|Game Design||By Shamus||Apr 18, 2013||90 comments|
A great deal of the fun in KSP comes from iterative building. If you looked far enough back in its family tree, you might find a grandpa or great-uncle named The Incredible Machine. You build a thing, press go, watch it fail, and repeat until you’ve perfected your design. My best moments with the game happened when I hit launch and had the entire rocket fly in seven different directions, obliterating my intrepid Kerbanauts and leaving a huge mess for the cleanup crew.
You can find this kind of play in Angry Birds, Portal, and even some Tower Defense games. You’re not intended to succeed on the first try, and there’s just as much fun in developing a new design as finally seeing it work.
This is in sandbox mode. The game will eventually have a campaign mode of some sort, and I can’t help but wonder what that will look like and how it will work. Right now you have access to all parts and can build as many exploding rockets as you like without penalty. You can fly to any planet you want, as long as you can figure out how to design a rocket capable of making the trip. There’s no external reward or recognition for accomplishing anything.
It’s natural to look at this open system and want to add some direction to it, but if you do it the obvious way you might kill the fun of the game. I realize that there is no lower form of game commentary than armchair game design, but I ask that you’ll indulge me a bit of rambling analysis. Also, I haven’t read anything by the developers regarding their plans. I’m not trying to out-guess them or make demands, I just thought this was an interesting design problem and I wanted to talk about it.
The most obvious way to make this into a game is by adding an economy. The player will somehow get money and they will spend the money building rockets. There’s already a cost value built into rocket parts in the game, suggesting that this is what they have planned.
I hope they do not do this.
The problem is that this would make the systems of the game run directly counter to the fun of the game. We saw this in Dishonored, where the game castigated you for using the most interesting powers. Combat powers were fun, varied, exciting, and if you used them you’d get the Bad Ending. An unspoken message of this sort of setup is “if you’re having fun, you’re doing it wrong”. If the player has to pay for every rocket part, then landing-pad detonations aren’t a hilarious mishap, they’re a ruinously expensive setback.
This will fundamentally kill the playful experimentation of shipbuilding. Instead of launching a ship to see if it works, you’ll be obliged to check and double-check your work to avoid mistakes. You will be avoiding one of the most entertaining aspects of the game. Instead of fast iteration, you’ll be forced to engage in slow analysis. When they have a mishap they won’t laugh because the command module went up a hundred meters, fell off and smacked into the explosive fuel tanks, they’ll curse because now they can’t afford to make another rocket and they’re going to have to do whatever it is you’ll do to make more money in this game. The player will be mandated to engage in focused, low-risk play.
The game could easily (inevitably, in fact) end up in killjoy situations where the player has sunk a fortune into a voyage only to find they’re just a bit over-weight and under-prepared once they’re a long ways into it. Instead of scrapping their design and making a better one, the economics will push them into overcoming the obstacle with tedious save-scumming because the financial loss would be too severe.
This sort of system wouldn’t ruin the game, but it’s not usually a good idea to purposefully introduce tension between what’s fun and what’s required. No matter how you look at it, charging for mistakes would drive players away from the things that make sandbox mode so rewarding. Yes, players could get around this by just playing sandbox mode, but that doesn’t really fix the problem. Most players will naturally gravitate towards structured play first, so they’ll be learning the game under punishing conditions.
In any case, there’s another way we could handle this…
I propose that instead of paying for each individual rocket and component with money, that we pay to unlock specific parts. Once you pay to fabricate or develop the Rockomax Mark 55 Radial Mount Liquid Engine, you can place as many of them as you like on as many Kerbal-murdering fireworks as you like, and the only cost is the time you spend making it happen.
If we really need some sort of in-game justification for this, we could assume that you’re paying SCIENCE POINTS! to unlock parts, but the rockets themselves are paid for by the Kerbal taxpayers. It could even be a running joke in the game to keep track of how many billion $Kerbux of taxpayer money you’ve wasted. (A running tally of crashes, fatalities, and taxpayer spending would be a nice touch.)
Doing things this way lets the player start off with a simple, straightforward list of parts. They can ease into the game by building orbiters that will teach them the basics without them needing to wade through 7 pages of exotic and esoteric gizmos. As they progress, they can choose which parts to unlock using the SCIENCE POINTS they’ve acquired. By doing this they can focus on unmanned or manned spaceflight, deciding if they want to go for reach (paying cursory visits to lots of far places) or depth (by building up bases or stations in one or two places.)
The penalty for building a rocket that doesn’t get the job done is that you’ll have to try again. Removing the direct financial penalty doesn’t make the game easier (because you still need to accomplish the goal) but it makes the iterative learning playful instead of painful.
Since we’re a thousand words into this and you’re still humoring me, let’s put another suggestion on the dry-erase board: The ability to earn SCIENCE POINTS by completing goals and the ability to spend those SP improving your technology.
You might have a few branches of technology where you can improve your metallurgy, construction, chemistry, solar cells, etc, etc. Put points into these to make your ships lighter, stronger, more energy efficient, etc.
See, the problem you want to avoid is the proliferation of parts. If their progress is entirely part-driven, then once they unlock the Rockomax Mark 70 they’ll never have a reason to go back to the Mark 55. The designers would have to script more parts, the modelers have to design more, and the player would have more to scroll through. This is a lot more work and it doesn’t make the game more fun. (Plus, it makes re-using and sharing ship designs a pain. If Josh sends me his early-game rocket, I’ll have to pull off all the outdated parts and replace them with the upgraded versions.)
Instead, I suggest keeping the part list as it is in sandbox mode: Each part is situational and every part has a use. There’s a time and place for the poodle and a time for the mainsail, and there isn’t a more advanced part that supersedes them both. Instead of acquiring better parts, you’ll acquire technology that makes your existing parts perform better overall.
Early in the game your ships will be heavier, less efficient, and more fragile. Near the endgame your ships will be lighter, more fuel efficient, more durable, even though they’re using the same basic parts. The sandbox mode we’re already used to would probably represent the mid-point on this spectrum. When you’re starting a new game you’ll barely be able to put a satellite in orbit. Late in the game things like manned round-trip voyages to the Mün should be trivial.
These improvements don’t need to be big. I assume we’re not going to have a endgame where you’re using Star Trek level technology. But even something reasonable like a 25% improvement across all systems can change the game in a big way. In late game when you’re designing a rocket that weighs 25% less, can endure 25% more stress, needs 25% less fuel to generate the same thrust, with engines that can output power 25% faster, and batteries that can hold 25% more power, you’re going to have a massive increase in what you can accomplish and how far you can go. This lets us give the player meaningful rewards that don’t break the semi-realistic tone of the mechanics. (No warp drives or antimatter engines or whatever.)
I really love this game and I’m hoping we get a campaign mode that allows for the fun of failure instead of punishing us for it. If I know game developers, the KSP team probably already has their own vision for the campaign mode and they don’t need me scribbling all over their design docs. But I offer this design as a conversation-starter and food for thought.