If I look at my own reputation, 99% was gained from answering, and only 1% was gained from asking. Shouldn't this be a bit more balanced?

Asking a good question should obviously be rewarded, and indeed is when someone upvotes the question explicitly.

From my very little experience, it seems that, when I asked a good question that helped someone unrelated, they tend to upvote the answer and not the question.

What if the person who asked would get a tiny bit of reputation when an answer gets upvoted? It will obviously have to be very little reputation, like +0.5.

Say what you think, please don't downvote if you disagree.

Edit: Obviously people think it's too easy to gain rep by asking (see The surest way to gain lots of reputation on Stack Overflow- ask questions), but still people should be more encouraged to ask questions or post questions+answers immediately for questions they know how to solve.

  • 15
    If you want a higher percentage of your reputation to come from questions, you have to stop answering so many questions and start asking more. Since most of your posts are answers it's only natural that most of your reputation comes from answers. You gained about 16 points per answer opposed to 7 points per question. That's not such a gigantic difference as 99 to 1.
    – sth
    Dec 8, 2011 at 10:56
  • It would help if I had more incentive to ask I guess. Right now I usually solve myself by digging into google or fighting it myself. Even so, when I do find an answer, my first instinct isn't run to SO and post a question w/ immediate answer - and it should be!
    – talkol
    Dec 8, 2011 at 10:59
  • 3
    Note that the conditions that prompted that post no longer apply. The rep you get from question up-votes has been halved and question down-votes are free to the voter.
    – ChrisF Mod
    Dec 8, 2011 at 11:17
  • I think you need to see the rewards balance as a whole, not just the rep. Askers may gain less rep, but otoh some repeatable badges are way easier to achieve for askers. E.g. popular, notable and famous question badges rely on just viewing the post, but there's no such answerer badges that rely on just views. Dec 8, 2011 at 12:28
  • @talfluxive - don't forget that you can answer your own question, if you solve the problem yourself. So if you are finding your own answers - consider still putting the question up and also putting up your answer.
    – Taryn East
    Dec 8, 2011 at 12:40
  • 2
    are you asking about something opposite to what was done per March 2010 global rep re-calc? meta.stackexchange.com/questions/42904/… - the one when question upvotes lost half of their original weight 10->5
    – gnat
    Dec 8, 2011 at 13:04
  • 5
    You want more motivation to ask questions? Isn't getting answers for your questions enough motivation?
    – Marcelo
    Dec 8, 2011 at 13:53
  • You're missing the point. If I already found the answer myself I need motivation to type the whole thing + the answer into SO
    – talkol
    Dec 8, 2011 at 14:00
  • 1
    Your motivation is to share acquired knowledge, or are you in here just for the rep?
    – Marcelo
    Dec 8, 2011 at 14:12
  • SO rep doesn't help me too much in life, it's more the recognition that I've actually helped someone. If rep is so redundant, why not drop it in the first place?
    – talkol
    Dec 8, 2011 at 14:17
  • 8
    "Say what you think, please don't downvote if you disagree." um.. that's now how meta works Dec 8, 2011 at 15:33
  • personally I don't think downvoting is a good method to show disagreement.. a negative number next to my question makes me regret asking it. let's start a question about this issue and see how many downvotes it gets ;)
    – talkol
    Dec 8, 2011 at 19:38

6 Answers 6


The data suggest that good questions already get "rewarded" almost as well as good answers.

We can test this by comparing each question to all its answers:

  1. Do questions routinely receive a lower score than the highest-voted answer for that question?

  2. Do high scoring questions attract more answers and better answers?

Comparing "Question score versus answer score", we get this data:

Score Diff    Number of   Percent    Avg Answers 
(Ans - Qst)   Questions              per Question
-----------   ---------   -------    ------------
   -201             21      0.0%         80.81
   -101             31      0.0%         38.90
    -51            197      0.0%         24.82
    -21           1019      0.0%         13.11
    -11           3301      0.1%          7.68
     -6          11638      0.5%          4.89
     -2         131719      5.8%          2.58
     -1         239191     10.5%          1.79
      0         736656     32.3%          1.44
      1         501940     22.0%          2.06
      2         555956     24.4%          2.67
      6          76785      3.4%          3.95
     11          18183      0.8%          5.83
     21           4061      0.2%          9.40
     51            532      0.0%         16.74
    101            136      0.0%         28.58
    201             55      0.0%         58.69
-----------   ---------   -------    ------------
               2281421    100.0%          2.15

which shows that, on average, questions are scored about the same as their answers.

Only 50.7% of questions were outscored by their highest-voted answer. However, note that this table includes questions with no answers.

Answers only outscore their questions by 1.01 votes (as seen in this query) and high-voted questions attract more and better answers.

In short, the system seems to be working fairly well.
Write a high-quality question, on a subject people care about, and you will get the votes that you deserve.

  • 2
    How appropriate that this was posted on the same day as Don't be afraid to use the science.
    – Pops
    Dec 8, 2011 at 15:32
  • 1
    Good questions are rewarded by getting answers and thus not being stuck any more. I know tha'ts a bit of a cliche but it's true. Dec 8, 2011 at 23:00

Your suggestion to add a tiny bit of reputation to the questioner when an answer gets upvoted doesn't deal with the case of good answers to bad questions (situations that might lead to a "Reversal" badge). I might be inclined to downvote such a question just to make sure that there is no positive net amount of reputation for the questioner because of my answer upvote.

If you're really looking for SO-intrinsic motivations to ask good questions, remember that there are a lot of question-specific badges that have no answer counterparts:

  • Popular/Notable/Famous Question badges for questions with lots of views
  • Favorite/Stellar Question badges for question that are favourited by a lot of users

EDIT: With regard to your statement that "people should be more encouraged to [...] post questions+answers immediately for questions they know how to solve": For this case, rewards for asking questions need not be improved at all -- the reward will be the reputation from the questioner's self-answer.

