1. Introduction
[1.1] Before we encountered Mary Sue in fandom circles, we had already met her elsewhere: in our computer game and role-playing sessions.
[1.2] As fans and gamers, we are at home both in fandom and in the various story worlds of role-playing games (RPGs). In this essay, we take a closer look at the Mary Sue concept in both RPGs and fan fiction. Using concepts from ludology and narrative theory as a basis, we scrutinize this figure against the backdrop of our practical knowledge.
[1.3] We define RPG broadly as a character-focused game that features rule-based, not skill-based, action. Because we are drawing from our extensive experience of Western-style role-playing games in various media, we deliberately chose this genre angle. One reason why it works so well for our purposes is because, like many fan works, RPGs are character driven and by extension narratives play a significant role.
[1.4] Frauke has been a player and game master of pen-and-paper RPGs since long before she was absorbed by fandom. To her, both experiences always felt connected. She has been intrigued by the difference between writing her own original characters and playing them in role-playing sessions, and writing borrowed characters in her fan fiction.
[1.5] Although Julia has never role-played with pen and paper, she is a computer and console RPG geek and has also participated in free-form writing RPGs. One example of a computer/console RPG is Mass Effect, a successful science fiction shooter hybrid first released for Xbox 360 in 2007 by BioWare. It created a media storm when Fox News decried the game's supposedly raunchy lesbian sex scenes. The lesbian sex is a fact: the user may choose to play a female version of the central player character (PC), Commander Shepard, and one of the optional romance quests involves a female-coded alien. There is, however, no such option for gay romance when the PC is male. Disappointed by the missed opportunity to pair male as well as female Shepard with the human male character Kaidan, we now "ship" female Shepard with the alien male character Garrus. Sadly, Mass Effect fan fiction for this particular pairing has a limited appeal: as a customizable stand-in for the game's player, Commander Shepard in fan fiction functions by definition as the writer's Mary Sue.
[1.6] The typical red flags of self-insertion and wish fulfillment are necessary but insufficient criteria for identifying a Mary Sue. Using only these criteria is problematic because it places any original character who experiences outrageously exciting adventures under suspicion of being a Sue—especially if female. This is why we adopt Alara Rogers's (2003) definition of Mary Sue as "an original character who overshadows the canonical cast." That is, the Sue's defining feature is her habit of distorting other characters and warping the fictional world around her person.
[1.7] In fannish circles, the term "Mary Sue" is often used as a slur. In contrast to this derogatory connotation, our position is that she is not evil in herself, but rather her effect depends on the context in which she appears. The Mary Sue phenomenon functions quite differently if seen from a fan fiction versus a gaming perspective.
2.
[2.1] As the theoretical basis for our ideas, we use concepts from game studies, or ludology. Basically, we want to look at the contested relationship between narration and simulation. "Ludologists love stories, too" (Frasca 2003a), but they focus on game play and game mechanics over story aspects, and simulation is a key concept for modeling those. "[T]o simulate is to model a (source) system through a different system which maintains (for somebody) some of the behaviours of the original system" (Frasca 2003b:223). The rules governing this system—the simulation—are often invisible, but can be informed by ideology as much as the actions or content within (like scripting mainly heteronormative romance options).
[2.2] Narration, on the other hand, is commonly understood as "the representation of a story," meaning that a story exists independent of its rendering (adapted from Abbott 2007:39–40). What actually qualifies as narration versus story, however, is less clear-cut and is subject to individual opinion.
[2.3] Simulation and narration have much in common, but while "their semiotic sequences might be identical…simulation cannot be understood just through its output" (Frasca 2003b:224). We agree, but to turn this on its ear, we argue that story (the sequence of events) likewise cannot be understood only through narration (the output that we read). In our view, the two concepts (narration and simulation) do not oppose each other: instead, stories are rendered into narratives through simulation. To put it the other way around, narratives are the output of running story simulations. After all, the most basic simulational model is to calculate a situation in one's mind and play games of "what if?"—something that fan storytellers are more than proficient in.
[2.4] So far, so good. But what makes this combined simulation-narration model useful for theorizing the Mary Sue?
