I've been disappointed a few times when I tried to solve a promising-looking puzzle, given up, and waited eagerly to hear the answer, but no answer came because the asker revealed they don't know it and nobody could solve it. This made me feel like I wasted my time looking for something that wasn't there.

By default, I expect questions posed to be puzzles with nice solutions that were crafted for me to find and enjoy. It's fine to pose a puzzle whose answer you don't know or an intriguing question you tried to solve yourself without success. But as a solver, I'd like to be warned that the solution might be impossible or ugly or require brute force.

This largely seems to be a miscommunication due to the similar formats of puzzles and actual questions.

So, I'd like to propose the following policy:

When posting a question in puzzle form to which you don't know the answer, you must say so. Otherwise, it is assumed by default that you are posing a puzzle to which you know the answer.

  • 1
    $\begingroup$ How do you think it would be enforced? Also, is there a way to mitigate the impact on new-user site usability? $\endgroup$
    – user20
    Apr 22, 2015 at 0:09
  • $\begingroup$ @Emrakul On Meta.SE, you are required to tag questions with either "support", "discussion", "bug" or "feature-request". Maybe we could require questions here to be tagged either "challenge" or "non-challenge" (those may not be the best tag names, I know)? $\endgroup$ Apr 22, 2015 at 0:20
  • $\begingroup$ @pacoverflow I thought about the same, but it's probably more trouble than it's worth. $\endgroup$
    – xnor
    Apr 22, 2015 at 0:24
  • $\begingroup$ That seems backwards to me. Why isn't the default assumption that the asker does not know the answer (hence the asking)? $\endgroup$ Apr 27, 2015 at 22:44
  • 2
    $\begingroup$ @Gilles The status quo of the site has trended towards an increase in challenge questions. At this point, it's relatively safe to assume that if an OP didn't specify which it is, it's a challenge question. $\endgroup$
    – user20
    Apr 28, 2015 at 7:14
  • $\begingroup$ I feel like this would mostly be the case in puzzles with optimization or puzzles from an outside source, so maybe it should be assumed that with optimization the solution is unknown? $\endgroup$
    – mdc32
    Apr 29, 2015 at 0:44

3 Answers 3


I do not think it is justified to require every question to state explicitly whether or not the answerer knows the answer. This would clutter up the challenges in a way that is unnecessary in most cases.

That being said, if the author implies a question can reasonably be answered when he has no idea whether it can or not, it is likely to waste users' time without providing benefit to the site or any users. This definitely does not mean a question like that is not acceptable on this site but that this pertinent information should not be intentionally hidden from other users. This is much more important in cases where any good solution would take an exorbitant amount of time or effort.

As a result the following should be considered good etiquette on this site

  • Any user with more than 50rep can make reasonable request for a general origin of the challenge and OP's knowledge of the answer within the question's comments.
  • Someone (preferably the OP) should willing answer the request without being required to provide any hints.
  • It is sufficient to indicate that the challenge was published elsewhere with an indication that an answer is possible. It is not required to reveal the citation of the original challenge as this would likely also provide an answer or additional hints.

I am of the opinion that if a question is taken or even paraphrased from another site, it would be good etiquette to provide the source once an answer has been accepted. That is not, however, a discussion for this meta post.

  1. It can't be made compulsory to declare whether you know the solution or not. It is a matter of personal opinion, and a puzzle (not necessarily a good puzzle) is complete without a good solution. Voting will decide if it is a good puzzle.

  2. Often it takes only 3-4 days to know if a puzzle is solvable. If you are not sure, you can always ask the OP. It is upto him/her whether to reveal the fact or not.

  3. Puzzles with no pre-defined solutions are, in fact, harder and require thinking in different ways. For solvable puzzles, one tends to directly think out-of-the-box. An unknown puzzle will require you to first think of all common ways of solving, and then think of more complex solutions.

  4. There is nothing wrong in a bad solution, as long as it made you think. If it takes you half an hour to prove that a puzzle can only be solved via brute force, one could actually call it a good puzzle (again, opinion-based) because the ultimate objective of a puzzle is to make you think differently.

  5. Optimization questions (such as the recent 'binary poison' one) often have good solutions that the OP him/herself might be unaware of.

  6. If you do want a 'challenge' and 'non-challenge' tag, please be aware that a single question can have a basic puzzle with a solution, and an advanced variant(s) without one.


This is unfortunately a large deviation from how most SE sites work, where the default expectation is that you DO NOT know the answer to a question you are asking (and otherwise provide the answer yourself rather immediately).

This would especially be hard to enforce with people who orignially came from other SE sites. A Tag-Based solution may be best, but it can't completely solve the problem which is ignorance of culture.


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