One of the funkiest tasks in making a fantasy game has got to be the magic system. Rules for travel, combat, and any number of other mundane tasks can be modeled on real-world data, as in Justin Alexander’s famed breakdown of stats and skill use. But magic has no such analog.

…Actually, that’s not true, is it? The real world is full of “magical” practices. Some are just science obeying Clarke’s Third Law: botany, chemistry, and applied disciplines like metallurgy. Some are passive folk beliefs about properties of the natural world: the protective power of certain gems or the color blue, for example. And some are beliefs about supernatural powers that people can actively bring to bear, either through their own power (such as the Evil Eye) or through channeling, directing, or requesting the powers of spirits or gods.

Which explains why magic is so thoroughly all over the map when it comes to its manifestations as systems in games: it’s hard to unify something that’s based on such a widely disparate set of real-world phenomena and beliefs. That said, the various cultural influences on the form and feel of specific magical effects and practices in a game world are less important, in the game-design sphere, than the mechanics behind how magic works: while it’s possible to mechanically support a specific “flavor,” it’s also possible to design a mechanically generic system that can support a variety of flavors.

The fundamental question at the root of any magic system is, “How do you limit it?” Most activities are limited by simple reference to the real world: people often have a decent intuitive sense for things like swimming, running, even swinging a sword. But with magic, by definition real-world limits don’t apply. Anything that you can imagine, or even simply gesture toward with words, is possible. To prevent every game from becoming Calvinball, some limitations need to be imposed.

Which brings us to the question of how to limit magic in YAOSC. And the answer is, “There are a number of overlapping methods that I plan to use, to be honest.”

First, what we’re not going to do: “spell levels” per se. D&D has class levels and spell levels, and a character’s supply of the former limits their access to the latter. They’re not intrinsically bad; it’s just that after reading this post. I wanted to build YAOSC without them. The implied setting for the system is one where almost anyone can at least try to work almost any magic, if they have the know-how, and their skill in the art is what determines the result.

What does that mean specifically? Well, everything begins with the skill system. Most magic use is going to be a matter of making skill checks, or of having a high enough skill level that no check is necessary.

For a long time I was thinking that there would be three skills that governed magical power: “rote” (for something close to traditional “spellcasting,” requiring application of both special knowledge and the power of the user), “ritual” (for time-consuming algorithms or recipes that require no other knowledge and little or no personal energy), and “gnosis” (for intuitive manipulation of magic forces, depending entirely on the power, will, and intuition of the user). “Concentration” would also have been an important skill for magicians.

But in the meantime, I’ve thought a bunch about a “skill point economy,” and I like the idea of a system that allows and encourages skill points to be invested in the use of specific tools – including specific spells. And it doesn’t entirely make sense to me for “rituals” to be one unified skill either; by definition they’re going to be specific and finicky and situational.

So instead of that, we’re going to have just two skills: Concentration (the ability to focus on a task and cultivate a “flow” state) and Arcane Lore (possessing and comprehending a collection of metaphysical trivia useful in getting a feel for how magic works). There will still be rotes (standard “spells” that require power input), rituals (longer methods that use other in-game resources such as time, rare and costly props, etc.) and gnosis (working magic through sheer willpower), but every rote or ritual will be a mini-skill. Players who invest points in those skills can gain certain benefits (such as lower energy costs or greater effects). But also, if a check is required, the relevant skill levels are added together:

• Rolling a check to perform a ritual adds your Arcane Lore skill to your skill level for that specific ritual.
• Rolling a check to execute a rote adds your Arcane Lore OR Concentration skill (depending on the situation) to your skill level for that rote.
• Casting by gnosis uses Concentration only. It’s harder to get good results when you’re not using a recipe.
• The difficulty of a spell can be signaled by the default die size for a check to cast it. (This is as close as we’re getting to “spell levels.”) Magicians can work harder (roll a larger die) to get more power, range, duration, etc. out of their spells, or accept diminished effects for less difficulty (a smaller die).
• Spells demand energy from the caster, reflected by filling one of the three resilience meters. Magic strains your mind and/or body. Difficulty can be increased to decrease the energy costs, or decreased by spending more energy. Energy costs should be calibrated so that anyone can afford to cast almost any spell at least once, but reliance on frequent magic use demands careful resource management.
• Failure on a skill check to cast a spell should have consequences, ranging from wasting time and energy (on a close miss) to catastrophic magical weirdness (on a dramatic failure), to discourage players from simply rolling and re-rolling the dice until they succeed.

I anticipate several results of this setup:

• Even at low character levels, simple spells can be cast automatically by dedicated students of magic, who have invested skill points in both the base skill and the spell itself and so their score is equal to or greater than the die they’d be using for a check – no result can be a failure.
• Similarly, powerful magicians can reliably perform even difficult feats of magic, and have reasonably good chances to achieve effects that non-specialists wouldn’t dare attempt unless desperate.
• It’s possible to fall anywhere on a spectrum of magical ability, from dilettantes who have only invested a handful of skill points into a single easy spell, to savants with a few points in Concentration and spells, to casual students with a few points in Arcane Lore and spells, to specialists with maximum points invested in both of the governing skills and in a wide variety of spells.
• Novice magicians will tend to gravitate to simpler spells due to their ability to cast them reliably, but anyone can have any trick tucked up their sleeve. If you have access to a spell, the biggest limits on what you can cast are the risks you’re willing to take.
• As YAOSC comes together as a whole, one thing to watch out for is cheap and easy ways to empty the resilience meters; those could be abused to provide magicians with unlimited magic use. In and of itself this isn’t necessarily bad, but it’s not what I want for this system.

All of the above is subject to change as I continue to work on things, of course, but it seems like a solid foundation to start building on for the time being. Join us next week when I come back to some of the Necropraxis-inspired thoughts that started off this whole thing in the first place: counterspelling and other sundry magical tasks that won’t be modeled as spells.