Massively Multiplayer Online Games and Situated Learning: an Adaptive Model for Community Development


Games as Tools for Practical Problems


Community Development


Our aim is to demonstrate how we can use the social and technological affordances of online virtual communities, specifically Massively Multiplayer Online Games, to effect learning – and in turn to enable community development. This will be accomplished by marrying the intuitive user interfaces and engaging play activities of the online game world with the power and sophistication of highly developed social/cognitive models. Players in these games will learn through situated activities in a realistic physical and social environment; behavior resulting from that learning will effect change and growth in real world communities.


There are multiple ways to define community; for our purposes it will be useful to first decide whether we are speaking of communities in the physical world, or of virtual communities that exist only in the “world” of communication and computer networks.


Physical Communities


Communities in the physical world can be defined along two parameters: association in physical space, or association by commonality of non-spatial traits such as values, interests, business, religion, etc. Communities of the latter type are often referred to as communities of interest. Association implies that members of a community are in a dynamic and interdependent relationship, where actions of individuals can and do affect other individuals.


Communities of interest in the physical world possess characteristics that make them suitable as bridges to virtual communities. Although members of a community of interest might meet in physical space, e.g. at a movie fan convention or software developers’ conference, or even live together as in a golf neighborhood, such physical connection is not strictly necessary. They might only communicate at a distance, yet still feel they are a connected part of their community. Thus, we can say that the community of interest can form a link between the basic community types: it can have associations in physical space and also associations defined solely by communications between members separated in space. Very often it will possess both, which makes it useful for our purposes.


Virtual Communities


Like physical communities, communities in the virtual world can also be defined along two parameters of association. One parameter we have already encountered: association by commonality of interests. The other parameter, and the defining one for virtual communities, is association defined solely as networks of connections that are themselves riding on networks of interlocking computer and communication systems. This kind of association is very different from geographical association in space, with great implications for the nature of virtual communities. In particular, the fact that virtual communities exist as interconnected communications nodes makes them amenable to the techniques of information technology in ways that physical communities never could be. This property makes a virtual community a likely candidate for types of community-driven scenario development that would be difficult or impossible to perform on physical communities. The question at this point is whether results obtained in a virtual setting can be generalized to the physical world. There is an existing body of knowledge about modeling and simulating community properties and behaviors. Such simulations have produced much significant knowledge about the nature of actual, physical communities, and we feel confident that we will be able to take those same precepts and practices into our MMOG community and emerge with valid results. Additionally, we believe that the bridging concept of communities of interest allows a strong mechanism for transferring social meaning between the physical and virtual domains.


Most virtual communities tend to be communities of interest, and the ubiquitous nature of global communications, and especially the Internet, means that such communities are open to membership by persons living almost anywhere. These virtual communities of common interest span the spectrum from small fan clubs of a dozen or fewer people to some of the larger MMOG communities, which can have tens or hundreds of thousands of members. While individual members may come and go, virtual communities can be remarkably persistent over long time periods if they suit the needs of their constituents. e.g. the well ( Such long-lived virtual communities will experience growth and change over their lifetimes in much the same way that physical communities might.


While all virtual communities share the common quality of mediation through computer and communication networks, the actual underlying technologies supporting them can be remarkably diverse, as can the types of communities enabled by those technologies. We can consider four main virtual community technologies (Kim, A. J. 2000):


Mailing lists are the simplest and easiest forms of virtual community. Network-based mailing list software facilitates the basic member-to-member interchange that is necessary for any community to develop. Mailing lists are appealing to novice Internet users, because of the ready familiarity of email, and communities formed this way tend to be conversational in nature.


Message boards – also known as forums, discussions, bulletin boards or newsgroups – are, like mailing lists, asynchronous communities. Members do not have to be logged on at the same time to communicate, since members post messages that others can retrieve and reply to at a later time. Message board software includes facilities for filtering messages by subject, thus giving an element of structure beyond that available in mailing lists.


Real-time chat systems are an increasingly popular foundation for synchronous virtual communities. Although most are text-based, a chat system can be thought of like an extended conference call, where all participants are privy to all communications as they happen in real time. The immediacy and sense of presence available through chat systems are powerful reinforcers of community among users.


Virtual worlds is the technology that we will concern ourselves with the most. As the name implies, a virtual world utilizes graphics, animation and sound to create an interactive and navigable environment, in which the users are represented by graphical stand-ins called avatars. The suite of technologies deployed in a well-constructed virtual world creates an almost viscerally real feeling of “being there.” In particular, actions of the user are seen immediately as actions in virtual space by the user’s avatar, and this linkage of user and avatar allows a powerful sense of navigating and acting in the computer-generated space.


Virtual worlds typically employ one or more communication modalities: Chat, usually text but sometimes voice, enables members to communicate in real time. Message boards and email allow for asynchronous communication. Persistent postings in the world allow for announcements and important information to be placed where all can access it on a read-only basis. Well-implemented communications in a virtual world complement and enhance the sense of reality created by the graphics and interactivity, and can lead to a powerful feeling of place and community for its members. This can lead, in turn, to an increased level of participation by members, which will itself raise the world’s fidelity level in a positive feedback cycle.


