Providing solutions to one's own puzzles

Question

The community is in the process of deciding which questions of the form "Here's a puzzle" should be on-topic, particularly when the poster knows the answer (see, e.g., here and here). I'm not taking a position on this issue right now. I want to suggest an idea regarding posting puzzles as questions, to see what other folks think.

One of the great things about Stack Exchange is that it provides a permanent record of questions and answers, which can continue to be useful to people long after the question has been answered. If we want to allow puzzles as questions, then someone should be able to visit this site and browse puzzles (with solutions).

Right now, there are some issues with this.

• Some puzzles don't have accepted answers. Someone should not need to read the comments to determine the correct answer to a puzzle.
• Some accepted answers are not written like an answer. What I mean is, some answers have wording "I think the answer is probably blah, because blah". Even if the answer and reasoning are correct, it sounds like a guess, not an answer. There are sometimes superfluous arguments in support of these answers, because they're written like defenses.
• Accepted answers sometimes have minor differences/omissions from the answer the poster had in mind. The person who posted this question sometimes leaves a comment to this effect. Again, I don't think someone should need to read the comments to get the full solution.

It occurs to me that in many cases, the person in the best position to write a high-quality solution to a puzzle is the person who posted the puzzle.

Proposal: A policy that when someone posts a puzzle as a question, they should also post (and accept?) a solution, if they know one.

Pros:

• This would improve the quality of answers.

• In the status-quo, people asking questions deliberately conceal information. In my opinion this is against the spirit of the Stack Exchange network.

• This would make voting easier. The quality of a puzzle depends heavily on what the answer is.

Cons:

I'm sort of guessing here. If somebody else wants to rewrite or add to the list of cons, please feel free.

• This might make solving puzzles here less fun for people. I imagine part of the excitement is racing with other users to see who gets the answer first, and getting recognition and reputation for having the best answer.

Answer 1

I think a better approach would be for the asker or the writer of the "correct" answer to then tidy up the accepted answer. This means removing things like "probably"... Also what I, and others, have done is to actually edit the accepted solution with a full explanation of the riddle's content. The riddles that we've got lately are complete original so the author is in a special place to do this. I've noticed some riddlers will also post a separate answer will the details that the winner didn't manage.

I agree about how comments should not be necessary to understand what's going on.

In summary I think the best approach is for OP and the "winner" to make sure the answer is tidy at the end of the day - whether it be by editing the accepted answer or indeed posting an answer of their own to supplement missing information.

Additionally, answer should not really be framed as you describe. The practise of saying framing things like a shot in the dark are being discouraged and downvoted these days - also I've noticed the community punishing lack of explanation (and polite encouragement of fleshing out all clues) in a very effective productive way lately.

PS: Back to this heartbreaking issue: "The community is in the process of deciding which questions of the form 'Here's a puzzle' should be on-topic". As to the usefulness to future visitors, (and this may apply even to Sudoku's actually), how about the 1000s of questions on SO fragments of regex? I'm talking about "What does this regex do?"... there are literally an infinite number of these! (unlike Sudoku). The only different I can see with those is that they have an objective mathematical definition and win condition - which frankly, I see as possible (but hard) for the puzzles in question.

Answer 2

Interesting point. My take on the pros and cons are a little different than OPs as I'll highlight below. Ultimately I oppose this proposal.

Proposal: A policy that when someone posts a puzzle as a question, they should also post (and accept?) a solution, if they know one.

Pros:

• This would improve the quality of answers.

This depends how we view quality in answer responses. Is it higher quality for answers to be fine-tuned using a pre-determined and shared solution? Or is it higher quality for solutions to be discovered through the acts of puzzling.

On the note of sloppy or piecemeal answers, if one is the correct solution then I agree with @d'alar'cop that the questioner can clean up the one they give the check mark.

• In the status-quo, people asking questions deliberately conceal information. In my opinion this is against the spirit of the Stack Exchange network.

This ties in to my previous point on the quality of answers. In the spirit of puzzling, concealing information and discovering solutions is essential. Most Stack Exchange sites are not this way, and dare I say, it is puzzling how well-suited Puzzling.SE is for the Stack Exchange format.

Overall, SE appears to me (admittedly a novice) to be about building a high quality database of questions and answers, along with communities of Q&A'ers, about various topics. On the topic of puzzling, the mysterious approach to puzzling questions is both core to puzzling and still in tune with the core of SE.

• This would make voting easier. The quality of a puzzle depends heavily on what the answer is.

Like voting on answers for their effort prior to knowing if it is correct or not seems to work well and fair here, I think voting for questions for their richness as a puzzling question prior to knowing its solution is also well and fair. Sometimes questions turn out to be more like duds when they've been tried and the solution is revealed, and in those cases I suppose votes can be changed. I think this is the 'Pro' I agree with most anyway.

Cons:

I'm sort of guessing here. If somebody else wants to rewrite or add to the list of cons, please feel free.

• This might make solving puzzles here less fun for people. I imagine part of the excitement is racing with other users to see who gets the answer first, and getting recognition and reputation for having the best answer.

I would also add, and this is the motive for this response of mine anyway, that:

• Visiting 'Unanswered Questions' of Puzzling.SE is part of the fun and spirit of puzzling here. If all questions come with an answer to start with, the Unanswered Questions becomes much less mysterious and puzzling territory.

Answer 3

“Guess what I have in mind” is not a suitable question for Stack Exchange. Stack Exchange is for questions with lasting value. Once a guessing question is answered, it has no value. Such interactions should take place in a forum that deletes or archives content once it's no longer open.

If a “solve this” question is suitable, then it doesn't matter what solution the asker had in mind. What matters is what solutions match the question as written.

“I think the answer is probably blah, because blah” is a perfectly fine answer, or rather a better wording would be “a possible answer is blah, because blah”.

If “solve this” questions are in scope, then they should have an objective correctness criterion, and even an objective winning criterion. For example, if the question is a riddle, then any answer that provides an explanation for all the elements of the riddle is correct — never mind that the answer was “supposed to” be time, or teeth, or whatever. The lack of an objective winning criterion is really the sticking issue with riddles.