Making Simulation Games for Education: Emergent Systems

What makes an educational game work? What keeps the wheels spinning in a player’s head, keeps them engaged and learning? Well, if you’ve read my other musings, you’d probably be quick to suggest it’s the same principles that make commercial games successful. While this is certainly true, it’s not the most informative advice. I thought it would be useful to get a little more specific and concrete by reflecting on my experience designing Iridescent’s newest physics simulation game, The Gravity Ether.

So we’ll talk about simulation games, the genre which seems to have the most educational promise. And we’ll talk about the principle key to making a simulation game work: emergent complexity.

Meaningful interactions through emergent complexity

In their landmark book Rules of Play, Salen and Zimmerman argue that game design is really about meaningful interactions. A game is successful and engaging when it creates meaningful interactions for its players. Now, there are many ways to create meaningful interactions, which is why their book is so long, and also why we have so many different genres of games. For a simulation game, interactions tend to be meaningful when the simulation has emergent complexity.

What is emergent complexity*? In an emergently complex system, there is a relatively simple set of rules or interactions that govern the nature of the simulation, but the way that those rules play out leads to interesting patterns and processes that “emerge” from the simulation. An emergently complex simulation thus has elements and patterns that are not immediately obvious from the rules alone, which makes exploring and manipulating those unexpected patterns a very meaningful interaction. The fact that the underlying rules are pretty simple makes it easy for a player to understand the basics of the simulation, so the game doesn’t feel random and out of their control. But the number of elements and variety of interactions between them is complex enough to lead to surprises and a somewhat unpredictable end result. The real challenge in a simulation game is including enough complexity so that the game doesn’t feel predictable, but not so much complexity that the game feels random. That ideal middle ground is emergent complexity.

Salen and Zimmerman also note that uncertainty is vital for meaningful interactions: at a very basic level we only want to play games in which there is an uncertainty about whether or not we will win. Why play a game you know you will win? Randomness in games, like dice rolls, can create uncertainty. But more important than having actual randomness is having a feeling of randomness. This is exactly the case for emergent complexity–none of the patterns that emerge from a complex system are truly random, but because they can’t be predicted by the typical player they feel a bit random. This gives emergently complex games an edge or feeling of uncertainty, which helps craft meaningful interactions.

* Note: the Intelligent Design folks have co-opted the term emergent complexity for their own purposes. Their use of the term is not related to its use here.

Case Study: Angry Birds

So let’s switch to Angry Birds for a second. This game achieved success because it hit emergent complexity on the head. Other projectile motion games tended to be too simple and predictable. There is some complexity in choosing the angle and direction at which you launched an object. And there is some complexity to how blocks fall. But it wasn’t nearly complex enough, and the games felt too predictable. So Angry Birds added the extra element, the tapping on the screen to make the bird do something crazy. Since you could tap at any point in the trajectory to initiate the bird’s crazy motion, an interesting interplay is added between the type of trajectory you choose and when in the trajectory you activate the bird’s ability. It was often unclear how these two interactions would play out, requiring you to try it to find out. This added more complexity to the way you knocked down the structures, creating a game with more meaningful interactions.

Case Study: The Gravity Ether

Of course, creating crazy bird motions that defy laws of physics is not the only way to add complexity to a system. In the Ethers game series that we are producing here at Iridescent, we limit our search to natural phenomena when adding complexity to a system. Physics on its own is ripe with emergently complex systems. For example, we have the three-body problem in classical mechanics. Although from the Universal Law of Gravity it is quite easy to predict the motion of two planets, adding a third planet to the mix makes simple predictions of motion nearly impossible. Each planet depends on the position and motion of the other two planets to determine its own motion. But its own motion will also affect the position and motion of the other two planets, creating a system which is too complex to be solved for. There are regular patterns that emerge from the interaction of these planets, but those patterns are results of emergent complexity, rather than easily predicable phenomena.

We used this existing emergent complexity in gravitation forces to create The Gravity Ether. In this game, a player can create and remove black holes with a tap of their finger. These black holes cause gravitational fields which affect the motion of planets. By correctly manipulating the motion of planets, players can achieve simple level objectives, like breaking blocks or collecting coins. This is a system based on a simple rule: planets are attracted to each other by the Universal Law of Gravitation. With one stationary planet and one black hole, it’s relatively easy to predict what will happen–the planet will accelerate toward the black hole. But take a typical level that has 3-4 planets moving in different directions, with one or more black holes in play. Can a player exactly predict the motion of each and every planet? No, not really. Of course there are general patterns that emerge–planets tend to end up clumped together, circular and elliptical orbits appear quite frequently, etc. But these patterns are emerging from a sufficiently complex system, and thus allow for meaningful interactions.

Types of Emergent Complexity

As noted by the two case studies, emergent complexity can appear in many forms. In the first case study, an extra rule, or type of interaction, was added to the typical projectile simulation game to create emergent complexity. In the second case study, additional elements were added to the system to create more total interactions, resulting in emergent complexity. These are the two general pieces that make a system: elements and interactions. In a simple sense, complexity results from having a lot of connections between the different elements in a system. Add more elements or more interactions between existing elements, and your system becomes more complex. At a certain point of complexity, your system will start to exhibit emergent patterns. There is no general rule for how many elements or interactions are needed to make a system emergently complex: it is in part a subjective judgment that varies from system to system, and is one of many pieces that adds an artistic element to game design.

Emergent Complexity and Learning

So, let’s summarize.

Simulation games need emergent complexity to have meaningful interactions.
Emergent complexity results from including a lot of (but not too many) elements and interactions in a system.

The third piece of the puzzle is what happens when a player engages in such a game. The more we play with an emergently complex system, the more adept we become at predicting how the system will react to different patterns and manipulations. It becomes more intuitive. Players are naturally drawn towards understanding the simulation better and resolving the uncertainty present in the emergent complexity. Players begin to grok the system, they understand how to cause different emergent patterns to appear and how to strategically use those patterns to beat the game. Another way of putting this: players start to gain an intuitive understanding for how the complex system operates. So, to add in this third piece and complete the puzzle of learning in simulation games:

Simulation games need emergent complexity to have meaningful interactions.
Emergent complexity results from including a lot of (but not too many) elements and interactions in a system.
As players become drawn into the game, they gain an intuitive understanding of how the complex system in the game operates.

What does this mean for teaching students educationally valuable stuff? Well, I just argued that through playing an emergently complex simulation game, a student gains a better understanding of the simulation itself. If the elements and interactions of the simulation are real, or reflect real life elements and interactions, then students are gaining an intuitive understanding for a real life system by playing the game. Thus we end up with two prongs.

A simulation game is engaging when it is emergently complex.
A simulation game is educationally valuable when it contains real elements and interactions.

This leads to my biggest reflection after designing The Gravity Ether:

Being an educational game designer for simulation games just means being a good selector of which real elements and interactions to include in the simulation to achieve emergent complexity.

The key component is being committed to only using real elements and interactions. Most games will start with some sort of real physics, but then add in unrealistic actions, which help create meaningful play but reduce the educational value of the game. The crazy motions undergone by Angry Birds defy the conservation of momentum, and thus do not help a player build an intuition for real physical laws. To me, being an educational game designer is about not making such compromises. It’s about finding and using the emergent complexity that already exists in real physical systems.