Content structure is king : An empirical study on gratifications, game genres and content type on Twitch

Content structure is king: An empirical study on gratifications, game genres and content type on Twitch 1 Abstract: While video games have been widely investigated from the perspective of play, an emerging online media phenomenon is the spectating of video game play, captivating millions of users daily. This study investigates the relationship of video game genres, content type and viewer gratification in the context of live gaming. To study this phenomenon, we employ an online questionnaire study (N=1097) to investigate six categories of gratifications: affective, information seeking, learning to play, personal integrative, social integrative & tension release motivations and their relationship with game genres and stream types. The results of this study demonstrate that “the medium is the message”, highlight the importance of archetypal structure (i.e. the type of streamed content) over content topic (i.e. the genre of games being streamed), and help to build a better understanding of user generated content and the democratization of media. Keywords: Streaming, new media, genres, uses and gratifications, video games, social media, usergenerated content Authors: Sjöblom Max, Törhönen Maria, Hamari Juho & Macey Joseph This is the accepted manuscript of the article, which has been published in Computers in Human Behavior. 2017, 73, 161-171. http://dx.doi.org/10.1016/j.chb.2017.03.036 2 Content structure is king: An empirical study on gratifications, game genres and content type on Twitch The effects and gratifications from playing video games have been a widely investigated and debated topic within the last decade both in academic literature, e.g. in media psychology, game research and communication studies (Elson & Ferguson, 2014; Hamari & Keronen, 2017; Mäyrä et al., 2016; Quandt et al., 2015), and in popular discourse. While this debate is still ongoing, a new yet uninvestigated form of game consumption has emerged: watching others play games via YouTube and live broadcasts on services such as Twitch. Today, millions of people watch others play games on the internet (Needleman, 2015; Twitch, 2016). Services such as YouTube have spearheaded a major shift in the media landscape, moving production of audiovisual media from large corporations and organizations towards smaller entities and individuals (Cha, Kwak, Rodriguez, Ahn, & Moon, 2007; Sjöblom & Hamari, 2017). The democratized process of content creation on video game streaming platforms such as Twitch allows for the existence of many types of content. In this context video game –related video content such as “Let’s plays” and eSports (Hamari & Sjöblom, 2017; Newzoo, 2016; Taylor, 2012; Twitch, 2016), have become especially popular. The participatory and interactive nature of this emerging form of media serves to bridge the divide between games and traditional media, such as television, via the convergence of interactive, communal and passive forms of media. This evolution of the media landscape towards user-generated content also brings into question the legitimacy of genre as the primary means of classifying media content. In the realm of video game streaming particularly, it becomes apparent that genre might not constitute the defining means of classification. As little is known about the gratifications obtained by watching these online video streams, this study aims not only to investigate the general gratifications that people derive from watching online streaming content, but also the differences in various streaming content. Not all game streaming content is similar; instead it is highly varied, ranging from very competitive endeavors to 3 highly casual ones. This variation in content type affords an interesting angle of research, and can be compared to various types of programming produced for other broadcast media such as television. Few studies have yet aimed to approach this topic quantifiably, thus far the literature on video game streaming has focused mainly on communities (Hamilton, Garretson, & Kerne, 2014), technical aspects (Kaytoue, Silva, Cerf, Meira, & Raïssi, 2012) and competitive eSports (Hamari & Sjöblom, 2017; Weiss & Schiele, 2013). Thus, the primary research question of this study is: RQ: How are gratifications from viewing (as operationalized by affective, information seeking, learning to play, personal integrative, social integrative & tension release motivations) online game streams associated with game genres of streamed games and types of game streams? Background Video game streams organize and present their varied content through the use of genre labels, much like any other media (Chandler, 1997). The use of genre as a means of categorising video games is more than simply common practice, it is almost unavoidable no matter if the discussion be academic, commercial or popular (Arsenault, 2009). This situation is not simply unique to video games and to the practice of Games Studies, as any object-bound discipline must pay close attention to the tools that it employs to describe the particular focus of study. As such, while the notion of genre as a tool to aid categorisation is rarely questioned, the particular genre labels that have been employed have been subject to prolonged scrutiny. This also applies to the fields of literature (Frye, 1957; Williams, 2006) and film (Altman, 1984; Stam & Miller, 1999) as it does for video games (Clearwater, 2011; Wolf, 2001). Differing taxonomies have been created which include a range of factors as points of reference and can include: period, country of origin, franchise, distribution channel/platform, target audience, or content (Chandler, 1997). Added to this are the technological and