Are Complex Systems Learned by Interacting with Complex Systems?

Feb 28, 2011

Many videogames are complex systems. It follows, according to casual discussion of games and learning, that complex systems are what we learn by playing videogames.

This assumption leads to the idea that to teach a complex system, we should model it, and somehow make a game out of that model.

However there is a subtle though important distinction to be made about what is actually being learned: how to interact with a particular complex system.

SimCity teaches how mayors work, not how cities work.

In a simplified and highly accessible manner, the simulation puts players in a position to be the mayor, not in position to be the city. The challenge, and so the majority of the learning, happens in responding as a mayor must to the types of feedback a mayor may receive by adjusting the control levers (though exaggerated in this case in effectiveness for sake of illustration, understandability, and enjoyability) that a mayor can decide upon. Put another way: the player is tasked with translating the input a mayor receives into the output a mayor performs. The player is not tasked with translating the input a city receives (from its mayor, citizens, and environmental factors) into the output a city provides (its state of development, crime level, traffic congestion) back to those that provide it with input. So, again, the player is learning how to think like a mayor, not how to think like a city.

In the case of SimCity it maybe seems absurd to focus so much attention on the distinction, but in the case of trying to teach a complex system it’s the difference between learning the intended material and learning what amounts to the opposite of the intended material.

To try out another type of simulation example: a flight simulator teaches us to behave like a pilot, not how to behave like airflow, nor how to behave like a plane.

Admittedly, the two cases above seem fairly obvious, and both are straightforward examples on account of being literal simulations in the traditional sense. These demonstrate something different than modeling the complex system itself (the city, or physical aerodynamics and plane mechanics) as a means of better understanding of it. The part the player comes to understand through play – how to think like a mayor (again, wildly simplified), or how to operate a plane, is not the part modeled, it is the part left to the player, and the software is instead the complement of the player’s role.

One way to gain an understanding of complex systems is through exploratory experimentation, the scientific method, verifying or rejecting results as a means of verifying or rejecting hypothesis that we frame as part of our gradually refined model to make predictable sense out of what’s happening. We can make software to facilitate exactly that, built as an interface and means of interaction with a model of a complex system, even perhaps with fiction to motivate mastery of contextualized challenges, but by doing so, we’re right back to learning not about the thing modeled, but instead learning the player’s role opposite the model – it would become software about how to think like a scientist.

But the system can adapt to the learner’s needs, someone might counter, as an advantage unique to learning through games. Then, I reply, the game is teaching players to be reliant on having something outside themselves identify their weaknesses, and even worse, doing such an important thing (in many cases) silently without acknowledgment.

A videogame can be reset to an initial state, to be tried again in another way, for mistakes to be safely made without consequences, someone might offer as more general advantages (as I too have done before). Then, however, it allows unplanned, unconsidered, semi-random though perhaps reactively adapted behavior to brute force obstacles, a blind trial-and-error approach that rarely works outside the safety of the software.

Such behavior, as I identified in Real-Time Play: Tactical Patterns, or tried to explain more concisely but less clearly in Diagram: Player Actions and Memory, “is a way in which players improve at videogames without constructing a more accurate mental model of how the videogames operate.”

The challenges can be presented in multiple varying contexts, and recombined in unpredictable ways, which to a player with experience can become predictably unpredictable or unpredictably predicable, in either case having the learnable consistency to lead to a particular sense of symbol interpretation and decision/action prioritization. That is why I am turned my attention toward the designed experience for the player, rather than the system modeled, in Reading Between the Lines: Learning by Experience.

To teach a complex system, we don’t need to model it in software – that much is comparatively easy, pleasantly straightforward in literal cases, and I think the wrong direction entirely.

To teach the complex system we need to instead figure out how to develop the player experience as the complement of the complex system to be learned, so that the player’s success requires understanding, thinking, and operating as the complex system.

How does the city interact with the mayor?

Learn and practice team game development with Gamkedo Club.
Membership worldwide. Professional support. Proven process.

Subscribe by e-mail to receive weekly updates with Gamkedo.Community interviews and YouTube training videos for game developers!

All contents Copyright ©2017 Chris DeLeon.

Site production by Ryan Burrell.