I was invited by Jon Cogburn to submit a chapter to the new collection of essays combining Dungeons & Dragons with philosophy, but my chapter wildly overran. Here’s a segment I had to cut.

What do Dungeons and Dragons monsters have to teach us about fiction and prop theory?

What the character sheet does for the representation of the player characters, Gary Gygax’s 1977 Monster Manual – whose truly dreadful original cover is depicted to the left – and Don Turnbull’s 1981 Fiend Folio did for the creatures they are pitted against. As with attributes on the character sheet, the capabilities of the D&D monsters are laid out in numerical detail – including purely representative elements (size, weight) as well as the statistics (HP, THAC0) required to beat them into a bloody pulp, not to mention props for explaining their behaviour (alignment). In addition, an illustration of the monster helped provide a depictive prop to help either the players imagine their foe, or the Dungeon Master to describe it.

The Monster Manual and its less elegantly titled alliterative descendents are more than just reference books for game statistics (although this is their principal role), for the Monster Manual is also a bestiary of the fictional worlds of Dungeons & Dragons itself. Of course, specific campaign settings make variations one way or another, and there perhaps is no single campaign of D&D anywhere which has used every single monster in every single published guide, but by listing a menagerie of menacing monsters in one manuscript an impression of the kind of world in question is inescapably provided. The Monster Manual unequivocally prescribed players to imagine that the fictional worlds of D&D were deeply weird places, filled with an ecology of utterly bizarre beasts that would make no sense in any other context but high fantasy.

No fantasy novel ever written has contained such a heterogeneous hodge-podge of heinous horrors – the reader would simply have no way of dealing with such eclecticism in a conventional narrative. Yet somehow, the fantasy role-playing game dodges this criticism – or at least, Dungeons & Dragons (and the computer role-playing games it has inspired) avoid this complaint. For it must be said that a great deal of latitude is extended towards D&D’s ramshackle collection of foes; a commercial RPG published today with such widespread disregard for cohesion in the resulting fictional world would be subject to criticism. D&D is immune to it.

There is something about the megatextual collision of monsters from every conceivable mythological source that serves to buffer the inherent nonsense that results from criticism. To this day, I am unsure quite what it is. Is it that D&D was the first of its kind, and is thus afforded a certain latitude? Perhaps. But I rather suspect that there is a craving for this kind of massive intersection of otherwise distinct folklores. We see the same theme expressed in a movie such as Shrek, which combines all fairy tales into one fictional world, or indeed in Neil Gaiman’s adult comics The Sandman, which conduct the same kind of mythic collision with more delicacy and panache. Deep down, we can’t escape the feeling that all stories are one story, and the curious concordance of creatures in the Monster Manual speak of the same urge.

Cross-posted from Only a Game; comments accepted on either blog.

Last week we looked at how prop theory can clarify game design by identifying components of a toy chest containing the principal and central props used for the play of digital games. This week we continue to consider the first and third person fully-rendered fictional world but look at another aspect of the toy chest paradigm – the play sets within which the individual toys are used.

The Play Set

Beyond the dolls and models (and their various trimmings) lie the toys which make up the wider substance of the fictional world itself. These are like the child’s play set, within which individual toys are placed for specific games of make-believe. When a plastic Godzilla appears in the parking garage, we are playing monster attack. Similarly, the contents of the toy chest combine with the following props to dictate specific opportunities for play. In these cases, the toy metaphor often gives way to the paradigm of the film set, which in many respects is analagous to the child's play set.

