Monday, December 17, 2007

What Will They Say?

Totally ivory tower...

Sometimes there's nothing better than asking a ridiculous number of "whys".

Take player-generated content. A lot of people want their games to use player-generated content. Why? Because it's basically a renewable resource. Essentially it's free developers, even if much of the content is really poor.

But let's turn that "why" around. Why do players want to generate content?

Because it's fun? Okay, why is it fun?

It's clearly not the UI. Even the worst UI in the world (SecondLife) gets thousands of players making some extremely high quality content. The actual process of making content is about as much fun as eating tin foil. This is especially true when someone spends fifty hours polishing.

So why is it fun?

A lot of people would answer in a lot of different ways. I don't think there is much of a consensus on the matter. A common answer is that people want to express themselves - it's part of being human.

Personally, I think that's not quite it. I don't think people want to express themselves, I think they want to express what's in their head. There is a difference, although it might seem semantic.

For example, Oblivion. Oblivion lets you build a character. Actually, that's the best part of the game: the character generation. It's definitely player-generated content, although it's not really content that gets shared with other players.

But this is not self expression. When you build a character, you're not building yourself. Although I'm sure some people build characters that reflect parts of themselves, I know a lot of people build characters that they think will result in interesting gameplay. Interesting to the player, rather than from the player.

When I build a female drow alchemist, it's not because I have an inner female drow alchemist, it's because I have expectations of the game. I expect the game will play in a specific kind of way, and being female, drow, and alchemically talented will make my play experience more interesting than, say, being a dwarven warrior. If I replay the game, I'll probably create a different character because that particular interesting experience isn't terribly interesting any more.

You could say that I'm not expressing myself. Instead, what I am doing is starting a dialog with the game. After judging what I think the game will be like, I say "what if I do the alchemist thing?" Then the game replies, "well, those plants you ignored before..." and the game play changes.

It's not an expression, it's a dialog.

Obviously, this isn't quite as unchained as most kinds of player content. In a lot of games, you can build things that are considerably more freeform. SecondLife, for example, lets you build just about anything.

The biggest difference here is that the game itself is not always the other side of the conversation: frequently, other players are.

If you look through SecondLife, you'll see what are fundamentally three kinds of content.

The first is the same kind you would get from Oblivion: the player is having a conversation with the game engine. People trying to do cool things with prims, or figure out how to make a glowy thing that chases you around, or building that first building, or even building a self-propagating ecosystem.

The second is social content. This is a piece of content that acts as a message to other players. A billboard, a sex act, a castle: this is a statement of some kind that is trying to convey something to other people. This is the sort of self expression that artists tend to talk about... but it's basically obsolete, noninteractive. It doesn't take advantage of the media.

The third kind is also social content, but unlike the second kind, this content is actually a conversation. It isn't simply a message in a bottle, but something that invites replies and further conversation upon itself.

Typical examples of this last type are message boards, where a thread of conversation exists, fundamentally, as itself. Although it is in the media of the message board, it has a membrane around it so that the context remains clear and the conversation remains on topic.

In SecondLife, you get a few examples of another kind of conversation: morphing content. A lot of content gets produced, and then another player will produce a similar piece of content, and so on and so forth. This "evolutionary dialog" is really very interesting, as it's not held between specific people or in an existing language.


I'm not interested in this "message in a bottle" stuff. I can do that anywhere. If I want to send a message in a bottle, I'll draw a picture or write a blog post. Fundamentally, games are interactive, so the content and the conversations should also be interactive.

I think that it would be interesting to focus a game on the kind of conversation that can't be held in a language. Less a forum, more conversing with content.

How would you do that? How would you create a game that helped people talk in such a way?

Well, first, it's clear that you need to let people make content.

Second, that content needs to exist in the game. This means that people have to be able to share content in the game world, not just between friends. A big part of this kind of conversation is happenstance: people who stumble into your line of dialog.

Third, the game has to be worth playing, and that means the content has to change how it is played to some extent. Personally, I would suggest having more of an inherent game than SecondLife to give the content some reason to exist.

Fourth and finally, it has to be very easy to both comprehend and reply to a "message".

That last part is about language. Your conversations are limited by the language you talk in and how well you can express yourself in the language. It would be difficult to talk to a foreigner with a bad accent about the nuances of southern politics. It would be impossible to do it in Klingon, even if you were both fluent.

Our non-spoken language is the same way. What it can express will limit our player's topics of conversation, and how easy it is to understand will limit how nuanced the conversations will be.

The "language" of Oblivion is an example of one which is both limited and difficult to understand. If you passed a character to your friends and said, "this character is cool!", it would probably take them hours of play to understand the nuances of why the character is interesting. Similarly, all you can really talk about is navigating the Oblivion world.

Of course, then there's mods. Players add a wide variety of content to Oblivion, and in this content you clearly see a deeper language. Much of the language is still based around navigating Oblivion's world: mods that change how fast you level up, or what enemies there are, encumbrance limits, and so forth. These allow you to rephrase the game's challenge, giving a different focus.

It would seem like a limited language, but while it's limited in breadth, it's very deep, very nuanced. Whether you tackle a world entirely populated by enemies of your level or whether your world is more realistically populated with enemies of every level changes the game entirely and makes a surprisingly coherent statement about independence and pandering that is difficult to express in English.

