31
$\begingroup$

I recently flagged an answer as "not an answer", and what came back was "declined - flags should not be used to indicate technical inaccuracies, or an altogether wrong answer." I found that surprising, so I had a look at the faq on flagging and, sure enough, an answer being altogether wrong is not listed as one of the (main) reasons for flagging as "not an answer". That leaves me at a loss to know what I should do when I come across an altogether wrong answer. Downvoting, even downvoting and leaving a comment, somehow doesn't seem strong enough. I don't want the answer downvoted; I want it deleted.

So: is there a consensus that one doesn't flag altogether wrong answers? And, if so, what is the recommended course of action when one comes across an altogether wrong answer?

$\endgroup$
  • 17
    $\begingroup$ On the other hand, I would not want mods to unilaterally delete answers, wrong as they are. I'd rather we decide on answer deletion the way we also decide on whether we should close a question... $\endgroup$ – J. M. is a poor mathematician Jul 24 '12 at 10:20
  • 9
    $\begingroup$ I can't resist pointing out that logically, an answer that is wrong is still, as you call it yourself, an answer (unlike, say, mock turtle soup, which is not turtle soup that is mock). It doesn't make sense to flag as "not an answer" that which is an answer. :) $\endgroup$ – Rahul Jul 24 '12 at 10:39
  • 2
    $\begingroup$ @Rahul, I knew your comment was somehow familiar... $\endgroup$ – J. M. is a poor mathematician Jul 24 '12 at 11:25
  • 2
    $\begingroup$ It means that the moderators should not clear flags with a declination status without the input of three or more users. $\endgroup$ – Asaf Karagila Jul 24 '12 at 13:28
  • 5
    $\begingroup$ @Rahul, if someone asks for $\int x^2\,dx$, and someone else answers, "mock turtle soup," I think that's both an altogether wrong answer (as in, an entirely inappropriate response) and not an answer (as in, not engaging with the question). Am I being illogical? Perhaps. But what I'm asking is, what is the correct course of action to take on seeing such a response? $\endgroup$ – Gerry Myerson Jul 24 '12 at 13:34
  • 9
    $\begingroup$ @GerryMyerson, dowvote it; if your rep allows for it, vote to delete it. If there is a point to it, add a comment explaining the problem —in some cases, a good comment pointing the problem paired with a wrong answer is more useful than a deleted wrong answer no one can see! $\endgroup$ – Mariano Suárez-Álvarez Jul 24 '12 at 13:39
  • 7
    $\begingroup$ Wrong answers can be instructive. $\endgroup$ – MJD Jul 24 '12 at 13:56
  • 3
    $\begingroup$ @Mark: Yes, wrong answers that are useful to point out how and where a natural (or naive) argument fails are good; but some answers are not like that at all. Some answers are just... useless. $\endgroup$ – Asaf Karagila Jul 24 '12 at 14:00
  • 3
    $\begingroup$ Dear Asaf, I do not agree that such flags should not be dismissed. Such a policy should be implemented in software or not all. For such matters, «experience» is defined in this site in terms of rep points, and users with enough of it can vote to delete. Otherwise we end up having a handful of people selecting who is experienced and who isn't... $\endgroup$ – Mariano Suárez-Álvarez Jul 24 '12 at 16:52
  • 9
    $\begingroup$ I think that using flags as a covert channel for inter-user communications is a terribly bad idea :/ If you want something like that, propose something sensible to the SO people. (They will propably suggest using chat...) $\endgroup$ – Mariano Suárez-Álvarez Jul 24 '12 at 18:53
  • 1
    $\begingroup$ @Gerry: Fair enough; the flag should be interpreted to mean "not an answer to the question". My point stands that this is still different from simply being a wrong answer. I'm not entirely certain which one you are talking about in this meta question. $\endgroup$ – Rahul Jul 24 '12 at 20:43
  • 2
    $\begingroup$ @J.M.: There is no finite amount of money, beer or whiksy that will get me back on the chat servers. $\endgroup$ – Asaf Karagila Jul 24 '12 at 22:20
  • 4
    $\begingroup$ @Asaf, "whiksy"? It seems that just writing about alcohol is enough to impair your typing. $\endgroup$ – Gerry Myerson Jul 25 '12 at 0:02
  • 5
    $\begingroup$ @Asaf: W H I K S Y isn't Irish, and it isn't Scottish; it's typo-ish. $\endgroup$ – Gerry Myerson Jul 25 '12 at 9:59
  • 4
    $\begingroup$ Ohhhhhhhhhhhhhhhhh :-D $\endgroup$ – Asaf Karagila Jul 25 '12 at 11:22
16
$\begingroup$

There is a strong enough base of 20k users, which can delete downvoted answers.

