Recently, I have seen users posting questions and then posting answers themselves afterwards. Usually, I would not be concerned about this, as it is accepted - I've even ended up doing this once after someone gave me some starting insight in a comment.

What does concern me is that some users have done this repeatedly. For example, one user has posted answers to 4 out of his top 10 questions. As well as this, the users post their answers within a short amount of time of the question (within an hour). Altogether, this gives the impression that they simply post the question-answer pair for the purpose of posting another answer. I don't want to call anyone out, but is this practice allowed/accepted?

  • 4
    $\begingroup$ If they got upvotes for it, doesn't that mean it was beneficial? $\endgroup$ Oct 17, 2019 at 5:17
  • 1
    $\begingroup$ I remember this instance from a few years ago. Playing "jeopardy" can be difficult, and doing it has the potential to greatly annoy some people. (Possibly replace "doing it" with "doing it poorly".) $\endgroup$
    – user1729
    Oct 17, 2019 at 11:01
  • 2
    $\begingroup$ stackoverflow.com/questions/2041/… $\endgroup$
    – user624723
    Oct 17, 2019 at 13:08
  • 1
    $\begingroup$ In fact, it is possible to post the question and the answer at the same time. See this post from 2012: Recently rolled out SE Encyclopedia feature.. $\endgroup$ Oct 18, 2019 at 10:53
  • 1
    $\begingroup$ Posting an interesting question and then posting a clear or insightful answer seems to me like a great way to record one's knowledge and share it with the world. $\endgroup$
    – littleO
    Oct 21, 2019 at 23:46
  • $\begingroup$ @littleO But why not just include the poster's answer in the question? It just seems like the posters are trying to post for the purpose of getting rep. $\endgroup$ Oct 22, 2019 at 1:04
  • 3
    $\begingroup$ @automaticallyGenerated There's no need to make assumptions about their motivations. We might as well interpret their behavior generously. Posting an answer as an answer makes use of the functionality of this platform, and allows the answer to either gain or lose visibility according to the upvoting process. I think this system does a pretty good job of highlighting the most useful answers. Besides, I think any upvotes received are deserved because it's a sign that someone else found the answer to be useful. $\endgroup$
    – littleO
    Oct 22, 2019 at 1:26
  • 1
    $\begingroup$ The issue of reputation can be handled by marking the answer as community. I have done this a few times. The point is if you have something really worthwhile to share in Q&A format then do post it here. $\endgroup$
    – Paramanand Singh Mod
    Oct 22, 2019 at 15:19
  • 1
    $\begingroup$ Just to point out, you can answer your own question before you even post it. There's a checkbox below the tags field when asking a question saying, "Answer your own question – share your knowledge, Q&A-style". If you're answering your own questions about an hour after posting them, you're doing it wrong. :-) $\endgroup$ Oct 23, 2019 at 4:30
  • $\begingroup$ See this discussion here at meta.mathstackexchange. $\endgroup$ Oct 23, 2019 at 21:55

3 Answers 3


Yes. You are allowed, and even encouraged, to answer your own questions, not just on MSE but everywhere else on the SE network: see these questions on the main Meta and this page in the Help Center.

In short, the reason is because by self-answering questions, even in the case where you have asked the question intentionally to answer it yourself, you are sharing your knowledge with the world (assuming of course that the question and answer are of a high quality). This increases the repository of knowledge on the SE sites and helps the site grow as a whole by contributing high-quality content. So, as long as you ensure that everything you post is well-written, then there is no reason to refrain from posting self-answered questions.