interactive possibilities of video games, or as Aarseth phrases it, the fact that video games are essentially 4 simulations in which players create experience through their individual actions (Aarseth, 2004). For some, therefore, it is the nature of a player’s interaction with the game that is the primary characteristic of interest (such as First-Person Shooter), for others it may be the content (Puzzle), or even the business model (Free-to-Play) of the game. The academic quest for a definitive means of classifying video games, is at odds with the way in which genre is both conceived and used in wider society as well as with a more nuanced investigation of cultural artefacts. Whilst there is a constant evolution of what is commonly understood to constitute any particular genre (Arsenault, 2009; Dor, 2014), the labels and terms themselves seem to be relatively constant and durable (Faisal & Peltoniemi, 2015). The perspective offered by “social linguistics” (Gee, 2015) is, therefore, valuable in understanding genre labels as context-specific discourses that are driven by distinct communities. These labels are used to convey messages about the content or style of media products, as such they can be presumed to have a significant impact on the expectations of the consumers and, therefore, on the gratifications users derive from watching streams. The contemporary proliferation of media formats such as video streaming and on-demand television suggest that the individual needs and preferences of users are becoming more important when selecting media content. This is in contrast to the structural characteristics of traditional media that have previously shaped consumption habits (Giddens, 1984). When considering the increased choice available to media consumers, user “repertoires” have been found to be a reaction to ever-increasing options; a way of managing choice in an environment of over-abundance (Heeter, 1985). Applying this perspective to the consumption of multiple media, Reagan (1996) found that consumers utilise different media in order to access specific topics of interest. However, subsequent investigation revealed that user repertoires were not simply defined according to either medium or genre, but by a mixture of the two (van Rees & van Eijck, 2003). The medium of video games is one in which genre labels are complicated by the need 5 to reflect its inherently interactive nature (Wolf, 2001). Therefore, further investigation is required in order to understand the network of relationships between a consumer, the gratifications that drive consumption, the genre of the consumed product, and the medium by which it is accessed. Such work would serve to begin the process of integrating video game streaming into the canon of existing media forms, in addition to testing the relevance of current approaches for interpreting this new phenomenon. Understanding the reasons for media consumption are a primary focus within communication and media sciences. One of the leading perspectives in this area is the uses and gratifications (UG) theoretical approach (Katz, Blumler, & Gurevitch, 1973; Katz, Gurevitch, & Haas, 1973; Ruggiero, 2000). UG has been used in a wide range of communication research contexts, such as television (Krcmar & Greene, 1999; Schmitt, Woolf, & Anderson, 2003), personal communication (Ishii, 2006), and the practice of multitasking (Wang & Tchernev, 2012). Another area where UG has been heavily used is the online context (Papacharissi & Mendelson, 2010; Whiting & Williams, 2013), including online games (Hamari & Sjöblom, 2017; Sherry, Lucas, Greenberg, & Lachlan, 2006) and video streaming (Cha, 2014; Chiang & Hsiao, 2014; Sjöblom & Hamari, 2017). General media studies have also had an interesting in mapping the effect that various genres and content type has on consumption. Research on motivations to watch TV have been categorised into two groups, where relaxation and enjoyment motivate the more ritualized viewers and, learning and information gathering the more instrumental viewers (Rubin, 1981). The different viewing habits of consumers are highlighted in specific genres such as sports (Gantz, 1981; Raney, 2006; Westmyer, DiCioccio, & Rubin, 1998), news (Rubin & Perse, 1987) and reality TV shows (Ebersole & Woods, 2007). Consumption of film and television in general has been concluded to be highly motivated by personal gratifications such as enjoyment and entertainment (Ferguson & Perse, 2000). Similar trends have been discovered in digital media spectating habits, where consumption of media content 6 on channels like YouTube has been found to be motivated by entertainment, information seeking and learning (Cha, 2014; Hanson, Haridakis 2008; Shao, 2009). In addition to these motivators, social media consumption, such as use of Facebook, was found to be motivated by social incentives (Chen, 2011; Pai & Arnott, 2013; Papacharissi & Mendelson 2011, Whiting and Williams 2013,). In this study, we will use the UG approach to investigate motives for watching video game streams, which types of streaming content may afford which kinds of gratifications, and the gratifications that viewers derive from watching video game streams. In particular, we will focus on the relationship of video game genres, stream types and viewer gratification in the live gaming content. To study this phenomenon, we employ an online questionnaire survey (N=1097) to investigate six categories of gratifications: affective, information seeking, learning to play, personal integrative, social integrative and tension release, and their relationship with game genres and stream types. Based on previous research within media, games and video game streaming, along with the expertise of the researchers, we offer three expectations regarding