Nicholas Taylor

Dr. Nicholas Taylor is Assistant Professor of Digital Media in the Department of Communication at North Carolina State University. His work applies critical, feminist and socio-technical perspectives to experimental and mixed-methods research with digital gaming communities. He is co-founder of Circuit Studio, a collaborative maker space at NC State.

Contact Information:
nttaylor at
North Carolina State University,
2200 Hillsborough St.,
Raleigh, NC, 27695

Chris Kampe

Chris Kampe is a PhD student of the Communication, Rhetoric and Digital Media Program at North Carolina State University. His research falls within Games Studies, with attention to: the use of games as pedagogical tools or as a medium to discuss serious issues; the role of interface in participant experience of digital spaces; lastly, how games are changing the ways we tell and engage with stories.

Contact information:
cwkampe at
North Carolina State University,
2200 Hillsborough St.,
Raleigh, NC, 27695

Kristina Bell

Kristina Bell is an Instructor of Communication in the Nido R. Qubein School of Communication and a Ph.D student in Communication, Rhetoric, and Digital Media at North Carolina State University. Her work uses critical and interpretive methods to explore issues of gender and race in digital media and pop culture.

Contact information:
kwbell at
North Carolina State University,
2200 Hillsborough St.,
Raleigh, NC, 27695

Me and Lee: Identification and the Play of Attraction in The Walking Dead

by Nicholas Taylor, Chris Kampe, Kristina Bell


The Walking Dead video game offers a compelling site in which to explore whether, how, and through what processes players form associations with the game’s playable character, Lee Everett. Eschewing perspectives that view the player/game avatar relationship as an exchange between two fixed and stable entities, we draw from microethnographic explorations of participants’ playing and commenting on The Walking Dead (Telltale, 2012), to illustrate the ways in which their relationship to Lee shifts from moment to moment. We offer a provisional schema which accounts for the variety of “attractors” informing these dynamic engagements: simulated, relating to the gameworld; lived, relating to participants’ real world experiences and orientations; conventional, relating to participants’ prior engagements with other related media; and situated, relating to the localized and embodied settings of play. The utility of this schema is illustrated through in-depth analysis of one specific instance in participants’ play.

Keywords: microethnography; adventure games; avatars; identification; zombies; The Walking Dead


This paper reports on microethnographic research of players’ game-based interactions and decisions, involving Telltale’s The Walking Dead (Telltale, 2012; henceforth, TWD) and our subsequent discussions with them. Gameplay in TWD is driven by the increasingly vexing moral (and mortal) dilemmas of Lee Everett, survivor of a zombie epidemic, as he travels through Georgia with Clementine, the young girl he is tasked with caring for, and a small group of survivors. Borrowing extensively from both graphic novel and cinema conventions in its narrative structure and representations, gameplay is based around the point-and click style of adventure games, with an emphasis on verbal interactions and moments of crisis in which the player must make consequential and often split-second decisions, sometimes dictating the fate of other characters. Players must navigate through a series of ethically ambiguous scenarios, often with unpredictable consequences, involving characters who are far from black and white and whose own moral leanings unfold over the course of the player’s interactions with them.

Though embedded in a post-apocalyptic world, these scenarios arguably represent a high degree of “social realism” (Galloway, 2006) in terms of the interpersonal relations between survivors. As such, it is a compelling site to explore issues of whether, at which points, and through what processes “identification” - some form of projective affiliation between player and digitally-mediated character - happens. While conducting observations and interviews with TWD players, we noticed that, despite playing the same game, their described experiences had fundamentally different focuses: managing group dynamics and resolving interpersonal conflicts, doing everything to ensure the survival of Lee and Clementine, or simply killing zombies.

In her ethnography of a school-based gaming club, Carr (2005) notes that the multimodal characteristics of games mean that “a variety of simultaneous ‘reading positions’ are open (or closed) to different players” (p. 466). She argues for more work that theorizes how this multiplicity “might mesh with different aspects of player subjectivity (in particular contexts)” (p. 466). Following Carr, we suggest TWD offers a compelling site in which to examine what players identify with (player/non-player characters, situations, etc) and how they perform this identification. Rather than attempt to categorize and theorize particular types of player/avatar affiliation (see, for example, Yee, 2006; Banks and Bowman, 2013), we view identification as a continual, dynamic exchange between the ‘modes of address’ the game offers and the aspects of player subjectivity activated through play (e.g. prior experiences with games, emotional state during play, relationship with co-participants and researchers, etc). Seen in this way, player/avatar affiliation is a fluid association, not a fixed state.

Beginning from this central insight, we offer a micro-analysis of TWD players' experiences that documents these shifting affiliations within a single given moment of play. Our account develops a provisional theoretical toolset for understanding the choreography of agencies that inform whether and how players identify with particular game elements.


To address these questions, we first outline how questions of identification have been taken up in studies of digital play. We then describe our exploratory case study, emphasizing the ways it both draws off of and departs from “microethnographic” approaches to the study of human performance and interaction (Atkinson, 2013; Streek and Meerhaus, 2005), and digital games in particular (de Castell, Boschman and Jenson, 2008; Giddings and Kennedy, 2008); we also offer a rationale as to why we chose a particular pair of participants to focus on in the context of this paper, whom we name (for purposes of anonymity) “WD5” and “WD6”. From there, we turn to a discussion of the research session itself, in which we showed participants key moments of their play through the first two episodes of the TWD and solicited their reflection.

