CR: From a developer's perspective, it is kind of exciting, the idea of divorcing story from writing. That's something you could never do in any other form of entertainment, and for whatever reason, really, it's rarely, if ever, actually been done in games. That is kind of a cool thing.

PR: Right. In a sense, there's kind of been a weird and unfortunate development on that side, because, think about it, right? Writers in games just kind of got to this stage where they were able to make their point. Where they were like, "Guys, you really need to bring in good writers. Like, writing matters."

CR: A lot of people still don't believe that.

PR: Yeah, exactly, and so they kind of started the clock on it, and they're like, "Okay, now we have our window. Let's get in there and really be important, and start in there early, and get what we need, and get paid what we should be paid." All of this stuff that writers have been struggling to do in every other medium they've worked in. And I kind of feel like, "Okay, yeah, but now we're coming to the end of that."

CR: They're pulling the rug out from under you.

PR: Yeah! And now we're going to get to a point where it's like, "Okay, you guys realize that what you're doing is sort of the literary equivalent to being a texture artist." Right? And that's something I have to struggle with, because I don't want to say that to anybody. That feels patronizing, and condescending, and belittling to what is ultimately a very, very, very vital craft.

CR: Now you're offending texture artists.

PR: Yeah, exactly. There you go. I'm offending everybody. I will actually manage to piss off just about every discipline in this entire business with this one talk. No, here's my point. The art director comes in. The art director does not... he establishes a structure in which texture artists, modelers, character designers, 3D artists, level artists all have to follow certain thematic objectives and targets, and then be able to sit down and generate data that supports that. And that's what they're doing.

That's not to belittle it, but they are generating data in support of that target. Right? And my point is that writers have traditionally been people who tell stories, who construct stories, and narrative structure, and then build on that... and then, yeah, they write a bunch of dialog, and they write scenes that help reinforce that, and they iterate, and edit, and do all these things they need to do to make that stuff hang together.

And what I'm saying is, the problem, in a sense, we are taking away from writers, to a certain extent, the structural part of their job. Right? At least on this game, we are saying, "No. There is a narrative design that helps to kind of support the player doing what they want to do, and having meaningful consequences as a result of that."

The writer's job is to deliver the dialog, the scenes, the character... they still have to deliver that stuff. If we don't have that stuff, it sucks. Right? So it's still vitally important. Just like we have to have good texture artists. We have to have good shader systems. All of that stuff matters. None of it is more or less important than anything else. But, what we are doing is we are taking something that is in the traditional role of the writer, and we're breaking it out, and we're making it systemic, and we're letting the game handle it. Right?

BS: I don't think that's actually really... I don't think it's bad, and I don't think it's that different from how it is when it works well anyway. Because the writing has to match... has to meet the objectives of the game, and you can't just have someone, stream of consciousness, write out a story, and be like, "All right, now we're going to make our game."

PR: Well, yes. If you want to take it to its absurd limit, yes. But the problem is that, even if you dial back from that kind of scenario, that's not too far off from what a lot of game writing ends up being. It's like you get a game that's in production, where they go, "Shit. We're at beta. We need a story." Like, I mean, no joke, that really happens.

And so then they bring in some writer, who may be a great writer, and may even have done maybe a great game's writing, and say, "Okay, you need to help us make this good now." And they've been put in a totally untenable position. What are they going to do? All they can really do is write cinematics. All they can really do is write stuff that the player has no impact on.

And in that case, they're better have asserting a certain amount of authorial control, just so at least that stuff is good. Right? You look at GTA. I love the writing in GTA, but it's a damned talky game. They're still cutscenes. I'm still sitting through them. I can skip them, but they're still cutscenes.

CR: The interesting thing I was going to say about GTA in regards to what you were saying about... I think your texture artist metaphor is actually a really apt one because, looking at GTA, I think the part that is similar to what you describe is everything that happens when you're not in the cutscenes.

Like the dialog, when you're not in a cutscene, is essentially textural. It's essentially creating a texture within the world. As you walk down the street, and you hear conversations, and you bump into someone, and it generates something they can say, and there's all these, essentially actually a massive amount of dialog, that is embedded in the game that is not part of the core narrative. That's sort of what you're describing. You're extending it to the actual story of the game.

PR: Oh yeah. That's 100% correct. Yeah, I want the core plot to be delivered to me by me overhearing cell phone conversations while I'm walking down Mohawk Avenue in Broker during GTA IV. That's what I'm saying. And yeah, I'm still going to walk into a downstairs adult video store and still have conversations with people, but my feeling is that, as much as possible, there shouldn't be obvious seams between that content. It should feel like part of one cohesive continuity.

BS: The biggest difference with that metaphor is that you don't... players don't mind if they see the same texture a few times.

PR: Yes, in that situation, probably we are staring at literally 10 or 20 times the amount of content. But I think there are a number of technical hurdles we'll have to overcome to get to a stage where it's feasible for us to build out huge, huge, huge quantities of that stuff.

I mean, okay, let's abuse our texture art metaphor a little further. On Far Cry 2 one of the things that people have commented on is that, well, this is an amazing, large, open world, and yet you can walk right up to any tree, add incredibly close detail with your sniper scope, and see individual veins in the leaves. And you can see the bark, the texture of the bark.

And we did that - or Alex Amancio, our art director did that - by developing a completely different kind of graphics pipeline that put the emphasis, rather than having giant, high-resolution textures with lots and lots of data in the texture, on layers of shaders, using masking systems, and kind of procedural noise, that allowed us to generate very noisy, random, life-like kinds of detail in our rock, in our ground, in our leaves, in our vegetation, in our animal fur, and all these things.

CR: It's the opposite way that id is going with the MegaTexture thing, almost.

PR: Exactly. None of that stuff's going on there. It's the interaction that just kind of... it's super-pretentious for me to call it this, but it's this kind of highly granularized, fractalized sort of approach. But that's really the way it works. When I was doing the talk on narrative, I said a big goal for us is to offload the processing of the story to the player's own brain. Right? It's kind of like an exercise in distributed computing.

Players... human beings are predisposed to seeing stories everywhere they look. It's something we've genetically developed. It's part of our nervous system. We indulge in what's known as a "narrative fallacy". When we walk around and things happen to us out in the world, if we're listening to our iPods, and the context changes depending on what the soundtrack is. This is something that human beings just do, whether we are thinking about it or not.

And I think as story designers and as game developers, what we're trying to do is give the player lots of raw materials to help produce that effect. And I think that's kind of the same idea. We need to develop a new story pipeline and a new kind of story architecture that allows the player to supply a certain amount of the context, and also infer a certain amount of the intent, even if it's not actually there. Because we do that in real life, so we can apply it to our experiences in simulation, and in our work as well.