How games induce cooperation? A study on the relationship between game features and we-intentions in an augmented reality game

How games induce cooperation? A study on the relationship between game features and we-intentions in an augmented reality game Benedikt Morschheuser Institute of Information Systems and Marketing, Karlsruhe Institute of Technology, Germany Corporate Research, Robert Bosch GmbH, Germany benedikt.morsc[email protected] Marc Riar Chair of Information Systems IV, University of Mannheim, Germany [email protected] Juho Hamari Gamification Group, Tampere University of Technology, Finland Gamification Group, University of Turku, Finland [email protected] Alexander Maedche Institute of Information Systems and Marketing, Karlsruhe Institute of Technology, Germany [email protected] DOI: Reference: https://doi.org/10.1016/j.chb.2017.08.026 To appear in: Computers in Human Behavior Please cite this article as: Benedikt Morschheuser, Marc Riar, Juho Hamari, Alexander Maedche, How games induce cooperation? A study on the relationship between game features and weintentions in an augmented reality game , Computers in Human Behavior (2017), doi: 10.1016/j.chb. 2017.08.026 This is the accepted manuscript of the article, which has been published in Computers in Human Behavior. 2017, 77, 169-183. https://doi.org/10.1016/j.chb.2017.08.026 How games induce cooperation? A study on the relationship between game features and we-intentions in an augmented reality game Abstract Seamless cooperation between individuals is essentially a crucial aspect of any successful endeavor. A host of literature has been published in the academic realm about how cooperation could be cultivated. However, true cooperation often forms organically without external enforcement. Recently, there has been one special example of a context where cooperation seemed to have effortlessly sprung up between people who might not even have had previous connections. The context is video/online games; games such as Ingress, Pokémon Go, and World of Warcraft bind people together to work against insurmountable odds and to overcome jointly held challenges. Organizations of many types have recently begun to gamify their structures and services in order to cultivate such seamless cooperation. However, before this potential of games can be successfully wielded outside video games, we need to understand better how games are able to cultivate such cooperation. Therefore, in this study we investigate how games can induce and cultivate we-intention of working as a group. Specifically, we investigate how cooperative game features affect different forms of group dynamics and how they further translate into we-intentions. We employ data from users of the augmented reality game Ingress (N = 206). The results show that cooperative game features induce we-intentions via positively increasing group norms, social identity, joint commitment, attitudes toward cooperation, and anticipated positive emotions. The findings imply that practitioners who are looking to increase cooperation should find that gamification inspired by cooperative game design is beneficial and preferable over individual-based gamification efforts. Keywords: Gamification, cooperation, online games, location-based games, augmented reality, weintention 1 Introduction Cooperation is an anchor of our society and a key ability of people, who are working, studying, and carrying out most leisure activities together to achieve shared goals. During the past decade, modern information & communication technologies (ICT) that digitally connect people around the globe have birthed completely new forms of cooperation in organizations and beyond. A large variety of collaborative technologies, such as online communities (Hutter, Hautz, Füller, Mueller, & Matzler 2011), crowdsourcing platforms (Geiger & Schader 2014), or instant messaging services (Shen, Cheung, & Lee 2013) have emerged that facilitate working together anytime and across geographical borders. However, many studies report that it is challenging to motivate people to adopt and use such collaborative technologies (Arazy, Gellatly, Brainin, & Nov 2016; Hutter et al. 2011; C.-P. Lin & Bhattacherjee 2008; K.-Y. Lin & Lu 2011; Wasko & Faraj 2005; Zhao & Zhu 2014). Thus, much work has sought to understand which factors drive people to cooperate (Bagozzi & Dholakia 2002, 2006; Tsai & Bagozzi 2014) and how design features in information systems and services can support and cultivate cooperative behaviors (Jung, Schneider, & Valacich 2010; Straub, Gimpel, Teschner, & Weinhardt 2015; Valacich, Dennis, & Connolly 1994; Zhang, Venkatesh, & Brown 2011). Concerning video games, it can be observed that, in many games, cooperation emerges effortlessly; people start to pool individual efforts, cooperate seamlessly even against the most unimaginable odds, and express strong enthusiasm while acting together (C.-H. Chen, Sun, & Hsieh 2008; Cole & Griffiths 2007; Ducheneaut, Yee, Nickell, & Moore 2006; Scharkow, Festl, Vogelgesang, & Quandt 2015; Teng & Chen 2014; Yee 2006). Thus, today, practitioners turn to games for inspiration on how to design information systems, services, and organizational structures more cooperatively (Bui, Veit, & Webster 2015; Morschheuser, Maedche, & Walter 2017b; Ribeiro, Farinha, Pereira, & Mira da Silva 2014; Schacht, & Maedche 2015; Thom, Millen, Dimicco, & Street 2012). This trend can be understood as part of the gamification movement, which represents the use of game elements and mechanics outside traditional video game environments (Hamari, Koivisto, & Sarsa 2014; Huotari & Hamari 2017). Initial empirical studies indicate that applying game mechanics and features of cooperative games, such as point systems that reward cooperation (Blohm, Bretschneider, Leimeister, & Krcmar 2011; Y. Chen & Pu 2014; Siu, Zook, & Riedl 2014; Thom et al. 2012), team competitions (Peng & Hsieh 2012; Chen & Pu 2014), or virtual worlds with avatars (Rico, Martínez-Muñoz, Alaman, Camacho, & Pulido 2011) can positively influence cooperation in various contexts. However, we still lack a comprehensive understanding of how games cultivate cooperation (Liu, Li, & Santhanam 2013; Hamari & Keronen 2017b; Morschheuser et al. 2017b), which keeps us from successful wielding the potential of cooperative games in gamification (Bui et al. 2015; Liu, Santhanam, & Webster 2017; Morschheuser, Hamari, & Koivisto 2017a). Recently, the concept of we-intentions has gained attention in information system (IS) research concerning understanding cooperation with collaborative technologies. Compared to typically studied individual intentions, the concept of we-intention relies on the idea that true cooperation requires collective intentions and therefore cannot be analyzed only as the sum of individual intentions (Searle 1990; Tuomela 2000). The concept is increasingly gaining attention in IS research, in order to study cooperation and cooperative behaviors in online communities (Bagozzi & Dholakia 2002; Cheung & Lee 2010; Dholakia, Bagozzi, & Pearo 2004; Shen et al. 2013; Shen, Cheung, Lee, & Chen 2011; Tsai & Bagozzi 2014) and crowdsourcing systems (Bagozzi & Dholakia 2006; Shen, Lee, & Cheung 2014; Shen, Lee, Cheung, & Chen 2009). Owing to the strong similarities between such virtual communities and online games, this theoretical framework provides excellent support for investigating cooperation in games. Therefore, in this paper, we empirically investigate how games cultivate we-intentions of working as a group by drawing on cooperation theory (Tuomela 2000) and particularly the concept of we-intentions (Bagozzi & Dholakia 2002; Tsai & Bagozzi 2014; Tuomela 2000). On the basis of survey data, gathered in the context of the augmented reality game Ingress that engages people in generating an interactive map with cultural points of interest, we seek to enhance current understandings of how engagement with cooperative game features induce we-intentions via group dynamics, such as group norms, positive and negative anticipated emotions, social identity, joint commitment, and attitudes toward cooperation. Further, we investigate whether engagement with individualistic game features – such as private badges, points or levels – that are currently often used in the context of collaborative technologies (Hamari et al. 2014; Morschheuser, Hamari, & Koivisto 2016) influence these effects. This paper summarizes this study’s findings and discusses theoretical and practical implications. 