I looked in the FAQs section and searched, but maybe I'm missing it.

I've noticed a lot of questions that have great answers that have been voted up 5-20 times for the answer, but the original questioner has never accepted an answer in the question.

Does this just remain this way or is there an automated process to eventually auto-select the answer with the most up votes?

Or does an accepted answer really not matter overall?

Added tag per request and checked community wiki.

  • 4
    this should be tagged sffaq and be a community wiki entry
    – Server Horror
    Commented Jun 11, 2009 at 13:52
  • Duplicate - meta.stackexchange.com/questions/21043/…
    – ChrisF
    Commented Feb 19, 2010 at 12:09

6 Answers 6


As folks have said, there's no process to auto-accept an answer to a "normal" question. There do exist, though, "above-normal" questions that have a reputation bounty offered on them. These questions DO get accepted automatically after 7 days if the asker does not accept an answer him/herself. The top-voted answer will receive the acceptance and thus the bounty.

  • You mean that they do get accepted automatically.
    – Graeme Perrow
    Commented Jun 11, 2009 at 14:39
  • Ooh, yes. Thanks for the catch! Editing to correct that.
    – squillman
    Commented Jun 11, 2009 at 15:05
  • 1
    Unfortunately, it seems that more of those bounty questions are the type that don't have a real "answer" but a "more or less accepted opinion"... but that's a rant for another day.
    – AnonJr
    Commented Jun 11, 2009 at 15:34

I'm not sure where we add feature requests, but it would be nice if moderators could accept an answer (or someone with a certain amount of reputation).

There are many cases were a question has been obviously answer, but the OP hasn't returned. Other times the highest voted answer is good enough for the community.

Another suggestion would be multiple accepted answers.

  • 2
    I agree Doug, some way of being able to clean up old/abandoned questions with obvious answers where either a "clean up crew" in the Review area can recommend/vote which answer is "correct" or choose something like ignore or "leave unanswered".
    – TheCleaner
    Commented Aug 7, 2013 at 20:31
  • 1
    I'd like to understand why so few return to accept.
    – ewwhite
    Commented Oct 17, 2014 at 13:07

It's up whoever posted the question to accept an answer. Answers are not automatically accepted regardless of how many up votes they get.


An answer may be very popular and technically correct without actually answering the posters actual question - i.e. it is not helpful to the poster. This can happen for a variety of reasons but the poster is ultimately responsible for marking the answer as accepted.

  • True, but the value of a question & answer after a few weeks is more for the community (and Google) than the original poster.
    – DLux
    Commented Jun 11, 2009 at 16:17
  • @DLux, the value to the community/google is not changed at all by having an answer marked as accepted.
    – Zoredache
    Commented Jun 11, 2009 at 17:59
  • I see it from time to time that answers are given as comments to the question, and the OP even "accepts" it by saying "This answered my question, thank you!" as a comment. I think those cases benefit from being able to convert the comment to an answer, and accept it
    – MichelZ
    Commented Apr 19, 2014 at 8:19

It's been proposed any number of times to allow moderators to accept an answer on behalf of a user. This has also been shot down the same number of times.

The rationale is that accepting an answer is meant to indicate which answer was most useful to the person asking the question, while community consensus as to the "best" answer is expressed through votes.

Thus it can happen that the accepted answer isn't the highest rated answer, or even is voted below zero.

It doesn't really matter much that a question doesn't have an accepted answer; so long as it has at least one upvoted answer it will not count as an unanswered question.

  • 1
    Understood, but I agree with a comment on that link you provided. meta.stackexchange.com/questions/18312/…
    – TheCleaner
    Commented Aug 7, 2013 at 20:46
  • We have about 27,000 of 156,000 questions showing as unanswered. I've spot checked some of these before; it's my opinion that most of them have an answer worthy of being upvoted, or the question just needs to be closed. Commented Aug 7, 2013 at 20:51
  • I see the rationale but would like to respectfully disagree with "It doesn't really matter much that a question doesn't have an accepted answer". The effect is that there, as far as I can see, exists no way of extracting and answering old (as in not in the moment newborns) unanswered questions which really are unanswered. To me that is a substantial value subtracter on many levels. But please tell me if I'm uninformed and a way to filter these in fact exists.
    – ErikE
    Commented Jan 29, 2014 at 20:50
  • @ErikE Are you looking for this? Commented Jan 29, 2014 at 21:15
  • @Michael Hampton: Yes, I went for the 'Unanswered' category, only to find (at first glance) a high ratio of questions which appeared actually correctly answered and often acknowledged so by the OP, only not marked as answered. As I do this in my free time, I felt the effort of finding good question candidates to answer was far too great. I am unsure if this is because of poor understanding of the site on my part, which is how I ended up here.
    – ErikE
    Commented Jan 29, 2014 at 22:29
  • 1
    @ErikE By default it shows "no upvoted answers". Those need upvotes too! Especially when they're correct. You then click on "no answers" to get questions with zero actual answers. Commented Jan 29, 2014 at 22:30
  • @MichaelHampton: Thanks, I see what you mean now and I also noticed how useful the Tags are. This should do it!
    – ErikE
    Commented Jan 29, 2014 at 22:47

I have come accross some instances of this:

  • Question has no answers
  • Question was "answered" as comments
  • Original author even states in the comment that this "answered" his question

In situations like that, I'm currently "quoting" the comments that solved the issue and posting them as "answer". This will give myself rep however, and not the original answerer, but I still believe this increases the quality of the whole post (answers should not be burried in comments!)

I would like to be able to convert a comment to an answer, and then be able to accept this on behalf of the OP, I think it would increase the quality of the site.

This could even be done using the familiar voting system we have today.

Vote for "This is the answer" -> Reason: "OP states this answer answered his question as a comment".

Also, we could only allow this after a certain period of time, e.g. a wait period of 1 Month.

  • 1
    I'm far too lazy to dig it out, somewhere there's either a blog post or a meta Q&A where Jeff Atwood (co-founder of SE) says that "stealing" an answer from comments and posting it as an answer is not only allowed, it's encouraged. Commented Apr 19, 2014 at 17:42
  • Yes, I did find that one. But I'd still like to appoint rep points to the original author... would just be fair :)
    – MichelZ
    Commented Apr 19, 2014 at 17:44
  • meta.stackexchange.com/questions/117251/…
    – user9517
    Commented Apr 19, 2014 at 18:15

You must log in to answer this question.