Virtual Epidemic, Real-Life Scenarios

A scene from the online video game "World of Warcraft."
Blizzard Entertainment
An accidental virtual plague that decimated the virtual population of a popular online game has piqued the interest of real-world disease trackers and public health planners.

The "corrupted blood" epidemic that triggered an unplanned and unwanted die-off of players of "World of Warcraft" may have been a simulation, but it offers valuable clues to how people might respond in the event of a real-life global outbreak of disease, they suggest.

Researchers are already eagerly exploring the possibility of using these popular simulation games to probe scientific questions for which actual experiments could not ethically be mounted.

"By using these games as an untapped experimental framework, we may be able to gain deeper insight into the incredible complexity of infectious disease epidemiology in social groups," American researchers Nina Fefferman and Eric Lofgren said in an article published Tuesday in the journal Lancet Infectious Diseases.

Dr. Ran Balicer sees great promise in the notion of harnessing the virtual laboratory of online games played by millions worldwide. ("World of Warcraft" alone has 9 million registered players.)

"Basically, I believe this is the next evolutionary step in infectious disease modeling, and this step must be realized as soon as possible, by epidemiologists joining forces with the game industry," said Balicer, an epidemiologist in the faculty of health sciences at Ben-Gurion University of the Negev, in Beer-Sheva, Israel.

"In virtual games like this, you have lots of nature-based disasters that should harm you on your way. There are earthquakes, there are tornadoes. Why not infectious diseases?"

Fefferman is a mathematical ecologist with Tufts University in Boston and Rutgers University in Piscataway, N.J.; Lofgren was one of her graduate students and a gamer who plays "World of Warcraft" and who survived the corrupted blood epidemic.

They said this first major virtual pandemic and other gaming disease outbreaks that may follow could help bridge the gap between actual epidemiological studies — which can only observe events — and the projections of mathematical models.

While mathematical models are a valuable tool to help public health officials visualize how an event might unfold, even the most avid of modelers will admit the projections are only as good as the assumptions on which they are based. People's reactions to an event or disease threat can amplify or mitigate the ensuing damage, but figuring out how they will react in the first place isn't easy.

In "World of Warcraft," death moved swiftly through crowded virtual mega-cities when Blizzard Entertainment, the company that makes the role-playing game, released a new update Sept. 13, 2005.

It allowed the most advanced players to move to a new level where they could go into battle against a winged serpent called Hakkar. The powerful warrior was able to infect players with a disease called "corrupted blood."

For high-level players the ailment was trivial, little more than a cold. But when infected players teleported back to other, more heavily populated levels of the game, they infected players with lesser skill levels. For them, the disease was lethal.

Blizzard Entertainment tried to contain the damage by attempting to quarantine areas of the game, but to little effect. Soon the virtual world was littered with virtual corpses. The game's developers eventually had to reset the computers.

Fefferman, who is also a consultant on disease surveillance for the U.S. Department of Homeland Security, said piecing together the way players responded during the corrupted blood outbreak taught her some unexpected lessons about human behavior.

"In this outbreak we saw lots of behaviors that mirrored real life — which was great — in different proportions, which startled us," she said in an interview.

One was what she called the "stupid factor" — where some players, rather than remaining safely off-line until the outbreak was over, actually logged on to view the carnage up close.

"The stupid factor of 'Wow, this is a curiosity. I will take the risk upon myself and go see what it looks like,' " she said, likening it to something risk-taking teenagers, who are often imbued with a sense of immortality, might do in real life.

"I've never, ever seen that incorporated into a (mathematical) model," Fefferman said, adding she will factor it into future work.

Balicer said he believes disease outbreaks embedded in the scripts of role-playing games could both teach scientists about how social interactions foster the spread of infectious agents as well as help determine human reactions and the consequences of choices made out of them.

But Neil Ferguson, a leading infectious diseases mathematical modeler with the faculty of medicine at London's Imperial College, believes the virtual world won't teach the real world much about the dynamics of disease spread. He suggested the lessons would be sociological, not virological.

"Because in essence, these are electronic universes and the disease will behave as you've programmed it to behave," said Ferguson, whose modeling work is the basis of the World Health Organization's plan to try to stop a flu pandemic at source using drugs and quarantine.

"I mean, the behavior of the population will modify that and you can observe those behavioral modifications on that. So that might be interesting. But one can't assume you're going to be really inferring anything about the epidemiology of a real disease. ... It's the behavioral aspects which are interesting."