Hey, everyone. Let's talk about complex systems.
Modern strategy games are deeply economic in nature. Simpler games, like Tic-Tac-Toe, lend themselves to naive optimization, which is why a game between two competent players will always end in a draw. My undergraduate thesis was written on the strategic analysis of the children's game
dots & boxes, which while definitely not as trivial as Tic-Tac-Toe, had a single optimal strategy to control the endgame.
To avoid these kinds of naive and trivial optimizations, then, non-random strategy games have tended to move in one of two directions: Combinatorial complexity which resists mathematical analysis (Go, and until the advent of Big Blue, Chess), or presenting players with a series of choices that involve both short and long-term trade-offs (most computer-based strategy games). The latter, of course, also tends to involve a certain amount of combinatorial complexity and rock-paper-scissors guesswork to prevent players from concluding that one course of action is optimal. It's also a lot easier to do, for certain values of "easy" - it's nearly impossible to make a game that's as strategically robust as Go, much less one that is actually accessible to new players (which Go and chess really aren't - learning the moves doesn't mean that you can achieve competency without tons of practice & study, and if you want your game to sell, you want your players to
feel as if they know what they're doing, even if they don't).
Anyway, the point I'm driving at is that RTS design is all about presenting players with choices - preferably real choices, rather than false dichotomies where only one or a handful of options is viable if you want to prevail. The way RTS designers do this is by creating complex economic systems where players have to juggle time, in-game resources, how much attention they can pay to each of their units, and requiring players to make decisions about how they should spend these limited resources in light of their game's rock-paper-scissors combat chain. The drawback of this complexity (which exists to obscure the game's underlying logic so it's harder to subject to naive analysis, as well as to pique the player's interest, prompting them to analyze it) is that it often makes it hard to anticipate the results of making changes to said system. Sometimes a butterfly will flap its wings in the Amazon and the housing market will collapse.
Essentially, that's what happened on Friday at work. I made a single change to the core of the game's economic system, which was intended to achieve effect A. Effect A was achieved! However, as a result of effect A, the Care Bears (
note: not actual care bears) became overpowered, because the fact that they had the Care Bear Stare meant that they didn't lose their units as often as the Flutter Ponies did, which meant they could amass more moonbeams and rainbows to charge up the Caring Meter, which meant they lost even fewer Care Bears, which meant they could charge up the Caring Meter even more, and basically the whole thing spiraled out of control.
As a result, I'm being forced to re-examine how much health and damage the basic Care Bear unit should start with - it's probable that they were always a little OP, but it just wasn't as obvious before. It's also possible that the solution to this problem isn't to completely retune the basic Care Bear unit, but to restrict access to the Care Bear Stare until the Caring Meter is already partially charged. Oh, and because I'm working with a complex system, whatever I try will need to be tested extensively, to make sure that I don't break something else in the process of solving this particular balance problem.
Did I mention that I love my job?
