Before I get into this let’s be clear: I’m not trying to define “sandbox”. I’m trying to understand it. I might want to use the term to market a game and I don’t want to lead anyone astray. Unfortunately I’m discovering a lot of variation in acceptable definitions.
Over on reddit someone brought up an interesting point that seems to demand another axis on my graph for sandboxes. I don’t want another axis because 2D graphs are very tidy and easy to make but if there are three variables then so be it. Let’s hope we only need three so I don’t have to draw a hypercube.
So over there the point was made that one factor in a sandbox is a lack of player agency. Not character, mind you, but player. That is, the player’s impact on the world is constrained to the actions of their character: they have no narrative authority outside describing their actions. They don’t get to name cities, declare species friendly, or announce that they found a speargun in the tool shed. The commenter places “sandbox” in direct opposition to “story game” over this exact issue. Maybe I’ll try and find the boundaries of “story game” another time.
This further seems to imply ownership: in a sandbox game the ref owns the world and its evolution. The players try to make a mark on it using their characters but they don’t have any authority over it. And in a sense they don’t own their characters either — since they have no narrative authority, they can’t raise new background information in the middle of play.
This makes a sandbox (or let’s call this a hardcore sandbox) a very small place, to my mind.
No ref, no matter how dedicated, can create a world with the same detail as the real world.
Players will read and retain only a fraction of what the ref creates and offers them to absorb.
Consequently, players have a fraction of a fraction of the knowledge about the world of their characters that a real person would. Removing narrative authority therefore shrinks the world: we could have the illusion that the world as described by the ref is really detailed, more detailed than the bits they wrote down, but only if the players can say things (through the characters’ mouths) that are true about the world. And the characters must know vastly more about the world than the player. More even than the ref — they grew up there. They went to school there. They spent 20, 30, 50 years there experiencing it inside the fiction. Characters must have an enormous amount of world information.
But the player can’t. It has to be simulated or the world shrinks to the notes of the ref. It is exposed as tiny.
There are a couple of ways to let it grow though.
One is to have the players ask the ref questions about the world. “Do I know more about this from my upbringing in this region?” or even “Could it be the case that this is true?” And the ref can add more detail to their world based on responses. This preserves the ref’s ownership.
But that feels, to me anyway, positively draconian. It places us in a position of almost worshipping the ref’s vision of the world.
Now for a long time that is exactly how I played. The ref was God and the players were peons within it, begging for information scraps. I enjoyed that a lot. It was a little suspicious how much I enjoyed that. So my preference now is the alternative: players can freely declare facts that fall into the scope of their characters’ experience. There is a sense in which this implies more trust — not more trust around the table but more trust from the ref for the player input. Trust is always complicated and so is power and this solution upsets some fairly traditional ideas about how that should be distributed (hardcore sandbox: the players trust the ref; the ref need not (and maybe doesn’t) trust the players).
But I get it. I’m starting to get a feel for what the sandbox is. The ref is running a world simulation and the players are interacting with it only through the interface of their characters. There is a comforting way in which this reduces the social interaction of play–I mean, everyone is of course socializing but it is not part of play. Trust and power hierarchies are strictly enforced. The game (not the play — play is fluid, jovial, human, questions, answers, jokes, sidebars, arguments, secret caresses under the table) has a rigid structure, like a video game: the characters are the avatars and the ref is the computer software. Boundaries are not crossed. That’s both appealing and repellant for me. It has a powerful structure that does not invite a lot of argument, which is slick.
If you spell it out.
But if you spell it out you kind of want to kick against it, to refuse being dominated. You might find you want to at least be allowed to (if not actually) push the world around as a player. You might be inclined to believe that your creative input has as much or more value than the ref’s in some contexts. You might then be invited to run your own sandbox, I guess.
None of this is spelled out in the Soft Horizon games and it probably should be. I expect the players to volunteer fiction outside the scope of their characters. It is not a hardcore sandbox. It’s a softcore sandbox, I guess?