Game Studies

The Structure of Videogame Preference

Abstract

This study investigates the statistical patterns of videogame preference orientations based on a broad survey of media- cultural- and lifestyle preferences among Norwegian college and university students. It discusses the findings in light of previous research on game preferences and game culture. Particular attention is given to established notions of "core" versus "casual" gaming, including the gender aspect.

We used Multiple Correspondence Analysis to identify patterns of association and opposition within players' stated favourites. In our analysis, the most dominant dimension articulates an opposition between an inclusive and family-friendly orientation at one end of the spectrum, and an exclusive, high-tech and dark/mature preference orientation at the other. This primary dimensional axis correlates with a demographic divide between low-frequent and predominantly female players and high-frequent and predominantly male players. The second most dominant axis articulates an opposition between a niche and a mainstream preference orientation, which is not linked to gender distinctions.

Through cluster analysis, we see that players' game preferences can be grouped into relatively distinct and coherent orientations. Drawing on the wealth of data in the original survey, we also see how player profiles can predict players' wider media- and cultural preference orientations. Among core gamers, the "E-sporters" and "Roleplayers" profiles reflect a fairly unambiguous association with autonomous game culture. In contrast, a large cluster of dedicated action- and sport fans, which includes only 4% women, is located at the intersection of game culture and mainstream masculine entertainment culture.

Keywords: sociology, videogame preference, player studies, gender, correspondence analysis, game culture, player types, player typologies, stereotypes, subculture

Introduction

Between 2008 and 2010, Jostein Gripsrud and Jan Fredrik Hovden conducted two broad surveys of Norwegian higher education students' cultural- media- and lifestyle preferences (Gripsrud, Hovden, & Moe, 2011), including 90 main topics, from music and literature to TV consumption and leisure activities. In this survey, respondents were asked to mention up to three favourite videogame titles, and also how often they play videogames [1]. Of 2367 respondents, 758 students mentioned one or more favourite games, and 533 different titles were mentioned in total.

This study uses this data to look for statistical patterns of association and opposition among the favourite titles mentioned by the respondents. We are particularly interested in patterns that can throw light on distinctions between core versus casual gaming, and the gendered nature of gaming preference orientations.

For example, if a respondent states that one of her favourite games is Mass Effect (typically stated in general terms, at the series level), then how statistically likely will it be that she will also point to the Championship Manager/Football Manager series as a favourite [2]?

First, our analysis seeks to identify some underlying principles or dimensions according to which such interconnections within players' preferences can make sense. Second, we seek to identify types of preference orientations, articulated as clusters of gamers who tend to prefer the same kind of series. Finally, we want to see if such clusters also have some distinctive characteristics in terms of their cultural- and media consumption more broadly. To what extent can gamers be categorised not only into different player types but also into different cultural types?

The data were collected some years ago, and do not capture the particular state of the field at present, but they are nevertheless a rich resource for studying systematic relationships between different game preferences. Even if the games market has changed since 2010, basic conventions of popular game series and game genres evolve slowly, and we should expect underlying parameters and patterns of game taste to remain relatively stable over time. The comprehensive scope of the surveys also gives us a unique opportunity to connect game preference orientations to wider media- and lifestyle preferences. We will also be able to compare media use and cultural preferences between players and non-players, the latter being a large group, constituting 60% of the total sample.

Our choice of methodology is informed by the work of Pierre Bourdieu, who applied Multiple Correspondence Analysis (MCA) as a tool to analyse the underlying structure of a field of cultural distinctions (Bourdieu, 1984). Like Bourdieu, we ask how patterns of association and opposition reflect different cultural dispositions within a given field of cultural consumption. To what extent is there an inner logic to the field of videogame preference, and what are the main dimensions of taste differentiation? However, whereas Bourdieu's aim was to analyse the relationship between taste orientations and the reproduction of class-based hierarchies, our study is focused on the structure of videogame preference as it relates to established distinctions and expectations within game culture, in particular gender distinctions. Such questions have not been studied before using this kind of methodology.

Previous player preference studies

Our approach to the question of gaming preferences differs from what we have seen in preference studies so far, although there are some relevant similarities. There are studies and statistics that link age and gender to platform and genre categories, in some cases also including favourite titles. Numerous studies have confirmed that there is a tendency for girls and women to prefer lighter and more casual game experiences (Krotoski, 2004, Kerr, 2006). Crawford and Gosling (2005) found that female students were strongly inclined to prefer games on mobile phones — even if the study reflects the pre-smartphone era. The latest report on children and media use from the Norwegian Media Authority (2016) shows that 16 year-old girls are far less inclined than boys to mention any favourite game at all, and that when they do, there is a strong tendency to mention different games than boys do (2016, p. 57).

In a survey on gender differences among young adult's motivations for video game play, Lucas et.al (2004) also found systematic differences with respect to genre preferences: Female respondents reported less interest than male respondents in games that combine a strong competitive orientation with 3-dimensional navigable environments ("three-dimensional rotation"). This finding is consistent with Jesper Juul's analysis of casual games (2010). A typical feature of contemporary games with a wide appeal, Juul says, is that they either do not use or strongly de-emphasise navigable 3D space, thereby becoming more inviting to players, and more easily able to blend with the space in which the game is played.

A number of studies over the recent years have aimed to construct models and typologies of videogame player motivations and styles, linked to demographics and playing habits. Even if such methods are based on pre-defined motivational variables, their aims are similar to ours in that they look for some underlying patterns and types of preference orientations. Much of this research, however, is concerned with motivations and preferences in the context of particular games. The majority of studies focus on online game environments, and the need to optimise their design in order to appeal to diverse types of players, motivations and styles. Such studies are indebted to Richard Bartle's influential killer/socializer/achiever/explorer-typology of online RPG players (Bartle, 1996), and broadly follow a similar player typology approach. A highly influential study in this tradition is Yee (2006), who uses factor analysis to model player types based on 15 motivation variables. His analysis suggests achievement, social motivation, and immersion as overarching components of motivation types, and connects this finding to demographic variables.