Our analysis extends the application of post-humanist perspectives of digital play offered by Giddings and Kennedy (2008), while attempting to articulate processes of player/avatar affiliation according to the factors—more accurately, the Deleuzian “attractors” (DeLanda, 2011)—that influence these processes. Attractors represent particular tendencies sedimented over time rather than concrete things (DeLanda, 2011); beginning from the recognition that affiliation is a dynamic performance rather than a singular state of identification with an avatar (Giddings and Kennedy, 2008; Linderoth, 2005), our framework allows us to better understand how this performance is influenced at different times by different forces. This perspective enables us to chart the subtle shifts in the characters, processes and scenarios that players identify with at any given time, and to theorize the source of these shifts, rather than assuming that identification represents a stable relationship between a “coherent, bounded subject” (player) and a “coherent, bounded object” (avatar) (Giddings, 2007).

We offer a demarcation of the attractors that influence the interaction between human and non-human agents/logics in TWD. These are simulated: of the specific game narrative and mechanic; conventional: extrapolated from game and narrative (in general); lived: extrapolated from physical experiences and applied worldviews; and situated: the player’s embodied position, senses, and responses. We conclude by reflecting on the methodological affordances carried out in this approach to studying participants’ experiences with TWD, and we consider the theoretical contributions this categorization, however experimental, might offer to subsequent studies of players’ interactions with other games and gaming experiences.

Literature Review: Identification in Play

We begin by sketching current research on the relationship between players and in-game avatars. Our own study is aligned with qualitatively-driven analyses that characterize the player/avatar relationship in terms of flux and indeterminacy—arising from the “push-pull” of agency across an assemblage of human and non-human actors (Giddings and Kennedy, 2008)—rather than as a fixed state. The ways in which players come to identify with in-game avatars has been addressed from a range of disciplines including film and theatre studies, education, social psychology, and ethnography, each varying widely from the other in terms of how it sees identification play out through gaming. We examine each of these in turn.

Avatar as Reflection

Driven by cognitive scientific approaches to film, Rehak (2003), Filiciak (2003) and Lankoski (2011) offer an understanding of the avatar as a “reflection” of the self, viewed voyeuristically. This work is informed by theories that position film protagonists as the vessels through which audiences engage with of cinematic texts. As with Turkle’s (1995) early and influential explorations of virtuality and identity, game avatars are seen as externalized representations of our anxieties and desires.

Taking a more empirical approach, quantitative studies in social psychology and mass communication inquire into the effects of avatar play on subjects’ sense of self. These studies tend to focus on avatar appearance and behavior in massively multiplayer online game (MMOGS) such as World of Warcraft (WoW; Blizzard, 2004), which allow for fairly extensive avatar customization and alteration. Scholars working within this perspective have coined the “Proteus Effect” (Fox, Bailenson and Tricase, 2013) to describe the changes to self-perception that can arise from playing as an avatar, from seeing oneself embodied in a virtual form. This perspective informs much of the more positivist research on “gender-swapping” (Huh and Williams, 2010) and the assignation of gendered stereotypes in games (Yee et. al, 2011).

Avatar as Vessel

Gee’s (2007) work on the cultivation of “hybrid identities” in role-playing games is representative of an approach that views avatars as means for extending player intentionality into virtual spaces. In this view, avatars are agents whose specialized abilities offer players the chance to act out different roles, orientations and outlooks. In a similar vein, both Klevjer (2012) Newman (2012) describe avatars in prosthetic terms: Klevjer sees them as “proxies” which “transport” player perception and agency into digital environments, while Newman sees them as “instruments” that amplify and transform player input. Similar notions appear in studies examining how players use avatars to explore subjectivities and personalities different from their everyday identity. Nakamura’s (1995) articulation of “identity tourism” in role-playing games, for instance, offers a critical look at the ways players temporarily inhabit virtual bodies (which often reflect gender and ethnic stereotypes).

Altered Affiliations

While each of these studies offers a productive theorization of a specific way in which a player might relate to her avatar(s) under certain circumstances, the understanding we seek here is driven by our observations of players enacting and describing multiple affiliative states over the course of a single TWD session. Our theoretical framework therefore follows the lead of Linderoth (2005), Giddings (2007) and Giddings and Kennedy (2008), who provide micro-analytic accounts of players’ shifting affiliations with avatars (and other in-game elements) over the course of a single gaming session. Linderoth (2005), for instance, analyzes children’s communication as they play Super Smash Bros. Melee (HAL Laboratory, 2001) and notes how their perception of agency with regards to their playable characters changes from one moment to the next, depending on the game’s operations and on their interactions with other players. Similarly, in his study of boys playing Lego Racers (High Voltage Studios, 1999), Giddings (2007) explores how the object of players’ identification shifts from one moment to the next; at some points the boys express an identification with their on-screen Lego avatars, but at other points they identify with the cars, and even “whole gameworld” (p. 45). The game compels these shifting identificatory states, providing a number of different objects, processes and entities to identify with at any given point. For Giddings, then, the notion of identification as “ideological investment by a subject positioned in relation to a fictional protagonist” (p. 46), or even as a stable relationship between singular “bounded subject” (player) and “bounded object” (avatar) is incommensurate with his observations of the fluidity and complexity of agency and affiliation mobilized through digital play (p. 46).

