A good IF puzzle makes the player feel clever after solving it. A bad puzzle makes the player feel like the author hid the stairs and then mocked everyone for not levitating.
Fair puzzle principles
The goal is understandable
Players should know what they are trying to accomplish, even if they do not yet know how.
The solution uses visible information
A clue can be subtle. It should not be imaginary. If the player solves the puzzle only by guessing a verb no one would reasonably try, the puzzle is not clever; it is a trapdoor.
Failure teaches
Bad failure:
That doesn't work.
Better failure:
The magnet tugs weakly at the latch, but it cannot reach through the thick glass.
Now the player knows magnetism matters and distance or glass matters.
The world is consistent
If fire burns rope in one room, it should not politely ignore rope in another unless there is a reason.
The player can recover
Avoid unwinnable states unless the game is clearly designed around them. If the player can lose a vital item forever, warn them or provide another path.
Puzzle types that age well
- Using an object in an unexpected but logical way.
- Reading environmental clues.
- Understanding a character’s motive.
- Reframing a goal.
- Combining two partial observations.
Puzzle types that often curdle
- Guess-the-verb.
- Pixel-hunt translated into prose.
- Required jokes that only make sense to the author.
- Mazes with no mapping logic.
- Timed sequences with poor feedback.
The author’s curse
You cannot play your own game as a new player. You know too much. Testers are not optional garnish; they are the canaries, cartographers, and door-kickers of your little world.