What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

Today, when you accept an answer you did not provide, you get a +2 reputation, however, if you accept your own answer, you get +0.

I expect the reason for this is to stop people gaming the system by asking and answering their own questions repeatedly.

However, this has the downside of providing a disincentive of accepting your own answer when it seems the best and instead accepting an inferior one in order to get the reputation bonus.

Therefore, I propose that the accept your own answer is only worth zero when there are no answers provided by anyone other than yourself.

I feel this keeps the intent of preventing gaming the system for reputation, while removing disincentive from picking the best answer, even if it is your own.

share|improve this question
14  
Not getting 2 points can hardly be called a disincentive. –  Peter Ajtai Sep 13 '10 at 17:26
    
Ouch, -4? I don't see why questions such as this should be disincentivized so severely. Pleaese help me understand what I have done wrong. –  WilliamKF Sep 13 '10 at 18:02
11  
On the Meta site, a downvote to a feature request generally indicates disagreement with that request. Because highly voted requests are the ones that are looked at to be implemented, downvotes are a way to voice that "I don't think this is a good idea". –  Grace Note Sep 13 '10 at 18:06
    
Wow you really like the word disincentive, don't you? :) –  Kyle Rozendo Sep 13 '10 at 20:36
    
@Peter Agreed - not getting ANY points can hardly be called a disincentive. The amount I've learned from SO is shocking. –  Bryan Downing Sep 15 '10 at 5:17
    
@Bryan Downing I guess it depends upon your personality as to whether it impacts your choice of which answer to accept. –  WilliamKF Sep 18 '10 at 15:40

2 Answers 2

up vote 4 down vote accepted

IMHO, self-accepted answers are a feature that should never have been implemented, and definitely don't want to see used more often.

Normally, "acceptance" means one thing: the accepted answer was able to help the asker in some way.

But a "self-accept" checkmark on an answer has no meaning. Presumably, the user answered their own question because they found the solution on their own - but there's no reason to assume they were able to actually communicate that solution in the answer they posted after the fact.

Indeed, I've seen more than a few self-answers that amounted to little more than "the solution described in Bob's answer worked, with some minor tweaks" - indicating that they either should have accepted Bob's answer, or that Bob's answer has one or more critical flaws (the nature of which which shall now remain a mystery to future readers).

To be fair, some users do come back around and post good, detailed answers to their own questions. And that was probably the motivation behind allowing self-accept in the first place. But they don't get any special placement from it, and good answers should get up-voted anyway... so it all just seems rather pointless.

share|improve this answer
    
I don't know about never having been implemented, but I agree that they shouldn't award reputation. I say the bit about never having been implemented because one of the questions I asked over at Gaming was addressed in a TF2 patch later that same day (which I then pointed out). –  Powerlord Sep 13 '10 at 19:09
    
@Powerlord: just to clarify... I'm arguing that adorning your own posts with a green checkmark isn't useful. Posting answers to your own questions is another matter - I definitely agree that they can be useful. –  Shogging through the snow Sep 13 '10 at 19:14
    
It's not entirely pointless because self-accepts don't get placed at the top like other accepts. So if somebody does accept their own answer, if that answer is crap then it won't get any votes and will just stay at the bottom where it should be. I don't see the harm. –  Aarobot Sep 13 '10 at 19:34
    
@Aarobot - By pointless I think Shoq means that it does no harm...... nor good. ---------------------- (... it does affect acceptance rate though) –  Peter Ajtai Sep 13 '10 at 19:40
1  
It does do good, @Peter, it just doesn't always do good. I've posted a few self-accepts that I know are better than any of the other answers. Aside from the obvious issue of accept rate, it encourages people to post what actually solved the problem, so there's less of a chance that we end up with abandoned questions. Some people may misuse it, yes, but I really think those people are in the minority. –  Aarobot Sep 13 '10 at 19:45
    
