“More Than Just Playing a Character”: Gender Exploration and Expression in Avatar Customization

Abstract

Avatar customization in games is seen as a crucial avenue for players to express and explore gender. While gender options in avatar customization systems are frequently binary, players could still achieve diverse gender expressions. This study investigates how players express and explore gender through avatar customization systems in videogames. Through semistructured interviews with 20 players from different countries who self-reported diverse genders, we identified three major themes: (i) conforming to and/or subverting gender norms; (ii) symbolizing self-identity; and (iii) exploring and experimenting with gender. These themes reveal how players utilize customization options to express gender beyond binary constraints. This research provides critical insights for game designers to support diverse gender expressions. The findings suggest that avatar customization systems are dynamic spaces for ongoing gender identity exploration.

Introduction

Avatar customization has long been the norm in role-playing games (Harper, 2019) where players create their own unique virtual characters through an avatar customization interface (McArthur, 2017; McArthur & Jenson, 2014). Similar to the real world, gender influences how players reshape themselves in the virtual world (Rosier & Pearce, 2011). For transgender and gender-diverse people, research has found that avatar customization systems serve as important spaces for gender exploration and expression (McKenna et al., 2022).

In early text-based role-playing games like MUDs (multiuser dungeons), there was no need to have fixed genders or appearances. Players could express and experiment gender beyond “the plane of certainty” and be whatever they wished (Reid, 1996, p. 327). However, gender options in modern avatar customization systems are often limited to binary checkboxes (male/female) (McArthur, 2018), even though the gender expressions of players are more nuanced and unconfined to such binary choices (Whitehouse et al., 2023). This begs the question: How do players use avatar customization systems to accomplish more nuanced expressions and explorations of gender?

Players “playing” with gender are a well-documented phenomenon in videogames. This phenomenon refers to the negotiation of the discrepancy between their own gender identity and expressions and the gender identity and expressions of their primary avatar in the game (Huh & Williams, 2010). It has been defined by many studies as “gender swapping” (also “gender bending” or “gender switching”) (Lou et al., 2013; Martey et al., 2014; Todd, 2012). These studies interpret gender swapping as a strategic behavior for gaining advantages in the game (Paik & Shi, 2013) instead of an exploration or experimentation of gender identity (Jenson et al., 2015).

“Swapping” and “switching” are limiting. They reinforce a binary view of gender, implying an exchange between male and female and excluding nonbinary identities. For players with fluid or nonbinary genders, these terms fail to accurately capture their gaming experiences and gender expressions. Thus, using such terms risks oversimplifying and misrepresenting the complex gender experiences of players (Whitehouse et al., 2023).

In fact, players don’t always choose a female avatar to express or experiment with femininity, and vice versa. Despite limited avatar gender options in avatar customization, in certain games, such as Animal Crossing: Wild World, players can choose “opposite-gender” hairstyles regardless of their avatar's gender, allowing them to explore different gender traits through hairstyle changes (LGBTQ Video Game Archive, 2020). Consequently, research that limits gender identification to male and female also overlooks how players explore gender through various customization options.

With a growing focus on inclusive and complex gender identities, more scholars are examining the experiences of transgender and gender-diverse players in discussions of avatar gender identity. For instance, a study by Kosciesza (2023) focused on the experiences and strategies of noncisgender players in presenting their gender identities within game environments. These studies fill in missing perspectives by centering the personal experiences of transgender and gender-diverse players.

Building on these perspectives, this study incorporates a new discussion of gender expression and exploration. We consciously recruited 20 players interested in customizing avatars with a variety of gender identities, and invited them to self-report their gender. Drawing on semistructured interviews with these participants, this article applies thematic analysis to explore the potential of customization systems for gender exploration and expression. This research broadens academic discussions on gender construction and avatar customization while offering new insights to game designers on supporting gender inclusivity through avatar customization.

Gender Exploration and Expression

In recent years, research on gender exploration and expression in avatar customization has investigated the potential of avatars and avatar customization systems as sites that help transgender and gender-diverse individuals explore and experiment with their gender identities. One notable investigation by Morgan et al. (2020) conducted focus group discussions with 17 transgender and gender-diverse youth aged 11 to 22, and revealed that avatars play a significant role in facilitating their journey of identity exploration, development, and practice, with avatar customization at the center of this process.

Through in-depth qualitative interviews, McKenna et al. (2022) delved into the effects of avatar customization on transgender and gender-diverse adolescents and found that avatar customization not only offered a low-risk space for gender exploration but also afforded validation and support for participants’ gender identities and transition aspirations. That is to say, avatar customization allows youths to experiment with their desired appearances and offers an immersive escape from the real-world. These findings underscore that avatar customization plays an essential role in self-exploration and affirmation among transgender and gender-diverse youths.

