Thursday, July 12, 2007

Expressive Rules in Social Simulations

In Halo, one of the biggest ways you determine what kind of player you are is by what kind of weapons you use, since you can only carry two. Your fire patterns and method of movement are closely related to the weapons you carry.

The choice of weapons is not a complex, high-grain choice with a thousand details. If you want the weapon, you drop the old one and pick up the new one. No "adverbs", no place on a slider, no skill challenge.

Furthermore, you can see the choice transparently long before you make it. In fact, you usually don't bother making the choice at all unless you see it is a choice worth making. Furtherurthermore, you can always come back and make the choice again, or reverse your choice.

There's no stress, no tension, no confusion, no cleverness, no secrets.

But it is one of the most important parts of the game, letting a player choose how they will play. It allows the player to express themselves in the simplest way: picking their favorite of two choices.

The rest of the game is like that, too: if you move right, you can almost always undo it by moving left. If you fire, you can usually adjust your aim and fire again. But firing and strafing and turning and walking are all excruciatingly simple choices. There is no complexity to them individually. It is only when you make a large number of them that the personality of the player begins to come through. And that's only because each choice slightly changes the context of the game - including relative to other kinds of choices.

Most of the "social simulation" systems out there - storyworlds and so on and so forth - rely almost exclusively on the idea that a few very precise choices is the best approach. "If we can give the player enough precision..." they guess, "if we give the player the ability to express himself exactly..."

That's like playing Pac-Man by letting players put the Pacmeister anywhere on the board whenever they want. "It'll give the players the freedom to tell the story of Pac-Man in any way they find interesting..." It will also be boring and pointless and completely futile because it short circuits the feedback loops. What little play remains may contain feedback loops, but only the most simplistic, useless kind.

Instead, a few different kinds of interaction should be given which allow the player to make a large number of small choices to subtly change the context of the world. Instead of the player performing a big action like giving roses to Sue, the player would perform a larger number of small actions. Perhaps managing poise, personal space, facial expression, and gaze target. Obviously, these choices would have to be more clearly represented, so the player can really get a feel for them.

But the same four basic kinds of action could also be used in bar-room standoffs, talking to the cops, preaching to a room full of southern Baptists, playing with children - any kind of social action you can imagine. Instead of fifty different kinds of "social action", have four smaller kinds of interaction and let them "load up" on any equipment they might feel they need. (Roses, spiky collars, brightly colored plush toys...) This would be the equivalent of picking a weapon in Halo: a choice made rarely that changes the way you "move" in your more common interactions.

I'm not saying that those four are the best four, or that this approach is the best approach. I'm saying that a few very exacting choices are vastly inferior to a large number of tiny choices.


Patrick said...

I think while your analysis is on point and your proposed paradigmn for social play is one I want to explore with R&D, because I think its the most profitable, your comparison implies that goal-orientation is always superior. The context that underlies your Halo and Pac-man examples is one of an explicit, goal, an openly but implicit goal, or an even more open range of aesthetic goals that the user can bring into their experience, are not accounted for in this model.

Indeed, you're reflecting my principle of conservation of agency. Halo has lots of local agency, but its global agency is constrainted to hitting point A, B, C, D, and so on. Pac-man has the same goal, where the sequential alphabet consistst of all dots on the screen.

Conversely, and you might as well come out and be explicit that you're disparraging Storytron, you can greatly exponentiate global agency by constraining local agency. I think that appeals to a more niche audience, but its not nessecarily inferior. The fact of the matter is; you don't have any data to support this assertion of superiority - outside of table-top games, which according to you is outside the scope of the agency ratio principle.

So alls I'm saying is, let the data come in, then judge.

Haha, I just remembered our bet. Two games off Manifesto, right?

Craig Perko said...

Sure, two games.

Look, I'm disparaging Storytron because it happens to be the biggest name. But ALL the social/story simulators make this same mistake. I'm disparaging a whole field, with Storytron simply being the most obvious name.

Goals are part of play. The goals can emerge from the player just as easily as the game: nothing in this says otherwise. Using this system, nothing would stop you from trying to give roses to a biker gang or intimidate poor little Suzy out of her lollipop. It's just a system. The world is separate, the goals are separate.

It has nothing to do with global vs local agency. It's easy to do a game like this giving more global agency. For example, Civilization uses this same kind of rule set but gives fairly strong global agency. Creating things in a 3D modeling tool follows these rules but gives very strong global agency.

There are no "goals" inherent with 3D modeling software, but the tight feedback loops allow a player to follow his own goal. This is the same.

Patrick said...

Ok, lets just say I'll hire you when I want to/can R&D something like this, and leave it at that.

The bet's still on though. :P