32

First off, I like the idea of accept rate. I'm tired of giving good answers and not having them accepted. The idea of adding another incentive to accept answers was another great idea by SO.

I have found a problem though (in my experience at least) which I'm wondering if other people would like to see tweaked.

My issue is that I have asked several questions where a few answers have been given, but none of them has answered the question or solved my problem.

The answers have no upvotes but SO still complains that I haven't accepted an answer and brings down my accept rate for not picking an incorrect answer.

My accept rate has not suffered terribly, but I've asked several difficult obscure questions just to probe the community for knowledge and those questions usually don't receive answers that merit acceptance. I don't think I should be penalized at all for NOT accepting an incorrect answer.

Maybe the question should have at least one upvoted answer before being counted?
What do you folks think?

4
  • 3
    dupe? meta.stackexchange.com/questions/16634/…
    – Troggy
    Commented Oct 27, 2009 at 16:36
  • 2
    Not a dupe but sounds like an alternative solution to the problem..
    – snicker
    Commented Oct 27, 2009 at 16:40
  • 1
    I thought it was probably a dupe, but I couldn't find an exact match or one that had positive things to say about it. Besides, I kept finding the same answers and I wanted to find new ones if possible. Commented Oct 27, 2009 at 16:52
  • 2
    I'm voting to close this question as off-topic because the concept of "acc​ept rate" has long been removed from the system. Commented Aug 8, 2018 at 6:22

7 Answers 7

34

I think there should be a "not an answer" flag just like there is an "accept as answer" flag available to the author of the original question. I've observed the same thing where all the provided answers do not actually answer the question, or they don't work. Sometimes they may have valid information, and may even get up votes, but they just are not an answer to the question.

So the author of the question can accept as answer the one correct, answer, or can flag each of the questions as not an answer, and both behaviors would count as accepting an answer. The whole point of the accept rate is to encourage question authors to followup with the people who provided the answers for them. One way of following up is to say "nope, that isn't what I am looking for, but thanks anyway!"

Also, the "not an answer" flag should not have any negative penalty attached. If the answer is actually wrong, then it can be down voted via the normal mechanism. This is purely to fine tune the answered vs. unanswered vs. accepted status of a question.

4
  • 4
    there should be a "not an answer" flag << I totally agree.
    – Shyam
    Commented Apr 13, 2010 at 22:28
  • 2
    A "kludgy workaround" tag would be pretty useful, too.
    – intuited
    Commented Jun 15, 2010 at 16:32
  • Should this not be implicit by a negative score?
    – AngryAnt
    Commented Aug 25, 2010 at 16:27
  • @AngryAnt There are times I have asked a question, it has answers with up votes, but they are wrong. Since I asked the question I am the only one who actually did the leg work to validate the answers, while others up vote because it just sounds good. The question asker should have a veto on answers, and on second thought it could actually carry a down vote penalty. Commented Aug 30, 2010 at 18:22
8

I'd just ignore it. Acceptance rate is just one datum next to many others about you (your number of upvotes and downvotes, total questions and answers posted, not to mention your reputation itself), so on its own it's not really a big deal: if your reputation is good then it shouldn't deter anyone from answering a subsequent question you ask. If someone comments on your original question asking why you haven't accepted an answer yet, it's okay to say that you haven't yet found a satisfactory solution.

Having no accepted answer on the question can be more of a draw to getting more answers later, which is what you want!

1
  • @CSharper, yeah, but as Ether said, don't pay too much attention to it. If it's over 50%, or something like that, you can consider it high enough (see Jeff's comments in the linked post). All this was quite thoroughly discussed here: blog.stackoverflow.com/2009/08/new-question-asker-features
    – Jonik
    Commented Oct 27, 2009 at 19:42
8

As Jeff points out in his answer to this question:

Bear in mind that any accept rate of 70% or higher is considered "excellent", so you don't have to sweat every single one. There are occasions where accepting isn't practical.

An accept rate of 100% should not be something you should necessarily aspire to.

2
  • 7
    True enough, but having a 70% is a C- in most grading scales and that is a fail for me. We poor perfectionists always go for 100. It's our nature. :) Commented Oct 29, 2009 at 15:33
  • @CSharperWithJava: Maybe, but in this case it is the SO community, not the question poster, who is being graded.
    – intuited
    Commented Jun 15, 2010 at 16:33
5

Maybe the question should have at least one upvoted answer before being counted?

No, that would discourage me from upvoting answers that are interesting, helpful, or in some other way deserving of the upvote. As the question asker I generally do go back to my questions, upvote all answers that are useful and not wrong, even if none of them are the answer.

The unintended consequence is worse than the "low accept rate" problem you perceive.

I have some advice on what to do if you think your accept rate is too low.

4

I don't think I should be penalized at all for NOT accepting an incorrect answer.

You are not penalized when you accept rate is below 100%. The accept rate is just a number and nothing more. If it is well above 0% there is absolutely nothing to worry about. And even folks with 0% get answers to their questions.

2

The purpose of the site is not "to probe the community for knowledge", the idea is that upon finding an answer, you could actually answer the question yourself. If not, just wait... after all, it's not about the accept rate, it's about getting help to your problem.

3
  • 3
    When I said that, I meant that I would ask a question I had no clue about and hope someone out there had the same problem or knew what to do. That's what I see as the purpose of the site, using community knowledge to answer questions you can't. Commented Oct 27, 2009 at 18:24
  • -1 - I totally disagree, I constantly "probe the community for knowledge" - which usually leads to an acceptable answer.
    – NTDLS
    Commented Feb 23, 2010 at 16:46
  • Yes, @NTDLS, but why complain then when it doesn't lead to an acceptable answer? After all, if you were "probing" while knowing the answer and the community does not, you should provide it yourself.
    – luvieere
    Commented Feb 25, 2010 at 21:31
2

My threshold for what constitutes a reasonable accept rate is 2/3. IMHO, if you are not getting a satisfactory answer to more than a third of your questions, that means either you do not know how to formulate questions or you are asking too many questions with no right answer.

When I work hard to figure out a good answer and feel like I have come up with a good answer and it is not accepted, well, I take that as a serious signal of lack of reciprocity on the poster's part. (I still hope that one day my answer to this question will be accepted).

When I see a semi-coherent post from a person with higher than 2/3 accept rate, I am inclined to try to fix up the question first. That helps the poster gain rep. On the other hand, when I see a 42% accept rate, I don't bother. If the question is relatively straightforward, I might fire off a quick answer. If it is a difficult question, I act all petty and move on to the next one.

So, no one is proposing penalizing people for not accepting an answer to one question. However, one is usually given the proverbial cold shoulder for establishing a pattern of not reciprocating. Big difference.

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