Massively multiplayer online role-playing games (MMORPGs) trace their lineage back to pen and paper Dungeons and Dragons games; later versions introduced a variety of props such as game-boards, ornate player tokens, and alternative rule sets [1]. The D&D lineage established role-playing games as multiplayer, interactive, social events; thus the arrival of personal computing and the development of stand-alone PC games where the player either adventured through fantasy lands by themselves or with a series of AI companions can be seen as a necessary but unfortunate interruption in the “natural” state of role-playing games. While the advent of text-based MUDs (Multi-User Dungeon) and MOOs (MUD Object Oriented) kept alive the social element of multiplayer gaming (and heavily influenced the text-centered communication present in most online games to this day) these programs lacked the sophisticated, complex and frequently arcane sets of rules of the earlier tabletop systems. Far from being off-putting to players, these complex rules made them feel as if they were both building and inhabiting an alternative world. It had been clear from the introduction of the earliest personal computers that computers could enable players to work inside these rules without the tedium of keeping track of their implementation and to do so on a large scale. However, such games had to wait for server and network technology to catch up so that they could combine the social experience with complex world-building in an accessible mass market format.
Thus, the arrival of Ultima Online (UO) in 1997 was an event eagerly awaited by the fantasy role-playing community. While not the first networked role-playing game it was easily the most successful to that point, in part because it was drawing upon a well-established gaming franchise, but also because it offered players a (for the time) truly massive world to explore, and one that actually changed in response to their actions [2]. In addition, the designer of the Ultima franchise, Richard Garriott (also known by his gaming acronym, Lord British) had developed a devoted following in large part because the later titles in the Ultima series had moved beyond the simple visceral thrill of adventuring in order to explore ethical and moral dilemmas. Garriott’s games were renowned for asking players to confront the effects of prejudice and stereotyping and emphasized what Garriott would come to call the eight virtues that formed the moral and ethical underpinning of the mythical world of Britannia: compassion, valor, honor, honesty, spirituality, sacrifice, justice, and humility. Thus when UO launched many players were enticed by the promise not simply of living an alternative fantasy existence in a fully realized world shared with other players, but by the chance to build a utopian medieval fantasy realm.
In hindsight, what happened next was all too predictable and was chronicled to brilliant effect by Amy Jo Kim in a 1998 article for Wired:
There's no shortage of realism in this game - the trouble is, many of the nonviolent activities in UO are realistic to the point of numbingly lifelike boredom: If you choose to be a tailor, you can make a passable living at it, but only after untold hours of repetitive sewing. And there's no moral incentive for choosing tailoring—or any honorable, upstanding vocation, for that matter. So why be a tailor? In fact, why not prey on the tailors? [3]
In short order the realm of Britannia was overrun by teams of player-killers (PKs) who slaughtered every player in sight, looted their corpses and used the proceeds to indulge in orgies of conspicuous consumption (flaunting elaborate weaponry and armor, building massive castles as guild headquarters, etc.). In short, the fictional realm of Britannia had begun to resemble the historical realities of the early modern period. . .or of our own period if we substitute corporations for player guilds. Kim herself investigated both sides of the player divide, creating a character called DarkStarr and spending time with one guild who reveled in the rich ceremony and ritual of Arthurian romance. She then spent time with another guild, the Insidious Brotherhood who, for entertainment, “killed a wandering healer, ate some of the victim's body parts, and prayed together over the corpse, offering it as a sacrifice to the evil Guardian” [4]. This apparently disastrous turn of events in Britannia was in fact quite consistent with Garriott’s design, which allowed for players to be able to explore the implications of evil and antisocial behavior as fully as they could explore their more virtuous counterparts. What seems to have caught Garriott off guard is that so many people would have found evil so attractive. Furthermore, offering the option of exploring the implications of antisocial behavior in a game world only works, as Kim notes, if everyone is playing the same game: “what happens when players who think they're attending an online Renaissance Faire find themselves at the mercy of a violent, abusive gang of thugs?”
