My standards for fantasy magic: it should, I believe, be mysterious, powerful and rare. This is my preference in all fantasy RPGs, especially Blade & Crown.
As I was designing B&C, the powerful part wasn’t hard to work out. Magic gives players a degree of control over the narrative, which is in some ways the ultimate power in an RPG. And the bigger the magnitude, the bigger the narrative control.
Mysterious? My biggest effort here was deliberately keeping the effects of magic vague. The spell descriptions are a couple sentences long at most, and I deliberately kept quantification to a minimum. I find that quantification is the enemy of quality, when it comes to magic.
Rare is defined largely by the difficulty in finding nodes. As I wrote in B&C, nodes shouldn’t be found just anywhere, and not even randomly. It’s important that the GM place them as is appropriate to the drama of the game. Finding a high-powered Crystal node should require, for example, journeying through a snow-covered waste for weeks — not just happening upon the node lying in the middle of the street.
However, even these standards can still lead to mechanistic, rigidly quantified magic. To be honest, I think it’s hard to quantify magic without losing some of the mystery. Once the mage knows that their Magnitude 2 node will produce exactly 517°C in temperature change per cubic meter of water, magic is no longer magical — it’s become engineering instead. (And if that’s the kind of magic you want, cool! But that’s not my preference, nor my design principle for Blade & Crown.) There is a strong inclination in gaming to quantify as much as possible, but I think that, especially when it comes to magic, quantification is usually at the expense of mystery.
Another thing that can allow magic to become mechanistic is convenience. If magic is easy to access, then it can become not just engineering, but commonplace technology.
In Blade & Crown, one of the places where I think this manifests most clearly is how nodes recharge: at a rate of 1 day per Magnitude level. This makes node regeneration quite regular and predictable — perhaps too regular and predictable.
A way to deal with this is to make the precise moment of regeneration unpredictable, or mystical. If the mage doesn’t know whether their node will recharge at midnight or at dawn, at the rising of the Moon or at the setting of the Plow, then they can’t rely on their nodes in a mundane way.
However, an experiment I’d like to try changes node regeneration in a more fundamental way. In this experiment, recharging a node requires sacrifice. The higher magnitude the node, the greater the sacrifice.
The type of sacrifice should clearly be aligned with the node’s element. Perhaps a Life node would require you to donate a bit of your own blood to give the node renewed energy; perhaps a Dark node would require meditation in a completely darkened room. A Light node might require candles to be burned in a particular way; an Earth node might require planting or nurturing something.
The degree of sacrifice should, of course, equal the magnitude of the node. A 1st magnitude node might require only a few hours of work; a 9th magnitude node might require a lifetime of dedication.
And of course the sacrifice needn’t be physical. Instead, a Crystal node might require you to act in a cold, calculating way; a Fire node might require sudden passion; a Metal node might require resolute endurance in the face of extreme danger.
And if that kind of sacrifice is required, then nodes resemble Traits even more than they already do. In exchange for harnessing mystical power, the mage is required to drive the plot forward. Great control over the narrative power then requires great contribution to the narrative — a good way to keep magic mysterious, powerful and rare.