A few studies work with player motivation typologies across genres and platforms. Bateman, Lowenhaupt and Nacke (2011) have done a sequence of surveys looking for correlations between a range of factors linked to playing style and preference, with the aim of developing a personality-type typology of players. Their findings suggest that systematic differences with respect to a preference for complex challenges may be a relevant dimension of gaming personality. Vermeulen et al. (2011) go further in this direction. They suggest that a preference for core and complex genres is a distinctive kind of acquired videogame taste, and that this unique taste of the experienced gamer cuts across genre preferences as well as gender divisions.

Through a series of comprehensive surveys and qualitative interviews, Kallio, Mäyrä and Kaipainen (2011) investigate the way in which players' preferences, motivations and identities are shaped and develop within the practices of daily life. Their concluding analysis of nine gamer profile categories of "mentalities" is relevant to the kind of preference orientations that we seek to reveal in our own study. For example, they differentiate between a "Gaming for Fun" type of player, who emphasises "speed, progress, flow, skilfulness", and an "Immersive Play" type of player, who is a "fantasy-driven hobbyist" (p. 344).

A closer methodological sibling to our own study, and also an illustrative case in point with respect to the link between player preference studies and studies on game culture, is Ip and Jacobs (2005), who use multivariate factor analysis of pre-defined gamer-classification variables to analyse the main dimensions of segmentation in the games market. Their cluster analysis suggests two main groups, supporting a distinction between hardcore and casual player attitude and experience.

Gamer culture

One of the aims of our study is to contribute with relevant insight into cultural differentiations of core versus casual gaming, and the role of gender in such differentiations. Graeme Kirkpatrick's study on the historical formation of the notion of "gamer" (2012) is a relevant background in this context. Drawing on Bourdieu's sociology of culture, Kirkpatrick finds that the notion of the "gamer" as a point of cultural identification emerged from within hobbyist computer culture in the mid- to late eighties, and was key to the establishing of games as a distinct cultural field. The gamer identity, Kirkpatrick observes, was formed around the idea of the authentic gamer, associated with a characteristic preference orientation. The authentic gamer was implicitly thought of as a young and tech-savvy male, who is focused on "gameplay" as the ultimate criteria of value and appreciation. The concept of gameplay was thus established as we know it today, as an exclusive marker of a semi-autonomous identity and culture. Kirkpatrick's conclusions resonate with findings in other studies that digital technology and formal systems tend to be perceived as non-authentic and alienating by young girls (Turkle, 1988, Corneliussen, 2014).

At the same time, Adrienne Shaw finds that it is problematic to assume that videogame culture is necessarily "... something very distinct and very different from mainstream U.S culture" (Shaw, 2010, p.404). In a similar vein, Garry Crawford, in his work on the sociology of videogamers (2012), points out that established concepts like fan culture or subculture, although often explicitly or implicitly invoked in game studies, are too narrow and specific to capture the broad range of practices under study (Crawford, 2012, p.107ff). Both these points — on the presumed autonomy of game culture within broader mainstream popular culture, and on the range and diversity of gaming orientations — are highly relevant to the present study.

Method

The methodological approach in our study is different from existing preference orientation studies, which are also looking for underlying patterns and types of videogame preference. We do not use pre-defined variables to describe motivations and preferences, and we do not ask respondents to report directly on such motivations; we do not ask them, for example, whether story is important to their enjoyment, whether they prefer puzzle or role-playing, or whether they would self-identify as gamers. Instead we use a purely title-based and inductive approach, where we simply ask respondents to state their favourite titles. Interpretations are then being left to the researchers, who must make sense of the patterns as they emerge from the data.

The statistical analysis is based on the 32% within our sample of students who stated their favourite titles. As a first step in the coding, all unique titles which belonged to the same series/franchises and did not differ dramatically in gameplay (for example FIFA games) were coded into what we will refer to as game series. We had to exclude a handful of title-mentions that were incomprehensible or too ambiguous (for example "driving games"), but otherwise we were able to assign mentions as members of game series with a high degree of certainty — even if we had to make some educated guesses in the case of apparent misspellings (for example "Sin City" is very probably "Sim City"). In total, the 533 titles named by the 758 respondents were contracted to 328 series.

The first goal of our analysis was to extract from this large number of series a smaller number of underlying dimensions. For this we will use MCA, a form of factor analysis (Le Roux and Rouanet, 2004). Series that end up close to each other along a particular dimension can been interpreted as sharing some common, underlying characteristic. To improve the stability of the statistical solution, we retained only series mentioned by at least 5 respondents [3]. As we also needed information on which series tended to be mentioned together, we had to remove those who mentioned only one series. The final analysis was done on a matrix consisting of 64 series and 417 respondents. In the references below, we have listed the series with original developer only, as most of them will have changed developer and/or publisher several times over the years.

MCA presents antagonisms between game series; it tries to describe the dominant underlying patterns with regard to what kind of series tend to be favoured together and those which do not. However, we also want to get a clearer idea of the variations between groups of gamers. To what extent can we identify groups of players who tend to prefer the same kind of series? For this purpose we have used a cluster analysis technique (here: hierarchal agglomerative clustering on the axes of the previous MCA) to refine our analysis.

