That's one of the challenges. If you're making a big triple-A game, there is so much groundwork that has to be done regardless of whether there's any creativity or meaningful design work, and no matter what, you've still got to have the engine [and everything else].

CB: Think about all that, right? You're making a piece of bug-free software that goes on a disc, and you're making an experience like a movie... and it's about four times longer than a movie. GO! (laughs)

And so to act like that's not some sort of huge creative endeavor that dwarfs a movie production is just not realizing what's going on. So you have engineering staff, there's a gigantic engineering staff because you're trying to innovate technically. You've got platforms that are changing every five years, so those can't even stay the same... It's just a really challenging environment, then, to have this creative thing happen.

I'll have to give our creative director, Brett Robbins, a lot of credit, too; he's got a gameplay sensibility. He did a really great job of keeping his eye focused on gameplay, because a lot of times you forget that the creative director has the whole fuckin' game to work with.

He's got the meta-game, and story, and character, and level design, and what's happening in mechanics. And then you have to worry about what's happening in the gameplay -- the sort-of conflux of all that -- and that's where if you take your eye off the ball, you lose. So he's always kept his eye on, "Okay, is it fun? Is it working? What are we doing? Is it fun? Is it working? What are we doing?"

And by doing that, it helps drive all the departments towards one answer; little X moments, where you're like, "Okay, here's what's supposed to happen in this moment: the raptor's supposed to jump at me, I'm supposed to hit him with stasis, and I'm supposed to cut off his arms."

That, alone, invoked every department in the whole team. They all had their tasks to do, and they would build these gameplay moments, as opposed to just building a mechanic over here, and building another mechanic over here, and then hoping that at some magical moment their systems would interact properly. You start with the interaction that you want, and then you build the spiral out from there.

Are you familiar with the MDA Framework [PDF link]? It's a theory about game design. It's based on what you just said. You start with the thing you want to evoke, and then you work backwards to the more mechanical side.

CB: Well, and it sounds like common sense, but when you have so much work involved to get it done, that can get obliterated really fast. Now unless that's a spoken ideal, people just... and they don't even realize that they're missing it, right?

I've been on games where we thought we had it, we established it, and we went off and established all the things around it. But then we forgot to keep checking in, to see where we were all going. And then we were all like, "Wow, that didn't work very well..." So, you know, you live and learn.

That speaks to what you were saying about the concept of design documents becoming less of a rigid thing; because if you're trying to let your more base-level mechanics emerge from the sort of emotion or interaction that you're trying to create, it's not very useful if you've already decided that these are the mechanics.

CB: Well think about that. Think about the lightning in a bottle that you're trying to capture, right? You've got all these different mechanics finally coming together, and it's a big layer cake.

The first foundation is the mechanics, right? You get those in, and you put them together -- like stasis, and [the TK gun] -- and people start playing with them and doing setups.  In some random meeting, after hundreds of meetings, a flash of inspiration hits the team, and we're like, "Oh! That was cool moment! Did you see that? There was a flash and he ran around the corner, and you hit him, and this happened!

If you cannot capture those flashes of lightning, if your process doesn't let you capture those, and institutionalize those, then you're failing as a creative culture. So if you're basically saying that, at some point I take a snapshot of my thinking, which is a design doc, and it doesn't get a chance to evolve and move? You don't really get it, as a creative culture; you're not really understanding what it means to be creative.

Now the other half of that is, you've got to get the goddamn game out, so you can't keep changing forever. So there's a healthy [balance] between the two, but you have to be able to just take a snapshot and make a design doc, which means that you've done your homework and researched it. Then you can be like, "OK, but if we find something cool, we get to do it!", right? So that's a real challenge.

But I think we've done an admirable job with holding off our development guys, who are the scheduling and capacity people, and having them feel comfortable enough to let us innovate when we can, and catch those flashes of lightning in a bottle.

And, of course, it's all statistics, so the longer you have, the more that you can do; like a five year game catches a lot of lightning in a bottle, but a two year game doesn't have much lightning to catch in a bottle, since you're too busy with everything else.