Other studies also paid attention to the gender expression of noncisgender players within gaming contexts. For example, Kosciesza (2023) conducted interviews with 10 transgender and nonbinary players, drawing from their experiences to understand how games influence the process of gender exploration. The findings indicate that while some transgender players are able to use the binary gender framework to express their idealized selves, the space for gender expression provided by games appears significantly limited for those attempting to present identities beyond the binary gender system.

Research on player–avatar relationships overlooked the exploration and experimentation that occur during the avatar creation process by choosing to investigate the avatars that players have already completed. As Banks and Bowman (2016) observed, previous understandings of the relationship between avatars and players have typically relied on Horton and Richard Wohl's (1956) theory of parasocial interaction, which posited that players perceived and responded to avatars, and formed a faux social relationship. This viewpoint considered the connection between players and avatars as one-way and nondialectical (Lewis et al., 2008). However, research has begun to recognize that the relationship between players and avatars can be bidirectional (Banks, 2015; Czerwonka et al., 2021). Considering the significant differences between different games, avatar customization interfaces, and their affordances (Ducheneaut et al., 2009), it is challenging for players to create identical avatars across various customization systems. In other words, avatar creation is not a one-way process but an interactive process with players and the customization interface (McArthur, 2018).

This study uses the concept of “gender exploration and expression” to reflect the continuity in players’ construction process of avatar gender. Gender exploration, in this study, is defined as the process by which players construct the gender of their avatars through experimentation with various adjustable features within the avatar customization interface. This behavior permeates the entire customization process and gameplay, as players do not always have a predefined conception of the gender identity they intend to create upon entering the customization interface (McArthur et al., 2015).

Although gender expression may seem like a product of exploration, it is not a static endpoint—many games allow players to change features like clothing, accessories, and hairstyle as the game progresses. Players can also abandon and recreate avatars if dissatisfied with their gender expression. Therefore, when players perform their gender within the game, they are also engaging in a form of gender exploration.

Intersectionality and Queerness

This study is built upon two key perspectives: intersectionality and queerness. They offer nuanced positions for examining the complex ways in which gender, race, sexuality, and other social identities intersect and shape gaming spaces.

Intersectionality offers a comprehensive framework that moves beyond viewing gender as a standalone attribute. The concept of intersectionality was first introduced by Kimberlé Crenshaw (1989) and is closely related to Patricia Hill Collins’ (1998) work on Black feminist thought. As Chess (2017) describes, “Intersectionality is the theoretical premise that when trying to understand systems of oppression (gender, ethnicity, sexuality, class, etc.), the different factors should be seen interdependently rather than independently” (p. 19).

Kishonna Gray (2012) examined the experiences of African American female players and Latina female players on Xbox Live, revealing how social and cultural factors such as sexism, racism, and homophobia intersect to form layers of oppression that profoundly shape their experiences in online gaming. In her book Intersectional Tech: Black Users in Digital Gaming, Gray (2020) further explored how Black players are constructed as “illegitimate” participants by mainstream gaming culture. Her work highlighted how Black players challenged and subverted oppression through digital practices, employing technologies, and strategies to redefine their presence and to create spaces of resistance and solidarity within otherwise hostile environments.

In avatar customization, intersections of gender, race, class, and other identity factors are pervasive. For instance, developers often present what they consider to be a “standard” body type as the default option for players (Harper, 2019), but this standard is largely based on idealized male or female bodies, and is established from a white or Asian male-dominated power structure (Fron et al., 2007). As Gray (2020, p. 107) pointed out, “Gaming culture replicates gendered practices and normalizes gendered hierarchies, and these structures amplify oppressive practices when gendered identities are racialized.”

This study also draws upon queer game studies. Queer theory rejects rigid identity categories and critically examines the power structures that uphold such classifications (Chan et al., 2019). Scholars such as Adrienne Shaw (2015), Bo Ruberg (2019) and Todd Harper (2013) have demonstrated how normative structures of gender and sexuality are embedded in game culture and design. Applying a queer lens in this study has the potential to deconstruct normative structures and allow nuances to surface.

For decades, videogames have been dominated by a white, heterosexual, cisgender male demographic (Apperley & Gray, 2020). This technomasculine structure, which Fron et al. (2007) termed the “hegemony of play,” has long marginalized everyone else in gaming spaces. In Gaming at the Edge, Adrienne Shaw's (2015) analysis of gender and sexuality in the Fable trilogy illustrates how even seemingly inclusive games can perpetuate oppressive worldviews: “That certain clothing options are read as cross-dressing, that sexuality is inscribed into characters’ very code, and that gender and sexuality are statically defined demonstrate, however, that an oppressive worldview defines the very structure of the game” (Shaw, 2015, p. 32).

