I've been thinking about small, dense worlds for gameplay. Most games scale up, I've been thinking about scaling down. A good example of this approach is any game set in one (large) house, such as Maniac Mansion or Gone Home. Of course, me being completely obsessed with scifi means I'd probably think in terms of one small moon base or something.
Most of the time, in games with small spaces, the gameplay is simple puzzles. Pick up A, B, C, twiddle D, insert into E. Classic adventure game stuff. But why? Can't we do other kinds of gameplay?
It's a matter of how long the game can be. In general, you can think of the length of the game as being the density of the space times the size of the space. It's relatively easy to create more space, so most games use larger spaces. Games like Go and Chess have small, static boards and are about as dense as you can get, and still they are rarely more than a few hours long.
Some games use space densely, but still switch it out. For example, a tactical RPG where every meter matters, but you keep getting thrown into new maps. So they have both density and size, and this is reflected in their typically extraordinary length.
Anyway, adventure gameplay can be made extremely dense because it is polymorphic. That is, the same things mean different things depending on what else has happened. The kitchen has a microwave? That's not interesting at the moment, but it becomes interesting when you decide to microwave something as part of a puzzle, or need to unplug it and use the plug for something else, or carry it to the window and drop it on someone's head far below... there's a lot of things it could be used for, depending on the context. By changing contexts continuously, adventure gameplay pushes the play density way, way higher than other genres typically go.
Of course, there are many ways to change contexts, and you don't always need to be adventure gamey to do it. Katamari Damacy is a fun example of simply using size as context, and a single house ends up having many different gameplay paths depending on where you start and how large you are. They attempted to double this by adding in new modes, such as collecting hot items only, but in my opinion it didn't work very well because the whole setup was engineered with size as the primary context, and using other contexts ends up awkward.
I'm more of a Katamari sort than an adventure game sort, so I got to thinking about physical gameplay that changes context, rather than puzzle gameplay.
One aspect I like is the concept of size changing how you interact with the setting. However, rather than a destructive interaction, I started thinking about a navigational interaction. At its most basic level: you can only go through certain places if you're small enough or large enough, due to the layout. But more than that, I want "going through space" to be actual gameplay, not just a simple act of navigating. Like a free running game but upside-down.
For example, let's say there's a narrow space. If you're tiny, you can just go through fine. If you're human-sized, you've got to scoot through sideways, holding your breath. If you're in a space suit, you can't get through... but you can stick your arm through. If you're a quad-copter drone, you've got to flip vertical and float through on pure momentum. That sort of thing.
There's also the matter of gravity and rearranged rooms. The same physical space could be laid out very differently depending on which direction gravity is coming from, or if there's gravity at all. Even the difference between lots of gravity or small amounts of gravity can change how you might navigate a space: with low gravity, leaping and climbing walls is easy. In high gravity, you're stuck to the floor... but you can exert a lot more force without spinning or bouncing, so there are times you'll need that. And, of course, various kinds of obstacles might squash to the ground in high gravity, which can be important.
Back on the subject of topology, I was thinking how to make the layout result in complex gameplay. Simply gating the environment with spaces you can or can't navigate depending on your size is too basic. But I've already mentioned a path to a solution: the ability to stick your arm through.
It's not just your size which matters, but also your flexibility and reach. There are many places a human can't squeeze through, but they may be able to get their arm in, or even crawl in until their butt gets stuck or whatever comical situation you'd prefer. This gives them a certain amount of reach into the restricted space. Combined with a long stick or a bit of wire or something, this could in turn support some more gameplay. A nonhuman of roughly the same size as a human (say, a cargo bot) would not be able to do that.
I like the idea of force being useful. Even if the cargo bot had an arm capable of reaching through the slightly-too-narrow space, it is not a soft device and if it is pulled through by force, it will rip apart. Humans can squash a bit, especially if you play fast and loose with biology. This means that they might not be able to simply slide through, but if they could get their arm to the other side, or grab ahold of something, they could force themselves through. Maybe you use a robot to drag a piece of debris over so you can grab it and pull yourself through... There could be some kind of damage meter or something if you want to make it cost more than time.
