I: Introduction

The relationship between game and story, rules and fiction, is an avidly discussed and troubled one. It is also one that bears great potential for the game-play experience. It is the combination of rules and fiction that makes games such potent tools for meaning generation, because fiction adds a referential framework to a game, a context that renders the rules meaningful in a sense that exceeds the strict confines of the game’s system. With the introduction of worlds, characters and conflicts, fiction can expand the thematic range of games and address a variety of human source concerns, for example freedom, love, security, confirmation of social values and believes, thus potentially enlarging a game’s emotional bandwidth. The integration of fiction into games enables thought provoking, deeply moving narrative experiences that are unique to the interactive form.

Of course, this is easier said than done. We all have played games with elaborate backstories, which nevertheless failed to provide emotions outside the continuum of frustration and triumph, because the fiction lost relevance during game-play. Game characters often feel shallow and plain, relationships to NPCs (non player characters) lack emotional depth. It is one of the most taxing challenges of contemporary game design to couple rules and fiction in a way that reinforces their particular strengths instead of getting into each other’s way. This article asks how fictional elements can be made tangible to the player and kept in the foreground during game-play, in order to tap the narrative potential of computer games. Our goal is to identify strategies how to successfully couple rules and fiction.

But before we can offer solutions, the problem must be clearly stated. Drawing on emotion psychology as well as game theory, we will argue that the reason for the proclaimed lack of impact fictional elements tend to have on the game-play experience is that these elements have a considerably lower reality status than the rules from which the actual game-play arises. Lower reality status means lower emotional impact on the player. According to Frijda’s law of apparent reality »emotions are evoked exclusively by events that are appraised as real and their intensity corresponds to the degree to which this is the case.« [1]

Part of the solution will thus be to emphasize the emergent narrative of games [2], because fictional elements that are enacted rather than presented through more classical modes of narration (for example cut-scenes) gain higher reality-status, thereby raising the emotional impact on the player.

Therefore, fictional elements have to be integrated into the moment-to-moment game-play, must be made enactable through the player to get her attention. But game design is the art of abstraction. There will and always have to be parts of the gameworld which are not implemented into the rule-system, or the game would become unplayable. The ›magic‹ happens in the gaps between the real and the simulated world. [3] So, which elements of the fictional world are to be integrated into the game-play in order to deepen the emotional experience and enable an inside view that is such a rare effect in current computer games?

The answer to this question shall not be left to the taste and whims of game scholars and -designers, so a useful framework is needed to systematically identify the relevant aspects of the gameworld that should be transferred from the purely fictional to the realm of the virtual, for example enactable. We suggest that Roland Barthes’ research on the functional units of literary texts – which is to be adapted in regard to computer games later in this article – may serve as such a framework.

Barthes differentiates between functional units of doing and functional units of being. While functional units of doing describe narrative elements that advance or extemporize the narrative plot, functional units of being provide us with explicit or implicit information about the state of mind of characters, about philosophical or ideological implications of the gameworld.

Discussing the various possible ways to integrate functional units of doing into the game-play in other than the usual ways might open quite a different perspective on questions regarding the structure of computer games and may provide the subject of a future paper.

As this current article focuses on ways to expand the emotional rather than structural possibilities of computer games by translating aspects of the narrative into the game-play, the emotional aspects, covered by Barthes’ functionality of being, will serve as starting point for our explorations. Finally we will attempt to identify different strategies to successfully integrate these fictional elements into the game-play and illustrate them with examples from existing games.

