Lesson 12 · Part 3: The Machinery

Emergence

How a handful of simple rules can produce a whole world of behavior, and why players imagine even more than you built.

11 min read

Emergence

You’re up in the cheap seats of a stadium with forty thousand other people, and a wave is coming. You can see it a quarter of the way around the bowl: a ripple of raised arms rolling toward you at about the speed of a jogger. It arrives, you jump up and throw your hands in the air (you’d feel silly not to), you sit, and it rolls on.

From the top row the wave looks like one enormous creature lapping the stadium. But nobody’s in charge of it. Nobody inside the wave can even see its shape. Every single person is following one tiny instruction: when the people next to you stand, stand up, then sit back down. The whole spectacle comes out of a rule short enough to print on the back of your ticket.

Level 11 made the case for building games out of systems, not scripts. This level is about the strange bonus you get when you do. A system can give you back far more than you put into it, and then the player’s imagination adds even more on top.

More comes out than goes in

The stadium wave is a case of emergence: complicated behavior in a whole that comes from simple parts following simple rules, with no part holding the plan. A flock of starlings swirling at dusk is emergence. So is a traffic jam that creeps backward down a highway while every single car in it is trying to go forward.

Ants are everyone’s favorite example, and for good reason. A single ant is close to a tiny robot. Smell this chemical, do that. Smell that one, do something else. There isn’t a lot more going on in there. The biologist E. O. Wilson spent a career working out those rules, and what he found is wonderful. Put enough of these little robots together and the colony finds the shortest path to food, sends extra workers where they’re needed, and picks a new home. No ant knows any of that. The cleverness doesn’t live in any ant. It lives in the system.

For a designer, emergence is a fantastic bargain. You write a handful of rules and get a whole world’s worth of behavior back, including situations you never thought of. The catch is that you don’t get to pick exactly which behavior shows up. More on that in a minute.

A grid and two rules

The cleanest demonstration of emergence needs nothing but squared paper and a pencil. A cellular automaton is a grid of cells where each cell has a state, like on or off, and decides its next state only by looking at the cells touching it. The stadium is a rough one: each seat watches the seats beside it.

The most famous cellular automaton is the Game of Life, invented by the mathematician John Conway. Each square on the grid is either alive or dead, and each has eight neighbors, counting the diagonals. Two rules cover everything:

  • A dead cell with exactly three live neighbors comes alive.
  • A live cell with two or three live neighbors survives. Every other live cell dies.

Every cell updates at the same moment, and then they all do it again.

That’s the whole game. It sounds like nothing much could come of it. Then you sprinkle a few live cells on the grid, let it run, and it goes bananas. Some shapes blink in place forever. Some sail diagonally across the grid like little ships. Some crash into each other and throw off debris that assembles into new shapes. People have been poking at this tiny universe for more than fifty years and still find new creatures in it.

Notice there are no dice anywhere. Life is pure determinism: start from the same pattern and you get exactly the same history, every time. Even so, Conway couldn’t have predicted from his two rules what the grid would get up to. The only way to find out what a pattern becomes is to run it and watch. So a world with no luck in it at all can still surprise you, including its inventor.

Fewer rules, deeper play

Ask a room full of game designers to name a game they admire and someone will say Go. It’s a board game, thousands of years old, with black and white stones on a grid, and its rules fit on an index card. A child can learn them in an afternoon. The strategy that grows out of them has kept people busy for entire lifetimes. Go and Life have the same quality, and plenty of designers hold it up as the thing to aim for.

“What is the simplest set of rules that you can create that leads to the widest range of complexity…”

Will Wright

This course calls that quality elegance: a lot of depth out of very few rules. Think of it as a ratio. Tic-tac-toe has very few rules but very little depth, which is why you stopped playing it around age nine. A sprawling board game with a forty-page rulebook might be deep, but it’s paying a fortune in rules for it. The sweet spot is a short rulebook and no bottom to the strategy.

Elegance pays off in very practical ways. Fewer rules means less to learn, so onboarding is quick. There’s less to build and less to break. And every rule has to pull its weight by making the other rules more interesting. A rule that just sits next to the others, adding one more thing to remember, is dead weight. Cut it.

Keep the hood closed

Here’s something odd about emergent worlds: players nearly always think they’re more complicated than they really are.

When SimCity, a computer game where you build and run a city, first came out, its designer Will Wright got letters from players describing what had happened in their towns. A plane crashed here, which knocked out a power line, which caused a surge over there, which started that fire across town. Long, detailed chains of cause and effect. Wright knew the code, and he knew none of those links existed. The players had built their own explanations, their own mental models, and nothing in the game told them they were wrong. The cities in their heads were richer than the one on the disk.

He called this the simulator effect: players imagine a simulation is more detailed and clever than it actually is. His advice about it is short and a bit cheeky. Don’t talk them out of it.

To make that work, the machinery has to stay out of sight. Engineers call a device like that a black box: you can see what goes in and what comes out, but not what happens inside. A vending machine is a black box. So is a Magic 8-Ball. So, frankly, is a cat. If you showed players every formula, the world would shrink back to its true size and the spell would break.

Scaffolding, and the price you pay

Keeping the hood closed doesn’t mean keeping players confused. Level 9 drew the line: when the player does something, the game’s answer has to be clear feedback. What stays hidden is the plain arithmetic underneath.

What the imagination needs is scaffolding, enough visible, consistent cause and effect to build a story on. The little people’s gestures and tone of voice are scaffolding. So is a plane that visibly crashes near a power line. A world that answers inconsistently, or gives no hints at all, just produces confusion, and confused players don’t tell a story about what happened. They shrug and wander off.

Emergence also sends you a bill. Because the behavior grows out of the rules, you can’t work backward from the result you want to the rules that will produce it. Simulation designers know the feeling well: you sketch what you expect the traffic to do, run the thing, and get something that looks nothing like your sketch, with no obvious knob to turn. The way through is to make changing the rules cheap, so that tuning becomes change something, run it, look, and go again, dozens of times in an evening. Level 17 is all about that loop.

Try it · Pocket Game, step 12Break a rule

Play your Pocket Game three times. First, remove one rule completely and play a round without it. Second, put the rule back, pick one number in the game (pieces per player, tries per turn, points to win) and double it. Third, play it as it was. After each round, write one line about anything that happened that you didn’t expect. If a broken version gave you a better surprise than the original, keep the change. You’re hunting for the smallest set of rules that still gives you something to think about.

Carry forward: Look for the fewest rules that give the most to explore, and let the player’s imagination finish the job.