There are one or two users that I have come across who continually and persistently post fake answers. Why do I say fake? Because they always do one of the following:

1. Give a mathematically completely wrong answer that shows without doubt that he/she knows nothing about the topic! It is not simply mistakes. One very recent example claimed that any property of the domain of a continuous function carries over to the image! Worse still, he/she explicitly said that if the domain is open, the image is too!

2. Teach others via the comments below their wrong answer! How they have the cheek to do that without knowing what they are talking about I don't know.

3. Provide a computer program that checks only from 1 to $10^7$ or something, when the question explicitly asks for a proof!

My question is what to do about it. I've tried flagging some such answers in the past but it gave mixed results. Even for reason 3, which is clearly a valid reason (does not answer the question), my flag got rejected TWICE by moderators who apparently don't read comments on flags.

Note that I have on many occasions tried to tell them what is correct and also where their logical errors are, and once in a while some others also do that, but they always instead claim that people are subduing their correct mathematics. So it appears that I cannot stop such nonsense. Why is it that we cannot flag an answer for being wrong and have maybe 5 or so users with higher reputation judge and remove it if it is really wrong? I don't get why we allow non-answers to confuse viewers. I am very tired of downvoting their answers just because they persist on pretending to know things and teaching others complete falsehood.

So this question is not the same as Don't flag wrong answers?, because I don't mind having wrong answers with honest intentions (in fact such answers with their comments could be useful for later readers to understand the mistakes made and also sharpen their understanding), but I don't see any use whatsoever for wrong answers with clearly wrong intentions.

• As far as 3. goes, numerical verification is often useful to give someone confidence that they are probably not wasting their time attempting to prove that something is true. While it is probably the case that numerical verification is more appropriate as a comment, sometimes the result is too long or complicated to give in a comment, so people leave the results in an answer. In this case, they should preface their post with a disclaimer that it is a comment, but too long or complex to be a comment. – robjohn Dec 2 '14 at 14:12
• One issue is that such answers may superficially look ok and some people probably base their flag review on a quick glance when they should actually read the post. Sometimes a glance suffices, but not always, and it's sometimes hard to tell what the flagger thinks is wrong. Perhaps we should have a way to communicate to reviewers when the post should be read in detail by someone who understands the field. – Joonas Ilmavirta Dec 2 '14 at 14:16
• @robjohn: I fully agree with having those kind of answers that you mention. However, the kind I am disgusted with are those for which the question is an elementary one (in some cases already having existing correct complete solutions!) but then this user comes along and says something along the lines of "It's very easy to write a program to solve it"... – user21820 Dec 2 '14 at 14:23
• @JoonasIlmavirta: I'd be very glad if your suggested mechanism were in place. That would be at least 1 out of 3 problems down, and the most objective one too. – user21820 Dec 2 '14 at 14:25
• In general, don't flag wrong answers using a custom flag (those for $\small\blacklozenge$-moderator attention), as only moderators see them. Voting (downvoting, voting to close, and voting to delete) are how these posts should be dealt with. Moderators are not guaranteed to know every topic and so it is better for others to vote for closure or deletion. Thus, it is best to flag these so that non-mods can see. – robjohn Dec 2 '14 at 14:34
• @robjohn: I did use "This is not an answer" flag for most of them, but stopped for (1) and (2) when I was told that presence of wrong answers is accepted by the community and they will not be removed based on incorrectness. Which is why there is nothing we can do except to downvote, which I feel is really ridiculous. Many such answers don't get viewed by people who know enough to judge the correctness, and so even get upvoted much more than people like me can downvote. – user21820 Dec 2 '14 at 14:43
• @user21820: people with enough reputation can vote to delete such answers. The reason that moderators don't usually vote to delete is that their votes are binding; that is, they delete the answer immediately. The community would rather have several people decide these things rather than a unilateral decision by a moderator. In some cases, however, a moderator will delete a question or answer unilaterally. – robjohn Dec 2 '14 at 15:49
• @robjohn: The reason I don't vote to delete such answers is not because my vote is binding, but rather that it is only reversible by another moderator, and so should be used in more extreme circumstances. That being said, the "low quality" queue is where delete votes go to die. One may consider using a chatroom (perhaps Jury Duty, or even the main chatroom) to organise deletions of this kind. This is not exactly uncommon across SE. – user642796 Dec 2 '14 at 19:31
• @gnometorule: The downvotes don't dissuade them as I've already pointed out in my last comment before yours, because more people upvote their answers than people downvote, despite most of their answers being useless. – user21820 Dec 3 '14 at 0:48
• @user21820: low-quality answers should be downvoted. There is no flag that specifies low-quality answers. If they do not attempt to answer the question, then there is a flag for that. – robjohn Dec 3 '14 at 2:09
• @NajibIdrissi Yes, but that only works if the answers get downvoted. Math users are stingy with downvotes in general (especially on answers, where they cost 1pt), and are generous with upvotes. Bad-faith semi-answers upvoted for the entertainment value is nothing unusual. – user147263 Dec 3 '14 at 17:43
• @Raff: If your hypothesis is true, it explains the phenomenon I am seeing. I'm more than glad to use up my downvotes on clearly nonsensical answers. All I have to do to get back the points is to answer one or two questions... – user21820 Dec 4 '14 at 1:59
• @Did No, I don't hold these views anymore. Some users post garbage day in day out and aren't blocked. And due to the +10/-2 rule, they even accrue reputation in general. – Najib Idrissi Jan 18 '17 at 6:30
• @Najib OK, glad to know. In my view the site has an incredibly lenient approach to day-in-day-out-garbage-posters, shared by at least a part of the mod team (but probably not by every other se site). On math.se, it seems that no amount of crap answers is enough to have any actual consequence to their author. – Did Jan 18 '17 at 6:54

