Good puzzles are hard to craft, challenging to solve, and give the person who does eventually figure them out a great sense of satisfaction.

Good puzzles are also hard to come by. Back in college, Microsoft would sponsor a yearly puzzle competition that required a full-day commitment and a team of 3 or 4 people. I only participated in it one year, but it was a very fun day.

I have, since then, judged puzzles against the ones that were a part of that competition. There are, to me, several factors that take an OK puzzle and make it great. Here they are, in descending order of importance.

Minimalism: the best puzzles are ones where every element is integral to the solution, meaning that by definition, it is a minimalist puzzle. Put another way, this could read “no red herrings.” Not every aspect of the puzzle has to point directly toward its solution, but no part of the puzzle should be there for no reason. Even more importantly, no part of the puzzle should be there to intentionally throw the puzzle-solver off-track. That’s mean and, for a good puzzle, completely unnecessary.

Multiple Epiphanies: all good puzzles should contain moments when the solver suddenly understands something; a certain code or representation will suddenly click, or the relationship between two elements in the puzzle suddenly becomes clear. The trick is to have multiple epiphanies. Conventionally, these will build upon each other, but if solving portions of the puzzle independently also has its own reward. Solving one code and applying it three times isn’t too fun. Solving three integrated codes and having to apply each one differently is much more rewarding.

Strong Causality or Interlocking Parts: so your puzzle has multiple parts to give the solver multiple chances to discover things. Take it up one more level, and create a way to interconnect them, so that a solver will be given hints on how to solve section B after solving section A. If you’ve designed the puzzle really well, solving section B first will also give hints on how to solve section A.

Originality: this is the X factor and isn’t something that is as easy to engineer as some of the other factors. But the most brilliant puzzles are those that use mechanics and hooks that haven’t been seen before. Or, perhaps, a mechanic that has been used before but tweaked in a new way. At the very least, you want to avoid a puzzle that makes something think, “oh, this is just like that other puzzle.”

So, go make me some puzzles! And maybe I’ll make you some.