II: Problem Description

Reality-status & Emotions

Because dramatic computer games contain rules and fiction, they have the potential to produce a wide range of deeply moving, thought-provoking experiences. The fiction addresses human source concerns, for example freedom, love, security, confirmation of social values and believes. In the course of events these source concerns are endangered and the wish to restore the desirable state provokes a change in action readiness, which psychologists have defined as emotion. [4] For a game to become meaningful outside the strict confines of the game system and to expand the emotional range of the playing experience, it is necessary to include fictional components to a significant degree. We follow Jesper Juul’s definition of fiction and the distinction he makes between fiction and storytelling:

›Fiction‹ is commonly confused with ›storytelling‹. I am using ›fiction‹ to mean any kind of imagined world, whereas, briefly stated, a story is a fixed sequence of events that is presented (enacted or narrated) to a user. Herman Melville’s Moby Dick is a story and a fiction, whereas a painting such as George Seurat’s La Grande Jatte is a fiction but not a story since it only presents one moment in time. [5]

What makes games special compared to traditional media, however, is the additional operational level of the rule-system (the other two being the level of interface and the level of fiction) [6] >from which arises the game-play. The rule-system introduces the game-specific source concerns agency and cognitive/sensorimotor competency. Jesper Juul accounts for the concurrence of rules and fiction in computer games by describing games as half-real:

Video games are real in that they consist of real rules with which players actually interact, and in that winning or losing a game is a real event. However, when winning a game by slaying a dragon, the dragon is not a real dragon, but a fictional one. To play a video game is therefore to interact with real rules while imagining a fictional world, and a video game is a set of rules as well as a fictional world. [7]

For a systematic exploration of computer games’ emotional potential, Juul’s observation is crucial. According to Frijda’s law of apparent reality »emotions are evoked exclusively by events that are appraised as real and their intensity corresponds to the degree to which this is the case.« [8] Thus, games have an intrinsically high reality status, which makes them so extraordinarily involving. It is important to note though that »[t]he law of apparent reality makes no pronouncement on reality in any objective sense, but rather on the impression of reality created in the subject.« [9] This explains why it is possible that fictional events can also elicit genuine emotions, if a) the recipients willingly suspend their disbelief, and b) the medium as such employs strategies to ensure the verisimilitude of the fictional world and events. These strategies do not have to aim at real-life realism. [10] According to thought theory, emotions can spring from imagination. [11]

Reality-Clash

Taking the high reality-status of the third operational level into account and the fact that fiction, too, can evoke genuine emotions, how come that the mere presence of both components in a game is not enough to tap games’ potential to provide a broad range of emotional experiences, not just those within the continuum of frustration and triumph? In his article Genre Trouble: Narrativism and the Art of Simulation, Espen Aarseth claims that there is not only an arbitrary connection between rules and fiction,

Any game consists of three aspects: (1) rules, (2) a material/semiotic system (a gameworld), and (3) game-play (the events resulting from application of the rules to the gameworld). Of these three, the semiotic system is the most coincidental to the game. [12]

…but that the fictional elements of a game somehow dissolve during game-play:

Likewise, the dimensions of Lara Croft’s body (…) are irrelevant to me as a player, because a different-looking body would not make me play differently (…). When I play, I don’t even see her body, but see through it and past it. [13]

He ascribes the irrelevance of fiction for the experience of game-play to the fact that the pleasures of computer games are primarily kinaesthetic, functional and cognitive. [14] We have no intention of regurgitating the whole ludology versus narratology debate here, and although we do not share the ludologists’ standpoint that games principally resist meaning or that fiction is unimportant for the playing experience (a position ludologist Jesper Juul himself is revising in his book half-real [15]) it has to be acknowledged that the relation between rules and fiction is not a straightforward one. This is especially true in heightened states of play, when one is struggling with a particularly difficult task. In these situations »contextual background is likely to be reduced to the more distant background, the game-play situation taking the shape of an abstract problem to be overcome rather than one that retains much in the way of contextual depth.« [16]

>From the perspective of emotion psychology, the reason that fictional elements of a game tend to slide into the background while the game-play occupies most of the player’s attention is the difference in reality-status between these two components. The realization of action tendencies evoked by the events one encounters in a game via game-play is more real (immediate, salient) than the fictional context, which in most games is conveyed in a non-enactive way. The intrinsic reality-status of the rule-system further explains why the game-play experience is not spoiled by incoherencies in the fiction, as long as the things one can do are sufficiently rewarding and fulfill the game-specific source concerns for agency and cognitive/sensorimotor competency. [17]

So, one could argue that there is an obvious clash of realities between the two main components of a game and that the fiction, due to its lower reality status, has so much less impact on the player than the game-play that its potential to expand a game’s emotional bandwidth stays unrealized.