While all virtual worlds deliver some degree of immersion and involvement, they differ as to the degree and the methods used to obtain immersion. Only recently some worlds were still offering a 2-dimensional world interface, but these have largely been supplanted by varying levels of 3-dimensional representation (none is actually 3-dimensional in the true sense). On the simpler end, such worlds as Whyvillle ( offer the user spaces that could be most aptly described as 2.5-dimensional. That is, although the space presented looks 3-dimensional, navigation within it is limited in such a way as to make it impossible to go behind objects, for instance. At the other end of the scale, worlds such as those created by Active Worlds ( or There ( are convincingly 3-dimensional, allowing users to wander at will through the computer-generated environments. It is important to note that greater visual and interactive fidelity does not necessarily lead to a greater sense of community; other factors can be just as important. For instance the quality of the communications channels, the social structures in place, the way in which the world is laid out for navigation, the other users present in the world – all can reinforce a strong feeling of community in a world. And in the case of Massively Multiplayer Online Game worlds such as Ultima Online ( or Star Wars Galaxies (, we have the additional complexities afforded by game play, which can create powerful user communities, as we shall see.


Entrepreneurship in BabylOnline: a Scenario


L10n355 (aka Nasim Begum) has a problem. She has Babel-on T-shirts piling up at the back of her shop at an alarming rate and no one, simply no one, is buying them anymore. What is worse, she is still paying three workers to produce those shirts. What should she do?


M00nr4k3r (aka Ahmet Aziz) also has a problem, but a different one. His Magma Lamps have become wildly popular, and it seems that everyone in BabylOnline wants one. His Babylsheks account is growing daily and he has just hired two extra workers to keep ahead of the demand for more and better Magma Lamps. M00nr4k3r’s problem is this: one of the designers he originally employed has just quit, stating that he intends to make similar lamps and sell them for less.


Meanwhile, out on the Street there is a party going on. It is the 6 months anniversary of the founding of BabylOnline and the place is all decorated for holiday; there is a even a carnival set up near the souk . The sysops have encouraged everyone registered with BabylOnline to log in today and the town is packed with hundreds of avatars, more than anyone can ever remember.


This gives L10n355 an idea. She goes to the souk and seeks out the carnival emcee, a tall male avatar with crystal eyes and glowing red neon hair. She makes a bargain – for 3 free Babel-on shirts and 5000 Babylsheks, the emcee will wear only those shirts during the carnival and mention L10n355 in his patter. Since essentially the entire population of BabylOnline is going to attend the carnival at some point, L10n355 feels like it is a good deal. Her hunch pays off and soon she is selling more shirts than ever before and has even expanded with a line of matching parasols.


M00nr4k3r is not so fortunate. His former employee’s lamps actually have more design flair than the Magma Lamps, and they are cheaper. He sees his share of the market gradually decline, until he lets all his employees go and then finally ceases to sell lamps. It is time to summarize the lessons learned from this experience and plan a new venture. He schedules a meeting with his two advisors at the Coffee Tree in TigrisTown.


During that debriefing, his advisors, post docs at the University of Texas McCombs School of Business and at the University of Texas IC² Institute, offer insights into how he might have been able to save his failing business. Although their avatars are whimsical, zoot-suited characters, the advisors are quite serious and they are concerned for their charges. They offer solid advice to further the entrepreneurial skills of these budding business people and they listen carefully to see how they can make the game world more effective. Later they write up a report and send it to the BabylOnline web site, along with some video clips of the party. The PI is in Singapore all week, but she logs into the web site daily to keep current with the project.



Our Theoretical Framework


Activity Theory


Our definition of community implies a dynamic and interdependent interplay between community members and the social constructs they create. Likewise, our emphasis on situated learning argues for a theoretical framework that will allow us to study actions and their contexts as whole units. Activity Theory seems to be a very good fit for our purposes. As developed by Soviet psychologists A.N. Leont’ev, Lev Vygotsky and others, Activity Theory replaces the dyadic subject-object relationship with a triadic subject-activity-object relationship, thus including a link that mediates the connections between subject and object (Leont'ev 1978). To Soviet psychologists, conventional Stimulus-Response psychology was seen as a rather passive view of human existence, and by contrast they presented their view of a human being as an active process, whose “inputs” and “outputs” could only be analyzed in terms of the activity of being. This emphasis on active processes makes it an ideal framework for analyzing systems, and indeed the theory has gained widespread currency within the field of Human/Computer Interface (HCI) design (Barthelmes and Anderson 2002) (Bødker 1991).


Activity theory is not a firmly fixed and established body of statements, and in some ways it is more like a set of heuristic rules than a predictive theory. One of the foremost current interpreters and practitioners of activity theory, and whose interpretations we find useful for our project, is the Finnish educational psychologist Yrjö Engeström, who has done much to expand the original model to include more complex activities and also networks of interconnected activities. In his view, there are three main principles involved in applying activity theory (Engeström 1996):


·         The entire activity system as the unit of analysis: An activity system integrates subject, object and instruments (material as well as linguistic and symbolic) into an integrated whole. This gives context to seemingly random events.

·         Historicity as the basis for classification: The notion of progress is redefined away from a linear path leading from “primitive” to “developed and civilized”; instead progress in seen in terms of mastery of complex processes. Additionally, the historical development of an activity system becomes an integral part of its interpretation.

·         Inner contradictions as the source of change and development: Activity systems are characterized by inner contradictions, and these are seen as a moving force behind disturbances and innovations, and eventually behind change and development in the system.


This set of guiding principles provides us with a very powerful tool for analyzing and modeling the complex systems that are communities, whether in the physical or in the virtual domain. In particular, we can deal with whole communities as they have developed historically, which is important for discussing the context of any learning and change that we may expect. Additionally, we have a method for accounting for such learning and change as a natural result of contradictions within, or added to, the community under study.


Situated Learning


Since we are ultimately concerned with learning and community development, we will focus our attention tighter on a theory of learning conceived within the context of activity theory: Situated Learning, as developed by Jean Lave and others (Lave and Wenger 1991) argues against the long tradition of learning researchers to consider learning as a process contained in the mind, while ignoring the lived-in world – the infamous mind-body dichotomy. Instead situated learning aims to create a model that encompasses both mind and the lived-in world as vital parts of the learning process (Lave 1996). Viewed this way, learning is an activity to be analyzed in terms of its dynamic interrelationships and inner contradictions, a perfect approach to apply for understanding learning in complex communities.


