I participate a lot, but almost exclusively in the (HTML5) Canvas tag. Because I don't answer much else I've got a smallish reputation for a frequent visitor(12K).
I'm the top answer-er for the Canvas tag, having answered around ~10% of all questions ever asked about Canvas on StackOverflow.
I would say that reputation numbers don't necessarily measure engagement per se in S.O. very well.
I've found that if I answer a general JavaScript question, even if its an absurdly simple one, I garner more reputation in a few minutes than if I I give a detailed answer to 2-3 in depth canvas questions.
If I wanted to pad my reputation I could definitely answer more general JS questions that crop up, but I really enjoy helping people with their canvas projects, so I sorta stick to that domain and try to help with my slightly-more-unique knowledge.
Anyway, reputation numbers aren't that important to me. At the end of the day, this is the only encouragement I need:
Could not agree more. I have the same relationship to the [floating-point] tag. It's very satisfying to help out with the detailed questions that no one else is answering, instead of trying to beat the rush to answer the silly operator-precedence questions.
> I would say that reputation numbers don't necessarily measure engagement per se in S.O. very well.
I've found that if I answer a general JavaScript question, even if its an absurdly simple one, I garner more reputation in a few minutes than if I I give a detailed answer to 2-3 in depth canvas questions.
Here's why: SO points are not based on complexity or thoroughness, but based on value - value that is perceived by the community. A quick answer to an simple Javascript question may seem rudimentary to an experienced programmer, but if it has hundreds of points, the community has determined that the answer is helpful in a big way. Chances are there are many, many people that have experienced the same problem, which makes an answer to that problem - no matter how simple - incredibly valuable to the community as a whole.
Edit: This is probably obvious for SO users, but it's worth mentioning for anybody that is confused or wondering why the point system works the way it does.
Reputation points in niche tags are definitely harder won than in the popular language tags, but (as you've demonstrated) it's easier to climb to the top of a niche tag and establish yourself as an expert. The real-world reputation you gain from that is more valuable than the reputation score on the site. Great job!
The high-traffic tags are full of low-quality or frequently asked questions. The amount of duplication in e.g. the C++ tag is staggering, but there doesn't seem to be a conscious effort to enforce a no-duplicate policy of any kind. Picking out worthwhile questions is difficult with all that noise. Worse, the "unanswered" section is full of questions that have been answered, the asker just never bothered to click accept. After a few upvotes to one of the answers and enough elapsed time that should probably happen automatically. There are also a lot of questions which cannot be answered given the information provided, and the original user is evidently no longer interested in getting it answered.
As you say though, you really can help people who have difficult questions in the niche tags (I've been spending far too much time in 'iokit' and 'objective-c++'). One upvote plus your answer accepted (total +25) is all you can hope for, though.
Regarding duplicates: they should be closed as duplicates of a canonical question, but the incentives all work against that so it rarely happens.
There's a ten-month-old widely-supported proposal on meta meant to address this, but it hasn't received any comment from the Stack Exchange Inc. folks. http://meta.stackoverflow.com/q/90620/134300
* in a nutshell, people ask the same conceptual question in different ways, it's good to have all those ways around to help Googlers
We've also improved finding duplicates in the close dialog since that meta post (it's a hard problem, so it's not a perfect suggesting system), so it's hardly like we've done nothing.
Also, as written that feature request is unworkable. Incentivize closing over asking, madness. Even incentivizing over editing (+2 up to 1k rep) is harmful IMO. I suppose we could just decline that post, but what'll probably happen is it'll be status-completed when we've come up with a better solution (which will be documented in an answer).
^ Based on the most recent data dump, somewhere around a 15 - 20% of closed questions (which are about 3-4% of all new questions). That sounds about right, honestly. I'm sure some stuff is slipping through, and we could make it easier to maintain these rates; but it doesn't seem like a pandemic of duplicates.
Oops, forgot the disclaimer: Stack Exchange Inc. employee, etc. etc.
The duplicates issue is something a little more complicated than just having a canonical answer for everything and it comes down to the naming problem. People understand things in different ways, so having multiple ways of asking the same question, and different explanations of the answers can be very beneficial.
Before looking for duplicate questions, it should be easy to reduce them by improving the Search.
I believe it is very hard in SO to find if a question: adding words in may request should reduce the number of results instead of increase it. I mean "C# Dictionary performance" should be same than "+C# +Dictionary +performance".
> the "unanswered" section is full of questions that have been answered, the asker just never bothered to click accept.
Answered questions appear in Unanswered only if no answer has been voted up. If you see a question with a good answer, just vote it up and clear the clutter for the next person looking at the Unanswered section.
