# What about a question calling for examples of a certain type of puzzle?

(Please see this question and comments on it for where this proposition originates.)

How would people feel about a question of the form 'What are some good examples of puzzles of type X?', inviting answers of the form '[puzzle] >! [solution]'?

There'd be no accepted answer, and the most popular puzzles would rise to the top. According to Josh Caswell and the meta references he linked to, this is not a desirable form of Q&A on SE as a whole - but the same could be said of challenge questions, and we discovered that those are certainly allowed here! I think this would be an excellent way of stopping too many of the same type of puzzle building up (imagine if we'd had a question 'What are good examples of Security-to-the-Party puzzles?'), provided there are only going to be a reasonably limited number of answers: for instance 'What are some good examples of riddles?' would be far too open-ended, but 'What are some good examples of alphabet-splitting puzzles?' would be fine IMO.

Discuss! Or just hit the UV/DV button to show your (dis)approval, if that's the way things work on meta.

• I'll work my comment there into an answer here when I get a chance. – Josh Caswell Dec 27 '14 at 18:36
• No, i just upvoted this; I am in favor of Meta discussion even when (sometimes especially when) I don't necessarily agree with the question author. I haven't voted at all on the main site question. – Josh Caswell Dec 27 '14 at 18:39
• I've upvoted too! – warspyking Dec 27 '14 at 23:58

# I think that it is a bad idea.

A question like 'What are some good examples of alphabet-splitting puzzles?' is a opinion-based question, so this do not start smelling very well. Further, since the alphabet-splitting puzzles could go in a wide range of different ways, it may be too broad as well.

And, what does the accepted answer means? It is not the correct solution anymore, it is no more than the OP's preferred challenge.

This could only work if you just want to get a lot of unsolved challenges or require that the answerers post the solutions together with the challenges itself, and both options don't seems to be very interesting.

Further, see this somewhat related topic, including my answer: Are the following mass-producible puzzles acceptable on puzzling.SE, and is it appropriate to ask for partial answers?

• I don't agree with your first paragraph: we shouldn't set too much store by standard SE catchphrases like 'opinion-based' or 'too broad' if we're trying to set our own precedents for this site. Agreed that such a question would be basically open-ended, but this seems to be OK. And you're right that having answers including solutions (which was my idea) would cut out the fun of solving puzzles, instead focusing on building up a repository of puzzles. [cont. in next comment] – Rand al'Thor Dec 28 '14 at 11:33
• It would probably only work for a very small class of puzzle types, e.g. those whose answers are very short and don't need much explanation. In the alphabet-splitting puzzles I've seen, the answer is just a few words and is clearly right without explanation. Ditto with most of the infamous spaghetti puzzles. Maybe we should set some strict guidelines for what sort of puzzles are OK for this type of question? – Rand al'Thor Dec 28 '14 at 11:39
• I edited the title to not be red; it was burning my eyes. – xnor Dec 28 '14 at 11:57
• @xnor - Split infinitive alert! :-) – Rand al'Thor Dec 28 '14 at 12:47
• If we share puzzles we'd also share answers, afterall an answer is just as much part of the puzzle as the question itself, so your argument is invalid. – warspyking Dec 28 '14 at 14:36
• @randal'thor Ok, even if you disagree with my first paragraph, I still think that it still is not a good idea. If you can create a good derivative or variation of something, go on and post an entire new question. I agree that your new proposed style is ok only for a very few special cases, but I think that they are as easily ok if each challenge is a question. [cont in next comment] – Victor Stafusa Dec 28 '14 at 17:34
• However, if you really want to keep this format, you might change it to something where the challenge is to find novel and interesting ways of splitting the alphabet, or novel and interesting ways where the guards at the party devise password-guessing algorithms. – Victor Stafusa Dec 28 '14 at 17:35
• @warspyking No, what you are saying only applies if you intend to post a challenge with its answer. If you intend that other people solve them, you will simply not post the answer. And even if you intend to post the solution to your own challenge, you might just self-answer your own question. – Victor Stafusa Dec 28 '14 at 17:38
• @Victor YatharthROCK's answer is a 'reply' to yours; I thought someone should notify you! – Rand al'Thor Dec 30 '14 at 17:58
• @randal'thor Ok, thank you. - This reminds me something: If you are in some question format where an answer replies some other answer, people won't be notified. I am not saying that you are proposing that, I am just reminding that if someone proposes a format like that, then it is probably a bad idea. – Victor Stafusa Dec 30 '14 at 18:40

## I think it's a good idea

Further, this moves the puzzle challenges to answers. So, if I want to solve a challenge posted as an answer, what could I do? Answer the challenge in a comment? Comments are not suitable for this purpose, since they must be short and are not editable after five minutes and will get mixed with comments about the answers and comments about other comments.

A question like 'What are some good examples of alphabet-splitting puzzles?' is a opinion-based question, so this do not start smelling very well. [...] And, what does the accepted answer means? It is not the correct solution anymore, it is no more than the OP's preferred challenge.

CodeGolfing.SE has a popularity-contest tag for questions like these where the criteria is subjective, generally "interestingness". Accepting answers could either work like there, where the most popular answer satisfying some criteria in some time limit is accepted, or it could depend on the OP, or it could be discouraged. A special tag for such questions should probably exist too.

Further, since the alphabet-splitting puzzles could go in a wide range of different ways, it may be too broad as well.

Well, every example of a 'good question' on the example page can be potentially mis-used ;) and it is up to the community to filter them.

• "Each answer should be a link to an already-asked question" ... not sure if I like the sound of that. That sort of question would no longer solve the problem of the site getting flooded with questions of the same type (the famous example here being the spaghetti parties). – Rand al'Thor Dec 30 '14 at 18:02
• @randal'thor Sorry, I don't see why; it would be like those old-style web directories. – Yatharth Agarwal Dec 30 '14 at 18:04
• Nice profile pic BTW! You should be on SciFi & Fantasy SE ;-) – Rand al'Thor Dec 30 '14 at 18:05

This sounds to me like a great idea to improve the efficiency of the site. it could serve as a question sandbox where not only can people ask a bunch of questions revolving around the one type for future readers, but can also gain feedback from professional solvers!

Perhaps we could even make a new tag

Tag a question with this if the answers you expect are puzzles themselves in which questioners can gain feedback from professional solvers. A question tagged with this should not have an accepted answer, and after time, will have a collection of puzzles.