Kenneth on Games: Horizontal Scaling in “Offworld Trading Company”

I’m going to confess that I’m a little bit scared of RTS games. They’re always so difficult to get into and you have to memorize so many build paths and then you just get killed by a strategy you didn’t know was possible. If only there was an RTS that was about reactions and decision making, rather than memorization. Luckily, Soren Johnson has been thinking about this problem too, and the result is Offworld Trading Company. This game takes heavy advantage of horizontal scaling to simultaneously ease the learning curve and encourage reactive strategies at the same time.

Horizontal vs Vertical Scaling

The term “scaling” as I use it is actually more commonly applied to computer science. In simple terms, horizontal scaling is when you add more components, and vertical scaling is when you add more things for a single component to do.

Most RTS games emphasize vertical scaling, and Starcraft 2 is the quintessential example. You only have two basic components, minerals and vespene gas, but then each of those components branches out to build so many different things. These units require X minerals and Y vespene gas, but in order to build them you have to research them which costs another chunk of minerals and gas, and then you have to actually build the research center, and so on and so forth. Everything branches off of just two types of components. Even the Civilization series, Johnson’s more well-known projects, worked this way.

But OTC has a massive thirteen unique resources. Six of them can be harvested from the map, whereas the remaining seven have to be synthesized using some combination of the others. However, each of these resources also governs fewer possible tasks. While minerals in SC2 are used in practically everything you could build, resources in OTC do one or two things and that’s it.

In order to get what you want in SC2, you have to navigate a complex tree of interdependencies and prerequisites. In order to get what you want in OTC, you just build it. Direct input to output conversion. OTC focuses on what you want rather than how you get it. Instead of wondering “how do I ensure that I’m going to get this objective,” you wonder “do I really want this objective, or another one?”

If a SC2-style vertically scaled RTS could be described as a game of sticking to your decisions, OTC would be described as the opposite, a game of being flexible and dynamic. However, then you run into the problem of power levels. In a world of perfect horizontal scaling, every resource would be completely equal, and that doesn’t make for interesting gameplay. You can turn on a dime and switch strategies whenever you want, but if every option is equal then there’s no point. Some options need to be stronger than others in order for the decision to be meaningful at all.

The fluctuating economy system acts as a way to simulate power spikes for brief windows of time. When one resource is rare, players want to commit and produce a lot of that resource before its price goes back down. Contrast this to SC2, where you can look up any race’s vertical scaling (or tech tree) and it stays that way forever. In OTC, everything is built on horizontal scaling, and then power spikes are artificially applied during runtime, effectively creating the same effect as vertical scaling but constantly changing throughout the match.

Anticlimactic Victories

One main problem with this system is the sense of progression, or lack thereof. In a vertically-scaling RTS, the game has a feeling of growing power. Your base gets larger, your units have shinier armor, you control whole armies rather than just small patrols. As you go up the vertical scale, you look back and see how far you’ve come. Then you send your units to attack the enemy and a giant epic battle happens and there are explosions everywhere, and one rugged victor rises from the ashes.

Matches in OTC has no such conclusion. You win by buying out your rival company’s stocks, but you can do this at any time as long as you have the money. So the result is that there’s no buildup. The game is going, people are making money, and then suddenly one player buys out everyone else and wins. It’s literally clicking a button in a spreadsheet.

On one hand, you could say that this is because spectators probably aren’t as familiar with the game, so they don’t have a sense of how the cash flow changes over time, and there’s actually a really subtle progression curve happening within the economy. On the other hand, this is a reaction that actually happened to the developers themselves when they spectated the OTC release tournament:

“…all that said, PB, he’s the one who stops getting his water frozen first… and suddenly the game’s just over. Just like that.”

“Well, that was quick.”

“Just suddenly ends. Nevermind. Game’s over. Forget it.”

“Wow. I didn’t see that coming.”

Horizontal scaling helped OTC get away from the memorized-build-order problem that plagues most RTS games, but it also took away the opportunity for climactic victories. It’s a lot like the problem with Dominion in League of Legends, where your win condition is disconnected from what you are actually doing in the game.

Victory feels like it should be the result of hard work and dedicated commitment. This goes directly against OTC‘s philosophy of adaptive gameplay. Is this a solvable problem? Could there be a way to get the best of both worlds?

If Soren Johnson couldn’t find an answer, then it’s probably impossible. But did he want to? It’s entirely possible that the OTC development team was perfectly fine with having a spreadsheet win condition. The current stock system is probably the closest they can get to a transparent progression bar without compromising the game’s vision.

But if it had to be a decision between dynamic gameplay or epic victories, I’m glad they chose the former. OTC is a new type of game that solves a lot of problems, and even an RTS-ophobe like me can get into it. Too bad I can’t blame my losses on build orders anymore though.