13 min read

Game Mechanics

In this post I summarize my current thinking about the fundamental elements of games. I consider the characteristics of the physical and logical parts: play spaces; game pieces; player roles; actions and legality; and resolution states.


Many years ago, I started taking an interest in the analysis of games in order to understand how they functioned. I wanted to learn something about game design, but mostly, I wanted to know how to construct games in software. While I consider myself a creative person, my ways of thinking about the world as it exists are literal, practical and scientific. I am an engineer. When people make things, they make them out of parts. Each part has a specific purpose and role in the game’s workings.

I didn't have a lot of luck. I found the various discussions of game development and construction either too too theoretical, or not theoretical enough.

On the one hand, ludologists were exploring games as cultural phenomena. They saw games as cultural artifacts, imbued with layers of symbolic meaning. They were interested in relating them to other art forms, to the social sciences, and to human nature. While they like theory, the didn't seem interested in theories about how games function internally. They seemed to be more interested in how games functioned socially. While I find it interesting, I don't find it helpful in trying to make games.

On the other hand, there were game developers. Either as designers or implementors, they were mostly interested in how their games were like or unlike other games that already existed. They didn't seem to worry too much about their internal nature. At least, they didn't much talk or write about that aspect, or in terms that I recognized.

Granted, almost everything I read was related specifically to computer and video game development. They were often focussed purely on the needs of a specific game and how it was made. At best, the more general works examined genres, but they were heavily dependent upon examples of games that fit into them. Genres: the commonly recognized stylistic groupings into which all games are shoehorned in one way or another. It was the same whether the authors were principally designers or engineers. Certainly they admitted to, and sought to exploit, similarities between different parts of the software used to make computer games: graphics, physics, AI, sound, networking, etc., but they didn't seem to care how those related to more general ideas of games and the way games work, independently of the medium of implementation.

In both cases, I found extremes. One group's focus was far too wide, and the other's was far too narrow. I was looking for something in the middle. You might say, I was looking for the Goldilocks zone. I never found it. Possibly, I'm not looking hard enough. But as far as I'm concerned, I can do it myself. Even though game design theory is not my passion. Software is my passion: specifically, software whose purpose is to bring to life imaginary worlds (which are mostly found in games).

Reducing Games to Their Component Parts

When I think about games, it is as I think about any complex subject. I start with a scientific outlook, and the most basic tool of science: reductionism. In other words, I think about the parts of games: the rules, the role of players, the pieces, and the game space. Games, and therefore their component parts, exist primarily in the user’s imagination. But in many formally defined games, some of the parts also exist in an external space, either physical (examples: a board game’s playing board, a table top for card games, and a field for athletic sports) or informational (examples: a score sheet in golf, the entirety of a computerized game).

Play Spaces

I would like to simplify by proposing that all games take place in a physical space. I will do this by treating informational game play spaces as simulations of real spaces. Of course, in some games it would actually be better to assume the opposite: that the physical manifestation of a game is not its primary manifestation, but is used as a kind of memory helper. It's also possible to imagine a game played in a game space with more than three dimensions, which would be difficult to construct in the real world. Such games can only exist in information space. But such games are rare, because our brains are adapted to thinking about the physical world.

If the internal state of a game's pieces are considered additional dimensions, virtually all computerized games are hyper-dimensional. However, we can get around that. Consider a team sport, like baseball. A description of the game of baseball will usually include a list of the equipment, the player positions, and rules about hits, outs, base running and scoring. It will not necessarily mention the details about how individual players affect the outcome of a particular game, but the players determine how a game proceeds as much as, if not more than, the rules do.

Think of a computer game. Now, imagine playing that game in the real world. If you need examples, look at live action role playing (LARP). In LARP, the players are characters, and all the characters are represented by players. The players keep track of the internal state of their characters, both mental (personality, attitude, allegiances, desires, preferences) and physical (health, abilities, talents).

In sports, the players are not (usually) role playing; they don't need to separately remember a character's mental and physical state: they are their own character. All I want is to demonstrate that sports and other games which take place in physical game spaces are also heavily influenced by informational content. Real games are easily as complex as computer games. When you take players' mental states into account, they are immeasurably more complex! What sets computer games apart is mostly the thematic dressing and the sheer number of participants, even if most of them are simulated.

But for the sake of discussion (even though any discussion implicitly treats its subject as information), it is still easier to use physical terminology. So, even when talking about computer and video games, which often feature fantasy settings and creatures which don't (or can't) exist in reality, as far as we know, it's simpler to treat those things as real, physical things, for the moment. We can reduce them to information states and representational tokens later. 

Game Participants

Let's review the list of things that make up a game: rules, pieces, and the game space. But don't forget players! In fact, I prefer to use a more general term, so that we can include virtual characters and other entities that so often populate computer games: participants. "Participants" also includes referees, score keepers and other people (or programs) which run the game and enforce its rules.

It is possible, for the purposes of discussion, to treat the non-player participants as a single entity, known as the "game master". Despite being a collection of different participants (people, computers or programs), they all act with a common goal. Any time a group work together to achieve a single aim, they are acting as one, so if it's easier to treat them as one, we might as well.

In most computer games, the game master is a single program running on a single computer. There are lots of variations on use, but generally, the collectivism principle applies: one goal (enforcing the rules of the game), one identity. (You could try to argue that players to a great extent internalize the rules of a game, and so also act as part of the game master identity. But that is a side effect of being watched by referees, and not wanting to get penalized or thrown out of the game. It's easier to assume that player's want to win, and that adhering to the rules is an imposition for them, at least while playing. We have to simplify.)

