I was just curious about this as I currently have a question that only has one answer, and it already has a -1. Now, this question is new, so I am hopeful that somebody will respond appropriately. However, it got me thinking about this.

Per this blog post, any question with 1 answer counts towards your accept rate. But, what if that 1 answer is wrong, and has been acknowledged by the community as wrong?*

*Obviously, there would have to be a threshold. IE. you can't use 1 or else the poster could downvote every answer just to keep their accept rate higher.

  • 2
    If your accept rate truly mattered, I think you would have a point. But if you accept all correct/helpful answers you received, there is not much to worry about for the loss in accept rate for the questions you haven't accepted/couldn't accept an answer for. – Bart Apr 19 '12 at 20:08
  • 4
    100% accept rate should not really be a goal. As long as someone's accept rate isn't zero or abysmally low (indicating that the user doesn't understand or doesn't care about how accepting answers works), I doubt anyone seeing it will consider the actual percent meaningful. – Ben Lee Apr 19 '12 at 20:08
  • possible dupe: meta.stackexchange.com/questions/16634/… – forsvarir Apr 19 '12 at 20:09
  • 3
    @BenLee There are many people who bitch about a 50% accept rate. Never seen those comments? – Daniel Fischer Apr 19 '12 at 20:10
  • I don't agree that this is a duplicate. It's highly related, but not counting questions with score less or equal than 0 and not counting questions with score less of equal than -2 is a pretty big difference. – Dennis Apr 19 '12 at 20:11
  • 1
    @DanielFischer, honestly no, I've never seen that for 50% accept rates. But I believe you (maybe it depends on the tags one frequents?). I've only ever seen people bitch about zero or near-zero accept rates (and usually I flag comments when they do that -- I think it's okay to kindly explain how accepting answers works to the poster in case they just don't know, but there's no excuse to bitch about it). – Ben Lee Apr 19 '12 at 20:12
  • @DanielFischer If the term "bitching" accurately describes their behavior (and it just might) I think we can safely ignore it. ;) – Bart Apr 19 '12 at 20:14
  • @BenLee Yes, it's tag-dependent. I hardly ever see remarks about accept-rate in haskell, but quite a few in c or c++. Kindly explaining is fine, of course. – Daniel Fischer Apr 19 '12 at 20:16
  • @BenLee While I do strive for the highest possible rate, I only ask this because it seems unfair that it would be a ding against you in any capacity, whether you have 10% or 95%. – Justin Pihony Apr 19 '12 at 20:16
  • 1
    @JustinPihony Per Jeff Atwood's answer, it should be an incentive to try harder to get a good answer. I don't quite agree, but I can see his point. – Daniel Fischer Apr 19 '12 at 20:19
  • @JustinPihony, I can understand that. After all, I think that while accept rate can possibly (if low) be a red flag to other users, which my comment addresses (a decent accept rate, even if not perfect or near-perfect, is not a red flag), is it also useful as an incentivizing tool. When you accept an answer not only do you get 2 rep, you get an increased accept percentage. It's another number to strive towards. So I can understand why it would be annoying on that side -- unable to get past a certain point for a reason at least somewhat outside of your control. – Ben Lee Apr 19 '12 at 20:27
  • @DanielFischer I agree with that, and would even bounty my question right off the bat :). But, what if there is no answer? I guess the right answer would be one that points to some sort of proof that it is not possible? That is where mine is currently headed.... – Justin Pihony Apr 19 '12 at 20:28
  • 1
    @JustinPihony What? Then why ask a meta question about your accept rate? And you said "what if there is no answer..."... – agf Apr 19 '12 at 20:41
  • 1
    @SomeHelpfulCommenter You're in the same boat now. – Daniel Fischer Apr 19 '12 at 21:54
  • 2
    @DanielFischer: I only see those comments once. One flag and they're gone, thanks to some magic on the server side. I encourage you to do the same. No reason to resist the urge like some helpful commenters do. – Cody Gray Apr 20 '12 at 1:49

Questions with only negatively-voted answers can be deleted by the OP.

  • +1 and accepted! I never knew that :) – Justin Pihony Apr 19 '12 at 20:28
  • 7
    Is this really the solution that we want?! Removing potentially useful content just because it hasn't attracted a good answer yet? – jscs Apr 19 '12 at 20:34
  • 1
    @IuliusCæsar If the answer is down-voted, then is it really useful? – Justin Pihony Apr 19 '12 at 20:39
  • 3
    @Justin: The question may still be useful to future readers, because it may still get a good answer. Robert, you'd be first in line, if this were a question of duplicates, to say "dupe answer doesn't mean dupe question". How can you say "downvoted answer means you should delete the question"? That's like chopping down your rose bush because your garden doesn't have enough bees and you're worried about the Apiological Society giving you a low Buzz Rate. – jscs Apr 19 '12 at 20:48
  • @lulius: Presumably, the OP has waited a sufficient amount of time that it has become apparent a proper answer is not forthcoming. – user102937 Apr 19 '12 at 20:50
  • 1
    Wait a minute, maybe I misread this answer. Robert, are you saying that the answer can be deleted, or the question can be deleted? @IuliusCæsar thanks for pointing that out... – Justin Pihony Apr 19 '12 at 20:50
  • @JustinPihony: The question can be deleted. The OP doesn't own the answer. Answers can be speedily deleted by other means, if they are not answers at all. – user102937 Apr 19 '12 at 20:51
  • 2
    I have removed my checkmark as I agree with @IuliusCæsar . The question might get answered at some point in the future and then be helpful to future users, which is the point of the site AFAIK. So, it is better to leave it open at that point...just my two cents – Justin Pihony Apr 19 '12 at 20:53
  • 1
    I also totally misread that. This doesn't really answer Justin's question. Agree with @IuliusCæsar. – Ben Lee Apr 19 '12 at 20:57

But, what if that 1 answer is wrong, and has been acknowledged by the community as wrong?

What if it does?

If you have exactly 3 answered questions (the point when accept rate kicks in), and all of them have wrong answers, and only wrong answers... that's generally a sign that you're doing something wrong in your question.

If only one out of 3 answered questions are like this, that's still 66%, which is a perfectly legitimate accept rate. The purpose of the accept rate is not to enforce "perfect acceptance." The purpose is to identify persons who don't know what accepting means and/or use it improperly.

You shouldn't be looking for 100% acceptance. Conversely, if someone's bugging you about a 60% accept rate, flag them.

  • 2
    I was going to post something similar to your first substantive paragraph; If someone's accept rate is low because they only get bad answers regularly, I think it's time for them to consider if the questions are the root cause. – Andrew Barber Apr 20 '12 at 0:34

You must log in to answer this question.

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