2 Theoretical Background and Research Model 2.1 We-intentions and their antecedents Researchers have put much effort into theoretically conceptualizing and studying the phenomenon of cooperation from different perspectives, including philosophy (Gilbert 1989; Tuomela 2000), game theory (Nash 1953), and social psychology (Johnson 2003). Commonly, pure or full-blown cooperation is typically considered to consist of collective social actions, in which more than one person act jointly toward a common goal (e.g. carrying a table jointly) (Gilbert 1989; Tuomela 2000). According to Tuomela (2000, 2011), such cooperation is characterized and determined by a collective we-intention of group members towards a shared goal. Thus, recently, studies have commonly drawn on the concept of we-intention (Searle 1990; Tuomela 2000, 2011) to operationalize cooperation in groups. In contrast to typically investigated personal intentions, which capture individual commitment to an action, we-intentions involve a ‘we-perspective’, expressing a collective commitment to participate in a cooperative action (Bratman 1997; Searle 1990; Tuomela 2000, 2011). Therefore, we-intentions are explicitly formulated with reference to a collective entity of we or us, which expresses the intention to jointly perform an activity together with others: “We intend to do X jointly” (Bratman 1997; Searle 1990; Tuomela 2000, 2011). Since the inception of these conceptualizations, the we-intentions-operationalization has been applied in the study of cooperation in several technology-mediated contexts, such as wikis, crowdsourcing platforms, instant messaging services, and other social communities (Bagozzi & Dholakia 2002, 2006; Shen et al. 2014; Tsai & Bagozzi 2014). These studies have shown that we-intention is a strong proximal determinant of cooperation and provides a more comprehensive explanation of user participation in group efforts than traditionally investigated personal intentions. Although the adoption and use of collaborative technologies have been frequently investigated in IS research (C.-P. Lin & Bhattacherjee 2008; K.-Y. Lin & Lu 2011; Majchrzak, Rice, Malhotra, King, & Ba 2000), studies have mostly investigated individual intentions (i-intentions). The we-intention concept has been largely overlooked by the repetitive application of theories such as the technology acceptance model (Davis 1989), the theory of reasoned action (Fishbein & Ajzen 1975), and the theory of planned behavior (Ajzen 1991), all of which solely investigate intentions from an individual perspective. However, it is obvious that participation and cooperation in online communities and other collaborative technologies are commonly a group activity. Typically, users adopt and use such technologies for a common reason or to achieve a shared goal. Thus, cooperation in such collaborative technologies is increasingly investigated through the concept of we-intention in recent IS research (De Oliveira & Huertas 2015; Shen et al. 2013; Tsai & Bagozzi 2014). Guided by Bagozzi and Dholakia (Bagozzi & Dholakia 2002, 2006; Dholakia et al. 2004; Tsai & Bagozzi 2014), a variety of efforts have been made to empirically investigate the cooperation of users in social communities, crowdsourcing approaches, or instant messaging services (Table 1). A synthesis of extant research indicates that individual factors, such as attitudes (Tsai & Bagozzi 2014) and anticipated emotions (Bagozzi & Dholakia 2002; Tsai & Bagozzi 2014) as well as social antecedents, such as joint commitment (Shen et al. 2014 2009; Tuomela 2000), social identity (Bagozzi & Dholakia 2002; Tsai & Bagozzi 2014), and group norms (Bagozzi & Dholakia 2002; Oliveira & Huertas 2015; Tsai & Bagozzi 2014) influence we-intentions to work together with collaborative technologies. More importantly, it has remained unclear which features in these technologies are responsible for invoking we-intentions and how specific features can, in the end, engage more cooperation. Moreover, the particular context of cooperation in games and gamified approaches has been largely ignored in prior we-intention research. Table 1: We-intention and its antecedents in collaborative technologies Work Context Significant antecedents of we-intention De Oliveira & Huertas 2015 Participation in the social network Facebook Group norms, social identity, subjective norms, social enhancement, maintaining interpersonal interconnectivity Tsai & Bagozzi 2014 Participation in a Chinese social network Group norms, subjective norms, social identity, anticipated emotions, attitudes toward contributing, desire Shen, Lee, & Cheung 2014 Crowdsourcing participation in wikis Commitment to the community, team trust Shen, Cheung, & Lee 2013 Collaboration via instant messaging Group norms, social identity, perceived critical mass Cheung, Chiu, & Lee 2011 Participation in the social network Facebook Group norms, social enhancement, entertainment value, maintaining interpersonal interconnectivity, social presence Shen, Cheung, Lee, & Chen 2011 Collaboration via instant messaging Group norms, social identity, desire Cheung & Lee 2010 Participation in the social network Facebook Social identity, subjective norms Shen, Lee, Cheung, & Chen 2009 Crowdsourcing participation in wikis Joint commitment, mutual agreement, personal outcome expectations, communityrelated outcome expectations Bagozzi & Dholakia 2006 Participation in Linux user groups Social identity, attitudes toward contributing, negative anticipated emotions, perceived behavioral control Dholakia, Bagozzi, & Pearo 2004 Participation in virtual communities Group norms, social identity, mutual agreement, desire Bagozzi & Dholakia 2002 Participation (chatting) in different virtual communities Social identity, positive anticipated emotions, desire 2.2 Cooperation and games Games are a pinnacle form of hedonic systems that invoke rich motivational experiences and excite masses of people (Hamari 2015a; Ryan, Rigby, & Przybylski 2006; Vesa, Hamari, Harviainen, & Warmelink 2017; Yee 2006). Since the first video games, researchers were fascinated by the psychological and behavioral outcomes of games and made efforts to understand the design characteristics, which are responsible for the rich motivational experiences and the different behavioral effects of games (Choi & Kim 2004; Hamari & Keronen 2017b; Hamari & Tuunanen 2014; Malone 1981; Ryan et al. 2006; Yee 2006). One particular behavior that can be observed in many games is cooperation. Especially in the context of online games that utilize the internet to bring people together, cooperation seems to effortlessly spring up between people who might not even have had previous connections (Cole & Griffiths 2007; Velez & Ewoldsen 2013; Yee 2006). Several studies have revealed that popular online games, such as World of Warcraft, Star Wars Galaxies, or Ultima Online attract masses of people by providing them with rich social experiences while playing together (Przybylski, Rigby, & Ryan 2010; Rigby & Ryan 2011; Yee 2006; Zhong 2011). Research has showed that players greatly enjoy the social interaction and cooperation in such games (Przybylski et al. 2010; Velez & Ewoldsen 2013; Yee 2006). In all these games, people help each other voluntarily to accomplish