Wednesday, November 19, 2008

Cityscapes and Immersion

It sounds silly, but one of the things I enjoyed most about Mirror's Edge was the menu sequence. The city in the background was beautifully done, and the music was perfect. I've always had a thing for cityscapes, even if they're only fleeting: the intro for Megaman and Streets of Rage, for example. The games had nothing about cities in them, but I liked the intro sequences. Don't even get me started on Blade Runner.

There's something very vibrant about a cityscape that vanishes when the game actually starts. Even if the game is about the city, such as Simcity, it doesn't contain the same feeling of potential and vibrancy.

So I was thinking about what kind of game could still give that kind of feeling. If we zoom in, we can get the same kind of feeling from unique things in the city. Not simply traffic patterns and police stations scattered in among the commercial zone, but unique bits: what makes this police station different from other stations? Why does the foot traffic linger at that particular toy store? Why is that particular person unique?

You can see it in games like Assassin's Creed, Crackdown, and even City of Heroes. There is an inkling that THIS roof is particularly interesting, or that the crowd is interesting, or that any number of other bits of city are interesting. However, these all tend to fall flat if you get close: the crowd is just a milling of identical NPCs, the roof is a nice enough view, but gets boring instantly, the graffiti is fun as background noise, but if you actually look at it, it's one of five designs repeated a thousandfold.

I think that one major reason for this is because video games are not particularly immersive. We talk about immersion, but one reason we talk about it so much is that we're really pretty BAD at it. Video games are marginally okay at immersing us in a "flow" sense - getting us involved in the mechanical rules and so forth - but they're pretty bad at getting us immersed in a more classical sense. Movies are much better at it, because a skilled director will draw our attention to immersive details, such as a unique piece of graffiti, an interesting view from a roof, a crowd that contains real individuals...

A big component of these tricks is that they calibrate our judgment by showing us what "people" think and do (even if those people are imaginary). Sometimes this is extremely straightforward: they simply show people thinking, doing, and feeling. This is especially common in children's shows (and scifi), with their hundreds of close-ups on people's smiling or unhappy or angry faces. As the audience matures and learns the thousands of cultural clues, the directors simply linger on those cultural clues instead of the faces proper.

For example, we know how to feel about dingy underlevels of the spaceship in Alien because the director uses the same cultural clues we've learned about dingy industrial areas. He then plays it up by focusing on the interactions of the workers, showing how THEY feel about the place (and the crew above them). Using these cues we can also judge the lower levels strongly, and when the fight starts raging down there, we feel a lot more immersed.

Video games often use these methods, although usually pretty clumsily. However, they are very limited by camerawork. The only ways to zoom in on someone are to either have a cutscene and take movie-like control or to have pop-up portraits. Neither of these is ideal: one is extremely limited, the other makes the game into a non-game, however temporarily.

But if we acknowledge that this is something we need to implement, we can use our limited methods to give judgment clues a lot more adaptively and precisely, perhaps (in the long run) even better than a movie director.

For example, lets say you're playing a game where you're moving around a city, similar to Assassin's Creed or Crackdown (or Mirror's Edge if you could actually move around the city). When you discover a particularly nifty place, you stop for a moment to enjoy it. But there is no real sense of immersion, nor any reason to dally, so the enjoyment quickly wears off and you move on.

But what if there was someone with you up on that neat rooftop, someone who could say "wow!" and thoroughly enjoy the view? In a movie, that's what would happen: the actors involved would stand at the edge of the roof and look around, and we would mirror their judgments. In a video game, that doesn't happen. But it could.

In theory it could be our own avatar that does these things, but assuming you have any kind of direct control normally, I think that stealing control away would BREAK immersion rather than reinforce it. So we are more or less limited to having another character present.

There's no reason it has to be human. Maybe it is - maybe it's your sidekick, or your girlfriend, or whatever. But it could also be an AI, a ghost, a memory, a flying dog... they don't have to be human, they just have to make judgments that we can agree with.

Pacing is an issue, too: most modern games are paradise for people with no attention spans. If we're deeply engaged in the rules and mechanics of a game, taking time out to think about how pretty the view is would be distracting.

On the other side of the affair is the difficulty in actually getting characters to judge things in a way that aligns with the things our player wants to judge. In a movie, this is easy: the audience is damn well going to judge whatever you point them at. In a game, it's more difficult, because you can't point the bastards. You can't even rely on where they point themselves: in a first-person shooter, I might be running along looking down the street, but as a player I'll be noting the cool sunlight effects in the corner of the screen. If a character behind me pops up and says, "wow, this street sure is dirty!" I'll reply, "durrwhocares?"

Add on to this the difficulty of actually getting characters to have the breadth of judgment required, you've got quite a feat. After all, these characters will have to be able to not just comment intelligently on a nice view, but also have to feel immersed in that view. Then, ten steps later, they've got to be equally immersed and judgmental about, say, a traffic jam. Or a man wearing a pink tutu.

I think it might be possible to do this to some extent, although I'm not sure how suitable it would be in any modern game. These days, all the games we play are horribly egocentric and impatient: the city doesn't exist, it's just an excuse to provide us with city-themed levels. We don't like slowing down to smell the roses because we've been trained to think in terms of advantage, and there isn't any in smelling roses.

I'm not sure that the characters' judgment clues would actually MATTER, either, since we don't often consider them to be worth thinking about. They inhabit a lesser world and are obviously not intelligent beings, especially if we save and load and replay this mission a lot. It might be necessary to radically change that in order to make this sort of thing work.


I guess that's my rambly essay. What do you think?


Olick said...

I think its a little rambly.

And I think it applies to GTA4. I was not immersed at all. I spent almost the entire game trying to get as fast as I can from point A to point B, wherein I would shoot some people, then run from the police, then shoot at them. I would have appreciated traffic jams, or random brawls/arguments.

As early as GTA3 you could see gang members fighting in-street. But I don't remember seeing that in GTA4. Occasionally I would take a break to drive around, or fight some cops, but it was still fairly goal-oriented, not environment-oriented.

To this day I wonder if its the games fault, or mine for having the wrong attitude

Craig Perko said...

I would consider it less something's fault and more a mismatch. I'm not interested in the GTA series specifically because of that: the "sandbox world" is truly a sandbox. As in, there's nothing there but sand. It's all the same.

I wouldn't say it's a bad game because of that, but I will say that I don't like it because of that.

Ellipsis said...

I think that feedback is important here. The reason players usually don't stop and smell the flowers is that they don't see the purpose of stopping to smell the flowers, while the game gives them strong incentives to progress.

It's just like having an awesome looking sword that's useful - if you have an awesome looking sword that the character can't actually use, it's not that interesting, but if it's a sword that they can make good use of then the player will look at it and appreciate it if the sword looks awesome.

I think the user needs some kind of vindication that they just spent a second observing X or Y - in a medium where almost everything offers some kind of feedback or even advantage, things that don't are perceived as filler.

Fallout 3 tries a very simple version of this by having NPCs comment on what you're looking at, especially if you're looking implies dubious intent, as in "Yes, that's locked, and yes, I can see you looking at it." The problem is that it, like everything, is repeated too often and comes up in too formulaic a manner.

Greg Tannahill said...

Completely missing the point of your post, I'd like to direct you to The Darkness, which has several hundred pieces of specially commissioned graffiti, none of which appears in more than one location in game. Very nice attention to detail, and it shows.