What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 127 Stack Exchange communities.

I've been using Stack Overflow sites for a few months now and, overall, I must say that the mechanisms implemented to glue users and activities to the interface and systems are nothing less than impressive.

Every now and then, still, I feel like a particular question did not receive enough attention. One of the reasons I believe is the fact that Stack Overflow (for example) had become a bit flat: people specialize in a few subjects, monitor questions, provide quick answers and move on.

There is not so much incentive to review answers or to or sort out existing answers (using the existing tools, I think).

Also, a bit too much emphasis is given to the speed of the answer, IMHO.

I believe that on average it would be very productive to have experienced and knowledgeable users not write answers, but review answers provided by others (to act like editors).

Still there is not much incentive for such work.

So, I was thinking what if there was a mechanism similar to bounties, but for review work or for voting?

Has this been considered and discussed earlier? If not, what mechanisms with the framework of existing so-like sites sounds interesting to you?

EDIT (trying to clarify a bit):

  1. I think current system is great and though I subscribe to "if it is not broken don't fix it", I don't subscribe to "if it is not broken don't question it".

  2. Here's a use case (of a very particular situation in which current system might be improved; but a use case nevertheless):

    • assume you are knowledgeable, have some reputation on Stack Overflow and have some time; would you ever go and read a question with less then 10 votes which has an accepted answer on a subject that you feel you have not much to learn on?
    • I think (please do correct me if I am wrong) that it can happen that no real expert is interested in reviewing it. I realize that normally it will happen that someone who is knowledgeable in the subject will look at the question, but I am just trying to establish that it will not be a priority
    • now it is not so hard to imagine that some percentage of such questions don't get looked at by an expert at all and end up with sub-optimal answer and maybe there is something that could be improved here
    • also, people are not so eager to read old questions

I was first thinking about a flag that would say disputed, but it is easy to imagine abuse and/or noise on such a mechanism. The second idea was a sort of reopening an interesting question with a bounty (this can, I think, be done now, but then the connection between questions is weakened). Finally, there's an idea of spending reputation (funding review with reputation), but then there is a question of bias. Therefore I've asked the question here :)

And one more note, I thank anyone who took the time to read this and even more thanks to people who commented and/or answered. Still, I think it would be beneficial to first to try to think outside of the box and then to try to put it inside the existing things.

share|improve this question
    
You want everyone to stop posting new answers and review already posted answers like some sort of clearing house of quality control? –  random Apr 19 '10 at 10:02
3  
@random: no not everyone, I'd only like to throw in a bit of incentive towards users that are inclined to do so. –  Unreason Apr 19 '10 at 10:30
1  
I reckon that until SO starts to reward editors with reputation then there will be no motivation to avoid answering questions on the basis that someone with less experience might come along and answer it in a minute. I've suggested before now that editors should be allowed to award themselves 10% of upvotes on questions that they have significantly improved - I suppose I could suggest the same on edited answers! Sheep - lamb, lamb - sheep. –  amelvin Apr 19 '10 at 16:02
    
+1 for the analysis you put in, though not for any particular one of the features that has been suggested. Accurate answers may be more important than quick answers, but the best answers are both. I think that many SO answers are both, at the moment. –  Pops Sep 27 '10 at 14:22
add comment

5 Answers

up vote 4 down vote accepted

I think the idea is fair, and taking care of unanswered questions - whether by answering, or closing them - is an important issue. I'm in favour of rep rewards or other things encouraging this.

However, there are measures aimed at old questions already. The community user grabs a number of unanswered questions, and bumps them onto the front page, on an hourly basis. Experienced (and other) users do sift through tags, and answer old questions. The way SO works, it makes no real difference how old a question is, as when you add a new answer, the question will be bumped to the front page - regardless whether it's a day old, or a year.

share|improve this answer
2  
I've suggested before now that the unanswered questions be given an automatic bounty by the community user after a certain amount of tumbleweed has blown by - this might clear up questions that aren't bad, but aren't mainstream. –  amelvin Apr 19 '10 at 15:58
add comment

Also, a bit too much emphasis is given to the speed of the answer, IMHO.

The person asking the question usually cares about the speed.

I believe that on average it would be very productive to have experienced and knowledgeable users not write answers, but review answers provided by others (to act like editors).

IMO that already happens; the "experienced and knowledgeable" users generally leave feedback in the form of comments to help add value (or point or errors) in other replies.

Or another way: there is value in providing a correct, quality answer quickly. Would you have us say:

I know the answer, but I just want somebody else to post first, so that I can pick them apart like a petty vulture; just sit back and watch the show, and when the blood stops flowing you might get an answer to your question.

Sounds a bit sadistic to me (of course, that might just be the way I phrased it),m and not especially valuable to the person with a question.