• Sets and Locations: like a movie set, the environments that the avatar can explore can be thought of as sets, within which play can be staged. And similarly to a movie, it is also possible to make sets of such complexity that they constitute locations. What’s the difference? Principally just scale, but there is also the question of connectivity. If the game consists of a sequence of enclosed locations (as most FPS games do) or as a map which collects multiple enclosed locations (as many RPGs do) then set is the most appropriate term. But if the game is closer to an open or playground world then location is a more appropriate term. Either way, we’re referring to the environment of play purely as a spatial entity. Anything that can be interacted with within the set or location can be thought of as a different kind of prop.
• Scenery: anything in a set or location that doesn’t interact with dolls or models in any way beyond being physically present is simply considered scenery.
• Doors, Keys and Switches: connecting sets together (or embedding sets in locations or connecting locations together)  are doors, and these may require another toy – a key or a switch – that opens the door. The key prop has the characteristic of granting an additional ability to the avatar (it can open certain doors), while the switch prop simply changes the states of other props. The latter always risks creating puzzle-solving play, and restricting the appeal of the play to a smaller audience, since challenging puzzles always have a dominating effect on play.
• Climbing Frame, Slides and Trampolines: this first term refers to a toy that allows dolls to ascend or descend (including ladders, ropes, vines and anything else), the second a toy that allows dolls solely to descend on a specific path (including rope slides, chutes etc.), and the last a toy that allows dolls to ascend but not descend. These may influence play in certain ways, but rarely if ever dominate it.
• Obstacles: a specific kind of toy within a set or location that is overcome in a particular fashion, such as a gap separating a part of the set, a hole in a wall that must be crouched through, can be considered an obstacle. In many cases, the player’s doll may have the ability to bypass the obstacle using its basic abilities (in which case the obstacle is primarily a control challenge for the player), in others a specific tool may be required (in which case the obstacle may have a structural role in arranging the shape of the world). Doors and keys might also be thought of as a kind of obstacle, albeit one with a distinctive character of their own. The presence of obstacles has an influencing effect on play, but it does not dominate play in the way that weapons will.
• Map and Compass: the map is a prop which depicts the whole of a set or location, while a compass is anything (including a radar or GPS) that guides the player within the world. These two props are of vital importance! Games that use locations must have maps (and probably should have compasses) unless there is a good reason for omiting them, while games that have sets may benefit from maps but are more likely to be able to survive without them. The key issue in respect of these props is the complexity of navigation - and while more complex geography probably requires a map, the map will heavily influence the resulting play. It is worth remembering that just as not everyone can remote control a toy airplane, not everyone can read a map, and the presence of a map is a sign of a game unlikely to reach a true mass market audience. However, the gamer hobbyists who buy and play many different games are almost all map-literate, and thus most digital games will require decent map and compass props.

Other Props

There are other props which have radical effects on the play of games but do not fit into the toy chest or play set metaphor in quite the same way as those discussed in the previous two sections.

• Shops and Money: the two kinds of prop go together. If you have shops, you have money to spend in them. (You may also have a shopkeeper – a simple robot doll who mans the shop). The presence of money is a heavy influence on play, and is likely to encourage collecting and stockpiling. 
• Experience and Skills: Experience and character screens in RPGs can similar be understood as shops and money, albeit a different kind of money that represents acquired skill. Experience dominates play in almost every context – and modern console games are including these props in everything from FPS games to sims these days. 
• Tasks, Puzzles and Stamp Collections: typically consisting of a verbal objective (a narration prop) and an optional prize (a toy paid out for completing that objective), the task is a basic structural element of any game, usually presented to the player within the narrative as a mission or quest. Tasks may not seem like a toy, but we can easily find examples of toys with intrinsic tasks and puzzles. Consider the ball maze and its implicit task of guiding the ball to a target spot, or the Rubik's cube with its intrinsic ‘mission’ to solve the puzzle, or an I Spy book with its “stamp collection” to be completed. When the task is more like a Rubik’s cube, the play will be dominated by puzzles and the potential audience narrowed, while the “stamp collection” style has wider appeal (provided each “stamp” is comparatively straightforward to acquire) but can still heavily influence play. Indeed, in modern games Achievements or Trophies attach a stamp collection of tasks and puzzles to every hobbyist console game, which carries a severe risk of dominating the play of these games.

Embracing or Avoiding Domination

The props identified last week and this week are in no way an exhaustive list of the toys to be found in digital games, but collectively they demonstrate the general framework being proposed – one which draws parallels with children’s toys on the one hand and movie productions on the other to encourage thinking about the elements of games from an explicit viewpoint of play rather than of game.

Furthermore, the perspective brings to the fore the way in which various props can dominate or heavily influence the play of digital games. Weapons and toy soldiers dominate the play in any fictional world whenever they are a part of the toy chest, and models and obstacles carry similar (although less pronounced) risks. Money props – especially in the form of experience points – have a severe dominating effect. Another significant risk for domination of play occurs with tasks and collections, and particularly puzzles, each of which makes the play more rewarding for players willing and able to co-operate, but at the cost of closing down accessibility from anyone not open to that kind of play. (My comments on the way in which enduring hardship ultimately enhances enjoyment is relevant here).