Giddings and Kennedy (2008) develop this further, portraying identification with an avatar as a contingent and momentary outcome of associations between actors in the heterogeneous network of digital gaming — actors that include input devices, avatar appearance and abilities, in-game opponents and barriers, players’ embodied and cognitive competencies, gaming platforms and monitors, and so on. The authors argue that “in the heat of a battle or the tangles of a puzzle the player may be less concerned with the appearance or intertextual connections of his or her avatar, but at other moments these factors may be primary, and at times the two will be inseparable” (p. 25). When and how players identify with avatars largely depends on what the game is doing to them at any given moment. Our study extends this line of exploration by not only acknowledging, but accounting for the shifting, messy and contingent sets of relationships that inform how we interact with and interpret in-game characters. In doing so, we seek to better understand how players shift from one affiliative mode to the next, and how this shapes their decision-making processes—particularly in morally fraught and emotionally challenging games such as TWD. First, we describe our study and the microethnographic methods employed.

Description of Study

Our case study was conducted over ten weeks in the summer of 2013. Participants (8 in total) included 4 males and 4 females, ranging from age 20 to 35. After expressing interest in the study, potential participants were invited to come in pairs to a research studio on our university campus for three separate sessions over a 2 to 4 week period. We encouraged participants to pair up with friends or loved ones so as to make them feel more comfortable. In the first session, participants were invited to undergo an intake interview in which they were asked about their basic demographic information, their prior experience with gaming, and the extent of their familiarity with the TWD franchise (whether the graphic novel, television series, or games). This enabled us to gain understandings of participants’ prior and ongoing involvement with games and gaming culture as well as zombie narratives, and to share any pre-conceived notions/perspectives they had on gaming, horror and survival genres, and TWD. Following this, participants were invited to play through Episode 1 of the five-episode game. Participants’ on-screen activities and verbal reactions to play were recorded via screen-casting software. Participants returned to the studio after one to two weeks to play Episode 2, involving the same recording process. For each session, at least one and usually two researchers were present, facilitating the session and generating fieldnotes.

After the two gameplay sessions, we asked the participants to send a list of moments that were particularly important to them during their gameplay. These, in combination with researcher fieldnotes, formed the basis for editing participants’ gameplay footage into a shorter highlight reel that we played for them in the third session. We invited them to discuss their reactions to these moments, their rationale for making particular in-game decisions, and their theories regarding the motives and actions of non-player characters. As we argue below, this format engaged participants in a process of microethnographically analyzing their own play.


With its small scale, exploratory aims, and combination of semi-scripted interviews, observations, and discussion with participants, the goal of this research is to generate descriptively-driven insights into the ‘play’ of affect generated by TWD, rather than to produce generalizable claims about players. As such, this work is largely situated within the interpretive tradition (Angen, 2000), in that the understandings we develop result from observations, reflection, and dialogue with our participants. The particular analysis we offer here is built on a combination of semi-scripted interview and microethnographic techniques. We did not engage participants in conversation during play — primarily out of a concern that doing so would distract them from the fraught and often time-delineated decisions the game demands — instead eliciting reflections from participants by showing them highlights of their recorded gameplay. The data we draw from was recorded during the third (and final) session for two study participants, WD5 and WD6, as they shared their insights and perspectives on clips from their play through the first two episodes.


According to Streeck and Mehus (2005), microethnography has its roots in Goffman’s dramaturgical “micro-studies” of social order, as well as in Garfinkel’s ethnomethodological approach to understanding the ways people reproduce larger social relations in their interactions (p. 385). Microethnographers build off of these sociological traditions, working with recordings of short segments of data - paradigmatically, audio-visual recordings (Erickson, 1982; Streek and Mehus, 2005) - to document how realities and relations are enacted through localized interactions. The approach involves “close reading” of both speech and gesture, and attends to “multimodal analyses of spoken and other actions, incorporating visual and linguistic data” (Atkinson, 2013: 358).

Microethnography has more recently been adopted by digital games researchers as a means of mapping the “choreography” of human and technological actors in a given gameplay context (Taylor, 2011). Giddings defines it as an “opportunistic approach to recording, describing, and analyzing brief moments of everyday technocultural activity” (2008: 149). Video-based microethnographic analyses have been employed to explore nonverbal interactions between Second Life avatars (Antonijevic, 2008); kinaesthetic play in school-based gaming clubs (de Castell, Boschman and Jenson, 2008); the performance of masculine subjectivities at e-sports tournaments (Taylor, 2011); and, the circuits of pleasure and agency produced through co-operative gameplay (Giddings and Kennedy, 2008). Each of these studies generates novel insights into the human and machinic apparatuses of digital play in different contexts, and each relies on microanalyses of particular gameplay moments in order to generate these insights. Where these studies adopt a more conventional microethnographic approach to digital gameplay, the present study engages in an analysis of participants’ reflections on video-recorded excerpts of their play. On one hand, this marks a departure from contemporary microethnography, as defined by Streek and Mehus, the focus of which is “the analysis of recorded specimens of interaction, usually without consulting participants’ judgments” (2005: 381); clearly, by inviting participants to watch and discuss their own play in TWD, we are interested in their judgments and interpretations of their gameplay, rather than direct observations of their in-game activities. The process we describe here involved participants analyzing their own minute decisions in, and reactions to, game-based scenarios. Therefore, we consider the participants’ input itself as microethnographic, as we engaged them in close readings of their play, and specifically of those moments in the video recordings of their TWD play that both researchers and participants deemed most unsettling, difficult, emotionally-charged, and/or significant to either their own understanding of the game or to the unfolding storyline. It is the outcomes of this engagement are what we report on here.


