Fate of Io
Good Puzzles vs Bad Puzzles
2004/01/24 21:29:41 PST by Temporal [manager]
[Temporal's avatar]

Lately I've been playing some of the games in the Myst series, and from them I have learned a good deal about how to make good puzzles. This is all relevant to Fate of Io, of course, so I thought I'd post what I've learned here.

There are two fundamental principles to making good puzzles:

  • A good puzzle should never require that the player perform an exhaustive search.

  • The player should know that they have the correct solution to a good puzzle before they try it.

Avoiding exhaustive searches

Good puzzles are solved through insight. Bad puzzles are solved through trial and error. The worst thing you can do in a puzzle is force the player to perform an exhaustive search. What is an exhaustive search? This can mean a number of things, such as:

  • Searching through a large area of a map looking for a hidden object.

  • Testing a large number of switches to see which one does the desired action.

  • Reading through a large amount of text to find the answer to a puzzle.

  • etc.

There are some implications of this rule that aren't so obvious, though, so I'll discuss some more specific cases.

The result of a particular action a player takes should always be visible or somehow indicated to the player. For example, say a player flips a switch which unlocks a door in another room. If no indication is given that the door is now unlocked, then the player is forced into an exhaustive search to find what the switch did. This is bad. The switch should either be clearly labeled, or some other indication of what it does should be given when the player uses it.

Puzzle designers often like to make clues to the puzzle as obscure as possible. This is not as good of an idea as it sounds. First of all, if a clue is placed in a location where the player wouldn't expect it, then the player is again forced to do an exhaustive search just to find the clue. Therefore, it should always be possible to track down clues in a directed manner. Furthermore, once the player finds the clue, it shouldn't be too hard to figure out what it means. What seems obvious to the puzzle designer is not always obvious to the player. For instance, just because there's only one object in the current level which has a hexegonal shape does not mean that simply drawing a hexagon on a clue is enough to make the player understand that the clue relates to that object.

Deriving answers logically

The solution to a good puzzle is always found through logic and reasoning. The player should always know that they have found the correct solution to a puzzle before they test the solution. They should know this because they derived the solution logically, rather than simply trying things randomly.

Now, don't confuse what I'm saying here. I am not saying that the player should always succeed when they think they have the right solution. I am saying that they should never succeed unless they think they have the solution. If it's possible to just accidentally solve a puzzle without figuring out how it works, then it's not a very good puzzle (either it is too easy or it required an exhaustive search).

Making a puzzle that requires insight and deduction to solve can be tricky, but it's not as hard as you might think. Start out by giving the player a set of basic tools or abilities. These abilities form the building blocks of the solution to the puzzle. The player's goal is to determine how to use them together to create a solution.

Now, the abilities given to the player should never be hard to understand. It is ok to force the player to test the abilities in order to determine their function, but the function of each ability should be easy to determine. As mentioned earlier, the results of pressing a switch should always be clearly indicated to the player. The player's job is not to determine what the abilities do, but how to combine them effectively.

In Fate of Io, most of these abilities will literally be the character abilities. However, that won't always be the case. Some situations will present other abilities to the characters, such as flipping switches to cause specific events and whatnot.


Myst III: Exile contained some very well designed puzzles, particularly in the world with the giant marbles. In that world, there were several sort of control panels which controlled various machines in the path of the marble. You would use the consoles to set up the machines, then press a button to send a marble through. It was very easy to see how the console affected the marble, but it took some deduction to figure out how to set it up right so that the marble passed through. Here, the "abilities" were the different ways you could manipulate the machines. It did take some trial and error to determine what each ability did, but this was ok, because there were very few abilities, and the results of each one were obvious once you tried it.

Probably the absolute worst puzzle you could ever put in a game is one in which you have to search a large area for a particular object, with no clues as to where the object might be. In Uru: Ages Beyond Myst, you are apparently required to search out seven small pieces of cloth in each area of the game. To make it as bad as possible, these pieces of cloth are often tacked on to the back sides of rocks, hidden in dark corners, or even placed on the outside of doors which can only be closed from the inside. This is, quite frankly, an example of horrible puzzle design. However, it is easy for developers to fall into the trap of creating such puzzles without thinking about it. Unless you've spent a lot of time thinking about it, undirected treasure hunts sound like perfectly good puzzles. Let's not fall into such traps.

fateofio.org © Copyright 2001-2005 Sam Pierce, Kenton Varda, and contributors
Powered by Io Community Manager, Evlan, and FreeBSD