common missions and goals, strongly interact with one another to discuss in-game strategies and opportunities, share comprehensive knowledge about game contents in forums and wikis, and form groups (known as guilds, clans or factions) that persist over time. Further, several empirical studies in the context of video games and gamified systems highlight the positive effects of cooperation in games compared to competition. For instance, Y. Chen and Pu (2014), Marker and Staiano (2015), Peng and Hsieh (2012), and Plass et al. (2013) found that cooperative approaches can lead to higher engagement with games or gamified systems compared to competitive approaches. Positive effects on goal commitment (Peng & Hsieh 2012), motivation (Marker & Staiano 2015), and intentions to recommend a game (Plass et al. 2013) have also been found to be effects of games that are designed to support cooperation. However, little research has analyzed the design features and dynamics in games and gamified systems that enable and promote cooperation (Bui et al. 2015; Liu et al. 2013; Morschheuser et al. 2017b). Zagal et al. 2006). Consequently, the use of cooperative game features and the personal success that may be attached to using such features may lead to a positive evaluation of an individual’s own value in the group and thus may affect his or her group-based self-esteem. Previous research on we-intention showed that a social identity positively influences weintentions (Table 1). Therefore, we hypothesize: H3a: Engagement with cooperative game features positively relates to social identity. H3b: Greater levels of social identity are associated with higher levels of we-intention. 2.5 The roles of joint commitment Cooperative games typically apply features that allow players to communicate and interact with other players (Choi & Kim 2004; Ducheneaut et al. 2006; El-Nasr et al. 2010; Yee 2006). This involves features for verbal and written communication, such as virtual chat (Ducheneaut et al. 2006), but also functions that can be used for non-verbal interactions. Examples of the latter are in-game abilities that can be used to interact with other players in a virtual world (e.g. virtual objects and virtual gifts) or abilities that can be used to positively influence other players (e.g. healing abilities) (El-Nasr et al. 2010; Rocha et al. 2008). Some cooperative games, such as Portal2 or LittleBigPlanet, also give players opportunities to use the body of an avatar in order to express feelings and communicate with other players via virtual body language (cf. C.-Y. Hsu, Tung, Wang & Wang 2014). A key aspect of these features is that players can use them as a medium to express willingness and commitment to participate in a joint action. For instance, a player who buys a virtual item that increases the attack value of all the players in his area shows his willingness to go into battle together with others. Also, avatar-based body language, which encourages other players to stick together, can serve as an instrument to express commitment to a joint action (C.-Y. Hsu et al. 2014). Theory on cooperation (Gilbert 1989; Tuomela 1995, 2000) emphasizes that participants’ joint commitment to a collective action is crucial to the cultivation of cooperation. Thus, joint commitment has been identified as an antecedent of we-intentions (Shen et al. 2014): “Weintention involves a joint commitment to contribute to the realization of the content of the weintention” (Tuomela 2000, p. 63). Compared to personal commitments, joint commitment involves a mutual agreement by every member to participate in a joint activity (“we do X jointly”). In other words, joint commitment can be defined as the common knowledge that all participants jointly express their readiness to be under the obligation to participate in a joint action and are jointly committed to do their part of “X” (Gilbert 1999; Shen et al. 2009; Tuomela 1995, 2000). We assume that communication features in cooperative games may support the joint commitments, similar to other virtual communities (Bagozzi & Dholakia 2002; Shen et al. 2014, 2009), since they enable players to communicate and express personal beliefs and obligations. Further, we expect that cooperative game features that enable and support cooperative interaction as described above, can be an important instrument to express joint commitment. Based on this discussion, we assume that: H4a: Engagement with cooperative game features positively relates to joint commitment. H4b: Joint commitment positively relates to we-intention. 2.6 The roles of attitudes In the context of games and gamification approaches, many studies have investigated the relationships between individual attitudes and the behavioral intention to use a game or a gamified system (Hamari & Koivisto 2015a, 2015b; C.-L. Hsu & Lu 2004; for a review, see Hamari & Keronen 2017b). An individual’s attitude toward a behavior represents the psychological evaluation of a planned behavior along a positive-to-negative dimension (Ajzen 1991). Attitudes are cognitive variables that influence decision-making and lead to behavioral intentions (Armitage & Conner 2001). Research has revealed that hedonic aspects of games (Hamari & Keronen 2017b; Wu & Liu 2007) and social features that increase social recognition (Hamari & Koivisto 2015a, 2015b) influence attitudes to use gamified systems and games. However, very few studies have investigated attitudes and intentions to play together with others. According to research on we-intentions, attitudes toward a cooperative act has been identified as a key psychological predictor of we-intention (Bagozzi & Dholakia 2006; Tsai & Bagozzi 2014). In the context of games, this relationship between attitude and we-intention is expected to be similar. Thus, we posit: H5a: Engagement with cooperative game features positively relates to attitudes toward playing together with others. H5b: Greater levels of attitudes toward playing together with others are associated with higher levels of we-intention. 2.7 The mediating roles of group dynamics Extensive research that has been conducted in order to investigate the antecedents of weintention (Table 1) revealed that group norms, positive and negative anticipated emotions, social identity, joint commitment, and attitudes toward cooperating together with a group explain we-intentions very well (Bagozzi & Dholakia 2002, 2006; Dholakia et al. 2004; Shen et al. 2009; Tsai & Bagozzi 2014). Especially concerning collaborative technologies – such as online communities, crowdsourcing platforms and messaging services – these group dynamics seem to predict we-intention and cooperation (Table 1). We extend this research by investigating whether and how cooperative game features induce we-intentions. As hypothesized, we suspect that cooperative features in games influence all these group dynamics in order to invoke and cultivate we-intentions. Consequently, there is no obvious reason to suspect that these variables do not fully mediate the relationships between cooperative game features and we-intentions. Accordingly, we posit: H6: The effect between engagement with cooperative game features and we-intention is fully mediated by group norms, positive and negative anticipated emotions, social identity, joint commitment, and attitudes toward cooperating with others. 