Gotland University, Cramergatan 3, SE-621 57 Visby, Sweden
Computer games can involve narrative and story elements integrating different forms of interactivity and using different strategies for combining interaction with non-interactive story and narrative elements. While some forms of interactive narrative involve simple selection between fixed narrative sequences, computer games more typically involve the integration of narrative with game play based upon a simulation substrate. These three forms, simulation, game play and narrative, involve pre-authored time structures at different levels of time scale. Simulation involves the lowest levels of time structure, with authored principles specifying how time develops from frame to frame based upon physics, the representation of game objects and their behaviour, and discrete event simulation. Games involve pre-designed game moves, types of actions that may be realized as abstractions over patterns of low level changes at the frame level. Linear and interactive narratives form the highest level of pre-designed time structure, framing low-level simulation processes and intermediate level game moves within a high level structure typically based upon classic models of narrative form. Computer games may emphasise one or more of these primary forms as the focus of meaning in the play experience. Story construction within computer games is a function of how these different levels of time structure interact in the play experience, being the result of pre-designed narrative content, story potential and the actual unfolding story created by the actions of the player. There are many strategies for integrating these forms. However, a crucial issue in the design of story content is the relationship between how the resulting game experience relates to user play preferences. In particular, categories of play style can be extended to include preferences for how story content is experienced, based upon audience, performance and immersionist orientations to story. Perceived tensions within computer game form, such as the tension between game play and narrative, are explained, not as fundamental formal issues, but issues of player preferences and how these are satisfied or not by different strategies for story content within a game system.
Integrating story elements in computer games has been a source of heated debate focussed upon fundamental questions about the nature of a game and the nature of a narrative (eg. ). The association of interactivity and narrative is often regarded as a fundamental paradox within the conception of interactive narrative (eg. ). This ongoing conceptual debate can be attributed at least in part to two significant factors. The first factor is that any debate over definitions tends to weaken or undermine strong interpretations of the meanings of the terms involved, since the terms do not have any absolute meanings and the boundaries between complex phenomena are generally vague. This factor can be eased by the use of definitions that are not interpreted as the realmeanings of terms, but instead provide distinctions that lead to a coherent strategy for design. In this case, we do not search for any absolute conceptions of “game” or “narrative”, but use definitions that lead to clearly distinguished and systematic ways of analysing, conceiving of, and designing game systems. From a design perspective, the best definitions are those that lead to the most clearly articulated and distinct design options.
A second and perhaps surprising factor in the discussion about the relationship between gameplay and narrative is that the issue has not generally been considered in relation to the preferred play/interaction styles of players. The “tension between game play and narrative” is manifest in statements like “cut scenes break immersion in the game” or “the game play is repetitive and has nothing to do with the story”. Another player referring to the same game may find its strategy for intermixing story and narrative to work perfectly well.
While this problematic relationship has been interpreted as a fundamental formal problem (, ), the diversity of views involved suggests that the problem is really a matter of fundamental but unacknowledged differences in the kind of experiences that different players are seeking. This paper presents a model in which different strategies for realising story and narrative within games are related to specific player preferences. The result is a framework in which perceived tensions between game play and narrative are accounted for as clashes between game play stylistic preferences one on hand and the formal structures of a specific game on the other. It can also be seen that these tensions have no privileged status in relation to other tensions between play preferences and game forms. Player preferences can be associated with specific kinds and combinations of game mechanics, as well as different strategies for integrating story forms with game play, including linear, nonlinear and emergent narrative structures. Hence this framework can contribute towards a structural approach to well formed game design aimed at specific player styles and preferences.
Player Types, Motivations and Play Styles
Different players may want very different kinds of experience from a game, and the interrelationships between player motivation, game mechanics and play style may be complex. A particular concern in game design is that a player may play in a style different from the play styles that a game is designed for. The play style goals of a game design should be explicit as a basis for deciding not only what styles to accommodate in the game mechanics, but also if and how to create the freedom for players to improvise around the mechanics if their play preferences are not strongly accommodated.
An influential scheme for categorising play styles has been developed by Richard Bartle (, ). Bartle’s scheme2 is derived from discussions between highly experienced MUD3 players on the topic of “What do people want out of a MUD?”. Bartle’s categories represent forms of player enjoyment identified in these discussions. The categories are:
- Achievers driven by in-game goals, usually consisting of some form of points gathering (eg. experience points, levels, or money).