Tapping the Potential

Clearly, we are facing a dilemma: on the one hand, it seems to be very hard for fiction to have an impact on the players who are absorbed by the things they can actually do in the game, on the other hand, integrating fiction into a game is the precondition for these things to become emotionally meaningful in a more encompassing way. The solution to this dilemma seems fairly obvious: for the fiction to become tangible to the player, it must be rendered relevant for the moment-to-moment game-play. We thus argue that the key to emotionally rich, thought-provoking and meaningful games is to emphasize the emergent narrative of computer games. According to Eric Zimmerman and Katie Salen, emergent narrative »arises from the set of rules governing interaction with the game system. Unlike embedded narrative, emergent narrative elements arise during play from the complex system of the game, often in unexpected ways.« [18]

In short: fiction must be implemented into the rule-system in order to become tangible to the player. This is not a new idea. In fact, it has already been practiced for a very long time, although only in regard to very specific aspects of a game’s fiction.

Game Design: the Art of Abstraction

The claim that fiction tends to recede into the background during game-play to the point of becoming irrelevant for the playing experience is imprecise and too generalized. There are a lot of fictional elements that stay in the foreground just fine during game-play. We cannot ignore the wall that blocks our way, the turnings of the labyrinth or the monster that attacks us. We will make a deliberate decision about the particular car we steal in The Godfather, because we soon learn that some are much faster than others, while the faster ones are harder to steer. All these elements clearly stem from the realm of fiction. So, obviously, the reality-clash explained above concerns fictional components to varying degrees, and the degree depends on whether the element is implemented into the rule-system or not. In his article A Certain Level of Abstraction, Jesper Juul explains:

If we assume the perspective that games have two complementary elements of ›rules‹ and ›fiction‹ all content in a game can either be purely fictional and not implemented in the rules (such as in the case of a game’s backstory), purely rules and unexplained by the fiction (such as multiple lives of a player), or in the zone in between, where the rules of the game are motivated by the game’s fiction (cars that can drive, birds that can fly, et cetera). The combination of rules and fiction is sometimes described as ›virtual‹ or ›simulation‹. [19]

The aspects of a game’s fiction that are most commonly integrated into the simulation are those which define the course of events. By drawing on Roland Barthes’ functional units of literary texts as a framework for our explorations, we will later refer to these elements as ›functional units of doing‹. They constitute correlations in the game’s structure, for example, finding ammo correlates with its later use; deciphering a code correlates with the opening of the safe door; encountering the monster correlates with battling it. [20] Barthes’ model is going to be explained in detail later in this text.

The result of emphasizing the implementation of functional units of doing into the simulation is that the emergent narrative of the game becomes very plot-driven. Accordingly, the game offers only an ›outside view‹ on the gameworld and its various characters and objects and thus lacks emotional depth and scope.

The fictional elements that are rarely integrated into the rule-system and therefore suffer most acutely from the aforementioned reality-clash are those which would enable an ›inside view‹ on the game’s events. Again in reference to Barthes, we are shortly going to introduce these elements as belonging to the category of ›functional units of being‹. While the functional units of doing constitute structural correlations (Barthes also calls them ›distributional‹), functional units of being are ›integrative‹, because they do not correlate with consecutive or complementing actions, but convey implicit or explicit information about game characters, their personality, motives and their sentiments towards other characters, as well as all kinds of cultural, philosophical, and ideological implications that constitute the atmosphere and significance of the gameworld. [21] Currently, these are almost always condemned to the backstory, if game designers bother to think about them at all. In the following we will argue that it is exactly this category of fictional elements, especially the subcategory of indices, which contains computer games’ potential to expand their emotional range and provide players with insights into the human condition.

