Genuine question stemming from Should the 'accept' icon be updated?

Assuming SE will go ahead and unpin accepted answers. Is there any value in showing the selected answer (most helpful to the asker) to the public (apart from the asker/answerer)?

I'm sitting here looking at the answer I just accepted, and imagined it was a lower score; what value does a reader gain from seeing a lower scored answer having the check mark that only means something to the asker?

The question can still be shown as having an accepted answer in the question list (one of the answers was helpful), and that's it. Why does it matter which one it was if unpinned?

  • 3
    Good question. Personally as a reader I would probably put extra consideration into the accepted answer and think about why the OP accepted it. It might still have some value. Sep 10 at 23:48

It's still the answer that OP felt was most useful at that point of time. In many cases, it's still one of the first/best answers to take a look at given the same issue. Ideally OP would have tried that actual answer for their problem and picked it because it solved their issue.

Don't forget - the problem we're trying to solve is newer solutions potentially being eclipsed by old, selected answers, not the selected answer system being useless.

Where opted in or out, the sort no longer emphasises the selected answer over the other questions. This doesn't mean "this worked for me" is a meaningless signal to other users - it just lets people also obviously see "this was the answer that was helpful to most people" ahead of it.

  • "OP felt was most useful at that point of time." Or he couldn't decide it, all looked equally good and solved the problem, and then he had to fallback to some questionable default such as who posted it first and so on.
    – Quasímodo
    Sep 11 at 1:25
  • 2
    Well, that's what unpinning the selected answer is supposed to solve amongst other things. It doesn't change the intent or potential value of ye olde tickmark - it merely ups the visibility of other answers
    – Journeyman Geek Mod
    Sep 11 at 1:29

In many cases, when one is looking for a solution to a problem and they find a question, the problem presented in the question is not necessarily the exact same problem they're trying to solve. Yet, it's close enough that the answers still apply.

Then, there are other cases where one stumbles upon a question which indicates (based on the wording, etc.) that they're in the exact same situation that the OP was in when they wrote the question (obligatory xkcd). In a subset of those cases, what the asker is trying to do might not be recommended and/or they might receive (good) answers addressing the bigger picture or the more general problem. For example, someone might ask:

How do I Foo the Bar?

And they get (again, good) answers like this:

You shouldn't really Foo the Bar; that's bad practice. Consider bla bla bla instead. [insert more explanation here]

In some situations, the OP's problem is so specific that they do really want to Foo the Bar and they don't care about anything else. So, when they receive an answer that tells them how to do that, they accept it, but it doesn't get as many upvotes as the other answer(s) (for good reasons).
Another example (specific to Stack Overflow and technical sites) is when the accepted answer only applies to an ancient version of a piece of software.

In both of the cases above, if someone finds themselves in the very same situation as the OP, they would really want to know which answer worked for them (which can mostly be inferred from the fact that the answer is accepted).

That's why unpinning the accepted answer --while leaving everything else as is-- renders the "accept" feature almost completely useless. Don't get me wrong, I'm not against unpinning it; on the contrary, I believe that the old (still current on most sites) system where the outdated answer or the answer that is too specific / suggesting something against good practice is listed as the top one is not a good system. However, there needs to be a way for users to somehow "jump" to the accepted answer if they choose to do so.


On all sites I've used, answers can get a lot of votes when the question is new, but better answers don't get too many votes when they are posted on an old question. It is extremely rare that new, great answers will find themselves above an existing answer that already has a lot of votes. This will be one of the reasons I'll check the accepted answer.

The accepted answer is sometimes also the only answer with feedback of any type on it. On Stack Overflow, I find plenty of questions where all the answers have a score of 0, no telling comments, but one answer is accepted. Other than commenting (which may get deleted), this is the only way that very low rep questioner can give feedback on an answer, and I put some trust in them because many people try an answer out before accepting it.

Some sites also have some questions with one and only one correct answer, for example [identification-request] and [story-identification] on Literature and Science Fiction and Fantasy respectively. On these two sites, the accepted answer tells people that the question won’t need another answer, and to allow the question to be closed as a duplicate (if another question has the same accepted answer).

As for the most extreme, there's Puzzling, which I assume won't want to see any changes to how accepting works. There are only a few questions there that don't have a definitive answer.

  • 2
    I don't follow the first paragraph. You'll look for the check mark; for the great late answer. But isn't the whole point of unpinning is that askers rarely update the check mark?
    – ymb1
    Sep 11 at 0:35
  • @ymb1 The last necromancer badge I remember getting is for an answer that sits at 6 vs the previously accepted answer at 15, but mine got accepted because the questioner is highly active on many SE sites.
    – Laurel
    Sep 11 at 0:49
  • Yes that happens, sadly rarely though, and that's why they're thinking of unpinning. So the common case is that the late good answer won't get a check mark, and the check mark will be on a lesser answer in that case.
    – ymb1
    Sep 11 at 0:52
  • @ymb1 I support unpinning the accepted answer (except on a site like Puzzling). But the lower scoring answer isn't always the better answer, so users should be able to take all the signals into account, especially when we've been discouraging other signals (comments) under the assumption that the checkmark would be there.
    – Laurel
    Sep 11 at 1:09

Given only one answer may be accepted and the asker uses that privilege appropriately, if

Case A: There is more than one answer, then

  • We know that

    • That answer solved the asker's problem at the time it was accepted.
  • We do not know

    • If other answers failed to solve the problem, if some of them also solved it but were not so good as the accepted answer, or if some of them solved it and were equally good.
    • If the asker ever checked back when receiving a new, late answer and considered changing the accepted answer.

Note also that some askers avoid changing the accepted answer because it will take back some points from the accepted answer.

Case B: There is no more than one answer

Trivial case.


In case A we can see that the check-mark is not useful insofar as you cannot judge whether the selected answer is of greater quality than other answers. In case B, it could signal that the (only) answer works for the asker, but who cares? Just leave a vote and I can tell by the number of votes for how many people that solution works.


It is furthermore questionable that so much weight is put on the hands of a single person. Although this will be now attenuated with the unpinning of the accepted answer, it is still a distinction.

The only drawback I can see if it is entirely removed is: Users with less than 15 points cannot vote. The solution is obvious: Always allow users to vote on answers to their own posts.

You must log in to answer this question.

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