In their seminal book of 1991, Situated learning: Legitimate peripheral participation, Jean Lave and Etienne Wenger further refined the concept of situated learning to include Legitimate Peripheral Participation, a key idea that will help us explain how learning obtained in the virtual domain can be meaningfully applied in real life communities. In this scheme, learning is situated within Communities of Practice. Such a community “[implies] participation in an activity system about which participants share understandings concerning what they are doing and what that means to their lives and for their communities.” (Lave and Wenger, 1991, p. 98) Such a community has shared goals, knowledge, techniques and values – in other words, it can also be considered a community of interest, according to our classification scheme.


Lave and Wenger’s communities of practice are conceived as having various roughly-defined zones in which members reside and function, according to their level of experience and legitimacy. The center is occupied by the “old-timers”, those who know the ropes and have been around to witness historical changes in the community. These adepts or masters literally embody the collective knowledge of the community and perpetuate it through their actions and interactions. Moving away from the center, the community contains successive zones that are occupied by members with increasingly less experience and knowledge, until we get to the very edge, a zone into which newcomers are placed to begin their journey to the center. Acquisition of knowledge or learning in this situation is determined by a centripetal participation in the community of practice, moving ever closer to the center as knowledge and skills accrue, on the way to attaining “old-timer” status. Each zone passed through will have its own level of knowledge, proficiency or experience, and these are obtained through activity situated at that level. Legitimate peripheral participation implies access by newcomers to the community of practice, not just as observers, but as empowered members of that community, and with limited but real access to its resources. It also implies that continued participation creates a path to fully invested membership. This is a subject we will elaborate when constructing our model for transferring learning and experience from the virtual to real domains.


Why use games to address practical problems?


What is a game? This definition becomes remarkably hard to pin down exactly; it seems that everyone has a slightly different take on what a game is or what it means to play a game. Here are a few samples:


·         Games are an exercise of voluntary control systems, in which there is a contest between powers, confined by rules in order to produce a disequilibrial outcome (Avedon and Sutton-Smith, 1971, p. 405).

·         A game is a form of art in which participants, termed players, make decisions in order to manage resources through game tokens in pursuit of a goal (Costikyan 1994).

·         A game is a system in which players engage in an artificial conflict, determined by rules, that results in a quantifiable outcome (Salen and Zimmerman, 2004, p. 80).


The last definition is a deliberate effort by its authors to synthesize the definitions of several other authorities, so we can parse it further: Systems provide a powerful conceptual category for thinking about how games work, and they also fit very well into our activity theory approach. Players interact with the systems of a game in order to play the game; they are the subjects in our activity triad. Games are artificial; they maintain a border with the “real” world that is analogous to our separation of the physical and virtual domains. Games embody conflict, a contest of powers that can take forms ranging from competition to cooperation. Rules are crucial to games; they create the structure without which games could not exist. Games have a quantifiable outcome or goal; at the end players have a score or ranking of some kind, differentiating games from less formal play activities.


The Massively Multiplayer Online Game


The above definitions are for the most general case for what a game can be. Let us turn to the particular nature of the type of game we have chosen to use for our project: the Massively Multiplayer Online Game (MMOG), also known as the Massively Multiplayer Online Role Playing Game (MMORPG), since it actually is a type of Role Playing Game (RPG). The distinguishing features of an MMOG are very large numbers of concurrent players, who are logged into the game via distributed computer networks and who play with each other and with system-generated artifacts in real time. Importantly, the nature of game play changes as a result of actions taken by the players, allowing for complex feedback mechanisms capable of simulating the kinds of social systems we wish to study.


In any RPG (including MMOG’s), players choose to become characters that evolve and act within an ongoing, designed, narrative structure, and their play is determined by the abilities and qualities available to the characters they have chosen. The definitions given above all agree that a game tends to have some goal or quantifiable outcome; it is a win-loss situation in which the players are ranked according to some agreed upon criteria. RPG’s (and of course our special case of MMOG’s) differ in that they are open-ended, at least at the large scale, with no set final state – in fact, for the case of commercial MMOG’s it would mean financial disaster for the game to reach a final state. At first glance, this lack of a defined goal would seem to be at serious odds with any reasonable definition of games, but an RPG is, as its name implies, very definitely a game. A useful way to get around this seeming conundrum is to think about RPG’s generally, and MMOG’s specifically, at several levels. At the very highest level, it is true that there is no ending state and that the game can continue in principal forever. At lower levels, however, at the granular level of player involvement, things look very different. At this level players are engaged in conflicts, strategies, negotiations, etc. that do have meaningful and disequilibrial outcomes: the raid on the castle was successful or not, the quest to find the energy stone did or did not reach its goal, the blacksmith did or did not succeed in becoming a sword maker.


We can think about the structures and activities in MMOG’s as coming from two directions: The overarching design of the world is top-down, and that ultimately determines the kinds of activities that are available to the players. If the system design is flexible and robust enough, if it allows players a sufficiently large repertoire of actions, then another type of system organization will materialize. This bottom-up, emergent structure is the result of players’ actions and reactions combining and reinforcing in ways that simply could not be foreseen or planned. It is this rich field of emergent player behavior, in conjunction with top-down direction, that provides the social space for our proposed experiments in situated learning and adaptive community development.