Because our respondents are students, and also a self-selected group (— those who have mentioned a favourite game title), our data are doubly skewed towards frequent and hobbyist players. This is obviously a significant limitation, but it also means that the data is well suited to studying differences within segments of players that may otherwise be lumped into a common category of enthusiasts or "hardcore gamers", and who may too easily be associated with a more or less homogeneous gaming culture. At the same time, our population of title-mentioners is large enough to also include a significant slice of less frequent players.

In stating their favourite videogame titles, we must expect that respondents will also express what kind of gamers they want to be, rather than accurately reporting the games they spend most time playing. While this potential aspect of self-expression implies a degree of uncertainty with respect to the exact nature of the respondents' reporting, it is also consistent with the aim of studying player preferences in the context of game cultures and player identities.

The primary dimensions of videogame preference

First, a note on the demographic data. If we look at videogames in relation to use of other media (Table 1), we see that playing videogames is a strongly gendered interest among Norwegian higher education students. While male students are roughly two and a half times more likely than females to watch a movie at home every week, they are 13 times more likely to play videogames every week.

Table 1: "Heavy" use of videogames, literature, and movies. Logistic regression (controlled for age and year of study). N=1830 (click to expand).

Let us then proceed to the analysis of the series titles data. We will interpret primarily the first two dimensions (Figure 1), which account for most of the variance in the matrix.

Figure 1. The space of videogame series. MCA, axis 1-2 [4] (click to expand).

Note that the Super Mario series of platform games are associated with a preference orientation to the left on the map. This means that it is relatively frequently mentioned together with series like Mario Kart and Zelda, and relatively infrequently with, for example, Sim City or Gears of War. Because it is also one of the most popular series among our respondents, Super Mario is a relatively strong marker of game taste, suggesting that a preference for, or lack of preference for, these games are part of what defines the underlying opposition in axis 1. World of Warcraft (WoW), in contrast, is placed towards the Origo (centre), which means that it has no clear profile with respect to the two primary dimensions of gaming taste. In other words: while listing Super Mario as a favourite series is clearly indicative of a particular gaming taste, listing WoW is not.

In our interpretation, this primary axis (left-right in Figure 1) appears to articulate an opposition between inclusive and exclusive appeal. On the left we find broadly popular classic series like Super Mario and The Sims, socially- and family-oriented series like Wii Sports, Singstar and Buzz, puzzle games like Tetris and Sudoku — and Angry Birds, which was at the peak of its popularity during the period of the surveys. Around the centre and towards the right we find games that are generally more time-demanding. They strongly tend to have adult and violent themes and styles, except for the sport titles, and express traditional motives in computer gaming: football, driving, shooting, and fantasy/Sci-Fi. These motives are dominant within hobbyist gaming culture, and traditionally linked to male-dominated popular culture more broadly.

However, even if the primary axis generally confirms commonly established assumptions of feminine and family-friendly versus masculine and specialist preference types, the left side cannot unambiguously be defined as a casual preference orientation. Super Mario and The Sims, which contribute significantly to the construction of the axis, due to their high mention-rate and clear profile, are both relatively big and time-demanding titles, requiring a level of effort and dedication that is not significantly different from, say, Call of Duty (CoD) or the Tycoon series. They differ however from games at the right end of the axis by their either non-gendered or feminine market appeal (Jansz et.al, 2010). On the left side of the spectrum we also see the Zelda-series, Tomb Raider, and Resident Evil, all which are similarly recognised for their cross-gender market appeal.

When we proceed to check the primary axis against the social characteristics of the players, we find a strong correlation with gender and frequency of playing: the titles on the left are mentioned more often by women than the titles on the right, and by players who play less frequently.

Nintendo games contribute significantly as a differentiating factor along this axis, indicating the popularity and relatively coherent style of its first-party game universe, as well as the strength of its brand, in particular (at the time of the study) the Wii platform.

Many of the titles on the left are also designed for social and offline party-style play, unlike the titles on the right side of the spectrum, which are primarily designed for solitary and/or online competitive play. This pattern seems to reflect a technological dimension associated with hobbyist gamer culture. The sophisticated 3D-graphics of game series like Bioshock, Gears of War or Call of Duty are designed to work with the cutting edge of consumer technology, and have an appeal typically associated with male tech culture. This distinguishes them from Nintendo, which prioritises gameplay and family-friendliness over graphics performance and high-tech appeal.

The pattern revealed by the primary dimension confirms the finding from Lucas et.al (2004) that female players tend to avoid games with "rotating" 3D worlds, as well as Juul's analysis (2005) of the non-intimidating and casual nature of 2D game spaces, in contrast to the exclusive appeal of advanced 3D game worlds. The distinctive dual-axis navigational interface of 3D-navigable games, using either two analogue sticks or mouse & keyboard, can indeed be off-putting to new players. The "tunnel vision" offered by a navigable virtual camera, which is particularly characteristic of First Person Shooters, is arguably more perceptually overwhelming and potentially intimidating than games played in two-dimensional space (Klevjer, 2012).

Super Mario games, and also The Sims, are nevertheless part of traditional and "core" gaming culture insofar as they have developed an "expressive and creative language of their own" (Juul, 2010, p. 11), far more so than Buzz or Singstar. Still, when compared to Fallout at the far right end of the axis, Nintendo games require less of the specialised knowledge and familiarity that is typical of hobbyist cultural practices. Even the more advanced Super Mario titles are relatively easy to pick up and play, and have a consistently friendly appeal in terms of interface, setting and style.