- Explorers driven to find out as much as they can about the virtual world, including mapping its geography and understanding the game mechanics.
- Socializers use the virtual world to converse and role-play with their fellow gamers.
- Killers use the virtual world to cause distress to other players, and gain satisfaction from inflicting anxiety and pain on others.
Bartle notes that these categories are fuzzy and that players may cross over, although one play style tends to dominate the preferences of any given player. The categories represent differences in preferred activity, and these differences are reflected in the language and discourse patterns of the different respective player types. Bartle suggests that the categories derive from two dimensions of distinction, one between acting upon (killers and achievers) and interacting with (explorers and socialisers), and the other between players (killers and socialisers) and the world (achievers and explorers). Bartle also identifies specific game features favouring each category while suggesting that achieving a balance between the play styles is an important factor in the ongoing success of a virtual world.
John Kim  has examined player style in the context of live-action and table-top role-playing games (RPGs). This has let to the development of what he has referred to as the Threefold Model, regarded as a way of grouping many aspects of “group contracts” into logical categories. The group contract is a kind of collective agreement between a specific group of players covering every facet of how the game is to be played: mechanical rules, how scenarios are constructed, what sort of behaviour is expected of player characters, how actions not covered by the rules are resolved, the allowance of outside distractions, etc.. The categories of the Threefold Model are:
- Drama. The dramatist style values how well the in-game action creates a satisfying storyline.
- Game. The “gamist” style values setting up a fair challenge for the players (as opposed to the player characters). The challenges may be tactical combat, intellectual mysteries, politics, or anything else. The players will try to solve the problems they are presented with, and in turn the game master (GM) will make these challenges solvable if they act intelligently within the contract.
- Simulation. The simulationist style “values resolving in-game events based solely on game-world considerations, without allowing any meta-game concerns to affect the decision. Thus, a fully simulationist GM will not fudge results to save PCs or to save her plot, or even change facts unknown to the players. Such a GM may use meta-game considerations to decide meta-game issues like who is playing which character, whether to play out a conversation word for word, and so forth, but she will resolve actual in-game events based on what would "really" happen.”
These categories have both similarities to and difference from Bartle’s. In particular, Bartle’s system does not have a category corresponding to the dramatists. Kim’s gamer (or gamist) category appears to be the same as Bartle’s category of achievers, while Kim’s simulationist category appears more loosely analogous to Bartle’s explorer category. Bartle’s socialisers and killers are missing from Kim’s system. Although Bartle includes role-players within the socialiser category, in a general sense all of Kim’s categories cover role players and, as seen below, role playing cannot be simply equated with the Kim’s dramatist category. These differences may be indicative of a different purpose behind the systems, largely resulting from the different form of games involved. Live-action and table-top RPGs require social agreement in order to effectively realise the game, since the rules are implemented manually. However, it is plausible that the socialiser and killer categories could equally well provide the foundations of the social contract required for manual implementation of game mechanics. Bartle is dealing with computational worlds in which the social contract is generally implicit and enforced by the software mechanics of the virtual (game) world engine. In this case the categories become ones of play preference within the space of possibilities supported by the system, and the dramatist style becomes a plausible category omitted from Bartle’s scheme.
Yee  has built upon Bartle’s scheme, adding just such a category (the immersionist), modifying the other categories and also dropping the explorer category. Based upon empirical data provided by questionnaires answered by 6700 players, followed by statistical factor analysis, Yee identifies the following categories representing the primary high level grouping of responses:
- Relationship (socialisers): This factor measures the desire to develop meaningful relationships with other players in the game - usually in the form of a supportive friendship. Players who score high on this factor usually make good friends online, and tend to have meaningful conversations with their online friends, which usually involves talking about real-life personal issues. In times of need, these players can usually count on their online friends for emotional support. These players also tend to feel that they have learned things about themselves from playing the game, as well as gaining a better understanding of real-life group dynamics.
- Immersion: This factor measures the desire to become immersed in a make-believe construct. Players who score high on this factor enjoy being immersed in a fantasy world they can wander and explore. They tend to role-play their characters, and use their characters to try out new personalities and roles. They enjoy being in the company of other role-players. They also appreciate the sense of being part of an ongoing story, and oftentimes will think up a personal history and story for their characters.
