The Many Ways to Make a Game

Slide Show >>

Sounds like the setup for a geeky joke: What do you get when four computer-programming grad students ponder gravity for seven days straight?The answer: four bouncy video games. In reality, though, the foursome produced more than 50 highly original games in the course of one academic semester -- roughly 10 weeks.

It's the true story of the Experimental Gameplay Project, the brainchild of four 2005 graduates of Carnegie Mellon University's Entertainment Technology Center master's program. Three are now employed by leading game developer Electronic Arts (ERTS

), and the fourth is developing a game at CMU to teach kids about Internet security. The lessons they learned in game-design innovation are worth passing on.

READY, SET, DESIGN. "The idea of making small games is the best way to learn game design," says Eric Zimmerman, co-author of the textbook Rules of Play: Game Design Fundamentals and the founder and chief executive of independent gaming company gameLab. "Art students won't tackle a big canvas from the start, but instead first make small drawings in a sketchbook. The idea of the small sketch is especially germane to games because it's important to explore a range of approaches quickly and learn from your own mistakes."

"In creating games, it's impossible to predict interactivity," Zimmerman adds. So, the CMU grad students' approach of intensive learning-by-doing forced them to quickly devise and test, without wasting time preplanning what would be engaging interactive games.

The endeavor was first launched in January, 2005, based on an initial idea by students Kyle Gabler, now a programmer and game designer at Electronic Arts, and Shalin Shodhan, now a graphics programmer at EA. The two collaborated with classmates Kyle Gray, a level designer at EA who is working on the Superman game due out in June, and Matt Kucic, a programmer and designer at CMU's Information Networking Institute). The foursome set themselves a daunting task: to each make a new game in less than seven days based around a common, challenging theme like vegetation.

The concept is similar to the Indie Game Jam -- a yearly gaming industry event in which programmers and designers try to come up with as many original games as possible within four days -- though Gabler didn't know of it when they conceived of the project.

MIME PLAY. One Experimental Gameplay Project game, Tower of Goo, which requires users to construct a skyscraper-like structure made of oozy, gravity-sensitive glop, has been downloaded more than 100,000 times. Such quirky games don't fit the mold of popular computer titles, like the big-budget, franchise-friendly games Grand Theft Auto: San Andreas or Halo 2, 2004's overall top sellers, according to the NPD Group. Still, after the quartet presented their work at the Experimental Gameplay Workshop at the 2005 Game Developers Conference in San Francisco, the CMU students were "approached by a lot of publishers," Gabler says.

But don't look for any of the titles created at CMU -- such as On a Rainy Day, which requires users to play the role of a tree made of hands and grab umbrellas falling from the sky, or Mime After Mime, in which players chase cartoonish, Marcel Marceau-like characters only using sound (the mimes just flash briefly on-screen) -- to appear commercially any time soon. Despite their current industry jobs, Gabler and his cohorts don't plan to sell their experiments. "This project isn't about the final games," says Gabler. "It's about the creative process."

To emphasize that point, the team recently co-wrote a white paper that includes a "handy cut-out list" for others who want to adopt their recipe for rapidly prototyping new game ideas. The list, which includes practical points and Yoda-like aphorisms, such as "rapid is a state of mind," is a useful guide for students and commercial developers looking to nurture innovative and efficient game design.

JUICY TIPS. According to the CMU grads, it's important that the player develop a sense of emotional involvement with the game, and that requires the right art and music to inspire a game's mood. They caution against wasting time with complicated effects, like subtle shadows and lighting, if a similar aesthetic result can be achieved with simpler programming.

Another prescription is to give players a sense of creation or customization by allowing them to build, draw, or make something as part of the game. It's also important to have a clear game-play goal, even one as simple as gathering objects, and to make a game "juicy" -- that is, infuse it with maximum action and reaction (like lots of splatters) resulting from minimal player input.

And, finally, they advise designers to cut their losses if a game's design isn't working. It's more efficient to drop the project than to mask its faults with tacked-on visual or other flourishes.

"CREEPILY HAPPY." Gabler says he applies those lessons to his work at EA. He cites a project in which the movements of the main character, a girl, prompts everyone around her to sing. "I'm working on the aesthetic of a creepily happy place by creating a video first, which I can go back to when designing the actual game," says Gabler.

While such intense focus on preparation might distract a designer from working on a final product, Gabler believes that, if done quickly, creative exercises can save time in the long run. "Large companies and big studios pour lots of time and money into developing huge games, only to later discover that they're not fun," he says. "A little bit of time spent early on in design saves time later."

Although Gabler, Shodhan, and crew have all graduated, the Experimental Gameplay Project lives on, headed by a fresh team of CMU students. By midterm, they had completed 27 games. At that rate, they should make their goal of 52 games in one semester. Like their predecessors, the current students are "pushing to make games that will make people think, 'You can make a game about that?'" says Shane Liesegang, a member of the new team.

ANYONE CAN JOIN IN. Some of the most recent offerings include games based on themes such as "temperature" and "chains," interpreted with both abstract and literal aesthetic elements that defy the best-selling models of shooter or sports games. Gabler still manages the site, which now encourages anyone, even non-CMU students, to submit their experimental games.

GameLab's Zimmerman believes that models like the Experimental Gameplay Project could be applied to commercial game design. "When you look at other design industries, like architecture and industrial design, research and experimentation are encouraged," he observes. "But in the gaming industry, there's a sense that we're in a creative crisis."

READER COMMENTS