[2.5] As per our definition, a Sue story is not characterized primarily by self-insertion of the author but by the distortion of the canon to accommodate her wish-fulfillment fantasies. If viewed through a ludological lens, the Sue's text fails as narration because it is the representation of a highly personalized simulation that has been "inadequately transformed" into "normal story values" (Nielsen Hayden 2003). To be more precise, the Mary Sue writer is playing an RPG of one. It is not the output—the narration—that is important to her. It is the simulation, the state of playing, a game of make-believe that is at the heart of the Sue experience. Clearly, we are talking about the realm of daydreams and personal fantasies. But in a fannish context, these fantasies often take place in Middle-earth (Tolkien), on Faerûn (Dungeons & Dragons: Forgotten Realms), or on the SSV Normandy (Mass Effect). The record of this experience, the resulting game log of sorts, is usually only of interest to the Sue writer herself, and to those few who share her very specific preferences.
[2.6] Thus, classifying the Sue story first and foremost as if its primary purpose was to function as narration is analogous to reading a record of a group's epic Dungeons & Dragons campaign, or the role-play journal of a LiveJournal Hogwarts simulation, as if they were meant to function independent of their game contexts. Much like watching someone else play a racing game on their computer, these recorded experiences are of limited interest to nonplayers. The tragedy of the Mary Sue is that her game logs are classified as, and read just like, "primary" narrative texts.
3.
[3.1] If Mary Sue fan fiction is an unintentionally misclassified game log, then looking at her from a gaming context should result in a kinder assessment. But unlike the Sue story, not all games are single-player experiences, and that introduces a new set of catches.
[3.2] We use pen-and-paper RPGs here as an example of cooperative creation of a shared game experience. In fantasy RPGs like Dungeons & Dragons, self-insertion and wish fulfillment are both not only unproblematic, but they are actually encouraged. Player characters are referred to as heroes, and the genre expectation is that they are forged as heroic characters, not modeled after average people.
[3.3] So if we have a group of five larger-than-life storybook heroes and heroines who go on new adventures every week, how can the presence of a Mary Sue still disrupt this experience?
[3.4] The most important factor here is that there is not a lone hero, but a group that constitutes a heroic team. This would also be the case in, for example, a LiveJournal fandom RPG, where every character belongs to the main cast of the game; no one is an extra or supporting character per se. The roles the PCs enact in the group hierarchy are developed during the game session in constant negotiation. All players cooperate because they have a common goal; they agree to work in a balance that allows everyone their own wish fulfillment. Rules and the game master's supervision ensure retaining this balance.
[3.5] What we describe here is, of course, the ideal RPG group rather than a representative case. There are several player types who tend to spoil their coplayers' fun. One of those types, a certain incarnation of the "power gamer," is looking only for her own wish fulfillment without realizing that there is a balance to maintain. With her domineering presence, she overshadows the rest of the group and tries to focus the game master's attention on herself and her character. Just like the Mary Sue in fan fiction, this power gamer warps the RPG universe, including the other players' characters, around her own gaming goals. She tries to dictate everyone's behavior in order to shape the other characters' actions according to her wishes.
[3.6] This is what makes the pen-and-paper Sue problematic—she infringes on the other players' autonomy. In an RPG, the player's control over her own character is essential to generating identification. Whether the character is a self-insertion or simply represents the author-player's choices, she is an extension of the player's persona. Taking control of this persona interferes with the game experience and may even feel like an assault on its creator.
4.
[4.1] In an interesting shift of perspective, many original authors report a similar feeling of violation at the thought of fan fiction writers "playing" with their characters. In a way, they are pen-and-paper role-players outside of a gaming context, and they expect readers to honor the RPG contract of not trying to control others' characters. They cannot, or do not want to, sever the bond between character and player/writer that is forged during the simulation stage of story creation.
[4.2] We argue that this individualized bond is at the heart of the RPG experience, and that Mary Sue is just a more extreme, more personalized form of the player's persona. This is also why any Mary Sue is more complex than self-insertions or original characters. Because Mary Sue is someone else's highly customized, unmistakably personalized wish-fulfillment fantasy, she is a static, proprietary character.
[4.3] By contrast, in fandom, it is the consensus that everyone is welcome to add her own interpretation to the mix: the source text's characters are considered open and available material. Different versions of any given canon character coexist peacefully, while there is only one exclusive version of each Mary Sue.
[4.4] But it is not only the characters created for role-playing that do not invite transformation. The very form of the game log prevents the reader/listener from actively engaging with the text.
