I am more than a month here on stack exchange. I noticed that not even half, but more than half of questions even having multiple answers have none accepted.

Most of these answers should be accepted, but most of the times, the OP just abandons the question leaving helpful answers uncredited.

Wouldn't it be better if answers for questions that are few days old, or maybe few weeks old where no answer is accepted to be automatically accepted based on score? Would that work out?

If this is not good question don't hesitate to vote down, but at least explain why you think this is bad idea.

  • $\begingroup$ Acceptance is for the OP to decide whether their answer has been adequately answered; this cannot be decided by a computer. Good answers that the community likes are rewarded with upvotes instead. $\endgroup$
    – user296602
    Apr 27 '16 at 5:11
  • 1
    $\begingroup$ But what if the OP is just a one-time-ask;never-come-back user, which happens a lot. $\endgroup$
    – KKZiomek
    Apr 27 '16 at 5:16
  • 1
    $\begingroup$ I don't see how that should affect things. Acceptance is not a measure of community acceptance as upvotes are, but rather of the individual. $\endgroup$
    – user296602
    Apr 27 '16 at 5:16
  • $\begingroup$ Ok, thanks. By the way, you strangely comment under my posts one by one lately ;) $\endgroup$
    – KKZiomek
    Apr 27 '16 at 5:17
  • 1
    $\begingroup$ Since this seems to be your first post on meta.math.SE, I will mention that voting is different on meta. The downvotes on your post does not necessarily mean that the post is bad. More likely they express that the users who voted do not agree with your proposal. $\endgroup$ Apr 27 '16 at 5:29
  • $\begingroup$ @martinsleziak Thanks for telling me that. I actually didn't know that. I see you're Polish too by your nickname. $\endgroup$
    – KKZiomek
    Apr 27 '16 at 5:32
  • 5
    $\begingroup$ I will add a link to an older related feature request: Allow the community to accept answers to abandoned questions. (It was listed among related question in the sidebar on the right.) In the comments to that question you can find some links to related discussions on meta.SO. $\endgroup$ Apr 27 '16 at 5:46
  • $\begingroup$ There are motives to answer questions besides getting the Accepted green check mark. Assigning that mark is a privilege peculiar to the person who originally posted the Question, and even after an Answer gets accepted, the check mark can be withdrawn weeks or even months afterwards. Life is uncertain, and so are votes and check marks. $\endgroup$
    – hardmath
    Apr 27 '16 at 12:29
  • $\begingroup$ Just small statistics: There are 603,107 questions as of now, of which 510,667 have at least one answer, of which only 189,082 do not have an accepted answer. That is roughly 37% of the questions with an answer do not have an accepted answer, and 31% overall questions do not have an accepted answer (despite being answered). (This is true for the time stamp of this comment, is bound to change in the future.) $\endgroup$
    – Asaf Karagila Mod
    May 3 '16 at 9:08

Suppose you wrote a question asking to find where your mistake is. It might be possible that there are 2 answers - one showing your answer and other showing another very beautiful method and OP doesn't want to accept any. It is more likely that the beautiful method will get upvote and by your feature that answer will be accepted however the other answer should had been accepted as it is answering the main question.

Now the problem with your idea is that the answer not actually answering the question will also be upvoted.

Now see there might be two cases why OP isn't accepting any answer. First of all new users aren't aware of how to accept answers so whenever you see a good answer that answers the real question you can ping the OP to say

If this answer has helped you you can accept it by clicking the grey button to make it green showing that you have accepted this answer.

In most cases OP isn't aware that he/she has to accept answer too and so don't do. However if there is a user who is constantly not accepting answer even after several times telling him to accept politely. If this behaviour continues for 5-8 question you can flag any one question for moderator intervention but try to do that minimum as moderator shouldn't be disturbed for accepting answer unless it is very urgent. What you can do is invite the OP to chat and ask him why isn't he accepting the question and if suppose he says he doesn't want to or reply you in a hurtful way you can definitely flag that for abusive.

However do remember it is fully up to OP whether OP wants to accept the answer or not and he/she should not be forced to do so. "Accepting an answer is purely voluntary and there's no obligation to do so."

This might help you.


How about after a certain amount of inactivity the OP is prompted to accept or not within a few days, or to give a reason why no answer should be accepted. If they fail to do so a moderator could then check if they have a valid reason, or be free to accept an answer on their behalf.

  • $\begingroup$ Yes, let's put more work to the moderators: brainless clicking through a list of "the current answers don't satisfy me". This is just not what accepting an answer is for. There is no problem to be solved. $\endgroup$
    – Lord_Farin
    Apr 30 '16 at 11:08
  • $\begingroup$ @Lord_Farin Thanks for your kind feedback. There would be no obligation to do so, only the option to do so if a moderator wanted to. Often a moderator is involved in a debate and can clearly see a question is answered. How about instead, they could at this point vote for a "potential answer", which upon failure of the OP to indicate otherwise, would become the accepted answer if the OP doesn't object. $\endgroup$ Apr 30 '16 at 11:49
  • $\begingroup$ @Lord_Farin The problem to be solved is that an end-user searching for a solution to a certain problem will currently find many similar questions listed as "unanswered", yet which have been answered. One or more will contain the answer they are looking for, while others will not, so they will have to search through them or ask the question again. $\endgroup$ May 2 '16 at 11:23

You must log in to answer this question.

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