Game Mechanics Taxonomy

I'd like to do rare or novel and unique games, just like they used to do in the eighties. To do this I need a way to differentiate, identify and classify computer game mechanics. The differentiation between different game mechanics will hopefully reveal infinite amount of niches to fill.

Why structural taxonomy?

I want find the unused or forgotten game mechanics. Therefore I need a map that describes where those are and what they are. Plotting games into genres doesn't form a proper taxonomy for this use because the genres are too arbitrarily selected and too coarse. This classical categorization of games plots diversely different games into the same genre and the games that do not fit into the chart end up into multiple genres.

In practice the genre categorization fails to describe anything relevant about a game. You'd have to go through most of the genres to find all the games you might like to play. In this sense genre categorization is as useful as categorizing the games alphabetically.

The game mechanics shouldn't be muddled into the narrative of the game because narrative is a distinct and separately studied subject. For example, if the behavior of a projectile weapon is exactly same, it shouldn't matter whether the artist depicts them as rocks, shurikens or arrows.

Broad categorization

If the taxonomy is based on the structure, it may be used to uniquely describe any game mechanic because the classification of the game is the structure of the game.

The structural taxonomy is likely not foolproof. The classification still depends on how do you interpret the game mechanics. But it's likely safe to say that it is more reliable than categorization of the game by the genre.

Because of the need to differentiate between games it makes sense to organize the taxonomy by how much the game mechanics would change if the mechanic were mutated. Broader categorization the mechanic matches to, more game mechanics should be suppressed or damaged by the change of the mechanic.

The taxonomization described below cannot be assumed to be whole, because the game design space is likely infinite. Also if a change to specific aspect of mechanics represents a large change to the game mechanics, it should be large change in the taxonomical location accordingly.

Modes

The games usually consists of several distinct modes that contain distinct gameplay each. The individual modes form high-level structure of the game and can be distinctly recognized by the player.

An example: In the early Final Fantasy, there was battle mode, world map mode and game sheet mode. The game sheet embedded over the world map and could be activated any time by the user.

No modes: There is only one distinct mode in the game. Some games exemplary for this category: Chess, Missile command, Galaga, Solitaire, Snake, Frogger, space panic, Lemmings.
Modes: There are multiple distinct modes in the game. Games: Mega man, Pac-man, Epyx summer games, R-Type, Gradius.
Embedded/Isolated modes: If the modes are embedded into other game mechanics they are embedded modes. Pac-man powerups introduce such mode. Opposite of embedded is isolated.
Hierarchical/Flat modes: If there are submodes, the modes form a hierarchy, otherwise the modes in game are flat.
Parallel/Individual modes: Modes in game might activate simultaneously and be parallel modes.
User activated/Event activatable/Environment activated modes: If the mode can be always activated by the player, it is user activated. If the mode can be activated only at special condition related to the player, it is event activated. Otherwise it is environment activated.

The description of mode activation is bit uncertain on games such as Gradius, where the modes seem to be both User and Event activatable. Although it can be argued whether Gradius powerups form actual modes.

Playing Fields

The playing fields are places where the entities are positioned to. The properties of playing fields describe how entities can be positioned on them or how do they move there:

No playing field: There simply isn't an element that holds any entities. Some really simple games might fit this category. Something like rolling a dice.
Discrete/Continuous field: Discrete playing fields determine fixed points where entities may be, whereas continuous fields define areas or volumes where entities may be positioned in.
Linear/Planar/Volumetric field: Determines how the entities may be position in the playing field.
Overlayed/Independent fields: Game may have multiple active playing fields. If overlayed, they interact with each other regularly. Otherwise they're independent.
Slotted playing field: There are fixed amount of slots. Such as in solitaire. To distinct from graph playing fields, the fields may be different by type but not uniquely different. The traversal between slots should be relatively freeformed.
Graph playing field: If the playing field forms a set of nodes and edges to relate between them, it's a graph playing field. Examples: Duke Nukem 3D and some dungeon games.
Grid playing field: Special kind of graph playing field where the organization of the nodes is regular. Giana Sisters, Super Mario, Space Panic, Pac-man.
Stacked playing field: Again originating in solitaire, if discrete fixed field can stack items, such that only the highest element or group forming of the highest elements are available at a time, the playing field is stacked. Also appearing in Tower of hanoi puzzles.