This “oppressive worldview” imposes normative structures within avatar customization interfaces. While this study primarily focuses on the relationship between avatar creation and gender identity, we cannot overlook these intersecting factors such as race, class, gender, and sexuality, how they are situated within the “hegemony of play,” and in turn how they impact the player experience. To uncover the nuanced ways in which players express and construct their identities through avatars, our theoretical framework acknowledges the racial, regional, sociocultural, and political backgrounds of players.

Theoretical Framework

Modern avatar customization systems often restrict gender options to a binary framework (Drenten et al., 2019). For example, a study by Czerwonka et al. (2021) examined avatar customization in six exergames, focusing particularly on gender and body type options. The study found that these games offered only binary body type or gender choices, and conflated sex and gender in their option identifiers. In another study, Jenson et al. (2015) argued that in MMORPGs, the gender choices players make in avatar customization were fundamentally based on a biological classification. They concurred, “There really is not much that is ambiguous about selecting ‘male’ or ‘female’ from an avatar selection screen when players are choosing the biological category of sex for their avatars” (Jenson et al., 2015, p. 865).

Controversy over gender terminology persists (Hyde et al., 2019). Some researchers argue that “sex” describes the biological differences between female and male individuals (which is still limiting as it leaves out intersexed people), while “gender” is a sociocultural construct (Torgrimson & Minson, 2005). This “sex/gender distinction” was developed by Robert Stoller in 1968 and has been widely adopted by gender studies (Brickell, 2006). For example, Candace West and Don Zimmerman, in their article “Doing Gender,” explored how the differences between women and men are created by social interactions and roughly follow the distinction of sex and gender (West & Zimmerman, 1987).

Yet, Judith Butler (1990) criticized the “sex/gender distinction.” In the book Gender Trouble, Butler argued gender not as a stable identity, but it was that the immutability of sex which maintained the distinction between sex and gender. The “sex/gender distinction” implied a fundamental discontinuity between sex and gender, and that gender should not be considered to be derived from sex as if it was composed of the cultural meanings of a sexed body. In that way, gender turned into a “free-floating artifice” as the construction of gender became isolated from sex. Then, as Butler suggested, “If the immutable character of sex is contested, perhaps this construct called ‘sex’ is as culturally constructed as gender; indeed, perhaps it was always already gender, with the consequence that the distinction between sex and gender turns out to be no distinction at all” (Butler, 1990, pp. 9–10).

Butler was not trying to deny biological or anatomical differences, but rather argued that interpreting sex independently of culture simplifies and dichotomized it (Morgenroth & Ryan, 2018). They pointed out that the stability of “sex” relied on being set in a preset framework, while the only “intelligible” concept of sex/gender/desire in the culture followed a binary and compulsory order that assumed people's being feminine or masculine being naturally aligned with their body's sexual characteristics, for instance, being attracted to the opposite sex/gender (Butler, 1990, p. 23). This assumption placed the binary “sex” as essential and natural, taking precedence over culture to become the basis of the order.

This research draws its theoretical framework from Butler's theory of gender performativity, positing that gender is not a stable concept but “a stylized repetition of acts” with a politically enforced performativity (Butler, 1990, p. 191). In avatar customization, players experience and express their gender identities not only through the gender option, but through a series of customization options. While gender options in customization systems frequently force players to select a specific gender, in the process of a player exploration and experimentation through various options, the avatar gender is in a state of unfixed fluidity. Therefore, we use the term “gender” in this research to emphasize the performative and constructed nature of identity. By applying Butler's theory of gender performativity, we position the construction of an avatar's gender as an interactive process between the player and the avatar customization system. In other words, gender identity in avatars is not a given but actively constructed and negotiated through avatar customization.

Method

Recruitment

We initially posted a survey via social media (e.g., X and Instagram). The survey gathered demographic information, such as gender identity and pronouns, avatar customization experiences, and the willingness to join semistructured. We received 166 survey responses, with 72 individuals expressing interest in interviews. After screening for experience and ensuring diversity in gender identities, we contacted potential participants by sending them the explanatory statement and a consent form. After coordinating interview schedules, in the end, we interviewed 18 participants. Additionally, at the “Pride at Play” queer games exhibition in Melbourne (June 3–24, 2023), we recruited two more interviewees. ¹

Participants

Between June and September 2023, we conducted semistructured interviews with 20 participants representing a diverse range of gender identities. Of these, 10 self-reported as cisgender (five male and five female), three as nonbinary, two as transgender, one as gender fluid, and one as gender apathetic. Three others provided nuanced self-descriptions: “cis woman with a history of questioning,” “biological woman without a gender identity, queer,” and “between nonbinary and female, closer to female.” We retained these detailed descriptions as they contribute to illustrating the participants’ nuanced understanding of their own gender identities, especially how these understandings are reflected in the avatars they create. Participants resided in different countries and regions: Australia (2), Canada (3), China (8), France (1), Hong Kong (1), Italy (1), and the United States (4), offering insights into diverse sociocultural gender norms. Table 1 summarizes the demographic data.