Barthes revisited

Roland Barthes’ research on the functional units of literary texts provides us with a useful framework for a systematic identification of fictional elements which – if implemented into the simulation – would potentially enable aforesaid inside view into the gameworld and therefore expand games’ emotional scope and intensify the player’s experience.

According to Barthes, every element in a literary text has a specific function, is meaningful for the text to a greater or lesser degree and in different ways. Barthes distinguishes between ›nuclei‹, also called ›cardinal functions‹ and catalysers on the one hand, and indices and informants on the other hand.

The nuclei and catalysers both correspond to a functionality of doing. The nuclei constitute the hinge-points of the narrative. [22] They refer to actions which open (or continue, or close) an action alternative that is of direct consequence for the subsequent development of the story, in short: they inaugurate or conclude an uncertainty. Because a story that consisted only of nuclei would be quite boring (boy meets girl, she sees him with a blond supermodel, they quarrel, separate, reunite and live happily ever after), catalysers are integrated to fill the room between the nuclei and provide the retardations that are necessary to create tension. The catalyser’s functionality is merely chronological and not consequential, they do not have the power to change the course of events.

The functional units of the second category, the indices and informants, correspond to a functionality of being. They fill the text’s structure with implicit or explicit information about the state of affairs in the fictional world, they offer an inside view on how things are. Indices refer to the identity of a character, a feeling, an atmosphere or a philosophy and have to be interpreted. They are not straightforward, but refer to an implicit meaning. A trembling hand can indicate fear, cold, sickness or excitement. The description of a stormy night can contain more than the information that the weather is bad. It constitutes an atmosphere that might indicate approaching danger. Informants on the other hand are pure data, whose main function is to orient the recipient in space and time or provide explicit information, for example, the exact age of a protagonist.

Certainly, the emotional impact of any text is constituted by a combination of all these functions. The functional units of doing provide the dramatic arc, the structure that regulates our interest by systematically revealing and withholding information. But no tension can arise from withheld information if we have no one to root for, and we will only root for the heroine if we emphazise with her. The human source concerns the fiction potentially addresses must be made tangible to us, and this is the task of the functional units of being. To a certain (not entirely small) extent, thanks to the phenomenon of narrative comprehension, we accomplish this ourselves (not just when reading a book, but in everyday life, too). We tend to fill the gaps in a merely chronological account of events by reasoning about cause and effect, thus narrativizing events and constructing our own inside view on the world. [23] But when sitting down to enjoy a designed experience, we expect the author/
designer to at least meet us halfway.

In short: while functional units of doing construct the plot of the narrative, it is the functional units of being which turn the consecutive events into a story, by adding an inside view that explains the causes for events and the reasons and motives for actions. It is important to note that units can have multiple functions, and this is indeed often the case. A nucleus is a turning point in the narrative, giving the plot a new direction, but it can also convey information about, say, the protagonist’s childhood trauma that prevented him from rescuing the drowning fiancée. Catalysers, too, can have additional indexical function. And indices themselves certainly always start as informants, because before an element can aquire a connotative meaning, it first must have an explicit, denotative one. A stormy night might be a metaphor for the uprising revolution, but on a more basic level it still is a stormy night.

By now, it should be sufficiently clear why we decided to focus on Barthes’ functional units of being in order to explore games’ potential to expand their emotional range and provide players with an inside view on the fictional world and its elements. But since Barthes built his model with the medium specific characteristics of linear, non-ergodic texts in mind, some adaptations must be made in order to instrumentalize the functional units of being to tap this potential in the context of computer games.