This finding is consistent with Juul's definition of the appeal of casual gaming. The casual attitude, he observes, is not necessarily related to brevity or lack of dedication, but a preference for games that are easy to learn (although they can be hard to master), conveniently interruptible, and, most importantly, have fictions that that are "emotionally positive" rather than violent, dark and intimidating (Juul, 2010, p. 50). According to our analysis, it is precisely this type of family-friendliness and positive aesthetic valence, including a de-emphasising of the role of tech appeal, which most consistently characterises the left side of our primary axis of videogame taste oppositions.

The second, vertical axis in Figure 1 reveals another familiar dimension of oppositions. This is most clearly articulated at the right end of the horizontal axis. Here the lower end of the vertical axis is dominated by role-playing games, and the upper end by sport and action-adventure. Also, whereas the lower end is dominated by fantasy and Science-Fiction of a particular flavour (e.g. Mass Effect, Dragon Age, Bioshock, Fallout, Zelda, Final Fantasy), such game fictions are all but absent from the top end, which is instead populated by distinctively mainstream and cross-media action fare: racing (Gran Turismo), shooting (Call of Duty, Crysis, Battlefield) and football (Championship Manager, Fifa, Pro Evolution Soccer). It appears that this vertical dimension differentiates between a niche-oriented or "geeky" taste at the lower end (although admittedly a large niche), versus a mainstream-oriented taste at the top.

The concept of niche in this context refers to the relative specificity and uniqueness of a game's thematic appeal and/or gameplay characteristics in comparison to other media, including traditional games. In this sense, The Legend of Zelda and Starcraft are more niche than Solitaire and FIFA.

We note that games in the top right corner of the dimensional map tend to be less complex in structure (football management games being an important exception), instead relying heavily on style and imagery from mainstream televised sport or Hollywood action films. The niche appeal games in the bottom right corner are generally more complex and strategically oriented, and also typically feature more elaborate mechanics of storytelling (Starcraft being the exception to rule here). In addition, the vertical dimension seems more consistently platform-agnostic than the primary dimension.

This interpretation does not conflict with the pattern we see on the inclusive left side of the horizontal axis. While the typically casual titles like Solitaire and Sudoku, which appear towards the upper left corner, are decidedly mainstream in terms of their appeal and imagery, Nintendo games in the bottom left are more distinct and elaborate in style and gameplay, and identifiable as a unique niche within popular entertainment culture.

If we compare to role-playing game series like Dragon Age, along the horizontal axis, we may categorise Nintendo as an inclusive niche, with lower barriers of entry, more family-friendly appeal, and less focus on technology. Correspondingly, if we compare to casual game series like Angry Birds in the top left corner, we may categorise CoD and FIFA as the specialist and exclusive variant of mainstream gaming, strongly focussed around a narrow set of style and gameplay preferences, and heavily coded in masculine entertainment culture.

Our analysis also produces a third and less dominant axis of preference differentiation (not shown). This axis mainly appears to bring out different preferences according to platform rather than inclusiveness (axis 1) or niche appeal (axis 2). Bioshock is here positioned close to GTA in the map. Both are console hits, positioned opposite series titles that are primarily associated with the PC platform (Battlefield, Diablo, War of Warcraft, Age of Empires). So the most likely interpretation is that what we are seeing along this third and less dominant axis is a PC versus console dimension of preference distribution.

Cluster Analysis

The cluster analysis suggests 7 clusters of gamers. The profile of their respective game preferences and some other characteristics are given in Tables 2a and 2b below. These clusters reflect the oppositions described above, while also unveiling new patterns and nuances.

Table 2a. Clusters of gamers. Favourite game series. N=557 (click to expand).

Table 2b. Clusters of gamers. Selected characteristics. N=557 (click to expand).  

Strategists cluster around a preference for The Sims, Civilization and other classic PC series that offer non-violent and slow strategic challenges, and are located close to the Origo on the map. The Sims is the title in our sample with the broadest appeal, across otherwise fairly distinct taste orientations. It is also the only title with an approximate "double footing" in two clusters, Strategists and Casuals.

Casuals and Nintendos emerge as distinct clusters, although each with a small base (10% and 7% of those who have mentioned titles). Casuals, located in the top left corner of the dimensional map, love Angry Birds. Nintendos, in the bottom left corner, are heavily dominated by Super Mario and The Legend of Zelda fans. The Sims as well as Super Mario are points of overlap between the two groups.

Partygamers, the smallest cluster in our sample, probably reflects the impact of the third axis, hardware platform. Partygamers have a preference for Guitar Hero, Buzz, and Singstar. At a higher level of analysis, this cluster is branching off from the Roleplayers cluster. The most likely interpretation is that these are players who own a videogame console, and who also use their consoles to play other, preferably fantasy-oriented, games.

The most clearly defined cluster in our sample, Lads, are located in the upper right corner of the dimensional map. They are differentiated from Casuals along the exclusive/inclusive dimension, but not along the niche/mainstream dimension. Lads are players who mention FIFA, GTA, and Call of Duty as their favourites.

In comparison to the other three large clusters — Strategists, Roleplayers, and E-sporters — Lads are much more focused in their preferences; titles outside their preferred orientation hardly get any mention at all. The relative rigidity of their preference for mainstream blockbuster titles may indicate that they are domain-focused rather than gameplay-focused in their preference orientation. The series Call of Duty, Football Manager, FIFA, and GTA are quite different from each other in terms of gameplay. It appears that what we have here is a group of players who love games because they love action and football, showing little interest in other kinds of games. At the same time, the games they prefer, especially CM/FM and Call of Duty, are also loved by many in the other clusters.

