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.

Joel Spolsky makes some good points about needing to have experts engaged to make a site like this work:

The power of the Stack Exchange platform is detailed, expert answers to extremely rare, "long-tail," highly technical questions. To get expert answers, you need experts. To attract experts, you need a site where people are asking very interesting and hard questions, not the basic questions, so that it's clear that this is a PRO site, not a consumer/enthusiast site.... and remember, the pro sites WILL attract the enthusiasts, but not the other way around.

There are some very good questions on stack overflow with brilliantly detailed answers from amazing experts but there are also a subset of questions that are too hard for the majority of users to answer which are also uninteresting for experts.

I've discovered this from my own experience. I've spent the last few days trying to find answers to an NHibernate question. In the end I got an answer by digging through the NHibernate source code, coming up with a pretty poor solution, going to the NHibernate developer google groups, asking some dumb questions about my solution until I finally got enough information to show me how to fix the problem in a way that was far superior to my original attempt.

I'm now happy but my problem is that the very question I wanted an answer to had been sitting in Stack Overflow for nearly a year with just a single poor quality answer.

The people who know this stuff are already members on stack overflow, like Fabio Maulo (who helped me find a solution to my question) and Ayende, but they hadn't answered this question.

So what needs to change improve the quality of answers to questions like mine? Is Stack Overflow always doomed to have this hidden mass of questions that are uninteresting to experts and too hard for the rest of us or is there a better way?

share|improve this question

6 Answers 6

I had a hard question on Super User about the intricacies of Adobe Reader and its ever changing security. Putting a bounty on it had no effect on getting better answers at all so the bounty was completely wasted.

Part of the problem is that the Stack Exchange sites turn over so rapidly that any question which is unanswered within a couple of days is buried under a deluge of new questions. Thus any bounty placed on an older unanswered question gets lost too.

Allowing a bounty to be available for longer on older questions might give them a chance at being answered.

However it makes me sad to say that I doubt that anything will work as it seems that the average Stack Exchange user is more concerned with garnering reputation than answering questions.

share|improve this answer
    
Why should a bounty on an older question be available longer than one on a newer one? I don't understand that part. Their age doesn't factor into how long they are visible, does it? They are visible on the "featured" tab, ordered by the date of the bounty, not the age of the question –  Pëkka Dec 10 '12 at 10:50
1  
Re your bountied Superuser question... I don't disagree with you in general that most people are here for the rep, but the question you posted is very tough. IMO, that's the main reason why it didn't get an answer, rather than Stack Exchange sucking. There are probably very few people in the whole world who can answer it. –  Pëkka Dec 10 '12 at 10:52
    
@Pekka It is indeed a very hard question, thats why I placed a bounty to try and get a response but the bounty expired before anyone answered so effectively it was wasted. If I'd been able to place a bounty that lasted longer then maybe there would have been more chance for an expert to come across the question and provide an answer. There is still an underlying problem with Stack sites in that it's easier for users who want to garner reputation to scan the new questions and cherry pick the low hanging fruit (to mix metaphors). That's a problem that Stack Exchange hasn't resolved yet. –  Clara Onager Dec 10 '12 at 11:51

One solution for this might be to have questions start increasing in value (slightly? a lot?) after time goes by.

There are a lot of adjustables in a proposal like this, but I'd imagine something like a question that has a score of at least 0, that has no upvoted answers would start increasing its answer value after a month has gone by. So after a month, upvoted answers on this question would be worth 15 instead of 10, while accepted answers would be worth 30. (Just to throw example random numbers out there.)

The key to this, of course, would be that there would be a search mechanism to find questions that would qualify for this increased value, so experts who want to maximize their rep profit for their time could still do the tougher questions. More difficulty, but more rep.

share|improve this answer
    
+1, I was thinking exactly the same thing that there should be some kind of scale which would adjust rep reward with time. Few responses to answers I didn't pick: 1) "there's no way to judge if a question is easy or not" -- absolutely false. If the question has 10 answers in the first 5 min, that's an easy one. Unanswered for days - not easy. 2) "experts" should answer all questions regardless of difficulty - most experts give up on repetitive beginner problems because those never stop. However, they don't mind at all sharing difficult experiences they had once to actually help someone out –  DXM Apr 24 '11 at 4:33
1  
You should make a feature request of this. –  Asad Dec 10 '12 at 10:55

One thing that kept me away from stack overflow was that, almost every project out there has a mailing list or forum. Asking questions at stack overflow "hides" them from places where actual specialists are.

