KEEPING IT SIMPLE

a long-winded article about efficient game design
by James Ernest

Let's say you've got a block of marble, and you want it to look like a horse. There are essentially two ways to do this: you can add enough clay to build the marble up into the shape you want. Or you can take a chisel and chip away everything that doesn't belong, until nothing is left but the horse.

Now, it's possible to compromise, adding a little clay here and chipping away a little stone there. You¹ll wind up with something in between the two horses, with enough clay to finish out the shape, and enough marble to remind you what you started with. Of course, nobody really makes sculptures like this, but for some reason it¹s a good metaphor for game design.

The marble slab you start with includes everything you¹re sure you want in the finished "game." It's the back story, it's the way you want the players to interact, it's the tools you want to give them. And it's all the unconscious connotations of "game" which you bring to the table. The clay is the rules you must add to your core ideas, to make them into a functional whole.

If you've ever tried to design a game with some depth, you've probably caught yourself adding a little more clay than you needed, either to improve the shape of your original slab, or to cover problems in an older layer of clay. Unfortunately, there's no upper limit to the amount of clay you can slap on, although your game can become unwieldy pretty early in that process. Maybe you gave up when the game got too complex to play, or maybe you decided it was finished because it just couldn't sustain any more "fixing." Either way, most beginners who try to fix one rule by adding another, eventually become frustrated when their games bog down without really becoming more interesting.

Yet some of your favourite games, and many of the "classics" you grew up with, seem to have almost no rules at all. The rules for Chess fit on the inside of the lid. So why are there so many volumes written about how to play it? In my opinion, the strategy in Chess is a product of the simplicity of the rules. This is a game which has evolved through generations of play-test, and so these aren't just any rules, they are exactly the right rules and nothing else. Can you say that about your latest game?

As you develop a game, keep asking yourself: is this new rule carrying its weight? Am I asking people to remember this rule just because it fixes another rule? Or does it genuinely add strategic depth to the game? You will need to ask these questions about your basic principles, too; are you assuming something unnecessary, without even knowing it? Deep inside that marble slab are rules you don't even have to write down, like "players are trying to win" and "cheating is against the rules." And there are other "rules" about game design which you often need to hunt down and bring to light, like "the game should be fair." You probably follow these rules without even knowing they are there, which makes them impossible to break, even when you want to. In any case, you shouldn't include anything unnecessary, if you really want your game to work.

Let me give you an example. I¹m working on a game called "Spree!" in which the players are looting a shopping mall. They are all carrying stun guns, because I want Spree to be a quasi-wargame, with missile fire being a vehicle for player interaction. That basic intent, along with some other game conditions and my preconceived understanding of "games," is the marble slab on which I begin.

"To shoot or not to shoot" is one of the strategic choices I want in this game. There should be definite rewards of relatively equal value for shooting at someone, and not shooting at someone, to make the choice difficult whenever it is presented. If the choice is always clear (i.e. always shoot if you can) then it¹s not really a strategic choice, it's just another rule. If the choices are always functionally equivalent, it's not much of a strategic choice, either; "Heads or Tails" doesn't require much thought. Ideally, each time the option is presented, the player should have to do a little analysis to decide whether shooting at someone is the right thing to do.

There are other strategic choices in this game, too, like deciding when to run to your car, when to rob a standing player, how best to take your move, and so on. But let's talk about the evolution of the shooting rule.

Here's the rule I start with: if you take a shot, it is the last action in your turn. Roll a die to see if your shot hits; the closer someone is, the easier they are to hit. Players can play defensive cards to make your shot fail. If you hit your target, they fall down. If you miss, you fall down instead. While players are lying down, they can't play cards. Specifically, this prevents them from playing defensive cards when people walk over and try to rob them. Fallen players also can't be shot at. On their next turn, fallen players stand back up, and proceed.

All of this is pretty intuitive; once you've heard it, you don¹t have too much trouble remembering it, which makes the rules fairly lightweight.

However, after a little development, I begin to see this problem: because it¹s the last action of my turn, when I shoot someone, they are going to have the chance to get up and run away before I can come and steal anything from them. Shooting that player has only made them easy prey for the players who go just after me, but not for me. That's hardly enough incentive to use my weapon.

The first solution: well, if you've been shot, you lose a turn. That gives the player who shot you the chance to come and hit you before you move. Still, the other players get the same chance, and those who got one chance before now get two.

So, we invented "dizziness" to prevent people who were in the process of losing a turn from being defenceless for too long. You stand up after your lost turn, and now you can play cards, but you're "dizzy", so you can¹t be shot at again. This prevents an infinite loop in which one player repeatedly shoots another, takes a turn, and then returns to shoot him again. But it fails to address the niggling central problem: shooting someone is still better for the other players than it is for you. So there's still no real incentive to use your gun.

Now the fixes start to get so ridiculous that I suspect a core problem. We are suggesting things like, "only you can steal stuff from the guy you shot." "you can shoot in the middle of moving but not more than once," and here¹s my favourite: "There's a special card in the deck that allows you to (insert overcomplicated patch for rules here)." Oh yeah, I love that one.

By now there's a lot of crap piling up on top of this non-functional rule, and shooting still doesn't function like it should. Rules that we have added to compensate for the deficiencies in earlier rules are failing to do so, so the game is becoming more complicated without really improving. The choice is becoming "difficult," not because there are several attractive paths, but because it's just incomprehensible.

Finally it occurs to me that the real problem in this mess is a big chunk of the marble, a preconceived notion of mine which says "players shouldn't be allowed to earn extra turns." An extra turn would have been a great reward for a successful shot, if only I hadn¹t dismissed it as "too good." This is an insidious section of my core principles, because I don't even realise that it's there. I'm just convinced that players should never get extra turns. After working with TCG's (specifically Magic) for so long, I have somehow become convinced that every unregulated extra-turn scenario leads to infinite regression, degenerate game conditions, and hundred-dollar trading cards. Now, ignore for a moment the fact that I've played Monopoly, and a dozen other board games with "roll again" conditions. Forget that I wrote Kill Doctor Lucky, in which players can easily earn three turns in a row. I'm still subconsciously sure that extra turns are bad bad bad.

When I suggest an extra turn, the play-testers' first reaction is pretty much the same as mine: "Oo, we don¹t know, isn¹t that dangerous?" But with a little play-testing we realise that breaking that "rule" and rethinking my original notions lets me chip away mountains of superfluous shooting rules, and also makes the game work smoothly. You shoot somebody, they fall down, they can't play cards, you take an extra turn, and presumably you go rob them. That's pretty much it.

The rules we cut away? The victim doesn't need to lose a turn. We don't have to keep track of who's "dizzy." No more extra rules about who can do what to whom when, just old "the guy lying down can¹t play cards." And no special card in the deck that lets you do something wonderful. We also exorcised all the game problems which arose from those rules, including the overwhelming fact that shooting someone just wasn't very attractive. And as it happens, this particular extra-turn rule doesn't lead to an infinite loop, because the engine quickly runs out of steam: once you've shot someone, they fall down. You can't shoot them again, so you can't earn another turn off that player. Even in the longest shooting spree, everyone falls down eventually.

When I'm done with Spree, it'll be a smallish clay statue, with a good amount of marble showing through; that's about what I was hoping for. Its just complex enough to make people think, but not too bulky to be understood. It's not Chess, but then it was never supposed to be. For one thing, Chess isn't very funny. For another, you never get to shoot anyone.

James Ernest is a freelance writer and game designer and the head honcho at Cheapass Games. He lives happily in Seattle on the money he earns by being married to a rich person.