  • 6
    $\begingroup$ I think self-answering should only be used in exceptional cases. When you want to share something surprising. A mere homework assignment will never do. $\endgroup$ Oct 17, 2019 at 8:58
  • 7
    $\begingroup$ Quoting quid Briefly and roughly: I feel in the standard scenario the burden of proof should be with those that think something is not useful. By contrast, in a self-answered scenario the burden of proof of usefullness is with the one posting. I agree with this 100 per cent. The bar to acceptable self-answering is very high. $\endgroup$ Oct 17, 2019 at 8:59
  • 1
    $\begingroup$ @Jyrki: Indeed, I agree with you, which is why I'd emphasized that the posts must be high-quality in order to be acceptable. $\endgroup$
    – YiFan
    Oct 17, 2019 at 9:56
  • $\begingroup$ @YiFan What would you consider to be "high quality"? A positive vote tally? $\endgroup$ Oct 18, 2019 at 5:20
  • 1
    $\begingroup$ @auto: this sort of thing is hard to define, of course. You should see the link provided by Jyrki to see others' opinion on the matter, but in my view a positive vote tally is certainly not enough. The sort of thing I would consider high quality is, for example, a very detailed exposition and explanation regarding a particular concept which is of broad applicability (e.g. a well-written explanation on, say, how one should intuitively think of the group isomorphism theorems), or questions containing significant surprising results (up to the judgement of those involved of course). $\endgroup$
    – YiFan
    Oct 18, 2019 at 8:40
  • $\begingroup$ (continued) To elaborate, if something is unlikely to be interesting or useful to a large audience, then that should not be shared as a self-answered question here. For example, routine homework problems which are really only interesting to someone taking the exact same course with the exact same homework problem. Also, one should always check if their self-answered question proposal has been covered by other questions before---in other words, check if what they are posting actually adds new value to the site. That's what's important in such instances, I feel. $\endgroup$
    – YiFan
    Oct 18, 2019 at 8:43
  • $\begingroup$ @YiFan Following is an example that perhaps you can analyze to help explain your point: math.stackexchange.com/q/3299721/343701 (The example is my own question which I partially answered). $\endgroup$
    – Jaspreet
    Oct 18, 2019 at 19:40

I posted an answer to the only question I have posted on MSE so far. Blue asked me to show my work and, since my solution was complete (I was looking for a more elegant solution) and the question was well-formed without the answer, I posted my work as an answer instead of editing the question. It seems to have been well-received by the community (in fact, it was my most popular answer for quite a while), and I would do it again in similar circumstances.

I also thought briefly about self-answering a few days ago. Someone asked a graph theory question, and my answer involved a lemma that bipartite graphs don't have odd cycles. I searched for that question and was broadly disappointed with the results of my search. In the end, I linked to the least disappointing result, but I was close to a Plan B of posting the lemma and a solution and then linking to that. Again, the community will downvote and flag any inappropriate use of that, but a complete, searchable, and well-labeled solution to a problem is generally in the community's best interest.

  • 4
    $\begingroup$ Note that ones motivation for posting such a "lemma" as a Q&A can be given as context for a post, which should help avoid the appearance of a gratuitous act. $\endgroup$
    – hardmath
    Oct 17, 2019 at 11:26
  • $\begingroup$ I'm fine with this - self answering a relatively low number of times. What irks me is writing a question for the purpose of answering it multiple times. $\endgroup$ Oct 18, 2019 at 5:17
  • $\begingroup$ Here is a discussion on meta related to the "lemma as separate question" scenario: Self-answered questions as lemmas for an existing question. $\endgroup$ Oct 20, 2019 at 14:39

Rather than answering my own question, I add an "update" to my question. That way, it still offers other people the possibility to contribute with a full answer, and my "updates" are usually of good quality, helping others solve my problem (usually in a more elegant way than I would do myself) and not be discouraged by the occasional apparent complexity of my question. And also, my update is visible at the very top, so people don't have to browse through lots of material to access the content. An example is the following (see here). I waited 2 days before updating, and I did it after discovering something interesting worth adding, yet my update is not a full, traditional answer, leaving other people the opportunity to build on it.

On a different note, accepting answers from other people (rather than your own) makes you appear nice and further encourage people to answer your questions. Answering yourself does the opposite.

If you have an answer to your question even before asking it, why not post your material as a question, asking "can my proof of xyz be improved?" and you list your proof in your question.


You must log in to answer this question.

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