2.8 The roles of other, non-cooperative game features Popular discussion pertaining to online games such as World of Warcraft or Pokémon Go highlight that people easily and organically form highly cohesive teams in games (Brown & Thomas 2006; Ducheneaut, Yee, Nickell, & Moore 2007), indicating that games seem to uniquely cater for the development of we-intentions between individuals. However, games are often complex and characterized by the uses of various game features (Björk & Holopainen 2005; Vesa et al. 2017) that allow a user to play them alone or together with others (Marker & Staiano 2015; Peng & Hsieh 2012). Currently, most game companies develop games in a way that allows the player to play both in single-player and/or cooperative multiplayer modes by integrating various game features. While we mainly hypothesize that games’ positive effects on group dynamics and we-intention stem from a game’s cooperative features, it is possible that the single-player features also have a role in the formation of cooperation in games. According to Morschheuser et al. (2017b) and Liu et al. (2013), only cooperative game features are designed to invoke cooperative goal structures that have been defined as a key antecedent of ‘full-blown’ cooperation (Tuomela 2000). In contrast, non-cooperative (cf. single-player/individualistic) game features are detached from cooperative goals and characterize individual goal structures, invoking personal intentions that lead to individual play rather than we-intentions and cooperation. Several studies have indicated clear differences in behavioral outcomes between cooperative, individualistic, and competitive designs in games and gamified systems (Y. Chen & Pu 2014; Goh & Lee 2011; Marker & Staiano 2015; Peng & Hsieh 2012; Plass et al. 2013; Siu et al. 2014). Based on these considerations and empirical results, we hypothesize that cooperative game features induce we-intentions and positively affect group dynamics, hypothesizing that individualistic game features have no effects on group dynamics that invoke we-intentions. Accordingly: H7a: Engagement with individualistic game features is not associated with we-intentions. H7b: Engagement with individualistic game features is not associated with group dynamics. Figure 2: The research model 3 The Empirical Study 3.1 Data We gathered the data from players of Ingress, a popular game with over 11 million downloads in more than 200 countries (Smith 2015; Takahashi 2014). Ingress is an augmented reality game that relies on a map of the real world and extends reality with location-based virtual objects, the so-called portals, at places of public art (Hulsey & Reeves 2014). Ingress players can use an app on their mobile phones, the so-called scanner, to find the location of such portals and create new virtual portals, which are connected by their location to landmarks in the real world. The game rules define that these portals should be created at places of public art, for instance, at statues, graffiti, or buildings with outstanding architecture. Main goal in the game is to collect points by creating and conquering these virtual portals and linking them together in order to create triangular fields in virtual space. The scoring system is based on these fields (the more space is covered by a field, the more points one can get). Two factions (Enlightened vs. Resistance) compete against each other for the portals and fields. The winning team with the most points (Mind Units) in a predefined period is determined at regular intervals. Since the portals and other game content are created by the community of players, Ingress can be also specified as a gamified crowdsourcing approach (Morschheuser et al. 2016; 2017a) that creates an interactive map with cultural points of interest (Hulsey & Reeves 2014). The user-generated database has been used as source for other services, such as the Pokémon Go app. Ingress seems apt for this study, because participants display strong cooperative behavior in their faction. Research has identified that, besides factors such as fun and exploration, the community in a faction and their shared progress is of great importance to Ingress players (Sheng 2013). Further, there are many regional associations of Ingress players on the internet that organize regional meetups, teach one another, and discuss strategies to achieve regional supremacy1. All of the above mentioned cooperative design features are implemented in Ingress. First, the game defines clear goals for all members of a faction. The common goal is to obtain more Mind Units (MU) than the other team in predefined time intervals (Checkpoints and cycles). Second, different mechanics and rules are implemented that allow players to cooperatively support each other, such as applying Mods, upgrading portals, and recharging Resonators of other players or the possibility to share items among faction members with Capsules. Third, players can interact with each other in Ingress using an integrated chat feature (COMM). In addition to these technical features, the Ingress community regularly hosts events where people can meet to socialize, exchange strategies, and play Ingress together (e.g. Mission Days, XM Anomalies and First Saturday events). Ingress also uses several more individualistic features, such as a level system, badges, and personal (agent) statistics, which are designed to invoke individualistic goal structures and allow to play the game as a single player. We tested our proposed model with data from Ingress players gathered via an online questionnaire. We invited the participants in different open Ingress communities on Google+ by posting a short description of the survey and a link to it. In addition, we asked the moderators and owners of several communities to re-post the link in their private groups so as to extend the range. The survey was active from May 4, 2016 to October 30, 2016. In that time, about 800 Ingress players responded to our call, and 233 players participated in the questionnaire. Similar to previous we-intention research, we asked the participants to identify a group of Ingress players they regularly play with (Bagozzi & Dholakia 2006; Tsai & 1 https://www.ingress.com/community/directory. Bagozzi 2014) so as to induce respondents to think about a real group of people and to better prepare them for the questions to follow. Since our study focuses on cooperative play, we excluded players who reported that they don’t regularly play in a group from the analysis, resulting in 206 valid responses. Table 2 outlines the sample characteristics. Respondents were entered into a prize draw for one of three €15 Amazon gift coupons. Table 2: Demographic information about respondents, including gender, age, days since last playing Ingress, and time playing Ingress Gender # % When did you last play Ingress? # % Female 62 30.1 today 134 65.0 Male 144 69.9 yesterday 59 28.6 > 1 day ago 13 6.3 Age # % Time playing Ingress # % < 20 19 9.2 < 1 years 54 26.2 20 to 29 49 23.8 1 to 2 years 60 29.1 30 to 39 70 34.0 2 to 3 years 54 26.2 40 to 49 50 24.3 > 3 years 38 18.4 50 and over 18 8.7 The questionnaire was answered by people from 15 countries. The average respondent age was 34.6. The sample consisted of 30.1% female and 69.9% male participants; 26.2% have been playing Ingress for less than a year, 29.1% between 1 and 2 years, 26.2% between 2 and 3 years, and 18.4% for more than three years. Respondents typically reported playing Ingress multiple times a day (68.4%, on a 6-item scale anchored between rarely and multiple times a day), play for 48 hours (median) per month, and have at least completed high school (89.4%) or a Bachelor’s degree (43.8%). 