Warren Spector’s Saturday morning Keynote for the SIGGRAPH Sandbox Symposium focused on a topic near and dear to my heart: emergence. Spector has been a master of emergent gameplay with titles like Thief and Deus Ex, the latter of which took some bold steps in terms of trying to leverage “emergence as a design material,” as I term it. Spector was, as always, delightfully curmudgeonly, exploring the topic with equal measures of humor and cynicism, but also providing very specific examples.
He took the audience through a narrative of early design and gameplay experiences that influenced his particular brand of theology on emergent gameplay. Ever since we invited him to speak at Entertainment in the Interactive Age in 2001, I’ve considered Warren was one of the more sophisticated thinkers about games and narrative. He really understands that in order for these two apparently oxymoronic forms to work together, they need to be deeply intertwined. As Tracy Fullerton put it at Living Game Worlds (now my new mantra): “The mechanic is the message.” Spector understands that the story is the mechanic and vice versa. He also appreciates that playing around a story and playing with one are to entirely different things, and he is among a very small group of designers who understands how to create playable stories in a deep and meaningful sense.
In his keynote, Spector asserted that the only way for videogames to transcend being marginalized is to leverage their unique properties. He invoked Henry Jenkins, but this has also been Janet Murray’s message for over a decade. Some of those qualities, as Spector sees it, include:
• The power to transport players to other places where they have active engagement and agency, where they have a vested interest in the outcome
• Immersiveness in increasingly believable, not necessarily realistic worlds
• Unique, player-driven experiences, where each player’s story is different
• Systemic design, thinking in terms of rule-based and procedural techniques rather than linear or branching stories/experiences
• “Shared authorship,” in which players become co-creators through gameplay
Spector said games are most fun when players are given the opportunity for varied strategies, where they can work towards, or even create personally meaningful goals, formulate and execute plans, and use information and resources in the game in unique ways. In Deus Ex, for instance, Spector said, he didn’t care if players fought their way, talked their way, or snuck their way past obstacles. In fact the ability to succeed (or fail) through a personalized strategy was the entire point.
By way of example, Spector described an early insight into emergent gameplay when working on Ultima VI with Richard Garriott, aka, Lord British. In most of the game’s puzzles, he said, there were multiple solutions. But one puzzle had only one solution but a player figured out a DIFFERENT solution that the designers had not considered.
Another example from his own gameplay experience was Doom. In Doom, he pointed out, if an enemy takes damage from another enemy, he will attack that enemy. Players could therefore pit enemies against one another by ducking out of the way and getting enemies caught in their own crossfire. Once they started fighting, they would kill each other and reduce the threat to the player. One really interesting aspect of this was the insight that emergence often arises out of a deep understanding of rules and behaviors, which players can then exploit to their own ends.
With Deus Ex, he said, he actively tried to leverage this in any way he could, and relished when players devised their own emergent strategies in the game. He gave two examples, but I will just repeat the second, which I thought was fascinating. Players could place mines on any surface as an obstacle to enemies. One property of these mines was that they would not detonate if the player stepped on them. Because they had dimensional properties, players figured out that you could use them tactically. They would place a mine on the wall, use it as a step, place the next mine, step onto it and remove the mine below. This basically allowed them to create an infinite ladder at no cost in resources (the mines could still be placed later to thwart enemies). Players could thus ascend to any part of the world, including those ares that did not exist! As a result, players often got stuck in some sort of back-of-house part of the world they could not get out of. This reminded me of some of the stories from Uru, when players would exploit collision flaws for hide-and-seek to hide inside a wall, ceiling or inside the trunk of a tree.
Spector’s definition of emergence:
• “Apparently complex structures or behaviors emerge from systems characterized by simple rules.” (Murry Gell-Mann, The Quark and the Jaguar)
• “Much coming from little” (Emergence, John Holland)
• “The program is capable of surprising is programmer” (Holland)
• “Perpetual novelty” (Holland)
How do we achieve emergence in games?
• Rules, objects, input, feedback
• Rules operate consistently on a variety of diverse and versatile objects…
• Rules can be manipulated by players (although one of his colleagues posits that this is not always necessary)
• If a player doesn’t do something, if there’s not state change, there is no gameplay; my interpretation: the game cannot play itself
Spector described the “old school,” non-emergent way of designing games as scripting the player’s very move. This results in the player essentially “playing the designer,” trying to figure out what the designer wants her to do at every moment. This brought to mind my earlier Virtual Cultures post about WoW drawing from Bernie De Koven: “Am I playing the game, or is the game playing me?”
What happens when emergence goes wrong?
• Arbitrary forces thwart players
• Things, characters and environment behave differently from one instance or case to the next
• The game seems inconsistent or incomplete
• Plans fail and players don’t know why, either
o Nothing happens
o Something illogical happens
The key to solving this is coming up with robust, universal rules that are consistent throughout the game world. In my own research, I’ve also found that bugs can often turn into features when players discover and figure out how to exploit them for fun.
Spector expressed a few pet peeves. One was that if he heard one more designer say he wanted to make his game cinematic, he would go on a killing rampage. “If every player of your game is having the same experience, go make a movie…please.” Sports games, he also added, just crib rules from existing games.
He also said something I found really interesting: “If we didn’t have to put pretty pictures on the screen, our jobs would be a lot easier.” The fact that game content has to be conveyed visually adds a level of complexity that sometimes takes the focus away from more substantive content issues.
Causes for optimism? The Internet has created new ways of reaching audiences, including those that have not traditionally played games. Smaller more experimental games are also finding an audience, which is opening a door for lower budget games.
Nonetheless, Spector recently allowed him and his company to be acquired by Disney, a decision he relishes, in part because he is a fan, and in part because after working as an indie, he discovered he likes to make, huge epic games that take big teams and a lot of money. No big surprise there, and no doubt he’ll have his wish.
The only criticism I really had of Spector’s talk was something sort of endemic of industry talks in general: Although he did talk about other designers and their games, he failed to mention others’ ideas and contributions to the thinking on this topic. Jesper Juul and Salen/Zimmerman for instance have both written extensively on this topic, and Janet Murray’s unique properties of digital media went totally unmentioned. This type of critique is always tricky: Was this an oversight, or is Spector simply unfamiliar with the writings of these authors? It’s hard to say, but it would be refreshing if industry folks got in the habit of citing others more. As dreary as a convention as this may seem to academics, it provides a means to engage with larger dialogues on these topics. I did appreciate his references to Gell-Mann and Holland, and appreciate that designers are reaching to other disciplines for influence. (When I do my Game Studies interviews, I like to ask game designers what books are most influencing their work.)
It’s ironic that Spector’s interest in emergence hits a wall at multiplayer games, my personal passion. When I asked him later if his new “top secret project” was likely to be multiplayer, he answered “God, I hope not!”
</p>
I asked Warren about that later in the day and he basically agreed with me that when there is more than one player involved, "emergent" behavior is not only expected, but easy to create. It sort of comes with the package. Multiplayer implies emergent behavior.
Posted by: Jose Zagal | August 09, 2007 at 04:34 PM
I agree completely! Where I'm going with this in my own work is thinking about crafting emergent social behavior, what Joseph Beuys called "social sculpting." To me there is something inherently contradictory about this, but that's what makes it so interesting: how can you plan for that which is unplannable? And yet we can look for patterns and hopefully learn from those patterns, methods of "second order design" that will enable us to enrich emergence for players in various contexts.
Posted by: gamegrrrl | August 25, 2007 at 11:17 AM