Players as Coresearchers

Approaches in Game Studies

Analyses of games and how they are designed tend to be based on the researchers’ own gameplay. This approach has become an accepted method for games research, and it is based on the idea that to understand how a specific game works, the researchers need to experience it themselves (Mäyra, 2008, pp. 165-167; Mortensen, 2002). Game analyses based on self-play typically take a structuralist approach in which game mechanics and other features are analyzed according to their functionality. Relying on the researcher’s gameplay only, such projects tend to study games from the perspective of an implied player believed to be intended by the designers and suggested through the design.

Scholars have made a number of attempts at outlining frameworks or typologies for how to analyze computer games (e.g., Aarseth, 2003, Consalvo & Dutton, 2006; Konzack, 2002). From a design-oriented perspective, Järvinen (2007) introduces a methodological toolbox that takes into account the elements of a game as well as the designed interaction options. He calls this a systematic methodology for practice-oriented game analysis and design that “explain[s] the inner workings of games and their players” (2007, p. 143). In Rules of Play, Salen and Zimmerman (2004) investigate game design principles from an analytical and theoretical perspective. The book allows game design students to understand the background for good game design, but even though it discusses players’ dynamic interaction with games, the arguments are not supported by any empirical data. These methods are able to describe the design of a game in great detail, but by referring to an implied player, they do not take into account the fact that games are experienced differently by different players.

Studies that include player participants are commonly carried out in sociologically oriented research into player interaction and the play activity. While researchers interested in multiplayer phenomena and the social structures of games often lean on ethnographic methods and participant observation (e.g., Hung, 2007; Mortensen, 2002; Steinkuehler, 2005; Taylor, 2006), interviews and observations have been popular in research on player behavior and activities (Egenfeldt-Nielsen, 2003; Enevold & Hagström, 2009; Klastrup, 2003). It may be argued that this research presents an indirect approach to investigating game systems, as they study how a specific design guides player behavior by focusing on the social processes of gameplay and the interaction between players and game system. However, with focus on player behavior and interaction, these studies rarely attempt to say anything about games as designed systems, and are interested in the system only because of its impact on player activities. Any attempt of using this research to make any claims about game design would only create knowledge about the consequences of design choices, and not about the design itself.

Without any method that investigates how game systems are interpreted by players, we have an empirical shortcoming within research that tries to understand the dynamic aspects of game systems. In order to develop a fruitful approach to studying game systems from a broader perspective, we need to combine analysis and studies of social play with an approach that takes into account how different players interact with the game system. Joining together the analytical and player-oriented approaches presented above may help us on the way. This kind of triangulation of methods may not only increase the rigor of the study by being designed to investigate what we want to investigate (Patton, 2002, p. 14), but will also increase the accuracy and trustworthiness of the study (Patton, 2002, pp. 93, 247-248) by illuminating the research from different angles.

Other Relevant Approaches

In media and communication research, the study of the subjective experiences of media users is well established. The British cultural studies tradition put a particularly strong emphasis on popular culture and the idea that media use must be understood from the sociocultural perspectives of the media users themselves. Hall’s (1980) studies of how media users could reject or negotiate the intended meanings of a media text is a prime example of this approach. The study of subjective experiences is also well known in experiential psychology and the psychology of religion. In the 1950s, Allport studied subjective religious experiences through reports by individuals (Wulff, 1991, pp. 231-235, 586-590), and the original aim of experimental psychology was to measure human experience by having individuals report on sensations in a controlled environment (Ashworth, 2008, p. 5).

Today, we see a growing interest in qualitative methods in psychology for the study of human experience (Smith, 2008; Smith, Flowers, & Larkin, 2009). Interpretative phenomenological analysis (IPA) is a branch of qualitative psychology that focuses on how people make sense of major life experiences. As the goal of IPA is to get insight into the participants’ subjective interpretations, it uses methods that allow the participants to “offer a rich, detailed, first-person account of their experiences,” such as in-depth, semistructured interviews, diaries, and focus groups, all of which let the participants to reflect freely around their experiences (Smith et al., 2009, pp. 56-57). The researcher takes a double hermeneutic position in which the researcher is trying to make sense of the participants, who are trying to make sense of their own experiences (Smith et al., 2009, p. 35). Similarly to the study presented in this article, the IPA researcher only has access to the subject of study through the participant’s own account of it.

