GDC: Inside The Experimental Gameplay Sessions 2009
“This is the eighth year in a row we’ve been doing this,” opened moderator Jonathan Blow (Braid
), introducing the Experimental Gameplay Sessions at Game Developers Conference 2009, before claiming that this was the “most interesting year yet.”
The reason? A “drastic, discontinuous change” had occurred in the games industry in recent years -- an explosion of creativity that had led to “the most consistent collection of designs that push the boundaries in the most interesting and thoughtful ways.” After briefly introducing each, he allowed the designers to speak for themselves.
The Unfinished Swan
- Ian Dallas (Giant Sparrow)
Though Dallas has been working on The Unfinished Swan
-- a first person title that made an impact late last year with an online video showing its central gameplay mechanic of using black paint to navigate around an entirely white world -- he stated that it was “nowhere near done,” but that he at least now knows where he is going with it, using his time on stage to discuss his processes.
According to Dallas, game development is about offering the player experience that they have never had before, but while The Unfinished Swan
did do that, after only a couple of minutes, the gimmick became flat. It was “too pure,” he contended, offering little more than the exploration of a space as a driver.
His challenge, he found, was in figuring out what kind of space existed, what kind of framework (or context) the game mechanic existed in, and what he could offer over and above throwing paint.
Structurally, Dallas knew he didn’t want to “create a puzzle game, an arbitrary series of challenges that tweak the mechanic,” but instead something “continually wonderful across 60-second increments.”
After wasting six months on dead ends, he found himself reading a childrens' book and, captivated by the “gentle relationship” between the reader and these books, he was inspired to create a title in which the player is a young boy chasing after a swan, inspired directly by the story of Alice in Wonderland.
Dallas admitted, though, that in fleshing out this story, he had to “cave in” on avoiding puzzle-based gameplay, showing us one example of an “Escher-like” space that formed an early puzzle in the game.
- Steve Swink, Scott Anderson (Flashbang Studios)
In an amusing footnote to the previous session, Steve Swink showed a prototype that he had created that was astonishingly similar to The Unfinished Swan
except with more of a spray-can aesthetic, revealing that after finishing the prototype, they discovered that The Unfinished Swan
was in development independently. Swink moved on to display their project, Shadow Physics
, which had a similarly unique take on the concept of in-game space.
The player takes the role of a shadow, therefore, a 2D character, cast upon a wall that exists in a 3D world. Other shadow objects, cast on the wall by 3D objects, could be moved, and in turn move the real world objects.
This had “really interesting ramifications,” with objects able to be hidden in the 2D space by hiding them behind objects in the 3D space; light sources moving around to warp and modify the shape of shadows, and the player character able to be made larger, smaller, or even create multiple player characters due to multiple light sources.
“Our game is taking a technology that has been developed to be subordinate to the characters, just casting their shadow,” said Swink, arguing that the effort required from technology seemed such a waste. “We’re reappropriating this technology and using it as part of our game design. It’s sort of ‘punk rock’.”
- Marc ten Bosh
A puzzle platformer in four spacial dimensions -- where Marc ten Bosh took special care to note that time doesn’t count as the fourth -- Miegakure
had to first be demonstrated to the audience in only three dimensions for the high-level concepts at work to start to make sense.
Bosh displayed a character living in a 2D plane within a 3D world, similar to Flashbang’s Shadow Physics
, except in this case, the player could switch the plane across the X, Y and Z planes, allowing him to move across or around objects that he couldn’t move around (through 2D movement) on another plane.
So, in his 4D demonstration, the 4D space featured a character living in a 3D plane (able to move around and jump over things) that he could swap his dimension across. Initially mind-boggling, the audience began to grasp the concept more clearly as it was explained that the 4th dimension was simply another way to move between the other three dimensions, even when he added more complexity by allowing objects in the fourth dimension to cast shadows in other dimensions!
“I could actually do five dimensions,” said Bosh, “I’m a programmer, so to me, I know locations are just a list of numbers.”
- Chris Hecker
Chris Hecker’s project was inspired by the simple fact that “spies are cool”, particularly early Bond-esque spies who have the ability to “hide in plain sight.”
