In this blog post, I will analyze the advantages and limitations that MMORPGs can offer to the study of collective behavior by examining their physical and structural similarities, the projection of player behavior, and real-world examples such as the “Tainted Blood Incident.”

Motivation for Choosing the Topic

Since the late 1990s, the development of gaming technology has led to a surge in online game users, and MMORPGs were a genre that was particularly beloved by many players. MMORPG stands for Massively Multiplayer Online Role-Playing Game, referring to games where multiple users interact and perform their respective roles within the same virtual space connected online. In the years since the emergence of first-generation MMORPGs, numerous incidents and accidents have occurred within them; interestingly, the nature of these events resembled situations in the real world or exhibited characteristics that could easily occur in reality. For this reason, I believe that the behavioral patterns observed among users within MMORPGs hold value for systematically elucidating human behavior. Therefore, based on the characteristics of MMORPGs and real-world examples, I intend to examine how this genre can contribute to the field of human behavioral analysis. The reason I chose MMORPGs as the subject of study among various game genres is that they best reflect collective interaction and social structures compared to other genres.

Similarities Between MMORPGs and Real Society

A question arises as to whether in-game communities and real-world society must be sufficiently similar for MMORPGs to be useful in the study of collective behavior. I argue that MMORPGs share similarities along two axes. First, the physical and structural attributes of virtual space can reproduce real-world social conditions to a significant extent. Second, the psychology and behavior of the “real people” playing the game are projected onto their avatars and actions.
The similarities between MMORPGs and reality are evident in several aspects. First is the diversity of settings. Some games are set in the modern era, while others are set in the past—such as the Middle Ages or the Joseon Dynasty—or take place in entirely fictional fantasy worlds. Regardless of the setting, users experience a sense of immersion, as if they truly exist within that era or world. Second is the player’s freedom. While traditional storytelling-centric games imposed a linear progression, MMORPGs offer a high degree of freedom, allowing users to design their lives within the virtual environment however they wish. Players can make various choices, such as setting sail to explore other continents or settling in one region to live as a merchant. Third is the diversity of professions and roles. Many MMORPGs offer a variety of professions (classes), and each profession creates a division of roles through its unique skills and functions. In this process, preferences and aversions toward specific professions emerge, resulting in a phenomenon similar to social division of labor. Fourth is the potential for communication and social interaction. Various social activities commonly seen in real society—such as conversation with other players, trading, forming parties, and dividing roles during combat—are faithfully reproduced. These four elements all hold significance beyond mere “realism,” and each requires explanation as a distinct concept.

The Potential and Limitations of Online Games as “Samples”

Online games provide a favorable environment for studying collective behavior, as they connect many real people. The way each user plays the game reflects, to some extent, their personality, values, and preferences. For example, choices regarding a character’s appearance or decision-making styles in specific situations can serve as signals revealing an individual’s tendencies. Therefore, MMORPGs contain a large number of “samples” with diverse tendencies, and if analyzed appropriately, meaningful data on collective behavior can be obtained.
However, the concept of “sample” discussed here can be confused with the issue of representativeness required in statistical and epidemiological research.
It cannot be assumed that online behavior corresponds exactly to offline behavior, and the context of the game itself can alter behavior. Furthermore, quantitative research is needed to determine how many users are actually participating and in what ways, as well as which groups they represent in real life. For these reasons, while MMORPGs offer a major advantage (the potential for large-scale data), they also have limitations that require caution in interpretation. The question remains as to whether a real person playing a game will always behave in the same way as they do within the game, and this issue will be addressed later as a limitation of the research.

Research Case Study: The Tainted Blood Incident

Among the incidents that have occurred within online games, the so-called “Tainted Blood Incident” in “World of Warcraft” (WOW) in 2005 is a notable case from a behavioral science perspective. This incident began when “Corrupted Blood,” an in-game debuff (status effect), leaked outside the dungeon due to a bug and spread rapidly through towns and cities like an epidemic.
While this debuff was not fatal to high-level players, it proved lethal to low-level players within seconds. As a result, many player characters died, and significant chaos ensued, with cities becoming littered with the corpses of characters. This incident attracted the attention of researchers as a case study that went beyond a simple in-game bug, allowing them to observe real-world epidemiological and behavioral responses—such as the fear that arises when a real epidemic spreads, attempts at isolation, problems with information dissemination, and the behavior of some who intentionally sought to expand the spread.
The “Contaminated Blood” incident demonstrated that MMORPGs can serve as useful “field experiments” for understanding collective behavior and risk response, transcending their role as mere entertainment platforms. However, this case has limitations in that the game’s context and the characteristics of the player group influenced the results; therefore, it should be used as material for exploring similarities rather than being directly equated with real-world epidemic research.

User Behavior Revealed During the Epidemic

From a behavioral analysis perspective, what makes this incident particularly interesting is that the response patterns exhibited by users when the epidemic spread closely resembled those of people in the real world. Low-level users took on vulnerable roles, much like the elderly or the infirm, while high-level users behaved like healthy adults. Healers with healing abilities treated infected players like doctors, and guild masters made decisions regarding quarantine and access control like community leaders.
Specifically, some guild masters demonstrated calm leadership by quarantining infected users or regions, while healers performed medical roles by repeatedly providing healing. On the other hand, some users avoided contact after becoming infected and met their deaths, while others intentionally came into contact with others to spread the infection. Furthermore, fraudulent activities—such as creating and selling fake medicines instead of actual treatments—emerged, revealing the full spectrum of human behavior in a disaster situation.
In this way, the events within the online game provided a platform to observe how people’s behavioral patterns—including leadership, cooperation and selfishness, institutional responses, and fraud—manifest when a real disaster occurs.

Limitations of the Study and Future Research Directions

While MMORPGs and real society share many structural similarities, we must acknowledge that fundamental differences also exist. For example, while the principles of supply and demand operate in the in-game market much like in reality, the value of in-game currency tends to depreciate and inflation occurs because it is continuously generated through activities such as monster hunting. Unlike the real economy, where money circulates and prices remain relatively stable, it is unreasonable to directly apply the economic behaviors observed in games to the real world.
There is also the issue that game characters do not always reflect real-life tendencies. A person who is passive in real life may act extroverted in a game, or commit criminal acts they would not commit in reality under the rationale that “it’s just a game.” If such non-reflective actors constitute a significant proportion of the group, this will act as a major variable in the study of group behavior.
Therefore, behavioral analysis through games must establish methods to control or adjust for the economic and psychological factors unique to games. Specifically, this requires research designs that account for in-game incentive structures, play motivations, and the disconnect between avatars and real selves, as well as a triangulation approach that combines other data sources (field observations, surveys, experiments).

The Significance of Game-Based Research

Why, then, are online games effective tools for studying human behavior? There are clear ethical limitations to directly experimenting with human behavior in extreme situations. For example, it is impossible to artificially create and expose people to war or large-scale disaster scenarios. In this context, games provide an environment where various extreme situations can be safely simulated without causing physical harm.
Numerous studies have observed that human nature is revealed more explicitly in extreme situations, and analyzing the behaviors people exhibit in contexts such as disasters or wartime is crucial for understanding the deeper aspects of human psychology. While many games feature unrealistic settings, a significant number deal with extreme scenarios that are entirely conceivable—such as nuclear war or pandemics—thereby providing insights that can inform the development of real-world contingency plans.
Finally, game research to date has tended to focus on verifying addictiveness or harmfulness. By exploring the diversity and structural significance of user behavior, as in this case, we can consider the academic value of online games themselves more broadly. Regardless of positive or negative perspectives, games serve as a valid laboratory for studying human behavior.