Kim notes that other designers had criticized UO and Garriott himself for not having learned lessons that the industry had learned about creating online worlds. To be fair, however, UO was creating an online world whose scale and complexity had rarely been attempted before. In retrospect it is also clear that many developers and professional game critics could not afford to be so smug, as the industry has persisted in making similarly problematic design decisions in both single-player and multiplayer contexts since then. The problem in fact is that the designers of UO fundamentally misunderstood the nature of the world they were creating, and this misunderstanding was the product of the interaction between two assumptions: one related to gameplay, and one related to ideology. In ideological terms the designers advocated the importance of “virtues” assuming that those characteristics were transcendental absolutes. This did not mean that they expected all players to follow those ethical and moral precepts—hence the design-permitted behavior of PK—but they clearly assumed that those who held to those virtues would be the “good guys.” What was missing was any realization that “virtues” tend to be highly relative, dependent on community norms, and are maintained largely by a system of rewards and punishments meted out by a centralized authority within that community. In this respect, the developers were making exactly the same mistake (albeit without the malicious and self-serving intent) of those self-appointed guardians of our civic virtue that routinely advocate for “family values” as if everyone agrees upon what those values are. If we look at Garriott’s eight virtues, for example—compassion, valor, honor, honesty, spirituality, sacrifice, justice, and humility—it is entirely possible to have all those virtues on display between the members of a player guild that is devoted to waylaying others and using them as sacrificial offerings. After all, street gangs and organized crime syndicates have their own versions of honor, justice, valor, sacrifice, etc. This “values” problem naturally didn’t manifest itself in previous games in the Ultima franchise. In earlier Ultima games the player operated as a community of one and the game was designed to support his or her experience, ensuring the player never had a meaningful encounter with another group who defined values differently than themselves. The game itself functioned as the centralizing authority that reinforced the “rightness” of the belief system that was shaped by the player's interaction with the game. This is still the limitation of most single-player games and the absence of this limitation is one of the things that makes online role-playing games potentially a valuable learning environment.
This “virtue problem” had more serious implications, however, to the degree that it intersected with a fundamental misunderstanding of the game world. It is clear that UO was initially understood by players certainly, and perhaps even by the designers, as a representation rather than a simulation, to use Gonzalo Frasca’s terms [5]. The game’s attention to realism is highlighted by Kim, but it was also emphasized in the game’s marketing campaign which showcased the level of visual and aural detail. However Kim’s article reminds us that one of the most striking things about the game was something that gamers now pretty much take for granted in their games: realism of effect. Kim notes that, “With practice, players become progressively more powerful and skillful—and it's reflected in their outfits. Indeed, seniority and in-game savvy is expressed in an immediate, graphic way: The clothes make the man or woman. The visual correspondence between what things look like and what they can do is exhilarating.” When we are noticing these kinds of things, however, we are still firmly within the domain of representation. This was a game that was built to represent a medieval fantasy kingdom, to create a world that looked and sounded believable. However when you create a world that is designed to look like a society and then invite real people to come in and treat that world like a real society, you are no longer in the domain of representation but of simulation. Unfortunately, if you as a designer are not fully cognizant of that fact, it can create problems. It is common for designers, game critics and players to talk about open-ended play environments as a sandbox, a blank space within which players can create their own games and forms of interaction. However, a sandbox is still a simulative structure: it has boundaries, for starters, and just about every sandbox I’ve seen already has a few toys lying around. In other words, a sandbox is still a constrained environment. A sandbox without an effective structure is, basically, a beach, which is perhaps a more accurate characterization of UO in its early days. And on this particular beach it was clear that a lot of people were having fun kicking over others’ sand castles.
Responses to the issues that Kim highlights were not long in coming. The “good” players began to band together into vigilante squads to protect themselves or one another against the PKs. This naturally produced a whole array of interesting moral problems that have always been associated with such behavior. Can you really condemn someone for being a PK if you are killing the players who are PKs? If you are offering to escort new players around in exchange for their gold, aren’t you just robbing them in another way? However this solution did not satisfy the many players whose vision of a medieval fantasy existence did not extend to being in constant peril of life and limb or in having to deal with the kind of realism inherent in unequal wealth distribution and consequent power imbalances. As a result, later iterations of UO split the realm of Britannia into (initially) two realms: one where PK activities were allowed and one where they were not. This solution however also proved unsatisfactory. Even those players who inhabited the non-PK realm and were relieved not to be in imminent danger of being offered as a sacrifice to pagan gods found that the moral and ethical complexity of the world was now missing. Without the danger and unpredictability of player-created mayhem, individual actions and decisions became less meaningful. This tension (often characterized as a conflict between PvPers (as in Player versus Player) on the one hand and “Carebears” on the other has remained a constant challenge for designers of massively multiplayer worlds.
1. On the evolution of early tabletop role-playing games see Mona, E. (2007). From the basement to the basic set: the early years of Dungeons and Dragons. In P. Harrigan and N. Wardip-Fruin (Eds.), Second person: Role-playing and story in games and playable media (pp. 25-30). Cambridge, MA: The MIT Press.
4. Darkstarr’s evening went downhill rapidly from this point, and led to what is surely one of the greatest lines in gaming journalism:
In a display of brute force, Bubba turned himself into a gorilla and threatened to sodomize Xavori with a thigh bone taken from one of the victims. Unintimidated, Xavori cast a blazing firewall at Bubba—but because of server lag (or perhaps bad aim) it hit DarkStarr instead, killing her instantly.
The hosts gathered around, staring down at the corpse in momentary dismay. One of them cried out, "You've killed the reporter from Wired, you moron!"
New Missions