3.2 Measurement We adopted the operationalization of we-intention and its antecedents from previous sources (see Table 1; Table 3). The items to measure the engagement with cooperative game features in Ingress were developed in two stages: First, we played Ingress ourselves and conducted semi-structured interviews with eight experienced players to identify cooperative game features in Ingress. The interviews were conducted via telephone and lasted approximately 40 minutes. The participants (5 male, 3 female, average age 28, all had been playing Ingress regularly for at least six months) were recruited in Ingress user groups on Google+. We guided the interviews along a questionnaire that contained a large set of Ingress game design features, which we derived from the official Ingress user guide2. Following Morschheuser et al.’s (2017b) classification framework, we asked the experts to separate the Ingress game features into individualistic, competitive, and cooperative game features. First, we comprehensively introduced these different feature categories. Next, we asked the participants to specify the nature (individualistic, cooperative, competitive) of the identified Ingress game design features on a 5-point Likert scale (strongly disagree – strongly agree) (see Appendix 1). We asked them to think aloud so we could take notes on their reasoning for placing the game features in the various categories. Several game features were perceived by the participants as having both competitive traits (at an intergrouplevel) as well as cooperative traits (at an intragroup level). For such cooperative-competitive features (e.g. factions), we carefully identified the cooperative aspects. Based on the interviews, we noted that 12 features were mainly perceived as cooperative features and eight features were classified primarily as individualistic (Appendix 2). Based on these identified cooperative and individualistic game features, we developed items that measured a user’s engagement with these features. Together, we used these items to operationalize the engagement with cooperative / individualistic game features in Ingress. We treated the constructs as formative, as its exposure happens through using those features or by perceiving them as important. Further, the identified features were not explicitly interchangeable, which contradicts reflective use (Diamantopoulos & Siguaw 2006). 2 https://support.ingress.com/hc/en-us. Second, we conducted a pre-study to evaluate the developed measurements and the constructs we adopted from the existing we-intention literature. We conducted the survey on two days in October 2015. In sum, 110 participants took part in the pre-study, which were recruited on Ingress user groups different to the groups of the main study. Since the pre-study revealed largely a high validity level of the used items, we made only small adjustments. We improved some of the items that had been developed to measure engagement with cooperative game features and added a third we-intention item based on Shen et al.’s work (2009). Table 3: The measurement items We-intention (WE-I) Tsai and Bagozzi (2014); Bagozzi and Lee (2002); Shen et al. (2009) (1) I intend that our group (i.e. myself and the group that I identified before) play Ingress together sometime during the next 4 weeks. (7-point “disagree” - “agree” scale) (2) We (i.e., I and the group that I identified before) intend to play Ingress together sometime during the next 4 weeks. (7-point “disagree” - “agree” scale) (3) We (i.e., I and the group that I identified before) plan to play Ingress together sometime during the next 4 weeks. (7-point “disagree” - “agree” scale) Group norms (GN) Tsai and Bagozzi (2014); Bagozzi and Dholakia (2002) Playing Ingress together sometime within the next 4 weeks with the group you identified before can be considered a goal. For each of the people listed below and for yourself, please estimate the strength to which each holds the goal. (7-point “very weak” - “very strong” scales) (1) Average of the strength of group members’ goal (2) Strength of own goal Positive anticipated emotions (PAE) Tsai and Bagozzi (2014); Perugini and Bagozzi (2001) If I am able to play Ingress with the group I identified before during the next 4 weeks, I will feel: (1) excited, (2) delighted, (3) happy, (4) satisfied, (5) proud, (6) self-assured, (7) relief, (8) glad (7-point “not at all” - “moderately” - “very much” scales) Negative anticipated emotions (NAE) Tsai and Bagozzi (2014); Perugini and Bagozzi (2001) If I am unable to play Ingress with the group I identified before during the next 4 weeks, I will feel: (1) angry, (2) frustrated, (3) guilty, (4) ashamed, (5) disappointed, (6) depressed, (7) worried, (8) uncomfortable, (9) anxious (7-point “not at all” - “moderately” - “very much” scales) Joint commitment (JC) Shen et al. (2009) (1) We (i.e., I and the group that I identified before) all know that all members are jointly committed to performing their parts of the common tasks. (7-point “disagree” - “agree” scale) Negative emotions 0.021 -0.145 0.252 0.830 Social identity 0.092 -0.029 0.261 0.212 Joint commitment 0.218 0.047 0.395 0.014 Attitude 0.097 -0.076 0.328 0.338 The relationship between group dynamics and we-intention Group norms We-intention 0.253 0.145 0.356 0.000 Positive emotions 0.254 0.078 0.392 0.002 Negative emotions 0.063 -0.026 0.169 0.207 Social identity 0.155 0.003 0.320 0.055 Joint commitment 0.151 0.024 0.263 0.011 Attitude 0.205 0.078 0.347 0.003 The total effect between engagement with game features and we-intention Cooperative game features We-intention 0.495 0.352 0.647 0.000 Individualistic game features We-intention 0.019 -0.075 0.243 0.814 4 Discussion In this study, we have investigated how games can induce and cultivate we-intention of working as a group. Specifically, we investigated how engagement with cooperative game features affect different forms of group dynamics and how they further translate into weintentions. We employed data from Ingress users (N = 206), a popular location-based augmented reality game. The results show that cooperative game features can induce weintentions (directly explains one-quarter of the variance), whereas individualistic game features seem to have no direct effect on we-intentions. Further, this study revealed that the relationship between cooperative game features and we-intention is fully mediated by group norms, social identity, joint commitment, attitudes toward cooperation, and anticipated positive emotions. The study findings extend previous research in several ways. First, by empirically investigating how game design features can invoke we-intentions in games, this study closes a gap in current game (Goh & Lee 2011; Liu et al. 2013) and gamification research (Bui et al. 2015; Y. Chen & Pu 2014; Morschheuser et al. 2016). In addition to previous research that has focused on the behavioral outcomes (Goh & Lee 2011; Marker & Staiano 2015; Peng & Hsieh 2012; Plass et al. 2013; Siu et al. 2014) and the design features of cooperative games (El-Nasr et al. 2010; Rocha et al. 2008; Zagal et al. 2006), this study connected both research streams and provided a comprehensive understanding of how engaging with such features can invoke collective intentions towards cooperative behavior. Further, this study has contributed to better understanding the effects of different game features in the sense that the results show that cooperative game features are positively related to we-intentions, while individualistic features are not. Therefore, these findings provide not only a novel contribution for understanding cooperation and participation in multiplayer games, such as Ingress, Pokémon Go, and World of Warcraft. Moreover, the knowledge we gathered can easily be transferred to current gamification research that investigates the differences between various game features in order to derive more precise design principles for designing successful gamification approaches (cf. Morschheuser et al. 2016; 2017a; 2017b; Hamari et al. 2014). Second, our results provided support for the postulations that enriching ICT with gamification that promote shared goals instead of competition or individualistic goals may support digital cooperation (Morschheuser et al. 2017b; Peng & Hsieh 2012). The present results support this notion by demonstrating that engagement with cooperative game features significantly relates to group dynamics, which operationalize the presence of shared goals. Especially, the strong relationship between group norm and cooperative game features are an indication for the presence of a shared goal between