In most cases, all you need to do for those is downvote the answer and leave a comment explaining what is wrong with it. Partly for the benefit of the author, partly so that others will read the comment and downvote the answer too.

One can vote to delete an answer, but such votes require 20000 reputation and are available only for answers with negative score (accepted answers are protected from this too).

In egregious cases, one may be able to summon a flash-mob of downvoters and delete voters from the chat.

As for flags, the two flags you may want to consider are Very Low Quality and Not An Answer.

• Very Low Quality is available only for posts with non-positive score. It is meant for unreadable or incomprehensible posts, where reader's reaction is not "this is wrong" but "what is this I don't even".

• Not An Answer is meant for texts that do not attempt to answer a question. E.g., something that should be posted as a comment, or as another question, or some irrelevant remark that should just be deleted.

These flags are not suitable for posts that honestly attempt to answer the question, even unsuccessfully. For those, see the first paragraph.

• Yes I agree with all your points, however the users I am referring to are not honestly attempting to answer the question. If my judgement is right, they are merely seeking attention. This would not be a problem except that they have completely no qualms in pretending to teach other, resulting only in confusion or in some cases false understanding. I am very glad to see honest attempts at an answer, and for those I very gladly give comments detailing what went wrong. But it is simply a waste of time and effort to do so for the problematic users I am referring to. – user21820 Dec 3 '14 at 2:21
• As for the summoning part, I'm not sure if that kind of behaviour is acceptable to the rest of the community, though I don't mind doing that if only to get rid of sheer nonsense quickly. – user21820 Dec 3 '14 at 2:22
• Just drop a link [Trolling/bad-faith answer](url) into the chat, and let others decide what to do. – user147263 Dec 3 '14 at 2:40
• Ah okay sure thanks! – user21820 Dec 3 '14 at 3:07
• I did it in the reopen/undelete/close/delete chat-room. Nothing happened. As I expected. No one cares. <sarcasm> Falsehoods are fine. </sarcasm> – user21820 Sep 30 '15 at 11:52