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.

I've been doing a lot of reading on Meta about this subject and I cannot find a definitive answer or view of what people think regarding this question. So, apologies if this has already been covered/closed/down-voted, etc!

I'm finding myself hesitating over whether or not to try and answer questions on stack overflow for users with very low accept rates. I notice there are a number of people experiencing this feeling as well.

My thought is thus; should users with a very low percentage of acceptance be banned from asking new questions? Maybe 'banning' is too harsh. Perhaps the user should be nagged by one of the dropdowns at the top of the page, i.e. "Have you considered accepting any of the 46 questions that you have not accepted an answer for?".

In my opinion, this should only apply to users with a low accept rate, i.e. less than 10%, WHEN the user has over N amount of questions within a particular timespan.

But, don't get me wrong on the whole accept rate functionality. I know it's been discussed to death from other angles, but I think users who generally can't be bothered to accept and clean up their questions should be made to, or encouraged to put some effort in. I know myself how frustrating it is to find an unaccepted question that potentially contains the answer to my problem. It's always a bit of a pain having to trawl through comments to see if the author has put a message along the lines of "Thanks, it worked!".

share|improve this question
    
    

5 Answers 5

up vote 26 down vote accepted

That feeling should come naturally. People are less inclined to answer questions from people with low acceptance %. And the user should be able to see that. If he does not, then it's a loss for him.

So I would say no, just let it come naturally.

You should also never restrict a user, since he could just log out and ask as a anonymous user, or create a new account. And then you're just creating a whole other type of a problem.

share|improve this answer
5  
Agreed. So why not something along the lines of a nag feature? It's not restricting, it's just hinting? –  GenericTypeTea Nov 6 '09 at 13:04
    
There was a nag feature, when you clicked the profile and got a list of the questions, there was a text above all the unaccepted ones with a text asking if you wanted to accept an answer on that question. It's gone now for some reason. –  Ólafur Waage Nov 6 '09 at 13:06
2  
+1 I hesitate over low-acceptance-rate users as well- which probably means that these users are actually being "punished" via community. I don't think anything needs to be changed. Now, new users on the other hand... =) –  Joel Goodwin Nov 6 '09 at 13:08
    
Strange, I don't recall ever seeing this. I've been using SO for almost a year now. –  GenericTypeTea Nov 6 '09 at 13:09
    
@Joel - I suppose that makes sense - why don't you post that as an answer? –  GenericTypeTea Nov 6 '09 at 13:09
    
GenericTypeTea - look for the "XX% accept rate" against the question asker's name on an SO question. There are certain criteria (I don't recall) where it won't be shown, like new users who haven't enough rep to be Judged. –  Joel Goodwin Nov 6 '09 at 13:38
1  
Aye, I've noticed that. My gripe is those who have been using the site for 8 months and have about 40-odd unaccepted questions and very few answers. –  GenericTypeTea Nov 6 '09 at 13:44
    
Removed my answer as John Smither's comment was upvoted more than my original, and if meta-meta is generating more interest than meta, well it's all a little self-defeating. –  Joel Goodwin Nov 6 '09 at 14:00
1  
@Joel: I would agree that the user is likely being naturally punished by the community, but they may well be unaware of it; certainly, people aren't likely to post a comment saying something like "I would answer this, but you don't accept answers, so I won't bother." More likely, they just won't get as many responses, and think nothing of it. –  beska Nov 6 '09 at 14:03
1  
People just approach things differently. You for instance select an answer to accept for all your questions, which is a very good thing to do. So you either ask questions that can be answered, or select an answer that you felt helped you the most. Not everybody does that and it's a shame. –  Ólafur Waage Nov 6 '09 at 14:06
    
@beska - it is a good point, but I wonder if something more "heavy-handed" is likely to push people away. I'm not saying it will - but I like SO's reluctance to chide directly. –  Joel Goodwin Nov 6 '09 at 14:11

I don't think that users should be out right banned from asking questions unless we are very careful about where the threshold kicks in. For example, someone might have a 10% accept rate, but they also only asked 10 questions where as someone else might have asked 100 questions.

However, I do think that it might be a good idea to add a short message addressing the user saying something like "You might encounter a lower response rate until you accept some answers!" This could either be on the user page, or it could show up when they go to create a new question, similar to the existing message for subjective questions.

share|improve this answer
2  
+1 Definately a good idea. –  GenericTypeTea Nov 7 '09 at 9:22

I don't think restricting a user's ability to ask more questions is a good idea. It may be that they just aren't getting good answers, particularly if the technology they're asking about is niche.

I know that there's nags in the question list on your profile, but perhaps once a month they could be nagged with a dismissable pop-up on the main question page.

I try not to use the answer percentage statistic but, as Ólafur points out, you do get a feel for whether it's worth spending a large amount of time attempting to help someone who doesn't play by the spirit of Stack Overflow. If I see a user who has a very low acceptance rate then I sometimes nudge them with a friendly reminder in the comments under the question. Quite often that's just enough because I think there's some folks who have just genuinely forgotten or haven't quite 'got it' with the way Stack Overflow works.

share|improve this answer
    
Great point regarding the niche technology. –  Ólafur Waage Nov 6 '09 at 16:06
    
+1 for the same reason. I think prodding a user with comments is the way to move forward. –  Joel Goodwin Nov 6 '09 at 16:39

Preventing them from asking new questions is taking things a bit too far. I guess gentle reminders through comments would work. Disallowing new questions might drive them away from SO.

I normally would answer the question irrespective of the accept rate - but if the user has enough (acceptably answered) questions to his credit, I would post a comment with links to meta posts about accepting/upvoting.

share|improve this answer

Marking answers as "accepted" is a fundamentally bad feature and should be removed, so acceptance rates are worse than useless.

