Monday, July 29, 2013


So, recently my big interest has been with stages - by which I mean Kerbal-style stages, where you discard each stage in turn as the next phase of your mission progresses. So, not stages as in "levels", but stages as in "constructs which perform a task and then are discarded".

Well, I've done a lot of testing and some prototyping and I've come up with some interesting discoveries about the concept that maybe you'd like to listen to. Well, I hope so, since that's the rest of this essay.

The first thing I tried was a simple "cargo bay" system. So you could design a vehicle and land it somewhere, at which point you would simply build a base out of the components you brought with you. Concrete, inflatable habs, prefab rooms, solar panels, just stick 'em together however you want.

In terms of resource constraints, it's actually pretty interesting. But in terms of the vehicular gameplay, it's actually really dull, because there's no interaction between the later stages and the earlier stages. You're literally just shipping a box full of stuff. If you want a different kind of base, you just ship a box that has different stuff in it... but the vehicle doesn't care. It's just a box.

Moreover, if you are going for deeply nested stages (3+ stages), you end up with a string of boxes, and you use each box to build a transport section for the next box... and again, nobody cares whats in the box for the next stage. No interaction.

I was thinking about ways to make the stages more interesting. There are, as far as I can tell, two ways to make the stages more interactive. One: make the components required by the next stage affect this stage. Two: recycle components used by this stage for the next stage.

For example, let's say you're building a space station. Instead of discarding the final rocket stage, why not use it? Send some crew in to scrub it down and weld in some room partitions and, viola, you have a nice, spacious science module! Yes, unrealistic, but we're talking about how to make stages more interesting.

In the other direction: you can't simply assemble a moon base out of component buildings, because they all need to be connected to the same framework of power, life support, and walkable tunnels. So you pack in all the inflatable habs and solar panels you want to use, yeah, but you also have to take a massive "+" shaped framework along. How do you make it work out? Maybe you make it fold into a kind of stool-shaped shape and attach rockets to the "legs". When it comes time to deorbit and land on the moon, you unkink those legs and land politely in the proper formation. Then you can attach the habs and solar panels and stuff to it. Maybe some work is needed on the joints to make it all properly connect up after all that folding and unfolding.

And, hey, yeah, you ditched that deorbit rocket on the way down, but it's good metal that you could use. Go out and find where it landed, drag it back to base. You can always use another large, heavyweight building, even if it requires weeks of astronaut-labor to clean it up and put in doors and stuff.

Basically, my thought is that it's really difficult to create strong, complex structures off-planet. So while a lot of the stuff you'll want to use is basically just packed up in a box and shipped around, the big core elements that hold the stage together need to be carted around more or less intact, aside from some minimal joints and reassembly requirements.

The joy of this approach lies in the N-stage situation, rather than the 2-stage situation.

Right now, we're talking rockets. Let's switch over to a different kind of game world.

This game world is on the sea, and the primary objective is to populate the various islands and set up subaquatic bases. So our major means of transport are submarines, ships, and aircraft. To push for as many medium changes as possible, there are teleport rings hidden in various places at the bottom of the ocean. When you go through, you appear in the air very far away. Of course, you can also go into the airside teleport ring and come out at the bottom of the ocean.

As such, we'll spend a lot of the early game just using ships and populating islands, with light aircraft for scouting. But as we get further along, we start to need to transition. A ship that carries a submarine. A submarine that carries a plane. A plane that carries a ship. Well, maybe not a plane, but an air-lift vehicle of some variety.

Each stage is disposable, but there's a lot of overlap. Unlike a rocket, fuel is not an overwhelming primary resource. Certainly fuel is important, but the medium switch is the primary mechanic here, and therefore the concern is all the pieces you needed to make the old vehicle work are now working against you in the new vehicle. That big hull you used to offset water and carry a heavy load is now either annoyingly heavy (for lifting into the sky) or continues to displace water annoyingly (when attempting to sink). The jet engines you flew with are now submersed, and a massive drag...

So there would be a lot of fun details to work with. Do you discard the wings when you suddenly find yourself underwater, or do you seal off those engines and open up the sub engines next to them, or maybe there's a whole upper stage containing wings and balloons that gets discarded. So... do you separate into sections? Seal and deal? Break off flanges? How much can you reuse? Do you deal with a crappy middle stage because you need a component from the first stage for the third stage?

Of course, this is all made more interesting by the ability of the crew to modify the vehicle (and resulting facility) with whatever you have in cargo. Maybe you go through an air portal and pop in under the sea... but instead of jettisoning the jet engines, you send a crew out to detach and stow them, and then attach turbines to that spot instead. Is your cargo bay arranged to allow that? Is the undersea condition okay for that kind of activity?

These are the thoughts I've been having. Stages.

I've also been thinking of stages in other kinds of games. For example, in a Shadowrun game, maybe you have several "stages" to your op, featuring decking, rigging, magical scouting and perimeter-establishing, crashing through the front gates in giant trucks, then breaking into the perimeter using an intrusion team so you can do a bit more decking and spirit-calling.

Instead of thinking of a Shadowrun run as a combat mission where you shoot everything, you could think about it in terms of constructing the mission out of stages. And just like with Kerbal, the stages can be built by you out of the components and tasks you think will help best with your mission.

And, like I've done, you can have complex relations between stages. Maybe the initial rigging you used to spy on the enemy and transport your heavy server farm to the location isn't simply discarded! Maybe it has a server-farm retrieval stage later on, and perhaps even supports a later stage by repurposing the spy drones. Perhaps even on the fly, as things go south.

Doesn't that sound interesting?


Ellipsis said...

Even without very much complexity, you can make stages interesting by early choices set up later choices. An example of this I recently ran into is in the mobile game Puzzle Forge. It's basically Triple Town (put down three of item 1, and they turn into one item 2, and three of those make one item 3, etc.), but with one major difference: instead of staging up endlessly at some point you "cash in" on what you've built by turning the components you've made into a finished product (a weapon, since you're a blacksmith) which you sell. The longer you spend staging up your components, the more valuable your finished product will be (and some picky buyers will only accept high-quality products), but it takes more and more room to build fancy products, and if you run out of room, you'll end up spending a huge amount of money buying items that let you clear out space.

The thing is, once you've created a higher-stage component, it's only useful to you if you a second one (so you can match it into a finished product). That means that you have to make strategic decisions about how much staging up you can do when you first start on a project - and you might still be wrong. Unfortunately, the game kind of devolves after that and becomes easy to do arbitrarily well at, but that particular mechanic is really engaging.

Craig Perko said...

Hmmm... I like the basic idea, but right now I'm really stuck on the idea of letting the player build arbitrarily complex things and then "run" them.

So Puzzle Forge is outside my focus, at the moment.