I think it would be more interesting if each layer had a different basic shape, like one requires a zig zag, one a spiral, one a triangle, etc. It would also make it more of a puzzle if the "peg" was an irregular shape, meaning it needs to be rotated to fit through certain gaps. Repeating the puzzel many times can get tedious, so maybe the final layers should have fewer moves associated with them.
Honestly, I don’t think it would be all that fun for me to repeat the same sequence of motions hundreds of times even if the layers each had their own distinct maze. Feels more like a fidget toy than a puzzle. Nothing wrong with a fidget toy and perhaps improvements could be made to improve motion and tactile feedback for idle fidgetability.
I designed a puzzle similar to this once, but framed… it used different mazes on each piece, so it was easier to get lost. I love seeing how it could have been done frameless, nice idea!
A single straight line. This puzzle is all about repeating the same step over and over again, so the more complicated you make that step, the more annoying the puzzle gets.
@@scudleeI was thinking about the Hilbert curve as well. Sounds like it could work. Also fits the theme of the puzzle quite well as a self-similar fractal.
Hi Oskar, I've been watching your channel for a long time and I just got a 3D printer, eager to finally print some of your puzzles! I'm quite new to this so can you maybe point me to some sources for printer setup suitable for printing these mechanical puzzle pieces? Like what filament types to use. Thank you!
Your question got me thinking of some interesting answers, I would suggest using a space filling curve like the Hilbert curve or Peano curve at some low number of iterations. Perhaps if this design allows for variation between layers the easiest to move layer could have have the highest complexity and the hardest to move layer could have the lowest complexity.
It's a a labyrinth, not a maze. Have you tried a spiral or other designs such as the classic labyrinth? Those could be interesting, but I think your zigzag may be pretty ideal so that it can be operated more easily. Even so, it clearly catches a lot, and I think that optimizing it for smooth operation will be a good investment.
@@rosebustersome make the distinction that since the Labyrinth in myth was a single path with no dead ends.labrynths should refer to similar one pas constructions and maze refers to those with multiple paths and dead ends
My mind instantly went to finite approximations to space filling curves. Maybe it would be a bit too curved and take too long, but it would retain a repetitive element without being too simple.
These exponential puzzles are lovely as mathematical concepts but probably too tedious to really play with. I'm wondering are there puzzle types that grow polynomially? Like quadratically?
Challenge for you Oskar. You've been making puzzles with exponent 2 and 3. But can you make a puzzle with a lower exponent? There exists such a thing as a mechanical "finary counter" where successive numeral places are the Fibonacci numbers, and the carry rule is that 011 = 100, ie, two adjacent 1's cause a carry. And we know that each Fibonacci number is about Phi times as large as the last. So can you design a "Phiggurat puzzle" that grows by the golden ratio? Can you make an exponential puzzle that grows by the Plastic Ratio? The Plastic Ratio has a corresponding carry rule too, so it should be possible. Low ratios are beneficial in a way, because they're still exponential but it's gentler.
Did you have a look-ahead in my playlist? ruclips.net/p/PLwdtSpXcZNXnWGKjE7U1r5zIjbeH5J_-4 Because you are describing exactlu Zigguphi, Fibonary, and a bunch of other recent prototypes. There will be a full article in the Cubism For Fun magazine later this month.
I think it would be more interesting if each layer had a different basic shape, like one requires a zig zag, one a spiral, one a triangle, etc. It would also make it more of a puzzle if the "peg" was an irregular shape, meaning it needs to be rotated to fit through certain gaps. Repeating the puzzel many times can get tedious, so maybe the final layers should have fewer moves associated with them.
To clarify, I didn't mean rotate just the peg, but rather rotate the entire top half of the puzzle in order to get the peg oriented correctly.
Honestly, I don’t think it would be all that fun for me to repeat the same sequence of motions hundreds of times even if the layers each had their own distinct maze. Feels more like a fidget toy than a puzzle. Nothing wrong with a fidget toy and perhaps improvements could be made to improve motion and tactile feedback for idle fidgetability.
I designed a puzzle similar to this once, but framed… it used different mazes on each piece, so it was easier to get lost. I love seeing how it could have been done frameless, nice idea!
What would be a better maze for this puzzle?
Sliding squares
A single straight line. This puzzle is all about repeating the same step over and over again, so the more complicated you make that step, the more annoying the puzzle gets.
Would a space-filling curve, like a Peano or Hilbert curve work?
@@scudleeI was thinking about the Hilbert curve as well. Sounds like it could work. Also fits the theme of the puzzle quite well as a self-similar fractal.
Hi Oskar, I've been watching your channel for a long time and I just got a 3D printer, eager to finally print some of your puzzles! I'm quite new to this so can you maybe point me to some sources for printer setup suitable for printing these mechanical puzzle pieces? Like what filament types to use. Thank you!
Your question got me thinking of some interesting answers, I would suggest using a space filling curve like the Hilbert curve or Peano curve at some low number of iterations. Perhaps if this design allows for variation between layers the easiest to move layer could have have the highest complexity and the hardest to move layer could have the lowest complexity.
It's a a labyrinth, not a maze. Have you tried a spiral or other designs such as the classic labyrinth? Those could be interesting, but I think your zigzag may be pretty ideal so that it can be operated more easily. Even so, it clearly catches a lot, and I think that optimizing it for smooth operation will be a good investment.
I thought those words were synonyms.
Surprisingly, the Ziggu* puzzles have dead ends. So even though individual layers are labyrinths, the whole is a maze.
@@rosebustersome make the distinction that since the Labyrinth in myth was a single path with no dead ends.labrynths should refer to similar one pas constructions and maze refers to those with multiple paths and dead ends
@@TheArcv2 I guess it could also a language thing. In Polish, which is my first language, there's only a single word "labirynt" for both of these.
It's a really interesting design. It seems to slide in a very satisfying way.
Thank you!
My mind instantly went to finite approximations to space filling curves. Maybe it would be a bit too curved and take too long, but it would retain a repetitive element without being too simple.
maybe if it was a different maze on each level it would make it more interesting having to remember and solve each maze multiple times
These exponential puzzles are lovely as mathematical concepts but probably too tedious to really play with. I'm wondering are there puzzle types that grow polynomially? Like quadratically?
Oskar van Deventer is the geared ziggucuboid of Bram Cohens
Challenge for you Oskar. You've been making puzzles with exponent 2 and 3. But can you make a puzzle with a lower exponent? There exists such a thing as a mechanical "finary counter" where successive numeral places are the Fibonacci numbers, and the carry rule is that 011 = 100, ie, two adjacent 1's cause a carry. And we know that each Fibonacci number is about Phi times as large as the last. So can you design a "Phiggurat puzzle" that grows by the golden ratio?
Can you make an exponential puzzle that grows by the Plastic Ratio? The Plastic Ratio has a corresponding carry rule too, so it should be possible.
Low ratios are beneficial in a way, because they're still exponential but it's gentler.
Did you have a look-ahead in my playlist? ruclips.net/p/PLwdtSpXcZNXnWGKjE7U1r5zIjbeH5J_-4 Because you are describing exactlu Zigguphi, Fibonary, and a bunch of other recent prototypes. There will be a full article in the Cubism For Fun magazine later this month.
Ooh awesome!
You should do a maze that looks like a lobster
This is what my normal dresser thinks it is.
A spiral maze