Hecker’s design in turn was inspired by the Turing Test, and the difficulty in which a computer has in fooling the user it is human when asked to do natural language processing. But what if, Hecker considered, you only allowed the player a “simpler, more responsive” form of interaction with the computer? Could he be fooled?
The perfect setting, Hecker decided, would be a cocktail party, where a variety of people interact in an already very stylized way and the social rules are an already tight subset. So Hecker’s prototype asks one player to take the role of a spy attempting to complete missions at a party filled with other AI characters, while an opposing player observes the party (as a “sniper”) and must attempt to observe the player enough that they can work out who it is and kill them before they complete their missions.
In Hecker’s current, “very early” prototype, the gameplay was formed around “tells” -- observable, if subtle, occurrences such as a brief look to the left when stealing a book rather than returning it to a shelf -- but he hoped that, in time it would evolve to become a game about observing subtle differences in behaviour.
- Daniel Benmergui
Daniel Benmergui opened by showing his trilogy of work I Wish I Were the Moon
, Trials and Storyteller
I Wish I Were the Moon
used a “snapshot” mechanic to move characters so they would interact with each other in a scene, as did Trials
(with more of a puzzle-orientated design), whereas Storyteller
relied more on moving characters around to modify the story across three scenes simultaneously.
in particular featured some interesting bugs or, more accurately, accidental features -- the ability to make a tombstone fall in love, or another, described by a comment on Kongregate Benmergui showed, as “HA HA GAY GUYS.”
This idea of changing scenes to modify a timeline intrigued Benmergui, so he began to prototype, creating a game based on Super Maro Bros.
where you would choose where Mario would jump in the timeline and that would affect his future possibilities, a game in which you would “play together with fate,” (fate wishing for you to “win”).
If you tried to jump into a block to die, fate would move it. If there was a block fate could not move, it would attempt to move other blocks to stop you hitting it. If fate couldn’t stop you committing suicide, it would create a parallel state where fate would “give up” on keeping you alive.
However, this remained a prototype concept (despite its potential) and so Benmergui closed by demonstrating his next game, Today I Die
. The title allowed the player to change the words of a poem -- “dead world, full of shades, today I die” to change the world of the game, and in turn change the world to find new words.
- Jenova Chen, Nick Clark (That Game Company)
“If you played Flow
, you may think of Flower
as ‘adding 3D movement, but removing AI’ so wonder ‘what the heck have they been doing for the last two years?’” said Jenova Chen, explaining that this session would cover exactly why.
“We start from the experience that we want to deliver, then work out what we want the gameplay to be,” explained Chen. “Not very efficient, but it’s the way we do things. So, we wanted to create a game that offered a safe, free space ‘filled with love’. So we wondered how we could create gameplay that would generate those feelings?”
Chen revealed that they used Microsoft’s XNA for their rapid prototyping stage -- “at the Sony office, too,” he quipped -- as they quickly worked their way through different designs, including a procedurally generated flower growing sim before they set upon the idea of flying through a field of grass.
“But who are you?” Chen considered. “Our first attempt was to create the player as the consciousness of the space, but that was very ‘out-there’.”
After discarding that, they tried the role of a seed travelling on the wind, before deciding it felt “too much like a golf game,” turning the terrain into a goal rather than a backdrop and frustrating the player when they didn’t hit key areas.
Next they tried a swarm of petals, collecting more petals by flying into flowers. “We liked this prototype because...” said Nick Clark, being cut off by Chen joking that it was “just like Flow
“We presented this to Sony, saying ‘this is just like Flow
, but cooler,” continued Chen, but he noted that they quibbled over the lack of depth and challenge.
Further prototyping led them to a survival challenge with spells, a time limit, the ability to boost and deadly areas that led to the loss of petals, and another prototype that required players to deposit petals into orbs to unlock checkpoints.
But “sometimes cool mechanics go against your goals,” said Chen. In their attempts to make something classically fun (hard and challenging), they found it went against their original plans to create something peaceful. “Graphics and music evoke different feelings, so gameplay should be able to do the same thing,” concluded Chen. “Game developers need to think of other kinds of experiences other than just ones which are ‘the most fun.’”
