A good puzzle: Observation -> Experimentation -> Understanding -> Execution
If you’re making a puzzle game, it can be useful to examine your levels by this structure. If your game contains levels and one is too easy, it might be because players come to an understanding too quickly, or they don’t need to experiment and understand enough to execute its solution.
Here is the breakdown for Circles:
Observation happens quickly because there are little elements and the goal is always the same. This is learned early on in the first few levels.
Experimentation is where you try to figure out the functions of the puzzle elements. The significance of this step is largely dependent on the type of puzzle you’re making, but it was one of the biggest parts Circles focuses on. You cannot understand how the circles work without experimenting/interacting with them.
Understanding comes in when you found all the behaviors of the circles to reach the end.
Execution is then an (hopefully) easy path to the end.
The hardest part for Circles is to make a puzzle that, before execution, requires full understanding, so you get the most out of experimentation.
Many levels failed because players could easily reach the goal without even observing, experimenting or understanding the elements. What was supposed to be an introduction to a new circle failed because there was a way to complete the puzzle without having to understand it. And if there is an easy way, players will find it and won’t look any further.