For instance if I were to ask a question about GWT, I would probably do so at the gwt's google group, since that's where gwt developers are.

share|improve this answer
2  
So, who's the developer of C++, to whom you'll address your C++ question? –  Pavel Shved Apr 23 '10 at 14:30
    
I think for smaller niches like GWT those groups will always exist, unless most or all of the specialists in a group decide that Stack Overflow (or a Stack Exchange site) is the platform where they want to share their knowledge. (By "small" I mean "smaller than C# or Java.") –  Bill the Lizard Apr 23 '10 at 14:33
    
Agreed. Maybe I just need to accept that not everyone is watching Stack Overflow to answer questions about their project and that I will need to work harder for the more difficult questions. –  MikeD Apr 23 '10 at 14:41
2  
The reason I love StackOverflow is that I've used many of those forums, and they don't always work or work fast enough. SO might not have the answer to everything, but when it does it's usually fast. –  Lance Roberts Apr 23 '10 at 17:26
    
Questions/answers on stackexchange tent to show up better on google then most mailing lists do. –  Ian Ringrose Dec 10 '12 at 10:52

I've run into this a bit. Some experts enjoy answering the lower level questions - they are very easy for them to answer without doing any research. They'll hit the harder questions as well, but their time is very limited, so sometimes those questions are left behind, even if they have seen them and could answer them with some effort.

There have been proposals in the past to weight questions differently, but at the moment there's no obvious incentive - generally because it's hard to detect whether a question is hard or not.

At the moment there's no good solution. I'm not even sure that splitting the site in two (a pro/theory/research/etc. vs 'regular', for instance) would solve the problem, but it might be the right way to handle this, especially given the volume of questions on SO.

I've argued in the past for beginner, intermediate, and advanced tags, but only beginner has gained some traction, and I agree with the reason not to include them - it's entirely subjective, and would result in a lot of edit wars that we currently don't have. Without such a system, though, all the experts can do is troll the unanswered questions list and hope there is more than 1 advanced question for every 10 that are simply bad questions.

share|improve this answer
    
I can see where you are coming from with this. In fact the question that I refer to had been given one, not very useful, answer so it wouldn't even have turned up in the unanswered questions list. Gnoupi has a point that it probably became invisible given that it had one answer however it was also given a very cursory treatment in the NHibernate jira (216.121.112.228/browse/NH-1353) which lead me to think that no-one from that project was really interested in solving the problem. –  MikeD Apr 23 '10 at 14:14

The best you can do within Stack Overflow is to put a bounty on it. Either one of the experts who already knows the answer will take notice, or one of "the rest of us" will step up our game and do the research necessary to answer it.

share|improve this answer
5  
In this case I didn't ask the question and the original questioner appeared to have given up having found another way. I guess I could have duplicated the question but that didn't seem right. Also what if getting points isn't what motives everyone? Ayende and Fabio don't have a huge rep. –  MikeD Apr 23 '10 at 13:54
    
@MikeD: No, unfortunately you can't put a bounty on someone else's answer, so that's a good point. It looks like in this case you were the one who stepped up is game and did the research. –  Bill the Lizard Apr 23 '10 at 14:28
1  
I'm afraid that my motives were purely selfish though. I only did the research because I needed to know the answer too :) –  MikeD Apr 23 '10 at 15:09
7  
@MikeD: That's the best kind of answer though! Your interest in finding the correct solution goes way beyond reputation or badges. (By the way, thanks a lot for taking the time to post the answer to SO once you found it.) –  Bill the Lizard Apr 23 '10 at 15:51
    
I tried putting a bounty on a hard SuperUser question to try and get an answer - no luck at all and wasted bounty to boot –  Clara Onager Dec 10 '12 at 9:11

From all the FAQs:

No question is too trivial or too "newbie".

I think we shouldn't bother about "quality of questions" in the manner you are describing. Any question is good to answer. No matter what Joel says about attracting "pro" users, it doesn't mean that "easy" or "uninsteresting" questions shouldn't be asked.

There is room for all kind of questions, and no one can judge what is "interesting". What is not interesting to someone will be interesting to someone else.

share|improve this answer
    
I'm not saying that the question shouldn't be asked, in fact I'm glad that is was, it's just that it wasn't answered for whatever reason despite the fact that people capable of answering it are members. –  MikeD Apr 23 '10 at 13:51
    
@MikeD - most of times, the reason is simply that the question fell from visible questions, not that someone was reluctant to answer it. And then, Bill's answer is the way to go. –  Gnoupi Apr 23 '10 at 13:57
1  
@Gnoupi Can I offer a bounty on someone else's question then? –  MikeD Apr 23 '10 at 14:00
    
@MikeD - some day, maybe: meta.stackexchange.com/questions/1409/… –  Gnoupi Apr 23 '10 at 14:14
    
@Gnoupi So in this scenario is it better to duplicate a question? –  MikeD Apr 23 '10 at 14:16
    
@MikeD - the "what to do when you want to revive a question which is not yours" is a topic which was debated several times here, I think (though I can't find right now). In some cases, if the question is badly asked, or is obviously abandoned (for example the asker never came back), I would say that it's ok to ask your own, point to this other one, explain why you do that. –  Gnoupi Apr 23 '10 at 15:15

You must log in to answer this question.

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