- Grief (killers): This factor measures the desire to objectify and use other players for one's own gains. Their means may be both overt or subtle. On the overt side, they may enjoy dominating other players by killing them on the battlefield, or by taunting and annoying them. On the more subtle side, they may enjoy manipulating other players for their own gains, such as deceiving other players through clever scams, or begging for money and items. In either case, satisfaction comes from some form of manipulation of other players for personal gain.
- Achievement (achievers): This factor measures the desire to become powerful within the construct of a game. Players who score high on this factor try to reach the goals as defined by the game. They try very hard to accumulate rewards. For example, they try to optimise their experience point gains to reach the next level as quickly as possible. Or they may try to accumulate as much high level gear as possible. Or they enjoy doing massive amounts of damage to non-player characters (NPCs). The underlying theme is a desire to get bigger numbers. But the satisfaction comes from feeling powerful.
- Leadership (which can be regarded as a subcategory of socialisers): This factor measures the gregariousness and assertiveness of the player. Players who score high on this factor prefer to group rather than solo. They are often assertive individuals and usually drift to leadership positions when in a group. Because a group led by an indecisive leader often gets fragmented, the assertiveness of these players probably allows them to be effective group leaders in the game.
The game provides an important context underlying successful engagement and immersion for players with all of these motivations. It is in particular the presence of game play and unfolding stories that provide a sense of depth and purpose for socialisation, as well as the immediate input and play mechanics for achievers, killers and immersionists.
Interestingly, Kim’s categories of Drama (story and narrative), Game and Simulation as driving principles for social gaming contracts also underlie the independent isolation of these factors as formal systems within computer games. Aarseth , Frasca  and Juul  focus on the question of the relationships between narrative and game play in computer games. Frasca  focuses on the nature of a game as a simulation. Lindley (, 18]) brings these elements into a three-way relationship and highlights the three formal systems in terms of existential dependency, a hierarchy of temporal design concerns, and different and complementary sets of methodologies for design. The mechanics of a computer game may realise the designed formal structures at all three levels, but players may be more or less free to play creatively in a style of their preference, in tune or at odds with the design emphasis in the computer game artefact.
This freedom also underlies the many roles that a player may have in relation to their character and the game world, discussed by Pohjola  in the context of live-action role-playing as a question of identity and immersion. Different strategies for identification and immersion suggest very different strategies for narrative and story construction in computer-based role-playing games. Based upon Pohjjola’s discussion, it is possible to remodel the dramatist and immersionist categories into at least the following three attitudes towards drama/story/narrative within a game:
- the audience: passive reception of a narrative, ie. being told a story; this is the model implicit within the use of predefined cut scenes in commercial computer games to convey story elements designed by the game developers
- the performer: active performance of a character role within an unfolding story; further distinctions here might be made in terms of the degree to which the role and/or the story are predefined, as opposed to being created by the performer prior to or during the performance
- the immersionist: immersion of the player in the character, ie. the player/character distinction is dissolved into a unified persona within the game world; here too there is a question of the degree of character predefinition required to encourage immersion
While Bartle  has proposed specific game features supporting and encouraging the different specific play styles in his system, since he had no dramatist/immersionist category the features required to support this play style remain unspecified. Bartle’s later work  nevertheless presents a model of the degree of immersion of a player with their character along a continuum ranging from the avatar at one extreme to the persona at the other, reflecting the performer/immersionist distinction. The avatar is a dramatically empty shell, being an instrumental vehicle for the player within the game world. The persona is, as described above, the immersionist state of total player identification with their character, representing the player’s being within the game world. Between these extremes lies the performed character.
Bartle’s  categories of play style emerged from what he describes as heated discussions about what makes a game good, the resulting distinctions accounting for differences of taste underlying player conflict. Similar controversies around the use of cut scenes in computer games can be seen to reflect different player preferences in relation to the story aspects of games, not only between gamers who are not interested and others who are interested in the story, but also between the different tastes for how story elements are manifested, a conflict between preferences for being told a story (the audience play style), performing a story (the performer play style) and immersion within a story (the immersionist play style, specialised here from Yee’s use of the term in order to align the term with its use in role-playing theory).
These distinctions provide a perspective for the following analysis of methods and game features addressing issues of story and narrative. It is first necessary to define some basic terms in narrative theory, and then to consider how these terms map onto interactive computer games. It is then also necessary to consider the overall design of time structures in computer games, interrelating the simulation substrate of games with higher level time structures involved in the design of games and narratives. Different strategies for realising narrativity in relation to the player preferences in relation to story are mapped out in terms of these different levels of temporal design.