by Nick Yee
Are people polite to computers? Given that computers are inanimate objects without feelings, this question may seem ridiculous. But a study conducted at Stanford University in 1996 showed that people interact with computers as if they had feelings. Communication scholars Clifford Nass and Byron Reeves had college students take a tutoring session from a standard desktop computer. The session was about different facts of American culture, such as the percentage of American teenagers who kiss on the first date, and included a quiz on a set of questions the computer had not tutored students on, followed by a scoring session in which the computer went through the students’ responses and let them know how they performed. The students were then asked to complete an evaluation of the tutoring session either on the same computer or on a different computer.
When a family member asks you what you think of his or her cooking at a family gathering, you tend to be polite and avoid offending that person. If someone else pulled you aside and asked you the same question, you’d probably be more honest. It turns out that people obey this politeness rule even when interacting with computers. The researchers found that students gave more favorable evaluations if they filled the form out on the same computer that tutored them. Students who filled out evaluations on a different computer gave less positive responses. Given that these college students were all familiar with computers, they did not consciously believe that computers had feelings. Instead, as Reeves and Nass argue, users rely on existing social norms when interacting with new technology. And we do this because our brains lack the cognitive resources to create and follow entirely new social protocols for every novel class of technology we encounter. When a computer asks us to evaluate its cooking (so to speak), we subconsciously treat it as if it were a person asking us, and the politeness rule is triggered. Without even being aware of it, we treat computers as if they had feelings and could be hurt emotionally by our remarks.5
As another example of how we fall back on existing social norms even in new technological spaces, consider the notion of personal space in virtual worlds. In the physical world, the amount of personal space we give another person depends a lot on whom we’re talking to and what we’re talking about. Intimacy, for example, can be expressed either with eye contact or by moving closer to another person. When one of these cues is accidentally triggered, such as when we are crammed next to strangers in an elevator, we modulate the other cue to maintain the appropriate level of intimacy. Thus, in an elevator, we turn away from and avoid eye contact with the people next to us to defuse the cues of uncomfortable intimacy. In a study my colleagues and I conducted in Second Life, a virtual world in which users can create their own content, we wondered whether people moving around with the mouse and keyboard in digital avatars would nevertheless conform to these physical norms. And it turned out that this modulation of eye contact and personal space indeed occurs in Second Life; people standing close to each other in Second Life are less likely to be looking directly at the other person. Instead of developing new social norms, we fall back on the ones we’ve learned from the physical world.6
The same is true for superstitions. False contingencies trigger superstitious behaviors around highly desirable rewards, whether we’re talking about pigeons in Skinner boxes or people in online games. For pigeons, this is food pellets. For online gamers, the rewards are magical items, rare monster spawns, or over-enchanting equipment. When a superstitious idea emerges, it can be inadvertently reinforced, and then social dynamics such as low relative cost help it spread across a community. And once a superstitious ritual spreads, it takes on a life of its own, and not even the direct refutation by game developers can quash the superstition.
Our digital bodies are fluid, mutable with the click of the mouse. Our fantasy worlds, with their elven druids and galactic starships, seem far removed from the physical world and infinitely malleable. But the reality is quite strange and sobering. Even if virtual worlds were tabula rasa, we are encumbered with a great deal of cognitive baggage. Our brains are hardwired with many mental shortcuts to help us make sense of the world. We simply do not have the time to carefully process every piece of information that comes our way. To cope with this inundation of information, our brains have developed automated heuristics that filter and preprocess this information for us. Thus, when we encounter new media and technological devices, we fall back on the existing rules and norms we know. We react to computers as if they were human and had feelings. And when we enter virtual worlds, this mental baggage hitches a ride with us. We react to digital bodies the way we react to physical bodies. And the same psychological triggers that lead to superstitions in Skinner’s pigeons lead us to develop superstitions in online games. This is an example of the Proteus Paradox: how our brains work doesn’t change when we slip into a digital body. In a fully digital technological construct meticulously built from rational, precise program code, the irony is that superstitions persist and flourish.7
The Supernatural in Online Games
The likelihood is high that most of these habits are just superstitions. On the other hand, although it is easy and standard for game developers to use random number generators throughout a game, it takes only two lines of program code to make a four-leaf clover bring you good luck on Tuesdays. And although we are unlikely to settle any debates about the existence of God in the physical world, there actually is an omnipotent, omniscient god in online games known as the game developer, who can and does change the rules and bend the laws of nature. Causal relations in online games can be magical, defying physical laws. There is no logical or scientific reason why a four-leaf clover would bring you good luck in the physical world, but there is a rational and scientifically sound reason for why this might happen in an online game.