users in a group, since group norms represent the “attempt of individuals to meet idealized values or goals, shared with others” (Tsai & Bagozzi 2014, p. 147). Further, the identified influences of cooperative game features on anticipated emotions regarding achieving a shared goal and joint commitment to contribute to the realization of a shared goal, support this theory. Thus, our study can act as an anchoring point for future research into the effects of singular cooperative game features such as shared goals on psychological and behavioral outcomes of individuals in groups. Third, our findings extend previous we-intention research (Dholakia et al. 2004; Oliveira & Huertas 2015; Tsai & Bagozzi 2014) that have neglected we-intention formation in games. Although the influence of design features on individual intentions in games and gameinspired approaches (i.e. gamification) have often been investigated (Hamari & Keronen 2017b; Hamari & Koivisto 2015a; Hsiao & Chiou 2012; Wu & Liu 2007), to our best knowledge, no other study has investigated we-intentions in this context. More importantly, little knowledge exists about how design features of information systems can influence weintentions. Our study results suggest that cooperative games are particularly efficient at invoking we-intentions. This finding not only contributes to the general understanding of weintentions, but also extends previous research in the sense that the results highlight the use of cooperative game features as a promising approach to invoke we-intentions. Considering our results in the context of recent we-intention research (Table 1), we confirmed that group norms, social identity, joint commitment, attitudes toward cooperation, and anticipated positive emotions explain changes in we-intention reasonably well, also in the context of games. This significant overlap between our study and previous we-intention research contributes to the generalizability of the presented results and provides a foundation for bringing community research and game research closer together. This implies that game and gamification researchers should also turn to existing community research, and vice versa, in order to extend understandings of digital cooperation. 4.1 Practical implications Clearly, cooperation is of great importance in society today; we work, study, and carry out most leisure activities together. However, true cooperation often forms organically without external enforcement and is hard to cultivate. Many practitioners face the challenge of designing organizational structures and digital communities that promote cooperation, without knowing how design features can promote cooperation. When looking at video/online games, one can observe that cooperation effortlessly arises between people who might not even have had previous connections, and not just between individuals, but across a wide spectrum of people. To better understand how the potential of games can be successfully wielded in nongame organizations and information systems, this study has investigated how games are able to cultivate cooperation. The study results support the hypothesis that engagement with cooperative game features can influence collective intentions (we-intentions) of working in a group. This can provide practical insights for those seeking to cultivate cooperation in digital communities and organizations. Our results suggest that cooperative design features in games – such as shared goals, cooperation enabling game mechanics and rules, and communication possibilities – offer great and as yet untapped potential for cultivating cooperation outside games. Designers of information systems and collaborative technology who seek to cultivate cooperation should consider the implementation of such design features known from cooperative games when gamifying their systems (Huotari & Hamari 2017; Morschheuser et al. 2017a; 2017b). In particular, our findings recommend the implementation of gamification features that influence group norms, invoke positive emotions when cooperating, and increase a member’s identification with a group, in order to induce we-intentions and cultivate cooperation. First, designers and managers of communities should consider increasing the overlap of individual goals by carefully intertwining goals (El-Nasr et al. 2010; Rocha et al. 2008) or defining shared goals, such as team challenges or missions that can be completed by a group of users (Morschheuser et al. 2017b), to strengthen group norms. Second, the implementation of features that provide motivational rewards, such as the experience of mastery, competence, and social relatedness (Ryan et al. 2006), for achieving shared goals should be considered so as to increase the anticipated emotions of joint actions. Third, those seeking to cultivate cooperation should work on the representation and attractiveness of groups (cf. Tsai & Bagozzi. 2014), and should implement features that can trigger self-reflection processes (Bagozzi & Lee 2002; Ellemers et al. 1999) in order to support the shaping of members’ identification with a group. Cooperative games such as Ingress, Pokémon Go, or World of Warcraft provide a treasure of examples of concrete design features that can easily be transferred to other environments. For instance, Ingress uses competing factions and defines the victory of the own team as a clear shared goal for all players in a faction. Several game features, such as leaderboards that visualize the regional supremacy of a faction and its members, as well as story and media items that explain and motivate the shared goals, are implemented to support the cooperative goal structures and may influence group norms in factions, as well as in their locally organized subgroups. To support members’ identification with a group, we recommend the use of uniform colors and logos that can convey a clear group affiliation, statistics for direct comparison between players that enable social categorization (Turner 1975), and visual representations of users by avatars that can initiate self-reflection processes (Bessière, Seay, & Kiesler 2007; Christoph, Dorothée, & Peter 2009), similar to Ingress. Examples such as virtual goods that have been found to influence a user’s social identity (Huang 2012) and could be used in some games as gifts to express commitment to other players, have already found their way from games (Hamari 2015b; Hamari & Keronen 2017a) into online communities, such as Facebook (Huang 2012). Further, our findings imply that practitioners who seek to increase cooperation should find that gamification inspired by cooperative games is more beneficial and preferable to individual-based gamification. We found that individualistic game features can have negative effects on we-intention, which in our study have been compensated through small positive effects on group dynamics. Individualistic game features are designed to motivate personal goals and try to promote egotistical behavior (Morschheuser et al. 2017b). Previous research in the context of sports and education (for a review, see Johnson & Johnson 1989, p. 173) indicate that mixing cooperative goal structures with individualistic goal structures can negatively affect the relationships between a group’s members, since egotistical motives and behaviors may harm cooperative efforts and trust among group members. Such aspects could explain the negative relationships between individualistic game features and we-intention. On the other hand, our findings revealed that individualistic game features are positively related with positive anticipated emotions to play with others and joint commitment. We suspect that intertwining personal goals with the goals of others (El-Nasr et al. 2010; Rocha et al. 2008) – such as in Ingress – may increase the desire to play together with others when playing individually. For the same reason, players