Monday, June 8, 2009

Game Theory

There was a long discussion a while back at the end of the comments section of ActionButton.Net's review of Metal Gear Solid 4, which was also the penultimate entry in their list of Best Games of All Time (but not on the list itself; Tim Rogers likes to do things like that). After a couple hundred posts it degraded/evolved into a broad debate on games as an art form. Feeling compelled to add my two cents (as I always do), I typed up a few posts that came out pretty good. About three people read the whole thing, but I think it's worth reproducing here for posterity - and so I don't have to repeat myself when asked for an opinion. So, for anyone who cares, here's what I think about the medium of games in general:

Games cannot have an integrated story, because they are not books or movies; they are really Possibility Space Architecture. Well, Tetris & co. are closer to Sculpture I guess, but we’re talking about MGS4 and other Big Games here.

Game designers are therefore architects and not writers. Most modern big-budget games choose to ignore this, and turn themselves into the equivalent of a ride. Really, how many times have you heard a game described as a ‘theme park ride’ (or maybe a ‘roller coaster’)? People intuitively get this. The metaphors line up nicely, I think; in this case, non-interactive cutscenes would be like a big video screen that you have to pass by on your way to the next room/level (’level’ is also a fitting term). They provide context.

You might think of better ways to provide context to your exploration of a building than video screens at every ten feet telling you where you are; maybe that’s fine by you. It’s the same with cutscenes. But the result is the same: you get context, not story. Someone way up in the thread said something to the effect of ‘I don’t care about the details, they’re demons from Hell invading Earth and I’m a badass Space Marine’. That’s context, plain and simple.

So, most games today are designed in such a way that the player’s narrative combined with the context results in a retroactive story. They tend to do it by straitjacketing the player into an overly detailed context (masquerading as ’story’), where we just have to play our little predetermined part, and everything will be OK. Others try for some sort of compromise. Hell if I know what the answer is. But I believe the parts of the puzzle are pretty clearly laid out.

Also, for bonus points, let’s get into the Games as Art debate by stating that of course they can be because they can (attempt to) use possibility spaces as tools to convey an emotional state or frame of mind. What the fuck more can you ask of an art form? Authorial control? I don’t recall Frank Lloyd Wright crying ’cause “people don’t pass through my houses in the right way”! You buncha fuckin’ linear media pussies.

That being said, I happen to think stories can be awesome in creating context for games; they’re simply not necessary (like I said, Tetris). [Regardless of Carmack,] porn with story really is better - story makes shit better, because we care.

[Poster CubaLibre here noted that I'm separating “context” from “story” where he'd consider them more as two different types of stories; also he mused on whether/how the act of playing influences the retroactive story. He used improv theater as a 'story-making' example where the context and story are closely entwined.

GilbertSmith opined that better AI, allowing for complex drama during gameplay, would help here while still fitting the original definition.]

Yeah, I don’t want to get all semantic about it, that’s just the word I use in my head; it’s as good as any other, I think. What I described as ’story’ could probably also be called ‘plot’. But as long as you understand what I’m talking about.

As far as influence on playing goes, on the simple end of the scale people will show off with cool but unnecessary moves while playing any game, even if they’re alone, mostly I think to make the experience cool - which is just another term for ‘improving the retroactive story’.

On the complex end of the scale, what I think you’re asking is ‘will the player feel pressured to play a certain way, so as to improve the story?’ And I believe the answer is yes, in fact a number of game design techniques rely on this pressure in order to subtly direct the player.

Like the chess example [where a number of posters considered what imparting predesigned narrative into chess would be like], it’s one of those tradeoffs where you diminish the game (because certain parts of the possibility space are now preferred, so you’ve lost the freedom/equality of choice of the “ideal” game) in order to improve the experience (the designer will obviously choose for maximum impact). On the one hand, ‘diminish’ sounds like it’s a bad thing, but I think it’s really okay, because most modern videogames (again, unlike Tetris, Quake 3, or Soul Calibur) really aren’t games at all, like you guys noted earlier, but interactive experiences that use specialized game fragments as artistic tools. Pac-Man would be harmed by a designer preference, but Shadow of the Colossus isn’t, because it’s trying to make a point.

Someone upthread already mentioned that, as game mechanics get more complex, people start asking for reasons, so you need context to keep players engaged. That’s a pretty good way of putting it, but I would add that context is more important in directed experiences. When the designer directs your choices, you feel limited and you want to know why. When there are no limits, the game is easier to accept on its own. Soul Calibur, like I said, is a true ‘game’ (at least in its classic VS mode), even though it’s pretty damn complex. Does anyone give a shit why Sophitia is fighting some French dude? It’s just a game. They’re fighting because you want to fight.

