Tuesday, July 12, 2005

Social Games

One of my interests is in making social games. For example, my various tabletops and LARPs are usually as social as I can make them.

For computer games, social is hard. You have to use some pretty keen emergent behavior, and even then, the limitations will burn in the player's mind because he's used to reality. Most games are about things people aren't actually familiar with. There's a few exceptions: sports games, mostly. Those exceptions are within the realm of accurate simulation, at least in certain facets.

However, everyone is familiar with talking to people. And simulating it is beyond available science. This means that games naturally shy away from this - they aren't good at it, and everyone can tell. Some games try this, but end up with an interaction system which is embarrassingly bad and inhuman. As in, every game I've ever seen try it.

Perhaps another question should be posed. Rather than "how do we make social games", we might ask "why do we want to?"

There are, as far as I can tell, two reasons.

The first reason would be to allow for a massive breadth of play. If you have social interactions, then every play-through can be radically different based on who socially interacts with who in what ways. This reason I can see as valid and not needing further examination at this time.

The second is that realistic human reactions will give the player a level and depth of immersion never before realized. This will, in turn, make the player get really into the game, at least in theory. I'm not sure this would actually work. Let's look a little closer.

In most narrative-filled (and even non-narrative) games, I tend to like specific characters for their character design and personality. I'm sure most everyone is the same way, given anecdotal evidence. Which characters I like varies from which characters other people might like. For example, FFIX's Quina actually made me stop playing: he/she/it/moron was incredibly, pathetically, idiotically stupid beyond all reasonable capability for anything outside a black hole to suck quite so hard. Quina was the warp drive of stupid: capable of being more idiotic in more dimensions than the speed of stupid is supposed to allow. I'm pretty sure that very few other people felt quite as strongly about it.

Putting in some level of social malleability would have saved that game, because I could have spent the whole time Quina was with me being brutally mean. It would have been even more nifty keen if I could have said "get the hell off my team, ass-hat!"

I see this as one of the primary REAL draws of social gameplay. Allowing someone to delve deeply into what they feel are the most interesting relationships in the game, while avoiding the pieces they dislike. It's an enhanced version of "wolf pack" memology... which, looking back, I haven't explained here, so I guess I'll have to do so in the near future.

Now, when most people think of social gameplay, they think of something like, "I want to walk up to any random person I meet and be able to strike up a conversation, be their friend, etc, etc."

There's some serious blindness inherent in this idea. First, it may require some serious algorithmic innovation before we can get it to work. Second, it negatively impacts gameplay.

Imagine a game set in a college. In real life, it is fairly unlikely you ran around talking to everyone. Chances are, you had a limited circle of friends you hung out with most of the time, and then their circles, who you would occasionally meet but weren't really 'friends' with. Almost all your social interaction was constrained to these people. You grew to not only enjoy their company, but have some level of empathy with their troubles and triumphs. And that's what you want in the game, right? Empathy?

Now, think about my earlier posts on socializing and making friends. Extrapolating, it is clear what would happen in a game: the player would approach characters based entirely based on who has the highest 'stats' - whether those stats are in-game or meta-game (such as how cute you think they are). Assuming it was the normally poorly-programmed game, you would make friends without difficulty, have fun for a bit, then discard them as you became jaded with the stats which brought you to them.

Even if it was more difficult to make friends, the fact remains: without anything to cement the player into a relationship, the player will drift as soon as he has plumbed the 'depths' of these inch-deep puddles we call characters. Hell, most real people aren't any deeper.

The only real way around this problem is to very strictly limit the player's ability to find new fodder. If he's 'stuck' socializing with people, then these people will be around long enough to lodge in his mind as "real people". Of course, for this to happen, they have to behave like "real people" - making jokes, having fun, doing things, interacting. That's really, really hard to make emergent. Light years beyond simply being able to form relationships.

So it's significantly easier to add socializing to an existant plot/characters, and "simply" write in slack to allow for varied social progressions. While more difficult than a strictly linear system would be to write, it is actually DOABLE, as opposed to this INSANE idea of having an UNLIMITED number of characters.



Darius Kazemi said...

Imagine a game set in a college. In real life, it is fairly unlikely you ran around talking to everyone.

Hey, speak for yourself.

Craig Perko said...

We want a very powerful emotional connection with each person we meet, so we want our players to be Darrens, not Dariususes. :)

Craig Perko said...

Waitaminute... that's not quite right... because you might be BORED with that kind of system, and Darren-types might be OVERWHELMED with your kind of system...

Do we want to match the system to the player's normal preferences, so as to use the optimal type of attachment to draw their emotions?