The Alternate Reality Game (ARG) remains a topic of great interest to me and to my students at MIT and USC. Through the years, we’ve discovered that the ARG falls at the intersection between our recurring interests in participatory culture, collective intelligence, new media literacies, transmedia entertainment, and civic engagement. In my Fandom, Participatory Culture, and Web 2.0 graduate seminar last spring, Jeff Watson wrote a provocative essay which reviewed and challenged the current state of ARG theory and design, proposing some of the limits of this still emerging genres, as well as identifying some experiments that stretch the ARG in new directions. I immediately knew that I wanted to share this essay with my readers, who have a range of different investments in this space, in hopes that it might serve as a catalyst for enlarging the conversation around ARGS and might give him useful feedback as he hopes to prepare this essay for publication.
Watson comes at this topic as a student in the USC’s Cinema School’s innovative iMAP program, which is designed to bring together students who are interested in both media design and theory. I am going to be teaching a seminar through the program this fall on Medium Specificity, and will be sharing the syllabus here shortly. Each of the students I have met through this program have impressed me with their creative insights and their willingness to test their ideas through experimental practice. The Cinema School as a whole is exploring how to break down the silos between theory and production and between the different craft specializations within production, because the media maker of the future will need to think and create across media platforms. This is yet another of the many reasons I am excited about being at USC right now.
by Jeff Watson
As marketing instruments, alternate reality games (ARGs) are powerful tools for generating buzz and fostering audience engagement. Their capacity to initiate and maintain playful and creative dialogue between producers and fans signals the immanence of interactive and participatory transmedia entertainment. However, the established structure of the ARG as a time-sensitive and event-driven experience managed by the behind-the-scenes machinations of “puppet masters” (PMs) forecloses a number of important design possibilities. Consequently, ARGs often lack the qualities of accessibility, replayability, and scalability that are so crucial to the adoption and impact of other kinds of socially-articulated interactive systems. In instances where the objective is to create or engage an elite class of “in the know” participants, such a lack may in fact be a strength; but for other use cases, accessibility, replayability, and scalability are critical. This paper outlines the significance of the absence of these characteristics from many “first generation” ARGs, and points toward an emerging “2.0” iteration of the form through reference to recent projects and practices in both industrial and institutional contexts.
In contrast to more capacious terms such as “pervasive game” or “big game,” the term, “alternate reality game,” refers to a very specific and well-defined form of interactive transmedia storytelling that “[takes] the substance of everyday life and [weaves] it into narratives that layer additional meaning, depth, and interaction upon the real world” (IGDA ARG SIG 2006). In an ARG, players discover the game through an encounter with one or more access points embedded in real world contexts. These access points, known in the parlance of ARGs as “rabbit holes,” lead players into a dynamic matrix of story components distributed across various kinds of digital and physical media.
By exploring these components, players discover puzzles and challenges that serve both as impetus to connect with other players, and as time- and context-sensitive content bottlenecks. In order to advance the narrative, players typically work together, first assembling into affinity groups via both official (i.e., game-sanctioned) and unofficial (i.e., player-created) social media structures; then tackling puzzles and challenges collectively, leveraging the range of competencies, geographies, and biographies inherent in a large and distributed player base. As puzzles are solved, the game’s content producers, or “puppet masters,” release successive cycles of story and interactivity, tweaking their approach along the way based on the observed behavior and emerging collective intelligence capabilities of the players.
This process repeats itself until the narrative concludes, typically with the launch of a product or service. At this point, official support for the player community is usually terminated, primary online game assets are deleted or otherwise rendered inactive, and the game ends.
Early participants and producers of ARGs compared their emergence to watershed moments in pop music (Phillips 2010) and cinema, with some going so far as to suggest that the ARG is effectively the defining narrative mode of our present communications epoch (Stewart 2004). Indeed, especially in the context of the early 2000s, ARGs represented a uniquely transmedial mode of interactive storytelling, and as such were seen as being natural and inevitable outgrowths of the burgeoning network culture. When playing an ARG, fans consume story in a variety of modes, via a range of devices, channels, settings, and practices. This platform-independent nonlinearity and fragmentary or distributed consumption-participation pattern was seen as a logical outcome of millennial shifts in media habits, and was referenced by various futurists as a model for how stories and games could be designed in the coming era of ubiquitous computing and social media (Gibson 2005; Vinge 2006).