Games, Models and Simulations


While we do not see gaming as being a replacement for existing processes of model and simulation building, we do believe that games and game-like products can incorporate the intellectual content of such models and simulations in a way that will increase their effective utilization. In the words of David Rejeski, Director of the Foresight and Governance Project at the Woodrow Wilson International Center for Scholars, “The challenge we face is to accelerate the capacity of decisionmakers to learn and, ultimately, to learn before doing. It is easier to connect the dots if you have seen them before, virtually or otherwise. If Boeing can design an entire airplane in the computer (the 777) surely we can do a better job of simulating the choices and dilemmas facing today's public sector managers.” (Rejeski 2002) He continues, “Used properly, games could improve public sector policies and management by allowing decisionmakers to improve their understanding of complex systems, lower their risk of being surprised by unexpected events, and exercise corners of their minds seldom explored in their day-to-day lives until a crisis.” This is a powerful and cogent argument for utilizing games as tools for policy exploration and development, and one that is repeated with variations by a number of thinkers, both within the game community (Sawyer 2002) (Crawford 2002) and within the public policy community (Committee on Modeling and Simulation 1997).


 Standard models and simulations used for social research tend to be designed for use by a relatively small audience that is intimately familiar with the target domain and is also adept at synthesizing information from outputted tables, graphs and charts. Although such tools may employ very accurate models, algorithms and approaches, simulation building and model construction is as much an art form as a science (Sawyer 2002). Given the importance of such simulations and models, it seems natural to use the best methods available to improve their accuracy and effectiveness, wherever those methods might be developed. We believe that the techniques of the interactive game industry, and in particular the online multiplayer game industry, can provide the creative and technical means for effecting such improvement.


Ben Sawyer of Digitalmill lists three advantages of game-based models and simulations (Sawyer 2002):


·         Advantages of design that encourage wider and repeated use, and amplify learning opportunities and strategic thinking among users.

·         Advantages of technology and approach that include utilizing off-the-shelf consumer hardware, high-end visuals and graphics, and intuitive interface design.

·         A background in developing both non-fiction and fiction-based models with incomplete data, or empirically derived data.


Especially when applied to MMOG’s, these advantages combine to give game-like models and simulations a more robust and intuitive interface that will encourage and facilitate participation by large numbers of players. These players will be able to engage the models longer and more thoroughly, and such massive and exhaustive participation should yield solid experimental data with good predictive properties. And beyond prediction lie the goals of situated learning and adaptive community development, real-world effects of virtual world participation.


Who is Using Games Already?


There is already a significant body of work created by organizations employing games for simulating public policy and social interactions. We aim to extend the existing state of the art by employing MMOG technologies and practices, in order to enable participation on a massive but coordinated scale. Here are some currently existing examples:


In 1997 the Virtual U (VU) (  was conceived by the Alfred P. Sloan foundation, in conjunction with William Massey, former CFO for Stanford University and president of The Jackson Hole Higher Education Group. The idea was to capture, as much as possible, the breadth of decision-making necessary to guide a major university. They believed that a comprehensive simulation of the complicated components of a university might help decision makers understand the effects their decisions had on those component parts. In effect it was a “SimCity” for universities, and they hoped it would deliver a novel approach to teaching university administration. VU 1.0 was released in the fall of 2000, and it has sold more that 1000 full copies; the demo has been downloaded over 15,000 times, with multiple users experiencing each copy sold or downloaded. Currently more than a dozen programs use it as part of their university administration program (Sawyer 2002).


VU’s strong development team contributed to strong features, including a budgeting system that is one of the most advanced seen in a game. The program’s simulation of a large organization’s decision making has received much attention for its realism. Instead of players’ decisions leading to immediate results, they initiate trends and behaviors that move toward the result desired by the manager players. On the down side, the model and simulation engine has too much weight, at the expense of playability. The developers are working to remedy that, as well as problems with overall complexity of the game, in an upcoming VU 2.0.


SimHealth ( was produced in 1993 by the Markle Foundation, working with Maxis and Thinking Tools. Its aim, truly ambitious at the time, was to simulate the health care policy of the United States. SimHealth gives players the chance to patch up the U.S. health system, while balancing the needs of citizens, the medical community and insurance companies, or to start over with a completely new system and see how it fares. Players must answer complex questions about health policy and their virtual communities will flourish or wither, depending how well their plans were thought out and implemented. The simulation software employed in this product was cutting edge for 1993 and employs adaptive automatic agents that act on their own to a certain extent.


This sort of policy simulation software offers numerous possibilities. For instance, using SimHealth, one can experiment with different factors to achieve the formulation of a comprehensive health care plan. Looking beyond SimHealth, it is feasible to conceive that government could use this tool to construct legislation and implement public policy in many different areas (Hiles 1995).


Joint Force Employment (JFE) ( is a CDROM based game developed in 2000 exclusively for the United States military to convey the concept that joint warfare is team warfare. It allows players to deploy forces in a highly realistic manner, while playing against state-of-the-art AI opponents. This game is designed to fill a specialized niche, training the top of the war fighting command structure. Although the number of personnel in a Joint Task Force's staff is relatively limited, the ability to rapidly integrate military personnel into a JTF and prepare for unanticipated missions on the fly justifies the use of online digital games (Prensky 2001).


JFE is every bit as much a game as such classics as Command and Conquer and Warcraft II, but at its heart lies an extremely accurate simulation of actual U.S. war fighting doctrine, and this creates a level of reality entirely appropriate for its mission. It is interesting that the main reasons the U.S. military chooses to train this way are cost – normal simulations are simply too expensive – and the motivating factor of engaging military personnel with the same kinds of media they have become accustomed to using recreationally (Prensky 2001).