I have argued elsewhere that we may fruitfully turn to the field of human-computer interaction and usability studies when investigating how game features work in practice (Jørgensen, 2008, pp. 207-209). Usability is a term that measures the user’s experience with a product in terms of how effective and easily it can be used. Usability testing is “the process of learning from users about a product’s usability by observing them using the product” (Barnum, 2002, p. 9), and it refers to techniques that can be used to evaluate a system. Usability testing is not so much a methodology as it is a collection of methods for testing the effectiveness of a product. The background for usability testing is in the practical needs faced by product developers. These needs are not only connected to whether the product is able to do what it is supposed to, but also to whether the users find it effective, efficient, and satisfying (Cairns & Cox, 2008, p. 213). In many kinds of product development, a gap tends to open between what the developers believe is a functional design and what the users actually find functional; this advocates the need for evaluation from actual product users in order to decide whether a system is matching the needs, wants, and levels of expertise of the intended users (Nørgaard & Sørensen, 2008, p. 10). Usability testing methods focus on the users’ experience with the product, and rely on the product users’ problem-solving processes and live feedback. During testing, the developers may observe how actual users interact with the product and what problems they encounter.

Usability testing is increasingly popular in game development. When adapting the method to game contexts, however, one must keep in mind that games are entertainment and therefore qualitatively different from task-oriented software. Usability criteria must therefore be adapted accordingly (Hoonhout, 2008, p. 75). As game-oriented usability studies have a specific focus on the relationship between how a game is designed and the players’ subjective experience of it, this is an interesting approach if coupled with relevant academic perspectives.

Players as Coresearchers

Based on the approaches discussed above, this article argues for the use of subjective player experiences in researching the game system. By including empirical players as coresearchers, the formulated method will have a solid foundation based on a meaning-making partnership (Hesse-Biber & Leavy, 2006, p. 128) between researcher and participant, in which interpretations are discussed and evaluated collectively. Both researcher and participant are active in the research process (Smith et al., 2009, p. 58), where knowledge is constructed through a discussion in which the participants are invited to interpret specific game features from their perspective as experienced players. Using open-ended conversations or semistructured interviews, the researcher allows the participant to take part as an equal partner in the analysis and construction of knowledge. This happens in a situation where participants are invited to interpret and evaluate, and where the researcher’s role is to challenge the participant and discuss analytical alternatives.

This hybrid between player and researcher is not a new concept; Copier has emphasized that this is an important combination that should not be overshadowed by the designer-researcher hybrid (Copier, 2003, p. 410). The difference between this and Copier’s account is that the hybrid in our case is not a professional researcher who also is an experienced player, but an experienced player who takes on the analytical role of a researcher. As a player-researcher hybrid, the participant changes status from an object of study to a knowledge-producing source. What the player-researchers lack in academic experience, they make up for in practical game understanding.

The partnership between participant and researcher, therefore, has the strength of producing sophisticated knowledge based on the combination of the researcher’s academic experience and the player’s game expertise, as the player becomes an expert guiding the researcher in understanding phenomena (Lucero, 2009, p. 88). This approach emphasizes IPA’s view on participants as experts on the specific experience that the researcher wants to study. It also puts weight on the bonds to constructivism and its understanding of knowledge not as a “truth” that exists objectively in the world, but as constructed by human beings. From this perspective, understanding a phenomenon comes into being through interpretation and analysis (Guba & Lincoln, 1998, p. 211; Patton, 2002, pp. 96-98; Schwandt, 1998, p. 236; Smith et al., 2009, p. 35). The researcher’s role is therefore not to discover the objective truth of a phenomenon, but to create a theory of how the phenomenon in question can be understood. Combining the researcher’s analytical and methodological expertise with the game experience of empirical players, the approach described in this article broadens the frame of understanding and constructs knowledge based on both a practical and an analytical understanding of games.

Important to the method is the idea that the players’ own interpretation and experience are central to understanding games. As the article puts heavy emphasis on player interpretations as an important source of data, the proposed method may be seen as a reader-response approach to games (Abrams, 1993, pp. 268-272). The methodology is certainly inspired by this and takes many of the same premises for granted, such as the idea that interpretations are subjective and the meaning of a text is constructed by the interpreter during the reading process. At the same time, however, the method acknowledges that games, through their interactive nature, are experienced in different ways by different players. For this reason, the fact that meaning making is subjective is not a surprising observation. Taking this stance, this article relies heavily on empirical players and separates itself clearly from Iser’s (1978) focus on an implied reader. Although good reasons exist for criticizing the idea of the implied reader in traditional media, any notion of an implied player of games is a more serious flaw as the dynamic nature of games presupposes that different players approach the games with different strategies and interaction patterns. Instead of focusing on an abstract player figure, the orientation toward empirical players is closer to cultural studies and Janice Radway’s (1984) approach where she introduced ethnographic methods into reader-response by interviewing readers to understand their reception of romance novels.

