Ethics in Game Design

The difficult topic of "ethics in game design" (not to be confused with ethical considerations about making games) was addressed by a panel headed by Bernard Yee, Director of "Gamer Programming" at Sony Online Entertainment -- another behaviorist maybe. Yee was joined on the panel by Bob Bates, Toby Ragaini, Austin Grossman, and Doug Church. Beginning with a brief introduction to ethics, the discussion moved swiftly but didn't really break new ground.

Altruistic decisions require choice and consequence within the game. Yet in-game consequences cannot include rewards, only penalties. As one panelist put it, generosity points defeat the purpose.

The idea of player decisions not guided by the cruel equations of in-game economy seemed very troublesome to some. Other voices from the audience dismissed the topic outright, claiming that games without real-world consequences could not possibly have an ethical dimension at all. In my view, such an attitude denies that our thoughts and reactions have ethical -- or other -- relevance. A truly ethical mind does not stop evaluating just because it has entered the reality of fantasy or daydreams.

Addressing first audience comment (which, predictably, failed to separate the designer's ethic from the ethical implication of the design), the questioner cited an Infocom game that included the option to torture an NPC. In an ironic twist, it turned out that panelist Bob Bates had himself suggested adding this element to the game in jest -- and had resigned from the project when it was actually added. Entertaining distractions aside, the panel somewhat sidestepped the "Torture? Y/N" thought experiment. In my opinion, the resemblance to classical psychology experiments on obedience and ethical choice is striking. Games will be considered a medium of self-expression and self-exploration only when players question their actions before and after the fact.

Similarly, there was dissent on the possibility of ethics in single-player games. Toby Ragaini claimed that single-player games could merely be educational, as ethics requires an affect on a human being. I oppose this assertion, based on findings on how children explore ethics by searching out entities on the border between living and dead, be it insects or (as detailed in Sherry Turkle's The Second Self: Computers and the Human Spirit) computers. Of course, a linear single-player game which doesn't allow any player choices leaves little room for ethical decisions (but does not rule out internal response to perception). The black-and-white world of Half-Life offers "questionable" design, not ethics. Doug Church pointed out that even Ultima 4, "poster child of ethical behavior," offers no other decision but to either go with the (linear) game flow, or simply stop playing at all.

For me, one of the most intriguing observations was how the panel and audience quietly subscribed to Manichaeism in its purest form. From the quip about DOOM as a "Christian shooter" (we are shooting demons, after all) to a brief exchange about Lionhead's Black and White, it seemed as if games allow only for right or wrong at most. Bob Bates was the only exception, conjuring an example where there is no "right" decision, just a legitimate conflict of (NPC) interests. If the mold of first person games can be described by "If all you have is a hammer, everything resembles a nail", then the confinement for "ethical" games might be expressed by "If everything is either reward or punishment, everybody looks like a dog."

It was quite telling that the panelists had to struggle for a response to the question as to whether they had ever "played a game that had ethical choice." Most examples (like cheating on AI players in Alpha Centauri) were not convincing. The panel and audience finally settled for the sacrifice of the sidekick "Floyd the Robot" in Steve Meretzky's Infocom game, Planetfall. Unfortunately, the designer himself pointed out that the sidekick himself initiates the act that leads to its demise, and that the player does not know the outcome. There is seemingly a lot of power in "perceived consequence" as opposed to actual choice -- powerful enough to make a roomful of game designers blame themselves for something the game designer had plotted.

Other aspects covered in the discussion included whether ethics in games requires the presence of a (human) audience (and therefore only multiplayer games qualify in that regard), whether the in-game ethical problems that plague online worlds (like player killing or looting corpses) should be adressed inside the game or outside, and whether players actually want to play "bad guys." Yee pointed out that TIE Fighter didn't sell as well as X-Wing, but the audience countered by pointing out the success of Bullfrog's Dungeon Keeper. At one point, Yee decided to blame the lack of ethics in games on the limited number of actions available to most players. In his view, WALK, SHOOT, and RUN might not be sufficient to respond to an ethical dilemma. Saved games and replayability were other scapegoats.