Whyville ( is not really a game in the usual sense, but rather an interactive virtual world with a large base of young (mostly 8 to 15 years old) users. Whyville was created in 1999 to help students explore science and social science concepts in an interactive learning environment. The virtual world is constructed to engage students in meaningful yet entertaining learning experiences through humor, curiosity, relevance and scientific enquiry (Ascherbacher 2003). One interesting feature of Whyville is that almost three fourths of its members are girls, a figure at odds with most web-based communities.


Whyville has several features that appeal to young people, particularly girls: chatting and social interaction, modifiable avatars and other design opportunities, a simulated community with creative and entrepreneurial opportunities, and science games (Ascherbacher 2003). According to Whyville co-founder Jennifer Sun, the primary digital experience for young people should be the experience of learning within a digital community (Sun 2003). To this end, Whyville incorporates a series of science games that facilitate learning according to the Constructivist model of learning by doing in a collaborative environment.


A very important feature from our perspective is the provision of entrepreneurial opportunities, leading to a robust virtual economy. Whyville has its own currency, the clam, and users can earn clams by performing a variety of tasks. In addition to these system-based chores, Whyvillians can set up shop in the world, for instance, to sell face parts for modifying their avatars (avatar modification is a high profile activity). These virtual businesses can create a fair amount of wealth in clams, and some in-world designers are running virtual design studios with several workers and complex budgets. We believe that participation in this virtual Whyville economy can provide useful training for participation in the real world economic sphere.


MMOG Technologies and the Construction of Social Systems


Our technical approach


The current state of the art for MMOG type games involves massive backend programs running on clusters of servers, connected through the Internet to client programs running on individual players’ computers. While an MMOG world may encompass many thousands of players simultaneously, they are not necessarily playing in the same virtual space. The logistics of handling such large numbers of players currently dictates that the game world be divided into sections, often called “shards”, each with its own cluster of servers and running its own version of the game. Thus two players may be playing the same MMOG, but if they are logged into different shards, they will not be visible to each other. So far this problem has proven fairly intractable and it means that in some sense the games are not quite as massively multiplayer as they are advertised to be. Nevertheless, one or two thousand players can easily coexist simultaneously in one shard of a typical MMOG, which is significant from our perspective.


Given the nature of the community issues we wish to address and the projected populations required to suitably game those issues, we feel that we should initially create a game world that will support up to around 200 simultaneous users. Although this number works well for early development of our social modeling processes, it is a rather unusual size requirement for online gaming. Most games are written for very small groups of 1 to 10 or so, or they are truly massively multiplayer as we have described above. Due to economic considerations, no one has really put much effort into creating game engines designed for 100 to 200 players, but we see that as being a creative challenge for our team.


Technology a la carte


We face some basic choices in the technological platform for our game world development. We can use an engine, such as Torque from Garage Games, that is really meant for single mission to small group play, and try to expand the capabilities of that engine. This strategy assumes that we will find no show-stopping limits to the underlying technology when we attempt to expand it, an assumption we do not feel comfortable making. At the other end of the spectrum lie the truly massively multiplayer game engines developed for such MMOG’s as Everquest from Sony and Ultima Online from Electronic Arts. While appropriate for future versions of our project, we feel that such massive worlds are overkill for our early development and might furthermore lock us into feature sets that we do not need, while leaving out features we want. Additionally, we have no assurance that we would be able to obtain a license for such a platform.


Our solution to this problem, and one we believe provides us with maximum flexibility at the lowest cost, is to use off the shelf products wherever possible, and integrate those products with custom code where needed. There are readily available and tested solutions for many of our needs: 


·         We must have a robust and easy to use graphics engine to create our world, a requirement easily met by Renderware or a similar product. This versatile graphics middleware package is a de facto standard for much of the mid-level game market and there is a wide body of knowledge about implementing efficient solutions with it.

·         Our Graphic User Interface (GUI) can be created using the Microsoft Foundation Class Library (MFC) a collection of software objects that includes the kinds of windows, buttons and other objects we need to get our game world onto the computer screen.

·         We can use an AI middleware system such as Simbionic to give our Non Player Characters (NPC’s) realistic motivations and capabilities, as well as applying AI to system rules.

·         There are many possible platforms for such necessities as chat, sound, video, etc. By building our own system, we are free to pick the packages we feel provide the best functionality in their class.

·         We plan to have a well-developed web component to the project as well, with discussion forums, chat rooms and real time downloads of information from within the game.


Some minimal amount of coding will be required to knit together these off-the-shelf products, but the end result should be a robust and functional platform at a size and cost that suits our needs.


Advancing the state of game networking


One vital area of functionality we will need to provide ourselves is the network wrapper that allows clients and servers to communicate efficiently. Available network solutions are currently either too small scale or too large to suit our needs. Also, the specialized nature of our MMOG means we have different network requirements than might be expected for a normal game environment. For instance, since we do not have a “shooter” game, we have no need to employ the tricks necessary to overcome network latency while minutely tracking moving objects. Writing the network code ourselves actually gives us an opportunity to advance the state of the art for networked games.


We believe that the size and nature of our game means that we may be well served by running the game platform in a grid computing environment, rather than on a normal server or server cluster – something that has not yet been done. Grid computing is a new paradigm for harnessing computer power, in which applications are run on groups of networked computers. Special software is used to aggregate the processing power of the individual computers in such a way that supercomputer strength is attainable, given enough component computers. While we do not need a virtual supercomputer to run our game, a certain amount of computer power is necessary and grid computing can provide that power by connecting existing machines and using their wasted cycles. Using the inherent redundancy of grid computing will also insure that we never have a server outage.


