GDC 2012: Tiger Style's rocky path to inventing new gameplay in Waking Mars
The studio's mantra is to stress innovative mechanics and meaningful artistic design into every title, but designing that content for this year's Waking Mars - currently Game of the Week on iTunes - required thoughtful iteration and careful curation of ideas, as Smith explained in his GDC 2012 talk.
Waking Mars is a platformer is set in a cave on the Red Planet.
You play as an astronaut trapped underground, who uses a jetpack to move around the environment. The player has to create ecosystems in order to pass through the game, throwing seeds into fertile group to grow plants that assist them in progressing to the next level.
Smith's has a background in triple-A game development, working on titles such as Thief, and while Spider did well, Tiger Style suffered "sophomore anxiety" (in Smith's words), following the success of the IGF winner.
Incidentally, Tiger Style doesn't want to make violent games, and cares about having artistic goals and being native to iOS hardware and the touchscreen. It stresses sophisticated interactions delivered to a casual audience.
The concept phase of Waking Mars took four months. Tiger Style prototyped and threw out two games in that period.
World Creator had the player drawing a world in two dimensions by tracing a line across the screen, and then planting trees to begin making a world that would grow on its own.
Octobot featured a multi-legged creature that players would navigate around the environment by moving its legs.
A third prototype game called Descent involved climbing down into a cave, which Smith called "Doodle Jump in reverse."
The player had a headlamp by which they could illuminate dark spaces, and the primary mechanics were crawling, hanging off ledges, jumping, sliding, avoiding boulders and collecting objects.
A rope mechanic offers players a limited number of ropes by which to slide down through large gaps, and assist in getting to the bottom of the cave without falling from a height and dying. The prototype of Descent was programmed in April 2010.
Yet Tiger Style decided the Tomb Raider games already existed, and that caves were boring.
This led it to consider imagining the caving game as set in outer space. The sci-fi thought experiment led to the concerns that the game would be less relatable and appeal more to a core gamer demographic. But the sci-fi angle did open the team to the idea of creating lifeforms, which led to the concepts of gardening and ecosystems which would prove so important to the finished game.
The first playable was called Mars Descent.
While it was still a caving game, Tiger Style sold it as "hard sci-fi, not soft sci-fi," in the hopes of keeping the game relatable. The National Geographic book Our Universe served as an inspiration to Smith, and he drew upon its material speculating what life forms might develop on other planets.
A jetpack power up was also introduced, and the headlight was used to repel creatures in the cave and encourage plants to grow.
The player began in a wrecked outpost on Mars, which served up some environmental storytelling, and then the player was funneled into an unexplored alien cave.
Room to breath
One of the primary mechanics of this first playable was an oxygen resource. Ice chunks found in the cave could replenish the oxygen timer. The jetpack had limited fuel which renewed itself when the player was on the ground, creating a Lunar Lander- style movement.
The player had to land carefully via short bursts of the jetpack if they descended from a height. Eventually the player encountered lifeforms, which would grow if they shined the headlamp on them.
Tiger Style introduced plants into their first playable such as oxygen plants, which served as permanent recharge stations for the oxygen resource and would also wake up creatures in the cave.
More sophisticated levels had players seeking light plants to obviate the need for a headlamp in a level, weed plants that would kill other plants, and water plants which spat out water seeds that were really difficult for the player to catch with the jetpack's limited fuel.
And if the player had to throw a water seed somewhere to nourish or grow a plant, and missed, the difficulty of getting another seed was problematic.
Exposing this playable to external feedback revealed issues.
The time pressure caused by the oxygen mechanic made players reluctant to explore. The need to ration jetpack fuel gave the players a constantly distracting thought to consider whenever they moved around the cave.
The headlamp was a great touchscreen-native device, but took an already tiny screen and filled it with black pixels.
Tiger Style didn't put an inventory into the game because it was worried it was a hardcore gaming tendency, but not having an inventory meant no storage, and the aforementioned problem with gathering water seeds.
At this point, Smith said it considered the question of Interaction Density, a measurement of meaningful choice divisible by number of player actions.
Lower interaction density was something like growing a single plant that required seven composite actions. High interaction density would be throwing one seed and growing an entire plant from that single action.
The second playable build forced Smith and his team to confront core gameplay. They began working on mid-level design ideas, for instance requiring a certain number of oxygen plants to open doors. They also introduced a mechanic where certain plants could only be grown in certain parts of a cave.
