UK Edition
Corrupted Blood and coronavirus: what World of Warcraft can tell us about the spread of disease
One small coding error caused a pandemic that killed millions in the online game – and it could still be of interest to scientists today
At first glance, you wouldn’t think World of Warcraft has much to teach us about life in 2020. The massively multiplayer online role-playing game, released in 2004, features elves, dwarves, and giant bipedal bulls – not to mention teleportation and self-resurrection.
And yet, thanks to one little coding error, the game has a unique place in the annals of disease management in the real world. It could even offer an insight into how coronavirus is currently spreading around the globe.
To understand how, you have to go back to 2005, when the game was still a few months old. It had already found a positive response among gamers, who interacted with each other in this online world through avatars, while completing quests and battling monsters. Blizzard Entertainment, the game’s developers, decided they would introduce a new villain for players to fight: a monster who drained players’ blood.
The monster could only be defeated by players afflicting themselves with a spell called Corrupted Blood. They could pass this ‘disease’ to other players by standing beside each other – but only in a sealed-off area of the game world.
Or at least, that was what was meant to happen. Due to a coding error, Corrupted Blood infected the animal companions of players. And while players were healed upon defeating the boss, their pets were not. Soon ‘cured’ players were reinfected, and able to pass Corrupted Blood onto other players away from the sealed-off areas. And thus, Corrupted Blood escaped into the wider world (of Warcraft).
So far, so fantastical – to you or I, at least. But to World of Warcraft player Eric Lofgren, Corrupted Blood had fascinating real-world implications. Lofgren is an infectious disease epidemiologist; seeing the bodies that were piling up on his computer screen, he picked up the phone and called Nina Fefferman, his colleague at the time, and co-director of the Centre for Modeling Infectious Diseases at Tufts University. “You should log in,” he said. “That thing we were talking about is actually happening.”
What interested Lofgren and Fefferman is that despite Corrupted Blood being a fictional disease in a make-believe world, World of Warcraft was very much influenced by human behaviour.
The difficulty with the study of how diseases spread is that research tends to be either observational and retrospective, in which case the data is hazy; or it relies on mathematical modeling, which some experts contend is unable to accurately predict human behaviour. Here, however, was a chance to study the reactions of real people in a uniquely complex virtual environment, without any actual life-or-death results (although, given how much time and energy players had invested in their characters, they were evidently keen to survive – just as they would be in real life).
“Unlike previous ‘virtual plagues’ that had been officially planned, [Corrupted Blood] was a local effect that went out of control — a naturally occurring virtual outbreak,” wrote Ran Balicer of Israel’s Ben-Gurion University, in the journal Epidemiology.
It’s a theme that Fefferman, who published an academic paper on Corrupted Blood in 2007, picks up on. “If someone tells you a natural disaster is heading your way, do you leave?” she said. “The answer is `Yes, I leave immediately.' And then when it actually happens, a lot of people have the emotional response of `You know what? I'm going to guess it's going to be OK. I'm gonna try and stick it out.' And it's a very reasonable emotional response, but it's very hard to predict even in yourself.”
Fefferman notes other things players did in response to Corrupted Blood that could inform mathematical modelling. For example, there was a rubbernecking effect: players would head into disease-afflicted towns to witness the devastation for a moment, before fleeing in the hope that short exposure wouldn’t lead to them catching the disease. Scholars have compared this to international journalists rushing to interview witnesses in disease-filled areas, then leaving again and potentially taking the illness with them.
What’s more, non-playable computer-generated characters were able to become infected by Corrupted Blood without showing symptoms. They were just like asymptomatic super-spreaders in the real world.
Researchers noticed that, as virtual cities became breeding grounds for the disease, littered with the corpses of players who’d succumbed, healthy gamers escaped to countryside areas where low population density made them less likely to contract the disease. Meanwhile, other more altruistic players would act as first responders, putting themselves at risk. Some went as far as setting up hospitals to help deal with the outbreak.
Even the games’ more fantastical elements offered real-world insights. “Attempts by game administrators to quarantine infected areas proved futile due to the ability of characters to rapidly teleport to distant lands,” wrote Balicer. “This is similar to the role of air travel in the rapid global spread of severe acute respiratory syndrome (SARS).”
Fefferman, Lofgren, Balicer, and other researchers began to posit the theory that the behaviours displayed in World Of Warcraft could be used as a basis for research into the spread of illnesses in our world too. But would it work?
“It is very interesting,” says Christl Donnelly, professor of applied statistics at Oxford University and professor of statistical epidemiology at Imperial College London. “It would be interesting for facilitators to have discussion groups with people afterwards to say, ‘why did you choose to do this?’”
However, Donnelly warns that any attempts to predict behaviour can be fraught with difficulty – even if they’re based on a sample from a game. “There are cases where what people tell you they would do is not the same choice they would make in the real life situation, and that's still true in this game as opposed to real life,” she explains. “There was a study for an emergency situation and they looked at traffic patterns and they had a completely different idea of what would happen because they didn't count on all the parents going to pick up their children from nursery and school as a priority.”
Another problem would be that people’s behaviour tends to morph as more information becomes available. “The decisions that people make will change over time depending on how things have unfolded and what evidence comes forward,” Donnelly explains. Still, she says, studying a game could “certainly help to anticipate behaviour change,” when it comes to calculating the potential impact of a disease.
While Blizzard were initially enthusiastic about researchers using their data to study the spread of disease, they subsequently cooled on the proposals and reiterated that World Of Warcraft was "first and foremost a game” and never “designed to mirror reality or anything in the real world."
Unfortunately, what the Corrupted Blood incident will never tell us is how to cure infectious diseases. After a week of pandemonium in the game, Blizzard Entertainment resolved the problem by resetting the games’ servers and patching the bug by making pets immune to the plague. In the real world, there are no such resets. Vaccines take a lot longer to develop than lines of code.