What makes a good puzzle?

After a recent trip to an escape room, I was musing on what made me like certain puzzles more than others, so I thought I’d write down some of my thoughts – partly so that I can get them clear in my mind, partly so that I can point game operators at it and partly to get other enthusiasts thinking. As always, let me know your thoughts by commenting below!


We have to start somewhere, so let’s begin with theming. A great puzzle should fit with the theme of the room. The more it fits the better. At a simple level, this might be using chemical elements in a scientist’s laboratory, or blood samples in a serial killer game. Where possible, don’t introduce objects that shouldn’t be in the game, and ideally, don’t have markings on those objects that wouldn’t naturally be there.

That’s only the first step though. Puzzle theming goes way beyond that. The next step is to make the puzzle fit naturally within the context of the room. Why would a scientist use the sum of the atomic weight of lead, iron and zinc as the code for his safe? It’s perfectly plausible, but make sure that there’s a reason given in the game – don’t expect me to just find the three elements and a periodic table and jump to that conclusion. Make the serial killer’s password be their childhood pet, the scientist use the atomic weight of an isotope they discovered for a padlock code, and the witch require the ingredients for a potion that does what you need to happen.

The final step is to make the way the solution is used fit naturally too. Why would the professor lock up a random cupboard? This one’s tricky, because in general an escape room needs a fair number of different “locks”, so you have to be pretty imaginative if you don’t want to end up with a million arbitrary keys and codes. “Magic” rooms make things a bit easier here because using hidden technology you can detect the proximity of a particular objects, but even for normal rooms, you can come up with a series of options – maybe there are three or four different passwords required to log on to a computer, rather than just one, or maybe you add a key and a combination code to a door where you were only going to put one or perhaps the code allows you to access a card in a filing index with extra information. This angle is truly hard work to get right, but I think it pays dividends.


A good puzzle should be savoured. I really don’t like games that give me a locked box with a pretty obvious number inside, which just opens another locked box. My five year old might think unlocking a box is fun, but for me it’s pretty unexciting, so I’d prefer to condense that to a single step. I want puzzles where I slowly move towards the solution, but how do you achieve that?

First off, don’t give me all the puzzle elements straight away. String me along, but make it obvious that I’m short of information. I don’t want to waste time trying to solve the puzzle, but I do want to be intrigued at to what it might mean. Give me a few more pieces as I progress, constantly teasing me.

Try to stop me brute forcing the solution – never give me three out of four numbers from a padlock for example, or all the numbers for a three digit padlock, but not the ordering. I don’t want to have to decide whether to “circumvent” one of the clues.

If you can, try to lead me astray when I’ve got most but not all of the clues. When I’ve got all the pieces, make it a puzzle I have to do a little bit of thinking about. I’m particularly fond of puzzles where there’s an interim step to the solution – words to colours, colours to numbers or similar, and talking to other players, I don’t think I’m alone.

Once I work out the answer though it should be clearly the right one. I should enter it in the keypad/combination lock etc, with absolute confidence it will work. The hallmark of a good puzzle is that it’s immediately obvious when you get a solution and you feel like you’re an idiot for not spotting it sooner.

Team work

Finally, involve as many of the team as possible. That might be in the mechanics of solving the puzzle – whether it involves someone in one room reading out a code for another player to input in a second room, two people controlling different aspects of a game or something as simple as multiple team members pressing a button at the same time. Alternatively, it might be in all the puzzle pieces you have to collect, or information you have to notice, which inevitably will be found by different people. I particularly like information, because it makes it much more likely that people have to work together, rather than just dump clue pieces in a corner for someone to put together.

The point here is that this is a team game, but so much of what we do is actually done in isolation – as a game designer, you should be trying to encourage us back to the team as much as possible.

Anything else?

That’s all I can think of. What about you?


  1. // Reply

    The bit you describe at the start of Progression we call ‘key for a key’. It’s without doubt one of the laziest and worst design decisions someone can make. The only ‘saving grace’ to it now is that rather than stand around and grumble about it our team has taken to running around/playing with the props, singing ‘key for a key’ (to no particular tune) while the person holding the key does the tedious unlocking.

    1. // Reply

      I like that idea. I’m going to come up with an annoying jingle for every bad trope we come across, and then get the team to sing it.

      Do you have something similar for UV? I’m going to start singing “scan the room” or maybe “turn out the lights”…

  2. // Reply

    I haven’t done one for UV yet, but would be perfect for that collective sigh when you find one and have to do a scan of the entire room.

    I’m thinking Blankety-Blank might be the tune for Key for a Key.

Leave a Reply

Your email address will not be published. Required fields are marked *