Another mid-level idea was to give planets relationships to one another. A light plant could shine down on an oxygen plant and help it to grow. This latter idea caused Smith to question whether the game required a player avatar at all, and whether it could be turned into a God game.
In the end, it settled on sticking with a personal adventure game, but this episode illustrated it was thinking broadly.
Smith also thought about a mechanic in which players had to reach acceptable environmental conditions such as oxygen, nitrogen and temperature levels by growing plants to adjust those readings, or requiring a certain biomass (read: number of plants) before moving on to the next level.
At this point Tiger Style "dropped anchor" on some of these ideas, which Smith called "A useful technique in the innovation space."
At a certain point, game designers need to simply build and test an idea in the playspace in order to inspire the next iteration.
The fuel mechanic was removed from the second playable. The headlamp and the search in the dark gameplay remained. Instead of only being able to grow certain plants in certain locations, fertile areas were introduced where any plant could be grown. This led to players essentially creating their own levels. Barriers called airlock plants remained in place until the player grew enough oxygen plants.
An inventory was introduced, as were more environmental hazards. Light plants also spat out dangerous projectiles. Stalactites that could fall from the ceiling were added, as were creatures like flying bats which could be repelled with the headlamp.
This mid-level focus didnt solve Tiger Style's problems, however, and so it had to shift to the low level design.
The team had to consider why combat works in games. It provides high stake drama (life or death), offers clarity and feedback on win/loss conditions, and included intermediate progress feedback in the form of hit points. Combat also made nuanced movement meaningful, for example in a fighting game where actions and specific timing led to success. This, in turn, supported the concepts of depth and mastery of a game.
Smith and company began trying to apply these principles to the second playable of Mars Descent.
The fuel limit for the jetpack was removed so that players could focus on building the skill of picking up seeds spat out by the water plants. If the player missed, having to fly over to get the seed was enough to reinforce the desire to master the skill.
Rules of engagement
Action gameplay tends to have lots of dynamism and unpredictability, but results are acceptable and clear.
The player makes a choice to take an action, exercises the proper skill and execution, and if they don't while the results may be unpredictable, the player understands how they happened, so the results are acceptable.
The headlamp mechanics of growing plants were a player choice, but skill didn't factor into it and the results were predictable. Rather than keeping that mechanic in the game, Tiger Style decided to lean more on physics challenges, like throwing seeds where they needed to go and letting them grow at once if the throw succeeded.
Throwing seeds was better because it was based on what Smith called "the simulation foundation."
A ton of complicated stuff goes on in simulation, but players understand the results of physics. They understand how fire propagates or how liquid flows, for example.
The player already knows these rules so the actions of throwing a seed, or knocking an object into a stalactites which then dropped and killed a plant, were arbitrated by physics which could lead to unpredictable but acceptable results.
Tiger Style continued working with the physics and added more meaning to collisions in the game.
Bat creatures could pick up seeds. Players were given the option to throw things at the bats to make them drop the seeds. Special plants called fisher tentacles would produce different effects if hit with different kinds of seeds. Player were given the ability to shoo crab creatures away from seeds before the crabs ate them. Acid drips were added that could injure the player.
Core to begin
These changes began to lead to more emergent gameplay.
A projectile might hit a bat in mid-air, which then dropped a seed, which a crab might come along and eat, which could be the player's fault if they allowed the crab to be there in the first place.
Emergent play created unpredictability, which gave players the motivation to master skills, for instance throwing seeds, so that they didn't accidentally lead a crab creature to a location that could alter the player's plans, or so that they didn't accidentally hit the bat which dropped the seed and set up the following chain of events.
The third playable that incorporated these changes was pretty much the version released.
Focusing on the lower level design turned out to be the right call. The red herring that Tiger Style almost fell for was that complexity lay at mid-level design. That, said Smith, was more appropriate for strategy games.
It scrapped the atmosphere balance idea but kept the biomass idea, which fed into the lower-level design of planting lots of plants, which in turn played into the design ideal of players creating their own level - but it was no longer a gardening game about tending to plants.
The darkness and the headlamp were removed, and Waking Mars was no longer a game about caving, either.
The game still had the ecosystem idea in place, which was made easier to play because Tiger Style had removed the Tomb Raider and Lunar Lander gameplay. Tiger Style learned that you cannot support three different games in a single design.
By removing the extraneous elements, Smith and company made it easier to focus on the ecosystem mechanic that turned Waking Mars into such a rousing success.