@Aarobot: as Peter surmised, I just think it's meaningless - the ranking behavior you describe happens whether or not any self-accept mark was placed. I suppose it might help as a way of removing clutter from the "unanswered" list, but... so will an up-vote. If it can't get that, then what indication is there that the answer actually solves anything? –  Shogging through the snow Sep 13 '10 at 19:58
1  
People can't upvote their own answers. And I suppose you're right, a non-upvoted self-accept doesn't prove much, but the questions most likely to get (legitimate) self accepts are exactly the questions which are least likely to be getting a lot of views - otherwise, the author wouldn't have needed to solve it himself. Tumbleweed questions are, as far as I know, the primary use case for this feature. –  Aarobot Sep 13 '10 at 20:15
    
@Aarobot: right, you can't up-vote your own answers because that would also be meaningless - if you didn't think it was good, presumably you wouldn't post it. Voting and (normal) acceptance indicate that someone other than the author thought it was worthwhile. Self-acceptance just indicates... an absence of self-loathing? –  Shogging through the snow Sep 13 '10 at 20:25
1  
@Shog9 - or, it indicates that you continued the research and others didn't, and found the answer yourself. Regarding lack of upvotes: if it's in a niche, then it's entirely possible that it won't attract any votes, where other answers do simply due to initial acitvity. And I just don't get this: "...what indication is there that the answer actually solves anything?" See my self-answer - it's correct. It is the only correct answer. It shows how to solve the problem. Where's the ambiguity? –  detly Sep 14 '10 at 3:14
    
@detly: you're missing my point: of course you think it's correct - otherwise, why would you write it? And I'm not saying it isn't - but allowing the author to assert the validity of his own answer doesn't give him anything he didn't already have. Heck, you could just start the answer with "THIS ANSWER IS CORRECT" in a big, bold font and it would mean just as little as "accepting" it would. It's the difference between getting your research published in a peer-reviewed journal and just dumping it on a web page somewhere... –  Shogging through the snow Sep 14 '10 at 3:37
    
@Shog9 - oh, I'm not arguing I should get rep from it, sure. But I do think it should be elevated above the other posts. I mean, in my example there, it is correct, but for some reason I can't signify that as I would if someone else sent exactly the same set of emails that I did. The reason no-one has upvoted my answer is because I was probably the only person in the world doing exactly that thing with exactly that build of that particular compiler. Besides, if I'm not qualified to judge my own answer to be correct, why would I be qualified to judge someone else's answer? –  detly Sep 14 '10 at 4:47
    
(Love Timecube though... that page freaks me the hell out.) –  detly Sep 14 '10 at 4:48
    
@detly: to answer your last question, consider an answer you've provided to someone else's question where, although you did your best, you were unable to communicate to them the information necessary for them to solve the problem. Although the answer was clear in your own mind, you were unable to put this into words the asker could understand... It isn't enough for an answer to signify that you know the solution - if someone else can't read and understand it, then it's useless. An up-vote, and especially a normal "accept", provides the reassurance that you've communicated effectively. –  Shogging through the snow Sep 14 '10 at 5:35
    
@Shog9 - okay, I see your point. –  detly Sep 14 '10 at 5:56
    
Shog9: You of all people should know that accepted answer != correct answer. Accepted answer means that it was the answer that solved the problem, and self-acceptance is no different. The author has every right to claim that none of the other solutions were satisfactory and that he came up with his own; naturally, the system recognizes that his solution has not been peer-reviewed by de-prioritizing it in the sort order and eliminating the rep reward (including the +2 that the question author normally receives). I really don't see the problem. –  Aarobot Sep 14 '10 at 14:20

I think the main reason for not giving points for accepting your own answers, is that you are least able to judge the quality of your own answers due to personal bias (conscious or unconscious) and just being wrapped up in the whole problem emotionally.

share|improve this answer
    
I'm missing how this would allow gaming the system. How would I benefit my reputation by preferring my own answer over another under this proposal? –  WilliamKF Sep 13 '10 at 18:05
2  
If you've drafted a question that other people have answered and all you want is to game a measly 2 points of reputation, wouldn't just accepting one of those answers be a lot easier? –  Grace Note Sep 13 '10 at 18:16
2  
There are easier ways to game the system, but I definitely agree that answerers - even when answering their own questions - are ill-suited judge their own answers. –  Shogging through the snow Sep 13 '10 at 18:43
    
@Willam - edited –  Peter Ajtai Sep 13 '10 at 19:29

You must log in to answer this question.

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