I know that myself and two others often delete answers. If more 20k users will take a look at the moderator tools page, and in particular the deletion votes page, we can delete these answers as a community.

I was planning on bringing this up, but you asked this question.

Simply, we should take more action as trusted users, someone downvotes and votes to delete, leave a comment to argue this point, and other users can argue against or vote to delete as well.

$\endgroup$
  • 2
    $\begingroup$ We certainly have a lot of 20k users now; this entire page lists all of them. $\endgroup$ – J. M. is a poor mathematician Jul 24 '12 at 10:19
  • 1
    $\begingroup$ Thanks. I guess the part I was missing was, under what circumstances does the vote-to-delete option appear? Is it the case that once an answer has more down than upvotes it becomes eligible for deletion? $\endgroup$ – Gerry Myerson Jul 24 '12 at 13:15
  • 3
    $\begingroup$ @Gerry: Exactly. If the score is negative, we can vote to delete it. In most cases, the answers that needs to be deleted are already downvoted or have zero votes. $\endgroup$ – Asaf Karagila Jul 24 '12 at 13:18
  • 1
    $\begingroup$ I'm inclined to accept this answer, on the grounds that I learned something useful from it. But @Henning has answered that only the author should delete a wrong answer, so I think I should encourage the discussion to continue while there seems to be disagreement on a major point. $\endgroup$ – Gerry Myerson Jul 25 '12 at 0:25
  • $\begingroup$ @Gerry, I agree. Discussions are good, but sometimes you are forced to act without choice. $\endgroup$ – Asaf Karagila Jul 25 '12 at 8:16
  • 5
    $\begingroup$ I guess that doesn't really answer the question whether one should delete wrong answers. If someone answers the question of whether a certain subset of $C_\mathbb{R}$ is compact by "Yes, since it is closed and bounded." I would be opposed to deleting this blatetly wrong answer. But I think how a poster reacts to corrections in comments is a good indicator of the level of "good faith". $\endgroup$ – Michael Greinecker Jul 25 '12 at 14:36
  • $\begingroup$ @J.M.: The first 7 pages are now exactly the 20k+ users (unless some of which are suspended right now). $\endgroup$ – Asaf Karagila Oct 18 '16 at 17:52
  • 1
    $\begingroup$ Since the question has been bumped, I will mention here that there is c.r.u.d.e. chat room, where deletion (and undeletion) of specific questions can be brought up and various related issues can be discussed. (It was not very active lately, but making users aware of this chat room improves the probability that it gets restarted.) $\endgroup$ – Martin Sleziak May 3 '17 at 7:19
  • $\begingroup$ @AsafKaragila I appreciate your answer and all the help that 20K users could provide! But, deletion is contingent upon accumulation of enough downvotes, and downvotes bring rep loss to users who cast them. I think there is something inherently wrong with this model. It might look like I just feel better when I find something to complain about, but (as much as the latter might be true, and nobody, me or someone else would win from it), I feel that there is a genuine issue here that deserves some discussion and search for a possible solution: math.meta.stackexchange.com/q/30935/188367 $\endgroup$ – Mirko Nov 25 at 1:23
  • $\begingroup$ @Mirko: It seems odd that you're so invested in your own reputation, but not willing to "risk it" to maintain the integrity of the site. Also, other than directing me to something that I've already seen, what was the point of your comment? $\endgroup$ – Asaf Karagila Nov 25 at 4:17
  • $\begingroup$ @AsafKaragila You are right, I do not want to risk my reputation to maintain the integrity of this site. I see nothing wrong with this. If this site cares for its integrity, there should be appropriate policies that would promote "correct behavior" whatever that might be. I don't want to, and I do not need to be the hero that would sacrifice points to protect the integrity of MSE. If MSE wants its users to protect its integrity, this should be implemented in a right balance of penalties and rewards, and not rely on my (or anyone else's) heroic and self-sacrificing behavior. $\endgroup$ – Mirko Nov 25 at 4:27
  • $\begingroup$ @Mirko: And how would the site maintain integrity? If downvoting costs nothing, then I can start downvoting correct content with impunity out whenever I feel like ruining someone's day. So the next step would be for you to request that downvotes to your answers won't cost reputation either, and then what? If all you want is just positive feedback I hear Facebook and Twitter are platforms where you can only "like" things. $\endgroup$ – Asaf Karagila Nov 25 at 4:28
  • $\begingroup$ @AsafKaragila I have not requested that downvotes on my answers won't cost me reputation. If and when I do that, then you could comment. Other than that "preventive action" arguments and not worth much. I don't think you or most users feel like ruining someone's day. Even if that is someone's intention (and you probably have seen enough), would it not be reasonable to allow one downvote per day without penalty? (And, to make things even, the second vote could cost two points penalty to the downvoter.) At present, mediocre answers are posted with impunity (as +rep often much outweighs -rep) $\endgroup$ – Mirko Nov 25 at 4:42
  • $\begingroup$ @Mirko: No. It would not be reasonable. You're suggesting to add a significant complication to a system that is already not too simple. What if I retract my "freebie" downvote? What if the post gets deleted? Does the next one become a "freebie" automatically? Other than you being more reluctant to downvote (which is a good thing for people who only care about reputation to be reluctant to vote), what is the harm of paying a single point per downvote? $\endgroup$ – Asaf Karagila Nov 25 at 5:02
  • $\begingroup$ @AsafKaragila If it is not reasonable, then it is not reasonable, I could live with that,and at least it sounds like you are trying to answer my concerns. If a user retracts their "freebie" downvote then that user could use it on another answer (within 24 hour time limit). If the post gets deleted the next one does not become "freebie". (I am a proud owner of a "Peer-pressure" award for an answer to now deleted question :) It is my one point and it would not harm you if I pay it. (Don't you laugh,you 300+K :) I do not want to be judged I care about what,and act based on what motivation $\endgroup$ – Mirko Nov 25 at 5:22
21
$\begingroup$