The playing fields the game presents provide quite major part of this kind of taxonomy. The playing fields by large part also determine what can happen in a game, so it is likely that many details of the game depends on the specific kind of playing field.

Entities

Entities are anything within the playing field. Gaming cards, playable and nonplayable elements are entities. Anything that can move on the playing field is an entity. Playing fields and entities likely never appear without each other.

Individual/Parallel/Multipositional entities: The entity may exist in only one playing field at a time, or it may exist multiple times at multiple fields. Or multiple times at a single field.
Newtonian entities: The entities try to simulate the newton's three rules of motion inside the volume.
Transitional entities: It is likely that sometimes an entity may affect the playing field by other ways than just reserving space in the field. If the entity merges with the playing field somehow, it is transitional.
Complex entities: Some entities may themselves contain one or more distinct playing fields.

Player representation inside the game

The representation of the player in the game, who sees the player and what is visible for the player.

Controlled avatar: There's a single entity in the game that represents the player. Usually appears in FPS, and some puzzle, action or platformer games.
Controlled entities: There are multiple entities that are controlled by the player. Lost Vikings and RTS games for example.
Control via altering parameters in entities or positioning/adding/removing/mutating them: Angry birds, Cut the rope, The Incredible Machine, Sim City, Sims, Lemmings.
Continuous control: The player can make control decisions all the time.
Intermittent control: There are designated modes or situations when the player may have control.
Setup and run control: The player can only make control decisions at the start.
Resource bounded control: The player can make control decision only if he has the resources to do it. The player loses the resource after he does the decision.
Fog of War: Some games limit the view of the player to the places where the player can observe, but conveniently also show the places the player has explored.
Fog of War with timestamps: I haven't seen a game this far that'd show the last time when the player saw the certain place.
Hidden information: The player doesn't detect everything on the playing field and must deduce indirectly whether the entity is in the place. This mechanic isn't common, but it appears in Battleship -game.

Winning/Losing/Evaluation/Progression conditions on the game

Very many games have some concept of progression built into them and it is commonly important for the game. This is potentially broader than other categorizations. Also it contains not only the conditions but the rules by which the conditions appear. I'm unsure whether this taxonomical group can be handled exhaustively.

Transport condition: The game progresses when you transfer an entity or entities to a correct location or locations.
Count condition: The game progresses or you win when you have activated, obtained or destroyed all or some of the specific entities.
Checkmate: The game progresses when certain entities no longer have possibility to move anywhere.

An example of classification: Chess

Chess has no modes. It has a discrete grid playing field. It has Individual transitional entities that block the movement of the king. The players have controlled entities on each side with intermittent control. The game has checkmate progression condition on king at each side.

Specific details the categorization doesn't tell are the size of the playing field and that there are two players, that the entity types determine movement patterns that are available to them. Some of these details are crucial for the gameplay and for the chess engines playing the game.

Possibly the details still allow some degree of differentation between other games. I'm leaving it up for exercise to classify the pong and tetris the same way.

How did I realise I need this?

My Ludum Dare entry seems to have been successful in a way that people have liked it. It made me to wonder why? The game didn't really have much of graphics or audio. It only had the game mechanics. I ended up to working on the game mechanics due to limitations and guessing what people would most likely submit as their entries. It can be the game ended up being diverse enough to be equally challenging to everyone.

Since the design space is practically infinite, it is easy to get paralysed by the seemingly infinite amount of possibilities. There comes the motivation for a mental map. That map will help in implementation of the games as well as in finding plausible ideas for testing.