the results of this study. Firstly, we expect different types of gratifications to be catered for by different stream types. For example, competitive and speedrun streams will both afford affective, learning and tension release gratifications while competitive streams will additionally afford information gathering and social integrative gratifications. Secondly, different genre types will similarly afford different gratifications, for example slower-paced genres such as sandbox games will afford more gratifications relating to social interaction than faster-paced genres such as FPS games. Finally, we expect that as an individual game can be featured in virtually any stream type, it follows that it is the stream type which is the most significant factor that affords the individual user’s needs to be gratified. Data & methods 7 Sampling The survey was initially piloted with 19 respondents, and after no major concerns arose, the primary survey was launched on the 26th of February 2015. The survey was online for approximately four weeks, ending on the 23rd of March. In order to incentivize participation, a prize draw of six video games from the online store Steam, each worth $50, was offered. The majority of respondents were from the online news and social networking sites Reddit, Twitter and Facebook, with additional respondents from individual online game forums. A filter question was included in order to identify and remove invalid responses, resulting in 1091 valid responses. The amount of invalid responses comprised a 3.2% decrease in dataset size. The average age of respondents was 22.9 years (M = 22, SD = 5.9). The data showed a strong bias towards male respondents, with female respondents comprising 7.7% of the data, in line with thirdparty estimates of gender distribution among video game stream users (Quantcast, 2016). Of our respondents, 93.2% reported that they had registered an account for the Twitch service, had used the service for an average of 22.1 months (M = 21, SD = 14.6) and 38.7% had streamed at some point. The demographic details of the respondents are displayed in greater detail in table 1. [Table 1 Demographic distribution of survey, here] Measurement Existing UG research has identified five high-level categories of needs: cognitive, affective, personal integrative, social integrative and tension release (Katz et al., 1973; West & Turner, 2010). We identified both learning to play games, as well as information seeking on products, to be two important cognitive motivations, and chose to measure them separately. The scales used in this survey all comprised of items that used a 7-point Likert scale (1 indicating “strongly disagree” and 7 indicating “strongly agree”). 8 To measure affective motivations, the perceived enjoyment scale of Venkatesh (2000) and van der Heijden (2004) was used. For information seeking on products, the usefulness scale from van der Heijden (2004) was used and modified accordingly. For the learning to play scale, items from the information seeking scale by Papacharissi & Rubin (2000) were used, alongside an item taken from the van der Heijden (2004) usefulness scale. These items were modified to fit the context of video game streaming and learning strategies. For personal integrative motivations, we used the recognition by peers scale from Hernandez et al. (2011). Within social integrative motivations, the companionship scale introduced by Smock et al. (2011), and the shared emotional connection scale used by Chavis et al. (2008) were used. Relating to tension release, the scales of escapism, relaxing entertainment, and habitual pass time previously introduced by Smock et al. (2011) were used. We measured the consumption of different video game genre streams using a frequency scale with values 1-5 indicating how frequently the person watched said genre or type (never, once a year, once a month, once a week, daily). The most fundamental issue that determines the way in which genre is conceived depends on the uses to which it will be put (Chandler, 1997; Elverdam & Aarseth, 2007), as such this study utilised genre definitions that are the product of established use in wider society and in research. The approach to the selection of game genres was, therefore, twofold. Firstly, we adapted a commonly used classification of video game genres (Lee, Karlova, Clarke, Thornton, & Perti, 2014). Secondly, we observed the 50 most popular games on the Twitch video game streaming service. We then cross-referenced these findings, resulting in 11 genres. Examples of games were provided alongside the genres in order to aid the survey respondents. The game genres used were: action, collectible card games (CCG), fighting, first-person shooter (FPS), massively multiplayer online (MMO), multiplayer online battle arena (MOBA), rhythm, roleplaying game (RPG), real-time strategy (RTS), sandbox and, lastly, sports. Types of video game streams 9 In addition to genre categories, we also classified types of streams according to the structure and content style of the stream. These adhered to the same frequency scales with values 1-5 as mentioned previously. The types of streams were: casual, let’s play, competitive, how to play, review, speedrun and talk show. The stream type classifications were based on a systematic review of Twitch content in which the defining characteristics of individual formats were identified. While there is no guarantee that the classification is exhaustive, there are no clear precedents to classifying streaming content in this form. It is worth noting that although Twitch presently supports the streaming of activities not related to video games, at the time the survey was conducted, Twitch had a strict policy allowing only for video game content. [Table 2 Video game stream