So where are we? Games are played by players, according to rules, overseen by a game master, in some kind of space. I haven't elaborated on the idea of game pieces or the nature of the rules, and I have yet to mention objectives. Oh, I just did.

Two Meanings of "Game"

Let me stop for a second to define the word game for my purposes. In fact, I have to define it twice, because there are two kinds of game. I will borrow terms from software engineering. The first type of game is a "class" of game, and the second is an "instance" of the some class of game. I think this is a normal way of thinking about games, but I am just adding some formal terms.

For any kind of game, you know it by the parts that I have mentioned: mostly made up of the rules, which define the physical parts of the game as a system: the different ways they can be arranged (the various states of the game system), how those arrangements can be transformed from one into another, and which arrangement—or sequence of arrangements—represents the winning or goal state (and, optionally, a tie state; any other state is a losing state).

Examples are easy. The game of baseball consists of a diamond (the play space); bats and balls, gloves and other equipment (the pieces); batters, infielders, outfielders and the catcher (players); terms including hit, bunt, strike, ball, base, run, out (terminology rules); and legal actions like pitch, swing, base run, throw, and tag (transformational rules).

The second kind of game, an instance, is an actual play through of a game, from start to finish. The start involves an initial configuration of the game pieces within the play space, goes through numerous stages of transformation as players perform legal actions (and possibly illegal actions, which may or may not result in penalties enforced by the game master), and finally arriving at an end state which may or may not constitute a resolution such as a win, loss or tie.

I will spare you (and myself, since I'm not a fan) a summary of a baseball game.

Definition of "Game"

A game is a activity which in players engage one another, or a system, with purpose. A game is defined by rules which describe the material and logical elements of the game.

The material elements include the play space, the pieces, and the roles occupied by players (which may also include equipment).

The logical elements describe the initial arrangement, or state, of the material elements; a set of legal actions for rearranging them; optional penalties for illegal actions; a set of possible consequences (some desirable, others not); and a set of states of the pieces which invoke one or more consequences when they occur.

A game constitutes a challenge for its players. By performing legal actions, they must attempt to rearrange the physical elements of the game into a desirable state: one with desirable consequences. Some games are composed a set of smaller games, played in sequence or in parallel (or both). Many games include the idea of scoring, which is a count of how often a player (or team) has reached a desirable state in their favour. Most games have one or more resolution states which determine when the game ends. Resolutions can declare winners, losers, or a tie, or may assign a final score. (In single player games, the player might win, lose, or neither. Often, they will have achieved a score by which they will compare their performance to other plays of the game in order to measure their relative accomplishment.) The resolution state of a game composed of smaller games is often based on the aggregate results of their separate resolutions.

Games as Mathematical Objects

In computer science, a game corresponds exactly to one of a variety of computational machine. (These "machines" are conceptual. They do not require an actual mechanical computer to represent them.) Depending on the rules of the game, the game might be classified as a state machine, a push-down automaton, or a full-scale Turing machine. The physical elements of the game correspond with the states of the machine. The actions which players can perform correspond with the language recognized by the machine which causes the machine to transition between states.

Every play of a game ends with an outcome, when that play is resolved. Outcomes exist in a space of all possible outcomes, and plays exist in a space of all possible plays. For many games, the set of outcomes is usually not too large (even including the possible scores), but the set of possible plays is often vast.

Every game (and here I mean the "class" which determines all possible plays of that game) exists within an abstract space of all possible games, but that space is completely unbounded. This space of endless possibility is the primary reason why it is difficult to understand the nature of games without depending on reductionism. If we don't narrow our scope, the best we can do in our analysis of games is to look at existing games, both classes and instances. (There is another meaning of the word "class" which we can use in games, which in software engineering would represent a "superclass" of games. A superclass is something like a category. It is an abstraction that, in this case, has no real instances, but describes common elements shared by a group of different classes of game. It might also be called a "meta-class". It roughly corresponds to the idea of game "genres".)


By reducing games to their component parts, we can at least enumerate a large number of possible game play elements, including both material and logical elements, and consider different classes of games based on the elements that compose them. By the act of composing these elements—these atoms of gameplay—we can invent new kinds of games for consideration, and hopefully discover entirely new games, and new genres. On the other hand, if we can't, it would be because all classes of games have already been invented. But we can only determine that by doing the work of examining all of the elements and all of their basic combinations. We can set an upper limit on the number of combinations in order to limit the set of possible genres that we can come up with that consist of games that are simple enough to describe and be played by regular people.

In a future blog post, I plan to begin a list of game actions which I consider the atoms of all game mechanics.

Latest Jobs


Playa Vista, California
Audio Engineer

Digital Extremes

London, Ontario, Canada
Communications Director

High Moon Studios

Carlsbad, California
Senior Producer

Build a Rocket Boy Games

Edinburgh, Scotland
Lead UI Programmer
More Jobs   


Register for a
Subscribe to
Follow us

Game Developer Account

Game Developer Newsletter


Register for a

Game Developer Account

Gain full access to resources (events, white paper, webinars, reports, etc)
Single sign-on to all Informa products

Subscribe to

Game Developer Newsletter

Get daily Game Developer top stories every morning straight into your inbox

Follow us


Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more