Studies

For purposes of, and as support to, the player studies, the following studies were conducted:

The main focus of the research was, however, the player studies. Seventeen individual interviews were carried out, as well as one focus group of five participants. All player studies took place in office spaces at University of Bergen and were conducted in Norwegian.

Focus Group

The focus group worked as a pilot study in which questions for the individual studies were tested (Adams & Cox, 2008, p. 25). During the session, five males between the ages of 25 and 30 were shown a selection of screenshots from the selected games, and were asked to discuss them based on questions about the gameworld, the interface, and the relationship between the two. The participants were selected to make the focus group as homogeneous as possible (Adams & Cox, 2008, p. 24; Smith et al., 2009, pp. 49-50). All participants knew at least one of the other participants in the focus group beforehand. They were invited based on their experience with games and the assumption that they would create a good balance within the group with respect to game preferences and verbal dominance.

Observations and Conversations

Observations of and conversations with individual player participants were the most extensive part of the study. Participants were recruited on the basis of self-selection, as information was posted in local game outlets, on the university campus, and on the web forum of a local role-playing and board game organization. The snowball method was also used to recruit participants from the outskirts of my own network (Patton, 2002, p. 237). Four females and 10 males were recruited; three of the males agreed to participate in the study of two of the selected games. The study was based on what I have earlier called the video commentary model (Jørgensen, 2008, Ribbens & Poels, 2009), which presents the participants with the cued recall and think aloud methods from usability testing (Barnum, 2002, pp. 94-95). Cued recall allows the participants’ gameplay to be recorded by video-capture software, before using the recording as backdrop for a conversation between participant and researcher. As an introduction to the study, the participants were shown screenshots from this and another game and were asked to compare them with respect to genre and interface. The participants’ gameplay was then observed and video captured while playing for approximately 30 minutes. After playing, the participants were invited to start the conversation by thinking aloud, thereby commenting on their own performance while watching the recording. With this as backdrop, the researcher added comments and asked questions as interesting features occurred, including the participant into an analytical process.

The main challenge of the player studies was to get the participants to verbalize their interpretation of the UI and its relationship to the gameworld. Semistructured conversational interviews with a high degree of open-endedness, and focusing on a specific gameplay context, was considered to be a fruitful approach as participants could talk freely, but still be restrained by a specific topic (Hesse-Biber & Leavy, 2006, pp. 125-126). For discussing more abstract topics, the researcher presented statements and questions that were likely to be opposed by the participants, such as “Can the military units see the health meters?” and “Who is the avatar talking to?” This positioned the researcher as a naïve, curious outsider (Smith et al., 2009, pp. 36, 64). This also created a setting for an active interpretative discussion, in which a meaning-making partnership between researcher and participants (Hesse-Biber & Leavy, 2006, p. 128) was established and the participants were included as equals in a conversation with the researcher. The focus on video captures and screenshots allowed the otherwise abstract conversation topics to be associated with specific examples that illustrated the integration of system information into the gameworld. In addition to allowing the participants to talk about specific features, these questions were also meant to provoke discussion and reflection, which they did. This will be elaborated on below.

Evaluation

In this section, I will evaluate the fruitfulness of using the proposed method in this research project. In spite of the exploratory approach that used participants as coresearchers and provoked critical discussion through the inclusion of biased questions, the study turned out to be successful, as the participants argued against the researcher and suggested their own reflective interpretations.

Before going into a detailed evaluation of the player studies, some remarks should be made about the game analyses and the developer interviews. The initial analyses of the games used as cases were carried out as a point of departure for formulating questions for use in the interview situations. In hindsight, I believe the project could have benefited from the inclusion of games from a wider selection of genres. Games with a stronger narrative and a player character with a more distinct personality would probably have provided a deeper insight into how players connect fiction and interface elements in games, and how character personality affects the sense of association with the gameworld.

The group interviews with developers were also carried out in order to contextualize the player studies, by creating a deeper understanding of the developers’ intentions as well as their thoughts on how the UIs should relate to gameworlds. The information gained from the developer interviews supported and illuminated the findings from the player studies, and worked, therefore, to emphasize the fruitfulness of the player studies. The developers were working on genres that are very different from those used in the player studies, but they still presented corresponding attitudes toward the integration of UIs in gameworlds. This implies that the method is able to create trustworthy and dependable results (Guba & Lincoln, 1998, p. 213).