Table 1.

Demographic Information of Interviewees.

ID	Gender identity	Location	Pronoun
P1	Cis woman with a history of questioning	The United States	She/her
P2	Gender Apathetic	Canada	Any pronouns
P3	Trans guy	France	He/him
P4	Nonbinary	The United States	They/them
P5	Male	Hong Kong	He/him
P6	Between nonbinary and female, closer to female	Australia	She/her
P7	Male	Italy	He/him
P8	Female	Australia	She/her
P9	Nonbinary	The United States	He/him or they/them
P10	Transgender woman	Canada	She/her
P11	Gender fluid	The United States	He/him
P12	Male	China	He/him
P13	Female	China	She/her
P14	Male	China	He/him
P15	Female	China	She/her
P16	Female	China	She/her
P17	Male	China	He/him
P18	Female	China	She/her
P19	Nonbinary	Canada	They/them or he/him
P20	Biological woman without gender identity, Queer	China	She/her

Interviews

Before the interviews, we sent an explanatory statement and a consent form to the participants. Participants filled out and signed the consent form, either on paper or digitally, for us to conduct the study, publish the study, and use the images or other relevant materials they provided. Each interview, conducted via Zoom video chat, was recorded with participant consent for both audio and video. The average interview duration was 74 min.

As a lead-in to the interview, we first asked the participant to customize an avatar in the game demo Our Life: Now & Forever ² while sharing their screen to open conversation. We invited the participant to briefly explain their actions and experiences, effectively helping us to quickly delve into the interview's main topics. Our research, however, focuses on the broader context of gaming rather than analyzing a single game. While the demo features a comprehensive and gender inclusive avatar customization system, it does not sufficiently represent the diversity of videogames and gaming experiences. Therefore, to maintain the broader focus in our research, this part of the interview served to initiate discussion but was excluded from data analysis.

The interviews primarily focused on participants’ experiences in exploring, expressing, and constructing gender through avatar customization. Some also shared screenshots of past avatars and related experiences.

Data Analysis

In this study, we adopted a thematic analysis approach to analyze the data, which has flexibility and capacity for an in-depth analysis of qualitative data (Braun & Clarke, 2006, 2022). We followed the six-phase approach proposed by Braun and Clarke (Byrne, 2022). This approach has been widely adopted by game studies research. For instance, it has been applied to explore the avatar–player relationship (Kartsanis & Murzyn, 2016; Szolin et al., 2023), and to capture the unique experiences of marginalized players (McKenna et al., 2022; Whitehouse et al., 2023).

Specifically, we first transcribed the Zoom meeting recordings into text and thoroughly read these transcriptions to familiarize ourselves with the data, making annotations during this process to record preliminary observations and concepts. Initial coding was conducted using NVivo R1 software, identifying key concepts and categories within the interview data. After reviewing the coded data, we gradually organized the initial codes into potential themes. These potential themes were carefully reviewed and refined to ensure they capture “the most important elements of the data in relation to the research question(s)” (Byrne, 2022, p. 1406).

We identified three themes: Conforming to and/or Subverting Gender Norms, Symbolizing Self-identity, and Exploring and Experimenting with Gender. These themes reflect the strategies players use in gender construction during avatar customization. In the next section, we will illustrate these themes with participant quotes and link them to broader research literature and theoretical frameworks.

Findings

Conforming to and/or Subverting Gender Norms

In videogames, avatar customization becomes a site for gender performance, where players express their identity by both conforming to and subverting gender norms.

Research indicates that there is a tendency for players to appear in ways that conform to socially expected gender roles and idealized gendered bodies, even in virtual environments (Guadagno et al., 2011; Martey & Consalvo, 2011; Triberti et al., 2017). Indeed, our interviews revealed that some players leveraged physical features that aligned with societal and cultural expectations of gender to reinforce their gender identity and expression. P8 (female, she/her, Australia) shared her avatar creation strategy, stating, “if there's options to add more curves to the hips and make it more like a feminine silhouette, then I’ll opt for that, as opposed to something more androgynous or masculine.” She illustrated this approach with examples from her avatars in The Sims, where she consistently ensured that the avatars were in “a dream body” or “like a supermodel.”