Beyond such utilitarian advantages of grid computing, we can also use it to investigate the relationship between Virtual Organizations (VO’s) and distributed computing resources. Researchers working under the auspices of the Globus Alliance have developed this concept of the VO and rationales for how it is best supported by grid computing architectures (Foster, Kesselaman, Tuecke, 2001). We believe our MMOG system can be thought of as a Virtual Organization, making grid computing a logical choice.


We plan to make extensive use of wi-fi hot spots for connecting to our game, since it may be distributed in areas lacking in normal Internet connectivity, but which have many wireless-enabled cafes and coffee houses. Locations that do have Internet connectivity will likely not have broadband, so we will call on the MMOG game industry’s long experience in dealing with issues of Network latency and slow connections.


Web integration


Another area in which we can advance the state of the art is by tightly integrating a web environment with the MMOG game world. We envision a highly interactive web site that is closely linked to the MMOG environment in several ways. The player community can use the chat and forum features of the web site to create a deep and meaningful dialog about the game. Persistent and important information can be posted there, such as users’ agreements, rules and policies, etc. It will be the public face of our project.


We plan to do more radical things on the web, as well. For instance, we can automatically feed screen shots from within the game to the web site, where users can look at them singly or in slides shows, or they can be archived. This means that interested parties can view what is happening in the MMOG, even without having the client running on their computer, a very handy feature for sharing the virtual world in the larger community. The same kind of feature could be implemented with streaming video as well. It is also feasible to integrate real time chat on the web site with the in-game chat system, once again allowing access to the game world by those lacking the client. We believe that such features as this will greatly enhance the utility of our MMOG for community development projects. For instance, an AI in the virtual world could send a message via chat or email to a project decision maker in the event that some important milestone had been reached in the game.


Using MMOG’s to Facilitate Community Development


Crossing the Line from Virtual to Real


A central tenet of our argument is that the gradual dissolution of boundaries separating the real world of physical space from the virtual world of mediated communication technologies will ease transfer of experience from one domain to the other. In particular, situated learning activities in virtual MMOG worlds will have greater relevance to real world community development, to the extent that they apply in both cases. For this to happen we need scenarios where virtual events impact the physical world and vice versa. Evidence of such will bolster our case.


Legal and economic implications


 We are already beginning to see activities, primarily in the economic sphere, that have counterparts in both domains, with people crossing the boundary freely in pursuit of wealth, both real and virtual. Some economists are even thinking about virtual world economies as alternative sites for productive activity that equate favorably with real-world jobs (Castronova 2003). According to Dan Hunter of the Wharton School of Business, “Even though these virtual worlds are nothing more than software coding residing in banks of computers, what goes on in them poses increasingly serious challenges to our notions about the nature of property, the legal rights of players in virtual worlds and even the presumed boundary between the real and the imagined worlds.” (Hunter 2003) When respected legal scholars say things like this, we may safely assume that change is taking place.


Virtual worlds are really a new kind of social order, and the economic and legal implications of virtual life are just beginning to be worked out. Compounding such issues is the fact that all such virtual worlds, whether MMOG’s or other types, run on software code that is proprietary property, and which resides on servers that are likewise property of the corporations that develop the worlds. These so called “god-corporations” (Lastowka and Hunter 2003) have so far been able to dodge serious questions about legal rights within their proprietary worlds by hiding behind their user agreements, but these issues will ultimately be worked out through recourse to litigation and legislation. A recent study of this problem of avatar property rights found that “From the above discussion of the three main normative theories of property we have come to a surprising conclusion: there are strong normative grounds for finding that property rights should inhere in virtual assets, whether these be chattels, realty, or avatars. Further, based on earlier discussion, we can conclude that there is no descriptive disconnection between our real world property system and virtual assets. From both descriptive and normative positions, it seems that virtual assets do, or should, attract property rights.” (Lastowka and Hunter 2003) These are strong words, but so far untested in the courts. Until such time as they are, we must consider emergent economic and property issues as provisional in nature, but that does not mean we cannot draw inferences from them. 


Virtual goods, real markets


Most virtual worlds, and essentially all MMOG’s, provide for at least a rudimentary economy, as developers have found this feature to be essential for attracting users. When entering an MMOG for the first time, one’s avatar typically is very minimally equipped and it requires a considerable expenditure of resources to gain the weapons, tools and accessories necessary to lead the “good life” in the game world. There is a definite progression in most MMOG’s from the almost naked and propertyless newcomer’s avatar to the elaborately equipped and invested avatars of experienced players. It is noteworthy that this progression to old-timer status parallels the centripetal progression of legitimate participation in Lave’s model of social learning (Lave and Wenger 1991).


Depending on the MMOG in question, there are numerous ways to acquire the virtual cash required to upgrade one’s avatar: Players are often rewarded for being logged on for certain amounts of time; players can accumulate feudal fiefs and require vassals to pay tribute; players can create special objects or services and sell them in-world, either to other avatars or to AI trading agents (Ascherbacher 2003). While these in-world economies were designed to function as hermetically sealed units, an interesting and provocative phenomenon has surfaced. At first a few players, followed quickly by ever-increasing numbers, discovered that they could auction items from MMOG worlds, such as magic swords, houses, avatars, etc. for real money at online auction houses like eBay. Although such trading so far involves only a small minority pf players, enough of them are trading that a few reputedly make 6-figure incomes (Castronova 2003). Currently eBay lists auctions for over 2500 Ultima Online related items, around 800 for Asheron’s Call and 200 for Anarchy Online. Even though Sony Entertainment feels threatened enough by this phenomenon to forbid auctions of Everquest items, there are nonetheless 150 or so of such banned items available (LeFevre 2004). Clearly this is a trend worth watching, blending as it does the economies of the real and virtual domains.


