All's Not Fair

Making fair puzzles is something of a black art. What's fair for one person is by no means fair for another. Even if I restrict my definition of "fair" to me, that definition will change from day to day. What mood I'm in plays a large role in how forgiving I am of puzzles which may straddle the line between fair and not.

Nevertheless, I think there are some general rules-of-thumb which I unconsciously use to determine whether or not a puzzle is fair. Unsurprisingly, most of them have to do with what kind of knowledge the puzzle expects me to have access to. I believe that a puzzle is fair if it:

• Doesn't use uncommon common knowledge. It's all too easy to assume that some certain bit of knowledge is common to most everyone when it's not. This is not to say that a puzzle can only use knowledge that most people are guaranteed to know. Using knowledge that can easily be looked up is perfectly fair, as long as the game indicates that it expects me to do so. A slight refinement of this requirement leads to:
• Doesn't depend too heavily on cultural knowledge. Do I need to know about the five-second guarding rule from college basketball to solve a puzzle? If so, it's probably a bad one. The baseball diamond puzzle from Zork II is a good example of one which is well-nigh unsolveable for many non-Americans due to the cultural knowledge required to solve it. If a puzzle does depend on some bit of cultural arcana, clues should be given to help the uninitiated.
• Gives clues. This is not an absolute necessity, but it makes puzzles more fair, especially if they use their own logic.
• Fits into the setting. Once upon a time, it wasn't uncommon for designers to put Coke machines in the middle of a dungeon. While anachronisms like that aren't too bothersome, anachronistic puzzles can throw me for a loop if I'm not expecting them. Having to switch gears in the middle of a game about magic to solve algebraic equations is a little annoying.
• Handles alternative non-solutions gracefully. Often there are possible solutions to a puzzle which don't quite work. In those cases, the game should indicate that you're on the right track, but that you don't quite have the right idea. Otherwise, I'm likely to dismiss the right solution because it's so close to the wrong one I just tried.
• Doesn't require knowledge from beyond the veil. It should be possible to solve a puzzle without having to die to find out some crucial bit of information.
• Makes sense after I've solved it. Don't you hate it when you solve a puzzle and you still don't know why your solution worked?
• Warns me it's unfair. If a game warns me that it's easy to render the game unwinnable then I'm more likely to save my game and to think that unfair puzzles are, in some weird, twisted way, fair.

Are there other criteria? Sure, but these are the most general ones I could come up with. Are these hard-and-fast rules? Of course not. You can find examples of good puzzles which violate one or more of the above. Like I said at the beginning, these are just rules of thumb. But if a puzzle is going to violate any of these rules, I'd suggest that the designer be sure that they have a good reason for doing so.