SIGGRAPH Keynote: 'Designing Games for Process Intensity'
The ACM SIGGRAPH Sandbox Symposium, a two-day video game event of keynotes, panels, and papers from Saturday, July 29 to Sunday, July 30, 2006, held in conjunction with SIGGRAPH 06 in Boston, Massachusetts, opened with a keynote questioning the current direction of game mechanics in mainstream game industry.
Greg Costikyan, CEO of Manifesto Games, a start-up devoted to creating a viable path to market for independently developed games, stirred up attention by referencing Chris Crawford’s theory of process intensity. Originally published in the first Journal of Computer Game Design in the 1980s, Crawford spoke to process intensity while reflecting on his game Balance of Power, a geopolitical simulator. Later in Chris Crawford on Game Design, Crawford refurbished the explanation as follows:
Process intensity is the degree to which a program emphasizes processes instead of data. All programs use a mix of process and data. Process is reflected in algorithms, equations, and branches. Data is reflected in data tables, images, sounds, and text. A process-intensive program spends most of its time crunching numbers; a data-intensive program spends most of its time moving bytes around.
“Crawford’s definition is a highly useful idea, partly because of the limitations of computers and data,” said Costikyan. He argued that, since computers are much better at crunching numbers quickly, games should be made for process intensity. “If you're not using [computers] for that purpose, you're just using a glorified DVD player,” he added.
“Game processing is 10% of what’s going on.” Costikyan equated the majority of current games with “poly pushers.” In a sprite based game, Costikyan explained, every time you want a sprite to have a different behavior, you have to enable that behavior with unique animations. In a 3D game, however, algorithms are in place for generating content dynamically, such as lighting. To break down the explanation of process intensity versus data intensity further, Costikyan drew from a comparative list: Simulations are process intensive and cut scenes are data intensive, while strategy games are process intensive, and adventure games are data intensive.
Costikyan called for more emphasis on processing, particularly in the area of interactive audio. While interactive audio is process intensive, canned audio is data intensive. He feels that interactive audio would create a more immersive experience for the player.
“Processing data is the very essence of what a computer does,” Costikyan quoted from Crawford. “Part of the reason we have plunged greater and greater into data intensity is that we wanted graphics that didn't suck.” However, notably, graphics are still what marketing widely focuses on in new game titles. Costikyan recalled the release of PlayStation and its 3D with dozens of polygons. He questioned whether or not what developers can do on the Xbox 360 is really more impressive than what is available on PlayStation 2.
“We're caught in a rut.” Costikyan asserted that although the development tools are getting better, they aren't getting better fast enough. He tied the rise of development costs to the desperation of publishers relying on sure-sell game titles. Currently, Costikyan estimated, a minimum buy-in for development of a console title can be anywhere from fifteen to twenty million dollars. Game genres with smaller but dedicated fan bases simply can’t be funded. “This market trend is not supportable,” he concluded.
“If we really want games that pull out interactivity, we need games that are more process intensive,” suggested Costikyan. He followed up by emphasizing that developers can achieve “infinite hours of game play through process intensity.” He then identified games which have been successful at integrating process intensity by his definition:
- Bounce Power is a game in which you play one of the world's great powers while trying to manage the world's geopolitical state. Process intensity comes into play with the huge amount of data for each country and variables of game play.
- Sim City is a complex state machine and a highly process intensive blank canvas for the players.
- Civilization is a series of randomly generated worlds with complicated economic models and multiple wind conditions. The games are infinitely playable and re-playable. Recent versions can be modded.
- NetHack is a single player dungeon exploration game with randomly generated dungeons intermixed with occasional design, as well as customizable characters.
- Gen MUD is an open source MUD game. Rather than having designed sections of the world, it's mostly randomly generated. Individual villages and individual non-player characters have different randomly generated objectives that react and interact on their own to create a dynamically changing environment.
- EVE Online is a Massively Multiplayer Online game (MMO) that is highly process intensive. Nothing in it is really designed, given that even the star system is algorithmic. Quests are player based and algorithmic.
- Spore is referred to as procedural content by Will Wright, but serves the same concept as process intensity. Players design creatures, which are algorithmically modified and generated, then developed into algorithmic civilizations. Civilizations can be shared across players' games.
Costikyan further elaborated on the possibilities of combining process intensity with player interaction. He believes World of WarCraft is not the future, because the environment is static and pre-designed. However, he expressed, Spore and EVE Online can offer glimpses into the possibilities of player immersion and expansion of content using process intensity.
“In essence, the experience of playing EVE Online is the experience of interacting with user generated content.” EVE Online is one universe, except for a recent shard split to support their Chinese players with better ping time. Necessary shard splits to handle data in MMOs of the past, such as EverQuest, will fade if process intensity is incorporated. When EverQuest first started, Costikyan explained, it was about the size of Rhode Island. “Imagine having to hire people to place every boulder and tree in Rhode Island.”
Costikyan encouraged incorporating research by computer science academics rather than those focusing on games exclusively. “If you want to take AI seriously, think, what kind of games can be created around them?” Academic research into areas such as Artificial Life can offer interesting trade-offs, he feels.
“All games are fundamentally algorithmic. Now that the visuals are caught up, we need to focus on what computers are good at.” Costikyan referenced the basis of games from the Neolithic to the modern. “It seems to me that we've barely touched the surface of what's possible with games.”
“We need to start thinking about process and true interactivity,” Costikyan concluded. “Go create!”