  • 2
    Neither of these are about good questions, they're just about how many other people view it, generally. That makes them not very rewarding- at all. If you post the world's best C++ template metaprogramming question, you won't get any of those.
    – DeadMG
    Dec 8, 2011 at 11:33

Shouldn't this be a bit more balanced?

Getting reputation only by answering is not bad, as the most important part of an SE site are good answers.
That doesn't mean you should avoid asking questions, though. Avoiding asking questions that you resolved by thinking 10 minutes more is a thing, but avoiding to ask questions in general is another one, and your reputation is not less valid, if you ask some questions too.

Clearly, meta sites are different. On meta sites, the questions's purpose is reporting bugs, suggesting a task that should be done on the main site's tags, proposing feature requests that would improve the main site, and asking which questions should be acceptable in the main site; without those questions, the main site would not "live" and become a better place for the users, and future visitors.

As for rewards about asking good questions, who keeps voting for answers is reminded to vote for questions too (the phrase says, "Questions are important too!"). This is not a direct reward for good questions, but it's just the users who can decide when a question is good, or worth a vote; the implicit sentence is "vote questions that are worth being voted."


If I look at my own reputation, 99% was gained from answering, and only 1% was gained from asking. Shouldn't this be a bit more balanced?

You answered 15 questions (gaining 230 reputation), but only asked 1 (gaining 7 reputation). On a per-question/answer base, this means that you gain about twice as much reputation from an answer than you do from a question. That sounds about right to me.

Asking a good question should obviously be rewarded, and indeed is when someone upvotes the question explicitly

The catchphrase Questions are important too! is slightly mistaken. Questions are actually more important than answers. Without questions, there cannot be any answers to begin with!

But questions and answers do not get upvoted in the same manner nor (necessarily) by the same people.


  • While browsing for new questions to answer, I find one that deals with a language that I know, but I cannot solve the problem. I find the question interesting, so it gets an upvote from me.

  • If I find an interesting (i. e., non-trivial) question that is already answered, I will also upvote correct answers.

  • I usually upvote answers to questions I am answering myself if the other answer was faster or is better than mine. In these cases, I usually do not upvote the question (unless the question is a good one).

  • If I found a solution to a specific problem I am having, I upvote the answer that helped me, since it is the answerer that solved my problem, and the question itself, since the OP did all the work for me.

  • I always upvote helpful answers to my own questions.

From my very little experience, it seems that when I asked a good question that helped someone unrelated, they tended to upvote the answer and not the question

As I said, I do this, but only if I am really facing the same issue. I do most of my upvoting while searching for questions to answer, so I upvote an answer or a question.

As of right now, I cast 48 upvotes on questions and 135 on answers, so for every upvote I cast on a question, I upvote 2.8 answers. Considering than one question has usually more than on answer, that sounds about right to me.

What if the person who asked would get a tiny bit of reputation when an answer gets upvoted? It will obviously have to be very little reputation, like +0.5

Good answer does not imply good question!

I have seen good answers to horrible (i. e., poorly written, a little too broad and somehow trivial) questions. More than once, I actually downvoted the question while upvoting the answer.

[...] people should be more encouraged to ask questions [...]

I agree that some people upvote too few questions. I don't think I do.

If you want to encourage people to ask questions, upvote more questions!

  • "Questions are actually more important than answers." No answers are much more important. What you said is like saying "a seed is more important than a tree". Just because the existence of one depends on the other doesn't confer relative value to the world at large. Dec 8, 2011 at 15:45
  • @ConradFrix: You are considering only direct value. Questions posses high indirect value because of the answers they generate. My highest voted answer ever took 5 seconds of googling and 10 seconds of reading Wikipedia to come up with. The question was very interesting, so it made the answer interesting.
    – Dennis
    Dec 8, 2011 at 17:20
  • I'm not "only considering direct value" I'm saying you're inflating the indirect values of questions. I'm also in good company From Important Reputation Rule Changes "While we value good questions (and asking a great question is absolutely an art), we want to explicitly encourage people to provide the best possible answers. Without people interested in providing good answers, the questions are moot. We know that answers have more intrinsic value than questions, and the reputation balance should reflect that." Dec 8, 2011 at 17:30
  • @ConradFrix: Your quote talks about intrinsic value, which is just better wording for direct value. I completely agree that reputation should be based on intrinsic value. And I think my answer reflects that. But when it comes to importance, you have to look at the whole picture.
    – Dennis
    Dec 8, 2011 at 18:04

Asking a good question should obviously be rewarded, and indeed is when someone upvotes the question explicitly.

I have seen some excellent questions that deserved a bounty. Normally I'm happy with just upvoting questions, but once in a while someone will ask an out-of-the-ballpark-excellent question and I wish I had the same tools to reward them as I can excellent answers.

But I don't want to tie question votes to answer votes -- each stands or falls on their own merits.

Once in a while, someone will ask a good question about implementing a horrible idea. The question itself might be very good but generate down-votes because people dislike the underlying idea. In those cases I can be sympathetic to the idea of tying question score to answer scores, but I haven't yet seen a case so egregious that I couldn't single-handedly bring the questioner to a net-positive reputation gain by my own upvote.


It should be improved.

But rewarding might be depends on the people who viewing it. People may see it from different points of view.

Also it depends on your questions/ answers ratio.

But when you answer someone else’s question, there is a high possibility of getting rewarded by the owner if your answer has some information that might helpful. But it will not be same with the questions.

You must log in to answer this question.

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