Designing a game is a weird sort of challenge. It is not necessarily because game design is intrinsically harder than other aspects of game development, but because it is hard to find a fixed theoretical foundation upon which the developer can start designing the game in a systematic fashion. Whenever somebody claims that there is one, it often turns out that it is merely a collection of vague, metaphorical ideas which do not quite fit together in a quantitative manner. Those who have seriously enjoyed learning mathematics during their academic years may share this impression.
Many books on game design, even ones that are written by people who have backgrounds in the field of engineering, primarily focus on surveying the cultural implications of a wide variety of videogames. They often talk about the history of videogames and how the authors' friends succeeded in selling their games when they were high school buddies back in the 1980s, etc, and never forget to present the reader with a list of casual methodologies which are supposed to encourage everyone to design one's own videogames in a highly "creative" way (along with a couple of cute-looking graphical illustrations).
These are all good, except that it is hard to discover any solid principle in such books. They are, to me, hardly anything more than a pile of self-help books with the topic of game design in mind - an amorphous cluster of random thoughts and motivational slogans which are poetically connected with one another in a rather optimistic manner, yet do not fail to continue feeding our shared daydream that this whole "soft-skill" narrative somehow pertains to the conceptual essence of game development.
To be fair, there are academic papers which approach the problem of game design in a much more rational way. Professional game designers who have studied the nature of in-game economics have a lot to say when it comes to outlining and refining mathematical models of gameplay, such as progression curves, mutual balancing of intransitive mechanics, means of exchange and their relative values, and so on.
These areas of research, however, focus rather on the perfection of details and their psychological optimization than on the initial conceptualization of the game as a whole. The aforementioned mathematical models may apply well on projects that have already established their core gameplay dynamics, yet the models themselves help only partially when a designer wants to come up with a brand new game concept. And this is oftentimes a crucial problem to solve because indie developers who are trying to make their own games from scratch are usually "lost" when they attempt to conduct such a grand act of creation. This usually forces them to stay away from rigorous academic aspects of game development and desperately try to squeeze their way out of the abyss of uncertainties by means of random bits of brainstorm and other fragmentary ideas which we all can sense on a subconscious level but cannot tell exactly what they are.
A common method of game design employed by many indie developers is highly experimental and is based off of a trial-and-error feedback loop. First, they devise a quick prototype which could be as simple as a few geometric shapes randomly moving inside the scene. Then they try to observe the "fun" part of it and proceed to add a new feature to that scene (e.g. coins to collect, enemies to avoid, etc) which is expected to render the prototype more playable. And then, they play with the updated prototype, add yet another feature on top of it, then test it, and so on, thereby forming a back-and-forth cycle between playtest and feature addition.
Such a method, which is purely empirical in nature, is not an invalid way of developing a game. One can definitely approach the problem of creating something out of nothing by adding gameplay features one by one, just as a game character would gradually unveil the player's FOW (Fog Of War) by moving outward from its base step by step. This approach, however, seems a bit too tiresome and always makes me wonder, "Shouldn't there be a better way?"
What if there are some kind of "Universal Laws of Game Design", just like there are Newton's laws of motion in classical mechanics? Wouldn't their presence enable us to establish a solid foundation of any game during its early development stage, thereby letting us bypass the necessity of spending considerable time desperately searching for core elements which make the game fun to play?
In order to find out such laws, we must first contemplate upon the essence of what a game is. From a practical perspective, a game is an interactive medium which consists of a set of goals. The player endeavors to achieve these goals by making a series of choices, which are often being referred to as "actions".
Such a generic sort of conceptualization, however, is way too vague to be considered useful for our purposes. There is an infinite variety of goals which can potentially exist within the game, and the designer must elaborately choose a subset of them unless he/she wants the game to be a mere concoction of random narratives which, when combined, will simply confound the player with their collective state of chaos.
And in order to reduce the problem of deciding which goals should be part of the gameplay system into a quantitative (i.e. rationally analyzable) one, we must first classify goals into their respective categories. The most challenging side of this is to figure out the level of abstraction we ought to pursue during this process. For example, it would be too cumbersome to simply keep listing a bunch of random goals and putting them in a large taxonomical tree - that is, goals which have their own thematic elements that sound interesting indeed, yet are too specific in their own context to be considered valuable for general usage. Goals such as, "Destroy the emperor's spaceship from the outer orbit of the moon!", "Stop the train by draining its fuel tank before the detective burns more than half of his cigar!", and many others, are only applicable in specific circumstances and are therefore not adequately designed to function as reusable components.
