Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 157 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

It would be nice if a question asker could somehow indicate that a bad answer is not an answer, or at least not a complete answer, so that the answer count would go back down to zero. This way, we (question readers) could distinguish between "the asker hasn't checked back yet to finalize this question" and "the asker has checked back yet but did not find any answers satisfactory."

When a question has an answer or two, even if it's not yet marked "answered," I won't read it as often as I read a question with zero answers. This post is assuming that many people exhibit the same behavior, with no evidence.

edit to add example:

If a question is 30 days old and has two bad answers, it looks the same as a question that is 30 days old, has two good answers, and has been ignored by the OP. I occasionally post in obscure topics that get few views, and I'd like a tool to distinguish my question from the latter group.

share|improve this question
Maybe we could somehow show the upvote count (of the most upvoted or all answers) together with the answer count in the question lists? – Paŭlo Ebermann Nov 23 '11 at 15:55
It might be sane to not show answers in the count if they don't have a positive score; I'm not sure. It would probably confuse people – Michael Mrozek Nov 23 '11 at 15:58
Unfortunately, the sheer mass of butthurt that would result from this would reach a point of singularity, thus pulling all of SE into its Butthurthorizon. We already have lots of problems with OPs hating on people who try to help them. Its usually over BS; they didn't like the approach of the answer, they don't care if what they want is completely wrong, etc etc. There are mechanisms already in place for the OP to indicate they have a good answer. – Won't Nov 23 '11 at 16:19
Oh, before I forget: There's a difference between "answered questions" and "questions with answers". – Time Traveling Bobby Nov 23 '11 at 16:54
The mechanism for an OP to indicate they have a good answer is fantastic. I want a mechanism for an OP to indicate that they don't have a good answer, because I feel like my questions get ignored after they have a bad answer or two. – Riley Nov 23 '11 at 17:09
Wow, I've misjudged what kind of questions meta was for. The downvotes surprise me. Sorry! – Riley Nov 23 '11 at 23:10
up vote 3 down vote accepted

In comparison with other question and answer sites, I can kind of see how a question which doesn't have an accepted answer may be misleading to a user who found it via Google and isn't otherwise familiar with how the Stack Exchange engine works.

When there's an accepted answer, the check mark serves as a flag to indicate that. But when there's no accepted answer, there's really no indication either way. A new user may not know what an "accepted answer" is until he stumbles across one.

Up-voting/down-voting is another indicator, but in my experience there are many questions and answers which have no votes at all, if for no other reason than they're just not popular enough to receive any attention. Many times when we discuss how the engine works we unintentionally think in terms of the entire community acting upon a single question and set of answers. But a lot of the content doesn't get that kind of attention.

I would be in favor of some kind of visual indicator on the page which alludes to the fact that these answers have been submitted, but the person who asked the question hasn't "accepted" one of them. Granted, this will draw more attention to the fact that a lot of drive-by questions never get accepted answers, even if they're perfectly good answers and even if they actually solved the person's problem. But that's another story entirely.

As an example, the MSDN forums have separate sections for "answers" vs. "all replies." The way they approach it isn't entirely friendly and intuitive either (I had to do a double-take the first time I was there because content was listed twice), but it's an interesting presentation of the data. I'm no UX expert by any means, but it would certainly be interesting to see some ideas from the Stack Exchange team on presenting more explicit cues about the state of the answers for a given question, even if it's nothing more than a language change on existing page elements.

share|improve this answer

A bad answer should get downvoted - that's the mechanism we use to show a bad answer is bad.

share|improve this answer
I think he's probably aware of downvotes; the important part was "so that the answer count would go back down to zero... When a question has an answer or two, even if it's not yet marked 'answered,' I won't read it as often as I read a question with zero answers." – Michael Mrozek Nov 23 '11 at 15:55
Do people really skip questions that have 1 answer? I don't; I know first-hand how bad some first answers can be. I do, however, tend to skip questions that have accepted answers. – Robert Harvey Nov 23 '11 at 15:57
A) Votes on answers are not shown on the question index. B) Not all questions get enough attention to reliably sort answers (especially after they've been answered once). C) At least one person skips questions that have 1 answer (me). D) I know how downvotes work but I think it's clear that I'm suggesting that they are insufficient. It's surprising to me that this answer would get 12 upvotes and my question would get 2 downvotes. – Riley Nov 23 '11 at 17:07

There are elements in place to facilitate handling these kinds of posts:

  • Not an answer: flag them as 'not an answer'
    • this is available through the dialog accessed from the 'flag' link under each post
  • Bad / incorrect answer: down-vote accordingly, which is accumulative
  • Incomplete answer: this is a judgement call, depending on what the value of the input is relative to the question posed. Sometimes it is worth of an up-vote, sometimes a comment probing for elaboration to an already decent input, or even merging with another nigh-complete or partial answer, making for a single authoritative post.

Ideally the colour co-ordination of the UI will be indicative of a question's state; but being (somewhat) colour-blind I can see how this could be a nondescript discriminator - though, separating the 'answer types' into dedicated views (so that you could otherwise distinguish whether or not it has +/- -voted or an accepted answer) would seem kind of absurd. Using symbol, or other separated indicator could be used, like displaying the number of answers and a tick (if answered), possibly.

share|improve this answer
Please clarify that Not an answer part, please. – Time Traveling Bobby Nov 23 '11 at 15:54
Not an Answer: answer that is really a comment, answer that is really a new question, "I have the same problem", etc. – Robert Harvey Nov 23 '11 at 16:00
@RobertHarvey: Okay, now I need to clarify what I wanted. I just wanted to see a clarification that "Not an answer" does mean "it is not an answer in a generic sense" and not "it is not the answer I was looking for". – Time Traveling Bobby Nov 23 '11 at 16:03
@Bobby In response to your comment I amended my answer, only to see Robert's take on it, and only then realised what you actually meant in terms of clarify; hopefully with both additions, it is clear now. – Grant Thomas Nov 23 '11 at 16:03

You must log in to answer this question.

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