Often more than one person contributes to the community coming up with the final accepted answer to a question. However, it is only possible to accept a single answer. Restricting to one final answer does make sense for the sake of referencing that answer later, but in many cases that answer is the product of the input from several individuals. I would like to be able to decide how much of a bounty goes to different users, or some other mechanism to recognize the contribution of multiple users. Of course I could always upvote their previous answers, but this is really incorrect if their previous answers aren't the complete, final answer I've accepted.

What think you metaSOrs?


7 Answers 7


The goal isn't to have multiple answers each containing a piece of the puzzle. It's to have a single, correct, comprehensive answer. Bounty provides a means of encouraging answers to tough problems by offering a reward: allowing you to subdivide this reward creates a situation where you're misleading those participating (who think they're competing for a big prize but only receive a portion of it) and potentially discouraging users from taking the time to write a complete answer.

In short, I think it's a terrible idea. Good, useful answers will be rewarded, by the voters who find them good and useful. If you aren't comfortable with the "winner takes all" bounty system, then just don't use it - there's no requirement that you do so ever.

See also:

Why Isn’t There a Way to Consolidate Answers?

  • 11
    To award increased, non-average effort. Answers on cogsci are considered best when they provide peer-reviewed journal articles. Most questions cannot be answered by most users from what they know, almost all of us have to do some research to find the answer, and this often takes quite a lot of time. Usually when there is one answer, no-one feels they need to provide a second one, because it seems like the effort has already been done. But different research might bring up different sources and perspective, all of them equally valid. Often there is no final answer, because science is openended.
    – user215347
    Commented Sep 18, 2013 at 19:35
  • Perhaps then, @what, this system is a poor fit for the types of questions you're hoping to find answers for?
    – Shog9
    Commented Sep 18, 2013 at 19:39
  • 3
    I just had this problem just now with one of my questions where I honestly think two people have just...nailed the problem problem frankly and I would like to add bounty to both of them :(
    – Sammaye
    Commented Aug 29, 2014 at 15:49
  • So how did this guy split the bounty??? stackoverflow.com/questions/18224184/…
    – NoName
    Commented Jul 18, 2019 at 0:56
  • More than one person offered numerous bounties, @noname: stackoverflow.com/posts/18224184/revisions
    – Shog9
    Commented Jul 18, 2019 at 4:16
  • 2
    -1 for the same reasons as user215347, especially when it comes to language responses. The wording of the question asked may conceal/hide the origin of the problem, and the solution may be the result of several answers.
    – cl-r
    Commented Dec 18, 2019 at 4:47

I think this is a good idea, because when you ask a question you have no idea how many answers there will be. Also there is a period where you have to wait until you can offer another bounty.

Another thing, there may be two answers that give the same technical answer but explain it in different ways, you understand one answer slightly better, but to someone else the other answer is easier to understand. And it's been said before the SO is more about being able to find answers through search engines, months after the question has been asked.

Also one answer may reference another answer and then build upon it. It would be inefficient, and create clutter if when you reference another answer you have to basically quote the entire thing. Also it's not fair on the person whose answer was built upon.

Getting an answer to a question is not just looking at one Answer segment but looking at all of them.

You should be able to set a bounty of say 200, and then give up to 200 reputation to a combination of answers.


Now that bounties are no longer tied to the answer acceptance system, it is possible to offer multiple bounties on the same question and award them to different answers.

Note that I agree with Shog about this not being a good idea. But your question was about whether it was possible, and the answer to that is now "yeah, kinda."

As ymar notes in the comments, this approach is less useful now than it was when I suggested it. As of September 2011, repeated bounties on the same question by the same user "cost more."


I asked about answers having multiple contributors. That would keep the accepted-answer down to 1, but offer the original author the ability to share some of the booty with his helpers.

  • 2
    Yes the more I think about it, the accepted answer would have to stay 1, but it would be nice to be able to either allow the original question author or the author of the accepted answer to split the bounty amongst the contributors.
    – defines
    Commented Jul 6, 2009 at 15:58

Splitting a bounty seems unfair to me.

A possible solution is to allow users to award the whole bounty to multiple users at once, if my current reputation allows that. Please see these two questions on meta.math.SE:

Why can't I offer a bounty of less than 500 reputation?

Giving a bounty to several answers at once.


The ability to accept multiple answers alone would require a pretty sizable change to the underlying concept of the StackOverflow mechanism.

If it weren't for the way bounties are currently decided when there is no accepted answer (i.e. the winner is the one with the most votes after the time limit) then I would agree with you.


It is very hard to comply interests of askers, responders and audience.

Asker interests:

  • Solution for the specific problem
  • Quick answer
  • Valued interesting question (non-FAQ, nontrivial)
  • Subjective aspects
  • (etc.)

Responder interests:

  • Valued... efforts, even if non-complete
  • ~ a unique solution
  • ~ fast response
  • ~ essey with a higher perspective
  • (etc.)

Audience interests:

  • A frequently asked question
  • A very specific question
  • Easily understandable answer
  • A short example
  • Complete explanation with references
  • (etc.)

Currently, main goal of SE/SO is serving the audience. However, only the asker can accept an answer, even because of a subjective thing.

Of course, some interesting situations easily occur, for instance, when "each answer containing a piece of the puzzle", is it acceptable to copy into one and wait for accepting (reputations $$$)?

Anyway, upvotes can deliver nice sum of reputation to the house. So you can value any answer with an upvote, and IMO this is adequate in the most case.

Nevertheless, there are questions with multiple valid answers, for example:

  • different complete solutions (algorithms, recipes) to the same problem
  • general vs. specific ("pure java solution", "vegan recipe")
  • simplified for-beginners answer vs expert TLTR explanation
  • A correction vs "Why do not this"/"Do this instead" answer
  • (etc.)

You must log in to answer this question.

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