively and where it was miscalculated.

To sum up, I think it was an educative-to-read wrong solution. Is it better to erase such an answer or to leave it be?

Now for the other thing about the same question - and I have a feeling that I should maybe start a new thread regarding that issue, if yes please tell me and I will do so, but it also seems appropriate to tell it here.

One user posts a smart and correct solution saying "no" to the original question, a day after some other user posts wrong solution saying "yes" to the original question, and the guy that originally posted the question accepts the wrong solution as the correct one. It was virtually impossible he didn't see the correct reply saying "no" since it was older.

Two things are confusing me with this. Firstly, the user that came second and wrote wrong solution. I mean, he (could have) read the correct solution. Why post another wrong solution after that? Or if he thinks his is correct, why not trying to find an error in the previous solution and commenting on it?

And of course, other thing, why did the guy who posted the question accept the wrong solution? Oh, and also, since it was a question with bounty on it, and accepted answer now has 2 down votes, did he still receive the bounty?

Also, to the people that already got into New year, happy New year! To others, have a great Eve! :-)

$\endgroup$
  • 2
    $\begingroup$ I have added the tag "specific question" because I think the issues here are not typical of "wrong answers", esp. in regard to the level of argument and exposition on offer. Of course your comments politely pointing out errors in answers are welcome, and if you have similar thoughts on the Accepted Answer (and the energy to articulate them), that would help current and future Readers. $\endgroup$ – hardmath Dec 31 '19 at 17:28
  • $\begingroup$ I see, it is a good tag since for the second part of my question I do address that one question in particular. And no, I don't have any "smart" comments on the wrong Accepted Answer, I just found the other (now deleted) wrong answer to be a "good" wrong answer about which I did have some comments which future readers might find helpful. $\endgroup$ – prosinac Dec 31 '19 at 17:52
  • 1
    $\begingroup$ If you're convinced "it was an educative-to-read wrong solution," then you could post it yourself, pointing out exactly what's wrong with it. $\endgroup$ – Gerry Myerson Dec 31 '19 at 17:57
  • 5
    $\begingroup$ As for the person accepting a wrong answer, see xkcd.com/386 $\endgroup$ – Gerry Myerson Dec 31 '19 at 18:02
  • 3
    $\begingroup$ If there’s something to be learned, I’d prefer it stayed there. Otherwise, it’s of no value. $\endgroup$ – URL Dec 31 '19 at 18:52
  • $\begingroup$ Somewhat related: Is sometimes keeping wrong answer reasonable? $\endgroup$ – Martin Sleziak Jan 2 at 2:43
  • $\begingroup$ It may be good to keep a wrong answer. Add at the top that it is wrong. But if it collects downvotes, that is an incentive for the solver to simply delete it. I have enough rep that I have on occasion left incorrect answers that I have written. But someone with low rep may not have that luxury. $\endgroup$ – GEdgar Jan 3 at 11:39
31
$\begingroup$

Regarding your question about keeping flawed answers: there are several reasons people delete their answers once a fundamental flaw has been pointed out:

  • the question will not appear in the Unanswered queue if it has an upvoted incorrect answer, so deleting the flawed approach allows the question to receive more attention;
  • the answer author may want to fix the answer, but needs more time to work on the fix. Deleting the answer is a cleaner way to avoid confusion than alternatives (like putting a "Don't look at this yet! It's wrong!" message at the top of the answer).
  • having a wrong answer attached to one's name can be embarrassing.

I agree that it might make sense to keep some wrong answers around, if the mistake in the answer is instructive (and the answer is edited to clearly mark it as a failed solution attempt). But I also think that the decision should be up to the answer author on a case-by-case basis.


Regarding why users accept incorrect answers despite being presented with a correct alternative: who knows? Users do all kinds of irrational things.

$\endgroup$
  • 14
    $\begingroup$ I wonder if it would be useful to add a feature where you can "deprecate" answers, meaning that users can still see it and comment, but voting is disabled and it does not count towards the answer count. In addition, it could be undeprecated if fixed. $\endgroup$ – PyRulez Jan 1 at 23:25
  • 1
    $\begingroup$ net 3 downvotes will grey it out. @PyRulez $\endgroup$ – Roddy MacPhee Jan 2 at 0:26
  • 1
    $\begingroup$ @PyRulez: great idea! It might be used for "not an answer but..." answers, which I think would be an improvement on the status quo. It would be nice to have statistics about how much a wrong answer discourages correct answers (I suspect there is a large effect). $\endgroup$ – Dap Jan 2 at 14:24
  • $\begingroup$ An alternate argument is that an instructive failed argument can be added to the question body - the question itself is the correct place for all relevant information that isn't an answer. $\endgroup$ – Dap Jan 2 at 14:38
  • 1
    $\begingroup$ @Dap no, it shouldn't be. That would give the false impression that the asker is looking for a comment on the incorrect answer, when they are in fact not asking that and it has nothing to do with their problem. If the answeree wants feedback on their incorrect post, then they should post a new question. $\endgroup$ – user64742 Jan 4 at 20:08
8
$\begingroup$

When someone points out that an answer of mine is wrong I usually delete it. But if it's wrong in an interesting or useful way I edit it to say so and leave it.

Here's one example: Symmetry group of a cube: $O_h \cong O \times C_2$ . I don't know whether it was accepted before or after my edit.

$\endgroup$
0
$\begingroup$

Yes and No.

Keeping an incorrect answer prevents the asker from getting the correct answer as people won't see it in the unanswered queue, so for that reason alone the answer should be removed.

However, there may be times in which a minor error in an answer once pointed out in comments or via an edit makes the answer good enough for the asker.

Ultimately it depends on the asker at least from a functional standpoint of the immediate goal of getting the asker their answer. From a general standpoint of wanting the site to be useful to future visitors though, it all depends on the quality of the bad answer.

I have been the victim of bad answers in the past, and I would have preferred the answeree asking "would you like me to delete it". I would have responded "yes, and if you'd like I'll alert you when a correct answer comes around so that way you can choose to undelete without causing problems".

Of course, as always the asker is completely allowed to delete their incorrect answer whenever they wish. If for some reason it gets falsely accepted (and then continues to accrue downvotes)... I'd ask for people with delete vote powers.

$\endgroup$

You must log in to answer this question.

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