The improv thing I’m not sure about. The way I deconstruct it, there is no ‘context’, because improv is basically a pure game, just like a Quake 3 deathmatch (Soul Calibur, etc). The only thing you get out of it is what the players bring in; there is no ‘designer influence’, so there’s no need for context. The story is not the rules - the rules shape the story. Think of the story like the blocks in Tetris: not every block goes together, so you have to align them in accordance with a few rules (the main one being ‘the blocks will never stop falling’ which is actually pretty close to that main improv rule that you must always accept what you’re given and go with the flow). It’s only confusing because the raw material you work with is so similar to what’s usually a part of the context. It’s not so far from Guitar Hero, really, which is also a game that gives you something usually reserved for the context and makes you play through it. And just like there’s no background music in GH, there’s no story in improv.

Also, GilbertSmith, that’s kind of where I went with the improv analysis, where you treat your co-players as objects to work with in the game.

And leaving aside true AI (which would really be just another player), a designer-specified pseudo-AI, the kind modern games are going for, really is sort of a tool for dynamic rearrangement of the possibility space, with the goal of maximum drama. It’s a good idea, but I think people often go too far with it, trying for realistic human characters which are also game elements, which is so fucking hard I don’t think I’ve seen it work yet. But make a big, dark horse, make him come to the player when called, and not fall off cliffs when you ride him, and there you go, empathy, works like a charm. So obviously it’s doable. People truly want to love game characters. The designers just have to stand out of their own way, and stop with the rude awakenings, like the voice-acted well-spoken well-articulated mysterious sexy stranger that repeats the same three fucking lines whenever you talk to him.


Well, actually, that last part may have been a bit messy. When I say designer-specified pseudo-AI, I mean that all NPCs in any game are really just representatives of the designer. When the sergeant screams “MOVE!!!” it’s really the designer who wants you to run. Real AI would just turn them into real people, and if that’s what you want isn’t it better (and less ethically dubious) to ask a friend to play with you?

Progress in games AI is simply trying to make the NPCs better and more faithful stand-ins for the designer. You’ll notice here that we already have games that do this perfectly, because the designers had the good idea of actually getting one of the players to co-design the game, and play all the NPCs in each session. That’s why CRPGs aren’t like the real thing; you lose the human factor, and computer DMs aren’t yet up to snuff. Maybe they can never be up to snuff, and someone should finally design a good multiplayer CRPG where one player is the DM (Neverwinter Nights was merely close).

The difference between D&D and improv, incidentally, is that in D&D one of the players, the DM, has an agenda beyond the situation at hand. That makes him a designer, and moves the game closer to an experience.

[CubaLibre then considered evocation: stories are designed to elicit things from readers - do games have a similar goal? Do they elicit different things than stories? If they overlap any, could a story-game be made in that area that respects both forms? In other words, why do people play games?

He disagreed about improv; watching the actors create within (or around) the rules is partly a joy in itself, regardless of the result - the ephemereal nature of it adds to the experience. The idea is that the audience is moved because the presented story is made temporary by the game-qualities of improv. He noted how he was focusing on the psychology of the audience.]

It does pretty much come down to ‘why do people play games’ - can we solve that one some other time? :-)

I will note there’s plenty of overlap between games and stories (all media overlap), they just have different strengths and weaknesses. Who was it that said novels delve into characters, while movies watch them interact? The particular strength of video games is something I can’t quite name (maybe there isn’t a name yet), but the closest concept is maybe ’sense of place’. I never really cared about Alucard, but I vividly recall the castle from Symphony of the Night. Except the ‘place’ in ’sense of place’ doesn’t have to be a physical location, so it’s not exactly right… I’m still working on that.

But that’s not important here. I’m convinced that a story-game properly placed in the overlap would be fascinating, and also probably awesome, but the evocation you’re talking about is a matter of player perception, while my concepts are mostly about the way games are created, and while I’d love to play what you’re describing, I haven’t the faintest idea how to make it.
And yeah, I was mostly thinking about improv comedy when I wrote that, but the point still stands. Improv comedy tends to make like Kojima and go all meta ‘n’ shit, while drama is like a good shmup, and works within the rules to perfect the result. And you’re right, I’m focusing on the actors, because after all the actors are the players; the audience is… well… the audience. But here’s what’s important, to me:

The audience is moved by the story because it’s a good story (its temporary nature is a cherry-on-top - if it were a bad story, the fact it was made up on the spot couldn’t save it). It’s a good story because the actors made it, and they’re self-aware, intelligent, emotional, social entities, which means they possess the massively complex mental toolset required to work those dramatic Tetris blocks into an impressive whole. It is a game, it is played, and the audience enjoys the result (you know Electroplankton? that thingy for the Nintendo DS where you, like, make microorganisms and they make music? show a particularly well-done blob to your friend and they might be moved by it, but it doesn’t change the fact that making it was a game to you). The thing that makes improv stand out among games is also why it has such an emotional impact - you will never (well, at least not for a very long time) be able to play it on a computer, because it manipulates emotions directly, and computers just don’t possess that toolset and understanding of human nature. Hell, they can’t even talk properly yet.

(That’s also why improv is a dead end to me, as far as video games are concerned. It’s a game, and it can be classified and all that, but the fact remains that it’s operating on a whole different level, and as much as we crave all that wonderful emotional resonance improv gets, we just don’t have the design tools to climb up there. We’ll have to think of a different way.)

No comments:

Post a Comment