Roleplayers, in the lower right corner of the map, are characterised by a preference for role-playing games and other large and complex single-player story-based games. The titles they mention have a large spread, and there is a comparatively broad overlap with other clusters; a wide range of games from other groups' stock of favourites are mentioned, including Civilization, CM/FM, GTA, Starcraft, Super Mario, Half-Life, Zelda, CoD, and Gran Turismo. No other cluster has a comparable diversity of taste. The contrast to Lads is particularly striking: while a typical CoD- and GTA fan is highly unlikely to also enjoy Fallout, the reverse is not true.

E-sporters cluster around World of Warcraft, Starcraft, Counter Strike and other favourites indicating a preference for online and competitive gaming. Based on what we know about the subsequent rise of the so-called MOBA-genre (Massive Online Battle Arenas), we may reasonably assume that at the time of the study, a fair share of the WoW-mentioners in this cluster were of the competitive PvP kind, probably Battlegrounds players. If we were to conduct the survey today, we would very likely see a considerable contribution of League of Legends and DotA to this cluster, probably at the expense of WoW. We also note that there is a significant association with the Lads cluster, via CM/FM and CoD.

The cluster analysis also indicates that single-player versus online game formats is a significant differentiator of gaming dispositions. In our interpretation, E-sporters are strongly online-oriented. The titles they share with other clusters are strategy and football games, as well as Call of Duty and Half Life, both very likely referring to online competitive play. There is no mention among E-sporters of any Nintendos or Roleplayers favourites, which are all single player. The other clusters are mainly offline-oriented in their preferences, with the exception of Lads, who judging from the titles are more mixed and pragmatic in this respect. At the time of the survey, the Grand Theft Auto series was mainly a singleplayer experience. Although we see that CoD is a big favourite, Counter Strike is not mentioned, in stark contrast to the E-sporters.

Looking at the supplementary variables, Lads, Roleplayers, and E-sporters are high-frequency players, in contrast to the less frequent players in the Casuals, Strategists, Nintendos, and Partygamers. Lads are overwhelmingly male, with the percentage of female players at only 4%. Roleplayers and E-sporters have a significantly higher share of women, at 16 and 17 percent. The Casuals have a 69% female dominance, and the Nintendos, Strategists, and Partygamers are all around 50/50.

The supplementary data also show that there is a strong direct association between gender and gaming frequency, which operates independently of the association with a preference cluster. Within all clusters, except Casuals, we see that women play significantly less frequently than men [5].

There are some points of contact between our findings and the findings of player typology studies. General motivational categories like immersion, sociality, and competition (Yee, 2006) would clearly be unevenly distributed across our 7 clusters; we can imagine, for example, Roleplayers scoring high on immersion, and E-sporters being strongly oriented towards sociality and competition. The 3 game-related "mentalities" of play suggested by Kallio, Mäyrä and Kaipainen (2011) do not at all seem far-fetched in light of our data; "Gaming for Fun" appears to match our own E-sporters cluster quite well, "Immersive Play" seems to fit Roleplayers, and "Playing for Entertainment" would not be a bad fit for our the Nintendos or Strategists. The Lads cluster, however, would probably escape categorisations of this kind, as it is not defined in terms of gameplay-related characteristics.

With reference to Bateman et.al (2011) and Vermeulen et.al (2011), we may ask to what extent these patterns are linked to gaming competence. The relative complexity of Nintendos and Strategists' favourite titles, and the relative accessibility and broad appeal of titles like Call of Duty or FIFA, speak against mapping a preference for complex challenges onto our primary dimensional axis. Instead, looking at the series titles, a relative preference for complexity appears to play significantly into the vertical axis, in other words that it is linked to a preference for niche "gameness" over mainstream entertainment. E-sporters and Roleplayers prefer titles with highly complex and challenging game mechanics, primarily in the role-playing and real-time strategy genres. In comparison, the Lads' favourite titles are far less daunting to a newcomer in terms of their interface- and structural complexity. This pattern resonates with the idea that a preference for complexity has been the mark of a self-identifying gamer. However it cannot be mapped onto the primary distinction between exclusive and specialist versus inclusive game appeal.

Game Taste in a Broader Cultural Context

To what extent do game preferences resonate with other cultural preferences? As the survey data also give a lot of information about the respondents' cultural- and lifestyle preferences, we may identify some wider cultural characteristics of each cluster. This allows us to see players' gaming preferences in light of their orientations in other cultural domains, and may give us a fuller and more nuanced picture of the differences between types of gamers.

Our assumption is that respondents who are similar in terms of their gaming preferences will also tend to be similar in terms of their wider cultural preferences, including media use. We must however also expect that gender will predict cultural profiles and leanings independently of gaming preference orientations. In principle, we may find for example that Lads share a set of cultural orientations not so much because they are like-minded as gamers, but simply because they are male. We must also check to see if cultural preferences are different between men and women within the same preference cluster. Even if they have similar taste in games, they may have systematically different preferences in other areas.

Table 3. Clusters: selected cultural preferences. Percentage, by gender. N=557 (click to expand).

Table 4. Clusters: selected media use and lifestyle. By gender. N=557 (click to expand).

First, we note that title-mentioners as a whole do have a certain profile, when seen against the large group of non-players. They score higher in "social media" and "surfing", and stand out through their high rating of Fantasy/Sci-Fi, comics, and X-files. Even Lads and Casuals, who are least likely among game players to enjoy Sci-Fi and fantasy, nevertheless rate more than twice as high as non-players in these categories.

Female title-mentioners, however, have an even stronger profile in this respect, when compared against female non-players. They read more, and spend far more time online, especially women in the Roleplayers cluster. In terms of cultural orientation we see that the categories poetry, literary classics, and hard rock music rate markedly higher among female gamers, although among Casuals to a lesser extent.