But how does this turn into gameplay? I need to make this a game which is interesting, and simply moving through space arbitrarily ain't gonna cut it.
I'm thinking of a "compact roguelike" - a small space station where the point isn't to move on to the next level, but to backtrack across the same rooms while changing their conditions so you can get something done. There would be a pretty loose set of conditions so you could approach it in a variety of ways - this isn't a puzzle game with only one solution. Similarly, the passage of time is often critical, so moving confidently and sleekly would be very valuable.
Balancing your human form, space-suited form, and a variety of remote operated robots would be the key, and a big part of the early game would be finding the places where you can change into a space suit or find and pilot various kinds of robots. A lot of the rest of the early game would be about figuring out what systems can be changed in what way from what locations to change the way the space station is navigatable - gravity, pressure, heat, wind, light, doors that lock or unlock, etc.
But in terms of movement, it's still a bit lacking. I really want movement to feel explorative, not just transitionary. I don't want the player to wedge themselves through a door just to get to the other side. At least, not while counting it as important gameplay.
To get the explorative movement I want, it's important to make the player have to explore constrained spaces "dark" - that is, without a clear idea of exactly what is coming down the line. This means that doors are only valid in terms of whether they are open or closed. So rather than simply exploring from room to room through doors, it's important that we have some very different paths to use. Which means our space station is built a bit unlike most game's space stations: the player will need to spend a lot of time in air ducts, crammed between walls, above ceilings, below floors, squeezing between endless computer pillars, clutching to the underside of walkways, flailing in midair in the wind of the climate control's main pipeline...
In addition, we can rely more heavily on rubble and debris. Collapsed ceilings, desks and chairs piled up against the door - anything to make navigation more difficult while simultaneously connecting the outside and the inside of the rooms. Vents and Jeffrey's tubes would abound, as would elevator shafts and so on.
There are two keys to making this interesting. One is to insure that the paths are never completely linear: the player always has something to actually explore, not just a corridor to squeeze down. There's always a challenge of whether you can turn and squeeze through a side passage, or navigate around a structural beam or something.
The second is to insure that there's always gameplay. When you crawl into an air vent, the standard movement gameplay would be "press forward to go down the air vent". However, this is a pretty dull thing to do if much of your gameplay involves crawling down air vents.
Fortunately, the first thing rescues the second thing. To make it so that there is open exploration in tight quarters, one of the things we'll end up doing is creating a lot of awkward, angled paths off the main path. The challenge is to go down them - not something that is easy, as you know if you've ever been in that kind of situation. The human body only bends in specific ways, and in a confined space even turning over is frequently impossible.
The gameplay needs to be fast enough that you never feel annoyed about the slow movement, and the best way to do that is to make the restrictions not about speed, but about orientation and flexibility. Therefore, the controls are not simply "push forward to go forward", because we want to allow the player much more control over how they are bent and pointed.
There's a lot of question as to what kind of interface would work best. Perhaps simple WASD with QE rotation is best. Perhaps a mouse-centric interface where you click on specific surfaces. Perhaps you do some kind of split interface where WASD controls you, but if you hold control it controls your hands instead of your legs. Lots of options, but all the options have to work well when in confined spaces or in the open, when a human and when a quad-copter.
You can also add a lot of complexity to the basic design. For example, having an inventory that actually exists in space, so it changes what you can go through and you may have to take off your belt and throw it through ahead of you. Or having a button to exhale and stay exhaled, so you can squeeze through smaller spaces... but after a little while you're going to get lightheaded, so be careful not to get stuck someplace where you can't breath! Also useful if the atmosphere has been contaminated someplace.
Anyway, just thinking about it. It might be an interesting game.