types, here] Validity & reliability The analyses were undertaken in SmartPLS 3.2.3 which supports component-based structural equation modelling which is the de facto method when the model includes both psychometric, latent constructs and formative constructs (Ringle, Wende & Will, 2005). Convergent validity was investigated through the AVE, CR and Alpha measures. As these were all above the recommended thresholds, we can conclude that convergent validity was met (Fornell & Larcker, 1981; Nunnally, 1978). We found discriminant validity to be met, as the square root of the AVE for each psychometric construct exceeded the correlation for any other construct, and each measurement item had the highest loading with its corresponding construct (Chin, 1988; Fornell & Larcker, 1981; Jöreskog & Sörbom, 1996). The results of these validations are presented in Appendix A. Results & Discussion We investigated how the consumption of various video game genres and stream types are associated with affective, tension release, information seeking, learning to play, personal integrative 16 gratifications, an implication which is borne out by the fact that RTS games have a negative influence on personal integrative gratifications. This negative relationship highlights the fact that the size of a stream is not the only important factor, but that the game genre also needs to support discussion in order to foster the level of communication required to feel gratified by receiving recognition. The fact that RTS games are both information dense and require a high level of existing knowledge in order to play effectively can lead to the situation where no fruitful input can be given for the streamer. We can see, therefore, how particular stream types and game genres serve to gratify specific needs of users. The results for social integrative motivations show an interesting facet of genres compared to the other motivational types. Popular eSports genres do not show associations with social motivations and instead we see two genres, that emphasize the community aspect, influencing social gratifications: rhythm and sandbox games. The fact is, that even though video game streaming as a whole can be seen as a social media, certain genres seem to naturally gravitate toward a user equilibrium more fruitful for social interaction. Previous research within streaming has exposed the fact that spectators experience smaller streams as being more conducive to social interaction (Hamilton et. al, 2014). These genres commonly fit into the casual type, which also showed a strong positive influence on social gratifications. While these are examples of small and possibly tight-knit communities serving as facilitators of gratification, competitive streams also influence social gratifications, and they are often large communities. We argue that the sense of community commonly experienced within these types of streams is a way of identifying as an eSports consumer or follower of a particular eSports title, rather than identifying with a particular stream or streamer, as can be the case in the less popular games. When inspecting the difference between Model 1 and Model 3, we can see this quite clearly for the FPS and MOBA genres which become less significant with the introduction of stream types, with most of the influence likely being accounted for by competitive streams. Social aspects have been found to impact the playing 17 of these types of competitive games (Jansz & Tanis 2007; Sherry et al., 2006; Williams, Yee, & Caplan, 2008; Yee, 2006), and as the current study indicates, are of significance for the spectating side of the spectrum. The findings are in line with previous work relating to social integrative motivations and streaming (Hamilton et al., 2014; Sjöblom & Hamari, 2017), eSports (Scholz, 2012) and social media (Papacharissi & Mendelson, 2010; Whiting & Williams, 2013). Theoretical implications The results of this study have interesting implications for future studies within user generated content, online media and video streaming. While this study focused on video game genres in the streaming context, we believe valuable insight can be gained not only pertaining to game genres, but also in relation to genres among media as a whole. The revelation in our results that the archetypal structure (i.e. the type of streamed content) of media has a stronger impact than the content (i.e. the genre of games being streamed) of said media is interesting both in the context of broadcasting but also for social video sharing through services such as YouTube, Facebook, Instagram and Twitter. McLuhan (1964) famously coined the term the medium is the message, and as we study an emerging form of new media, it is highly relevant to consider the role of medium and message in the context of video game streaming. In this study we looked at both game genres and stream types as predictors of motivation types. Both these could be considered as part of the message of the larger medium of the internet, indeed McLuhan argues that the content of every medium is another medium. However, we argue that stream type transcends the level of pure message, as these types represent archetypes of content, comprised by a higher level of structure. Genres, on the other hand, can be considered closer to being the message, even though they also play a role in facilitating interactions between streamers and spectators. The results of this study indicate that, in fact, the existence of video game streaming as a medium serves to shape the message, in this case the behavior and trends of the people using it. When considering the laws of media related to the 18 message is the medium (McLuhan & McLuhan, 1988), the tetrad, we see the implications that video game streaming enhances niche