Following the publishing of the literature review about character representation and diversity in games, the next step in the analysis phase of the Gender in Play – Representations of Gender in Games project consisted of a large-scale analysis of character representation in the Finnish game industry.
Setting the base: examined content and analysis guidelines
The base of the analysis were games made by Finnish developers, released from 2018 to 2020. This list of around 240 games in total has been provided by Neogames. Considering the large sample and the short timeframe available for the analysis, the first step needed to be taken was deciding on the selection (i.e. exclusion) criteria which would make the workload realistic, on the one hand, and bear fruitful results in line with our project aims, on the other. Thus, when taking into account the large number of games featuring exclusively animal or abstract characters, the decision was made to only examine games with human characters, or those humanlike enough, with the presumption that those will bring out valuable results considering aspects of gender representation and gender-based violence in games. “Humanlike enough” refers to when despite not being human, the characters expressed decidedly human behaviors and performed human actions, such as going to work and having relationships. One example of such characters can be found in the game STONE, with its protagonist being a “hungover koala detective [who] wakes to find his lover Alex has been kidnapped.” (STONE on Steam, May 8, 2021) Additional to the games with no human or humanlike characters, educational, virtual reality and iWall games as well as games based on intellectual property were excluded from the analysis. Limiting the number of games to one per company per year, our final sample size was reduced to the total number of 117 games.
With that still being quite a large sample, the following strategies were agreed upon: 1) the focus will be put solely on the controllable characters and 2) the games won’t be played. Instead, the information was gathered from sources such as the marketing materials provided by the developers/publishers and, when available, gameplay videos, limited to 15 minutes of gameplay. This information, although very limited, paints the picture of the Finnish industry “at a glance”, as well as shows the general insight and provides key information to any interested person before buying or downloading the game.
Once the list of 117 games was finalized and the examination materials agreed upon, the different aspects of the analysis were organized. As the name of the project says, our interest lies in understanding representations of gender in games. As we also understand that there is a number of different factors that can affect how gender is experienced in society, we opted for a more intersectional approach, which in the context of this research meant looking not only at gender representation, but also the sexual orientation, the skin color, age, religious symbols and disability connected to the characters. Those were followed by specific questions regarding the presence of violence and possibly violence against women. The choice of, at this point, focusing on skin color rather than ethnicity was due to the brief nature of this analysis, which focused more on the visual and the “apparent” features in games, and thus does not allow going deep enough into the games to understand the ethnicity of the characters.
The analysis was based on the visual, written or audio materials provided by the game companies with a special attention dedicated to finding as much publicly available information as possible about the examined characters. In terms of gender, for example, with this approach a character simply looking feminine, androginous or masculine was not enough for us to mark them as a woman, non-binary or man. Instead, an effort was put into looking mainly for character descriptions and pronouns used. This led to many characters being marked as “unspecified” as we could not assume the intent of the developers when the descriptions were not clear enough. With that in mind, after the analysis was completed, the results were shared with companies whose games were analysed with a chance to comment on the results and notify us in case of misinterpretations. Although the findings presented in this article are those resulting from the conducted analysis, the information gathered from the developers’ feedback will also be provided where relevant as separate insights.
Most of the games did not specify the characters’ gender (68). 39 games had character selection or creation offering at least one man and woman option. The ratio between games that offer men and women characters is almost equal, with the number of 63:51 in favor of men characters. Non-binary characters are recognized, however, in a small amount of only 3 games. Our analysis showed no indications of transgender characters.
The skin color of examined controllable characters is presented based on the colors’ hex codes (Picture 3). The samples were collected from the area of the skin most representative for the character (i.e. avoiding shaded areas, bright spots etc.). However, it is important to note that the samples were not collected from the games themselves but from the available materials and YouTube videos which indisputably had an influence on the quality of the sample and the overall end results. Additionally, in some cases the in-game scenes and characters were distinctively stylized, which influenced the skin color sample and might lead to possible non-realistic depictions of those characters (Picture 4).
Religious symbols were rarely depicted and often difficult to recognize and define. A total of 14 games included some religious symbols, specifically from Christianiaty, paganism, shamanism, or African and Ancient Egyption elements (e.g. Gods). Of particular note was the game Raanaa – The Shaman Girl which has the player playing as the titular Sámi girl as she goes through levels inspired by Sámi gods and shamanism. This was also the only game where we found Sámi characters.