In fact, the reason why players in Final Fantasy XI believe that cardinal directions and moon phases have an impact on crafting is because moon phases actually do have an impact on certain well-documented aspects of the game. For example, some magical equipment is enhanced when the moon is in its crescent phase (both waxing and waning).8
Final Fantasy XI’s crafting system was particularly ripe for superstitions, because the parts of the system that were verified were wacky enough that anything might have been true. [Final Fantasy XI, female, 23]
Hidden rules not only perplex gamers, they also perplex game developers. The interwoven complexities of program code make it difficult even for game developers to identify non-obvious bugs. This was the case in Asheron’s Call. When a few players began to complain that their characters were perpetually unlucky and unfairly targeted by monsters, it was easy for the game developer, Turbine, to brush them off and claim that it could find no such bug. On the face of it, the notion of monster-haunted characters did seem like digital hypochondria. In the game community, this rumor was referred to as the Wi Flag, named after a character named Wi who widely discussed his torment. And everyone had a good laugh out of it for a few months, until it was revealed by Turbine that there was indeed a bug.
Our developers at Turbine initially answered these complaints by saying that they could find no such bug. . . . Easy culprits, such as a malfunctioning random-number generator, were eventually dismissed. But our search went on. . . . And then one day, long after most people had learned to either forget or ignore the Wi Flag, the answer was found. . . . We hope it is of some interest to those of you who have long been afflicted with this terrible burden.9
In Asheron’s Call, monsters choose whom to attack at any given moment based on metrics such as who attacked it last or who is doing the most damage, but these metrics do not apply when a group of players first appears within a monster’s attack radius. Turbine’s code for the game had an error in this part of the decision-making algorithm. Monsters were more likely to attack players at the beginning of the group list. A character’s identification number, permanent and assigned during character creation, determined the sorting order in the group list by mistake and, thus, the likelihood of attack. In other words, some players
in Asheron’s Call were indeed perpetually unlucky and haunted by monsters.
Whether directly implemented by game developers or inadvertently introduced into the game via bugs, magical causality is not only plausible in online games, but we can actually point to specific instances of it. Moon phases can affect your performance. People can be born unlucky. And even though magical causality in online games is uncommon, believing in the supernatural in online games is not entirely irrational. Thus, online games not only hijack our psychological wiring to encourage superstitious beliefs and rituals, but the plausibility of supernatural beliefs helps propel and sustain these superstitions. Our social training and brain wiring follow us into these new worlds. We think of technology as something that promotes rationality, but technological constructs can actually promote superstition. And as we’re playing games, we’re also being played, driven to dance in virtual boxes.
Superstitions in online games reveal several surprising aspects of game design. First, game designers often don’t have complete control over the social systems they create. In the case of Dungeons and Dragons Online, players rejected the developers’ statements that conflicted with their own beliefs on negotiating with treasure chests. And second, players are actually creating a great deal of game content even in the virtual worlds in which they ostensibly cannot create game assets as we traditionally define them. Players are not able to create or modify creature models or change the code in an online game, but superstitions can create new experiences and social interactions for many players. For them, these superstitions are as much a part of their gameplay as the elements hardcoded by the developers. In a way, superstitions are free content for game developers; they are stories that require no additional resources or effort to create.
We tend to see superstitions as irrational, even primitive. And it’s hard at first to see how ritual dances can be a good thing. But the key to every good story is engaging with the audience. A predictable story with no surprises is boring. The television and filmmaker J. J. Abrams refers to the use of a “mystery box” as a storytelling device to engage the audience as a story unfolds. Thus, in his monster movie Super 8, the audience never sees the monster until the last ten minutes of the movie. Until that point, the audience is actively engaged in guessing what the monster looks like based on the clues left in its destructive wake. And when the fans of TV show Lost were not watching the show, they spent a great deal of time thinking, talking, and posting on forums about the show’s mysteries. Although superstitious rituals would be a bad thing in the jury room or the classroom, I would argue that they are indicators of engagement in online games. Whether it’s a story or a video game, nothing engages an audience like a mystery they can help solve. On the other hand, irrational behaviors are inherently hard to control. In chapter 5, we’ll see how volatile beliefs have led to racial profiling in online games, but to understand how those beliefs emerged, I first need to explain why online games are so much work.
CHAPTER 4 THE LABOR OF FUN
Until I played Star Wars Galaxies, I never knew how absorbing industrial entrepreneurship could be.
The late afternoon sun was making me uncomfortable under the hooded robe and thick Wookie fur. The wild wheat field had been slowly depleted over the past week. Other surveyors had come and left their own automated harvesters—ugly metal installations that slowly blanketed the river delta. This wheat strain had a superior decay resistance, and I had frantically stockpiled it for my production of biological effect controllers. My pharmaceutical factories had ground to a halt three days ago owing to the break in the supply chain, but now, with this new wheat stock, I was hoping my luck was turning around. My clients were flooding the mail console, and if I couldn’t deliver new stock soon, I would probably lose many of them to a competitor.
I turned to the sound of the approaching swoop speeder bike. It was that annoying donkey-faced Bothan again. I wondered if Laza ever slept. I had to planet-hop to find a new wheat source, and the quality was unlikely to match this wild strain. Perhaps a jump to Naboo first? I glanced at Laza, who was emptying his harvesters. I had a twenty-minute head start on him. I hopped on my own swoop speeder and headed toward the travel terminals. Earlier in my career, I had wanted to become a doctor, but the risk and adventure of pharmaceutical manufacturing certainly had its appeal.
