I wonder why some of the top answerers (> 1000 answers) on Stack Overflow have so few questions (< 100 questions). Has anyone ever caught them using multiple accounts?
Yes. Indeed, this is the dirty little secret of SO...
As you're well aware, the question/answer ratio for human beings is fixed at 283/7. It is physically impossible for anyone to answer more than seven questions without having first asked at least 283. Therefore, any user with a smaller ratio is undoubtedly maintaining one, if not several, shadow accounts used only to ask questions and thereby maintain their obscenely unnatural Q/A ratios.
The exception to this is Jon Skeet, who is not human.
Generally speaking I would say because they know more and need help on less. Additionally, I would anticipate that the people who can answer many questions, can do so because they've become adept at learning quickly, and therefore do not need to ask questions as frequently to learn more. And finally, it may simply be a matter of choice, that they're choosing to be involved in the site via answering rather than asking. I expect you would find a similar trend in the reverse, that people who have asked a lot of questions, have answered relatively few. Regardless, I definitely do not believe this suggests they're using a separate account to ask questions from.
I don't use any secret accounts for asking questions. One primary reason is that it's better for me rep wise to ask questions than to use a secret account. Secret accounts don't have any rep benefit :).
I ask few questions because I primarily use StackOverflow.com as a distraction tool. Whenever I'm waiting for a build, running suites, or I just need to let my mind wander I head over to stackoverflow.com.
I do actually ask lots of questions (how else do you learn?). But due to Visual Studios current place in the release cycle most of my day to day issues are dealing with my day job working on Visual Studio. Typically about strange component interactions, old and new API issues, performance etc ... These are answered quickest by emailing various teams or digging through source control and bug logs.
This is my subjective observation from answering Python questions for less than two years:
About 90 percent of the questions coming in today are simply lack of efford or lack of knowledge resulting from trying things like
Hey I gonna do webscraping (by copying this code I found elsewhere and replacing the url)
without needed base knowledge about
How to create a csv for the data I scaped by copying a tutorial code?
Even if you look for duplicates and don't "farm" reputation by answering all posed questions, you are accumulating reputation and answers fast.
I am less than two years on Stack Overflow and got ~1.5k answers with 9 questions.
I don't need a second account for asking questions:
If I have a professional question, most of the time researching it gives me what I need without me needing to post questions the umpteenth time because I was too lazy/inexperienced to get to my goal.
If we somehow could ensure that people read the documentation of the feature they are using (but not understanding) we would cut down questions by 2/3.
I'm mostly on another site than SO (I have about 5K on SO, about 180K on another site) but I think my answers will be relevant to what goes on there.
I wonder why some of the top answerers (> 1000 answers)
So this is me, I have nearly 4000 answers on the SE site I use most (more than 4K if you count the meta of that site).
have so few questions (< 100 questions).
Again, this is me, I have considerably fewer than that (17 I think -- and some of those aren't my questions so much as other people's questions that I posted in order to answer them, so I could point still other people to the answer)
Actual genuine questions asked by me because I wanted the answer are quite few -- and many of those remain unanswered.
Here's why I ask few questions: I can mostly answer my questions myself. I'm a researcher in my area of expertise, so I know how to find things out. If I don't know the answer already, usually I can figure out an answer from scratch; sometimes I have to go looking, and sometimes a little of both -- but nearly all of the time I can answer the question to a sufficiently satisfactory level.
I do sometimes have a question I can't immediately answer and sometimes prepare to post a question, but in doing the search and research (as required in [help/how-to-ask]) suitable for a good question, I figure out the answer anyway. I've aborted at least ten half-asked questions because I found or figured out an answer to it while preparing to ask. The discipline involved in preparing a question in order to be clear and specific and to present what I've tried already often makes the question much easier to answer, and since I'm pretty good at answering questions, it shouldn't be very surprising that it often suffices to just prepare to ask it.
I expect something similar will be true of many other high-rep answerers -- they'll mostly be experts in their area and will either have few questions they can't answer with a little effort, or will find that their questions are so esoteric/difficult that they don't expect they could get an answer. (As a researcher the questions I can't answer are often questions nobody has answered before; that's kind of what research is about.)
I wish I could come up with some more good questions. I do post them when I can.
Even though my posts on SO are considerably lower in amount than the site I mostly spend time on, the pattern there is similar (133 answers, 3 questions, almost all in the
r tag), and for the same reason -- when I need to I can mostly answer my own questions.
I also answer questions in forums outside the SE network and my record there is similar -- or even more skewed. I think I'd have generated over a thousand answers for every question I've asked on other forums.
Has anyone ever caught them using multiple accounts?
Having a second account is not specifically against the rules, but doing certain things with multiple accounts is (like voting for another of your accounts). It's highly inadvisable to have a second account because it's easy to inadvertently break a serious rule that way (I have many times been reading an old answer and thought - hey this is a pretty good answer - and clicked upvote to discover it was an old answer of mine that I didn't recognize (if you do it from the same account it tells you that you can't). If I instead did that from a second account it would tell me nothing but I just broke a rule and breaking that one can have very serious consequences.
In spite of the fact that I would much rather NOT have a second account, I do have one, with about 50 reputation (garnered from answers made from there). It has exactly one function, which is to let me check for sure what a low reputation user is seeing when they're having some kind of trouble with the site.
I have explicitly posted a question to meta asking that mods be able to lower their reputation temporarily so they don't need to risk screwing up with a second account but the response was underwhelming. Apparently we have to take the risk of using a second account if we want to see what a low rep user sees.
[I use other systems that let people set their privilege to lower privilege-roles (aside from retaining the ability to then raise it again) in order to see how the system looks like for them; it's very useful.]
There is no evidence of this.
A user has nothing to gain from asking questions on separate accounts from the answering ones.
Any serious gaming of the system will be automatically detected and shutdown.
Also, if a user chooses to ask questions on one account and answer them on another, there is no reason why this should be disallowed. (unless they are gaming the system)
And, think about this: If they have all of the answers to everyone else's questions, why do they need to ask their own?