Begin With General Systems Theory
General Systems Theory provides an established way to study interacting elements as a whole. A system's behavior depends on relationships among its parts and with its environment.
For game design, that directs attention beyond individual cards toward how the design works together.
The system is what we choose to examine. Its boundary separates it from its environment. Inputs cross inward; outputs cross outward. Processes transform inputs and existing state into new results.
Those terms become useful when we say exactly what they refer to.
In this article, I am examining each game's formal state and procedures: cards, piles, active effects, objectives, and the rules that change them. The people making decisions and interacting socially are outside that analytical boundary.
This is a deliberate model of part of a game session.
It is not a claim that people can actually be removed from play.
Open and Closed Need a Defined Exchange
The terms have technical histories. In thermodynamics, an open system exchanges matter and energy with its surroundings. A closed system excludes matter exchange but can exchange energy. An isolated system excludes both.
Bertalanffy's discussion of living systems emphasizes the continuing exchange that sustains organisms.
For the game comparison here, I am applying the boundary-and-exchange question to information that affects the formal game state. I am not treating cardboard as a thermodynamically isolated object.
A comparatively closed game engine resolves play mainly from its existing state and specified procedures. An open engine accepts consequential information from its environment through defined channels.
That distinction requires two answers: what is inside, and what exchange are we studying?
Without those answers, calling a game open or closed tells us very little.
WAR: A Comparatively Closed Engine
WAR gives me a useful example of comparatively closed resolution.
Its physical rules establish card values, conditional effects, the handling of ties, and the movement of cards between piles.
Players reveal cards. The engine determines what happens next.
A player cannot decide that a historical unit deserves a higher value because they admire it. A persuasive speech does not change a battle result. A bargain does not replace the printed tie procedure.
Players still have reactions, memories, and preferences.
Those belong to the surrounding experience. They are not ordinary inputs to the battle calculation.
There is a qualification: shuffling introduces uncertainty. If a model takes each new shuffle order from outside, that order is an input. If the model includes its randomization procedure and the random outcomes used, subsequent automatic resolution can be studied within that model.
Randomness does not settle the boundary question for us.
WAR is comparatively closed here because its resolution admits few ongoing player decisions or social judgments, not because its future is always predictable.
CHAOS: Defined Channels From the Table
CHAOS gives me a different relationship between the formal game and its environment.
Its developing card set asks players to choose cards, targets, and exchanges. Those decisions enter the formal state through the rules.
The production candidate also contains a particularly clear example: Straight Face.
The card requires a draw of two cards, with an additional card if its player smiled during the designated observation window.
A facial expression is not a value stored in the deck.
The rule observes behavior at the table and translates it into a change in a hand.
From the boundary chosen here, that is consequential information crossing inward.
The new hand then changes what the player can do. That result may influence their next decision or reaction. Information moves back toward the people at the table, who respond with further inputs.
This is why CHAOS is my open case study.
The design gives the social environment specified ways to affect the formal engine.
The example comes from a development candidate. It explains a design mechanism; it does not establish that the mechanism has already produced a balanced or enjoyable finished game.
Changing Objectives Is Not Enough
CHAOS also changes active win conditions and uses a changing Chaos Level.
Those are important parts of its behavior.
But an internal rule can change an objective without receiving anything new from the environment. An automatic program could do the same thing.
Changing conditions therefore do not, by themselves, prove openness.
Instead, ask what caused the change.
Was it an internal procedure? A player choosing a card? An observed action? A judgment that the rules explicitly require?
The difference is the path through which information enters.
That path deserves as much attention as the resulting change.
Feedback Can Exist in Either Game
Feedback occurs when a result returns to influence later behavior.
In WAR, the result of a battle changes card ownership. When collected cards return to circulation, earlier results influence later battles. Printed effects can also connect one battle to another.
That can happen within the formal engine.
In CHAOS, a changed hand or objective can influence a player's next choice. Their choice then changes the state again. With players outside our chosen boundary, this loop crosses between engine and environment.
Both examples involve feedback.
Only one needs that particular exchange with players to complete the loop.
A feedback loop is not the definition of an open system.
It is a relationship we can trace after establishing the boundary.
Emergence, Adaptation, and Complexity
Systems thinking also asks what appears through interaction: emergence. Complexity concerns the relationships that make a whole difficult to understand by examining its parts separately. Adaptation concerns adjustment in response to changing circumstances.
These ideas give me questions for each design.
In WAR, how do card circulation and conditional effects combine across repeated battles? Does a pattern arise that is easy to miss when reading a single card?
In CHAOS, how do players adjust when their hands, available targets, or objectives change? What patterns arise when several people respond to each other's choices?
Those are questions to investigate through analysis and playtesting.
An unexpected result is not automatically evidence of deep emergence. A game named CHAOS is not automatically an example of mathematical chaos.
The terms become useful when they help identify a relationship we can actually examine.
Look for Stability Without Assuming It
Equilibrium and stability raise another design question: does play settle into a condition, return after a disturbance, or continue changing?
For a card game, we need to specify the condition before making that claim. It might concern card distribution, repeated patterns, or the available routes to victory.
A recognizable turn sequence does not prove equilibrium.
Nor does frequent disruption prove that the whole system cannot develop a stable pattern.
For CHAOS, I would watch whether changing objectives create fresh decisions or simply make earlier decisions feel irrelevant.
For WAR, I would watch whether automatic circulation produces momentum, repetition, or an excessively long finish.
Those observations can reveal unintended consequences: results the design permits that the designer did not intend.
Openness Needs Procedures
Once a game accepts social behavior as an input, the designer has a practical responsibility.
What counts as the behavior?
Whose behavior matters?
When does observation begin and end?
How is a disagreement resolved so play can continue?
A smile-dependent card needs an observation window. A targeted effect needs a legal target. A player choice needs a point at which it becomes final.
The boundary can admit interaction while remaining precise.
Otherwise, the mechanism that invites the table into the game can also invite repeated argument about what happened.
Change the Boundary, Change the Analysis
Suppose we include the players, conversation, and room in the system under study.
Now the decisions and smiles previously treated as external inputs are internal events.
We would have to examine exchanges with a wider environment instead.
The classification changes because the question changes.
Every real game session exchanges energy and interacts with its surroundings. The comparison between WAR and CHAOS concerns their formal engines and information channels, not absolute physical closure.
That is why I return to the boundary before returning to the label.
Design the Exchange You Want
WAR concentrates much of the design work in the relationships among cards and procedures. Its ordinary battle resolution gives players little power to alter those relationships through choice or persuasion.
CHAOS makes player decisions and selected social observations part of what the engine receives.
Neither approach guarantees a better game.
They create different design problems and different opportunities for play.
General Systems Theory helps me make those differences explicit.
Define the system.
Identify its environment.
Trace what enters, what changes, and what leaves.
Then follow the feedback and watch what the relationships produce.
The useful question is not simply whether a game is open or closed.
It is what the design allows to cross its boundary—and what happens because it does.
Sources and Further Reading
For the theoretical foundation, see Ludwig von Bertalanffy's General System Theory excerpts, the Systems Engineering Body of Knowledge's system glossary, and its concepts of systems thinking.
For the physical terminology, see Daniel J. Berger's thermodynamic systems overview at Bluffton University.