As with other microethnographies of digital play, as well as studies that (like ours) seek to qualitatively understand the processes shaping avatar/player relations (see, for instance, Shaw, 2013), we focus on a small number of participants (two) in order to conduct a close reading of their gameplay experiences. Both participants are female graduate students in their mid-20s. WD5 was raised in the American Midwest, describing her background as “German” and “American”. WD6 was born in northern India and recently moved from New Delhi. While neither woman described herself as a “gamer”, both claim to enjoy certain gaming experiences: WD5 professed enjoyment of console-based role-playing and adventure games such as Okami (Clover, 2006), while WD6 said she plays mobile games. Neither expressed much familiarity with the point-and-click style of adventure games on which much of the action in TWD is based, although WD6 claimed she used to play many horror games as a child, stopping when she became more serious about school. Each had some familiarity TWD’s media franchise—they had both read the first issue of the graphic novel, and WD6 had started watching the TV show before losing interest and “googl[ing] the rest of the story” to find out what happened. In terms of their interest in zombie-based media more generally, the women diverge. WD5 said in her intake interview that she does not enjoy zombie movies or TV shows, and exclaimed that “zombies are overdone”; WD6, conversely, claimed to have seen “almost all” zombie movies, though she is now starting to tire of them.


We have opted to look closely at the third, highlight reel session involving these two participants because of the compelling mixture of similarities and differences we noticed as they recounted their rationales and reactions to moments in their TWD play. On one hand, they made many of the same decisions at key moments in the game, and thus from the perspective of the software (and the data on player decisions it automatically logs), the play experiences of these two women were quite alike. On the other hand, the rationales they offered for these decisions, and the dispositions towards the playable character, the game, and indeed, the study itself reflected in these rationales, were striking. The experiences of these two participants therefore offer a rich terrain on which to explore the different kinds of relationships players are enacting to digital games even when playing the same game, in seemingly similar ways. Based on this exploration, we offer two insights regarding player/avatar affiliations in TWD: the first is that such affiliations are fluid, shifting from one moment, or one account of play experiences to the next; the second is that in discussing their gameplay, participants instantiated very particular (though in some cases, common) associations between the game and other facets of their lives.

Lee as Me?

In all of our participants’ microethnographic readings of their own play, we noticed a fluid movement between different affiliative states: at some points they spoke as if they were Lee, projecting themselves into a given scenario or even explaining what Lee was thinking at the moment. In other instances they described Lee and themselves as wholly distinctive entities. To demonstrate these shifts, we turn to a particular incident in Episode 2:

Lee and his party are all locked in room when Larry, the antagonistic father of another party member (Lilly), appears to suffer a heart attack and falls to the ground. If Larry dies, he will rise as a zombie — putting the group in danger. In the moments that follow, Lee is asked to make a decision: Lilly wants Lee to give him CPR; Kenny (another survivor, and head of a small family), wants Lee to help smash Larry’s head before he can rise as a zombie.
Though the player’s decision will affect events slightly, all variations lead to the same outcome: Larry’s head will be smashed with a salt-brick — though this fact is hidden from players at the outset. Although the general trajectory of the narrative will remain unchanged, the player’s decision affects how other characters subsequently react to Lee.

WD6 decided to save Larry; whereas, WD5 decided to kill him. What follows is the justification WD6 provided to account for her decision. Admittedly, this account came after the occurrence itself, so it does not speak to what the participant felt in the moment — only her understanding of that moment after being asked to reflect upon it. Her description of the occurrence suggests that she did not operate within a single mode of identification; rather, as shown through her account, she moves freely between multiple perspectives.

WD6: Clementine was there, and I'm not going to kill anybody if Clem—I mean, I had this set principle, very clichéd—that there's a kid and there is like a way to save a life or kill a life I'm going to take save a life at all costs… I knew that it was taking Lilly's side or taking Kenny's side, but… if he survives Kenny will not have a grudge because he survived, but if he dies then Lilly would have—I mean—that grudge that you did not try to save him. So that was the best way to try to keep the group cohesive at all points… for me it was like Lilly was still upset because Larry gets killed by Kenny, but it left Lee in a very good position in a way because Lilly came and saved me later and Kenny was fine.
In the first sentence, “I’m not going to kill anybody [in front of Clementine]”, she speaks as Lee within the context of the game’s reality; it is impossible to grammatically separate the perspective of the player from the inferred perspective of the character. Embedded in that same sentence there is a momentary departure in which the player frames her identity differently: “I mean, I had this set principle, very clichéd”. The player speaks from outside of the game world, critically appraising the value system she chooses to employ while making decisions in the game. Within one syntactic unit, WD6 expresses her perspective as though it was held by Lee, but then references her own values as though she were consciously applying them to Lee, enacting a similar affiliative fluidity Linderoth (2005) notes in his account of children’s Super Smash Bros. Melee play. Her final description further emphasizes this fluidity: “for me it was like Lilly was still upset because Larry gets killed by Kenny, but it left Lee in a very good position in a way because Lilly came and saved me later and Kenny was fine”. In the first movement the player again refers to this more removed perspective as an observer of what transpired. Following the description of events, she refers to the decision as leaving Lee in a good position to manage group dynamics — here Lee is referred to as a separate entity. However, following that statement, WD6 describes an outcome—“Lilly came and saved me”—wherein she (player) and Lee (avatar) are conflated. She slips from what Banks and Bowman (2013) call “avatar-as-Me” to “avatar-as-other” — seen here not as expressions of discrete player “types”, but as different modes available to the player.

