Index |  Articles |  Links

By Mark Damon Hughes <kamikaze@kuoi.asui.uidaho.edu>

   To me, a good puzzle is one where: There's a problem set up by the author or emergent from the world; if designed, the author has presumably thought of at least one solution (while you can get away with not doing this in pencil & paper RPGs, it doesn't work so well without a human GM to fudge a solution if all else fails...); the solution is possible with the skills and objects and information available in the game universe, must make sense (even if only after the fact), and preferably requires no special-case coding to implement - the physics of the universe should be all the mechanics it needs. I'm dubious but sometimes tolerant of "undead puzzle-solving" - being expected by the author to try something reasonable, get killed or foiled unexpectedly and get some information, and then restore and use the information from your death to solve it. If done with humor or drama, they can be effective instead of annoying. A lot of Infocom games had these things, and IMO they're more a literary device than actual puzzles.

   Other kinds are, to me at least, not "puzzles" so much as random guessing games. I'm not adamantly opposed to puzzles with only one solution logically possible, but they seem rather weak and artificial - in the real world, at least you DO have the option of smashing the puzzle box or taking your Rubik's Cube apart and reassembling it or cutting the Gordian Knot. If there's only one path through and it's artificial, then it seems like I'm just doing a walkthrough - reading a book with hard-to-turn pages instead of playing a game. I guess there's an audience for that, I just don't get them (and they won't like my games, I expect).