communities and retrieves an act of gameplay involving social interaction that was in decline. Additionally, it serves to obsolete some previous mediums as primary channels for game marketing, while also reversing into a phenomenon in and of itself. One where users record videos of particular events that take place during live streams, subsequently making these highlights available of video sharing sites like YouTube. Furthermore, McLuhan’s position highlights genre, and its inherent usefulness, as being socially-defined and subject to constant revision (Chandler, 1997; Elverdam & Aarseth, 2007). We can see that for some users, those seeking social and personal integrative gratifications, genre labels are useful tools as they highlight any inherent interactive qualities and characteristics. For others, who seek gratifications related to tension release, it may be more productive to consider genre labels that communicate the theme of a game being streamed. Finally, certain gratifications such as information seeking, learning to play, and tension release may be linked to specific game titles, rather than any video game genre categories currently used. Such results highlight why game genre labels are a hybrid of different qualities and of the differing levels of abstraction provided by concepts of platform, genre, mode and milieu (King & Krzywinska, 2002; Järvinen, 2003). The fact is that different users seek differing forms of gratification which, in turn, require distinct information in order to effectively communicate a game’s qualities to the consumers. In relation to wider media, it is apparent that the declining influence of traditional structural characteristics (Giddens, 1984) suggests that established genre labels require adaptation. To some extent this already occurs as many genres are modified in order to reflect specific sub-genres (Chandler, 1997). However, it may be beneficial to utilise labels which stress elements unconnected to narrative or theme, ones which are instead based on sociallyemergent labels reflecting presentation style such as the “unboxing” videos popular on YouTube. 19 The emergence of video game streaming is a significant sign that the importance of user generated content and the democratization of the media ecosystem continues apace. While many technological developments have been necessary to make video game streaming a reality, the democratization of media is not just a paradigm shift within a narrow media landscape. Rather, it reflects a greater development in the democratization of production and economic coordination driven by recent IT developments, such as sharing economies (Hamari, Sjöklint, & Ukkonen, 2016; Sundararajan, 2016), gig economy/microwork (Irani, 2015), crowdsourcing (Howe, 2006; Morschheuser, Hamari, & Koivisto, 2016) and crowdfunding (Belleflamme & Lambert, 2014; Mollick, 2014). The results of the current study, particularly in relation to information acquisition, indicate that consumers turn to streamers as an important source of information, analogous to other peer-produced informational content, e.g. Wikipedia (Shao, 2009). This trend is also highly visible on services such as YouTube, where a multitude of videos aimed at disseminating information are available on various topics (Hanson & Haridakis, 2008; Haridakis & Hanson, 2009). When considering the phenomenon of video game streaming as a media form, we see the evolution of an interesting facet of sociability within the context of games. In the early days of the evolution of video games, in arcades spectators would gather around the person playing a game, as home consoles were not yet widespread (Newman, 2004). Later, LAN-parties facilitated the faceto-face interaction that many people had experienced while sitting on a couch with friends, playing games (Jansz & Martens, 2005). As games moved further to the online sphere, face-to-face social interaction was, in many cases, lost. Video game streaming has brought back some of the social interaction that had been absent from the domain of video games. Therefore, we are not only witnessing a convergence of interactive and “passive” media but also an oscillating development curve where aspects of the media experience come and go depending on the platforms on which they are consumed. 20 Not only does streaming offer interesting insights into the sociability of games, it also makes the important observation that particular games and genres might offer contrasting gratifications when played and when spectated. This makes games a highly interesting form of media, as the duality of spectating and play separates it markedly from television and movies. For example, competitive games might cause a high level of stress when played, as winning or losing play a large role in the game experience. This, coupled with the toxic attitude that many players display when a teammate has made what they perceive to be an error (Kwak, Blackburn, & Han, 2015), may make some players shy away from the games. However, when spectated, the same level of stress may not arise as you are not subject to the same expectations that the player is. The results of this study indicated positive associations for many competitive genres with tension release motives, suggesting that it may be a gratification that shows differences between play and spectating. Practical implications For the majority of motivations, stream types were shown to have the highest associations with gratification, displaying important implications for both service designers, streamers and game designers. The importance of content structure as a chief facilitator argues for designers of stream services to perhaps rethink the way many consumers discover new content. Currently, a majority of the explorative functions available on