WD6 does not seem to confine her perspective to that of the avatar, but there are moments in which she takes up his perspective intimately, just as there are moments where she applies her will upon the situation, without suggesting that this belief is embedded in his character. In this close reading, avatar identification appears as an ongoing process that is constantly being shaped by forces internal and external to the game itself, ranging from perceived norms of behavior, generic understandings of media, and even the player’s own bodily awareness. This recognition requires an epistemological shift, from trying to understand the why certain players identify with certain avatars, towards trying to understand the common “attractors” (DeLanda, 2011) that influence and shape a given expression of this process.

Forces of Attraction

We begin with the premise that when playing a videogame, identification is a fluid, ongoing process; the player does not occupy a stable subject position, nor is that subject position guided by a single logic. This leads us to ask, what factors shape the process of player identification in games—or to re-invoke Carr’s (2005) question, to what can we attribute such shifts in “reading position”, not just between players, but within one small moment of play? DeLanda (2011) argues that patterns/processes should be defined by the factors that affect their behavior, rather than by external characteristics they manifest. By focusing on characteristics, we idealize the process and blind ourselves to instances that deviate tremendously from this norm. With this in mind, we will expand upon player identification, by articulating the recurrent forces that influence it; to do so, we will employ Delezue and Guattari’s concept of attractors (1987). An attractor does not represent a single, knowable thing; rather, it represents the “inherent or intrinsic long-term tendencies of a system: the states which the system will spontaneously tend to adopt in the long run, as long as it is not constrained by different forces” (DeLanda, 2011: 14). In their most abstract sense, attractors are the indefinite origin of a force that explains the tendencies in observed processes to move toward predictable states; they can be more accurately understood as fields of attraction. Attractors do not represent singular objects, nor can the processes they influence ever ultimately collide with them. However, we can infer their existence because they have measurable effects on a process that is being analyzed. This force appears in the tendency to empower or constrain a way of interpreting a situation with relation to all other available ways.

From the highly specific context of this study, in which we both observed players and invited them to account for their actions during play, we infer the presence of four recurrent, attractive forces that affected their decisions and rationalizations. What follows represents a provisional articulation of the interplay between TWD and the range of attractors activated through participants’ play. The attractors that we articulate are large and vague: they can be shown to affect much of the play process, but a deeper examination, or an examination in a different context, would very likely lead to the identification of additional smaller (more nuanced) attractors, which are not articulated in this study.


The simulated attractor derives from textual, visual, audible, mechanical, and procedural cues given by the specific game being played. This force becomes apparent when players act in response to or justify actions with reference to diegetic elements of the game world (e.g. the argument made by a character for a specific course of action, or visual cues taken to indicate weakness in a monster).

In order to illustrate this force, consider this evaluation offered by WD5: “Lily kind of comes off as more abrasive [than Carly], however as.. kind of mean as she can be... I really admire her loyalty to her father”. The participant responds to the diegetic rendering of a character, but that does not appear to be the only force guiding her response. She expands to say, “you can't choose your family, and not everyone in your family is going to be the most likable person all the time but like, she loves her dad” (WD5). The participant personalizes the experience, suggesting a second force of attraction: that of her own, lived experiences.


The lived attractor derives from understandings of life outside of the diegetic gameworld: understandings of how natures and processes tend to or possibly could behave. This force comes from behaviors that have been observed or are believed to exist in players’ everyday (and/or past) experiences, but the lived attractor seldom (if ever) operates in isolation. Participants invoke lived experience because they perceive it has having bearing on the decisions made in the simulated moment.

Turning to a reflection from WD6, when explaining why she chose not to kill Larry: “I can take the ‘morally right’ choice; that can be justified [to the group] later... because I was trying to save a life. I mean people won't be pissed off at you because you were trying to save a life”. Here, the player couches her decision (to try to save Larry) in an estimation of peoples’ responses to it — effectively asserting that she knows how these simulated characters will respond because she knows how real people would respond. This is noteworthy, in part, because it is inaccurate: in the game, some characters will never forgive this decision, just as other characters would not forgive the opposite decision. In this instance, this stronger force exerted by the lived attractor causes the player to respond more to her lived experiences than the behaviors simulated by the game.


The conventional attractor derives from past experiences with representational type (i.e. game, narrative, mythos). This force comes from observances of past, (believed to be) fictional signs or behaviors that are perceived as having bearing on the decisions made in the simulated moment. This is exemplified by instances in which meaning or (predicted) behavior is inferred/deduced on the basis patterns/expectations established by previous media forms.

When accounting for her ‘risky’ decision to try to save Larry rather than killing him, WD6 asserts: “There is always this presumption that somebody will save Lee at least and Clementine is probably not gonna die... [because] they had to keep the story going.” In this instance, the player informs her decision with generalized, external understanding of ‘how zombie narratives work’; in a single moment, her choices are influenced by the specific constraints of the game, her understanding of this game genre, and her understanding of this type of story — thus it is a more critical, but (perhaps) less personal affective force.