Other observers, invested in visions of participatory and collaborative storytelling, noted that, unlike typical consumers of cinema, television and other few-to-many media forms, the players of ARGs are always already necessary and constitutive elements of the work. That is, in an ARG, audience participation is not a byproduct, but rather an essential and formative component of the text. To practitioners and theorists with a stake in participatory culture, the notion of an interactive storytelling form conceived from the ground up as a means of facilitating the collaborative production of media artifacts provided a “perfect illustration of all of the principles . . . shaping the media landscape at the present time” (Jenkins 2006).
Further, ARGs were viewed as fitting into a long tradition of spatially- and temporally-distributed narrative forms, and for some, their emergence indicated the arrival into the mainstream of practices that had hitherto been relegated to fan subcultures and marginal art movements. Like the critical interventions of Situationism, which sought to reconfigure public space as a “new arena for creation” wherein “unforeseen games will become possible through the inventive use of material conditions” (Nieuwenhuys, 1959), the ambiguously-bounded play of ARGs has the ability to produce dramatic shifts in subjectivity that “[sensitize] participants to affordances, real or imagined,” “[make] all data seem connected, or at least plausibly connected,” and “make surfaces less convincing” (McGonigal 2003, 43-44). Similarly, ARGs promised to do to mainstream storytelling what “distributed narratives”- experimental narratives spread out across “time, space, and the network” (Walker 2004, 1) – had done to avant garde and electronic literature:
Distributed narratives break down the aesthetics of unity we have followed for millennia. They take this disunity a step further than the bricolage of postmodernism, by collapsing the unity of form as well as that of content and concept. Yet perhaps they also point to a new kind of unity: a unity where the time and space of the narrative are in sync with the time and space of the reader. (Walker 2004, 11)
Finally, by bringing together once disparate practices such as game design, performance art, and cinematic narrative, ARGs were seen as being on the cutting edge of interdisciplinary new media thinking. Great things were forecasted, including the use of ARGs in establishing and leveraging collective intelligences in order to solve real-world problems (McGonigal 2003; Jenkins 2004).
IV. ARGs in Practice: 2001-2010
While ARGs have proven that they have the potential to mobilize elite groups of “lead users” who can co-create content and evangelize for a brand or cause (McGonigal 2003; Szulborski 2005; IGDA ARG SIG 2006; Dena 2008a) – and that they can quickly generate alarmingly efficient collective intelligences (McGonigal 2007; see also DARPA 2010) – they have, perhaps understandably, failed to live up to some of the high expectations set at the turn of the century. ARGs have not seen the kinds of growth in popularity that other forms of gaming and storytelling have seen over the past decade (Schell 2010; compare with Dena 2008a); they have not proven to be a particularly effective way of building lasting communities or collaborative practices, especially when compared to more systems-oriented approaches to organizing and maintaining collective action (see Shirky 2008); and they have failed to maintain the same kind of relevance to contemporary media habits and technologies that they arguably held in the early 2000s, ceding this territory to other kinds of pervasive interactivity such as mobile and social media games, casual games, and collaborative production games (see Montola, et al 2009).
The specific reasons for these shortfalls vary from context to context. In the media industry, for example, ARGs have largely been considered marketing tools, and as such have often not been sufficiently integrated into the development and production processes of the properties they promote, leading to disconnects between fans of the source material, ARG player communities, PMs, and producers. As writer Rich Silverman of the Transmedia LA message board (2010) puts it, “[it’s] been my experience that an ARG component of a film or TV property comes to the game too late to be really effective . . . we need to start seeing these things baked into the development process of any show or film they’re supporting.” In educational and institutional contexts, ARGs are similarly marginalized, typically employed as orientation tools (Down 2008) or experimental promotions (Goodlander 2009), but rarely meshed with the core operations or mandates of their hosts.
It is unclear whether the persistent design problems that constrain or preclude the accessibility, replayability, and scalability of ARGs are the cause or the result of this marginalization. Nevertheless, making such a determination is probably less important than identifying what those core design problems are, and suggesting ways in which the form can evolve or adapt in order to correct them. In general, these problems center on three overlapping and relatively unchallenged aspects of traditional ARG design, namely: 1) that, despite the decidedly playful and improvisatory character of the relationship between puppet masters and players, ARGs are ultimately not game systems but rather vehicles for delivering story; 2) that ARGs treat their core audiences as monadic “collective detectives” rather than groups of living, breathing individuals; and, 3) that ARGs are linear, event-driven experiences.