It seems to be just really hard. I see a ton of duplicates in the MySQL and Java tags as well. New programmers who are also new users of S.O. vastly outnumber the experienced, and they're exactly the ones who are going to ask questions that have been repeatedly answered.
Any forum that welcomes new users is going to have this problem. The same kind of thing happened on Forrst, which should probably have been a little more resistant.
> The high-traffic tags are full of low-quality or frequently asked questions. The amount of duplication in e.g. the C++ tag is staggering, but there doesn't seem to be a conscious effort to enforce a no-duplicate policy of any kind.
This is a tough one. There already is a system for dealing with duplicates, which is to vote to close as a duplicate, but requires 5 people to so vote before it will actually be closed. You could reduce that number, but there would probably be concern about abuse and mistakes if you reduced that too far. The thing is, it's fairly labor intensive to actually find duplicate questions, and ensure that they really are duplicates. In many cases, it's just easier to answer the new question, than to dig out the duplicate, vote to close, and then, if you actually want to be helpful, explain in a comment why the question is a duplicate (since beginners might not understand how the other question answers theirs).
> the asker just never bothered to click accept . . . and the original user is evidently no longer interested in getting it answered.
I find this very annoying, too. I can understand why it would happen from the point of view of the asker: he might have found the answer on some other website 5 minutes later, or he might have just decided to find a workaround. Still, I think SO should remind people to accept answers (penalize them for having low accept rates) or have the most upvoted answer selected automatically after a while. Nobody wants to write detailed answers for someone who isn't interested in the question anymore.
In theory, the system's "economics" encourage askers to accept an answer, as askers with low acceptance rates will receive fewer quality responses (and often comments reiterating the phenomenon). But there's little incentive for drive-by askers with no engagement with the community, such as it is, to choose an answer; a consequence of SO's growing popularity is that these users comprise the largest group (901K users with between 1 and 200 reputation).
There should probably be some mechanism by which moderators or other trusted users can, after a time, accept an answer for such abandoned questions. Perhaps questions thus answered would be treated slightly differently (and identified differently in the UI); the original asker could always return and override the proxy acceptance.
> * Still, I think SO should remind people to accept answers*
It does. Your accept rate is displayed under your username when you ask a question. Let it drop to low and other users will start complaining in the comments under your question, telling you to start accepting if you expect answers.
I completely agree about giving in-depth answers. There's nothing I love more than an answer that relies on either a) stuff I know or can do really well; or b) a bucketload of research. Several times I've went back to answers and edited them, even after an answer has already been chosen, because good answers are way more important to me than recognition.
Completely with you on the "thank yous" being the biggest motivating factor; I wouldn't care about the site if it weren't for those. That said, if you want to get your rep up without lowering your standards, one option for the high-traffic tags is to go for bounty questions. That way, if you give good answers, you get a reward which is commensurate with your time/expertise, and you know that the asker both appreciates the complexity of their question and is very keen to have an answer.
I've always felt that reputation is an imperfect proxy for how helpful you've been, which is the important thing in the end. That screen shot is a much more direct measure - you should be proud of it!
I wouldn't say 12K reputation is "smallish". I have just under 6K (mainly from ruby-on-rails and google-app-engine) and according to Stack Overflow that puts me in the top 6% of users overall.
I am similar to you (just over 10k rep). Often when I answer a question, it is a tougher one that others shun away from. I may put in some real work to help the person. However, because the the level of difficulty of the question is higher, the level of understanding of the voter also needs to be higher to know if he/she should vote for the answer. Of course, there are other factors like the narrowness of applicability to a more difficult question and the time taken to even read the question if it is a longer one.
It sucks we don't have good whuffie (subjective reputation). Sticking to a low-traffic tag is certainly a good way to see more challenging questions, and get more enjoyment from the site; filtering out some of the popular tags (which tend to be low-signal) from the front page also works.
I'm the top answer-er for the Canvas tag, having answered around ~10% of all questions ever asked about Canvas on StackOverflow.
I would say that reputation numbers don't necessarily measure engagement per se in S.O. very well.
I've found that if I answer a general JavaScript question, even if its an absurdly simple one, I garner more reputation in a few minutes than if I I give a detailed answer to 2-3 in depth canvas questions.
If I wanted to pad my reputation I could definitely answer more general JS questions that crop up, but I really enjoy helping people with their canvas projects, so I sorta stick to that domain and try to help with my slightly-more-unique knowledge.
Anyway, reputation numbers aren't that important to me. At the end of the day, this is the only encouragement I need:
http://i.imgur.com/POZmt.png