Portal to Fun
I saw first a demo movie and then a few screens of actual play for the recent Xbox 360 game Portal. I’m eager to borrow it and try it myself. Although it began as a design augmentation to the beloved first-person shooter Half-Life, Portal is essentially a puzzle game. There are no bad guys between you and the exit to shoot, nor a proper weapon to shoot them with; instead, you must cross various walls, pits, force fields, and similarly immobile obstacles, mostly through the use of your portal guns.
After the first few screens, you are introduced to first one, and then to a pair of portal guns, which you can use to create a linked pair of teleportation gates in the walls, ceiling, and floor. To give a very simple example of how to use the guns, you might need to cross a lava-filled chasm to reach the exit. You could shoot the blue portal gun at the wall behind you, then shoot the yellow portal gun at a wall on the other side of the chasm, then step through the blue portal to appear at the yellow portal. Voila! Naturally, the game is more complicated than that; often, you’ll need to employ chains of well-placed portals to move from point A to point B via points X, Y, and Z. You must also trigger switches to turn various room elements, like force fields and doors, on and off. You can use crates to trigger objects while you go stand somewhere else, and you can employ the portals to manipulate crates and other room elements.
What makes this game so intriguing is the creativity behind it. The whole project began as a new weapon for a shooter game. Like the gravity gun—a sort of telekinetic hook—for the same game, designers and players alike became more interested in what the new toy could do than in the general mayhem which the toy was originally meant to aid. Once they had the code in place to do something simple with the portal guns, designers got sidetracked by all the creative uses to which they could put it. Hey, look at this! I can use a portal in the ceiling to drop a crate on that guy around the corner. Hey, look at this! I can use a pair of portals in the floor to catch that guy in an inescapable loop, falling back and forth between portals. Hey, look at this! I can use two portals to see four copies of myself at once. Hey, look at this! I can use a portal to create a waterfall coming out of the side of that skyscraper.
And so on. The puzzles are neat because they started out as something neat, and a puzzle was built around the idea of taking advantage of that possibility, instead of employing a top-down design of creating the obstacles first, then adding the tools to bypass them. Any trick that was not sufficiently neat either appears as part of the early tutorial screens, or simply didn’t get included at all. Because every puzzle began with “Look what I can do!” allowing a player to duplicate that discovery is fun. Because every puzzle began with “Look what I can do!” there is every possibility for a player to think of some other use the designers didn’t. That’s a real payoff for creative thinking.
Contrast this with a puzzle in which starts with a trench that can’t be crossed, but must be crossed. The designers add a large lever on the wall; pulling it teleports the player to the other side of the trench. Problem solved, but in an artificial fashion that does not leave a player feeling like he’s doing anything more than going through a pre-programmed flowchart. Instead of playing, he’s following directions. Not fun.
This kind of exploitation of an existing game engine to do things it wasn’t really meant to do has a short but glorious tradition—short because computer games haven’t been around all that long. You can see it in the YouTube videos of game characters acting out short comedy sketches, and vehicular stunts initiated by blowing up your own car in such a way that it flips through the air. You can see it in the way the Katamari begins to tear up chunks of the countryside in Katamari Damacy. (I am morally certain that began as an unintended effect, but was left in for being silly and fun.) SimCity began as a utility for designing a bomber plane video game; Will Wright decided it was more fun placing buildings down as he designed the background countryside than it was to fly over them, blowing up bad guys. The mathematical study of cellular automata like Conway’s Life (not, technically, a game) began to mean something only after students began to play with the idea of crashing self-moving shapes called “gliders” into one another in a controlled way. Lord knows how many easter eggs are designed to reward players for doing something just because they can.
This is the essence of computer programs as toys, rather than computer programs as games. Conceiving an activity and building an engine to replicate it may or may not work. Building a robust engine first almost guarantees that you can do something fun with it.