The main difference between computer games and traditional texts is that games contain the additional operational level of the rule system (see above). Still, in games, too, every element serves a function. Games, just like any other form of mediated experience, are designed and everything that is included or excluded from the final product – be it on the level of fiction, interface or rule-system – can be regarded as the result of a deliberate decision and shapes the way in which the game is perceived, played and experienced. It is important to note, though, that while it is possible to unexceptionally explain every element in a traditional text with the help of Barthes’ units, one would be hard pressed to do so in regard to computer games – at best, it would become rather confusing, because one had to constantly renegotiate the function of a specific element in reference to a particular operational level. For example, there are two sorts of indices in a game due to the level of the rule system: those which follow Barthes’ definition and that refer to the state of affairs in the fiction, and a second form, which serves to communicate system information to the player.

The beauty of Barthes’ model, however, is that it is clear and unambiguous. We do not intend to stress it by forcing it onto games to explain their various elements exhaustively. We want to use it as a source of inspiration to systematically identify emotionalizing strategies in games. Doing so, it cannot be emphasized enough that while functional units in traditional texts are received and interpreted by a reader/viewer, functional units in games can be enacted by the players. It is quite obvious what this means in regard to functional units of doing. Instead of having an author describe how the protagonist recklessly drives a car from A to B, in games we expect to and can do it ourselves. Games offer numerous possibilities to enact catalysers. A little less frequently, they even provide players with the opportunity to steer the course of events in a completely different direction by letting them enact nuclei.

We claim that functional units of being can also be enacted by the player. To be precise, this concerns indices rather than informants, because if a certain condition or state of a fictional element shall be made tangible to the player through acting, this cannot be done directly but only through action that indicates this state of affair. Games are procedural systems. If we want to get a message across, it must be embedded into this system, or its information value will be lost due to the higher reality status of the rule-system. Conditions that are coupled to certain actions are always referential. That the heroine is claustrophobic is an explicit information (=informant). If we want to make this information tangible to the player, providing an inside view that offers a deeper understanding of and emotional engagement with the character, one must first identify the specific symptoms of claustrophobic fits, thereby creating indices that refer to the explicit information, which can then be transferred into concrete game-play challenges and actions. Such a procedure prevents the fictional information »this character is claustrophic« from receding into the background.

III: Strategies

Up to now, we have tried to show what elements of the fiction can be integrated into the rule-sytem and thereby made enactable in order to broaden the emotional range of player experience. Adapting Roland Barthes’ concept of functional units for use in computer games, taking the additional level of the rule-system into account, we have decided to leave the functional units of doing aside for the time being, as they are important for the game’s structure rather than its emotional implications.

Instead, we have focused on the functional units of being, especially the sub-category of indices, which, if made enactable, convey information about the feelings and beliefs of a character, about philosophical or ideological implications of the gameworld, in short: provide an inside view on a game’s fictional elements, thereby making the gameworld more tangible to the player instead of being mere background information irrelevant to the game-play.

The following examples illustrate strategies how to create indices through player action and can be divided into three categories. What distinguishes them from each other is the extent to which the game-system dictates player action in order to achieve a certain effect. Is it necessary for the player to behave in a certain pre-defined way to beat the game, is it possible to choose different playing styles at the cost of making the game harder or less satisfying to play, or is the player given true choice on how to realize her interpretation of elements like the avatar’s personality, the relation between characters or the atmospherical, philosophical and ideological implications of the gameworld?

Strategy I: Force

What defines this category is that certain player actions are made cardinal on the level of game-play in order to force the player to realize units, which correspond to indices on the level of fiction.

In the survival-horror-game The Thing, based on the John Carpenter movie, the player finds herself in an antarctical research facility, whose inhabitants have recently been wiped out by an alien viral infestation, leaving the place deserted except for numerous bizarre monsters that now threaten the player’s life and hinder her escape.