Some participants also used specific body features that conformed to gender norms as anchor points to ensure their gender identity could be identified. For example, P12 (male, he/him, China) considered a beard to be a “representation of masculinity” used to distinguish gender. He demonstrated some avatars he had created, which displayed pronounced diversity in several aspects, such as hair color, hairstyle, makeup, accessories, and skin tone (see Figure 1). Nevertheless, a shared feature among these avatars was their beards. As he said, “I don’t want the whole character to be presented as neutral, so that people don’t have a way of deciding whether it's a man or a woman […] that (beard) is one of the things I want to keep as a gender distinction.”

Figure 1.

Screenshots of avatars in Elder Ring (left) and Starfield (right) created by P12.

However, not all players adhered to gender norms tied to their sociocultural background. Several participants described expressing their gender by choosing features that defied traditional expectations. P18 (female, she/her, China) regarded the choice of skin color as a significant form of gender expression. She first explained how skin color was correlated with gender within her sociocultural context: “I personally think that in East Asia, most likely the majority here prefer to be fair, I mean, they prefer a woman to be fair-skinned […] fair skin is considered a very, very [emphasis] outstanding characteristic of a woman.” As a girl with a fair skin tone, she believed that choosing a darker skin tone for her avatar conveyed a more rebellious spirit and showed her personality to some extent: “If a woman's skin color tends more towards the [rebellious] type, which I would now choose, like coffee color, or chocolate color, then people will think her personality is tougher, and she will be more disobedient.”

P18, a cisgender woman, did not equate her identification with her gender to an acceptance of traditional femininity or societal expectations of women. Yet, aspects of her identity that diverged from conventional femininity had long been suppressed by the environment. Therefore, in virtual worlds, P18 consistently selected customization features that contradicted societal expectations and esthetic standards to subvert gender norms, thereby expressing those parts of her identity that did not align with traditional notions of femininity. Her choices illustrated how gender expression has been shaped by multiple intersecting factors—social, cultural, and personal—that influenced her desire to break away from restrictive norms (Gray, 2020). As she stated, “I’m already in the game, I don’t want to be bound by those rules and restrictions anymore.” For players like P18, this form of representation was profoundly significant—a sentiment echoed by P1 (cis woman with a history of questioning, she/her, USA) who stated, “[…] having access to short hairstyles but wanting to still be referred to as a girl in a game was important.” Both participants expressed a desire to break away from traditional gender norms, challenging the male-dominated power structures in game culture (Fron et al., 2007), which upholds binary categories and reinforces gender stereotypes (Ruberg, 2019; Shaw, 2015).

For these cisgender players, engaging in acts of subverting gender norms in virtual environments served as a way to rebel against and escape from the real world, and as a way to reveal more nuanced aspects of themselves. Conversely, for players who refused to be categorized in gender binary terms, creating avatars with “subversive” characteristics sometimes was merely a pursuit of finding an expression that aligned with their real-life gender identities. P2 (gender apathetic, any pronouns, Canada) used their friend's experience as an example: “Like one of my friends, he has a really high pitched voice and he's like, ‘I wish I could just play as a guy with a high pitched voice not having to play as like a girl with a high pitched voice.’ That one is an important one.”

Similar to Kosciesza's (2023) findings, we found that participants who attempted to present identities beyond the binary gender system in games faced more pronounced limitations and challenges than individuals whose gender identity could be represented within the binary. P4 (nonbinary, they/them, USA) shared their experience in matching avatars and their desired outfits: “Typically, in real life, I wear button-up shirts, T-shirts, that kind of thing with jeans, maybe shorts in summertime […] There are some games where the female characters don’t have access to that; but the male characters that can do so are so [emphasis] hyper masculine, they’ve got the big square jaw and the massive shoulders. And I’m like, that doesn’t look like me.”

Other participants seemed to get used to these restrictions. P10 (transgender woman, she/her, Canada) explained: “There's a lot of games where you get two choices, you get your default male, your default female and that's it. And sometimes that can feel a little … you have to fit yourself into one of these two boxes. And that can be a little frustrating sometimes. But I’m so used to games where that's what it is that you just tune it out. So it's not like a huge deal to me I guess when the games don’t have, I am never going to not play a game because it doesn’t have super in depth avatar customization or whatever.” This aligns with Shaw's (2015) observations in Gaming at the Edge regarding players’ attitudes toward representation in games: many players take a “nice when it happens” approach, viewing representation as a “bonus” rather than an actively pursued goal. For marginalized players, there is a desire to avoid being seen as the ones responsible for demanding diversity in games. Instead, they hope for game developers to include natural and realistic forms of diversity, rather than superficial or tokenistic representation.

The experiences of P4 and P10 underscore the specific pressures and realities gender-diverse players face within gaming environments, making it challenging for them to express gender by subverting gender norms. In this context, subverting gender norms was sometimes regarded by them more as an idealized vision rather than a practically achievable mode of gender expression.