Of course, it is a limited language, and there are therefore a fair number of mods that seek to talk on different subjects, including a large number of fashion mods... but these are all very limited, because the language of the game only supports them on accident.

Our language in our theoretical new game needs to have a wider language than that, while simultaneously allowing for dialog using it. (Mods aren't dialog.) And, ideally, it would be faster to understand: you wouldn't have to play for ten hours to understand the message...

A fundamental problem with this is that players don't usually have a whole lot to say. If you look at SecondLife, you'll find that the majority of content is painfully bland. Eighty thousand boxy houses and eight billion pairs of underwear, all of which say (or fail to say) basically the same things.

I would argue that they don't have anything to say because the language of SecondLife is crippled: it is reduced to what you bring with you. It is only fairly recently that it has begun to develop strong things to say on the subject of privacy, economy, ownership, and social well-being... and even now, the majority of the conversations on those subjects are held in English, outside of SecondLife. Seems odd that when a language develops opinions, you talk about those opinions in a different language. But when a language is so difficult to speak, it's not suitable for dialog.

Perhaps... perhaps our game should include a method of creating languages? Creating contexts in which things can be discussed? Fast, slow, wide, nuanced... on any conceivable topic?

This is a hard question. I'm tempted to say that the best design would be a deeply entangled set of games and content designed by players, often using other games and content designed by players.

And, no, I don't mean Metaplace. I mean a contiguous world...

Hrm. Any opinions?

I can't believe you read all of that.


Olick said...

This is gonna take a while to entirely digest...

Firstoff: content sorts. I see where you're coming from, but I think that some content can come from different sources that you mention. like someone who is experimenting, while creating something they want to have (rather than make).

I think that the three sorts of content you mention are very.. universal as well. I see the mentioned three sorts of content as: the techincal know how to build content, what you want to make with the content, and what others do with the content. Learning how to paint; putting down a message in your painting, or painting a subject you want to paint; and then having people see your painting in a gallery, and talk about it, and all the possible years of influence your painting may have on other artists.

In games this third social content is very strong because the content language is DESIGNED to be traded, used, and otherwise interacted with, whereas paintings are.. you look at them. And put them in expensive cases with special lighting so that the colors won't fade.

Olick said...

Um. Wow I should have checked that first paragraph better. Especially that last sentance. What I mean is, the content comes from someone simultaneously experimenting while creating something that they desire to create for whatever reason they have.

Craig Perko said...

While the first kind of content can be thought of as gaining technical knowledge, I believe that, instead, we should think of it as communicating.

It's not "merely" technical knowledge, because as you learn the knowledge, you learn what designer was saying. And you learn the nuances of the statement.

In something like SecondLife, it's merely technical because SecondLife's language has so little inherent context. I think that's the real problem, and I'd like to avoid it.

But... on another note, I don't think that the third type is very common in games. I don't think it's very common anywhere. A jet or a house or a piece of underwear isn't a dialog, no matter how many people buy and instantiate duplicates. It's only a dialog when someone else is inspired to make another piece of content "from" the first piece of content, and that new content is shared in a similar way with a similar context.

Olick said...

I guess I don't entirely understand what you mean by learning what the designer was saying. What the designer's saying with.. the method of creation? Or saying with the system its going to work in?

Maybe I'm just drawing a parallel between the game world and the real world, wheras the real world doesn't have someone creating a coherent design or statement for you to learn.. um.. unless you believe in a God as the Designer, or that the act of creating art can help you understand the universe.

And referring to the art that inspires other art, I think that does happen all the time. Take music for example, every band today didn't just come out of nowhere, it all descended from the music that came before it. The band heard all sorts of music as they were growing up, and they wanted to make more.

However in the media where it IS common (music. painting. writing. it is a long-term process. Even the shortest response time can be a month or two, and long term responses can take entire lifetimes to mature. And even then one single piece of content will combine disparate interests, so tracing where the conversation began and ended, if we are using the conversation metaphor still, becomes a complex discussion in of itself.

In this context (that may well be me taking an idea and running in a direction you didn't anticipate) the third type of content isn't content at all, but sort of.. a grand recursive loop for generating more content.

Craig Perko said...

Every system has something to say, even systems that were created at random. What they have to say is usually not very interesting if you try to express it in English: a Japanese RPG has things to say about exploration, danger, and about fundamentally assuming you are a good guy. These exist in even the worst designed RPG.

Some games practically shout. Katamari, for example, expresses something very loudly. But I can't really put it in English at all. Something about childhood and dreams, I think.

Of course, real systems have things to say, too. You can learn a lot from rainforests and the web of corrupt politicians we have ruling over us.

I'm not sure it's possible for a system to exist and not have something to say... it's not a matter of even creating art. It's a matter of interacting with the system freely.

Sometimes, though, what a system wants to say is drowned under a million tons of noise, like in SecondLife.

As to bands, yes, I think that music is probably a primary example of these evolutionary conversations, although science, movies, and genres are other examples.

But these languages are slow (as you said) and completely one-way to the majority of people. I'd like to make a more transparent system, one that everyone can talk in.

The end result may be a loop that generates more content, but it's not a random loop. The loop is a contained conversation, an exploration of some kind of concept.