The latest issue of Gamasutra sister publication Game Developer magazine includes a postmortem of Sucker Punch's Infamous, written by studio co-founder Chris Zimmerman.
Infamous, an electricity-themed open-world action game, was the first original property from Bellevue-based Sucker Punch since 2002, and the company's first title rendered in a relatively realistic visual style -- a distinction that led to surprising snags along the way, as revealed below.
The following excerpts from Game Developer magazine's recent postmortem, published in the October 2009 issue, illustrate how Sucker Punch grappled with the "cost of realism" and an early lack of vision, then focused on iteration to successfully launch the game.
Initial Lack of Vision
With the move into a considerably new tone of game, Sucker Punch started off with a hazy view of the game's underlying core -- which cost the team months of productivity early on:
"For the first year of the project, we really didn’t have much of an idea of what kind of game we were building. We were sure about some things -- that you played as a superhero, that the game took place in a city, that it was a T-rated game -- but beyond this, we weren’t sure of much. We described the game as a combination of Animal Crossing, the Sims, and Grand Theft Auto -- an action game with pervasive simulation elements.
"It isn’t surprising that this description didn’t exactly enlighten people -- we should have realized that our inability to articulate a crisp vision for the game meant that we didn’t have a clue what we were doing.
"More than anything, this lack of focus represented us misapplying the prototype-test-iterate cycle that served us so well in tightening up gameplay. We worked on prototyping a bunch of innovative little gameplay scenarios, expecting that a clear vision of what the game should be would arise from the prototypes. It didn’t; we just ended up with a bunch of little prototypes that had nothing to do with each other. We were nearly a year into the development of what was ostensibly a combat-focused action game, but still hadn’t done any work on letting the player fire a weapon, or on having enemies that shot back!
"Eventually, we wised up, realized how lost we were, decided that we were doing an action game, and started working on basic game play. Things progressed quickly from there; most of the innovative little gameplay scenarios were set aside, to be revisited once the core of the game was tight. A few of them even ended up in the game—like pedestrians spontaneously taking pictures of the hero. Overall, though, our lack of focus cost us months and months of work, pursuing what ultimately turned out to be discarded ideas."
Underestimating The Cost Of Realism
The switch to a different kind of game also had an impact on production, as Sucker Punch found that the realistic look targeted by Infamous required an entirely different -- and more expensive -- approach to what was used for its earlier, more colorful Sly Cooper games:
"We wildly underestimated how difficult the transition would be from the bright and cheerful PlayStation 2 graphics of the Sly Cooper games to the gritty and realistic PlayStation 3 look of Infamous. In effect, we tried to take two steps at once. First, we needed to catch up with the realism in the best-looking PlayStation 2 games; second, we needed to match the even higher quality bar being set by the best teams as they moved to the next generation of hardware.
"This was a struggle on multiple levels. First, we had a technology gap to clear; the engine we used for the Sly Cooper games needed to be substantially rewritten to run on the PlayStation 3. We needed to rework our whole pipeline to accommodate things like vertex and pixel shaders. Unfortunately, we didn’t start this effort until after we completed Sly 3, leaving us without a pipeline or engine running on the target hardware for the first 10 months of product development.
"The lack of a working pipeline and engine made it a challenge for our art team to learn the new skills and techniques necessary to build the realistic content we needed for inFAMOUS. All our assumptions about how to build content needed to be revisited to match our new expectations and the capabilities of the new hardware. In addition, our art team needed to hone their skills building realistic content, and the lack of a working engine made this painful."
Prototype, Test, Iterate
There is absolutely no theme more common to Game Developer postmortems than that of teams reinforcing the value of prototyping and iteration. Some teams learn that lesson the hard way; others learn earlier on. Fortunately for Sucker Punch, it fell into the latter category:
"One of the first lessons we learned at Sucker Punch was the difficulty of designing a fun video game on paper. We’ve never had much luck predicting which ideas will end up being fun, and which ones won't. The successes we’ve had have been the result of trying out lots of ideas, pursuing the ones that had a smattering of fun in them, then testing and iterating until we’d refined the fun into its purest form.
"For example, Infamous started with an “ammo” model much like other shooters—defeated enemies dropped ammo for your superpowers. Ammo came in multiple colors, and each superpower used a particular kind of ammo. One kind of ammo needed to be used right away, and timed out if not used quickly enough. Enemies also dropped health packs, which looked like a different color of ammo to the player.
"There were good reasons for all of this complexity: we wanted to limit the use of the most powerful hero abilities, we wanted to encourage players to use a broad set of their superpowers, and so on. We’d been uncomfortable with the way the system worked, but felt boxed in by all the constraints we’d placed on the design.
"The focus test trumped all of this—players hated the ammo model. They were confused, they didn’t like having to conserve ammo, they didn’t think picking up glowing balls made them feel like a superhero ... it was a complete train wreck.
"The focus test forced us to radically simplify things. After some false starts, we unified ammo and health around the hero’s core electrical powers. Drawing electricity from the environment replenishes the hero’s store of energy, which is then used up by his powers. Drawing electricity also heals damage to the character. The player ended up with an easy-to-understand model, and a simple rule to follow—when in trouble, look for electricity to drain. At the same time, we gained a powerful way to limit the player’s abilities—if an area of the city was blacked out, the player would be relatively weak in that area.
Overall, this change was a home run. Without our core belief in focus testing, we would have a hard time making this large a departure from shooter conventions."
The full postmortem for Infamous explores "What Went Right" and "What Went Wrong" during the course of the game's development, and is now available in the October 2009 issue of Game Developer magazine.
The issue also includes a countdown drawn from the seventh annual Game Developer Top 20 Publishers report, a feature on wrapping C++ into scripting languages, an examination of in-game tutorial methods, and our regular monthly columns on design, art, music, programming, and humor.