The situated attractor derives from mental or bodily responses resulting from or concurrent with the physical context of play. The forces come from a subject’s embodied awareness that she is engaging with an interface within a specific situation (e.g. making a decisions quickly because you need to go to the bathroom). This is exemplified by in-game actions which are purposed towards achieving out-game effects. A pronounced example of this force came from exchange between the interviewer and WD5.

WD5: The two episodes when we were playing, like one I was playing on a Saturday afternoon when I was relaxed, the other I was playing after working for about nine hours.
INT: So you probably played differently?
WD5: Definitely — so it was like, as soon as I can finish it.
Here, the player indicates that her bodily state (fatigue) caused her to play Episodes 1 and 2 in a fundamentally different manner: she admits to rushing through events in Episode 2 quickly so that she could get home and rest. The Situated attractor can be inferred when external factors present in a play-space (noise, observation, etc.) or known, internal responses to stimuli (fatigue, fear, etc.) affect play.

Each of these attractors operate on an inherently complex process, such that it is not productive to account for one’s isolated influence. A given player response might demonstrate the effects of a single attractor as more/most pronounced, but, in each instance, there are multiple attractors affecting action/reflection. The interplay between these forces of attraction will vary from case to case: in some instances one force appears to amplify another, in other instances one force counteracts or weakens the other.


The analysis that follows represents a provisional demonstration of this framework, applied to a single instance in WD5 and WD6’s reflections on highlights from their play in TWD. Here, we use the same instance described above, in which players must decide whether to kill Larry, thereby preventing him from rising as a zombie, or to try to save him. In comparing the participants’ rationales for their decision, our aim is to show the utility of this framework to understandings of the fluid, dynamic process of player/avatar identification.

In the middle of our interview, we asked WD5 which of the female figures she identified with the most; she responded: “[Lilly] comes off as more abrasive, however...I really admire her loyalty to her father, and I don't think that ever really came out in any way in the game, but just me personally like you know you can't choose your family... she loves her dad, don't mess with her about it.” The player demonstrates an extreme personal empathy with a non-playable character due to some unspoken, shared experience. This empathy did not, however, stop WD5 from crushing Larry’s head with a salt-brick; an action for which she was immensely remorseful. Despite this remorse, she never indicated a desire to go back and change what happened. In an effort to better understand the role she occupied in this momentary decision, we will articulate the interplay of the attractive forces that influenced her decision.

When we asked her why she killed him, she said, while curling up slightly, “because I didn’t want to kill him as a zombie... I get scared very easily”. This suggests that a situated attractor affected her decisions, but this was not the only force in play. She later elaborated by adding: “my worry was Kenny’s worry... Larry was gonna rise as a zombie and we all realized that Larry was stronger than pretty much all of us and we’re in an enclosed space and we have a kid with us and.. and I don’t like scary... it seemed ruthless of course and horrible but... we may be helpless trapped in this tiny enclosed space with this zombie monster if we don't do something”. This vivid description, coupled with references to the diegetic logic of a character, indicates the force of simulated attractor.

Examining this moment closely, it is clear that her decisions (her performance of Lee in this situation) are not the result of a single coherent desire, rather they are an entanglement of conflicting forces. Simulated factors clearly affect her decision: she has taken a number of cues from the virtual environment (small room, limited options,) the processes that occur in the game (zombies rising, possibility of death,) in addition to cues from the different characters (Larry’s physical strength, Clementine’s need for protection). At the same time, the emphasis that she gives to certain diegetic elements may be affected by her bodily responses (a situational set of attractors). She stresses earlier “I get scared easily” and later “I don’t like scary”, both of which suggest an awareness of a physiological response her body has to certain stimuli. She ascribes many motives to Lee, including concern for the well-being of others, but she does not ascribe fear. The player’s bodily responses explicitly affect her in game decision-making process: leading her to take in game actions (killing Larry) to avoid out game consequences (experiencing fear). However, it is impossible to fully know the extent to which each attractor affects how she engages the scenario, because of interactions between them. In this case, the player’s justification (to kill Larry) falls in line with Kenny’s justification — while his argument may have been persuasive, that persuasiveness was likely enhanced by the player’s emotional state. Rather than each existing abstractly, the force of the situated attractor amplifies the force of the simulated attractor, while weakening the influence exerted by the lived attractor.

The player’s own lived experience also influences her decision-making process, though not in a uniform manner. When rationalizing her decision, she notes: “of course Lilly wants to save her father, but if her father is going to kill her, she's not gonna be able to make that decision to kill him even if she has the opportunity... you have to just sort of take that decision out of her hands”. While this assertion has some basis in the narrative of the game, the player is clearly making predictions about what will happen based on her previous interactions with people (or potential characters of other fictional narratives) — not out of a definite knowledge of the game’s behavior. However, her descriptions of the act as “ruthless” and “horrible” combined with her multiple references to remorse (“I felt really bad about it”) point to an opposing force, though one still coming from prior lived experience. This conflict demonstrates that even within a single field of attraction, there may be multiple (potentially oppositional) sources of attraction.

In those moments where she perceives diegetic scenarios through the eyes of Lee, WD5 occupies the position sketched out by film studies approaches to player/avatar relations, imagined as a “mirroring” between the two (Rehak, 2003). Far from representing any kind of normative playing state, however, this positionality is produced by the absence of particular fields of attraction - namely, the conventional.