Indeed, many of the problems associated with ARGs can be traced back to their status as temporally-bounded and sequentially-unfolding experiences. As Jim Stewartson of Fourth Wall Studios puts it, “[ARGs have historically been] essentially rock concerts. Very large, real-time, elaborate experiences that were really cool and really fun for the people who were involved with them” (Morris, et al 2009). This event-like design clearly eliminates any potential for replayability, and it exacts almost equally dire consequences on accessibility and scalability. In a typical ARG, players who don’t have the time at the right time to dive into the game can find their experience spoiled by those who do. Even players with high levels of interest in the material and a strong desire to participate in the game’s challenges can be reduced to lurking on message boards or merely following along with puppet master- or player-created story summaries if they don’t have the time required to keep up with the more hard-core players. Consequently, the vast majority of the players of traditional ARGs aren’t “players” at all, but are rather more like spectators, albeit very multi-modal ones:
Of the millions of people who ‘experience’ an ARG only tens of thousands actually play them, the rest read the texts created by players. Now, as I have stated many times before, this is a very interesting model of audience tiering and shows a preference for player-created narratives above producer-created ones (indeed, the desire for a linear narrative above a fragmented one)…but the large numbers often claimed . . . are not indicative of the people who actually play these forms. They are hardcore games that only a (relatively) small amount of players can actually play directly (due to skill, time and access obstacles). I don’t see how a form with such accessibility issues is the ultimate form. (Dena 2007)
Dena (2008b), Montola (2009), and others have pointed out that this “pyramid of participation” enables transmedial designs wherein “different play modes contribute to each other and support an experience that is larger than its parts” (Montola, et al 2009, 121). In such an arrangement, spectators co-exist with variously-engaged players, with the hard-core participants effectively acting as “stars” of the game’s narrative ; puppet masters and serious players document the actions of the hard core in real- or near-real-time; and the rest of the player base consumes this documentation serially. This kind of structure has been experimented with to varying degrees of success.
However, since this and other kinds of “tiering” (Dena 2008b) demand the production and management of numerous additional layers of game assets, they also represent some of the most expensive and labor-intensive solutions to the accessibility problem.
Another aspect of the accessibility problem emerges from of the manner in which ARG designers traditionally address their players. As Sean Stewart notes in an interview with members of The Cloudmakers (2001), “[the] premise from Day One was that the entire Internet should be considered as a single player; that we could put an ad in a newspaper in Osaka in the morning and have some kid in Iowa using that information by supper time.”
That is, while individual players in an ARG are free to privately interact with characters or artifacts from the game, the puzzles and challenges are designed with such complexity that any information gathered from these interactions needs to be shared with and processed by a collective in order to be properly contextualized and rendered sensible in a timely fashion. While this design encourages the formation of collective intelligences, in the context of a time-based, event-driven, closed information system such as an ARG, it also somewhat counter-intuitively results in an increasing diminishment of the degree to which new players can easily access and enter into the activity.
That is, once a functioning “collective detective” (Cloudmakers 2001) has been established, it will tackle the challenges presented by the puppet masters with a self-refining efficiency that will largely discount the need for new members. Knowledge production structures populated by elite players with available time, an appropriate range of competencies, and relevant social capital will gather, process, and analyze data faster and more thoroughly than a non-integrated outsider ever could. Further, as the game progresses, prospective members without adequate reputation within the player community and in-depth knowledge of “the story so far” (see Dena 2008b, 41) will naturally find it increasingly difficult to find a role within the collective.
To illustrate this problem, consider the recent DARPA Network Challenge crowdsourcing experiment (2009). In this experiment, ten red weather balloons were placed in visible locations around the United States, and the public was challenged to find the balloons using any legal means whatsoever. Nine hours after the event commenced, all ten balloons had been found by a team from MIT (http://balloon.mit.edu/). In this instance, the team, which had conscripted around 5,400 balloon spotters via social media and various public entreaties (DARPA 2010), served its purpose and was quickly dissolved; but what if the DARPA Network Challenge had been only the first of many challenges in a long-term game – that is, if it was merely the opening puzzle of a three month long ARG. How would this emerging collective intelligence have evolved? Would it have become more broad-based like Wikipedia, exploring the diverse interests and passions of its user base, or would it have gravitated toward greater efficiencies, tighter working groups, task-oriented committees, and editorial sub-teams?