An even more interesting development is the online virtual currency exchange (Terdiman 2004). At least one such exchange is now in operation, The Gaming Open Market ( This functions just like a real world currency exchange, except instead of exchanging, say, U.S. dollars for Euros, one can exchange U.S. dollars for Simoleans (the currency unit for The Sims Online) or for currency from a number of other virtual worlds. Additionally, currency can be traded from one world to another, e.g. Simoleans for Therebucks, at a cost of 4% for each side of the trade. The opening of such currency exchanges makes it possible for persons who do not even have accounts or avatars in a world to nevertheless trade in that world’s currency, a clear example of virtual behavior and experiences crossing the line into the domain of everyday life. A door has been opened between the worlds and there will be no closing it again.


Legitimate participation in virtual worlds


Communities of interest


The opening between the real and virtual domains is mediated by the concept of communities of interest that we developed earlier. In the example above, we have a virtual community, consisting of players who have amassed in-world currency through trade, work or other means; we also have a real world community of people who wish to obtain virtual currency for whatever purposes. Although they exist on different sides of the real/virtual boundary, these two communities are both defined as well by association of interest – namely interest in exchanging virtual currency for real world currency or for other virtual currency. This community of interest provides a way for an individual to be both in the real community and in the virtual community at the same time. And since communities are made up of interacting individuals, the presence of such doubly-linked individuals provides the open door for experiences to transfer between worlds.


Situated activity in communities of practice


Our communities, both real and virtual, can be viewed as activity systems in which the members of the community conduct activities, mediated by tools and technologies, that engage the target domain in a process aimed at some desired outcome. If that target domain is the community itself, and the desired outcome is community development, then we have an activity system model for community development. In particular:


·         The community and the actions of its members are taken in whole as the unit of analysis. Complex feedback loops between actions and effects are expected and accounted for.

·         The community is considered as a historically emergent entity, with multiple overlapping effects of such historicity acting at any given time.

·         Change and growth are results of inner contradictions within the community. These contradictions can be systemic to the community, or they can be introduced. A powerful way to think about learning is in terms of such introduced contradictions.


The activities in question – in this case learning and development – are situated within the structure of a community itself defined in terms of activity. This recursive and operant definition of situation works, because in activity theory context cannot exist outside the activities of the system. Thinking about community as a community of practice further refines our model. In our example, the practice of currency trading defines the community, in both the real and virtual worlds. It is important to note that communities of practice can also be communities of interest, providing a powerful linkage between real and virtual.


Legitimate peripheral participation


Considering the community of interest as a community of practice implies a stratified structure with differing levels of involvement, expertise and investment, leading from newcomer to established veteran status. Such structure allows for legitimate participation at the periphery of the community by members who are not yet knowledgeable or fully functioning. Legitimacy is bestowed by allowing peripheral participants real insight and knowledge about the community, as well as opportunities to carry out community tasks at a level that is not critical to community survival (Lave 1991). This participation has historically involved actual engagement with the practice in question.


One of the major criticisms of education deriving from this position is this: that most educational systems involve peripheral participation in the practice of education, rather than in the target domain of that educational system, leading to students who are adept at being students but not adept in the desired target domain (Lave 1991). Simulation-based learning provides a way around this problem, and we believe that the application of MMOG technologies and practices will significantly improved the simulation of experience, leading to better learning outcomes.


Transferring experience and learning


Our problem is to realistically simulate the experience of a community of practice in a MMOG virtual world, and to be able to transfer learning gained from that situated activity into the real world of everyday life, leading to community development. We now have all the elements in place that we need to accomplish this goal:


·         The MMOG virtual world community, which has a counterpart in the real world – the

·         Target domain community. These two communities are linked, in turn, by a third

·         Community of interest, defined by the overlapping membership between the first two. This community of interest is also a

·         Community of practice, as defined by its activities, which allows

·         Legitimate peripheral participation as a site for learning. This learning transfers readily to the target domain through the

·         Community of interest aspect just described.


Increased verisimilitude of our models and simulations will lead to longer and deeper involvement by MMOG players, resulting in a richer and more lifelike experience. Learning thus situated has a better chance of being transferred to the domain being modeled.


Stakeholders and decision makers as legitimate participants


Community development implies the involvement of decision makers at the very least, and should imply the involvement of stakeholders as well. While these groups may have overlapping membership, often they do not. We believe that effective community development should engage both stakeholders and decision makers in an ongoing dialogue, and that situated activity in an MMOG virtual world provides a natural arena for such dialog.


Too often, community development decisions are seen, rightly or not, as being top-down with only cursory attention given to those people most likely to be affected. It is not easy, however, to establish meaningful dialog between decision makers and the stakeholders in a project. First, those stakeholders must be identified. By what criteria do we decide whom to consult and whom not to consult about potential impact? Having decided on a group of advisory stakeholders, the problem then becomes logistical – how do we maintain an ongoing dialog with them over a significant period of time? It also becomes a matter of communication – how do we convey information about the project in a way that is meaningful and constructive for all parties? One way that such committees have historically sought to convey complex planning information is through the use of models and simulations.


Current simulation and modeling software can do a thorough job of analyzing complex and often conflicting data, and then generating “what-if” scenarios. The usefulness of such software is limited, however, by several factors. Most models are built for small audiences of specialists, who understand the underlying math and science (Sawyer 2002). Very few policy planners have ever actually run such a model; they rely on others to do so and then interpret the results returned. Also, even if results from such models and simulations are widely broadcast, the stakeholder population is likely to have even less grasp of the underlying concepts, so the actual impact of disseminating such simulation data will probably be slight


A community of interest, a community of practice


Seen this way, the stakeholders and decision makers together form a significant community of interest in their desire to understand and act upon the results of realistic model and simulation programs. Such a community of interest can provide powerful mediation between a real community and a virtual community. Given this, we suggest constructing an MMOG virtual world as an interface to the simulation and modeling software. This will allow both community development stakeholders and decision makers to enter the game world and play through different scenarios to see how their decisions evolve. Such legitimate peripheral participation in the virtual community of practice will give players a real sense of how the proposed development will actually affect them.