When I was "learning the ropes", I learned masses by trying to help, but also reading the answers of the more experienced users. Why can't that approach work here?

share|improve this answer
3  
+1 I know your question was rhetorical but, in fact, I think that is the most successful approach to learning on SO -- jump right in and answer what you know then learn from comments and other answers. I've probably learned more in the last year or so from my wrong answers or answers that I started to write, then thought "wait, is that really true" and had to research, than I did in reading books about stuff for years. The act of writing an answer for everyone to see cements knowledge in a way that reading never will. –  tvanfosson Apr 19 '10 at 12:22
    
Oh, don't get me wrong, I think the approach works very well, indeed. My qustion was if it can be improved further? For example there are certain areas in which the questions repeat frequently and are not merged (not always a bad thing); also there are some questions where answer is really disputable (and it would be nice if there was a way to call for attention); and then there are some question where the wrong answer gets accepted (this is bad). I know none of these are simple problems, and so is quite nicely balanced, but still there is always room for improvement. –  Unreason Apr 19 '10 at 13:29
1  
Maybe there should be an "Expert" badge fora given tag and users with enough rep gained under a certain tag can mark an answer as "verified correct" I know I've seen plenty of half-answered and just plain wrong answers marked as the accepted answer ( and given plenty of upvotes as well) –  Jim B Apr 19 '10 at 15:59
    
@Jim B: leave detailed comments explaining why an answer is wrong. I've encountered the same thing, and generally found it feasible to reverse the score simply by applying some well-reasoned criticism. And also down-votes. Don't hesitate to down-vote anything incorrect. Yes, the authors can still make out like bandits from the knee-jerk votes, but at least the answer scores themselves aren't misleading future readers... –  Shog9 Apr 19 '10 at 16:22
    
@Jim B Speaking as someone with a lot of rep, rep != expert. I'd be very unhappy if SO started setting up some elite regarding expertise. –  nb69307 Apr 19 '10 at 19:54
    
@shog9- I'm a serverfault user and the comments don't allow enough space to give enough detail to explain exactly why an answer is fundamentally wrong,and "basic" questions with the fundamentals get moved to superuser (e.g. servers don't crash on serverfault only desktops so move that BSOD...) the only other option is to just post the correct answer and hope future readers are smart enough to check up on the answer before implementation. Issues like this make me wish serverfault was for admins & superuser was for home/consumer questions rather than server only/ anything that might be desktop. –  Jim B Apr 19 '10 at 20:24
    
@neil - I agree - I'm not looking for an elite either and I certianly agree that high rep doesn't mean expertise. It would be nice to be able to judge (easily) if a poster knew what he was talking about rather than just answered a whole lot of questions- it seems to me that would provide incentive to clear out alot of the older and incorrect answers on a topic –  Jim B Apr 19 '10 at 20:38
    
@neil just a brain dump: what about self-moderation? sometimes i answer questions where i am confident of results and sometimes i am trying to learn (i do try to note it in text), but maybe it would be useful to have an option to say, i am pretty, pretty sure (in fact, i am ready to bet some reputation :) ) –  Unreason Apr 19 '10 at 23:06
add comment

There are incentives already for doing "community-positive" activities such as voting and editing: the badges, which are orthogonal to the reputation system.

Do you have any specific suggestions for new badges to be added?

share|improve this answer
    
As you say the concepts are orthogonal - I was more interested in discussing improvements in the rep system. :D –  Unreason Apr 19 '10 at 15:40
add comment

assume you are knowledgeable, have some reputation on so and have some time; would you ever go and read a question with less then 10 votes which has an accepted answer on a subject that you feel you have not much to learn on?

The subjects that I ignore are those that I'm flatly not interested in, not ones that I don't think I have much to learn about. For example, I have a lot to learn about javascript, but zero interest in doing the learning, so it's on my "ignore" list.

i think (please do correct me) that it can happen that no real expert is interested in reviewing it. i realize that normally it will happen that someone who is knowledgeable in the subject will look at the question, but I am just trying to establish that it will not be a priority

I can't speak for other tags, but it's my impression that all of the questions relating to C++ get loked at, if not by an "expert" than at least by a high-rep user. Certainly I look at most C++ questions during UK daytime.

now it is not so hard to imagine that some percentage of such questions don't get looked at by an expert at all and end up with sub-optimal answer and maybe there is something that could be improved here

Once again, it doesn't seem true for C++ - a question may get a suboptimal answer because it is a bad question, a dupe or whatever. But it's my impression that the good, worthwhile questions get good, worthwhile answers.

also, people are not so eager to read old question

I get a steady stream of rep from my own old questions and my old answers. so someone is reading them.

Bottom line - I think you have identified a non-problem, at least as far as the C++ tags is concerned.

share|improve this answer
    
Ok, thank you for the input (+1). Still, I think that among what you describe as bad, dupe and other questions that get suboptimal answers might be a high percentage of my use case and depending on the number of questions I would not call it a non-problem (near dupes are typical examples). –  Unreason Apr 19 '10 at 16:40
1  
@Unreason It would be a non-problem if people simply voted to close bad questions - most of them are beyond redemption anyway. –  nb69307 Apr 19 '10 at 16:47
    
Ok, so can we agree that there is some room for improvement and that it would be good if there was some mechanism that would motivate people to take action on certain questions? :) –  Unreason Apr 19 '10 at 18:02
add comment

Generally if a question sinks with little active there are a few reasons, mainly it's a poor question or has limited scope.

Despite this, users (myself included) still browse older/unanswered questions in case it's possible to help.

share|improve this answer
    
Yes, I agree, but don't you think that if there was a bit more incentive to review questions that overall quality of posted material might improve? (not saying that I am right, just asking for opinion) –  Unreason Apr 19 '10 at 10:38
1  
I think the incentive to help, learn and improve is enough of an incentive already. –  Finglas Apr 19 '10 at 11:01
1  
truly noble dear sire... but realistically we are only exploring algorithms here and there seems to be something in the way so sites provide answers with lower ratio of noise/signal and faster then any usenet group, and I don't think it is the difference in the moral values of users that differentiate the two, but the sum of the user interface and mechanics of the system. –  Unreason Apr 19 '10 at 13:34
add comment

You must log in to answer this question.

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