What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 133 Stack Exchange communities.

TL;DR:

I believe there are still some issues with self-answers, particularly near-immediate self-answers, which are not addressed by the current system. This could be remedied by imposing an automated restriction on how soon self-answers will be displayed in the thread. Allow users to submit self-answers at any time, but do not display them until certain reasonable conditions (a relatively short timeout, or the posting of a competing answer) are met.


History

This thread covers a good deal of the concerns surrounding self-answers, and some mechanisms in place to prevent "bad" self-answers from new users and other abuse of the self-answer system.

Minimum reputation for answering your own question should be higher than what is needed to ask

As I understand it, the following restrictions apply:

  • Users with <100 rep cannot self-answer within 8 hours of posting a question.
  • Self-answers cannot be accepted within 48 hours of posting.
  • Self-answers do not provide rep from being accepted.
  • Accepted self-answers are not pinned to the top of the thread.

Problem

This does a good job of preventing users from posting answers which should probably be comments (or not posted at all), allows superior answers to rise above accepted self-answers, and prevents some other abuses of the system.

However, there is a less tangible element of self-answers which this does not address:

Self answers are a disincentive to other users who may otherwise be willing to post more accurate, detailed material stemming from broader or deeper knowledge and/or experience than the asker has.

When someone posts an answer to their own question, it presumably means that the problem which spawned the question is solved to their satisfaction. This is fine for the purposes of sharing information with the community, but it deters others from providing solutions to the question which may be more complete or accurate since there is not so much likelihood that the self-answerer will accept a competing answer.

Worse, in the case of near-immediate self-answers, this can remove the incentive for other users to gain reputation via up-votes because the asker has an unfair head-start. Unless everyone is required to view new answers to questions where they have already voted on an answer, most people probably vote on a good answer and never look at the question again. This can put (potentially) better answers at a disadvantage even when the head-start is within an hour.

Suggested Solution

Add a delay to the posting of self-answers. The delay should be the least of these:

  • Until a competing answer has been posted.
  • A reasonable amount of time to allow competing answerers a chance to post, without preventing self-answers outright. (Suggestion: 2 hours)

To better facilitate this, and avoid completely turning people off from appropriate self-answering, I suggest the delay be handled through automation more than user limitation. That is, the user should be allowed to submit the self-answer at any time. But the answer will not actually be displayed to other users until one of the above conditions have been met. This could also be applied to the existing 8-hour delay for low-rep users.

Example:

T:0 - A user finishes resolving a tough problem, just before he would have gone to SE for assistance, and decides to share his new-found knowledge with the community. He posts a question describing his problem.

T:5 minutes - The user then posts a self-answer. There has not yet been another answer posted to the thread. The user is given a notice describing the restrictions in place regarding self-answers, and the answer is submitted into the database but not displayed to other users. (Similar to how suggested edits are only displayed to the user suggesting them, until they are approved.)

T:1 hour, 15 minutes - Another user sees the unanswered question, and posts their answer. Their answer is displayed immediately, and the self-answer is now also displayed.

T:2 hours - Had the second answer not been posted, the self-answer would automatically be displayed after this point.

share|improve this question
12  
It works pretty well now. (IMO) –  user173320 Mar 27 '12 at 14:59

1 Answer 1

Self answers are a disincentive to other users who may otherwise be willing to post more accurate, detailed material stemming from broader or deeper knowledge and/or experience than the asker has.

No, they're not.

Or at least, they shouldn't be to you. If you come across a question to which you have a better answer than those that have been posted (or even accepted), then you should post that answer.

If it's actually better, it will earn upvotes, and possibly even be accepted by the original asker.

share|improve this answer
1  
I agree they shouldn't be. But, consider what one would naturally do when faced with such a situation. Would they rather spend their time composing and (if necessary) researching an answer for someone who's already found one for themselves, or would the time be better spent helping someone else still in need of a solution? –  Iszi Mar 27 '12 at 17:53
1  
@Iszi: Maybe I'm not a good one to ask. I frequently choose the first option and leave good, detailed answers to old questions or those that already have accepted answers. That person is still in need of a solution—at least, they're still in need of the correct solution. If I can provide that solution, then I'm perfectly content to do so. And if they're not in search of a solution, well then what's the problem? Just treat the question like one with someone else's answer accepted. –  Cody Gray Mar 27 '12 at 17:56
    
@Iszi How would hiding the self-answer be better in that case? If you don't want to write answers to people who have already found the answer themselves, isn't it better that you get to know that as soon as possible instead of having to wait until you've already done all the work? –  Juhana Mar 27 '12 at 17:57
1  
In a sense, immediately self-answering a question is a sort of anti-competitive practice. This helps avoid that, while still not prohibiting self-answers outright. The issue of where the answerers choose to spend there time is not something that can be fixed by the system. However, my proposal here is something that can be done to help prevent what is (I think) widely regarded as bad practice on the part of the self-answerer. –  Iszi Mar 27 '12 at 18:08
    
@Iszi: And that's precisely my point. What you're assuming is "bad practice on the part of the self-answerer" is not "widely regarded" as such. In fact, you're one of the few people who seem to think so. We strongly encourage people to ask and answer their own questions. I struggle to understand how a practice can be "anti-competitive" when you're working in a non-competitive environment. –  Cody Gray Mar 27 '12 at 18:10
1  
Perhaps it's just me, then. I don't know. I generally try to hold myself to a day or so waiting period before I post a self-answer, in case someone else can provide more details than my knowledge permits. I come here more for the interest in the knowledge of others than professing my own, so I really don't see how the site or community benefits greatly from allowing individuals to immediately self-answer a question and then just let the thread die. In the end, such activity is really not much more than blatant rep-whoring - the only real beneficiary of which is the self-answerer. –  Iszi Mar 27 '12 at 18:15
2  
@Iszi: Other people can still provide an answer, even after you've posted one. We don't block the submission of answers just because one has been provided or even accepted. I don't understand what problem you're trying to solve. Yes, the point is to gain information. Answering a question does not "let the thread die". This is not a forum. The checkmark doesn't mean "closed". –  Cody Gray Mar 27 '12 at 18:19
1  
Also, I disagree to a certain extent that this is not a competitive environment. Answers, by default, are displayed based on their acceptance and number of up-votes. It is natural that some answers will come later than others, and thereby be at a disadvantage, simply due to the schedules and availability of those providing the answers. –  Iszi Mar 27 '12 at 18:22
1  
However, an immediate self-answer does have an unfair advantage just the same as a race in which one of the racers is holding the starting gun. This can skew the resulting votes in favor of self-answers which, while perhaps accurate to some degree, do not show the complete picture. In such cases, the more appropriately valuable answers will never have the opportunity to rise above the self-answer to where they will more likely gather the attention of future users who might otherwise ignore them. –  Iszi Mar 27 '12 at 18:23

You must log in to answer this question.

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