If you know anything about game design, you know that it’s a process of iteration. You try new concepts and fail until you find something that works, then you refine it until you have something you can publish. The notion that we can take a single brilliant idea and turn it onto a masterwork is nothing but an appealing myth. Puzzle games are different though. To make a good puzzle game, you need to start with a solid mechanic and extrapolate it in all possible directions. You’ll need to have a significant chunk of your development process done before you even know if your mechanic works or not, and if it doesn’t you often need to start back from square one. There are aspects you can polish – timing, UI design, power-ups - but creating something elegant like Tetris or Bonza requires strong intuition and more than a small leap of faith.