In the process of avatar customization, each choice or attribute potentially carries gender expectations rooted in sociocultural contexts. According to Geraci and Geraci (2013, p. 342), these virtual bodies, which have been encoded by social values, offer players a unique opportunity to reclaim or subvert gender norms. They write:

The bodies in games seem to predispose men to believe […] that they must gender swap if they do not want to fall into traditional expectations of what a ‘manly’ man is like. So men might also benefit from additional body types in their games, ones that do not reinforce the belief that men must be either effeminate or hypermasculine.

While their research emphasizes the significance of body option diversity for gender exploration and expression, it overlooks how customization specifically contributes to gender expression. We argue that the unique potential of avatar customization in gender expression is manifested in the dual opportunity it provides: players can both challenge and adhere to gender norms within the same avatar. For instance, P2's avatar emphasizes masculinity through a beard, while incorporating traditionally feminine traits like long, pink hair, and makeup. This blend reflects the complexity of P2's self-description—he sees himself as a “very typical male” yet acknowledges having “some feminine personalities.” This means that players are not only able to select options that subvert gender stereotypes, showcasing unique gender identities, but they can also adopt characteristics that conform to gender norms, thereby making the avatar a site for complex gender expression.

Moreover, for players who do not fit into traditional binary gender frameworks, this complex and nuanced way of expressing gender is especially important. Geraci and Geraci (2013) discussed gender swapping as an opportunity to rethink gender identity in “a space open to non-normative versions of gender,” yet their perspective missed the needs of gender-diverse players. As P10 noted, the complexity of avatar gender they desire to customize in virtual worlds goes far beyond the “default male” and “default female” dichotomy. For gender-diverse players, “gender swapping” might not apply, as they seek to express their gender in richer ways rather than being forced to choose between two inadequate options.

In short, gender expression in avatar customization goes beyond the “male/female” checkbox. Players can create diverse gender expressions that transcend the binary by customizing features that conform to and/or subvert gender norms. This not only enables complex self-expression but also provides a vital opportunity for players challenging traditional binary categories to express their gender identities.

Symbolizing Self-Identity

In the study, we found that certain customization options employed by some participants in the construction and expression of their gender identity do not directly map to external factors such as gender stereotypes or sociocultural values.

Specifically, many participants considered their prominent personal characteristics as integral to their gender expression. When asked about the options through which she expresses her gender, P10 (transgender woman, she/her, Canada) emphasized that height is a significant aspect of her gender expression. In real life, P10 stands at a height of six foot four, which is one of her outstanding characteristics. She told me, “If I walk into a room, there's a 99% chance that I’m the tallest one in that room.” Thus, for her, being able to adjust height allows the avatar to better reflect her true self, “a lot of games it's just kind of, everyone is just the same height because that's just easier from a technical aspect. But being able to, you know, create characters that more closely reflects me in the physical world […] that's something that's always nice to be able to fiddle around with.”

Similarly, some participants elaborated on how they associated specific game choices with their gender identity from personal experience. For instance, P3 (trans guy, he/him, France) considered hair color as a most significant feature of expressing his gender identity (see Figure 2), “If I can, it [the hair color] is always pink, because I’ve spent a lot of time in my life with pink hair […] when I started transitioning, and when I started being more comfortable as a guy I had pink hair. It's very important to me.” Although hair colors, to some extent, are related to social expectations and gender norms, it is apparent that what P3 emphasized is the personal experience embedded within the hair color. To P3, the pink hair, accompanying his gender transition process, not only reflected an aspect of his current gender identity but also encapsulated the entirety of his gender journey.

Figure 2.

Screenshot of avatars P3 created in different games (from left to right: Splatoon, Cyberpunk 2077, and Stardew Valley).

P6 (between nonbinary and female, closer to female, she/her, Australia) also mentioned hair color, but she revealed an intrinsic emotional connection, “I think color [expresses the gender identity] as well. Just in the idea of choice, I think, because it kind of feels like something, not just your aesthetic choices, but it can be mood as well. Which I think is pretty tied to emotions and feeling like who you are as a person.” This suggests that an individual's identification with an avatar is not merely based on external resemblance or embedded personal experience, but extends into the realm of emotional resonance and experiential feelings.

Indeed, whether it's the player's personal experiences, external characteristics, or emotional connections, these fundamentally represent how they “feel like who they are as a person,” and thus can be considered symbols of self-identity. When these symbols are infused to customization options, the features become what Martey and Consalvo (2011, p. 166) refer to as “languages written on the digital body itself,” projecting group identities grounded in aspects such as gender, race, and class. As P6 further added: “I feel like who I am is tied to, not just my identity, but like my gender identity.”