It was not easy to tell the sexual orientation of characters, as it was mostly unspecified and irrelevant to the gameplay and story, with only one game having a character described as homosexual (Your Royal Gayness). Because of that, we chose to also look for the characters’ intimate relationships. This approach was not flawless, however, as a character being in a homosexual or heterosexual relationship does not necessarily mean they are homosexuals or heterosexuals themselves, for example. Taking all that into account, we found 8 games with playable characters involved in heterosexual relationships and 2 games with characters either described as gay or in a homossexual relationship. Finally, an interesting view on the expression of sexual orientation was found in the virtual world game Hotel Hideaway in which the players communicate and interact with others freely and have a possibility to express their own sexuality.
This was the category most difficult to determine and gather information on; it was found that specific age of the characters and their age group is something developers do not often bring up in discussions or clearly state in the promotional materials. Therefore most of the characters in our analysis were marked as unspecified (108 games in total). However, in such cases an estimation of age has been added, so that a general view and indication of the most represented age category could be brought up and discussed. It might not come as a surprise that most of these characters were estimated to resemble the age category of young adults or middle-aged persons. Personal interpretation, however, can be misleading as it was shown in the example of the character Audrie Smoothspy from The Spy Who Shrunk Me. Audrie, who based on our interpretation seemingly fit the adult category, is actually in her 50’s as it was confirmed by the developers.
Cases where characters’ age or their age category was explicitly mentioned in the game itself or the promotional materials were only a few, concretely: child (1), teenager (1), young adult (1), adult (2). Table 1 gives an overview of all age categories and their respective age ranges.
Table 1. Age categories with respective age ranges (as defined by United Nations for Youth and Merriam-Webster Dictionary)
The analysis showed only 2 cases of explicit disability, one game included a character with a physical disability (a wheelchair user) while the other incorporated a character with mental disability (depression). The existence of disability (among other features such as sexual orientation or age) was very difficult to determine without playing the games. It is, however, important to note that in some games we have indeed noticed potential representation of characters with disabilities, which we unfortunately could not tell for sure from the examined materials. One such example, Lotus, a monk character from Friends & Dragons, was actually confirmed to be blind by the developers of the game.
Violence and violence against women
Violence is a feature often present in games today so most of the examined games (87) also incorporate or depict some form of violence. Violent actions or behaviors found in these games mostly reflected the core game mechanics of fighting enemies and trying not to get hurt. When it comes to violence against women, violent actions or indications of violence against women were found in 8 cases. In-game violence is directed at women characters in different forms and it is important to make a distinction about how the character’s gender motivates these actions. Games were marked as depicting violence against women when they involved imprisonment of women characters (that needed to be saved), used gender-based offensive language towards women (e.g. “that old hag” or “the dread wizardess Galatax was rearing her ugly head”), or when violent actions were directed towards women heroes and/or villains. As it was the case in the previous category, this feature was also very difficult to determine based on the small sample of materials and the limitation of not playing the games. This limitation emphasized the need for a closer examination of the games where the examples of gender-based violence against women were found. Therefore, in the next phase of the analysis we will examine the aspects of gender-based violence against women characters in more detail, by playing the games and analysing them from a player’s perspective.
As already mentioned, our main limitation in analysing such a large sample of games in such a short period of time was finding the best and most representative source of information. Making a comprehensive overview was rather difficult considering we were not playing the games and were basing our information solely on published materials and walkthroughs. This means that the examined content was largely influenced and our analysis steered by the choice of the providers of gameplay videos (e.g. often the videos will start only after the character selection has been made, which made it difficult to determine which characters were available at start; the character selection often depends on the person playing and as most of the video creators were men, data gathering about women characters was scarce). Furthermore, getting to know the characters, their stories and identities was not possible with just 15 minutes of observed gameplay. For this to be possible, one would need to watch/engage in full gameplay, do a thorough research of published materials and/or even ask the developers directly.
Considering the limitations of the large-scale analysis, a need for a more focused approach in examining character identities and the narrative elements in the games has emerged. Therefore, the next segment of the analysis will be an examination of a smaller sample of games with a special focus on:
- representation of women characters and character diversity in Finnish games
- gender-based violence against women in Finnish games
The selected examples are games from the large-scale analysis which appeared as interesting and/or relevant examples of features mentioned above; they will be played and analyzed from a player’s perspective. As within the scope of this analysis we cannot play all 117 games and therefore will not be able to address all interesting cases and efforts devoted by the Finnish companies, we also find it important to bring this topic into a discussion with developers themselves. While the presence of some specific groups was found to be lacking in the games analysed, there is interest and initiatives (such as this project) aiming to improve that. Therefore, in the next stage of the project, we will conduct interviews with the developers, discuss their good practice examples and challenges in creating diverse characters in the gaming industry.
Our findings about underrepresented features, as well as good examples of women and diverse characters and possible cases of gender-based violence will be added to the information gathered from the interviews and will be a source to inform and assist developers in creating and designing their future characters.