- Chris Hazard, Mike Resnick (Hazardous Software
“Imagine an RTS where you can send your units back in time to destroy your opponent's units before he’s even built them,” opened Chris Hazard, thoroughly confusing everyone in the audience, a confusion that never quite seemed to lift.
A complex RTS which features a “timeline” across which players can leap across, the game worked as a “race to the past” by players who understood it. Hazard found it easiest to describe the gameplay through tales of previous plays, such as a battle over a mining base that he won only to find his opponent going forward in time, researching nuclear technology, sending that technology back in time to before the mining base battle and nuking the area—only for Hazard to go back in time, avoid sending his men to the location, and watch his enemy nuke his own troops!
The game features multiple aspects of time travel -- such as paradoxes and the requirement to send units either backwards or forwards in time to maintain causality; with characters even able to fight along side multiple future or past versions of itself -- under the knowledge that any damage their past versions receive they also do.
- Tyler Glail
Borne out of Tyler Glail’s frustrations with “dark levels” -- levels where you can’t see anything but are asked to navigate through them -- Closure
is a 2D “dark level” platformer where anything in the darkness does not exist. The player carries a light and can use other lights to illuminate the playfield, but a lack of light is important to players, as it allows them to jump over or through walls or obstacles that, illuminated, would cause a problem.
“I wanted to make a game that defies expectation,” explained Glail. “I wanted to play on the player’s prior knowledge of dark levels. Our brains fill in the blanks about where the walls should exist, but in this game they don’t.”
Where is My Heart
- Bernhard Schulenburg
A difficult to explain single-level prototype, Bernhard Schulenburg’s Where is My Heart
was a fairly classic character-switching 2D platformer, featuring three characters attempting to make their way to a “heart tree.”
“The world is not displayed coherently,” Shulenburg informed the audience. “That’s on purpose. It’s a comic panel effect -- you could say a representation of the idea these characters are lost in the world.”
Rom Check Fail
-esque retro game-themed Rom Check Fail
has been described as a “desconstructivist concept mashup,” according to Farbs, “but in my head, it’s something much nerdier -- it’s an experiment in gameplay variation.”
“Gameplay” is generally considered the interactions you ask the player to perform and what they perform them with, so for example creating a character that can shoot and then giving them a variety of monsters to shoot. If you create one player character and 14 monsters to shoot, you have 14 gameplay variations, but if you use all 14 monsters as playable characters too, you end up with 49 gameplay variants -- “over four times as much game,” laughed Farbs.
Of course, this amount of variation had an effect on the variance -- Rom Check Fail
has a “high level” of variance as each round both the player character and the enemy could be different.
“When you change what the player is doing, they have to re-learn what they’re doing, and when you prepare to chance what the player is doing, there’s also an anticipation. Both of these things have an important role in pacing,” concluded Farbs. “When you’re thinking about pacing, don’t just think about changing the speed and don’t ever just make your game harder. Think about what the player is doing, and how you are changing that.”
Derek Yu’s closing session was an exploration of roguelikes—procedurally generated dungeon exploration games. “The most exciting things about these games is that each time you play them they are different,” explained Yu, but he found it disappointing that so many roguelikes somehow offer similar experiences, relying on traditional fantasy settings and turn-based, top-down gameplay.
“The things I really like about roguelikes is their random level generation and that when you die you are dead and can’t continue. Even though I’m a huge fantasy nerd, D&D doesn’t really have very much to do with either of these things.”
When working on his own roguelike, Yu considered other genres, and thought about his own feeling that platform game design is entirely arbitrary—“why place that platform there?”—and that you would find yourself playing the same levels over and over again.
was his attempt to create a roguelike platformer. “My model for making this game is that death is fun! Players are not condition to think that way—after all, death is a bad thing for most living things—but the random generation means that you don’t have to play the same thing over and over again.”
Yu concluded, “there are lot of things you can do with roguelikes rather than make a typical dungeon crawl.”