The best way of avoiding such a rigmarole is to start from the most fundamental goals which we all share in common. These are the goals that are (either consciously or unconsciously) being shared by all lifeforms and are often indicated by terms such as "instincts", "archetypes", and kindred others which suggest their generic applicability. The idea is, once we have a clear notion of what are the most primitive goals of every living organism, we will also have a clear notion of what are the most primitive goals of the player as well because the player is a living organism, too.
Let us suppose that there is a single point in space which represents the simplest form of life we can ever imagine. It has no brain, no organs, no tissues, no nerves, and absolutely no cells. It is just a single mathematical point which only possesses a set of basic properties which would qualify it as a "living thing" and absolutely nothing else. This is the most abstract model of a biological organism we can manage to conceive in the domain of quantitative reasoning (Just like a "point mass" is the highest form of abstraction of a rigid body in classical mechanics).
This abstract lifeform has its own "region of influence", which is the set of all points in space that can be influenced by its existence (i.e. have a nonzero probability of creating a future event whose chain of causality can be traced back to the event which gave birth to the lifeform). Examples include:
(1) The organism's place of residence (e.g. A bird's nest, a person's house, etc).
(2) A territory that is frequently being visited (patrolled) by the organism (e.g. A cat's alley).
(3) The organism's lifetime (as long as we suppose that "time" is just another dimension in space).
(4) Portion of space which the organism's own body occupies (e.g. A plant's area of exposure to sunlight as well as the extent of its roots underneath the surface of earth, an area covered by a swarm of fungi, etc).
(5) A territory that is "owned" by the organism by means of power (e.g. military force, legal ownership, social consensus, etc).
(6) The organism's public reputation as a social being (e.g. popularity, brand values, credit, acquaintances, etc).
(7) The organism's biological heritage (e.g. distribution of its genetic information throughout the ecosystem, which can be defined as the weighted average of the regions of influence of its children, grandchildren, great-grandchildren, etc).
The most ultimate goal of a lifeform is to preserve and expand its region of influence as much as possible. There are multiple ways of achieving this. If it is an animal (i.e. something which can displace its own body to somewhere else), it will be able to temporarily expand its region of influence by visiting places that are outside of the region. Another method which applies to all kingdoms of living things (not just animals) is to reproduce itself for the purpose of generating yet another point which is equally capable of emitting the force of influence around its current location. This latter method also expands the organism's region of influence in the direction that is parallel to the time axis, since creating a younger copy of itself is the most quintessential way of delaying the expiration of its influence.
There are also occasions, however, after which the lifeform's region of influence may shrink in size. The presence of another lifeform's region of influence may overtake the existing region (since increased influence of others means decreased influence of mine), for instance. Also, the fact that the duration of influence has its own limit means that unvisited parts of the region naturally erode over time.
The implication of the aforementioned phenomena is that an organism's region of influence is permanently being exposed to the state of oscillation between expansion and shrinkage.
Environmental factors constantly disintegrate the weakest parts of the region, while the organism itself strives to protect and enlarge it - for it is a common sense that letting one's line of influence propagate through the spacetime continuum as far as possible is the most optimal way of prolonging the causal chain of its own existence.
This general description alone, however, won't lead us anywhere except an endless row of philosophical treatises. For the purpose of game design, what we need is a set of discrete building blocks which can be used for formulating the game's narratives. Such elements must be based upon the core instincts that are universal in every biological being (in order to appeal to the broadest audience as possible), yet we must arrange them in a systematic manner instead of deferring the necessity of their acute interpretation with poetic rambling. As a starting ground, let me come up with the 3 most fundamental goals which I think must be shared by all lifeforms. These are: (1) Absorb, (2) Expand, and (3) Secure.