These symbolic features, as core expressions of identity, often remain consistent across different avatars of the same player. This can be observed in the case of P3, who always creates avatars with pink hair color. Similarly, in the interview, P9 (nonbinary, they/them or he/him, USA) discussed how skin color carried their self-expression, showcasing some of their frequently used avatar appearances in Fortnite. ³ “My [avatars’] skins are mainly dark skin.” (see Figure 3). P9 explained, “I suppose having the correct skin tone […] [is] one of the main factors for me […]. I just feel I’m brown like I’m Filipino. I think that is a very important aspect of who I am, and a part of my identity as a person.” This aligns with previous findings that players often retain their core identity elements and physical characteristics across different avatars (Messinger et al., 2019).

Figure 3.

Screenshot of several P9's favorite avatars in Fortnite.

It is noteworthy that when discussing how they express their gender identity through avatars, participants emphasized the importance of “accuracy”—the need for the symbols of self-identity embedded in their avatars to be as close to their true selves as possible. This similarity between players and their avatars, and its impact on players’ gaming experiences, has been extensively discussed in scholarly work. For instance, Hooi and Cho (2012) noted that the resemblance in avatar appearance could enhance players’ self-awareness, which in turn affected their sense of immersion. Similarly, Paleczna et al. (2022) also found that the similarity between players and their avatars positively impacted immersion. Our participants shared similar points. As an example, P4 (nonbinary, they/them, USA) expressed, “My voice does not sound like a lot of cis men, but it also doesn’t sound like a lot of cis women. So usually what I have to do is like, try to play around with the levels of voice that doesn’t really sound like mine […] Being able to hear a voice that has the vocal qualities that mine does, would do a lot for immersion […] in general, most of the things I’ve wanted are better accuracy to life.” However, the participants’ pursuit of accuracy was not merely for enhancing immersion, but rather they desired that the characteristics imbued with self-identity more effectively reflected their own characteristics, experiences, feelings and emotions. As P4 further emphasized, “a lot of the trans people I know want to be able to more accurately represent their experiences in videogame form.”

Overall, this theme reveals another mode of gender expression in avatar customization: Players customize features for their avatars that accurately mirror their own characteristics, experiences, and emotional connections, transforming these features into symbols of their self-identity. As an integral aspect of this self-identity, players’ gender identity can thus be reflected and expressed through the avatars they design. In this mode of expression, gender is expressed in a more personalized manner that is based on the individual player's unique experience and feelings. The role of customization options in gender expression does not have to rely on gender norms or stereotypes. In other words, options directly related to gender are not the only pathway to construct gender identity. This implies that game designers might recognize the extensive potential of customization options in terms of gender expression and could enhance the possibility and inclusivity of customization systems in expressing gender by offering more diversified and detailed editable features.

Exploring and Experimenting With Gender

Many participants regarded the exploration and experimentation of gender as an essential process for gender expression, as evidenced by P2's (gender apathetic, any pronouns, Canada) experience: “I’ve played where you can be a female character and have short hair, and that's always really cool to me […] Yeah, I usually kind of explore the gender expression in characters through the clothing.” This suggests that the exploration of customization options in games is closely linked to the exploration of gender identity. Such explorations extend beyond the confines of the virtual world, inspiring players to reflect on and understand their gender identity in the real world. For instance, P9 mentioned how exploring customization options within games has played an enlightening and supportive role in their exploration of gender identity in the real world: “exploring titles like Miss or Mister or the Sir […] definitely helps […] [I will] think about what I really prefer? And if I were to transition, where would I go? I think personally, that's something I haven’t figured out yet.”

P10's (transgender woman, she/her, Canada) experience is another notable example. Unlike P2, who was still in the process of exploration, P10 shared her story of finding her name in both game and real world: while creating her avatar in Final Fantasy XIV, she chose a name at random that resonated with her. It wasn’t until a friend, with whom she had only interacted in the game, naturally addressed her by this character name during a voice chat (P10 calls this “a big moment.”) that she started to think: “maybe there's more to this than just playing a character in a videogame.” Later, as she started hormones, she was trying to pick a new name for herself and found “nothing was clicking” on name websites. However, when she logged into the game, a guild mate greeted her by her character's name, and in that moment, she thought, “oh, well, maybe that's what I’ll use for myself.” Eventually, this name became her obvious choice when she legally changed her name.

