What is the goal of accepting an answer? When I read a question and I see an answer accepted, how can I know if it is a good answer?


Same way you know if any answer is good:

  1. Read it. Does it make sense?
  2. Look at the votes. Do other people think it's good?
  3. Read the comments. Has anyone pointed out problems with it?
  4. Try to use the solution given. Does it work?

"Accepted" just means the person who originally asked the question liked it enough to accept it. Don't put too much stock in that...

  • 1
    I vote for this. The decision should be based as much as possible on the answer itself and not incidental factors such as its author (otherwise how would new users ever get votes?). – Aarobot May 4 '10 at 21:13

You don't.

The accepted answer solves the OP problem.

Usually the top voted answer is better, sometimes the top-voted is the OP accepted but this is not always the case.


Usually people accept the answers that solve the problem or answer the question that they asked.

If it's been accepted it likely accomplished that goal. You can also tell by the number of upvotes it may have recieved and the comments that were left, if any. People on StackOverflow tend to point out incorrect information rather quickly, and downvote any answers that they know are wrong.

If in doubt, you could always try the answer out yourself, if possible.

  • We can accomplish a goal in several way. Before accepting we need maybe to let time to others to express alternatives; one answer that is accepted one hour later let will have less than half of the questions, as the other part of the world is sleeping. – Vicente Botet Escriba May 4 '10 at 21:12
  • @Vicente, you are correct, usually questions that stay unanswered will receive more answers, which may be better. However a better answer doesn't mean that the accepted answer is bad. If it still solved the problem and sufficiently answered the question, it's fine to accept it. The accepted answer can always be changed, and if the OP was interested in more answers, they could always hold off on accepting one. Personally, I usually wait a day before making a decision. – Brandon May 4 '10 at 21:42

I could think of a few indicators that an answer is likely "good":

  • Does the author have much reputation?
  • Does their profile indicate much involvement in the tags associated with the question?
  • Have they received up votes?
  • Did they discuss their answer with others in comments?
  • Is their answer well-written, and professional?
  • Does their profile suggest they commonly receive several up votes on their answers?

These items aren't required for an answer to be good, but would definitely help identify good answers.

  • I will add, was the question accepted not too early? If a question is accepted to soon, better answers that are given later could not be accepted, or can the OP change?. – Vicente Botet Escriba May 4 '10 at 21:08
  • 1
    Tsk tsk, encouraging people to upvote answers from high-rep users, some folks are going to be very upset! – Aarobot May 4 '10 at 21:10
  • 1
    Of the above items, only “Is their answer well-written, and professional?” of any real use, and that's just because that makes the answer more readable. Ultimately, the true metric of an answer is the degree to which it answers the question. Nothing else pertains (the users who tend to produce good answers do so because they're good at addressing the question, but their reputation, etc., is not a predictor). – Donal Fellows May 4 '10 at 22:45
  • @Aarobot I never encouraged people to upvote merely on grounds of high-reputation. – Sampson May 5 '10 at 2:04

There are questions for which the OP is the best or only person who can test the value of an answer: if the question involves a large unavailable codebase, exotic hardware, ...

Voters may upvote the answers that they recognize as similar to answers that they have already seen, sometimes without checking if they apply to the question at hand. That's a trait of human nature that appears in many online forum with voting.

  • Often the OP accept too soon and the voters visit the questions only while these are on the top 5 newest and don't look if the question is already answered. – Vicente Botet Escriba May 4 '10 at 21:06
  • And quite often the question falls on deaf ears and you only get rubbish answers for a long time, e.g., that don't actually address the question and instead just push some trite approach from Central Casting that utterly fails in the questioner's case. The community is not all-wise, especially for expert-level questions. – Donal Fellows May 4 '10 at 22:53

You don't. You just know the questioner approved of the answer, or he accepted it because the answer-guy is a sock puppet of the questioner. It happens a lot.

  • 2
    true, except for upvotes and comments... And as for the sock puppet thing, anyone is welcome to look at the track record of the answerer. For instance I had a look at yours :) – Carson Myers May 4 '10 at 20:48
  • How do you know how often it happens? Do you have sock puppets? – Aarobot May 4 '10 at 21:11
  • 1
    @Aarobot, if he does, he's not doing a very good job using them. He can hardly stay at more than 1 rep for a day. – Brandon May 4 '10 at 21:44
  • that comment is insulting, and mean. say you're sorry. – Evan Carroll May 4 '10 at 23:13
  • 11
    @evan I am convinced your account is performance comedy / art, like Andy Kaufman. en.wikipedia.org/wiki/Andy_Kaufman – Jeff Atwood May 5 '10 at 3:08
  • 1
    @Jeff - come to think of it, I hear "Mighty Mouse" in the background every time I read one of his posts... – user27414 May 5 '10 at 20:03

You must log in to answer this question.

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