About AAA Development Process: BioShock and DOOM4

[Touchdown]

This is something I wanted to talk about a couple of times. Back in 2007 (or 2008?) I watched a fantastic post-mortem on BioShock by the Lead Programmer Chris Kline. He went through the major points in the development, presented some early versions of the game and so on. The video is currently lost from the internet and while I have a copy of it, I'm not 100% sure if it's ok to reupload it (trying to find out, just in case).

Anyway, here are some of the main points of interest. Pay attention to the process itself, it's not about BioShock the game but about it's development.


___________________________________
The game began development in 2002 but shortly after that Irrational had to move to other projects leaving BioShock for 2 years. The work resumed in 2004.


AUGUST 2005: The Vision Demo
The intent was to create a "Vertical Slice" demo where all major game systems would be implemented, even in a half-baked form if necessary. A proof that the ideas that the team had were on a right track.

The result: A couple successes, the demo was running on a 360, a lot of systems implemented. But everything else was a failure. Art direction was bland, gameplay lacked emotional connection, no sense of fear or wonder, the AI Ecology didn't make sense to the player.

The team felt like they've failed.


SEPTEMBER 2005: Full Production
The team intended to correct mistakes of the Vision Demo.

The result was that everyone was doing their job, the progress was being made across the board and no major problems occured. However THAT was the problem. Nobody was paying attention to the big picture of how it would all work together. Systems were being implemented and cut all the time. Levels were built with no consideration for story or gameplay.

They were making something but it wasn't clear if it was going to turn into a good game. The morale of the team was dropping because people could see it was not working.

The only way to fix it: threat of public humiliation. (sounds familiar?)


MAY 2006: Reveal
The E3 demo was critical, it forced the team to pull together. They had to prove that their ideas could work, both to the audience and to themselves. (sounds familiar?)

For the first time the team felt like they've built something that had a potential to be a game. (15 months before the release!)

The demo was mostly smoke&mirrors, triggers were scripted to show things that were supposed to happen dynamically in the final game.

The success of the demo yielded more time and money.


SEPTEMBER 2006: X06 Demo
The team was finally seeing an exciting direction for the game. (11 months before the release!)


OCTOBER 2006 - JULY 2007: Final Push
Presentation of the games systems was completely redesigned.

Refocusing the game, moving from stats-based to organic systems ("shooter-ification"). Only at this point the game was really starting to take shape and the 'real' full production began.

Up until this point Little Sisters were still mortal.

First two levels were redesigned to capture the player in the first 20min.

A lot of changes were made to the way systems worked with each other: fire melting ice, enemies reacting to being set on fire, generally making sure the gameworld is consistent and everything behaves as the player expects.

The story was not finished until January 2007 and the AI dialogs were redone in March.

In short: a lot of new features and the final shape of the game came to be in the last 10 months of development.
___________________________________


So, what does all of that mean? As you can see, there are some similiarities between the development of BioShock and DOOM4. The reveal is probably the closest thing, even though BioShock was unveiled with screenshots (the demo was for the press only).

But the thing I wanted to point out is that even landmark games like BioShock might have a very rocky development. The design kept changing and it was in the flux until the E3 2006 demo, the 'real' development took <10 months out of 5 years in total, and A LOT of stuff was made during that final stretch.

Point is: game development is not always as efficient and calculated as some might think. Hell, how about Deus Ex? I don't have the story on it but it supposedly had a very chaotic development and it was close to cancellation multiple times because of the 'craziness' happening on the team.

Someone mentioned (Shaviro?) that the QC demo might have been a 'vertical slice' type of a deal with no actual game beyond what was presented. Much like with Irrational, the demo might have been a 'get our shit together' / 'threat of public humiliation' situation (well, kind of public, in both cases) and the reception was supposed to be an indication of whether the current direction is good or not. Also to boost the morale. And since the QC demo was well received, we can assume the game is in full production right now. You know, the 'real' full production.

Obviously none of that means that DOOM4 will shake the industry. It might flop and turn out half-baked. All I'm saying is that just because there's chaos at id, it doesn't mean greatness cannot come out of it.

[again, a reminder, this thread is not meant for discussing BioShock as a game, we don't care about it, it's only meant as a comparison of the development processes]

[Marnetmar]

Didn't the original Half-Life nearly get canned for similar reasons as well? Development hell doesn't necessarily equal a bad game.

[RUSH]

I agree that a troubled development process doesn't necessarily mean a bad game. However, I'm still a little concerned when I hear Tim Willits claim that Rage "had a soul." That game was mostly an empty shell in terms of character design and plot, and while I don't expect amazing characters or plot in the new Doom, I still hope they can create a proper identity. I want the new game to FEEL like a Doom installment, not some Brutal Doom/Call of Duty/Halo concoction.