Unluckily, the research facility consists of several buildings that can only be reached by crossing snowstorming outdoor areas, still all of the buildings have to be visited in order to beat the game. Every time a building has been cleared of alien infestation, the player has to find her way in the freezing night, and the danger of being killed by alien monsters alternates with the danger of freezing to death. The cold is not a factor merely represented by narrative means such as sounds, animations or dialogue; as soon as the player leaves a building a little blue bar appears on the screen, which indicates how much longer the avatar can resist freezing to death. Simply being outside makes this ›cold meter‹ diminish, and when it reaches zero, the avatar will consequentially die.

As a result, the antarctic cold is a constant and threatening factor in the player’s fight for survival, it takes the form of an omnipresent enemy, and the seriousness of its effects require constant player action in order to avoid defeat. Weather conditions, which might be considered as merely indexical on the level of narrative, become an obligatory requirement to act upon.

Space is, to a certain degree, transferred from the fictional to the simulation in most games. Yet, the physicality of space is most commonly reduced to the game-play function of navigation. Sensible dimensions of space such as temperature, humidity or smell, which significantly shape our perception of real-life spaces, are seldom integrated into the rule-system.

While the cold meter is a rather simple feature, it transforms the narrative element of the antarctic cold into a key factor of gaming decisions, therefore making the atmosphere of the gameworld more tangible to the player than narrative means could achieve; the pressure of finding a warm spot just in time becomes omnipresent, and moments of disorientation in the open easily lead to an actual surge of panic.

Strategy II: Seduction

By seducing the player to act in a certain way, she is not forced to do things in order to realize a predefined indexical meaning, but is allowed some autonomy to act out her own interpretation of the game and its elements. It is not necessary to act in a certain way in order to beat the game, still, as we are talking about seduction, there are actions that promise an easier or more successful way to achieve the game’s goals.

In the games of the Hitman-series, the player adopts the role of a contract killer (the hitman) and has to beat several self-contained missions, commonly consisting of the elimination of a specified target person. On first sight, one might assign the Hitman games to the shooter-genre; yet, when playing the game it becomes clear that shooter strategies are not what the game designers had in mind. Mainly, the very low health of the avatar, the absence of med-packs and the tricky AI of the numerous bodyguards who usually watch the target person make the game a hard and unsatisfying one when played in the way associated with classic shooters.

But the Hitman-games offer a variety of more sophisticated alternatives to hit-and-duck strategies, one of the most dominant ones being clever disguises.

The disguise of a postman, for example, might enable the player to enter a house without raising the bodyguards’ suspicion, the janitor’s clothes might offer a safe way to the cellar in order to manipulate the elevator brakes, and dressed as a waiter one might even get close enough to the target person to switch his drink with your special Long Island cyanide potash. They never heard you coming, and when the target drops dead, the hitman has already left the building.

Following the trail of encouraged procedures, the player gradually adopts the contract killer’s way of thinking; the need to remain unrecognized dominates each of the player’s steps, the desire to cast surgical rather than devastating blows, while not necessary to beat the game, becomes a highly estimated goal in the emergent narrative, and patrolling enemies that would quickly be disposed of without the ideological implication of the professional’s code of conduct are sought to be tricked and avoided.

The fact that there is an alternative to this seductive course of action makes it even more emotionally compelling to the player, as there is a decision to make which requires the player’s action-readiness and grants her limited responsibility over the way the avatar’s personality is enacted. Still, there is not much of a choice in seduction, as it is always a choice between better and worse: the encouraged procedure contains the promise to heighten the player’s experience of the game by letting her enact crucial elements of a hitman’s trade, therefore granting an inside view of one of fiction’s most mysterious archetypes. Playing the game shooter-style, letting the many possibilities to become a highly competent master of deception slip, is still possible; yet, it would feel a bit like riding the ghost train with your eyes closed.