I did not find the attempts to define ethics as non-optimal decisions (with respect to personal gain) entirely convincing. Yee's claim that a "hero never reaps reward" falls short -- it may just be the definition of reward that changes. It is equally possible to say that ethical decisions optimize with respect to a different cost function (see Kant's "categorical imperative," or even the examination of apparent altruism in sociobiology).

The analysis of "tit for tat" in game theory is an interesting perspective when considering players that cheat on AI partners in strategy games. Conversely, "stable" strategies as defined by John Maybard Smith might be a source of inspiration for online games.

In the end, whether outside or inside the game world, the economy -- not ethics -- guides most decisions. If game designers expect to put ethical considerations into the heads of gamers who couldn't care less, their designs will fall apart, with or without online community.

One other observation suspiciously absent from these discussions on ethics was that ethics is "no fun." Ethical dilemmas hurt. Witness the sweetness of the classical Hollywood movie "kiss off" contrasted with the haunting quality of an open, ambiguous ending. All things considered, the audience was probably right on target in suggesting that the ethical dimension of a game is brought about by raising questions, not by providing answers.

Let me conclude by making some observations about Yu Suzuki's Shenmue keynote presentation. In my blessed ignorance, I experienced the presentation of this accomplished designer's work as a history lesson on computer games. Having set out to create his first game decades ago with a team that fit into a single room, Suzuki commanded 300 internal and external contributors and a staggering amount of resources for what he calls a "cinematic RPG." I could not help but compare the skyrocketing costs for "props" in the game development industry to the plummeting costs for making feature films (digital cameras and post-production technology have let people bring independent movies into theatres for less than $35,000). For Shenmue, computer-aided modeling was found insufficient, so life-sized head mockups were created, scanned at 50,000 polygons per face, and then reduced to much less. As Suzuki pointed out with a smile, he was "not making games for PSX2."

Motion capture is a prime example of how the limitations of movie production affect both the budget and artistic expression of games. Real-world props and actors are needed by games that rely on motion capture. Worse, the actors have to be taught and trained first (e.g., fighting games require accomplished martial artists for motion captured scenes). Like movies, games now have to create reality first. These "cinematic" games have given birth to a new sampling industry, as well as to games defined by a new kind of derivative design -- one that clones the real world.

A Sega promotional movie shown during Suzuki's presentation posed the question, Is this really a game? As my head filled with the lectures and panels of the previous days, I couldn't help asking myself the same question. This game has 350 characters and nearly as many voice actors (in Suzuki's words "too many, too expensive"). This game goes through painstaking efforts to fill the gaps of mundane tasks like opening and closing doors, tasks that movie economy "cuts" out of the experience. What can it bring us beyond the death of the garage developer?

To me, all the meticulous effort put into Shenmue seemed more appropriate as edutainment than entertainment. If we wanted to experience other lives in similar detail, it would likely have to come as a documentary, not a game or movie. Somewhere between the rigid harness of narrative and the pointless complexity of cellular automata, games have to find a way to create meaning and relevance outside and beyond the actual interactive experience.

Suzuki has fulfilled for himself a dream, one shared by many (if not the majority of) game designers. He strives to create a new genre by merging the imagery of movies with the interactivity of games. Only a third of the Shenmue team were Sega employees, the others were recruited from external industries, all of which presumably are at home in the movie business. Time will tell whether the "movie-as-game" is our future - for now, the console industry prepares the ground for the return of Siliwood.

Bernd Kreimeier is a physicist, writer, and coder, working as senior programmer and project lead at Loki Entertainment. Previous work for Gamasutra and Game Developer magazine includes "Killing Games: A Look At German Videogame Legislation" as well as "Rising from the Ranks: Rating for Multiplayer Games". See Graphics at GDC for more of his coverage of GDC 2000. He can be reached at bk@lokigames.com.