who are committed to their private goals may also be committed to their team’s goal if it helps them to succeed in a game with intertwined goals. Overall, our findings indicate that practitioners should always carefully consider possible side-effects of mixing different game feature types (Morschheuser et al. 2017b). Further, cooperation cultivators should seek to design and implement gamification inspired by cooperative games instead of implementing individualistic game features, such as private badges or level systems, which are currently often used in communities and other collaborative technologies (Morschheuser, et al. 2016; 2017a; 2017b; Hamari et al. 2014). By investigating Ingress, which is not only an augmented reality game, but also a crowdsourcing approach, this study provides particularly valuable insights for gamifying crowdsourcing initiatives and related systems where people work together to achieve common outcomes (Morschheuser et al. 2016; 2017a; Prestopnik & Tang 2015). Our findings reveal that the implemented game features in Ingress represent an effective approach to cultivate strong and persistent cooperation. More than 73% of the respondents reported that they play Ingress for at least a year, and more than the half specified that when playing actively, they use the app multiple times per day. One possible reason for this strong engagement could be the great variety of applied game features. A recent review of gamified crowdsourcing systems revealed that crowdsourcing approaches, which strive for collaborative, emergent outcomes, may benefit from implementing manifold gamification approaches (Morschheuser et al. 2016; 2017a). By providing the opportunity to satisfy a broad spectrum of needs, the use of manifold gamification designs can engage broad target groups. Accordingly, managers and designers of crowdsourcing communities should consider applying manifold game design features when designing incentive mechanisms. On the other hand, for the reasons discussed above, designers should carefully consider possible effects of mixing cooperative, individualistic and competitive game features. 4.2 Limitations and future research Similar to other studies conducted by online surveys, a shortcoming of this research is that the data were self-reported and that the participants were self-selected. Respondents of such surveys might not represent the entire population of a service or game, but might represent a population that shares a strong positive affinity to the object under investigation. In the context of our study, it can safely be assumed that the survey participants were highly engaged with the game and that less active users may be underrepresented in the sample. Thus, our findings may have neglected less active and less engaged players. This issue might be addressed in future research that particularly examines the intentions of less active players or players with little experience in cooperative games. Previous research indicates that less experienced users are inclined to follow suggestions and beliefs of those with whom they share goals rather than rely on their own knowledge, which may influence their group norms (Shen et al. 2011). On the other hand, experienced users are more likely to develop closer relationships with other group members and therefore perceive a stronger social identity with a group (Bagozzi & Dholakia 2006). The study results show a more substantial relationship between cooperative gamification features and social identity as opposed to group norms. Future research could consider the effects of players’ experience on group dynamics and we-intentions. Further, longitudinal studies will be needed to determine the psychological and behavioral outcomes of cooperative game features in relation to players’ lifecycles. Also, this study was conducted in the context of a particular cooperative game that seems to have implemented a set of successful game features that can invoke and cultivate weintentions. However, these identified and investigated cooperative game features strongly depend on the considered context. Although there are no a priori reasons to assume that our study context has influenced the findings on how games induce we-intentions, the results might be somewhat context-dependent. For instance, the strength of the effects between cooperative game features, group dynamics, and we-intentions may vary depending on the context and the applied cooperative game features. Future research that compares our study findings across different contexts could add to the generalizability of our results. Further, future research should try to increase the robustness of findings presented in this paper by combining de facto usage data with survey-based measured intentions and perceptions. Finally, we encourage researchers to refine the understanding of moderating influences that may impact the cultivation of cooperation in games and gamified systems. Prior research in the context of gamification indicate that demographic and personality traits moderate perceptions and intentions of game features (Hamari & Koivisto 2015a; Tondello et al. 2016). On the other hand, moderators such as group size (Brewer & Kramer 1986) or culture (Tsai & Bagozzi 2014) may moderate group dynamics and cooperative behaviors. In the special context of anticipated emotions, the moderating effects of gender and the extent of personal investment might be a reason why we have found no significant association between negative anticipated emotions and we-intentions among mostly male participants (cf. Bagozzi & Dholakia 2006; Taylor, Bagozzi, & Gaither 2005). We thus call for additional research into possible moderating effects. Future research that expands the understanding of factors that influence cooperation in games will not only contribute to a better understanding of this phenomenon, but will also support the design of gamification for cultivating cooperation in information systems and organizational contexts. References Ajzen, I. 1991. “The theory of planned behavior,” Organizational Behavior and Human Decision Processes (50:2), pp. 179–211. Anderson, J. C., and Gerbing, D. W. 1988. “Structural equation modeling in practice: A review and recommended two-step approach,” Psychological Bulletin (103:3), pp. 411–423. Arazy, O., Gellatly, I., Brainin, E., and Nov, O. 2016. “Motivation to share knowledge using wiki technology and the moderating effect of role perceptions,” Journal of the Association for Information Science and Technology (67:10), pp. 2362–2378. Armitage, C. J., and Conner, M. 2001. “Efficacy of the theory of planned behaviour: A metaanalytic review,” British Journal of Social Psychology (40:4), pp. 471–499. Bagozzi, R. P., and Dholakia, U. M. 2002. “Intentional social action in virtual communities,” Journal of Interactive Marketing (16:2), pp. 2–21. Bagozzi, R. P., and Dholakia, U. M. 2006. “Open Source Software User Communities: A Study of Participation in Linux User Groups,” Management Science (52:7), pp. 1099–1115. Bagozzi, R. P., and Lee, K.