The original poster isn't the right person to make the call. Almost by definition they don't know the answer to the question. I'm sure any subject matter expert can find questions on stackoverflow where an answer that is wrong has been accepted.

Here's my poster child for why this is a bad feature:

http://stackoverflow.com/questions/8327810/why-should-i-make-password-as-hash-code-then-save-in-database

It's a question about why you should hash stored passwords. The accepted answer does not use the word "salt". Nor do any of the comments for that answer.

The question is completely reasonable. The person asking has no clue about the area, but has enough of a clue to formulate a decent question. But he also has no way of knowing that the accepted answer excludes something that's incredibly important. Just ask linkedin.

Here's another sort of example:

http://stackoverflow.com/questions/10408073/how-to-obtain-jchar-from-stdstring

Talks about converting from a char* to a Java string without any mention of string encoding. OK, it's sort of correct, but any sensible Java programmer would be dancing up and down waving their hands and saying "NO NO NO!!! Wait! You're hiding some really important assumptions!" But someone asking the question isn't going to know that.

Questioners do not normally have the right skills to choose the right answer.

share|improve this answer
    
Someone might not know if it's the "most correct" answer, but they know whether it solved their problem and they know if it's the answer they understood the best. That's got to be worth something. –  Mark Ransom Jun 8 '12 at 14:45
    
James, consider asking some questions here before formulating too strong of an opinion. The SO system has worked really well for a few years now. Instead of saying how flawed this process is, perhaps you should have instead asked why we do this so you can understand the reasoning, then formulate your opinion. Good luck! –  jmort253 Jun 8 '12 at 14:47
    
Don't get me wrong - I think the whole system works great. I use stackoverflow all the time. But I think sometimes the system works in spite of some of the components, not because of them. @MarkRansom, yes, absolutely. It's worth a +1 and probably a comment. But I stand by my original position: someone asking the question is in the wrong place to choose the best answer. (OK, there's an exception for an expert who asks a question knowing the answer already. That's a small minority.) –  James Moore Jun 8 '12 at 14:51
    
And @MarkRansom, your comment highlights why I think this is a bad feature: you want to give extra credit for answer that the original questioner understood the best. Unless you have a way to make sure that's also the right answer, this is the wrong thing to do. Especially for programming questions, easy-to-understand and useful answers can often be wrong, since they leave out subtle-but-important side effects. –  James Moore Jun 8 '12 at 14:57
    
Asking a question and already knowing the answer may be a minority, but I wonder how small it is. There's a checkbox on the question form that you can check to leave an answer at the same time - I don't think they would have implemented that unless it was worth their while. I had a recent opportunity to use it: stackoverflow.com/questions/10838211/… –  Mark Ransom Jun 8 '12 at 14:58
    
That's really asking for a different feature though: a way to mark an answer as authoritative. Again, not something the original questioner can usually do. It'd definitely be valuable. –  James Moore Jun 8 '12 at 15:03
    
@JamesMoore I'm pretty sure the OP knows if they solved their problem or not. Also what do you do when the community misunderstood the question and upvoted the wrong answer? Having both accepted answer and upvotes help future visitors with their problems. And that's what really important. –  Some Helpful Commenter Jun 8 '12 at 19:06
    
@SomeHelpfulCommenter we disagree here. Yes, they know what worked for them in this case. But almost by definition they lack the experience that will tell them all the implications of that answer. After seeing the comments here, I'm sticking to my original point that the person asking the question probably has less expertise than anyone answering the question, and giving the questioner an special gold star to hand out makes no sense. –  James Moore Jun 8 '12 at 21:52
    
Actually, the only reason I really care is that any sorts of rules around marking questions as answered or not encourages off-topic comments where people ask questioners to mark things as answered or not. I'll go along with anything that makes that entire category of comments vanish and stop wasting the world's time :-). –  James Moore Jun 8 '12 at 22:03
1  
@JamesMoore hmm that's an interesting view point. There's a feature that comments will be deleted with one flag if they contain the words "Accept Rate". Perhaps you should open a feature request for a similar feature on comments that include "Accept My Answer" –  Some Helpful Commenter Jun 8 '12 at 22:17
    
+1 to @SomeHelpfulCommenter. I should go find some questions that are badly marked; that would make my point better. My feeling is that they usually tend to be in specialized areas like Java classloaders and memory management, where there are many, many answers that are useful, but only in the common contexts. Really the answers have to come with lots of caveats ("this only works if you have a URLClassloader - which you do, 95% of the time, but if you're in the other 5% of the time, this will die horribly"), but the questioner who's supposed to mark the right answer has no way of knowing. –  James Moore Jun 8 '12 at 22:28
    
Just added stackoverflow.com/questions/8327810/… as a timely example of why I think this feature is bad. –  James Moore Jun 8 '12 at 22:39
    
There's always the chance that the asker can (and sometimes does) unaccept an answer in favor of a much better one, later down the road. –  Makoto Jun 8 '12 at 22:51
    
Hi James. I wanted to say that I think it's awesome you're thinking about these sorts of issues. stackoverflow.com/q/8327810/552792 is a great example that highlights your point. However, consider for a moment that the problem here is more with the question and not the answers. This op could have googled this. There are many resources on this topic already, and the op really isn't describing a specific problem. IMHO, we might want to ask if the question is constructive. In my experience, constructive questions are less likely to produce answers such as the one marked accepted. –  jmort253 Jun 9 '12 at 3:23
    
You've got a point, but certainly this question is far, far above the bottom of the barrel. I've flagged questions that I thought were much worse and had the flags rejected. –  James Moore Jun 9 '12 at 4:31

You must log in to answer this question.

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