Looking at the clusters individually, they all to some extent display their own unique media and lifestyle profile. Lads and Roleplayers appear more clearly defined than other clusters, especially when contrasted against each other.

Lads are consistently less interested than other clusters in traditional expressions of high culture like poetry, opera and modern jazz, even when only compared against other male gamers. They are also less fond of reading, more interested in popular culture products like Norwegian Idol and chart hits music, and — predictably given their game taste orientation — rate significantly lower in preference for SF/Fantasy literature.

The Lads' orientation towards mainstream forms and genres of culture is distinct also when measured against male non-gamers, who score consistently higher across traditional high-culture indicators. Lads also watch significantly more TV per week than male non-gamers, and almost twice as much as male Roleplayers — who, in contrast, rate 3 times higher in "read literature" every week, and score 15% on poetry versus 2% among Lads.

Judging by weekly reading, their preference for classics and poetry, and lack of interest for Idol and chart hits, Roleplayers emerge as a highbrow cluster within student culture, among gamers and non-gamers alike, and irrespective of gender. At the opposite end would be the Lads cluster, and then the other clusters falling somewhere in-between — including, for the sake of comparison, the large group of non-players.

E-sporters would fall relatively close to Lads according to traditional high/low cultural distinctions, scoring only moderately higher on key indicators. On the other hand, they are still less mainstream in their orientations, scoring very low on Idol and Chart hits, and high on SF/Fantasy, comics, and X-files. This cultural profile in between Lads and Roleplayers seems broadly in line with the game preferences data, where we saw that E-sporters are less opposed to mainstream blockbuster titles than Roleplayers.

The clusters Nintendos, Casuals, and Partygamers are all constructed from relatively small groups of mentioners, so any patterns of cultural orientation are highly uncertain and tentative. We see, for example, that there are unusually sharp gender divisions within the Nintendos cluster, and that in general, males here are far less enthusiastic about traditional high culture than their female counterparts. They also rate inexplicably low on "going out on the town" weekly, at 4%. Of possible interest for further study is also the fact that male Casuals have a significantly higher media consumption overall than any other group, which would make sense on the assumption that high media consumers also play games.

Gender differentiations

Lads apart, the most distinctive cultural profile is found among female Roleplayers. They read even more than their male counterparts in the cluster (and hence more than any other group), and they spend 13,3 hours weekly on social media, which is more than double that of the second highest rating group, the female Strategists, at 5.8 hours. Female Roleplayers do not go out much, and are more likely than anyone (including their cluster brethren) to enjoy X-files. So overall we see a kind of "Roleplayers plus" profile: more clearly literary, online-oriented and outside the broad mainstream. This profile is in line with Pierre Bourdieu's theory that minority members of a cultural field, going against society's expectations, identify more strongly with its values than majority members (Bourdieu and Passeron, 1979).

Note also that Strategists emerge from our analysis as a kind of neutral ground for the gaming field as a whole. They are equal in gender distribution, and less frequent players, yet apparently fond of complex challenges. They are close to the Origo on the dimensional map, and have broad overlaps with other clusters. In spite of being the second largest group in our sample, they have a comparatively low profile in public discourse, which is possibly due to their non-gendered, classics-oriented, low-tech, and primarily offline preference orientation.

Given that Roleplayers and E-sporters both have close to a 20% share of female players, women are strikingly absent from the Lads cluster. This most likely reflects the absence of girls and women from a certain type of commercial entertainment culture more broadly.

The relative strength of female Roleplayers and E-sporters in comparison to the lack of female Lads, and the high proportion of female Nintendos and Strategists, speaks to the non-gendered nature of the underlying differentiation between mainstream and niche videogame preference orientations. As noted above, it is also this vertical axis of the MCA dimensional map that primarily differentiates between relative simplicity and relative complexity; Angry Birds versus the Legend of Zelda, or Call of Duty versus Dragon Age. This finding supports Vermeulen et.al's conclusion (2011) that a preference for complex challenges cuts across genre and gender divisions.

Conclusion: differentiating the gamer

Our findings confirm that there is indeed a pattern and a logic to computer game preferences. At the aggregate level, we find clear patterns of association and opposition within people's stated favourite series titles. Overall, given the messy nature of actual game playing practices, the students' favourites cluster more clearly into preference orientation groups than we expected. Game preferences do also to a significant extent predict cultural- and media consumption more widely, indicating that videogame preference is situated in a broader field of aesthetic and cultural dispositions.

The primary dimension of taste differentiation is recognisable as an inclusive-exclusive axis, correlating with gender and frequency of playing. Graeme Kirkpatrick's historical articulation of gamer identity (2012) resonates with the specialist and exclusive end of our primary dimensional axis, defined by technologically sophisticated titles, and dominated by male frequent players. Our analysis also suggests that niche versus mainstream sensibility is a strong factor in predicting gaming preferences. Niche appeal means that there is a relative autonomy of the cultural disposition that enables appreciation of a particular type of game.

The way in which our analysis shows how the inclusive-exclusive- and mainstream-niche dimensions operate independently of each other in the formation of gamer preferences is a central finding in our study. Without a two-dimensional analysis, all preference orientations, including the clusters we have named Lads and Nintendos, would collapse into a simple core-vs-casual dichotomy.

Given its central role in videogame history, we may speculate that Nintendo has been key to the historical development of a particular flavour of videogame taste, which is reflected in our data. On the dimensional map, we see that the "Nintendo effect" is pulling in the same direction as the The Sims, which is similarly a phenomenon of its own, or at least it used to be.