It may be illuminating to pursue the reasons that certain props dominate play, and the way in which this domination necessarily creates a barrier for anyone uninterested in playing this way, since somewhere in this perspective lies valuable lessons for commercial game development – and invaluable opportunities for art games seeking to carve out new play experiences beyond the obvious tropes. By using prop theory as a lens, game design may be able to tease out more playful combinations of toys, or at least better understand how it is that the kinds of play we are presented so often end up in surprisingly narrow forms.

How can prop theory assist game design? One potentially illuminating application is by narrowing the gap between what the developer sees and what the player experiences by conceiving both the assets and the play in terms of toys – specific props for play. In this piece (and its companion which follows next week) I present a slightly different way of thinking about game design – one in which every game is effectively a toy chest – and encourage game designers to think about digital games more playfully.

Asset Lists

When a medium to large digital game developer works on a project, there are sure to be 'asset lists. The term, inherited from corporate business in general, is intended to be a breakdown of all the resources that can be used – a catalogue of graphical data (e.g. poly models and textures) and animations, as well as the elements of the game engine, the various subsystems that make anything work (e.g. interface, renderer, AI). This is a very nuts and bolts way of seeing a game project.

The player, on the other hand, encounters all of this content not in the form of an asset list but in the context of the fictional world of the game. They do not see (unless they too are game developers) textures, poly models and a renderer, but they see a car, a gun, an alien, a building or what have you. They experience the game props directly, through the fictional world of the game. The assets used to generate those props are no more important to the player than the sets and props used when shooting a movie are for the audience in the theatre. They may want to see them at some other time in that context, say, when they are watching a “making of” documentary, but they aren’t interested in them in this way when they are enjoying the film.

One possible way for prop theory to feed into game design practices, then, is to present some aspect of the design paperwork – perhaps just the concept design (the high level description of how everything works), or perhaps a full game design (the detailed description of how everything interrelates) from the perspective of the props involved in the game. Or, to put it another way, rather than breaking down the asset list by the individual classes of asset, present the asset list how the player experiences its content. So, for instance, a Nazi soldier in a WWII FPS game appears in a catalogue of props as exactly that – a Nazi soldier. If the game has multiple humanoid opponents, they appear in the prop list individually – so a Zombie is a different prop to a Nazi soldier, even if they share certain common components.

For convenience, the props can be sorted by type, allowing common elements to be referenced in one place. Both the Nazi soldier and the Zombie are particular kinds of doll, in the sense I have contrasted with avatar, although they are enemy dolls (or action figures, for those who get held up on the gender implications of ‘doll’). In this sense, what I am proposing here is related to object-orientation in programming – but what is intended by pursuing the prop perspective is to take upon the object-oriented view from the player’s perspective, not from the perspective of implementation. (It is possible, of course, that narrowing the gap between these two points of view will be of benefit to the developer anyway).

Viewing play in this way reveals that certain kinds of prop dominate the play of a fictional world, such that their presence almost guarantees that the play will revolve around that kind of prop, while others strongly influence the play such that the flavour of the play is inescapably altered by their presence. These trends occur because certain kinds of activities are more intrinsically rewarding (“more fun”) than others – but for every kind of fun that might occur, there is a player who is effectively excluded when that type of play dominates.

A Toy Chest

What kinds of prop do we need to consider? There will not be a fixed and static ontology, because there are always an infinite number of different ways of interpreting ordered collections of entities. What prop theory suggests as a potentially valuable frame of reference, however, is the notion of a toy chest – that is, to catalogue the contents of game in question by way of looking at the kind of toys on offer in the fictional world of the game.

If, for simplicity, we curtail our perspective to just first and third person fully-rendered fictional worlds we find something like the schema that follows. Firstly, let us consider the way the player participates in the fictional world from the outside inwards:

• Remote Control: the interface from the player’s perspective is the equivalent of a remote control for a toy. Thinking about it in these terms helpfully discloses the issue of complexity. Few people could fly a radio controlled plane (twin sticks plus buttons), significantly more can drive a radio controlled car (one stick plus accelerator), considerably more can point a laser pointer at a target. It’s the same with a game. There are no blockbuster flight games, because the controls are too complex for most players, while the Wii appeals to a very wide audience precisely in the cases of simple controls. Whether we are talking about a gamepad, a Wii Remote, a stylus and touchpad, a mouse and keyboard or whatever else, the control device and the interface collectively can be considered the remote control.
• Dolls: what the player controls with the remote control is a doll (or, as described in the next bullet point, a model). Each doll is a humanoid character which has different inherent capabilities, and different abilities in the context of other toys. This ability to interact with the other toys meaningfully is important to the appeal of the player’s dolls, and modern games provide many different intrinsic abilities such as jumping, ducking, dodging, countering attacks and so forth. Thinking of these element of the game as a doll (or action figure) clarifies the player’s relationship with it – and also helps explain why the big videogame franchises are likely to also be selling physical versions of these toys to kids, like the Halo ODST action figure pictured above.
• Model: each model vehicle (e.g. a car, a plane, a spaceship) is principally characterised by a different mode of movement within the fictional world. Whereas dolls often have many different ways of interacting with the other toys, models usually only have one way to interact with the other models – crashing into them. This can have such a strong effect on play that having models risk dominating play – in the case of spaceships and planes, this can catastrophically restrict the appeal of the game (because of the complexity of control) unless these models are somehow made entirely peripheral. Even if they do not dominate, models always have a strong influence – the moment vehicles are permitted in a fictional world, certain other kinds of play are de-emphasised.
• Avatar: the avatar refers solely to what allows the player to take action in the fictional world of the game, and this is simply a kind of suture between the remote control and the dolls or models being controlled. The avatar is barely a prop, and what is relevant to the toy chest are the kinds of things that the avatar is allowed to ‘enter’ or ‘possess’. Part of the appeal of a game such as Grand Theft Auto is the ability to “be” either a criminal doll or a model car – for the avatar to link the remote control to multiple different toys. This flexibility is a large part of the appeal of this kind of game.
• Robots: as well as being remote controlled by the player (via the avatar), dolls and models may be possessed by computer-controlled agents – making them into ‘robots’. A Zombie is a simple kind of robot, for instance, which more or less runs towards the player or players without any more complex thought, a Nazi soldier may be a more intelligent robot, but anything that moves around the world can be understood as a robot. ‘Robot’ in this sense is the equivalent to the ‘avatar’ – the avatar bridges the remote control to the props in the world, the robot bridges the AI to the props in the world. Note that a robot need not be a warrior (like the ‘Toy Soldier’, below) – it could also be something that can be spoken to (like a Shopkeeper).
• Weapons: these are the toys that allow dolls and models to damage and ultimately eliminate certain other toys from the world. Characteristic of weapons are that the presence of weapons in the toy chest guarantees the play will entail violence. This is to say that weapons dominate the play of the world they appear in. If you want to have more interesting and unusual play, you must either restrict or exclude weapons from the toy chest.
• Toy soldiers: instead of offering weapons as individual toys, dolls may inherently function as weapons. In this case, instead of dolls (which can, in principle, do many different things) we have toy soldiers – toys that move and fight. Just as weapons dominate play, so do toy soldiers. Indeed, if the player is given a collection of toy soldiers to play with, they are essentially forced into tactics and strategic play against robot toy soldiers (narrowing the audience, since the complexity of these games requires a more cerebral player).
• Tools: just as a weapon grants a doll the ability to cause violence, a tool grants a doll an additional ability e.g. a torch is a tool that illuminates, a rope is a tool that provides the capacity to overcome certain obstacles, and a grapple is a tool that moves toys around in certain ways.

The above constitute the principal props of digital games, those elements of the toy chest which are more or less the centre of attention for players. In the other half of this discussion next week, we’ll consider those props that provide the elements of the fictional world itself, and that otherwise condition the play within that world.

Next week: A Play Set for Game Design

It is one of the strangest aspects of play that whatever can be endured will ultimately serve to enhance the enjoyment of the player who perseveres. Tolerating repetition adds satisfaction to the completing task. Tolerating difficulty in challenges turns mere success into glowing victory. Tolerating frustratingly obscure puzzles leads to smug triumph when they are eventually cracked.

Of course, each of these ordeals to be endured will also exclude certain players from reaching their eventual rewards. Not everyone is willing to endure tedium, difficulty or obscurity. But it is striking to note that the same things which cause certain players to give up a game are the very things which make it worth playing for others. This is more than just ‘different strokes for different folks’ – it seems as if whatever a player will endure ultimately ends up enhancing the reward they experience.