On Arcane Magic — and Why Spell Slots Were Never Enough

The spell slot model has a literary ancestor. Jack Vance’s Dying Earth stories feature wizards who carry spells in their minds like charged weapons — use one and it is gone, replaced only through careful re-preparation. It is an elegant conceit for a dying, exhausted world where magic is ancient and costly and faintly absurd.

Gary Gygax borrowed the surface behavior — the memorization, the forgetting, the daily reset — and made it the mechanical spine of AD&D magic. What he did not borrow, because it did not translate into a game system, was Vance’s implied physics. Vance’s wizards understood what was happening to them. The forgetting was a property of a specific kind of magic in a specific kind of world, with its own internal logic. AD&D kept the constraint and dropped the explanation.

The result is a system that behaves consistently but cannot be reasoned about from the inside. A Magic-User in the world of the game cannot think about why their magic works. They can only account for it.

A craft practiced for centuries produces practitioners who understand their craft. Not every practitioner understands it deeply — most working mages are probably like most working engineers, competent within their domain without necessarily grasping the foundations — but the foundations get worked out over time. People observe, experiment, form hypotheses, test them against results. Institutions accumulate knowledge. The smart ones ask why, and eventually some of them get answers that hold up.

If arcane magic is real and has been practiced in Oath-Bound for as long as the historical record shows, then Oath-Bound has a body of arcane theory. Practitioners have names for things. They have competing schools of interpretation. They have working models that account for most observable phenomena and acknowledged gaps where the model breaks down. They argue about the gaps. Some of those arguments are centuries old.

The spell slot model cannot support this. There is nothing to argue about. You have slots. You fill them. The checkbook does not generate theoretical disputes.

The Oath-Bound arcane system is built around nwyf — the medium through which arcane effects are instantiated in the prime material. Nwyf is not a unique concept. Analogues appear across fantasy literature and game design under various names: mana, essence, the force, vis. The underlying intuition — that magic operates through some permeating medium that practitioners learn to work with — is widespread because it is a natural way to think about how a craft-based magic system would develop.

What Oath-Bound has tried to do is give nwyf enough internal structure to be worth knowing about. Not so much structure that it becomes another checkbook in disguise, but enough that a practitioner in the world could have opinions about it. Enough that two mages could disagree about how it behaves in a given situation and both be reasoning from something real.

Nwyf is emitted by wyrd — the deep structure of obligation and consequence that runs through the prime material. It permeates the world unevenly. Practitioners learn to perceive it, to work with concentrations of it, to shape effects through it. Their capability is described not by a slot configuration but by a band — a level of working that reflects how much nwyf they can direct and what density they can achieve. Working at the edge of their band is costly and uncertain. Working well within it is reliable but limited.

The band system is not a disguised slot system. It describes a practitioner’s relationship with a medium, not a daily allocation of uses. A practitioner who understands nwyf understands why some days are harder than others, why certain locations make working easier or more dangerous, why a colleague with a different cast produces different results from the same working. There is something to understand, and understanding it is part of what it means to be an Arcane Actor in Oath-Bound.

One of the things the PHB magic system cannot account for is why two Magic-Users of equal level, casting the same spell, might produce different results. In the PHB, they would not. The slot is the slot. The spell is the spell.

In Oath-Bound, every practitioner has a cast — the characteristic quality of the nwyf they work with, shaped by their training, their instincts, and the particular way they have learned to perceive and direct the medium. Cast is not a mechanical modifier in most circumstances. It is the reason arcane practice is a craft rather than a procedure. Two practitioners of matched cast working together can achieve multiplicative effects that mismatched practitioners cannot. The same working, in different hands, can have a different texture.

This is the kind of thing practitioners would notice, discuss, and argue about. Arcane academies have opinions about cast. Masters assess it in their students. The vocabulary exists in Oath-Bound because the phenomenon exists, and the phenomenon exists because the underlying model supports it.

None of this requires a player to understand the theory in order to play an Arcane Actor effectively. The band system tells you what you can do. The GM adjudicates the details. Most sessions will not require a working knowledge of nwyf theory any more than a carpenter needs to understand materials science to build a chair.

What it does mean is that an Arcane Actor in Oath-Bound inhabits a craft. They have a model for what they do. They can think about it, talk about it with other practitioners, and make reasonable inferences about situations they have not encountered before. When something unexpected happens — a working that behaves strangely, a location where the medium feels wrong — they have a framework for asking why.

That framework also travels beyond the table. One of the less obvious costs of a sparse magic system is how it handles the unexpected question — the player whose Arcane Actor wants to research a variant of a familiar working, something adjacent to what they already know but pushed in a new direction. The vanilla AD&D system has almost nothing to say about this. Two tables asking the same question get two different answers, not because the GMs are making different judgments but because there is nothing underneath to reason from. The answers are invented from scratch each time, and they do not travel — a ruling that makes sense at one table is arbitrary at another because there is no shared model connecting them.

A coherent underlying system changes this. If nwyf has properties, if cast has a signature, if bands describe a real relationship between a practitioner and the medium, then a GM facing a spell research question has something to derive an answer from rather than invent one. How far is the proposed variant from the base working? Does it push against the practitioner’s band or sit comfortably within it? Does it require a cast that matches their instincts, or is it fighting them? The answers still require judgment, but they are judgments anchored in a model rather than pulled from thin air. And two GMs reasoning from the same model will tend toward compatible answers even on questions neither has faced before. The system is extensible in a way that a checkbook cannot be.

That is a different experience from accounting for spell slots. It is also, we think, a more interesting one.