11
$\begingroup$

Are puzzles that are "semi-interactive" on topic?

To define "semi-interactive":

A semi-interactive puzzle is one in which the OP may not post all of the puzzle in the question at the beginning - depending on the answers/comments received by the community, the puzzle progresses and more is added to the question to reveal different clues/locks/messages etc.

Imagine playing one of those online escape the room games, but it's (mostly) in words, and you have the community solving the puzzle with you.

For example: A puzzle that describes a locked box, and a detailed room. First, it asks for the community to find a key by solving various puzzles. When the key is found, the box is opened, and the OP adds a section to the question revealing what was in the box (and other clues/puzzles)... etc etc.

$\endgroup$
  • 1
    $\begingroup$ Do you mean that you will edit the question as parts of it is solved? $\endgroup$ – mmking Jun 10 '15 at 19:31
  • 1
    $\begingroup$ @mmking Yes. Exactly. $\endgroup$ – Aify Jun 10 '15 at 19:33
  • 4
    $\begingroup$ I see no reason not to create a new Question when the first puzzle's solved. $\endgroup$ – Masclins Jun 15 '15 at 15:41
  • 5
    $\begingroup$ @AlbertMasclans Consider this: Puzzle one is solved, then puzzles 2 to 4 - and somewhere in puzzle 5, it depends on several key parts in puzzles 1 to 3. Would it not be much better to keep all the parts in the same question? Or would you like to actively have 5 tabs open swapping back and forth to try and solve 1 puzzle? $\endgroup$ – Aify Jun 15 '15 at 15:55
  • $\begingroup$ @Aify you could just put the answers in a spoiler tag or a link? It seems that questions like this are like putting a square peg in a round hole. SE isn't made for it, but I guess there is no harm in trying. $\endgroup$ – Mark N Jun 30 '15 at 19:17
  • $\begingroup$ It's been six months and apparently there are hardly any puzzles tagged semi-interactive (three, to be precise). $\endgroup$ – ghosts_in_the_code Jan 19 '16 at 11:38
  • 1
    $\begingroup$ @ghosts_in_the_code No one really did it the way I imagined it, and I've been busy lately with other stuff. I may do this the way I originally planned in the future, but for now I think i'll have to pass. $\endgroup$ – Aify Jan 19 '16 at 17:07
8
$\begingroup$

While I'm relatively new here and not up to speed on all of the rules, I can say with certainty that I would be very interested in a puzzle of this format. There are a few multi-part puzzles that have some claim to fame here, so I would first be sure that your puzzle won't be just as well suited with multiple parts. Keep in mind that with only one question, you can only accept one answer.

I think the main issue you'll encounter is that the answers will be fairly scattered. If you have 5 parts, and three people attempt to answer each one, that's fifteen answers! Fifteen is a lot of answers for a question, especially when the answers aren't all (if any) conclusive for the whole puzzle.

At the end of the day I'd say go for it, but only if you're prepared to do two things (diligently): One, put together a Community Wiki answer that will compile all the parts that have been solved; and two, don't disappear like some people tend to do. There's nothing worse than solving 60% of a puzzle and not being able to get the final 40% because the poster decided to leave the site.

Also, you might consider something like a 50 rep bounty for each correct answer, to give the community a reason to answer if they have no possibility at the "correct answer" marker (since that will go to the community wiki answer).

$\endgroup$
  • $\begingroup$ Disappearing won't be an issue, if I were to write one of these i'd see it through till the end for sure.. after all, I have posted some really long puzzles.... Community wiki was definitely what I had in mind, but a bounty per correct answer is a bit too much of a toll on the measly amount of rep I currently have - (If i had over 2k, for sure I'd do the bounty thing). Upvotes to answers and credits in the CWiki will hopefully be enough? You've brought up some very interesting points, even though it doesn't actually answer whether or not this would be on topic XD $\endgroup$ – Aify Jun 10 '15 at 20:04
  • 1
    $\begingroup$ I don't for the life of me know if it's on-topic, but I felt like voicing my ideas on the subject because it sounds really cool! :) $\endgroup$ – Bailey M Jun 10 '15 at 20:08
  • 8
    $\begingroup$ I agree with this. If it's something y'all are willing to test out, I'd personally be okay with it (and it sounds interesting), but note in advance that it's definitely not something Stack Exchange is in any way designed to accommodate. Be prepared going into it that it may not/probably won't work the way you'd like it to. $\endgroup$ – Aza Jun 10 '15 at 20:18
11
$\begingroup$

In General

Now that a couple of these types of puzzle have been posted its worth list the problems that have been foreseen and experienced.

Bailey M noted "with only one question, you can only accept one answer" and that the correct "answers will be fairly scattered"

Which have proved true.

There is also the problem that the comments fill up with commentary on the first part of the question. That when a new part of the question it does not appear in the new questions list, makes it difficult to see that a new challenge is available.

Not having an easy way to see the progress of the questions (who has answered what so far) is a pain those details aren't added to the question. Posting a community wiki with the answers together is more useful but does need to be high up the page to get seen easily. The question has the potential to get very long very quickly.

Having no way to up or down vote individual parts of the puzzle is annoying when the parts could have been posted separately.

From what I've seen of the problems I would not recommend this form if the puzzles could be posted as separate questions. It only seems a useful format if there is information in the early parts required to solve the later parts.

In Practice

In the question You're invited! Now what? are some good examples of these problems. I think that question would have benefited from being asked over a shorter period of time to avoid congestion of the question and comment space. Issuing the short answer to each part of the puzzle (spoiled obviously) after a day or two without explanation and moving on to the next part might have cut down the total number of answers (currently 14)

If the verbose flavour text was trimmed that question could be reasonably well balanced. It's first and last parts providing interest to readers any time in the future with the second part only of interest to the person solving in real-time, this demonstrates the utility of the form in providing the opportunity for puzzles that are one shot (ie what is the current...?) while also providing a more durable puzzle in the other parts. This of course being the exception to 'rule' don't post parts together that could be separate.

Although I would recommend not using time dependant information in puzzles in general, here it is done well is it doesn't interfere with the other parts of the puzzle. I think I remember seeing a puzzle where a long calculation had to be performed and one of the numbers would have changed rendering the puzzle obsolete very quickly, which I think is something to be avoided.

If you want to see an example of a semi-interactive puzzle that reverted to a series of individuals look at Treasure hunt 'round the world! (prologue)

Personally I prefer the shorter independent puzzles. They can be tackled one by one or in any order. They have discrete sets of comments making things clear and easy to read and it is easy to find the correct answers.

$\endgroup$
  • $\begingroup$ This is further exacerbated by trying to browse around it on the iphone app. At least in the browser version you sort answers by age. The iphone app doesn't have this option. $\endgroup$ – LeppyR64 Jun 24 '15 at 13:56

You must log in to answer this question.

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