Strategy III: Choice

While seduction offers a feeble choice between better and worse, real choice aims at a decision between different outcomes of similar value. The question of free choice addresses highly complex matters of player freedom: are we thinking of choices never explicitly integrated into the game, or are we talking about choices between different pre-scripted paths? Does freedom of choice mean that the player can significantly influence the outcome of the game, or the freedom to enact the avatar’s path in a personalized way, thereby fulfilling the emergent narrative which results from moment-to-moment game-play, but is assembled by the players themselves?

It is, again, important to note that the strategies presented here do not aim at the structural possibilities of games, but at the enactability of indexical functions. While free choice regarding the game narrative’s cardinal functions could potentially initiate a completely different course of the game itself, right now it is important how the player’s decisions influence the way she takes control over the avatar’s feelings and believes, shaping the implications of the gameworld in regard to the avatar’s perspective.

Giving the player meaningful choices is a powerful yet tricky strategy to make particular aspects of the gameworld tangible. It is crucial to think about the game’s fiction as a system of interconnected elements that becomes palpable when the relevant aspects of the gameworld are integrated into the rule-system. Yet, it is sometimes necessary to think around corners to determine these relevant aspects. Which elements can be convincingly incorporated into the game-play and what they indicate in the larger context of the fictional world is not always straightforward. It might help to remember that units can have multiple functions on the narrative level, for example, indicate more than one implicit meaning.

The fictional scope of Everquest 2 develops from the conflict between two metropolitan cities, the City of Qeynos, »a beacon of hope in a sundered world«, governed by the just Queen Antonia Bale, and the City of Freeport, ruled by self-declared Overlord Lucan D’Lere and his corrupt militia. It is obvious that this conflict clearly alludes to a motif of good and evil. So far for the fiction; the actual game-play stays mostly untouched by the opposition between the cities, which is only referred to in quest-dialogues and represented in the audio-visual design of the two locations. Citizens of Qeynos and Freeport alike will travel the world hunting the same mobs, getting offered the same quests and even being able to form groups together. For Qeynos citizens, killing comes just as natural as for citizens of Freeport, the only difference being the reasons delivered on the narrative level to legitimate the act; Qeynos citizens are required to kill in order to keep the townfolk safe, while Freeport citizens are instigated to kill for money.

But there is one aspect where the choice for a home city becomes relevant for the game-play: the choice for a character class. While nowadays a class is selected before one starts the game, like in most fantasy RPG’s, in the early days of Everquest 2 the decision was made only after one had entered the game. After deciding for either of the two cities, players started out as untrained commoners, and only after they had become acquainted with the basics of the game, they had to choose their class archetype, becoming either a mage, a fighter, a priest or a scout. Later in the game, and regardless of the city they were living in, they were given the choice of specializing in a certain class of their archetype. Scouts, for example, now had the choice to become either a rogue, a bard or a predator. Finally, as soon as players reached Level 20, they advanced in a sub-class of their chosen profession, the specializations for a predator, for example, being either ranger or assassin.

The tricky part was that, contrary to previous choices, the sub-class professions were city-aligned: the ›good‹ rangers could only live in Qeynos, while Freeport was the natural habitat of the ›evil‹ assassin. Being a predator in the city of Freeport meant that you could either comfortably become an assassin, or you had to betray Freeport to Qeynos in order to become a ranger instead. The choice was not an easy one, since the betrayal quest line was one of the hardest in the game, still, the abilities of the two sub-class professions were quite different, allowing for two very different ways to play the game, so it would really matter to make the right decision: while the ranger was highly specialized in devastating bow-attacks and was able to move easily and quickly through the wild, assassins were experts in sneaking undetected after their prey in order to kill them quickly with cunning close combat dagger abilities.

As one of the developers’ most highly estimated goals was class-balancing, none of these two classes was ›better‹ than the other, but the different play-styles offered by the specialized abilities made quite the difference nonetheless for the game-play experience.

So, how was this decision made an ideological one during the game-play?

