home
puzzles
genres
studio
help
about
links
forum


Forums
  [Search] Search   [Recent Topics] Recent Topics   [Members]  Member Listing   [Groups] Back to home page 
How to make a nice puzzle?
Forum Index -> General Discussion
Author Message
Jeroen



Joined: 22/07/2011 09:01:37
Messages: 7
Offline

Hi everyone,

I've been enjoying PuzzlePicnic a lot for the last six years or so. I am very gratefull to all the people who make this website possible.

I'd like to pinch in and also design some puzzles, but I need some help there. Not technical help, I've had a look at the design studio which looks pretty straightforward, and I even made one (rather easy) puzzle once, but I am especially wondering how you puzzle designers come up with nice ideas, and how to make sure that your puzzle has only one unique solution.

Hope to hear some tips and tricks.
JHe



Joined: 01/12/2011 11:39:39
Messages: 87
Offline

Now there happen to be two of my puzzles in the silo. They exemplify different sides of this. Pity that they don't actually show except to me and the judges.

First there's a Cogwheels. The idea came from another Cogwheels I solved, and then I implemented it as simply as possible. So I'd say that existing puzzles are a good source to ponder. Of course there are also outside influences, such as the Valentine's Day for shapes of some puzzles, and George Michael's death that caused me to make an Alphametics - pity that my tribute wasn't good enough to be published.

Then a Fences is an example of how things can go wrong. I thought that I checked it rigorously, but at some point I must have written off a path that was possible after all. So there was another solution. And why is it in the silo? Murphy's Law worked so that it has slipped there and must be retracted. So the system can go wrong, too. Sorry for the trouble, folks.

To conclude:
Check, and check again.
About the ideas: As with (other) arts, study the previous masters and then create something of your own, possibly after imitating them for exercise. It would be ideal if you had a friend that could discuss this scene with you.
edderiofer


[Avatar]

Joined: 16/06/2012 16:15:22
Messages: 229
Offline

JHe wrote:
Then a Fences is an example of how things can go wrong. I thought that I checked it rigorously, but at some point I must have written off a path that was possible after all. So there was another solution. And why is it in the silo? Murphy's Law worked so that it has slipped there and must be retracted. So the system can go wrong, too. Sorry for the trouble, folks. 


Oops, I must have accidentally clicked the "accept" option when I shouldn't have. Silly me, this has now been fixed.
[Email]
JHe



Joined: 01/12/2011 11:39:39
Messages: 87
Offline

Great. I moved it to the sandbox and then it could be deleted, as you said.

About the original topic: I didn't give much actual tips, that's because there are real masters among the authors and I'm not one of them. But feel free to ask about certain genres, here or in private messages.
Jeroen



Joined: 22/07/2011 09:01:37
Messages: 7
Offline

Thanks JHe, for your answers! I'd love to hear more answers from "The Masters".
Of course I realize that my question is very general, and strategies etc. might differ greatly per genre.
Johan


[Avatar]

Joined: 22/12/2006 20:08:51
Messages: 1037
Offline

I usually either start by choosing a puzzle type that I feel like playing around with and then I try to think of some constraint to impose on myself (this could be as simple as rotational symmetry), or I start with some theme in mind and then I try to find a puzzle type in which I could explore it.

Then I draw an empty grid and repeatedly place more clues and derive whatever I can from what I have given away so far, all the while keeping in mind my self-imposed constraints, the fact that it needs to end up with a unique solution (*) and that it should have a fun solving path.

Every now and then I will need to backtrack because I can't make it work. This is most often because I realize some area of the puzzle will allow for multiple solutions (if any) if I don't give out certain clues that I don't want to give out, or because I have reached a contradiction. Sometimes it is also because the puzzle would become too easy or too hard.

Puzzle making experience and familiarity with a puzzle type can help to build intuition for spotting potential issues far enough in advance. Still, it has happened to me that I spent hours and hours on a single idea and eventually failed to deliver. On other occasions, puzzles came out much better than I had expected with much less effort than I had anticipated.

Creating a puzzle is like solving a different kind of puzzle - one that allows for some more creativity. And although at times you might fail, when you succeed, you are left with something others can also enjoy.

I hope this is helpful.

(*) Note: While solving a puzzle, you can sometimes use the assumption that it has a unique solution in your deductions, but when creating a puzzle you are not allowed to do this - at least not to prove the uniqueness.

P.S. I should clarify that for some of my puzzles I wrote computer programs to help me find or create them.
Jeroen



Joined: 22/07/2011 09:01:37
Messages: 7
Offline

Thanks Johan, that sure is helpfull! I really enjoy the puzzles you make.
I'm going on holiday now, but after my holiday I'll try to design puzzles as well, as it sure sounds very fun to do.
 
Forum Index -> General Discussion
Go to:   
Powered by JForum 2.1.6 © JForum Team