Moderators are not referees nor editors of the site. I personally have never deleted an answer because of incorrectness, and I think this should in fact never happen.

I imagine everyone agrees that mods should not unilaterally delete answers because they are wrong—why should they delete them when someone else tells them they are wrong? Suppose you flag an answer as incorrect expecting me to delete it because of that: of course I will have to read through the answer to see what basis you have for your opinion if I am going to act upon it. But then I have to decide myself if the answer is wrong or not, that is, if I agree with your assessment... But then a funny situation arises: mods would only be able to delete answers they find flat out wrong only if someone else flags them before!

$\endgroup$
  • 1
    $\begingroup$ "And, if so, what is the recommended course of action when one comes across an altogether wrong answer?" $\endgroup$ – user636532 Jul 26 at 11:36
18
$\begingroup$

As a non-moderator, I agree that simply being wrong is not a reason for anyone (except the author) to delete an answer.

Wrong answers should be downvoted (that is fundamentally what downvotes are for), and the wrongness possibly explained in a comment if the answer is not obviously wrong given what the other answers explain.

The only kind of answers that should be deleted (other than by their authors) are those where it is unlikely that the answerer in good faith thought he was answering the question correctly: joke answers, blatant hoaxes, not-an-answers, irrelevant rants, spam, abuse, and so forth.

$\endgroup$
  • 2
    $\begingroup$ Late-arriving one-line answers that do nothing but restate badly what has already been explained in a good answer are kind of a grey area. Personally I think downvoting is good enough for them, but there are people going around deleting them. $\endgroup$ – Henning Makholm Jul 24 '12 at 14:39
  • 3
    $\begingroup$ I would modify your first comment by saying "Simply being wrong is not a reason for anyone other than the person who posted the answer to delete an answer." $\endgroup$ – Arturo Magidin Jul 24 '12 at 15:16
  • 2
    $\begingroup$ @ArturoMagidin: Yes, of course. I remembered that qualification the second time around. $\endgroup$ – Henning Makholm Jul 24 '12 at 15:18
  • 1
    $\begingroup$ So, if someone answers "Please help me work out $\int x^2\,dx$" with a statement of the quadratic formula, and I judge that the answerer was befuddled but acting in good faith, then I should downvote, but not vote to delete? $\endgroup$ – Gerry Myerson Jul 25 '12 at 0:18
  • $\begingroup$ Of course, divining good faith from something written on the Internet can be a difficult task, most especially if the answerer fails to return to respond to any comments left on his "answer"... $\endgroup$ – J. M. is a poor mathematician Jul 25 '12 at 1:12
  • 1
    $\begingroup$ @GerryMyerson: Correct. $\endgroup$ – Henning Makholm Jul 25 '12 at 1:18
8
$\begingroup$

It is possible that wrong answers can actually, in some ways, aid one in understanding the material once the error has been pointed out. By that I mean, seeing an incorrect implementation of a theorem or method (i.e using Lagrange's theorem as if it were if and only if by assuming it's converse is true) may shine light on how one ought to use it or avoid that pitfall of thinking as long as the error is pointed out.

$\endgroup$

You must log in to answer this question.

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