MMOG based simulation has several advantages, as compared to standard modeling and simulation software:


·         The realistic modeling of physical and social communities allows players to identify more thoroughly with the problem set and with results in the game.

·         The intuitive, game-based interface allows average people to interact with sophisticated models, thus enabling both decision makers and stakeholders to actually run the simulations.

·         Engaging game play encourages players to spend sufficient time with the simulations to wring out results.

·         The sophisticated MMOG technology allows large numbers of players to test the simulations simultaneously, leading to greater confidence in results obtained.

·         The networked nature of MMOG technology means that both stakeholders and decision makers can log into the game remotely at times and locations of their choosing.

·         The MMOG virtual community of practice makes an ideal site for situated learning to occur, and provides proper conditions for transferring such learning back to the target domain of community development.


We believe these advantages will allow the creation of highly engaging virtual game worlds, in which large numbers of players will interact to form true communities of interest and communities of practice. Membership in these communities will in turn enable situated learning through active and legitimate participation, leading to effective opportunities for community development.




Ascherbacher, P.R.; Gender Differences in the Perception and Use of an Informal Science Learning Web Site; Final report to NSF, Caltech’s Pre-College Science Initiative, 2003


Avedon, E. and Sutton-Smith, B., eds.; The Study of Games; New York, John Wiley and Sons, 1971


Barthelmess, P. and Anderson K.M.; A View of Software Development Environments Based on Activity Theory; CSCW Special Issue: Activity Theory and the Practice of Design; David Remiles, ed., 2002


Bødker, S.; Through the interface: A human activity approach to user interface design; Hillsdale, NJ, Lawrence Erlbaum, 1991


Castronova, E.; On Virtual Economies; The International Journal of Computer Game Research, Volume 3, Issue 2, December 2003;


Castronova, E.; Virtual Worlds: A First-Hand Account of Market and Society on the Cyberian Frontier; CESifo Working Paper No. 618, December 2001;


Committee on Modeling and Simulation; Modeling and Simulation: Linking Entertainment and Defense; Committee on Modeling and Simulation: Opportunities for Collaboration Between the
Defense and Entertainment Research Communities; National Academic Press 1997;


Costikyan G.; I Have No words and I must Design; Interactive Fantasy #2, 1994;


Crawford, C.; Subjectivity and Simulation; Serious Games: Improving Public Policy through Game-based Learning and Simulation; Foresight and Governance Project, the Woodrow Wilson International Center for Scholars 2002;


Engeström, Y.; Development studies of work as a testbench of activity theory: The case of primary care medical practice; Understanding practice: Perspectives on activity and practice; Chaiklin, S. and Lave, J., eds.; Cambridge, England, Cambridge University Press, 1996


Foster, I., Kesselman, C., Tuecke, S.; The Anatomy of the Grid: Enabling Scalable Virtual Organizations; International Journal of Supercomputer Applications;


Hiles, J.; Lessons Learned from SimHealth: the Promise and Limits of Health Policy Simulation; Cyber Security and Policy Initiative Seminar, George Washington University 1995;


Hunter, D.; In Norrath, Tattoine and Rubi-ka, Just What Are Your Legal Rights?; Public policy and Management, Wharton School of Business, 2003;


Kim, A. J.; Community Building on the Web; Peachpit Press, Berkeley CA 2000


Lave, J.; The practice of learning; Understanding practice: Perpsectives on activity and practice; Chaiklin, S. and Lave, J., eds.; Cambridge, England, Cambridge University Press, 1996


Lave, J. and Wenger, E.; Situated learning: Legitimate peripheral participation; Cambridge, England, Cambridge University Press, 1991


Lastowka, G.F. and Hunter, D.; The Laws of Virtual Worlds; U. of Pennsylvania Law School, Public Law and Legal theory Research Paper Series, Research Paper # 26, 2003;


LeFevre, N.; How MindArk’s MMOG, Project Entropia, is planning to make you rich!; Stratics MMOG Weekly: Volume 9, 2004;


Leont'ev, A. N.; Activity, Consciousness, Personality; Englewood Cliffs, NJ, Prentice Hall, 1978


Prensky, M.; True Believers: Digital Game-Based Learning in the Military; ASTD Learning Circuits 2001;


Preston, J., Booth. L. and Chastine, J.; Improving Learning and creating Community in Online Courses via MMOG Technology; Clayton College and State University 2004;


Rejeski, D.; Gaming our Way to a Better Future; Serious Games: Improving Public Policy through Game-based Learning and Simulation; Foresight and Governance Project, the Woodrow Wilson International Center for Scholars 2002;


Salen, K. and Zimmerman, E.; Rules of Play: Game Design Fundamentals; Cambridge MA, The MIT Press, 2004


Sawyer, B.; Enhancing Simulations, Models and Their Impact Using Interactive Game design and development Practices and Technolgy; Serious Games: Improving Public Policy through Game-based Learning and Simulation; Foresight and Governance Project, the Woodrow Wilson International Center for Scholars 2002;


Sun, J.; Interview; conducted by Rob Renolds for, 2003;


Terdiman, D.; Virtual cash Breeds Real Greed; Wired News, 2004;,2101,61999,00.html?tw=wn_story