Additionally, some participants expressed the desire for going back to the avatar customization system to modify avatar characteristics during the gaming process. P11 (gender fluid, he/him, USA) elaborated: “Having a lot of facial and hair options for me personally [is important], just because I changed my hair a lot. So … I like being able to go back and change those [options], which is not something all the games do yet.” This reflects gender as an evolving and expressive process, rather than a fixed identity. As Judith Butler (1990, p. 45) posits in the book Gender Trouble:

Gender is the repeated stylization of the body, a set of repeated acts within a highly rigid regulatory frame that congeal over time to produce the appearance of substance, of a natural sort of being.

However, as P11 mentioned, in numerous games gender options are unchangeable. Once players make a choice, they become “locked” into a specific gender. ⁴ P8 comments that this is a privilege of cisgender players: “[…] if they have the option for pronouns, great. But if they don’t have the option for pronouns, and I click female, it will automatically put in the pronouns that I want. But I realized that's coming from a very privileged standpoint.” This echoes what Butler calls the “heterosexual matrix,” in which “a person assigned female at birth (sex) will ideally grow up to be a feminine woman (gender) and her desire will be directed towards a man (e.g., a person of the opposite sex/gender)” (McCann & Monaghan, 2019, p. 121). The heterosexual matrix not only reinforces the gender binary but also excludes all forms of gender expression that do not fit.

Therefore, we coin a new term “gender-lock matrix” to refer to the design which, once players make a choice within the “male/female” gender options, the avatar they create is indubitably endowed with corresponding pronouns, attire, and gestures that are immutable, thereby extending what Butler (1990) calls the “compulsory order” (p. 33) of sex, gender, and desire from reality into the virtual world. In the process of avatar customization, the gender-lock matrix both limits the possibilities of gender expression and reinforces the privilege of cisgender players. In this sense, avatar customization becomes a part of gender politics, encompassing broader issues of power, identity, and inclusivity.

The players’ desire to return to the avatar customization system and modify its features reveals an evolving process of exploring and expressing gender through avatars in the game. This process is not merely a reflection of personal choice, but also a practice of gender identity. As Butler points: “There is no gender identity behind the expressions of gender […] identity is performatively constituted by the very ‘expressions’ that are said to be its results” (Butler, 1990, p. 34). From this perspective, avatar customization is not just a tool but has the potential to be a dynamic and interactive space that enables players to construct and reconstruct their gender identities through performatively constructions, and thus constituting a significant source of affordance for gender exploration and expression in avatar customization.

Future Research

One potential limitation of this research is the lack of age data, leaving potential correlations between age and gender expression in avatar customization unexplored. Future research could incorporate age as a factor to examine how different life stages influence gender expression.

Furthermore, given that this study employed a qualitative approach aimed at in-depth understanding rather than generalization or statistical comparison, future research could expand the sample size to analyze whether themes are applied differently across various participant types. Insights from game developers could also be explored to complement the player-focused findings and provide a deeper understanding of design practices for gender inclusion.

Conclusion

In this research, we examined the experiences and strategies of players in expressing gender through the avatar customization systems of videogames. In the book Video Games Have Always Been Queer, Bo Ruberg (2019) points out that “even games that appear to have no LGBTQ content can be played queerly, and all games can be interpreted through queer lenses.” In other words, the queer potential of videogames doesn’t necessarily require pre-existing LGBTQ content to be activated by players; rather, it can be released through the player's perspective and interactions. As the first two themes discussed in this article have demonstrated, even within avatar customization systems embedded in a binary gender framework, players could resist and reconstruct this binary system in various ways, achieving diverse gender expressions that transcend structural limitations through their avatars. Thus, while avatar customizations are frequently lacking queer design, they have always been queer.

As we have proposed in this article, the “gender-lock matrix” in avatar customization results in a coercive compromise that suppresses the possibility of diverse gender expressions and explorations. Similar to the concept of the “heterosexual matrix,” the “gender-lock matrix” constructs a “compulsory order” through technical means: players must choose between male or female, after which they are assigned unchangeable pronouns (typically, he/him or she/her) and an avatar option library corresponding to that gender choice. Some games even program different dialogues and reactions from NPCs (nonplayer characters) based on the player's chosen gender. Although players can use mods or similar technical means to alter game data of avatar customization systems and disrupt this order (Scacchi, 2010), these methods are often considered “illegal” operations (Kretzschmar & Stanfill, 2019), thereby reinforcing and granting legitimacy to the binary gender framework.

Our findings offer critical insights to game designers: on the one hand, it is significant to recognize that players might use customization options for gender expression and exploration, either through the embedded gender norms or by integrating symbolic representations of self; on the other hand, breaking the “gender-locked matrix” in avatar customization would provide players more room to resist hegemonic oppression and reshape the ongoing processes of gender expression, exploration, and experimentation. We propose that avatar customization systems are not only the beginning stages of gameplay, but are transient wardrobes of opportunities where players can return to repeatedly try, explore, and experiment with gender identities.