“How do game events marking success versus failure affect a player's level of engagement?” The answer to this question marked the “number one” piece of research done in game studies, say Ian Bogost, Jane McGonigal, and Mia Consalvo.

Though game developers don't often hear much about it, there's a breadth of research taking place in academia that examines video games—and not just whether games are good or bad for the players. On the final day of the 2006 Game Developers Conference, three academics presented ten research projects from 2004 and 2005 whose findings could impact the way game creators think about and make games.

Ian Bogost, assistant professor at Georgia Institute of Technology, Mia Consalvo, associate professor at Ohio State University, and Jane McGonigal, a PhD candidate at the University of California, Berkeley convened to decide which research would make the top ten cut.

Although the research they considered was not initially limited to the sciences, they opted to leave out work done in humanities departments simply because it was not quantified and thus not easily presentable in the 60-minute time slot afforded to GDC sessions. McGonigal even called the presentation a “rapid-fire top ten countdown,” in reference to the four- to five-minute intervals in which each complex work of research was presented.

The top 10 research projects they find most valuable to game developers (followed below by a brief synopsis of the relevant findings) asked:

The presenters also half-jokingly included a “number zero” research study that asks, “How do game developers design for spectators as well as player experience?”

Number 10, by G. Cassidy et al., Glasgow Caledonia University, found the high emotional impact of music did not correlate to player effectiveness. However, when players picked their own soundtracks, their ability to succeed in the game increased and they become more emotionally responsive to the activity as well. McGonigal summarizes that “game music is not just about emotional impact or world-building. ... Player success actually hangs on it.” Developers should consider thoughtfully how and when they use game music to support or challenge players, she says.

Number 9, by K. Hew, M. R. Gibbs, and G. Wadley of The University of Melbourne, uncovered aural feedback that players actually found disruptive, such as noise, speech not intended for them, and trash talk. Poor voice chat usability, in other words, hinders the players' attempts to be social, subverting the very goal of voice chat. Consalvo explains that the research participants slowly could adapt to the ambient and distracting noises, but usually removed their headsets when the sounds became “too troublesome.” If developers choose to use voice chat, she advises it be targeted in a way that's very specific to the game.

Number 8, by S. Griffin, The Georgia Institute of Technology, discusses the advantages and disadvantages of having more physical interfaces, most notably peripheral or novelty controllers, such as the guitar used to play Guitar Hero or the as-yet unreleased joypad for Nintendo's forthcoming Revolution. Bogost explains that these devices should, theoretically, drive the player to reflect the same gestures being performed on screen. Unlike standard joypads, which abstract the movement of the player's body (pressing buttons with fingers) from the movement on the screen (button-pressing responses), controllers that take into account the player's own physical body and motions are less symbolic. However, simple button-pressing works best when the actions on screen are themselves complex and symbolic. Bogost suggests game developers use this research to question the “balance between a gestural control and a symbolic control.”

Using multi-player games such as Halo 2 for Xbox Live, City of Heroes, and EverQuest, the researchers of Number 7, C. Campanella Bracken et al. of Cleveland State University, discovered that “collaboration is an extremely powerful driver of emotional stickiness,” says McGonigal. The findings indicate (of gamers who played at least 12 hours per week) that players depersonalize their adversaries and do not feel a strong personal awareness of them. Player collaboration, on the other hand, resulted in the strongest sense of presence, meaning when gamers work together with other gamers, that's when they have the greatest sense of community awareness.

D. Miller et al. of Stanford University 's department of Personality and Social Psychology conducted the number 6 study. Although their work did not involve games, says Consalvo, the findings are straightforward enough to be transferable. The study placed observers into a room where young students were administered an exam. When the observers were informed that particular students had a high probability of cheating, the observers reported that they saw the students cheat or saw potential for the student to cheat, even though the student was instructed by the researchers to not cheat. In the game industry, developers can learn from this study the importance of perception, says Consalvo. “Perceptions are often more important than realism for fairness in multiplayer games.”

Michael Nitsche, who conducted the number 5 study at Georgia Institute of Technology, looked at player-controlled cameras. Although player-controlled cameras are the interactive equivalent of cinematic montage, its use in games as such is rare, says Bogost, who cites the “sniper” camera angle as an example. Bogost claims that Nitsche's research shows why points of view have meaning inside of the gameplay experience. Other games Bogost mentions as examples are GoldenEye (sniper), Siren (“sight-jacking”), Doom (overhead), and Fatal Frame II (third-person fixed perspective).

T. Manninen and T. Kujanpaa of University of Oulu, Finland, in the number 4 study looked at Battlefield 1942. McGonigal, strongly suggesting that developers read Manninen and Kujanpaa's work in its entirety, is particularly fascinated by their area of research. This ethnographic study compiled data on how players gave cues to one another in the game and found players wanted to communicate in ways that the game did not support. Players wanted to coordinate with their teammates in ways that were not supported by the game mechanics. For example, players were unable to gesture to one another or make other non-verbal cues. They also could not interact with other players without violent physical contact. McGonigal notes that the workaround strategies players invent, such as pointing with weapons instead of pointing with hands or arms, are still limiting. Because players are already inventing ways to communicate in these alternative ways, McGonigal says developers need to seriously explore these possibilities to better meet their needs.

Number 3, conducted by Erika Jonsson at the Royal Institute of Technology, Sweden, found players enjoyed playing Half-Life more when they used an eye-tracking device combined with their mouse control. “Use of eye-tracking could be a successful addition to your game, provided it has a useful function and is properly play-tested,” says Consalvo. Participants in the study actually earned higher scores in the game when they used the eye-tracking device.

Number 2, out of the University of California, Los Angeles (Y. Cao) and University of Southern California (W. Tien , P. Faloustos, F. Pighin), was a technical study of emotionally influenced facial animation used for speech. Speech-driven faces are common in games, says Bogost, but people look and speak differently under different emotional states, so the faces would be more expressive and embody realism better if the animation matched both the variable emotion as well as the speech. Bogost notes that developers may try manually specifying emotional content or try using a support vector machine to identify emotional content based on a script.

Failure is remarkably important to game players. That's what Niklas Ravaj et al. from the Helsinki School of Economics found when they compiled their research. Using the game Super Monkey Ball 2 for GameCube, they examined how player success and failure affected the player's level of engagement. McGonigal calls their findings “counter-intuitive,” noting the participants felt more pleasure and excitement in active failure than in success. Passive failures, on the other hand, leave players feeling less engaged. So the ways in which developers make failure possible—either active or passive—will have a significant effect on how players receive the game. “It didn't matter that within the game [the players] were doing really terribly,” says McGonigal. “There's a certain satisfaction of sending a monkey into space.”

Bogost, Consalvo, and McGonigal's presentation containing a full citation of the research summarized at the conference is available on McGonigal's website: www.avantgame.com/top10.htm.