From the post How should duplicate questions be handled?

Should duplicates be deleted?

... Duplicates that are word-for-word copies or that are so poorly written that they are not useful may be deleted by users with sufficient privilege.

How can one find those? Are there many of them?


One situation this is likely to happen is when the source and target (stub and master) are asked by the same OP. If someone is interested, I wrote a query in SEDE to find stubs with no answers, so if they are confirmed to be copies or poorly written questions, they can be deleted.

Self-duplicates with no answers (list of stub Qs)

Merged questions are also returned from the query, which are a special case because the stub is a locked post and does not qualify for auto-deletion.

For completeness, here are two more queries relating to self-duplicate questions.

The first one is from stub questions with answers, while the target is unanswered. Possible actions here are to remove the duplicate status and delete the target, or reverse the duplicate direction (stub becomes target, target become stub):

Self-duplicates: stub has answer, target does not (list of stub Qs)

The second one is about self-duplicates with answers both in stub and target questions. These are candidates for a merge with the possibility of subsequently deleting the stub:

Unmerged answered self-duplicates (list of stub Qs)


How can one find those?

Duplicates and plagiarism appear to be detected and reported in the SOBotics Chat, a room that is dedicated to moderating content on Stack Overflow with the use of bots.

Source code is available, as is the dump of Stack Exchange. Features not supported could be suggested or a better tool written if one is so inclined.

Are there many of them?

Not for long, usually; it's a matter of discovery and action. The tools don't require "word for word exact copies" but they do need someone to look at their output, a few tools do work without further prodding.

There are also other similar projects: SEBotics - A side expansion of the SOBotics project to the SE network.


Duplicates with many downvotes are often worth deletion. The downvotes might be because the question is not significantly different from its sibling duplicates, so any reasonable search would have found one of the others: lacks research effort. The downvotes might be because it is so poorly written that it does not provide a clear sign post to the canonical question: unclear, not useful.

As usual, read before casting a delete vote.

  • But what if it has a valid/good answer? Should it be deleted? Or merged? Will the system allow it to be deleted with delete votes? Also, I suggest you add in the answer how one could find those with a quick search in the search bar. Mayne like this: meta.stackexchange.com/…. Tks – Andre Silva Jan 13 at 10:06
  • Not really. Feature requests are sometimes perfectly written but just happen to have many people who disagree with them. That is not reason to delete. – Shadow The Princess Wizard Jan 13 at 13:20
  • @ShadowWizard, even feature-request which is a duplicate with no answers and that is (heavily) downvoted? Why? Because it can be useful for search engines (reaching the canonical answers)? And excepting Meta sites, what do you think about (heavily) donwvoted duplicates with no answers? Should/could they be deleted? – Andre Silva Jan 13 at 14:57
  • 1
    @AndreSilva well, it depends. If the feature request is worded in a similar way to the original and adds nothing new at all then yes, it might be better to delete. Otherwise, better keep it so others can find the original more easily. – Shadow The Princess Wizard Jan 13 at 15:20

You must log in to answer this question.

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