This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.
[Gamasutra is proud to be publishing notable Game Developer magazine postmortems online for the first time - starting with project lead Finley revealing the creation of 2K Boston/Australia's seminal BioShock.]
The story of developing BioShock is an epic one and isn't easily expressed in 10 postmortem points. The team and the game changed remarkably over the course of development. A company was acquired. The team size doubled. The product focus changed from RPG hybrid to shooter.
It's easy to talk about the processes we used to develop the game, but it's harder to describe the creative spark that somehow managed to turn the most unlikely of premises (a failed underwater art deco utopia set in the 1960s) into a marketable shooter. It took a visionary to make the creative choices to guide the game, and an incredibly talented and hardworking team to bring that vision to life.
1. Every demo tells a story.
Demos were galvanizing moments for BioShock. They led to a unified team vision, identification of problems and solutions, external excitement, and internal support. For example, the project was signed after GameSpot ran an exclusive feature based on a single-room graphics demo.
Since BioShock was a relatively unknown IP outside the game development community, the public's impression of it would be critical to building the buzz we needed to make it a commercial success. As a result, every time we took the game out in public, we put great thought into the message we wanted the demo to deliver and the level of polish of the presentation.
Our first public presentation was at E3 2006. We had developed a great deal of content before that point, but hadn't yet built a space that really demonstrated the game experience to our satisfaction. The E3 demo forced us to focus the whole team on what the user experience should be. We defined a message for the demo- player choice-and built a narrative around that message. Even though the experience was highly scripted at the time, it effectively demonstrated the feel of the game we wanted.
Another example of demo-inspired development was the "Hunting the Big Daddy" demo. Though Big Daddies and Little Sisters had been part of the game in some form since the beginning, initially the player could confront Little Sisters directly without necessarily needing to dispatch the Big Daddy that protected them.
During the development of this demo, the team discovered that with some polish and tuning changes the act of dealing with a Big Daddy could be a truly epic battle in itself. This led to the realization that Big Daddy battles should be the key to player growth, essentially providing a roving boss battle that players could undertake at a time and place of their choosing.
Another example is the graphical effects on the player's hands when using plasmids, which came out the first BioShock trailer created with Blur Studios. In that cinematic, the player uses a hypodermic needle to make his arm into a weapon; after the injection the protagonist's skin blackens and swells and angry hornets burst out of it to attack the Big Daddy.
When working with Blur to develop the trailer, we knew that the sequence didn't accurately reflect the game's visuals, but we did it because it really captured the vibe of what the "genetic modification" part of the game was all about.
2. Course corrections.
One of the true successes of BioShock's development was our ability to identify and react when the game was not shaping up to become what it needed to be. For example, the first vertical slice prototype we built was an non-navigable linear corridor shooter that looked like it took place in an abandoned box factory.
It didn't provide a compelling experience as either an RPG or a shooter. In response, we threw away that prototype and started again from scratch with the goal of building a single room that felt like the ruined underwater utopia we were trying to build.
First we did concept art passes. Once we got a concept that worked, we built it. Then we used it as a demo space. We used that single room (now Kashmir Restaurant in the first level of the game) as an artistic reference that guided us in creating an aesthetic unlike any other game on the market. (For more about the artistic style of BioShock, see the free art book download here.)
Each department went through a similar crisis moment over the course of the project. These frequently came as the result of the demos, but not always. At one point, when facing a shortfall of programmers and an overflow of tasks, we proposed removing physics objects from the game entirely in favor of having only large, constrained physics actors.
This would have allowed us to spend much less time tuning the physics of individual objects while allowing the world to seem somewhat dynamic. However, doing so would have removed a huge level of interactivity from the game, so that decision was corrected relatively quickly.
In terms of design, we created a depth and density of game systems that fit into a game about character building and choice, but would not have been competitive as an FPS. Around the time that the game went into alpha, we took a hard look at that gameplay and realized that, although there were many choices, they weren't very compelling.
This was because we hadn't been thinking as much about making a shooter as we should have, and many of our key interactions (weapons tuning, plasmids, length of AI engagement) were designed and tuned for a slower and more cerebral experience. To put it another way, nerdy RPG-like stat changes just didn't seem meaningful in the vibrant and dangerous world of Rapture.
Once we recalibrated the game to be more like a shooter, we simplified many of the deeper systems tremendously so that the user would be able to understand them. We also put more polish time into the core interactions of the game, such as the weapons, plasmids, and user interfaces. We ended up with fewer choices overall, but each one of those choices was infinitely more functional, understandable, and fun than the previous ones.
It was inevitable that we lost some progress due to these major corrections. But the team's ability to pull together and address the fundamental problems was amazing, and the results were well worth it.