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.

I would like to see a new measure added to user profiles, that of the percentage of questions for which the user accepted an answer he himself provided.

It's certainly true that every user has the right to find her own solution to a problem, to post and accept her own answer. And it's reasonable that she earns no reputation for doing so.

However, I also think every user has the right to select which questions he wants to spend his time addressing, and, more to the point, which users. Can we agree that it's a bit of a risk to spend time and effort in composing a high-quality answer for a user who is clearly new, has no account, and has no history of either asking or answering questions? Or who has a low accept rate?

In the same way, I think there is risk involved in answering a question by a user who has a penchant for answering her own questions, even after the community has given considerable time to providing advice, direction, and code.

So I propose the creation of a new measure, self-accept rate, equal to the number of self-answered questions divided by total number of questions multiplied by 100. I don't think it if necessary to post this figure with the user summary box on questions and answers. But it would fit well somewhere not inconspicuous on her summary or reputation page.

Update after the thrashing

Ok, I got it, the idea is not popular. But gentlemen, you must acknowledge a fact: reputation is the currency of StackOverflow. Like it or not, it is a prime motivator of much of the activity here. People quote it on resumes, people refer to it in technical talks, people request that it be accessible on their linked-in profiles. You can exchange it (via bounties) to get answers to your difficult questions.

I understand that accept rate was (ostensibly) created as an incentive for people to accept answers. And it's not the most popular feature. But let's face it, the incentive worked, and for a reason. People want their efforts to bear fruit. People don't want to feed a leech.

But enough is enough on this dinger of a suggestion. Now for a second question: does StackOverflow provide an API? Because I think it would be fun to implement this thing, you know, giggles.

share|improve this question
2  
Re Your update: Yep, there is an API =) Check out StackApps. As well as the API Documentation. Good luck! –  jadarnel27 Mar 8 '12 at 15:59

4 Answers 4

up vote 5 down vote accepted

I think you had some wrong assumptions here and there:

"And it's reasonable that she earns no reputation for doing so."

As far as I know, you earn no reputation from accepting your own answer, but you do earn reputation for upvotes on your own answer.

"Can we agree that it's a bit of a risk to spend time and effort in composing a high-quality answer for a user who is clearly new, has no account, and has no history of either asking or answering questions?"

It is risky, yes. But you implicitly accept that risk by answering. No-one forces you to answer, so if you do, you accept the consequences (down-votes, comments, etc).

"Or who has a low accept rate?"

A low accept rate is not necessarily proof that someone is being "malicious" (not sure of a better term here). Sometimes there is another simple reason for a user to have that low rate: None of the answers deserves to be accepted. It's my right as the asker to say if and when my question is answered. That's why no-one else, not even moderators, can accept it on my behalf.

"I think there is risk involved in answering a question by a user who has a penchant for answering her own questions, even after the community has given considerable time to providing advice, direction, and code."

Note that it is permitted and even encouraged by a badge (Self-learner) to answer your own question. You get that badge if when accepting your own answer you get at least 3 upvotes; which, in terms of SE, means that the answer did have some value.

The wrong assumption in your post is that "accepting your own answer too much is frowned upon" which is not true. Accepting your wrong own answer is indeed frowned upon. But if by chance all of my self-answers are the best, why not accept them?

share|improve this answer

Establishing another metric to "rate" the asker is contrary to what is still a basic rule on SO: The question is what's important, not the asker.

The "risk" you take when answering a question is not getting the accept mark. So what? Your answer is still there, to be upvoted and used by the community. If it is better than the OP's self-answer, it will get more votes. If it's obvious that the OP self-answered just to whore points, they will likely get downvoted.

share|improve this answer
    
The algorithm is anything but complex. I don't even think it would require an extra database field - it's two float operations on numbers which already exist. Furthermore, if it's the question and not the asker that's important, then what's the point of reputation? Let's face a simple fact - reputation is the one incentive offered for people to participate on SO. Seriously, you expect me or anybody to believe that tens of thousands of people come here to be altruistic? Some do, some don't. I'd like to find out who at a glance rather than by searching through their questions and answers. –  psoft Mar 8 '12 at 14:09
    
Rather than complex algorithm, I would say giving too much information. The percentage of self-accepted questions is not complex, but it doesn't give any information about the user accepting the answer given from another user; it would not be even enough to give the percentage of accepted answer for each weekday, or the last 5 days, to understand if a user would accept the answer given from another user. –  kiamlaluno Mar 8 '12 at 14:12
    
Wait wait wait, did you just say 'percentage of self-accepted questions' (you meant 'answers' btw) 'doesn't give any information about the user accepting the answer given from another user'? That's exactly what it gets at. –  psoft Mar 8 '12 at 14:16
    
