I recently posted a puzzle which seemed to get a lot of activity, people commenting, posting answers and seemed to be popular enough. However, one user had a bit of a rant about how I was "making up a new rule" when an answer was posted to justify certain logic.

I don't have a problem with that - I'm not trying to be petty or anything. I actually agree that some of the edge cases, I didn't specify up-front, and there is arguably some missing data which you could use to formulate your answer.

So bearing this in mind, I decided I'd be a little more rigorous with giving enough examples to rule out most chance-finds, and specifically mentioning a couple of edge cases.

The first question took about 4 hours to be solved, with 15 different people posting answers across that time.

The second question took about 75 minutes with the second person to guess at it getting it correct (and I thought it was a relatively unusual puzzle).

Clearly, I've got two questions which are very different in difficulty. The first one possibly had a couple too many people guessing at it (demonstrating a wide range of potential answers, even if several of them are wrong for the given clues), while the second one was solved relatively quickly.

Are there any good rules of thumb to adhere to when writing puzzles to ensure that

  • the difficulty you're setting is appropriate
  • (in a password puzzle) the examples given can realistically give one algorithm
  • (in any puzzle) you strike a good balance between giving clues and giving spoilers

I appreciate that there are a lot of different kinds of puzzle (and this may draw too-broad close votes because of that), but I'm trying to get an idea for how to determine the amount of hints I should give. Obviously it will differ from puzzle to puzzle, so please feel free to take specific examples off the site to illustrate any answer.


For password puzzles, it's much better to err on the side of giving too much information. Puzzles like this or this erred the wrong way. Compare with my first password puzzle.

I would say that, as a rule, if you post a password puzzle, and someone finds a simple explanation for the data that you didn't think of, then you should accept their answer anyway. Hopefully this rule scares people into making more fun puzzles.


Simply put, a puzzle should be specific enough to guide users towards your intended answer. I learned the hard way that a poorly-defined puzzle can have a significantly large space in which all of the provided solutions are valid, with the only difference being the method with which they were found.


You must log in to answer this question.

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