Saturday, January 15, 2011

Dark Past (part 2): On level design, hookers, cybernetic architecture, Tony Hawk and all that converges.


Last time, I mentioned a lot of games that we like for their emergence and "cool stuff that happens" -- but the difference is that games like Thief, or "immersive simulations," actually do something with that emergence.

If, in Grand Theft Auto, you beat up a sex worker to get your money back -- an immersive sim approach would have that hooker tell other hookers about you, and no hookers would ever service you again or help you in missions -- unless maybe you stole a nice car and left it in their parking lot as a gift, or ran over a pimp for them, etc. In this way, all your actions "listen" to each other.

A "hooker map" for GTA4. (Apparently they prefer the MET, Times Square, Long Island City...)
... But GTA doesn't do it because that's not what it's going for. (Too much thoughtfulness in my bowl of destructive mayhem.) And that's okay. But an immersive sim, like the Thief series, does it differently.

What I've always loved about Thief was the level design -- and by level design, I mean mainly the floorplans and complexity of the spaces -- not so much the visuals.

Not that the games are ugly or visually boring, but... Thief II, released in 2000, used essentially the same Dark Engine used in Thief I from 1998. Compare this to more powerful engines as those in Unreal Tournament or Quake 3 Arena, which were released months before in 1999... Thief II was already severely behind from the start. Plus, a lot of the models and animations remained untouched from Thief I, which was smart for production but limited its visual novelty.

Fortunately, "immersive sims" aren't "immersive" through their graphical fidelity or photorealism. (see: immersive fallacy) Rather, they're immersive in how they simulate / simplify the complexity of interlocking systems and the beauty of exploiting these systems or getting caught unaware by them.

Surely there's a connection to be made between Thief's mechanics, the way each system "listens" to each other, the interconnectedness of one node to another to create a new system greater than the sum of its parts, reacting in novel ways to player input...

... and Thief's unique brand of FPS level design? Rooms connect to others in countless loops; a guard might be "listening" in the floor above you and sound an alarm, which places the eastern wing of the level on a lockdown -- when you try to escape, some thugs are already waiting at your apartment. Rooms interact with rooms, walls interact with floors...

A level is a system.

... which is hardly a new idea. In his influential essay "The Architectural Relevance of Cybernetics" (1969), Gordon Pask theorized an architecture that could iterate upon itself and respond to user inputs. He was also involved with the "Fun Palace," a never-built plan headed by fellow architect Cedric Price.

The main conceit of the Fun Palace was that visitors could shift walls and ceilings around according to their needs, as a sort of public democratization of architecture. This dynamicism would reflect public use; sometimes a performance, sometimes as a museum, sometimes a playing field, etc. In this way, a user and a building engage each other in dialogue -- this is the basis of Pask's "conversation theory."

The Fun Palace.
It is also the basis of current video game design theory: meaning emerges from a player's "dialogue" with the system, rather than a player's dialogue with the developer. (... unless you're Tale of Tales, who reject the necessity of a system and support the latter.)

One possible direction for the future is to give the system more power; one of Pask's early projects was the Musicolour, sort of the first audio visualizer -- and most impressively it would eventually get "bored" of repetitive music and stop responding to it. Imagine if iTunes judged you for having Ke$ha in your playlist and refused to play any of her songs. "But iTunes, I like her ironically!..."

... This is, however, an improbable direction for the future of video games as conversation. Right now, even amazing immersive sims like Thief are based on how predictable the NPCs are or how limited the "system" is. Previously I said my interaction with guards were among the richest conversations I've had in games -- but if you told Pask about it then he might reply, "what conversation?" It's more like a one-sided interrogation if anything.

Often when we give the system comparable power to players, or even merely more power, we complain that it's "cheating." It's too unpredictable! It's too powerful! Single player games suck, etc. (Or in a game like Call of Duty, are we actually empowered at all, or is this a false consciousness?)

However, that's only if we see NPCs and AI as the only manifestations of the system's power; the level itself constrains you at all times, determining where you can go and where you can't. Maybe I wasn't actually talking with the guards, but rather the floor and the walls, as Pask posits?


And the only levels that even come close to this empowerment, I think, are the ones popularized by the Tony Hawk skateboarding games. (Which are, of course, indebted to the actual real-world act of skateboarding in places you aren't supposed to. Interesting how Tony Hawk kinda matches the aesthetic of Thief...)

In Tony Hawk, you do all sorts of mundane goal-oriented stuff like spelling out "SKATE" or collecting cash or video tapes. But much like Thief, stealing the loot isn't the real objective. Rather, it's about your process through the system, how you get there and when.

But this is where real-life skating and Tony Hawk's virtual skating diverge. While both are about finding interesting routes for performing tricks through the environment, or "lines" as the skaters call 'em, real-life skating is about short bursts of impressive combinations while the latter is about a long, uninterrupted chain of tricks to rack up ludicrous amounts of points -- which, in turn, places (undue?) emphasis on manuals. (A "manual" is when you raise your front wheels off the ground, but the tail of the board doesn't touch the ground either. They're difficult to maintain.)

The Chicago warehouse level from Tony Hawk 1; the skating equivalent of de_dust. Perfection.
So you're playing Tony Hawk and you want to score points. You "read" the level and try to find the best "lines" -- and if you mess up, you can "manual" to maintain your combo; a fail-safe, the skate equivalent of Thief's flash bomb -- and as you grind, kickflip, 360, tre bomb, etc. along your line, you try to be as "steezy" as possible. (Do the kids still actually say that?)

If you were competing with another skater, the smart thing would be to interrupt his line somehow.

... Which is more or less how pro-level 1v1 Quake 3 Arena works: you have a specific line throughout the level, timed precisely to weapon respawn times. If you know your opponent loves the lightning gun, you'll want to deny them the lightning gun by collecting it right as it respawns -- to the extent that it respawns right as you run past. Along the way, you rocket-jump, trick-jump, bunny-hop, 360, etc. (Or maybe diverge from your own route, briefly, to throw 'em off your own strategy.)


But Q3A takes place in a multiplayer context*, so I'd argue that single player Thief is closer to single player Tony Hawk. Of course, the two are still rather different -- one is about elegant, long stretches of movement (Tony) while the other is about infrequent bursts of movement with a great deal of waiting in-between (Thief).

Actually, let's just ditch nuance here: Thief is the best pro-skating video game series ever.

So let's design our steampunk castles like schools, malls and skateparks; our concrete bowls are gold-laden throne rooms, our rails are lush carpeted floors... and our ramps are shadows.


Next:
Dark Past (part 3): Letting Go of the Immersive Sim


* (People game critics wonder why there isn't more scholarship on non-MMOG multiplayer games? Well here are my excuses for shying away from the subject: (1) they're all basically rock paper scissors at their cores, (2) popular strategies emerge over years of play, so scholarship actually becomes "obsolete" and relegated to history as key players change the way the game is played, (3) almost all abandon any pretense of narrative, (4) to write adequately about the highest level of multiplayer play, you have to be a really fantastic player, which you probably aren't.)