WD6 articulates a very different orientation to the same scenario, one shaped less by attention to the internal dynamics of TWD and more by her previous experiences with zombie narratives. In contrast with WD5, WD6 seems guided by moral certitude drawn from lived experience, which is supported by prior experiences with zombie movies and games - what we call conventional attractors. Focusing on her rationale for trying to save Larry, she invokes a “set principle, very clichéd”, which is “[if] there is a way to save a life or kill a life I'm going to take save a life at all costs”. She articulates none of the self-doubt or second-guessing that WD5 exhibits with regards to the same scenario. After WD5 reflects on her own sense of ambiguity with the choice, WD6 counters with, “in all my zombie experience they take a little time to get up”. Here, WD6’s conventional attractors, derived from her familiarity with “how zombies work”, effectively trump WD5’s diegetically-driven justification: trying to save Larry’s life will not lead to catastrophic consequences because corpses are not reanimated immediately, giving the group time to act should Larry start to turn.

She goes on to assert the impossibility, from a narrative standpoint, of any dire consequences to her choice: “for me it was like basically we're not gonna all die if he comes back as a zombie…. obviously there was this presumption in my head that they had to keep the story going — at least Lee and Clementine will be safe.” This “presumption in [her] head” derived from prior experiences with zombie games and movies effectively legitimates her decision to try saving Larry; conventional knowledge eliminates doubt that would undermine the application of lived experience. Consequently, she is able to to adhere to a moral principle to always try to save a life “at all costs”, without concern that such actions would lead to the deaths of all the characters. At several points, however, the game pushes back against WD6’s expectations, shaped by these conventional and lived attractors. For instance, she admits to being surprised when, despite her choice to save Larry, Kenny kills him regardless: “I did not think that while I was trying to resuscitate him that Kenny would come... throw the brick at his face or something”. Here (as well as in other instances throughout TWD) a particular outcome is produced regardless of the player’s choice. Furthermore, WD6’s belief that Kenny will remain loyal to her/Lee regardless of her decision to save Larry (in her words, “even if he survives Kenny will not have a grudge”) proves inaccurate — if the player sides with Lilly and Larry in this instance, Kenny will continue to make life difficult for the player/Lee as the narrative progresses. The implication here is that while WD6 may have partaken of a wider array of attractors shaping her play - specifically, a moral principle derived from real life conviction, supported by an understanding of the ways zombie narratives “should” unfold - the game itself disrupted and undermined this moral (and conventional) certainty.


In conducting this close reading of participants’ accounts of their play in TWD, we offer two contributions to understandings of avatar affiliation in gaming. Both contributions extend the productive work carried out by post-humanist and ethnographic explorations of play, which have unsettled some limiting tendencies in game studies to read player/avatar relations as expressions or manifestations of an exchange between two fixed entities.

Methodologically, inviting participants to watch and comment on video recordings of their play generated a wealth of data about their affective states during play, their rationales for certain decisions and the points at which elements of their lived experience were brought to bear on their in-game decisions. Players’ own accounts of their gaming behaviors are often as, if not more nuanced and complex as our attempts, so engaging them directly in processes of theory-building and analysis seems fitting.

Theoretically, in drawing from the Deleuzian (via DeLanda) notion of “attractors” to micro-analyze the dynamic play of agency over the course of a single (albeit intense) moment in TWD, as recounted by two different players, we have attempted to explicate the deeper currents of affect that shape our experience and interpretation of gameplay in given moment. As we have shown, these currents can temporarily coalesce into something we might read as “identification”, either with Lee or another (non-player) character. Similar insights regarding the fluidity and contingency of players’ sense of agency and affiliation are offered by Linderoth (2005) and Giddings and Kennedy (2008). Expanding upon this work, we provide an account of the forces that instantiate these shifts in affect; we contend that certain constellations of these forces give rise to identification, be it with an avatar, a non-player character, or a situation. The pull of individual forces, emanating from these attractors, leads to the intensifying or disrupting of a particular affiliative state. A player’s capacities for identification (with a gamic element) is shaped by these attractive forces, and the resulting differences in identification will account for radically different experiences, despite the diegetic elements of the game remaining constant. What we offer here, therefore, is not a theoretical framework for better understanding identification, but rather a conceptual apparatus—an “infralanguage” (Latour, 2005)—for mapping the shifting agencies that produce moments or states of identification (among other affective states). The player, her background, the situation within the game, and the context in which it is played will affect not only the strength of each attractor, but on a more “molecular” level it will affect the attractors that operate within each “molar” field of attraction. The task this paper lays out, therefore, is to apply this provisional framework to other game contexts and experiences, and to develop more finely-tuned methodological and theoretical tools for articulating the complex, messy and dynamic interplay of attractors shaping identification.


We would like to offer our sincere thanks to the reviewers for their valuable suggestions. We are also tremendously indebted to all of the participants in this study, for their time, insight, and intestinal fortitude in the face of a zombie apocalypse.


Angen, M.J. (2000). Evaluating interpretive inquiry: Reviewing the validity debate and opening the dialogue. Qualitative Health Research, 10(3), 378-395. doi: 10.1177/104973230001000308

Antonijevic, S. (2008). From text to gesture online: A microethnographic analysis of nonverbal communication in the Second Life virtual environment. Information,Communication and Society, 11(2), 221-238. doi:10.1080/13691180801937290

Atkinson, P. (2013). The mastersingers: language and practice in an operatic masterclass. Ethnography and Education 8(3), 355-370.

Taylor. (2011). Play globally, act locally: The standardization of pro Halo 3 gaming. International Journal of Gender, Science and Technology 3(1).