Although generally successful, the more exploratory player studies are somewhat more controversial. Using participants as a source for collecting data about a third participant is a matter that potentially interferes with the reliability of the data (Smith et al., 2009, p. 36), as the researchers run the risk of not investigating what they were intending to (Adams & Cox, 2008, p. 18). I believe that securing reliability is dependent on careful design of the interview guide, as well as having information-rich participants (Patton, 2002, p. 46,; Smith et al., 2009, p. 58). In the case of this study, careful design of the interview guide means that the researcher’s abstract hypotheses and concepts wree operationalized into procise statements or questions that made sense from the perspective of players (Jørgensen, 2009). Discussion topics such as whether or not system information should be interpreted as part of the gameworld, and who the avatar is addressing when stating “I cannot carry any more,” were fruitful entrances for reflective conversations about the status of different game elements. As the interview guide was designed to make the participants question the statements presented to them, the participants were automatically put in a position in which they became research partners equal to the interviewer. Of course, asking naïve and biased questions could backfire, and the participants would support the idea proposed by the topics presented, thereby compromising the dependability of the data. However, the participants often smiled or burst out laughing when confronted with such questions and emphasized that interpreting the gameworld in the same way as the real world or any fictional world would feel wrong. ¹ Thus, the participants made it clear that they did not approve, and the discussion format ensured that other biased topics were exhausted before moving on.

In general, using player participants as coresearchers was a successful approach that opened up for a very open relationship between researcher and participant. I believe part of the success was adapting a student-tutor relationship (Adams & Cox, 2008, p. 23) in which the researcher takes on the role of a student who is discussing with an expert. This enabled the participants to open up and be confident that their views were as interesting and valid as any other interpretation, and that they were encouraged to argue against the interviewer.

Having information-rich participants turned out to be important. Selecting participants purposefully may lead to information-rich cases from which one can learn a lot about important issues central to the study (Patton, 2002, p. 230). In this study, it is the sources and not the cases that are information-rich, as the role of the participants was to be expert sources and not cases. Due to the hypothesis that experienced players would have the greatest analytical understanding of the games, the participants were recruited based on their knowledge of one or more of the games used in the study. A specific level of education was not required, but relying on self-selection, however, the study did end up with participants largely with a university or college background. In general, the participants had good interpretative skills and a high level of analytical understanding of the games, but some participants needed time to accept the premise for the topics presented to them. In these situations, the challenge was to introduce the participants to the mindset of the study without making them feel subordinate and to avoid forcing the hypotheses of the study onto them. This was solved by an introductory phase where respondents were asked to compare screenshots and talk freely about the usability of the game, but in hindsight, this may have made parts of the interviews too general for being directly applicable to the study. In addition, as many of the participants had academic backgrounds, they may have overinterpreted features, and their views may not be representative for the average gamer. This does not compromise the study, however, as critical reflection was desired, and as the study was carried out for the purpose of generating academic theory of game interfaces. Another issue is connected to the risk that self-selected participants were dedicated fans of the selected titles and that they therefore may have accepted a greater degree of inconsistency in UI design than the average player. As the data shows consistency among player interviews, and is also supported by the developer interviews, we have no reason to believe that this affected the study significantly.

In addition to research conversations, the individual player participants were subject to screenshot analysis, gameplay recording, and cued recall where the recording was used to discuss game features and the participant’s performance. Choosing these methods was connected to the hypothesis that specific examples would better support a discussion of abstract concepts, and that gameplay in context would create a sense of immediacy and be most appropriate for understanding the dynamics of the game and for capturing emerging events. In the study, however, the screenshot analysis turned out to be the most fruitful of these approaches. Although the discussion of the recorded gameplay very well highlighted the auditory interface, and emerging events, the screenshot analysis better captured the interpretation of the graphical UI, perspective, and comparisons between different games. Having participants comment on their performance through cued recall was only partly successful, as the discussion most of the time strayed away from the recording. However, as a way to keep the game fresh in the mind of the participants, as a conversation starter, and an opportunity for the researcher to take notes, the video capture was fruitful, although the same situation could have been achieved by letting the participant play without recording the gameplay. As the auditory interface was of secondary importance in this study, leaving out the video capture altogether could have been done to make the preparation time more efficient.

In the case of the individual interviews, using the video capture as an alternative to screenshot analysis for discussing specific features would mean pausing and rewinding the recording at appropriate times, which would be disruptive for the conversation flow. In connection with the focus group, using video capture was not considered, as this would demand focus on one particular person’s gameplay. Another challenge related to the focus group was having five participants who often got into heated debate with each other and consequently were speaking simultaneously. This made the transcription process hard, but had been accounted for as the participants were selected on the basis of differences in dialects and voice qualities. This was not, however, a problem in connection with the developer group interviews, as they rarely spoke simultaneously. This may have been due to familiarity with interview situations from the media.