streaming services highlight the game being played. However, it may be more fruitful to facilitate alternative means of exploration through the type of content being broadcast, rather than the topic. This is not an easy task to tackle, but perhaps utilizing crowdsourced categorical tagging of content dimensions could prove a valuable tool. Additionally, eSports organizations and broadcasters should think about the relation of their content to personal integrative motivations, as while competitive streams showed positive associations for other motivations, personal integrative motivations were missing. Regarding individual genres, the strong association between action games and information seeking motivations is something we feel is of note. While our research merely highlights the need of information 21 seeking among action spectators, and does not directly link this to purchasing, previous studies have shown the impact that video game streams have had on game sales (Hernandez, 2016). The findings also provide game developers with a better perspective on the spectating habits and motivations of their target audiences. Although the streaming and spectating culture naturally calls for cinematic game experiences, there are other important factors for game developers to consider. The streaming and spectating culture is heavily social and interactive, and can be utilized in game development by offering ways to better integrate streaming activities within the game though user-generated content sharing services, chats and integrated tournament structures. Usergenerated game content can also be a useful promotional and information tool for game developers, which can be utilised and embraced within the game development process. Limitations and future research As the dataset used in this study was collected through an online survey, where respondents are self-selected and data self-reported, this naturally poses some limitations for the study. Selfselected respondents may be more active users of streaming service than the average user, and hence the results are skewed towards the highly active population of service users. Furthermore, even though the sample size is adequate in all statistical senses, and also large in the context of streaming research, it still sets some constraints. Many of the genres represented in this study are not very popular, and hence quite a small percentage of respondents reported watching them. Should we wish to paint a more in-depth picture of the smaller genres, such as rhythm or sports games, a separate study should be aimed specifically at spectators of these genres. It should also be noted that our interpretation of the genres represented in this research may be subject to discussion, and that respondents may potentially interpret these genres differently. It may have been beneficial, therefore, to collect additional information about individual game titles. We chose to employ the UG approach to studying the subject of video game streaming genres and stream types, as we considered it to be abstract enough to build an informed image of 22 the phenomenon. This was particularly apt as there has not yet been much research on the subject, however, we do acknowledge the value of having scales developed specifically for this context. Hence one limitation of this study may be that the six motivations studied here do not cover all the motivations for spectating. As mentioned, the importance of stream types is a phenomenon that might be worth studying in additional contexts. For example, on YouTube, videos are commonly classified and found by what the core of the content is about, rather than the overall mood or structure of the video. Even though the specifics might be different, it would be highly interesting to see how the types of content used in this study would impact viewership within the context of YouTube. 23 References Aarseth, E. (2004). Genre trouble. Electronic book review, 3. Altman, R. (1984). A Semantic/Syntactic Approach To Film Genre. Cinema Journal, 23(3), 6-18. Apperley, T. H. (2006). Genre and game studies: Toward a critical approach to video game genres. Simulation & Gaming, 37(1), 6-23. Arsenault, D. (2009). Video game genre, evolution and innovation. Eludamos. Journal for Computer Game Culture, 3(2), 149-176. Beentjes, J. W., Koolstra, C. M., & van der Voort, T. H. (1996). Combining background media with doing homework: Incidence of background media use and perceived effects. Communication Education, 45(1), 59-72. Belleflamme, P., Lambert, T., & Schwienbacher, A. (2014). Crowdfunding: Tapping the right crowd. Journal of Business Venturing, 29(5), 585-609. Cha, J. (2014). Usage of video sharing websites: Drivers and barriers. Telematics and Informatics, 31, 16-26. Cha, M., Kwak, H., Rodriguez, P., Ahn, Y. Y., & Moon, S. (2007). I tube, you tube, everybody tubes: analyzing the world's largest user generated content video system. In Proceedings of the 7th ACM SIGCOMM conference on Internet measurement (pp. 1-14). ACM. Chavis, D. M., Lee, K. S., & Acosta, J. D. (2008). The sense of community (SCI) revised: The reliability and validity of the SCI-2. 2nd international community psychology conference. Lisbon. Chandler, D. (1997). An introduction to genre theory. The Media and Communications Studies Site. Chen, G. M. (2011). Tweet this: A uses and gratifications perspective on how active Twitter use gratifies a need to connect with others. Computers in Human Behavior, 27, 755-762. Cheung, G. and Huang, J. (2011). Starcraft from the stands: Understanding the game spectator. In Proceedings of the 2011 Annual Conference on Human Factors in computing systems (CHI’11). ACM, New York, NY, USA, 763-772. Chiang, H.-S., & Hsiao, K.