I was drawn back to the following question from activity in my feed:

How do we assign Game of Thrones houses?

As I looked at it again, I started to think of the question in a different light. It looks to me like a survey question, where the OP has posted a question and three different options, allowing the community to vote on which one they like more. (The catch here is that I stepped in the middle and posted an alternative.) Given the format, it would have also fit on SurveyMonkey.

For me, this seems like a misuse of the Q&A format provided by StackExchange sites. Should questions that function as survey questions be allowed on B&CG.SE? If not, what reason would be given for closing?

  • I think it is just a bad question as the solution appears to be spelled out in the rules.
    – Joe W
    Commented Dec 30, 2016 at 20:57
  • 1
    @JoeW I think the point I'm trying to make is that this is a survey disguised as a question. The OP knows that there is a solution in the rules, but is laying out alternatives and letting people vote on them.
    – SocioMatt
    Commented Dec 30, 2016 at 21:43
  • 1
    In this case it is really about house rules but it is overall just a bad question.
    – Joe W
    Commented Dec 30, 2016 at 21:57

2 Answers 2


No, we shouldn't.

Such questions are not in the spirit of a "one question gets one answer" ideal that Stack Exchange aims for.

Then of course, these are almost always primarily opinion-based if there is no way to objectively determine the best answer, and too broad if they either invite thread-style discussion or cannot be comprehensively answered by less than a chapter in a book.

  • 3
    I don't agree that SE has a "one question, one answer" ideal. Many questions have multiple solutions which could be packed into different answers, and part of graduation has to do with the question to answer ratio being 2.5 or higher. I also think that this question could be based on expertise, which is fine by site standards as outlined in help center. I don't think your answer is addressing the core issue I'm relaying, which is using SE as a survey tool.
    – SocioMatt
    Commented Jan 2, 2017 at 18:20
  • Graduation requirements are about usage and user numbers. Having more answers to a question is proxy for activity of answerers. You can't have more answers without either more questions or more answers per question, so the latter is how answer activity gets rated. Then of course, the tour is very direct: "Don't ask about... // Questions that are primarily opinion-based // Questions with too many possible answers or that would require an extremely long answer" And an ideal is not meant to reflect reality, that's why it's an ideal and not a target or requirement.
    – Nij
    Commented Jan 3, 2017 at 2:32
  • I upvpoted only because of the primarily opinion based part of this answer. The other part, "one question gets one answer", could use some justification. A link to a Help page, parent Meta, an SO blog, or anything else authoritative would work.
    – Rainbolt
    Commented Jan 3, 2017 at 13:57
  • I did a little research myself. It's possible to ask a good question that has multiple correct answers. I also found this blog post by Jeff Atwood that talks about polling questions. As Aarobot (the inspiration for that blog post) says, "real questions have answers, not items or ideas or opinions".
    – Rainbolt
    Commented Jan 3, 2017 at 14:15
  • @Rainbolt Those are interesting reads and worth the time. However, I think they are using "poll" in a different way than the issue to which I am referring. In the referenced case, the same user has posted the question and three different answers. In the cases your links discuss, answers to the questions are by different users. The difference is in intent and design; the OP for my referenced post designed the question and answers, whereas the poll questions Aarobot delves into become poll questions because of community involvement.
    – SocioMatt
    Commented Jan 3, 2017 at 18:48
  • @SocioMatt The first link had nothing to do with poll questions. My intent was to show that the Stack Exchange network does accept some questions that have multiple answers. The other link describes exactly the kind of poll question that you are asking about. It doesn't matter who writes the answers to a question like that. A self-answered poll question is still a poll question.
    – Rainbolt
    Commented Jan 4, 2017 at 14:23
  • I don't think it's helpful to focus too much on "one question one answer" or "objectively determine the best answer." Even for objective questions, there may be multiple solutions with no clear best one, or with the best one varying depending on circumstances. That's totally fine, it doesn't break the system, and they can be very useful questions. And this only becomes more true with good subjective questions (which are okay!). What we want to avoid is many answers, especially with no hints at all about which might be better.
    – Cascabel
    Commented Jan 10, 2017 at 1:43

There are many instances where I believe this would be appropriate. There are many games that leave choices up to the user and without any specific way of executing the choice. Having questions which give multiple ways of coming to a decision would maybe yield multiple right answers, but the person asking the question may in fact want multiple answers. This is something that I believe should be changed in the rules even if it's not my decision.

  • A question having multiple and different correct answers on it does not mean it is a poll question. This is about questions where the goal is not to get a correct answer but to rate multiple methods of solving the problem.
    – Joe W
    Commented Jan 24, 2017 at 23:56

You must log in to answer this question.

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