The term Microcosm in relation to game design has often been thrown around, specifically relating to how designers are designing a tiny little microcosm. This is garbage. Designers are designing a macrocosm. No matter how small they think it is, the largest boundary they provide a player is the macrocosm. Within the macrocosms each little system is its own microcosms (and macrocosms at the same time), and each system within such systems are a microcosms. The boundary set for the macrocosms should have an infinite amount of play within.
Play as we know involves a boundary and some rules. Play is essential for learning, and learning is essential for basic functionality of games. So we can really look quite deeply into this play concept, while getting horribly lost.
When one thinks of a boundary, without doing any surveys of any kind of research, one might be inclined to picture of a fence, or at least a square, perhaps a rectangular prism. Which as a result are the two major boundaries found in modern video games. Not really spheres, certainly nothing abstract like gravity. But boundaries are really important when it comes to play. We want players to learn the basics of X, then we want players to understand X, and finally we want players to master X, probably so we can show them Y. This can be seen a lot in games, but essentially it's slowly extending the boundary, allowing the player to experiment more, play with different combinations and different rules.
It would be best to think of a boundary in at least three dimensions, if you have the mental prowess to think of it in four dimensions, proceed (on your way to oblivion). The point here is depth. Does your world have depth. And please don't think for a second I am talking about plot, visual or aural indicators. Your world has to have working mechanical depth, and the player can never know how far that rabbit hole goes. Because it should be endless, if they can see the bottom, you need to re-evaluate.
Games are usually set inside a box, most games are just a series of boxes with stuff in them. And this is a physical constraint. But a designer should think of this on a higher plane, imagine if the box had no definable start or end. There is now unlimited potential in the box. This sounds like a loose concept but it's really not. It's allowing for the possibility that every time the player plays in the box the end result is different. And naturally most of the time things will be different. But it's the learning cycle that should be different, can the player learn something new every time they play in the box.
Filling a box of limitless depth sounds like a challenge. But essentially what you want to accomplish is provide the player with the opportunity for limitless learning given a specific set of tools within a boundary of limitless proportions. That instantly sounds like an oxymoron and impossible. Picture it like this, we have a boundary with dimensions of X Y and Z. This space is a vacuum, and one could say fill it with matter with air and it will expand to fill the space. But this would be come extremely thin the bigger the boundary gets. Simply just adding more air won't suffice. Instead we want to fill it with gravity. Things the player can be naturally pulled towards, "gravitate" towards if you will. This gravitational force is provided by, strategies, mechanics, behaviours, systems all the microcosm you're putting into your macrocosm allow the space to be filled. Now when you extend your boundary to 2X, 2Y and 2Z the systems and mechanics also need to naturally grow and fill the space, without thinning out or simply just adding more. Extensions later on simply won't do, as previously stated just adding more air will not suffice. they need to be in the base systems from the beginning and they need be nurtured as the boundary expands. Game mechanics are organic systems.
Now we have the idea of growth, two concepts that go hand in hand depth and growth. We want deeper, longer lasting game mechanics. However we cannot simply throw the player into the deep end and ask them to swim. Most will drown, they won't learn anything and become bored. When players aren't learning and they have become bored, they put down their controllers. You've failed to entertain their mind. So we need to grow our macrocosm with the player. Forever expanding the boundary and allowing the mechanics to grow together to take advantage of the space, and at every stage we need to be allowing the player to learn this new space
Remember that players are just a part of the macrocosm you’re trying build they are a very important microcosm. They don’t sit outside of it. Incorporate them.
That's part one done.