Our study may serve to illustrate the point made by Crawford (2012) and Shaw (2010): gamer identities and gamer culture are a complex web of interrelations, entangled in broader media culture. To the extent that our data does point to patterns indicative of autonomous game culture, including a preference orientation towards complex challenges, such patterns cannot be mapped onto the primary — and highly gendered — dimension of exclusive versus inclusive appeal. Two of our game preference profiles, Roleplayers and E-sporters, appear to fit the bill of autonomous game culture. These profiles are picking up the impact of historically specific game cultures, as they have characteristically formed around role-playing, arcade game tournaments, and the rise of online E-sports culture. Their generally niche-oriented cultural taste in other areas supports this interpretation.

However, the contemporary "exclusive mainstream" of FIFA and Call of Duty lovers does not at all display a similar kind of relative autonomy, operating instead squarely at the intersection of specialist gaming culture and masculine mainstream entertainment culture. The cluster analysis, supported by the media- and cultural consumption data, indicates that Lads, which is undoubtedly the commercial backbone of the triple-A game industry, play games because they love action and sport, not the other way around. Hence their massive investment in mainstream media content, and their comparative indifference to niche- or high culture genres in other media.

At this point it is also worth observing that gamers in general do not score lower on media- and cultural consumption than others, and do not have a less active lifestyle. The male title-mentioners in our study actually spend more time on TV and other media than male non-gamers, in addition to their large investment in playing games. Nor is there any indication that gamers are less fond of reading or traditional high culture forms than non-gamers, with the exception of Lads.

Finally, our findings confirm that gender is a strong predictor of videogame preference orientations. As a whole, female students strongly lean towards games with a friendly and low-tech appeal over games with dark, violent and high-tech appeal. This confirms previous studies on the gendered nature of gaming as well as technology more broadly. In Cornelia Brunner's terminology, we can say that the games move from a "feminine sensibility" on the left to a "masculine sensibility" towards the right end of our dimensional map (Brunner, 2008).

On the other hand, the niche versus mainstream dimension of videogame preference, which also appears to be linked to gameplay complexity, is not gendered in nature according to our findings. As expected, there are significant clusters of female Roleplayers and E-sporters, who also have culturally distinct profiles across media. In contrast, the near complete lack of female players among self-declared FIFA- and GTA fans is striking in our data. As it turns out, the largest and less distinctly "gamey" segment within specialist game culture is also, by a substantial margin, the most male-dominated.

Endnotes

[1] The surveys were done in Bergen and Oslo, the two largest towns in Norway. The first study was done in late 2008, the second in 2010. Both surveys were conducted by regular mail and sent to a random stratified sample of the student population. The response rate was 55% in Bergen (n=1223) and 44% (n=1144) in Oslo, giving in total 2367 respondents.

[2] All series titles extracted for the final analysis are listed below, with original developer. Exceptions are "Chess", "Solitaire", and "Sudoku", which are generic titles that come in a wide range of digital variants.

[3] For more details on the methodical reasoning for this procedure, see Le Roux and Rouanet (2004).

[4] Percent of explained principal inertia axis 1-5: 3.1%, 2.9%, 2.7%, 2.6% and 2.5%. Series in italics are mentioned by only four respondents, whereas the others are mentioned by at least five.

[5] Frequency of playing at least five days a week (female/males): Strategists (10%/24%), Roleplayers (19%/42%), Partygamers (6%/40%), Nintendos (9%/33%), Casuals (19%/13%), Lads (33%/47%), E-sporters 20%/43%.

References

Bartle, R. (1996). Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs. Journal of MUD Research, 1(1).

Bateman, C., Lowenhaupt, R., & Nacke, L. (2011). Player Typology in Theory and Practice. Paper presented at the The 2011 DiGRA International Conference: Think Design Play.

Bourdieu, P. (1984). Distinction: a social critique of the judgement of taste. London: Routledge & Kegan Paul.

Bourdieu, P., & Passeron, J.-C. (1979). The Inheritors: French Students and Their Relation to Culture. Chicago: UCP.

Brunner, C. (2008). Games and Technological Desire: Another Decade. In Kafai, Y. B., Heeter, C., Denner, J., & Sun, J. Y. (Eds), Beyond Barbie and Mortal Kombat: new perspectives on gender and gaming (pp. 33-46). Cambridge, Mass.: MIT Press.

Corneliussen, H. G. (2014). Making the invisible become visible: Recognizing women's relationship with technology. International Journal of Gender, Science and Technology, 2(6).

Crawford, G. (2012). Video Gamers. London: Routledge.

Crawford, G., & Gosling, V. K. (2005). Toys for Boys? Women's Marginalization and Participation As Digital Gamers. Sociological Research Online, 10(1).

Gripsrud, J., Hovden, J.F., & Moe, H. (2011). Changing Relations. Class, Education and Cultural Capital. Poetics, 39(6), 507-529.

Ip, B., & Jacobs, G. (2005). "Segmentation of the games market using multivariate analysis." Journal of Targeting, Measurement and Analysis for Marketing, vol. 13 no. 3 (2005):275-287.

Jansz, J., Avis, C., & Vosmeer, M. (2010). Playing The Sims2: an exploration of gender differences in players' motivations and patterns of play. New Media & Society, 12(2), 235-251.

Juul, J. (2010). A casual revolution: reinventing video games and their players. Cambridge, Mass.: MIT Press.

Kallio, K. P., Mäyrä, F., & Kaipainen, K. (2011). At Least Nine Ways to Play: Approaching Gamer Mentalities. Games and Culture, 6(4), 327-353. doi: 10.1177/1555412010391089.

Kerr, A. (2006). The business and culture of digital games: gamework/gameplay. London: Sage Publications.

Kirkpatrick, G. (2012). Constitutive Tensions of Gaming's Field: UK gaming magazines and the formation of gaming culture 1981-1995. Game Studies, 12(1).