Corellian Wheat
Yes, pharmaceutical manufacturing is a profession in a video game. We tend to think of video games as an escape from work, a mindless diversion. But just as superstitions find a way to flourish in virtual worlds, so, too, does work. As the technical capabilities of games increased, the complexities of play also increased. In 2003, two games launched that both had full-fledged player-driven economies: Star Wars Galaxies and EVE Online. In Star Wars Galaxies, which we will consider first, almost everything that was bought or sold in the game was created by another player. Apart from pharmaceutical manufacturers, players could specialize as tailors, architects, and even bioengineers.1
The game captured the mundane minutiae of the manufacturing process from geological surveying to retail advertising. Not only did players have to assemble the components to create the final product, but there was a player who surveyed for the raw resources, a player who harvested the resources, a player who created the product prototypes, a player who used a factory to mass-produce the product, a player who developed the advertising, and a player who created the retail store to sell the product. The player population determined entirely the supply and demand of all these raw, intermediate, and finished goods. This round-the-clock commercial activity occurred in a public market called the Bazaar as well as through privately owned merchant droids spread throughout the galaxy. Truly dedicated players might try to manage an end-to-end production chain, but more often than not, informal cartels formed in which players could focus on a specific segment of the production chain.
The manufacturing process in Star Wars Galaxies was bewilderingly complex for players coming from hack-and-slash massively multiplayer online games. First, the quality of manufactured products depended on the quality of the raw resources used, and the galaxy was populated with a broad array of raw resources. For example, the biological effect controller (an intermediate component) required an organic component, which could be avian meat, berries, or wheat. The quality of the controller depended on the potential energy of the organic component. The math underlying manufacturing outcomes was sufficiently complex to warrant many player-written guides, littered with equations, such as:
MAX_EFFECTIVENESS = ((Resource1_OQ + Resource2_OQ)
/COMBINED_MAX_OQ) * 0.66 + ((Resource1_ + Resource2
_PE)/COMBINED_MAX_PE) * 0.332
Knowing what kinds of resources to seek out was only the first part of the problem. Locating high-quality resources required a lot of legwork. These raw resources were randomly located throughout planet surfaces in the galaxy. Once located through trial and error with the surveying skill, these resources could be harvested by hand or by setting down automated harvesters, which accumulated the resource slowly over time. Thus, early comers to a resource patch could take the most resource-rich spots, which yielded larger harvests. To resolve the problem of veteran players taking up all the rich deposit locations, every seven to ten days after a resource had appeared, the game would replace that resource with a randomly generated resource of the same class at a different location in the galaxy. It was a never-ending game of musical chairs, and surveyors had to be on their toes to avoid breaks in their supply chains. Several third-party sites emerged specifically for players to input and collectively keep track of the emerging resources. Owing to the difficulty of maintaining high-quality resource stocks, many players made a living in the game merely from harvesting and selling raw resources.
Once players had located and harvested the resources, experimented with prototypes, and used factories to mass-produce the final products, they were faced with the biggest obstacle of all—each other. Because products had to be bought
by other players, retail marketing was the true endgame. Successful entrepreneurs would often create a memorable brand name across their product offerings to create loyalty, especially important for their new product lines. Another strategy was to hire architects to design stylish and eyecatching stores to attract customers and enhance the shopping experience. Hiring exotic dancers also didn’t hurt. Some players banded together to create centralized malls. And others took advantage of more devious tactics:
I set about disrupting the sales of the other master weapon-smiths. I targeted one at a time. I would find out the location of their shops and set up a shop nearby. I’d pay players to loiter near the other weapon-smith’s stores and point people over to my store with cheaper prices and I’d pay entertainers who were working in bars across the galaxy to advertise my store between sets. Finally, to really put the nail in their coffin, I’d pay a smuggler to use a special skill called weapon-splicing to improve my master-crafted weapons even more, thus giving my weapons the edge in quality as well as in price. [Star Wars Galaxies, male, 24]
Star Wars Galaxies created a living, breathing economy, and the hack-and-slash aspect of the game actually paled in comparison to the financial opportunities. Overwhelming an opponent in combat was one thing, but gaining the majority market share in a product segment while driving a competitor out of business certainly had its appeal.
Hadean Drive Yards
EVE Online, another online game with a player-driven economy, takes the work metaphor one step further. Player organizations in EVE Online are called corporations. Similar to the complex economy of Star Wars Galaxies, EVE Online players participate in a manufacturing-based economy within a dangerous war-torn galaxy. Resource-rich areas are constantly changing hands between factions depending on large-scale military outcomes. Furthermore, mining ships are under the constant threat of being hijacked by pirates and thus often hire mercenaries or have corporation members as bodyguards.