21

This question already has an answer here:

Background

Accepted answers are useful to show which answer helped the OP the most and are likely to be helpful to others; therefore they show at the very top of the list of answers; irrespective of the sort order or the votes on the accepted answer. This is usually has a good effect; showing the OP's favourite answer followed by the communities most upvoted answer (probably two very good answers) at the top.

However this can have a slightly strange effect when the community is highly against the accepted answer and downvotes it heavily. In addition to having a (strongly believed to be) wrong answer first it also turns grey; leading to an odd visual effect of something that looks like the top answer being difficult to read and looking like a footnote.

For example all of these, they're wrong, everyone knows they're wrong and yet they sit at the very top of each of their questions. Their authors can't even correct the situation because accepted answers can't be deleted.

Reasoning

Ultimately the sort order isn't for any political reasons; it is to show you what is likely to be helpful first. The accepted checkmark is a strong indication that an answer will be useful, on the other hand a negative score is an indication that an answer is unhelpful; at some point the downvotes should win.

Proposal

When an accepted answer is sufficiently downvoted to turn grey ( <=-3 ) then it loses its accepted answer privileges to be at the top of the answers list as the community is clearly strongly against it. This avoids the mixed message of putting it at the top (saying its important) and marking it grey (saying its unimportant).

What I’m not proposing

I'm not proposing removing its check mark or in any way undoing its accepted answer status, it was accepted by the OP and we shouldn't mess with that.

Related proposals

Don't put heavily downvoted accepted answers at the top has a similar thrust but seems to be saying that an answer with downvotes, even if it has an overall positive score should not be at the top; that seems slightly mad and I'm not proposing that.

marked as duplicate by Monica Cellio, Tobias Kienzler, Lance Roberts, Robert Harvey Oct 14 '13 at 19:36

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • 2
    Why are all the questions you refer wrongly accepted? Some contain only 1 answer that is accepted and has 1 upvote. – juergen d Oct 14 '13 at 10:16
  • @hims056 See the end part of the question; "[it] has a similar thrust but seems to be saying that an answer with downvotes, even if it has an overall positive score should not be at the top; that seems slightly mad and I'm not proposing that. " – Richard Tingle Oct 14 '13 at 10:16
  • 3
    @juergend I am not saying they are wrongly accepted, i'm saying they shouldn't be at the top of the list of answers (where there is more than one answer). I have no right to say what should or should not be accepted. However, I think the sort order should always be optimised towards whats likely to be useful; the checkmark is in favour of it being useful, downvotes are against, at some point the downvotes should win – Richard Tingle Oct 14 '13 at 10:17
  • The alternative proposal is; 'stop turning accepted answers grey' but that is definitely a poor second place under the; 'stop mixing our messages' plan – Richard Tingle Oct 14 '13 at 10:20
  • You might have a point but on this question I already wonder how it should be sorted? – rene Oct 14 '13 at 10:32
  • @rene I don't know Ruby on Rails so can't really comment, but the other upvoted answers seem to have solutions that avoid the race condition described in the accepted answers comments – Richard Tingle Oct 14 '13 at 10:35
  • Yes, but the highest voted answer doesn't really answer the question. It is a comment / discussion on the other better answers. (I'm not into RoR either...) – rene Oct 14 '13 at 10:37
  • @rene I'm enclined to agree and would flag it as "Not-An-Answer" if I knew enough Rails to know for sure. I still might – Richard Tingle Oct 14 '13 at 10:38
  • I found a "better" search, hope you don't mind me editing it into your post. – ShaWiz Oct 14 '13 at 10:41
  • @ShaWizDowArd Not at all, I'm liking the "score:-50..-20" syntax, I'll have to remember that one – Richard Tingle Oct 14 '13 at 10:42
  • @RichardTingle cheers, we learn something new every day! :-) – ShaWiz Oct 14 '13 at 10:43