If the player already resided in the city that best reflected her approach to the gameworld, her favored play-style would already find support in the cities ideology: feeling at home amidst the wily and selfish townsfolk of Freeport already qualified you as a cunning assassin, and the sneaky play-style and poison-dripping daggers you already came to know as a predator would much better suit your desires than betraying to Qeynos and becoming an altruistic treehugger (as qeynosians are often referred to in the game). But if the cunning play-style of the assassin left the player feeling guilty, desiring to make good to the world, becoming a qeynosian ranger might just be the right thing. During the betrayal quest, players had the chance to enact the ideology of their new class by the specific tasks one had to fulfill in order to become a citizen of a particular city.

While these tasks were the same for each profession (predators had to do the same things as clerics or warriors), the good- or evil-aligned tasks, in combination with the basic abilities of the professions, let the player enact the key ideology of her new profession: a predator collecting garbage in the woods surrounding Qeynos would feel much like a ranger, while a predator spying out conspirators against the overlord would make much use of his assassin-related skills.

This way, the narrated ideology of the particular city became closely connected to the enacted ideology of the corresponding professions, making the choice a key factor of experiencing the games’ ideological implications through the player’s action. But no matter what the decision, it would always be a deliberate one, a trade-off between one profession’s attitude and the other. While the decision would essentially affect the way the game would be played, it still left the player true choice, as none of the possible decisions would neither positively nor negatively affect the chances to beat the game.

By allowing the player a certain degree of autonomy in the choice of a particular way to enact different ideologies, she is encouraged to make her playing style matter in terms of the gameworld’s fiction. The connection between the fictional ideologies of the two cities and the different playing styles assigned to the two professions allows for an emotional assessment of the fictional situation, at the same time creating action tendencies to be acted out in the actual game.

IV: Conclusion

Starting point of this article was the claim that computer games have the potential to be deeply moving and thought provoking, but that this potential is currently almost untapped. The article suggested that a precondition to expand games’ thematic and emotional scope is the integration of fiction into the rule-system, but stated also the problematic relationship between these two game components.

By drawing on game theory and emotion psychology, the problem between rules and fiction was explained as an imbalance in reality-status between rules and fiction. We argued that due to their higher-reality status, the rules in the sense of game-play are simply more involving than the fictional components with their lower reality-status. The rules win the battle for the player’s attention and force the fiction into the background. Key question of this article was how fictional elements can be made tangible to the player and kept in the foreground during game-play.

Obviously, to tap games’ potential to provide a wide range of deeply moving emotional experiences, the fiction must be integrated into the rule-system. We suggested that integrating especially those aspects of the fiction which provide an inside view on the characters and events, would heighten the emotional significance of the playing experience.

To identify these elements, Roland Barthes’ functional units of literary texts provided us with a useful framework. According to Barthes, every element in a narrative text serves a specific function. He distinguishes between two main categories of functions: functional units of doing, which advance (cardinal functions) and extemporize (catalysers) the plot and which constitute the structure of a given text; and functional units of being, which give explicit (informants) or implicit (indices) information about characters, their personality, state of mind and relationships to other characters as well as the nature and ideological or philosophical implications of the fictional world.

As this current article concentrated on ways to heighten the emotional rather than structural possibilities of computer games, our focus lay on the functional units of being, specifically the indexical functions. We identified three strategies to realize indexical functions via game-play. The distinguishing element of these strategies is the element of player freedom in the realization of indices. In the first category player action is dictated by the game, thus allowing the designers a lot of control over the effect the strategy will have on the player. The second category does not force the player to do things in order to realize a predefined indexical meaning but rather seduces her to behave in a certain way. The third category is characterized by offering the player a true choice, thus encouraging her to act out the significance of different playing styles in the context of the game’s fiction.

We do not claim that these strategies present a fool-proof recipe for more emotional games, but they are a good starting point. They are intended to help game designers to think about game-play mechanisms in a more systematic way and hopefully help them identify the relevant aspects of the gameworld that should be incorporated into the rule-system in order to tap the fiction’s potential to provide a wider range of emotional experiences.