[4.5] Sometimes, when you encounter a fellow pen-and-paper gamer from outside your own RP group, she will start to tell stories about her character's brilliance. This is normal and understandable: the average role-player is immensely proud of her character (note 1). However, this player may talk and talk without ever noticing that she is beginning to bore you. She is recounting her personal gaming experience, reliving the adventure while sharing the story.
[4.6] But this story is not translated into a narrative paradigm and is lacking characteristics like structure, closure, or, most importantly, a place for the listeners to situate themselves. You, the listener, cannot relate to this text: neither a collaborative simulation nor an accessible narrative, it is closed to you. Like Mary Sue from a fan fiction perspective, the story of an RPG character is a source that does not invite a creative response.
5.
[5.1] So far, we have only talked about the negative effects of the Mary Sue character within role-playing and fan fiction communities. But the true home of Mary Sue is an area that, by sheer number of players, is bigger than fan fiction fandom and pen-and-paper RPGs combined: the computer or console single-player RPG.
[5.2] Indeed, what sets Western-style RPGs apart from Japanese RPGs (JRPGs) is the former's focus on creating a narrative vacuum for the player's self-insertion. As one enthusiast puts it, "I wasn't interested in playing somebody else's character. The bare minimum requirement I have for playing an RPG is that I get to choose what character I play—if not, to me, I may as well read a novel because that is the medium I'm comfortable having the characters chosen for me" (note 2).
[5.3] This is why amnesiac (Star Wars: Knights of the Old Republic), reincarnated (Planescape: Torment), or deliberately nondescript characters who quickly rise to fame (Might & Magic) are popular tropes for RPG protagonists. With limited connections to a narrative context, there is no need to actively warp the story world or original cast to make the protagonist shine: the game is designed with the single purpose of creating the perfect environment for Mary Sue. What is seen as a bug in fan fiction is a feature in computer RPGs.
[5.4] This is harder to accomplish than it sounds. From assigning physical traits to choosing a background story, many customizable features (like face generation) give the illusion of choice, but they often have little impact on actual game play. Without prescripted features and events, there are no set narrative goals, which defeats the purpose of the story-driven RPG. Yet each ready-made story bit limits the player's freedom to pursue—and create—her own personal narrative, like romancing whomever she wishes regardless of gender or (alien) anatomy (as in the aforementioned case of Mass Effect).
[5.5] This balance between prescribing too much on one hand, and offering too little narrative content on the other hand, is being continually negotiated by producers and critically commented on by players:
[5.6] [F]or me, that fact that I am participating—the […] protagonists [sic] is significantly made by me, is a very important part of my enjoyment, even if it is somewhat illusionary. This is the unique opportunity that video games offer. They aren't MEANT to be an empty cypher, they are meant to be created significantly by the player in more than just stats and race selection. (note 3)
6. Conclusion
[6.1] In this wild ride from fan fiction through pen and paper to single-player computer RPGs, we used the conventional concept of the Mary Sue as a focal point for highlighting the playful nature of story creation. Our aim was to support a more dynamic understanding of narrativity and its interdependence with simulation. At the same time, we wanted to show that, contrary to popular opinion, there is nothing whatsoever wrong with Mary Sue in and of herself. The problem is one of context.
[6.2] The Sue writer is simulating a single-player campaign, while both fandom and pen-and-paper RPGs are collaborative ventures, although they work very differently. In fandom, the Sue story is, essentially, a closed text, a relatively unfiltered log of wish fulfillment that has most of the trappings of typical fan fiction but is an uneasy fit in a community that is based on transforming and sharing texts. Within role-playing groups, the Mary Sue, or power gamer, disrupts the balance of wish fulfillment agreed on among players by trying to make the other characters' actions revolve around her. It is only in single-player computer/console RPGs that this version of the Mary Sue is truly at home, even if creating a narrative vacuum for the player to situate herself in creates its own set of challenges.
[6.3] We wholeheartedly enjoy the gratification through simulated wish fulfillment and personalized agency that is at the heart of the Mary Sue experience—by engaging us in this way, game franchises have won us as fans. This is why, unlike many fan fiction fans, as gamers, we have an overwhelmingly positive perception of a character type that is as ubiquitous as it is derided in fandom. Differentiating between "game logs" and "fan fiction" has helped us understand and appreciate the dynamic, cooperative creative processes that are at work under the surface of fandom, beyond the textual traces that they leave behind.