-L. (2014). YouTube stickiness: the needs, personal, and environmental perspective. Internet Research, 25(1), 85-106. Chin, W. W. (1988). The partial least squares approach for structural equation modeling. In G. A. Marcoulides, Modern Methods for Business Research (pp. 295-336). Lawrence Erlbaum Associates. Clearwater, D. (2011). What defines video game genre? Thinking about genre study after the great divide. Loading..., 5(8). 24 Courtois, C., Mechant, P., De Marez, L., & Verleye, G. (2009). Gratifications and Seeding Behavior of Online Adolescents. Journal of Computer-Mediated Communication, 15, 109137. Dor, S. (2014). A history of real-time strategy gameplay from decryption to prediction: Introducing the actional statement. In History of Games International Conference Proceedings (Therrien, C., Lowood, H. & Picard, M., eds.). Montreal, Canada: Kinephanos. Ebersole, S. & Woods, R. (2007). Motivations for viewing reality television: A uses and gratifications analysis. Southwestern Mass Communication Journal, 23(1), 23-42. Elson, M., & Ferguson, C. J. (2014). Twenty-five years of research on violence in digital games and aggression. European Psychologist. Elverdam, C., & Aarseth, E. (2007). Game classification and game design construction through critical analysis. Games and Culture, 2(1), 3-22. Faisal, A., & Peltoniemi, M. (2015). Establishing video game genres using data-driven modeling and product databases. Games and Culture. Ferguson, D. A., & Perse, E. M. (2000). The World Wide Web as a functional alternative to television. Journal of Broadcasting & Electronic Media, 4 (2), 155-174. Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39-50. Frye, N. (2015). Anatomy of criticism. Princeton, NJ: Princeton University Press. Gantz, W. (1981). An exploration of viewing motives and behaviors associated with television sports. Journal of Broadcasting, 12, 263–275. Gee, J. (2015). Social linguistics and literacies: Ideology in discourses. London: Routledge. Giddens, A. (1984). The constitution of society: Outline of the theory of structuration. Univ of California Press. Hamari, J., & Keronen, L. (2017). Why do people play games? A Meta-Analysis. International Journal of Information Management, 37(3), 125-141. DOI: 10.1016/j.ijinfomgt.2017.01.006 Hamari, J., & Sjöblom, M. (2017). What is eSports and why do people watch it? Internet Research, 27(2). DOI: 10.1108/IntR-04-2016-0085. Hamari, J., Sjöklint, M., & Ukkonen, A. (2016). The sharing economy: Why people participate in collaborative consumption. Journal of the Association for Information Science and Technology, 67(9), 2047-2059. Hamilton, W. A., Garretson, O., & Kerne, A. (2014). Streaming on twitch: fostering participatory communities of play within live mixed media. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 1315-1324). ACM. 25 Hanson, G., & Haridakis, P. (2008). YouTube Users Watching and Sharing the News: A Uses and Gratifications Approach. The Journal of Electronic Publishing, 11(3). Haridakis, P., & Hanson, G. (2009). Social interaction and co-viewing with YouTube: Blending mass communication reception and social connection. Journal of Broadcasting & Electronic Media, 53(2), 317-335. Heeter, C. (1985). Program selection with abundance of choice. Human Communication Research, 12(1), 126-152. Hernandez, B., Montaner, T., Sese, F. J., & Urquizu, P. (2011). The role of social motivations in elearning: How do they affect usage and success of ICT interactive tools? Computers in Human Behavior, 6, 2224-2232. Hernandez, D. (2016, July 13th). Game Creator Success on Twitch: Hard Numbers. Retrieved October 6th 2016, from Twitch: https://blog.twitch.tv/https-blog-twitch-tv-game-creatorsuccess-on-twitch-hard-numbers-688154815817. Howe, J. (2006). The rise of crowdsourcing. Wired magazine, 14(6), 1-4. Irani, L. (2015) The Cultural Work of Microwork. New Media & Society, 17(5), 720-739. Ishii, K. (2006). Implications of Mobility: The Uses of Personal Communication Media in Everyday Life. Journal of Communication, 56(2), 346-365. Jansz, J., & Martens, L. (2005). Gaming at a LAN event: the social context of playing video games. New Media & Society, 7(3), 333–355. Jansz, J., & Tanis, M. (2007). Appeal of playing online first person shooter games. CyberPsychology & Behavior, 10(1), 133-136. Järvinen, A. (2003). Games without frontiers. Unpublished Ph. D. manuscript. Jöreskog, K. G., & Sörbom, D. (1996). LISREL 8: User's reference guide. Scientific Software International. Katz, E., Blumler, J. G., & Gurevitch, M. (1973). Uses and Gratifications Research. The Public Opinion Quarterly, 37(4), 509-523. Katz, E., Gurevitch, M., & Haas, H. (1973). On the Use of the Mass Media for Important Things. American Sociological Review, 38(2), 164-181. Kaytoue, M., Silva, A., Cerf, L., Meira, W. J., & Raïssi, C. (2012). Watch me Playing, I am a Professional: a First Study on Video Game Live Streaming. WWW '12 Companion Proceedings of the 21st international conference companion on World Wide Web (pp. 11811188). New York: ACM. King, G., & Krzywinska, T. (2002). Screenplay: cinema/videogames/interfaces. Wallflower Press. 