@psoft It's self-accepted questions as in "questions with a self-accepted answer." You are requesting something that, at the end, would not assure you the user is going to accept the answer given from somebody else. I don't see the reason for giving that extra information. –  kiamlaluno Mar 8 '12 at 14:44
    
@Pekka: Isn't reputation a metric which rates askers/responders? –  psoft Mar 8 '12 at 14:45
    
@kiamlaluno: the same is true about answer-rate –  psoft Mar 8 '12 at 14:46
    
@psoft That is not a reason for adding information about self-accepted questions. It would be like saying that, because a site allows me to add my real name, my birthdate, and the place where I am, it should also allow me to add my shoe size in my profile. If you want to see if the user is probably accepting an answer given by another user, you can check the user's profile. –  kiamlaluno Mar 8 '12 at 14:52
    
@kiamlaluno: which is what I do and will continue to do. –  psoft Mar 8 '12 at 14:57

Why would you refrain from answering something due to the possibility of your answer not being accepted? Don't you run that risk when answering any question? I'm not a big fan of showing the accept rate that we do currently, but I see why some people find it valuable. I don't answer questions nearly as much as I used to, so I'm not really in a position to talk about it.

I'm strongly against adding yet another metric that deters people from contributing knowledge. Yes, we made a bit of a game out of it. However, the knowledge, not the game is what's paramount.

As we clearly state that asking and answering your own questions is perfectly acceptable behavior, I don't see why we would want to shine a negative light on it.

share|improve this answer
    
It's not just that possibility, there are other reasons to avoid certain individuals and questions. For example, if you have nothing to contribute to a discussion, why would you step in? By the same token, if you think a person is just leeching, why would you feed him? –  psoft Mar 8 '12 at 14:48
2  
How is someone asking and answering their own questions leeching? They don't get anything from it. I have mixed feelings about the current accept rate percentage, which I've voiced (I assume you were speaking to that). But the idea of adding an indicator that would cast positive behavior in a negative light just doesn't sit well with me. –  Tim Post Mar 8 '12 at 15:08
    
Your point is well-taken. Thanks for the input. –  psoft Mar 8 '12 at 15:15

And I suggest that....

  • We take all of the numbers in the profile.
  • Add them together.
  • Multiply by 42.
  • Then divide by 100.
  • And see what we get.
  • We will call it each individuals' Ultimate Answer Rate

See how silly it can become?

As soon as you start taking random numbers on a user's profile and manipulate them, everything else seems meaningless by comparison.

share|improve this answer
    
Who said anything about random? I offered an argument and supporting ideas, backed by a mathematical formula which is rather simple and not at all arbitrary. –  psoft Mar 8 '12 at 13:53
7  
My Ultimate Answer Rate is 8124.9. Is that good or bad? I'm feeling a bit insecure now. What can I do to improve the number? Please help me. –  Pëkka Mar 8 '12 at 13:55
    
@Pekka'sReputationBordello Just say it's a different one. :) –  kiamlaluno Mar 8 '12 at 13:56
    
@Pekka'sReputationBordello did you take into account all of the numbers on your profile page? (including anything not completely associated with you -- up and downvotes on answers, questions, etc)? –  amanaP lanaC A nalP A naM A Mar 8 '12 at 13:56
    
Why would divide the number by 100? –  kiamlaluno Mar 8 '12 at 13:56
    
@kiamlaluno Why not? –  amanaP lanaC A nalP A naM A Mar 8 '12 at 13:57
1  
@Pekka I have viewed your profile in order to try and improve you "Ultimate Answer Rate" (by increasing the profile view count). HTH TIA BRB THX LOL! –  jadarnel27 Mar 8 '12 at 13:58
    
ah, no, I counted only the numbers in my profile. What a relief! @jadarnel thanks so much!!!!! –  Pëkka Mar 8 '12 at 13:58
    
@Pekka'sReputationBordello: don't forget to multiply by the site revision number. That'll give you an "Ultimate Answer Vector" that's even cooler. –  Mat Mar 8 '12 at 14:00
    
@Mat Hmmmm How do I multiply by 2012.3.8.1521? That might be a little difficult... Do I carry the "."? –  amanaP lanaC A nalP A naM A Mar 8 '12 at 14:01
    
@Pekka'sReputationBordello hmmm this post seems to be detracting from my Ultimate Answer Rate. Doe the minus sign count towards it do you think? –  amanaP lanaC A nalP A naM A Mar 8 '12 at 14:18

You must log in to answer this question.

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