-H. 2002. “Multiple routes for social influence: The role of compliance, internalization, and social identity,” Social Psychology Quarterly (65:3), pp. 226–247. Bessière, K., Seay, A. F., and Kiesler, S. 2007. “The ideal Elf: Identity exploration in World of Warcraft,” CyberPsychology & Behavior (10:4), pp. 530–535. Beznosyk, A., Quax, P., Coninx, K., and Lamotte, W. 2012. “The influence of cooperative game design patterns for remote play on player experience,” in Proceedings of the 10th Asia Pacific conference on computer human interaction - APCHI’12, Matsue-city, Shimane, Japan, pp. 11–19. Björk, S., and Holopainen, J. 2005. Patterns in Game Design, Boston, Massachusetts: Charles River Media. Blohm, I., Bretschneider, U., Leimeister, J. M., and Krcmar, H. 2011. “Does collaboration among participants lead to better ideas in IT-based idea competitions? An empirical investigation,” International Journal Networking and Virtual Organisations (9:2), pp. 106– 122. Bratman, M. E. 1997. “I Intend that We J,” in Contemporary Action Theory, G. HolmstromHintikka & R. Tuomela (eds.), Dordrecht, Netherlands: Kluwer, pp. 49–63. Brewer, M. B., and Kramer, R. M. 1986. “Choice behavior in social dilemmas: Effects of social identity, group size, and decision framing,” Journal of Personality and Social Psychology (50:3), pp. 543–549. Brown, J. S., and Thomas, D. 2006. “You play World of Warcraft? You’re hired!” <https://www.wired.com/2006/04/learn/> (Accessed February 11, 2017). Bui, A., Veit, D., and Webster, J. 2015. „Gamification – A Novel Phenomenon or a New Wrapping for Existing Concepts?” in Proceedings of the 36th International Conference on Information Systems – ICIS’15, Fort Worth, Texas, USA, pp. 1–21. Chen, C.-H., Sun, C.-T., and Hsieh, J. 2008. “Player Guild Dynamics and Evolution in Massively Multiplayer Online Games,” CyberPsychology & Behavior (11:3), pp. 293–301. Chen, Y., and Pu, P. 2014. “HealthyTogether: exploring social incentives for mobile fitness applications,” in Proceedings of the Second International Symposium of Chinese CHI on - Chinese CHI’14, Toronto, Ontario, Canada, pp. 25–34. Cheung, C. M. K., Chiu, P., and Lee, M. K. O. 2011. “Online social networks: Why do students use Facebook?” Computers in Human Behavior, (27:4), pp. 1337–1343. Cheung, C. M. K., and Lee, M. K. O. 2010. “A theoretical model of intentional social action in online social networks,” Decision Support Systems (49:1), pp. 24–30. Sheng, L. Y. 2013. “Modelling learning from Ingress (Google’s augmented reality social game),” in Proceedings of the 63rd Annual Conference International Council for Education Media - ICEM’13, Singapore, pp. 1–8. Siu, K., Zook, A., and Riedl, M. O. 2014. “Collaboration versus competition: Design and evaluation of mechanics for games with a purpose,” in Proceedings of the 9th International Conference on the Foundations of Digital Games - FDG’14, Lauderdale, Florida, USA, pp. 1–8. Smith, J. 2015. “A Million People Around You Are Playing an Alternate Reality Game You Can’t See,” <https://mic.com/articles/119366/one-million-people-around-you-are-playing-analternate-reality-game-you-can-t-see> (Accessed February 11, 2017). Stets, J. E., and Burke, P. J. 2000. “Identity theory and social identity theory,” Social Psychology Quarterly (63:3), pp. 224–237. Straub, T., Gimpel, H., Teschner, F., and Weinhardt, C. 2015. “How (not) to Incent Crowd Workers,” Business & Information Systems Engineering (57:3), pp. 167–179. Tajfel, H. 1982. “Social psychology of intergroup relations,” Annual Review of Psychology (33:1), pp. 1–39. Takahashi, D. 2014. “Google's mobile game Ingress enables 7M players to create usergenerated missions,” <http://venturebeat.com/2014/09/25/googles-mobile-game-ingressenables-7m-players-to-create-user-generated-missions/> (Accessed August 27, 2016). Taylor, S. D., Bagozzi, R. P., and Gaither, C. A. 2005. “Decision making and effort in the self-regulation of hypertension: Testing two competing theories,” British Journal of Health Psychology (10:4), pp. 505–530. Teng, C.-I., and Chen, W.-W. 2014. “Team participation and online gamer loyalty,” Electronic Commerce Research and Applications (13:1), pp. 24–31. Thom, J., Millen, D. R., Dimicco, J., and Street, R. 2012. “Removing Gamification from an Enterprise SNS,” in Proceedings of the ACM 2012 conference on Computer Supported Cooperative Work - CSCW’12, Seattle, Washington, USA, pp. 1067–1070. Tondello, G. F., Wehbe, R. R., Diamond, L., Busch, M., Marczewski, A., and Nacke, L. E. 2016. “The Gamification User Types Hexad Scale,” in Proceedings of the 2016 Annual Symposium on Computer-Human Interaction in Play – CHI Play’16, Austin, TX, USA, pp. 229-243. Tsai, H.-T., and Bagozzi, R. P. 2014. “Contribution Behavior in Virtual Communities: Cognitive, Emotional, and Social Influences,” MIS Quarterly (38:1), pp. 143–163. Tsai, H.-T., and Pai, P. 2014. “Why do newcomers participate in virtual communities? An integration of self-determination and relationship management theories,” Decision Support Systems (57:1), pp. 178–187. Tuomela, R. 1995. The importance of us: A philosophical study of basic social notions, Stanford, CA: Stanford University Press. Tuomela, R. 2000. Cooperation, Dordrecht: Springer-Science+Business Media, Vol. 82. Tuomela, R. 2011. “Cooperation as joint action,” Analyse & Kritik (33:1), pp. 65–86. Turner, J. C. 1975. “Social comparison and social identity: Some prospects for intergroup behaviour,” European Journal of Social Psychology (5:1), pp. 1–34. Valacich, J. S., Dennis, A. R., and Connolly, T. 1994. “Idea Generation in Computer-Based Groups: A New Ending to an Old Story,” Organizational Behavior and Human Decision Processes (57:3), pp. 448–467. Velez, J. A., and Ewoldsen, D. R. 2013. “Helping behaviors during video game play,” Journal of Media Psychology (25:4), pp. 190–200. Vesa, M., Hamari, J., Harviainen, J. T., and Warmelink, H. 2017. “Computer games and organization studies,” Organization Studies (38:2), 273-284. Wasko, M. M., and Faraj, S. 2005. “Why Should I Share? Examining Social Capital and Knowledge Contribution in Electronic Networks of Practice,” MIS Quarterly (29:1), pp. 35– 57. Wu, J., and Liu, D. 2007. “The effects of trust and enjoyment on intention to play online games,” Journal of Electronic Commerce Research (8:2), pp. 128–140. Yee, N. 2006. “Motivations for Play in Online Games,” CyberPsychology & Behavior (9:6), pp. 772–775. Zagal, J. P., Rick, J., and Hsi, I. 2006. “Collaborative games: Lessons learned from board games,” Simulation & Gaming (37:1), pp. 24–40. Zhang, X., Venkatesh, V., and Brown, S. A. 2011. “Designing Collaborative Systems to Enhance Team Performance,” Journal of the Association for Information System (12:8), pp. 556–584. Zhao, Y., and Zhu, Q. 2014. “Effects of extrinsic and intrinsic motivation on participation in crowdsourcing contest,” Online Information Review (38:7), pp. 896–917. Zhong, Z.-J. 2011. “The effects of collective MMORPG (Massively Multiplayer Online RolePlaying Games) play on gamers’ online and offline social capital,” Computers in Human Behavior (27:6), pp. 2352–2363. Appendix A A.1. Questions asked to determine the nature of a game design feature in Ingress: I find that this game feature … («Description of the functionality») … is of individual nature (5-point Likert scale: strongly disagree - strongly agree) … is of competitive nature (5-point Likert scale: strongly disagree - strongly agree) … is of cooperative nature (5-point Likert scale: strongly disagree - strongly agree) A.2. Results of the Ingress feature categorization: Identified game feature of Ingress Individualistic game features Cooperative game features Competitive game features Factions X* Action Points (AP) X Mind Units (MU) X* Agent level X Medals X Mission badges X Agent stats X Deploy Resonators X* Recharge Resonators X* COMM (in-game chat) X* First Saturday (FS) events X XM Anomalies X* Personal avatar X Weapons X Power Cubes X Mods X* Playing of missions X Mission days (x) X Hacking of portals C C C Attacking portals X Takeover portals X* Upgrade portals X* Checkpoints and cycles X* X = primary perceived category of the game feature. (x) = secondary perceived category of the game feature. A minority of experts perceived this feature as part of the corresponding category. * = features that were perceived as having both competitive traits (on an intergroup level) as well as cooperative traits (on an intragroup level). For such cooperative-competitive features (e.g. factions), we sought to carefully identify the cooperative aspects before developing the corresponding survey items. C = feature that was perceived as a core game mechanic of the game. Thus, no clear assignment to one of the feature categories could be made.