What Went Wrong

1. Evolving Product Positioning.

The spec of BioShock changed so much over the course of development that we spent the majority of the time making the wrong game- an extremely deep game, and at times an interesting one, but it was not a groundbreaking game that would appeal to a wide audience.

We knew from the start that we'd have to make late changes to really bring the game to life-we had even built our original schedule to allow for six months of finalizing-but the amount of change that we ended up needing seriously exceeded our remaining schedule. Ultimately, we were very lucky to get an extension in the eleventh hour.

Part of the reason for the late course change came from not having our internal product message clear from the beginning. BioShock had initially been positioned as a hybrid RPG FPS. The decision to reposition the game as a focused FPS came later, after our initial production phase in summer of 2006. Had we been working with an FPS mentality earlier, we could have made better use of our time.

Another contributing factor to the late switch was that the game had been more or less proceeding according to plan throughout development, so there didn't seem to be any emergencies that needed intervention from higher levels of management. Milestones were completed, goals were met, development seemed to be proceeding uneventfully. But as the game neared alpha, key people began looked more closely and saw that BioShock wasn't on track to become an accessible and marketable game.

As mentioned in the first What Went Right point, the real turning point for BioShock came when we had to present the game to the outside world, which forced us to carefully consider the story and takeaway message. In retrospect, we should have tried to develop some of that thinking sooner.

2. Narrative content development happened late.

We had many drafts of the story over the course of development, but the final draft turned out to be an almost complete rewrite. To make matters worse, we failed to fully exercise the narrative production path in early versions, so once the final draft was complete and recorded, many implementation and pipeline problems appeared for the first time that should have been caught and resolved earlier.

The core issue was that a giant pile of content came online well past beta, and the team had to scramble to get that content correctly installed while also fixing bugs. Competing demands for time and resources meant that, unfortunately, some of the important narrative details of the game weren't created until the final rewrite, and therefore required quite a bit of work to retrofit them into an existing game.

To add to our woes, the first focus test feedback on the narrator's voice came back extremely negative. People found the character extremely off-putting, so we recast the part at the last possible moment. On the positive side, this allowed us to refine several areas of the game (including the intro sequence) to ensure that the player knew what to do at the right time. On the other hand, it was difficult to get all the content in, debugged, and polished in the remaining timeframe.

3. Scaling vision to team size.

Our goals and vision pretty consistently overreached our production capacity. Ideas that started out small turned out to require a tremendous amount of coordinated support to reach a polished state. Although we were able to add resources regularly and make some cuts late in the game, our ability to plan for how much work it would take to bring any single idea or space from concept to completion was poor.

In addition, we didn't have an effective internal review process set up until very late in the development cycle. We would work on levels to the definition of a milestone, get feedback, and then set it all aside for a while or leave it in the designer's hands to polish. It wasn't until we created a more regular review cycle, where all the key players sat in one room and watched the gameplay session and someone logged all the bugs, that we were really able to define the amount of remaining work to bring a feature or level to completion.

The ultimate reason we were able to pull off the game we made to the level of polish we did, was the sheer dedication of the team working on it. Even though we had padded the finalizing schedule, we still far exceeded it.

It's to the team's credit that they stepped up to the challenge with incredible dedication. The final crunch period on BioShock was long and hard. People who had been pacing themselves for six more weeks of work had to reset to three more months of work rather suddenly. Had we understood our polish cycle requirements sooner, or had known about the additional time earlier, we could have paced ourselves better.