Hideo Kojima gave an inspiring and entertaining GDC keynote describing the 20-year evolution of the Metal Gear series, from the very first Metal Gear game on the MSX2 platform up to 2008’s Metal Gear Solid 4: Guns of the Patriots.

In explaining how to make the "impossible" possible, he outlined the obstacles and times in his life when he thought he might be up against a “wall of impossibility.” To cross that wall of impossibility in game development, he changed his perspective or came up with new ways to attack the problem, whether it be hardware-imposed constraints or a self-imposed goal of excellence.

"Ninety percent of what is considered impossible is in fact possible," he claims. "The other 10 percent becomes possible with the passage of time and technology."

The original Metal Gear from the 1980s was born from hardware constraints. Working with the cartridge-based MSX2, Kojima dealt with the problem that only a very limited number of sprites could appear on the screen at once. This prevented the designer from creating an all-out Rambo-type experience, with bullets and enemies flying about.

He had to change his mission in design to get around, or over, the wall.

Kojima decided to approach the obstacle from different angles; he could not just throw up his hands and call the task "impossible" because of the limitations of the MSX2. The first idea he had was a combat game in which players don't fight, rather they just get held up when caught. That may not be much fun. Another idea was for a game that would be about escape, with a player running away from enemies. Kojima's train of thought eventually came to a combat game about hiding and sneaking, but he worried it wouldn't sell--such a game wouldn't be as "heroic" as the typical Rambo movie. But he was getting warmer.

Then he came to the idea that served as the basis for Metal Gear: Infiltration, hide and seek. The concept melded with the technical limitations of the MSX2, and the gameplay was fun. Constraints had brought about innovation: the first stealth game.

Kojima joked, saying that if he had given up, there would not only have been no Metal Gear, but also no Splinter Cell. Imagine that.

His main point: Don't fall into a rut where your preconceived notions automatically cause you to decide that something is impossible. Use those seemingly unconquerable challenges as tools or motivation to accomplish the impossible.

Kojima said that sometimes design can navigate a game creator around an obstacle, sometimes advancements in software or hardware technology can. Kojima is an adept designer, but he has never been one to shy away from technology. He and his team have pushed videogame hardware with virtually every iteration of the Metal Gear Solid series. The designer actually wanted to create a 3D Metal Gear on MSX2--that was, in fact, part of that "impossible 10 percent" at the time. The PlayStation eventually afforded him the means to create such a game.

In Kojima's case, not all obstacles are imposed by hardware constraints. Sometimes he imposes challenges upon himself and his team. For Metal Gear Solid 3: Snake Eater, he wanted to create a game in a natural setting with more organic surroundings--an environment that would change the sneaking dynamics of previous Metal Gear games. He could have made the game in an industrial environment like Metal Gear Solid 1 and 2--after all, that worked--but he chose not to. It's all part of staying out of the rut--all people often find themselves stuck in a certain way of thinking. He encouraged GDC attendees to avoid that.

"Let's join together, everybody here today, and make the impossible possible, because I think everyone has that responsibility in the game industry."

Sande Chen contributed to this report.