Banks, J. and Bowman, N. (2013). Close intimate playthings? Understanding player-avatar relationships as a function of attachment, agency, and intimacy. Presented at IR14, Denver, CO.

Blizzard Entertainment. (2004). World of Warcraft. USA: Blizzard Entertainment.

Simulation and Gaming 36(4), 464-482.

Clover Studios. (2006). Okami. Japan: Capcom.

de Castell, S., Boschman, L. and Jenson, J. (2008). In and out of control: Learning games differently. Loading 2(3). Retrieved from:

DeLanda, M. (2011). Intensive Science and Virtual Philosophy. London: Continuum International Publishing Group.

Erickson, F. (1982). Audiovisual records as primary data source. Sociological Methods and Research, 11(2), 213-232.

Filiciak, M. (2003). Hyperidentities: Postmodern identity patterns in massively multiplayer online role-playing games. In M.J.P. Wolf and B. Perron (eds.), The Video Game Theory Reader. New York: Routledge, pp. 87-102.

Fox, J., Bailenson, J.N., and Tricase, L. (2013). The embodiment of sexualized virtual selves: The Proteus effect and experiences of self-objectification via avatars. Computers and Human Behavior, 29(3), 930-938.

Galloway, A.R. (2006). Gaming: Essays on Algorithmic Culture. Minneapolis: University of Minnesota Press.

Gee, J.P. (2007). Good Video Games + Good Learning: Collected Essays on Video Games, Learning, and Literacy. New York: Peter Lang.

Giddings, S. (2007). I'm the one who makes the Lego Racers go: Virtual and actual space in videogame play. In S. Weber and S. Dixon (eds.), Growing Up Online: Young People and Digital Technologies. New York: Palgrave Macmillan, pp. 35-48.

Giddings, S. (2008). Events and collusions: A glossary for the microethnography of video game play. Games and Culture, 4(2), 144-157. doi: 10.1177/1555412008325485

Giddings, S. and Kennedy, H. (2008). Little jesuses and fuck-off robots: On aesthetics, cybernetics, and not being very good at Lego Star Wars. In M. Swalwell and J. Wilson (eds.), The Pleasures of Computer Gaming: Essays on Cultural History, Theory and Aesthetics. Jefferson, NC: McFarland, pp.13—32.

HAL Laboratory. (2001). Super Smash Bros. Melee. Japan: Nintendo.

High Voltage Studios. (1999). Lego Racers. Denmark: Lego Media.

Huh, S. and Williams, D. (2010). Dude looks like a lady: Gender swapping in an online game. In Bainbridge W (ed.), Online Worlds: Convergence of the Real and the Virtual. New York: Springer, pp.161-174.

Klevjer, R. (2012). Enter the avatar: The phenomenology of prosthetic telepresence in computer games. The Philosophy of Computer Games, 7, 17-38. doi: 10.1007/978-94-007-4249-9_3

Lankoski, J. (2011). Player character engagement in computer games. Games and Culture, 6(4), 291-311. doi: 10.1177/1555412010391088.

Latour, B. (2005). Reassembling the social: An introduction to actor-network theory. Oxford: Oxford University Press.

Linderoth, J. (2005). Animated game pieces. Avatars as roles, tools and props. Aesthetics of Play Conference Proceedings. University of Bergen, Norway, Oct. 14-15.

Nakamura, N. (1995). Race in/for cyberspace: Identity tourism and racial passing on the Internet. Works and Days: Essays in the Socio-Historical Dimensions of Literature and the Arts, 25(26), 181-193.

Newman, J. (2012). Videogames. London: Routledge.

Rehak, B. (2003). Playing at being: Psychoanalysis and the avatar. In M.J.P. Wolf and B. Perron(eds.), The Video Game Theory Reader. London: Routledge, pp. 103—128.

Schott, G.R. and Horrell, K.R. (2000). Girl gamers and their relationship with the gaming culture. The International Journal of Research into New Media Technologies, 6(4), 36-53.

Streeck, J. and Mehus, S. (2005). Microethnography: The study of practices. In K.L. Fitch and R.E. Sanders (eds.), Handbook of Language and Social Interaction. Mahwah, NJ: Lawrence Erlbaum, pp. 381-404.

Shaw, A. (2013) Rethinking game studies: A case study approach to video game play and identification. Critical Studies in Media Communication, 30(5), 347-361. doi: 10.1080/15295036.2012.701013

Turkle, S. (1995). Life on the Screen: Identity in the Age of the Internet. New York: Simon and Schuster.

Telltale Games (2012). The Walking Dead. San Rafael, CA: United States.

Yee, N., Ducheneaut, N., Yao, M., and Nelson, L. (2011) Do men heal more when in drag? Conflicting identity cues between user and avatar. In Proceedings of ACM CHI ’11 Conference on Human Factors in Computing Systems, pp. 773-776. doi: 10.1145/1978942.1979054.

Yee, N. (2006). The demographics, motivations and derived experiences of users of massively-multiuser online graphical environments. PRESENCE: Teleoperators and Virtual Environments 15, 309-329.

©2001 - 2015 Game Studies Copyright for articles published in this journal is retained by the journal, except for the right to republish in printed paper publications, which belongs to the authors, but with first publication rights granted to the journal. By virtue of their appearance in this open access journal, articles are free to use, with proper attribution, in educational and other non-commercial settings.