Laura Bliss is CityLab’s West Coast bureau chief. She also writes MapLab, a biweekly newsletter about maps (subscribe here). Her work has appeared in The New York Times, The Atlantic, Sierra, GOOD, Los Angeles, and elsewhere, including in the book The Future of Transportation.
Economists look to massively multi-player online games for insight into real-world money questions.
My friend is exasperated with her younger sister, who's been crashing on her couch. Recently she asked her sibling to get up early to let the electrician in, but the sister slept through the appointed hour. She'd been up until dawn, playing World of Warcraft.
What was she doing for so long inside the game?, I wondered. She might have been counting out bloody coins for a nice new tabard, or sifting through Silverpines in search of spellbooks. Maybe she was extracting a gas cloud for raw motes (a kind of raw resource), or taming sporewalkers (a type of beast). Or maybe she'd been chatting with some other player, one encountered among the 6.8 million that inhabit the Warcraft universe. These are all things you can do to develop your character, build relationships, and make money there.
"When will she learn?" my friend groaned. But some researchers are asking an inverse question: What can the real world learn from what's happening in virtual ones?
Massively multi-player online games (MMOs) like Warcraft, EVE Online, or Everquest are synthetic environments, but they function in ways that parallel real ones, socially and economically. Yes, sporewalkers and spellbooks are fantastical, but scholars have noted players spend real time and effort to produce, distribute, and consume virtual products and services.
A 2009 study led by Edward Castronova, an economist and media expert at Indiana University, took up the question of whether virtual worlds could be considered “economies” in the way a real nation might. The authors examined the MMO EverQuest2 and found that, “with the buying and selling of goods, banking, the creation of goods, and the equivalent of income, derived from taking currency and items from defeated monsters,” the characteristics of a functioning economy were there. Furthermore, a GDP could be calculated, and price levels of various items across the game responded to fluctuations in inflation. "People approach pricing and wealth in virtual environments in much the same way as they do in the real world," Castronova told me.
That’s part of the reason MMOs are so alluring to Castronova—who’s often cited as a founder of the field of virtual economics—as well as to a growing legion of social scientists. Another draw is that these virtual worlds are self-contained and user-driven: For the most part, game developers don’t intervene in the action; each player is following their own quest, forging their own path, building their own networks. People are making decisions on their own.
And all of those decisions, actions, and interactions are captured automatically. In the real world, economists would have to reply on statistics and sampling methods to estimate economic aggregates, but in a MMO there’s no such need. "A multi-player game environment is a dream come true for an economist," Yanis Varoufakis told Reason.com. Varoufakis is a Greek economist, recently hired by an MMO manufacturer to be an in-house economist (think Janet Yellen, but for a world of digitized zombies). "It’s like being God, who has access to everything and to what every member of the social economy is doing."
Not all MMOs have gone so far as to actually hire academics to regulate their games. But some have opened up their data, allowing outside economists to ask questions and even perform experiments that would be nigh impossible in the real world. A study published in PLoS One last month, for example, takes up this inquiry: How, in a contained economy, does wealth arise among individuals? Can it be a function of a person's behavioral patterns, or their position within social networks?
It is a variation on the difficult and nuanced question of how inequality originates. Here in the U.S., where income inequality widens by the hour, there's not a firm consensus on why it's happening, or how to even measure it. But with a complete, virtual data set, it's possible to tease out strands of causation.
Stefan Thurner, professor of Science of Complex Systems at the Medical University of Vienna, and graduate student Benedikt Fuchs examined how individuals accumulate wealth in Pardus, an MMO set in the galactic future. There, players traverse a universe of planets, space cities, resource fields, and even space monsters in a personal spaceship. Pardus' economy is robust, though highly unequal; Using a vast data set reflecting the wealth of all 40,785 players, second-by-second, since 2006, the authors calculated an average Gini index of .65. (That index represents the income distribution of a given nation's residents.) The graph below charts the measure of wealth equality among all players in black:
It also shows two others groups: Players who were in alliances (in blue), and those who were not (in red). Alliances are simply groups of players, usually no more than 140, that can be officially declared through the game's built-in tools. Members of the alliances pay into a common cash pool which is often used for exploration, defense, or for economic purposes, like building up production chains or coordinating production facilities.
What's clear about the graph is that those who formed alliances had a more equal distribution of wealth than those who were not in alliances. And that points towards the most important finding the researchers made: Using a data set reflecting the wealth of 12,186 distinct players, second-by-second, since 2007, the authors found that the most significant factor in terms of an individual gaining wealth was their position in social networks, particularly in their trade networks.
In the chart to the left, the wealthiest players are shown in red, and the least wealthy in blue. The richest are those with the highest degree of both "trade-in" and "trade-out," meaning they were participating more in not only selling goods that they themselves had produced (trade-in) but also buying more from other players (trade-out). Another chart also showed that the wealthiest players were selling to people less connected in the trade network than themselves.
The next charts, above, show how wealth-gain was related to a player's friend and enemy network. Here, the researchers were greatly helped by the fact that within the game, players actually mark other players as friends or enemies. In Chart D, the wealthiest players are those that are liked by more players (friend in) than they like themselves (friend out). Poor players, the authors note, mark other players as friends more often than others have marked them. And in Chart E, the wealthiest are very rarely marked as an enemy by others (enemy out), but do sometimes mark others as enemies (enemy in), but generally these are enemies shared by many other people.
In sum, the authors write, the wealthy organize their networks so they are the "hub of a star-like network." Wealthy players trade with many others, while their trade partners traded with fewer others, and very little among each other. In the friendship and enmity, the wealthy are well-liked and withhold animosity—except to public enemies.
Whether these correlations are traceable in the real world isn't really possible to say—and that's the point of this research. Again, the data here is so finely captured that it would be prohibitively difficult and expensive to collect in a non-virtual setting. But it is worth pointing out that Thurner has published previously on other kinds of social dynamics within Pardus, and has proven how some mimic real social mores—such as how men and women chose friends differently, how most people fundamentally behave compassionately rather than antagonistically, and how people tend to seek friends more than enemies. "We're making revolutionary use of new data to try and to understand Homo sapiens a little bit better," Thurner told me.
With that in mind, Thurner argues that the insights from his latest research can be applied in real-world settings. "To make money, position in your networks is so important," he said. "Job-searchers should be thinking proactively about where they are in theirs." Right now, his team is looking to secure funding to create an online job-placement app that would show a user their place in her real "trade" network (how many other bankers she knows, or seamstresses, swim coaches, etcetera). Think LinkedIn, but with a greater emphasis on mapping out where a user stands among colleagues in skill-set and connections.
There will be economists who disagree that it's appropriate to extrapolate from virtual economies, which, while they mirror real ones in so many ways, also bear obvious differences: These are economies where no one really subsists, where there aren't traditional tax structures, where male players dominate.
But harder to argue with is this: As digital currencies like Bitcoin become more ubiquitous (Ecuador's soon-to-be-named digital dollar was announced just this week), we have virtual worlds to thank.
"Games have always had the digital wallet," Castronova said. "By which I mean, different kinds of currency, where you pick what to pay with. We're getting closer to a real world version, where you go to Target and your wallet talks to the cash register. You've got reward points, credit card points, Bitcoin, credit, cash. How do you pay? It won't be long until these wallets get put into a system that helps users figure out how best to pay, making currency really fluid."
And perhaps, also, our perception of the value of these virtual worlds.