Klevjer, R. (2012). Enter the Avatar. The phenomenology of prosthetic telepresence in computer games. In H. Fossheim, T. Mandt Larsen & J. R. Sageng (Eds.), The Philosophy of Computer Games (pp. 17-38). London & New York: Springer.

Krotoski, A. (2004). Chicks and Joysticks: An Exploration of Women and Gaming. London: Entertainment and Leisure Software Publishers Association.

Le Roux, B., & Rouanet, H. (2004). Geometric Data Analysis. Dordrecht: Kluwer Academic Publishers.

Lucas, K., & Sherry, J. L. (2004). Sex differences in video game play: A communication-based explanation. Communication Research, 31(5), 499-523. DOI: 10.1177/0093650204267930.

Norwegian Media Authority (2016). Barn og Medier 2016. Oslo: Norwegian Media Authority.

Shaw, A. (2010). What Is Video Game Culture? Cultural Studies and Game Studies. Games and Culture, 5(4), 403-424. doi:10.1177/1555412009360414.

Turkle, S. (1988). Computational Reticence. Why Women Fear the Intimate Machine. I C. Kramarae (red.), Technology and women's voices. Keeping in touch (pp. 41-41). New York: Routledge & Kegan Paul.

Vermeulen, L., Van Looy, J., De Grove, F., & Curtois, C. (2011). You Are What You Play? A Quantitative Study into Game Design Preferences across Gender and their Interaction with Gaming Habits. Paper presented at the Think Design Play: The fifth international conference of the Digital Games Research Association (DIGRA), Hilversum.

Yee, N. (2006). Motivations for play in online games. Cyberpsychology & Behavior, 9(6), 772-775.

Ludography

2K Australia/2K Boston/Irrational Games. (2007-2014). Bioshock.

AcademySoft. (1986-2014). Tetris.

Bethesda Game Studios. (1994-2017). Elder Scrolls.

Big Blue Box/Lionhead Studios. (2004-2014). Fable.

Bioware/Snowblind Studios/Black Isle Studios/Overhaul Games. (1998-2016). Baldur's Gate.

Bioware/Obsidian Entertainment. (2002-2009). Neverwinter Nights.

Bioware. (2003-2004). Knights of the Old Republic.

Bioware. (2009-2015). Dragon Age.

Bioware. (2012-2017). Mass Effect.

Blizzard. (1994-2002). Warcraft.

Blizzard. (1996-2014). Diablo.

Blizzard. (1998-2015). Starcraft.

Blue Byte. (1993-2011). Settlers.

Bungie Studios/343 Industries. (2001-2017). Halo.

Capcom. (1996-2017). Resident Evil.

Core Design/Crystal Dynamics. (1996-2016). Tomb Raider.

Creative Assembly. (2000-2017). Total War.

Crytek. (2007-2013). Crysis.

Digital Illusions/EA Dice. (2002-2016). Battlefield.

DMA Design/Rockstar North. (1997-2014). Grand Theft Auto.

Domark Group/Sports Interactive. (1992-2017). Championship Manager/Football Manager.

Distinctive Software/Electronic Arts Canada. (1994-2017). Need for Speed.

Ensemble Studios. (1997-2016). Age of Empires.

Epic Games. (2006-2017). Gears of War.

Extended Play Productions/EA Sports. (1993-2017). FIFA.

Funcom/Red Thread Games. (1999-2014). Longest Journey.

Harmonix Music Systems/Neversoft/Free Style Games. (2005-2015). Guitar Hero.

Harmonix Music Systems/MTV Games. (2007-2015). Rock Band.

Infinity Ward/Treyarch/Sledgehammer Games. (2003-2017). Call of Duty.

Interplay/Bethesda. (1997-2017). Fallout.

Io-Interactive. (2000-2016). Hitman.

Ion Storm/Eidos Montréal. (2000-2017). Deus Ex.

Konami. (1998-2015). Metal Gear Solid.

Konami. (2001-2017). Pro Evolution Soccer.

Lucasfilm Games. (1990-2009). Monkey Island.

Maxis Software. (1989-2014). Sim City.

Maxis Software. (2000-2014). The Sims.

Maxis Software. (2008-2011). Spore.

Media Molecule. (2008-2015). Little Big Planet.

MPS Labs. (1990-2009). Tycoon.

MPS Labs/Activision/Firaxis. (1991-2017). Civilization.

Namco. (1994-2017). Tekken.

Naughty Dog. (2007-2017). Uncharted.

New World Computing/Nival Interactive/Black Hole Entertainment/Limbic Entertainment/Virtuos. (1995-2016). Heroes of Might and Magic.

Nintendo. (1985-2017). Super Mario platform games.

Nintendo. (1986-2017). The Legend of Zelda.

Nintendo. (1992-2017). Mario Kart.

Nintendo. (1998). The Legend of Zelda: Ocarina of Time. [Nintendo 64], UK: Nintendo.

Polyphony Digital. (1997-2017). Gran Turismo.

Relentless Software. (2005-2010). Buzz!.

Remedy Entertainment/Rockstar Studios. (2001-2012). Max Payne.

Rockstar San Diego. (2010-2011). Red Dead Redemption.

Rovio Mobile. (2009-2017). Angry Birds.

London Studio. (2004-2014). Singstar.

Square. (1987-2017). Final Fantasy.

Ubisoft. (2007-2017). Assassin's Creed.

Valve. (1998-2008). Half-Life.

Valve. (2000-2012). Counter Strike.

Valve. (2007-2011). Portal.

Westwood Studios/EA Los Angeles/EA Phenomic. (1995-2012). Command & Conquer.