32 Table 3 Regression analysis for affective and tension release motivations Model 1 Model 2 Model 3 Affective Tension release Affective Tension release Affective Tension release R2 0.160 0.178 0.221 0.252 0.248 0.273 β p β p β p β p β p β p Action 0.125** 0.000 0.152** 0.000 0.077* 0.016 0.071* 0.031 CCG 0.154** 0.000 0.166** 0.000 0.081* 0.008 0.097** 0.001 Fighting 0.022 0.472 0.031 0.340 -0.010 0.750 -0.004 0.911 FPS 0.206** 0.000 0.206** 0.000 0.061* 0.046 0.080* 0.010 MMO 0.036 0.219 0.010 0.733 0.015 0.610 -0.012 0.690 MOBA 0.177** 0.000 0.227** 0.000 0.037 0.211 0.110** 0.000 Rhythm 0.039 0.159 0.045 0.094 0.011 0.665 0.005 0.858 RPG 0.047 0.125 0.003 0.919 0.068* 0.027 0.006 0.860 RTS -0.069* 0.028 -0.054 0.087 -0.063* 0.048 -0.054 0.083 Sandbox 0.095** 0.001 0.083* 0.004 0.070* 0.021 0.030 0.311 Sports -0.014 0.673 -0.004 0.892 -0.001 0.972 0.004 0.893 Casual 0.279** 0.000 0.293** 0.000 0.232** 0.000 0.245** 0.000 Competitive 0.297** 0.000 0.262** 0.000 0.276** 0.000 0.204** 0.000 How to play 0.024 0.391 0.018 0.506 0.010 0.716 -0.002 0.941 Let's play 0.088* 0.003 0.153** 0.000 0.038 0.227 0.128** 0.000 Review 0.016 0.592 -0.041 0.150 -0.005 0.855 -0.047 0.109 Speedrun 0.061* 0.030 0.079* 0.005 0.026 0.423 0.060 0.053 Talkshow 0.046 0.086 0.080* 0.004 0.029 0.291 0.068* 0.018 Note: * = p < 0.05, ** = p < 0.01 33 Table 4 Regression analysis for information seeking and learning to play motives Model 1 Model 2 Model 3 Information Learning Information Learning Information Learning R2 0.160 0.094 0.221 0.168 0.304 0.179 β p β p β p β p β p β p Action 0.265** 0.000 0.009 0.807 0.162** 0.000 -0.008 0.834 CCG 0.044 0.157 0.078* 0.013 0.018 0.564 0.038 0.210 Fighting 0.030 0.299 0.025 0.427 0.014 0.637 -0.008 0.795 FPS 0.068* 0.024 0.166** 0.000 0.036 0.258 0.045 0.183 MMO 0.053 0.071 0.073* 0.022 0.029 0.301 0.042 0.161 MOBA 0.079* 0.007 0.237** 0.000 0.049 0.107 0.091* 0.004 Rhythm 0.049 0.068 -0.022 0.476 0.018 0.517 -0.027 0.345 RPG 0.081* 0.020 0.004 0.916 0.044 0.195 0.013 0.696 RTS 0.087* 0.003 -0.007 0.816 0.066* 0.033 -0.013 0.685 Sandbox 0.112** 0.000 -0.043 0.229 0.048 0.106 -0.050 0.162 Sports 0.012 0.628 -0.021 0.585 0.021 0.411 -0.007 0.859 Casual 0.183** 0.000 0.118** 0.000 0.127** 0.000 0.105** 0.001 Competitive 0.041 0.118 0.297** 0.000 0.011 0.703 0.252** 0.000 How to play 0.070* 0.015 0.230** 0.000 0.053 0.072 0.207** 0.000 Let's play 0.263** 0.000 -0.011 0.713 0.187** 0.000 -0.002 0.954 Review 0.116** 0.000 0.019 0.488 0.068* 0.023 0.023 0.419 Speedrun 0.120** 0.000 0.026 0.334 0.046 0.145 0.032 0.326 Talkshow 0.031 0.309 -0.062* 0.036 0.001 0.960 -0.056 0.062 Note: * = p < 0.05, ** = p < 0.01 34 Table 5 Regression analysis for personal integrative and social integrative motivations Model 1 Model 2 Model 3 Personal integrative Social integrative Personal integrative Social integrative Personal integrative Social integrative R2 0.100 0.169 0.114 0.209 0.144 0.242 β p β p β p β p β p β p Action 0.105* 0.003 0.141** 0.000 0.035 0.337 0.052 0.144 CCG 0.007 0.830 0.024 0.471 -0.031 0.347 -0.030 0.356 Fighting 0.057 0.073 0.021 0.502 0.028 0.376 -0.004 0.899 FPS 0.087* 0.010 0.111** 0.000 0.034 0.337 0.027 0.403 MMO 0.030 0.399 0.056 0.097 0.009 0.802 0.032 0.332 MOBA 0.040 0.188 0.098** 0.001 -0.014 0.663 0.018 0.575 Rhythm 0.074* 0.016 0.112** 0.000 0.038 0.226 0.073* 0.019 RPG 0.002 0.951 0.032 0.406 -0.012 0.736 0.021 0.579 RTS -0.071* 0.024 -0.055 0.076 -0.079* 0.017 -0.063* 0.036 Sandbox 0.168** 0.000 0.186** 0.000 0.126** 0.000 0.129** 0.000 Sports -0.013 0.652 -0.005 0.883 -0.011 0.719 0.002 0.948 Casual 0.212** 0.000 0.270** 0.000 0.178** 0.000 0.220** 0.000 Competitive 0.051 0.092 0.117** 0.000 0.058 0.108 0.114** 0.001 How to play 0.021 0.538 0.015 0.645 0.030 0.361 0.010 0.746 Let's play 0.077* 0.017 0.154** 0.000 0.037 0.271 0.100* 0.003 Review 0.035 0.321 0.032 0.345 0.032 0.371 0.020 0.542 Speedrun 0.086* 0.005 0.076* 0.009 0.068* 0.045 0.041 0.198 Talkshow 0.076* 0.020 0.105** 0.001 0.066* 0.044 0.088* 0.005 Note: * = p < 0.05, ** = p < 0.01 35 Appendix A: Discriminant validity AVE CR Alpha 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 1. Genre: Action 1,00 1,00 1,00 1,00 2. Genre: CCG 1,00 1,00 1,00 -0,02 1,00 3.Genre: FPS 1,00 1,00 1,00 0,21 -0,26 1,00 4. Genre: Fighting 1,00 1,00 1,00 0,33 0,08 0,11 1,00 5. Genre: MMO 1,00 1,00 1,00 0,27 0,25 0,02 0,15 1,00 6. Genre: MOBA 1,00 1,00 1,00 0,03 0,16 -0,10 0,12 0,22 1,00 7. Genre: Rhythm 1,00 1,00 1,00 0,20 0,09 0,10 0,30 0,21 0,15 1,00 8. Genre: RPG 1,00 1,00 1,00 0,39 0,27 0,01 0,30 0,46 0,12 0,22 1,00 9. Genre: RTS 1,00 1,00 1,00 0,14 0,26 0,08 0,26 0,30 0,23 0,16 0,29 1,00 10. Genre: Sandbox 1,00 1,00 1,00 0,42 -0,04 0,24 0,17 0,21 -0,03 0,18 0,25 0,10 1,00 11. Genre: Sports 1,00 1,00 1,00 0,33 0,05 0,19 0,13 0,14 0,03 0,21 0,22 0,08 0,29 1,00 12. Stream type: CASUAL 1,00 1,00 1,00 0,35 0,07 0,16 0,14 0,21 0,15 0,22 0,19 0,05 0,26 0,19 1,00 13. Stream type: COMPETITIVE 1,00 1,00 1,00 -0,04 0,11 0,30 0,13 0,07 0,36 0,07 -0,04 0,16 -0,02 -0,02 -0,05 1,00 14. Stream type: HOWTOPLAY 1,00 1,00 1,00 0,16 0,14 0,02 0,13 0,24 0,25 0,10 0,21 0,17 0,14 0,08 0,17 0,10 1,00 15. Stream type: LETSPLAY 1,00 1,00 1,00 0,44 0,07 0,16 0,20 0,24 0,05 0,18 0,32 0,20 0,38 0,17 0,34 -0,01 0,17 1,00 16. Stream type: REVIEW 1,00 1,00 1,00 0,31 0,13 0,07 0,20 0,24 0,06 0,15 0,29 0,25 0,21 0,08 0,18 -0,03 0,46 0,29 1,00 17. Stream type: SPEEDRUN 1,00 1,00 1,00 0,30 0,13 0,12 0,45 0,19 0,08 0,32 0,34 0,26 0,21 0,17 0,14 0,09 0,10 0,32 0,17 1,00 18. Stream type: TALKSHOW 1,00 1,00 1,00 0,25 0,15 0,11 0,20 0,19 0,13 0,17 0,21 0,23 0,26 0,14 0,26 0,11 0,31 0,27 0,39 0,15 1,00 19. Motivation: AFFECTIVE 0,75 0,92 0,89 0,24 0,13 0,20 0,15 0,18 0,19 0,15 0,19 0,09 0,21 0,13 0,32 0,29 0,15 0,22 0,12 0,17 0,20 0,87 20. Motivation: INFO 0,77 0,93 0,90 0,41 0,10 0,15 0,23 0,26 0,14 0,20 0,31 0,23 0,29 0,20 0,33 0,05 0,23 0,42 0,29 0,26 0,24 0,43 0,88 21. Motivation: LEARN 0,76 0,93 0,90 0,05 0,09 0,11 0,08 0,14 0,25 0,05 0,08 0,10 0,00 0,02 0,13 0,31 0,27 0,06 0,11 0,08 0,08 0,49 0,42 0,87 22. Motivation: PERSONAL INTEGRATIVE 0,72 0,91 0,87 0,22 -0,02 0,15 0,14 0,11 0,04 0,15 0,12 0,01 0,25 0,11 0,28 0,06 0,12 0,21 0,15 0,17 0,19 0,47 0,42 0,31 0,85 23. Motivation: SOCIAL INTEGRATIVE 0,54 0,90 0,88 0,29 0,02 0,18 0,16 0,19 0,11 0,22 0,19 0,07 0,31 0,16 0,36 0,12 0,15 0,31 0,18 0,19 0,26 0,66 0,48 0,35 0,70 0,73 24. Motivation: TENSION RELEASE 0,54 0,92 0,90 0,25 0,14 0,19 0,16 0,16 0,23 0,16 0,16 0,10 0,19 0,14 0,36 0,27 0,14 0,29 0,10 0,20 0,23 0,72 0,46 0,41 0,45 0,65 0,74