A lot of roleplaying games have rules that allow for characters to have a companion character who is in some ways an extension of the player character. They take up resources you would otherwise use for your character’s own abilities (points, slots, build choices) but more or less belong to you and act more or less at your direction, though sometimes they are under the game runner’s control.

The approach I’m taking to defining broad character qualities in A Wilder World encourages you to think of your character’s main defining qualities as a bit like a Green Lantern power ring to begin with. You still have to figure out which of your qualities best addresses the problem, and figure out how it addresses the problem, but the limits are the intersection of your imagination and the skepticism of the Storyweaver or group at large.

Magic takes that up a notch, in that there’s much less of a clearly defined line where “Okay, that’s just impossible.” Whether or not it’s possible to acrobat so hard you break down a reinforced stone wall is a matter of the game’s tone; if it’s at all realistic, then no, you can’t. Whether or not it’s possible to grow and animate plants with plant magic to take it down is more a matter of opinion, since “realism” and “verisimilitude” aren’t concepts that apply to high fantasy magic.

So magic needs to find its limitations elsewhere. I was talking about this back in June, and while I like the ideas I came up with there for different “magical prices”, I feel like they’re way overly mechanical in the way I imagined them being applied.

But while thinking about how to refine the idea of “low control” magic made me think of a general approach that I really like, and that is: treat magic the way you would treat an “ally” ability, one ultimately under the control of the Storyteller:

It shows up when the player character calls it, mostly.

It does what the player directs it to, mostly.

It does only what the player asks it to, mostly.

It’s the “mostlies” that matter, that make it interesting and that serve as a limitation. And the thing is, they all hinge on the idea that the magic use is successful.

Under this model, the use of magic would consist of three acts, which we might call summoning, directing, and dispersing. Each of them would have a chance for success or failure. If you fail to summon magic, nothing happens, or nothing significant. You might get a rustle of leaves when you wanted a gale of wind. You might get a brief patch of discolored air where you tried to form an illusion. You’re not left with any more problems than you started with, though.

When you direct the magic, you tell it what to do. Success means it does more or less that, but with the possibility of “wild” results built into the result mechanism. Failure would generally mean it does not do what you want it to do, with any significant negative downsides also coming from the play of cards.

You could continue to direct the magic as long as you maintain concentration on it, without having to draw or disperse it in between. This would be the normal state of affairs when a wizard is dueling or doing battle, or using magic to accomplish a large task that is actually many smaller ones.

When you’re done, though, you’d have to attempt to disperse the magic. Success signifies a clean ending. Failure would not mean your magic lasts forever, just that the magical forces you unleashed linger a bit longer and do some damage on their way out the door.

Even when magic is operating under your control, though, it would be like a charmed minion or a familiar or beast companion or cohort or point-bought ally, in the sense that you can tell the Storyteller “I have my magic do this”, but the Storyteller can interject, or interpret things a little differently according to the nature of the forces involved.

Making the equivalent of three checks to do anything with magic might be a little excessive. The concept of “trivial magic” (cantrips, roleplaying special effects, “I’m a wizard” demonstration) still exists. It’s only when you’re trying to achieve something that magic accumulates risk and price. Character gimmicks that take the form of specialized spells your character has mastered removes the need to summon magic before and disperse it after for that single very specific application, leaving only the control roll.

All this doesn’t completely supplant the idea of different “magic prices” I described back in June, but I think it makes for a better baseline approach. As always, my central idea when it comes to magic is to create an experience more like how powerful magic works in fiction than how it works in roleplaying games.