22
$\begingroup$

We have the policy that duplicate questions should not recieve answers and instead should be closed as duplicates. Now what shall I do in the following situation:

  • A question is asked, for which I have an answer in mind, but I think is likely to be a duplicate.
  • I use the search feature to verify that it is a duplicate.
  • After a few minutes I don't find anything, but I am still not completely sure.

Considered options:

  1. I write an answer nevertheless. There is the danger that the question gets marked as a duplicate later, and some people will even downvote my answer because of that.
  2. I leave a comment, saying "This question is possibly a duplicate."
  3. I don't do anything. I wait for one day or so and see what others have done with the question. If nothing has happened, I search again for duplicates and decide again.

Maybe there are other options. What is the recommended action?

$\endgroup$
5
  • 30
    $\begingroup$ I think your question may be a duplicate, Martin, but I will wait to see what others do with it. $\endgroup$
    – Gerry Myerson
    Commented Oct 12 at 11:19
  • $\begingroup$ One possibility is to have a chat room devoted to searching for likely dupes. But it may not get much use since many users are motivated by the SE rep game and they get no rep for locating dupes but plenty of rep for posting fgtiw dupe answers. $\endgroup$
    – Bill Dubuque
    Commented Oct 12 at 22:19
  • 3
    $\begingroup$ I think ideally there should be some mechanism that the first person to identify a duplicate which gets approved as a closed reason gets some kind of rep. Then there is a literal point based incentive which would probably lead to better outcomes. $\endgroup$
    – Sidharth Ghoshal
    Commented Oct 12 at 23:55
  • 1
    $\begingroup$ @BillDubuque There already exists a chatroom related to searching - I assume it could be suitable for this purpose, too? (Although looking at that room, one can see that it is not very active.) $\endgroup$
    – Martin Sleziak
    Commented Oct 14 at 10:56
  • $\begingroup$ @SidharthGhoshal There were similar suggestions on this meta: A possible solution to reduce answering of duplicate questions and Encouraging duplicate hunters. $\endgroup$
    – Martin Sleziak
    Commented Oct 14 at 10:57

2 Answers 2

Reset to default
15
$\begingroup$

I feel that even if a potential MathSE answerer suspects that the question is a duplicate, if they first make a significant (and unsuccessful) effort to find the duplicate, then they have done their due diligence.

In such a situation, I feel that they should then post their answer.

If I was in such a situation, and subsequently, a pre-existing duplicate was identified, the first thing that I would do would be to check the answers in the pre-existing duplicate. I would then copy my answer to the pre-existing duplicate if and only if none of the pre-existing answers capture my specific analysis.

Beyond that, I would delete my answer from the newer question.

The exception to this is if the OP has already accepted my answer, which implies that I can't delete my answer. Assuming that the newer question was subsequently closed, then I would navigate to the cured chat room, explain the situation, and suggest that the (newer, duplicate, now closed) question be deleted.

Note that it is a simple matter to wait 24-48 hours, then click on your account icon, and then click on your recent answer. This way, you can check whether the question was in fact identified as a duplicate.

$\endgroup$
1
  • 2
    $\begingroup$ Sounds very reasonable to me! I like that you cover all cases. $\endgroup$
    – Martin Brandenburg
    Commented Oct 14 at 7:24
14
$\begingroup$

I have run into this situation before (at least I concretely remember this happening to me in physics.stackexchange.com).

Besides the search bar, and google, you can try a tool such as approach0.xyz to help search the q/a set a bit more efficiently

Beyond that if it's hard for you to find it as a seasoned 170k rep user its definitely going to be hard for the asker so you might as well post an answer and include within that answer what efforts you made to try to research the problem (you can view that information as downvoter repellant).

Leaving the comment "this question is possibly a duplicate" doesn't help the asker at all (unless they are better than you at searching stackexchange, which is extremely unlikely) but it might encourage other seasoned users who see the question to try their hand instead of everyone looking at the question and thinking "this looks like a duplicate, but I'm sure someone else can find it".

A more useful comment would be "this question is possibly a duplicate but I did not find a match in ___, ___, ___" and that saves other folks trying to find the duplicate some time as well as helps build public consensus quickly if this is a genuine duplicate or not.

As always, you can do nothing when you aren't sure and this is usually wise but, (and this is my opinion) if there is something even slightly useful you can do I would say it’s better to do that than nothing at all.

$\endgroup$
5
  • 7
    $\begingroup$ It seems you presume a strong correlation between high rep and searching skills / motivation. That is far from true in my experience. In fact many of the highest rep users gained so much rep because they rarely search for dupes and repeatedly post duplicate answers - even after EOQS and after the site has evolved from population mode to organization mode. $\endgroup$
    – Bill Dubuque
    Commented Oct 12 at 22:15
  • 8
    $\begingroup$ @Bill, I think all that Sidharth is presuming is that one particular high rep user (namely, Martin Brandenburg) has searching skills and motivation to find, and vote to close, duplicates. $\endgroup$
    – Gerry Myerson
    Commented Oct 12 at 23:03
  • 3
    $\begingroup$ @Gerry My comment pertains to said justification: "as a seasoned 170k rep user" - in order to help dispel common misconceptions about such. $\endgroup$
    – Bill Dubuque
    Commented Oct 12 at 23:26
  • 3
    $\begingroup$ I did presume a correlation there, perhaps its wrong. I'm not actually sure what the data says, my belief/general feeling was that most higher rep users might care about such things. I do second Gerry's comment that even if its not true in general its probably true for MB specifically since he has bothered to ask a question about it. $\endgroup$
    – Sidharth Ghoshal
    Commented Oct 12 at 23:53
  • 3
    $\begingroup$ Perhaps this FAQ post is worth mentioning: How to search on this site? And there are a few other related posts. $\endgroup$
    – Martin Sleziak
    Commented Oct 14 at 10:58

You must log in to answer this question.

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