According to fieldwork conducted by McGonigal (2007), the latter is more likely: rather than becoming more inclusive or expansive, the group might in fact become increasingly specialized along particular “threads of investigation” tied to the core problems with which it was presented. After all, the puzzles in ARGs are ultimately very specific: unlike Wikipedia, which is almost completely open-ended, the knowledge production demanded by an ARG is focused on a particular story world and an associated set of puzzles with clearly-defined solutions – much like the narrow-but-complex balloon-finding task of the DARPA experiment. Further, since the puzzles in ARGs are often cumulative and informed by the solutions to earlier puzzles, those who were on board for the first discoveries – in this case, those who understood the methodology by which the original 5,400 balloon spotters were coordinated and the information they provided was processed – would arguably be more valuable and acceptable assets to the team than newcomers unaware of those practices and procedures. Somewhat ironically, then, this kind of collective intelligence design, when applied to closed information systems such as ARGs, has steeply diminishing returns when it comes to inclusivity and accessibility.
What all of these problems have in common is an origin in the “non-gameness” of ARGs. ARGs, despite their name, are not, in fact, games; rather, they are ergodic (Aarseth 1997) transmedia texts that, structurally speaking, are much more akin to scavenger hunts or group puzzle-solving activities like the annual MIT Mystery Hunt. Rarely in ARG design do we see the generativity, rulesets, and procedural rhetoric (Bogost 2007) that characterize games.
This is fine; not everything has to be a game. But for an activity which so often aspires to take place on a massive scale, both in terms of content and participation, to not use game mechanics as a means of generating and managing interactivity is arguably a recipe for disaster. Indeed, much of ARG design is reminiscent of early experiments in electronic literature and interactive filmmaking which sought to create vast narratives via branching story trees: very quickly, artists who took this approach discovered that to do so meant writing or shooting orders of magnitude more material than the reader or viewer would ever see.
An interactive movie-game like Dragon’s Lair (1983), for example, needed a total of 27 minutes of animation stored on custom-made laserdiscs to provide an interactive experience that lasted for a maximum of 6 minutes (The Dot Eaters 2007) – and even then, the gameplay consisted of little more than making a handful of left-or-right decisions about which direction the protagonist should move. Compare this outcome to an even older video game, Rogue (1980), a procedurally-generated dungeon-crawler that remains popular to this day. In Rogue, the virtual world is generated on the fly at runtime via an algorithm. Instead of devoting limited computational resources to storing and displaying pre-rendered content (as in Dragon’s Lair), Rogue‘s programmers used a compact ruleset to create their gameworld, producing an expansive and endlessly replayable realm of fantasy adventure and tabletop RPG-style interactivity that would have been technically impossible to produce using pre-made dungeon scenarios given the limited storage resources of early 1980s home computers. Despite being made for free by hobbyist programmers, Rogue‘s parsimonious use of algorithms rather than branching content trees resulted in much more interactivity and depth than was presented three years later by Dragon’s Lair‘s spectacular but simplistic left-or-right decision making interface. This is the real power of games: to create dynamic interactive experiences through rules rather than archives of pre-made content. As we shall see below, approaching ARG design from this perspective opens a range of new possibilities for producers.
Finally, because ARGs are so expensive and labor-intensive to maintain, media companies and institutions overwhelmingly abandon the communities they create once the putative purpose for their creation has been satisfied (McGonigal 2003; IGDA ARG SIG 2006). While this instrumental view of community may have short-term benefits to brands and creatives, and while many media companies are likely comfortable with the risk of “blowback” from disaffected ARG fans (especially since said fans will have long since served their purpose by the time their complaints come to the fore), in the long term, such a view effectively undermines one of alternate reality gaming’s most important potentials for generating value: the creation and maintenance of strong, persistent communities.
Jeff Watson is an interdisciplinary media practitioner with a background in screenwriting, filmmaking, and game design. His doctoral research in Media arts and Practice at the University of Southern California’s School of Cinematic Arts focuses on investigating how ubiquitous computing and social media can enable new forms of storytelling and civic engagement. For more insights from Jeff Watson, you can check out his website or follow his Twitter flow.