An organism must "absorb" resources from outside in order to gain enough energy to sustain its life and carry out its own sequence of actions; otherwise it will run out of energy and die. It must also "expand" its region of influence by means of movement and self-replication, while also making sure to "secure" parts of it which have previously been acquired.
One of the reasons why I have presumed the existence of these 3 particular instincts is that this trio are reminiscent of our popular conception of the way in which game mechanics can usually be classified. For instance,
(1) While playing a war game, the player has 3 choices: Whether to collect resources (i.e. "absorb" fuel and constructional ingredients), attack the enemy (i.e. "expand" one's region of influence by means of conquest), or defend the base (i.e. "secure" one's region of influence by means of defense).
(2) While playing Pac-Man, the player has 3 choices: Whether to collect tokens (i.e. "absorb" score points which will allow the player to proceed to the next level), spatially traverse each of the succeeding levels (i.e. "expand" one's region of influence by leveling up and climbing up the leaderboard), or avoid ghosts (i.e. "secure" one's region of influence by saving Pac-Man's life from the enemy characters).
(3) While playing a city-building game, the player has 3 choices: Whether to collect taxes from the citizens (i.e. "absorb" financial potential energy from the taxpayers), grow urban areas by building roads, public facilities, and other parts of the infrastructure (i.e. "expand" one's region of influence by literally expanding the size of the city), or prepare for potential disasters/crimes by installing fire/police stations (i.e. "secure" one's region of influence by means of first responders).
And the list goes on.
These three fundamental goals, however, will remain quite impractical if we are to regard them only as means of identification and nothing else. If we want to use them for game design purposes, we ought to represent them as mathematical procedures each of which has its own set of input parameters. This way we can allow ourselves to begin assembling bits of design logic with their functional implications in mind.
In order to figure out how to apply them exactly, however, one must beware that a game, despite being driven by goals and actions, starts its narratives from an arrangement of physical entities (aka "objects"). We have a world, and inside the world we have a set of objects. One could easily suppose that there must be a point in space which represents the game's central lifeform called the protagonist (aka "myself"). In addition, there could be other points in space which are quite isolated from the protagonist itself and can be grouped into two general categories:
(1) "Resources" - Objects which, when in touch with the protagonist, increase its ability to invoke actions based on its own will. Examples of resources include: Coins, Money, Fuel, Score Points, Victory Points, Action Points, Food, Crafting Ingredients, Collectible Cards, Keys, Territorial Ownership, Psychological Assertion of Dominance (e.g. Fame), and so on.
(2) "Obstacles" - Objects which, when in touch with the protagonist, decrease its ability to invoke actions based on its own will. Examples of obstacles include: Walls, Fences, Landmines, Enemy Characters, Explosives, Traps, Swamp, Locked Doors, Locked Chests, Diseases, Poison, Debt, Lawsuits (which are filed against the protagonist), Stigma, Accusations, and so on.
These two categories nicely reflect the dualism of our universe because they resemble the positive(+) and negative(-) energy particles of our physical reality. Resources are positively charged particles since they add up to the lifeform's store of potential energy. On the other hand, obstacles are negatively charged particles since they subtract from the lifeform's store of potential energy.
In order to grow/preserve its region of influence as efficiently as possible, an organism must maximize its number of chances to increase the region's current size. In order to fulfill this task, it must maximize its chance of colliding with a resource and minimize its chance of colliding with an obstacle. This means that the ultimate goal of every lifeform is to minimize its distance from nearby resources and maximize its distance from nearby obstacles as much as possible (while also making sure to keep expanding its region of influence by means of locomotion and procreation). This pair of motives, from a behaviorist point of view, can be modeled as a physical system made up of hypothetical point-charges and their surrounding hypothetical force vectors.
Such a purely mechanistic model, however, is probably not adequate for representing the design of a game because a game puts more emphasis on the semantics of motivation itself than its causal byproducts. Besides, representing the game's inner world in terms of a single continuous space filled with a fine grid of fields and their respective force vectors distracts us from constructing the game's individual components in a discrete manner, which makes it difficult to build gameplay systems in a highly modular (therefore robust) way. What we probably need is to identify a set of discrete rules of interaction between the organism and its surrounding resources/obstacles, rather than trying to analyze every detail of their ensuing physical phenomena.
(Will be continued in Volume 2)