[We've been in Shanghai at GDC China for much of the past few days, and over on Gamasutra, we've got a host of write-ups from the show. We'll feature a few GSW-relevant ones here, starting with Christian Nutt writing up an Indie Games Summit talk from Steve Swink.]

At GDC China, Shadow Physics developer and Independent Games Summit co-organizer Steve Swink explained the experimental game ethos to a Chinese audience, outlining a roadmap for thoughtful innovation.

Swink opened his presentation with demos of two games: Inputting and Scale, and then explained to the audience why one would fit the mold of "experimental games".

While Inputting tries and abandons many control schemes, Scale sticks to one mechanic and aims for as many "interesting uses of that mechanic as possible. That's sort of the fundamental precept of the design methodology," Swink said.

"Portal is the exemplar of this," he said, as is Braid. More examples include World of Goo, Fez, Closure, and Miegakure.

"What all these games have in common is that they apply a certain set of constraints."

1. They're new. "['New' has] kind of a fuzzy definition... People tend to know it if they see it, but people [will] argue if it's been done before... You can look at [Xbox 1 time-rewinding title] Blinx relative to Braid, but if you apply an eye to a certain mechanic in a way, it becomes new."

2. You take as long as is required to make the game. "There is no finite production schedule on a game because what you're trying to do is explore the mechanic as much as possible...You take as much time as you need to explore the mechanic as fully as possible."

3. Exercise craft and discipline. Swink's example here was non-games: Wagner wrote the Ring of the Nibelung, a 15-hour opera cycle; Brahms wrote "way better music that took an hour," said Swink. 


4. No filler content. "You have to have the discipline as the game designer to let the bad levels and parts of the design fall off the table."


5. Not recycling content. "You're not going to play the same level over and over again. Once a concept has been introduced, it's gone again immediately. You use it once and then you've sued that concept. You don't make five levels based on that concept.., You wrap those levels in one bigger concept."

6. You don't waste the player's time. "This is the fundamental philosophical underpinning" of experimental games, Swink said, with some passion. "This all stems from the finite number of irreplaceable hours in our lives. I want to spend those hours doing something meaningful for me, and i don't want to waste people's time."

In fact, said Swink, many games are instead engineered for length and intentionally wasting the player's time. "I think that's unethical and it's not a very good game." In Swink's view, people should aim for "activities which enrich your life and are meaningful experiences."

He admitted that indie darling Spelunky may not quite fit that description -- "it breaks down, but not for me. I have played an hour of Spelunky every day for a year and I have derived an extreme amount of pleasure from it."

When he beat a Spelunky ironman challenge, "it was the most amazing gaming achievement I've ever had." The crucial difference from other games? "It was a legitimate challenge; it was not a fake challenge that was engineered just to make me feel challenged."

"We want to create genuine challenges," Swink continued. Quoting Braid developer Jon Blow, he said that "challenge must be real." In Swink's opinion, "it's very insidious as a game designer to say 'I want the player to feel clever.' We want the player to be clever."

Swink said that creating games that are intrinsically meaningful, which inspire people to think and not just distract them, we will approach true meaning -- with the premise that Einstein's work on Relativity was the most important time ever spent by a human in history, and then extending that to say that the four-dimensional play of Miegakure might inspire a child to become a physicist.

Delving into Shadow Physics

"Perhaps by playing Shadow Physics, I can teach people the connection between an object, light, and shadow," Swink said, before diving into an extensive demo of his game.

The game is a 2D platformer that's influenced by 3D objects -- light sources cause 3D objects to throw shadows, which is what the player navigates. The character, being a shadow, can warp and distort on uneven surfaces. Shadows of objects inherit their properties -- balls are bouncy.

The player can manipulate the objects by pushing and pulling their shadows, and depending on the angle of light sources, this can change the environment drastically. Later, the game involves multiple light sources and other tricks to involve the player in more complex puzzle-solving.

The game fits the very definition of "experimental games" as Swink set out in the beginning, as it thoroughly explores the natural extensions of the basic shadow concept, increasing in complexity as the game progresses. "You introduce concepts, and then they start interleaving, and then you can do more puzzles with them," said Swink.

But there's another thing you need to really hook players -- "to make a really excellent game like a Portal or Braid is to integrate some form of narrative or story," he said. Without the battle against GlaDOS, he said, Portal would be "very tedious." In fact, Swink has assigned the creation of Portal levels as coursework, and finds playing the results dull.

It's not so much about storytelling, however; in Swink's view, "you have to develop some sort of context" that delivers meaning. In Portal, "much of what makes that game so compelling and the fact that you feel like you're pushing against an antagonist."

Another Essential: System Design

This is "another difficult thing to come to grips with," he said." He described Mario 64 (which, amusingly, the Chinese audience was totally unfamiliar with.) The rules which govern the stars and coins in the game are totally arbitrary (such as collecting 8 red coins for a star, of which there are 6 in each level.) However, it all hangs together with an internally consistent logic that gives the game structure and goals.

"That has to be applied as well. There has to be a particular structure applied to what you're making and you have to think really hard about it," said Swink. "That sort of context has to occur in order to have a game that is really in harmony with itself."

"If you want to make a really good game, you have to take care off the really invisible things that players do not think about and which are thankless," said Swink. This "is why so many experimental games start from an existing genre and play style that players are familiar with." You have to make the game feel good at its core so players can "receive" your game's inventiveness.

And tools are key to this -- not just because they allow for efficient workflow, but because "tools define experimentation," said Swink. Adding a flexible lightsource movement tool to Shadow Physics changed everything for Swink.

The way tools "affect the way you perceive the design of a game is really profound, and i'm still trying to come to terms with it. I think you can make those tools more usable and fun to play with -- i think that might be the most important part of game development."

The Look of It

Swink thinks that game aesthetics in mainstream games generally fall into boring, broad categories -- problematic because "not very many people think very hard about not just the style of their game and whether it's harmonious with their game and story... but how the visual style can impact the way the game is played and can interact with it."

Swink faces an uphill battle; it's difficult to get players to focus on both the shadow plane and the 3D objects in front of it. Players tend to want to focus on only one or the other. This is a challenge where the game's art style will define the very success of its mechanics.

The Substance Problem

And while Braid and Portal were challenging, they were very forgiving in letting the player retry their puzzles indefinitely they were solved, which "adds a very lovely softness. This is really interesting and really fun."

Since with puzzles, "once you've solved them they are not interesting at all. You no longer want to engage with them, because it has one right answer," the game can't be too frustrating.

Many games have a "all sizzle, no steak" problem, Swink worries. You need to discover "good meaty mechanics you can hang a whole game on... versus one that seems really appealing on the surface and as soon as you delve into it as a designer" you can't go further without falling back on contrived and cliched game design tactics.

You're in trouble, says Swink, when "you very quickly start to say 'now what?' when you're looking at that mechanic."