Handling puzzles with multiple steps and the posting of partial answers

Question

^{Note: This post was created specifically due to mod flags, and comments on answers in my treasure hunt question, however the discussion is probably more broadly applicable than that.}

Obviously, the ultimate goal of any StackExchange site is to collect high quality questions (in this case good quality puzzles and discussions thereof), with high quality, well explained answers. To quote the banner that got applied to my question, "We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations."

However, on a "challenge" type site (which, like it or not, Puzzling.SE has become), I would argue that a big part of the purpose of the site is for participants to compete in being the first to solve the puzzle. Therefore if the puzzle has a long, multi-step solution, posting incorrect or partial solutions may give away vital information allowing others to bypass parts of the challenge.

In my specific instance, there's probably a dozen or so smaller puzzles that need to be solved in order to reach the final solution, so if someone posted a fully explained attempted answer that happened to get one or two parts wrong, then the next person to come along could, with much less effort, get the correct solution (and thus the "prize").

With that in mind, I said in my puzzle, "if you want to keep your working to yourself initially (this is a race for the treasure after all!), feel free to [post unexplained/unjustified answers]", then clarified with "obviously, before being awarded the tick, you'll need to justify those steps with valid logic though".

Three possible solutions that I can see are:

1. Unexplained solutions are acceptable as "partial/incomplete answers", but should never get the tick until fully explained/fleshed out
2. People who don't want to give anything away in multi-part puzzles don't post at all until they have a complete solution
3. We force people to explain all answers and thus allow others to pick up where they left off and potentially bunny hop them to "victory"

I was going with option 1, as I feel there are issues with the other two (when applied to multi-part challenges). In option 2, you end up with potentially complete silence on the challenge giving the creator very little feedback on how it's being approached, where people are struggling, etc, and thus what hints to provide. With option 3, you end up in the situation as described above when the person who does 90% of the work gets pipped at the post, and someone else gets lucky and receives the "award".

Is there some community consensus as to what is acceptable in this specific regard (beyond the generic "answers must have explanations" guidance applied to all SE sites)? And if so, are there circumstances in which exceptions to that consensus should be applied?

Answer 1

Another thing you can do is to post and accept your own answer which collates/quotes from each individual poster's partial answers. Something along the lines of

The correct answer is:

Foobar!

---

First, you do this, as ghosts_in_the_code figured out here

....

Then, you do this other thing, as Kate Gregory figured out here.

...

And so on. Then you mark it Community Wiki.

Doing it this way pulls all the relevant information into a single post, links to the appropriate steps wherever they came from (regardless of upvote order, which can change), and makes it clear that you're just acting as curator, not trying to undercut everyone else.

Also, in addition to upvoting, you can award a bounty to someone if you feel like they solved the majority of the puzzle, even if they failed to take the last step.

Answer 2

You can give the tick to whoever you like. It doesn't have to be whoever got the complete solution first. The tick is for the "best" (in the opinion of the asker) answer - "the answer that helped you most" on problem-solving sites like SO or Travel. What's more, you can move the tick if a better answer is posted.

People who post soon may be "giving away" clues but they're also getting upvotes from everyone who likes the answer. Someone who posts later, piggybacking on those clues, may not get as many upvotes since not everyone comes back and "evens up" their votes.

People can edit their answers, too. So if you need to know A, B, and C to solve the puzzle, and I post that I have discovered A and B, with good strong explanations, then someone else says "aha in that case C and here's the solution!" I can edit my answer to include the C-related information (that I now understand/see) and make my answer even better. I will probably get the tick because my complete answer is better than their "complete" answer that doesn't have explanations.

I don't think questions should really include meta information like when to answer and when to comment, but I would discourage "answers" like "Is it 7? If it is, I'll explain why." Either comment "is the key the colour scheme?" or put an answer out there that shows why you think it's 7. I don't think there's a real risk that others will benefit from that, mostly because I don't see that as a risk, I see that as the point of the site. To loop back around to the pirate map question, I haven't upvoted either answer that is there now because they are of no use to me. If someone were to post an answer that actually attempted to decode or explain even one clue, I would probably upvote them.

Answer 3

The accept button is only to help new users and people who don't have time to spare, to know which is the simplest, but complete answer that makes sense by itself. I would consider giving it to a well-explained answer, even if it happens to be a more well-written copy of another answer.

Upvotes are what actually affect rep, so I prefer giving upvotes to people who answer first. I upvote well-written but late answers, only if they really have put some original effort into.

There is nothing wrong in accepting a low-scoring answer; in fact, I think there is a badge granted to those who do so.

Answer 4

Thinking about it more, I think the solution that provides the closest to the best of both worlds may just be:

4. I change my question to suggest that partial, unjustified answers be posted as comments on the question, rather than answers

_{(My only concern here is that the comments will become a cluttered mess, but I guess the mess has to go somewhere. In my specific instance, I'll just collate failed attempts for the purposes of hints in the bottom of the question itself...)}

Answer 5

Workable possibilities are best illustrated with examples. I may expand this answer as I think of more relevant cases.

---

In this excellent puzzle, the OP accepted the answer which solved most clues (also the oldest answer) and awarded a bounty to the person who found the final clue, even though that answer only solved one of the many clues. He also updated the question to include (spoilered) solutions to each clue solved so far, in order to help people know which they had to focus on to finish it.

In this absolutely brilliant puzzle, the OP created a community-wiki answer to summarise the (extremely long!) solution, crediting different answerers for different parts of it. He also awarded a 300-point bounty to one of the other answers.

In this very enjoyable puzzle, the OP accepted, and awarded a bounty to, the answer which solved the most parts of the (multi-layered) puzzle. He also edited the question to credit those who solved each individual part.

(Pardon the opinionated adjectives. I can never resist eulogising the second of these puzzles, so I had to say what I thought of the others too.)