I was wondering how one could come up with a brand new puzzle like 2048 that would be interesting and exciting. It seems like I succeeded in developing such a puzzle.
Given a small number of lines, one can solve the puzzle by randomly trying. As the number of lines increases, the complexity of the game grows not only quantitatively (you have to go longer to the goal), but also qualitatively: you have to invent new techniques and approaches because the old ones are not enough.
The game feels like a jigsaw puzzle, each piece of which fits any other piece, but the overall picture does not add up.
Arnold's puzzle is based on open mathematical problems in combinatorial geometry. Related topics: https://en.wikipedia.org/wiki/Arrangement_of_lines https://en.wikipedia.org/wiki/Braid_group https://en.wikipedia.org/wiki/Kobon_triangle_problem
Given a small number of lines, one can solve the puzzle by randomly trying. As the number of lines increases, the complexity of the game grows not only quantitatively (you have to go longer to the goal), but also qualitatively: you have to invent new techniques and approaches because the old ones are not enough.
The game feels like a jigsaw puzzle, each piece of which fits any other piece, but the overall picture does not add up.