A system-agnostic magic system?

System-agnostic seems like a contradiction in itself. Let it be a conundrum to solve (thanks @anahata).

What I’m actually pondering here is a set of magic rules for tabletop rpgs, that would work across a variety of systems. Say it should plug in effortlessly into OD&D and Classic Traveller, but also work with GURPS or Fudge, and maybe even more story telling focused games like FATE or PbtA-Games.

If you think of magic spells as entities that actually break or alter the rules of any given game, it doesn’t seem too far fetched, to have them interface in a sensible way with differing base rules … I hope.


The undeniable advantage of popular gaming systems is their relatively wide spread availability, a lot of players might already have the relevant book, or can access it easily.

As I like to play old games, or maybe not so popular game systems, I usually don’t expect my players to shell out money for a non-main stream gaming book, or to tediously search the second hand market for a copy.

For players of mundane types as fighting-men, rogues or simple adventurers this usually isn’t a big deal, but the magic users player will at some point want to reference the exact wording of some spell … so she’ll need a book … a spell book possibly.

For this exact situation a freely available (think Creative Commons, OGL …) and easily accessible (free indexed PDF, print yourself at cost at the local copy shop) would be ideal.

So let’s see what can be done …

What’s in a Spell?

I think if you look at the various spells in, say D&D and GURPS, there are certain recurring elements:

  • The spells effect
  • Time to cast
  • Range to cast
  • Duration of the spell
  • Restrictions

The spells effect usually conveys a narrative right, like legitimating the player to state how the magic user moves through a wall, conjures a demon or transforms into a giant mantis or whatever. Under other circumstances, this narration would break the rules of the game. That’s why you need magic. Sometimes the narration directly accesses gaming mechanics, like a fire ball doing 6d6 damage … that’s rules system specific … that’s going to be the difficult part.

Time to cast, range and duration are simple and universal. Even in the far future or in the most unlikely fantasy realm time will come in seconds, minutes and probably also hours days, and so on, and range can always be measured in terms of real world length measurement. Nothing gamey here, use SI units.

Restrictions are important – they tell you when, and only when it is allowed to break the rules by using magic and what kind of effect this has on the base game moving along. Those restrictions make the game fair and playable. Common restrictions are:

  • Spell slots
  • Spell levels
  • Spell cost, in terms of mana, fatigue points, gold, rare ingredients
  • Time and range also act as restrictions
  • agreement and / or support of some super natural being in case of clerical magic

As already said, time and range come in real world SI units, and gold, ingredients and a deities opinion about magic are part of the narrative. But spell slots, levels, mana points are rules-specific, so this is where a “system-agnostic” system would need to substitute alternative, self-contained rules.

First take at universal magic *rules*

This is a work in progress … two basic ideas:

Spells have varying impact and effectiveness, this could be modeled by an abstract spell complexity on a range from 1 (low complexity, easy to cast), to 10 (high complexity, hard to cast).

Magic users probably will be more or less apt at working spells. This could be modeled by some magical aptitude stat. This might be an existing stat, like intelligence (INT) in the original game, it might be derived from a combination of existing stats, or it might be newly added to a given game system.

I think a magical aptitude in the usual 3d6 range from 3 to 18 should be in order. I know, this is somewhat specific, but it also is a very common range for gaming stats, facilitating success rolls with 1d20, 3d6 or derivatives thereof like 2d6, 4d6 or 2d20 with advantage/disadvantage. Also, if multiplied by 5 the 3-18 range converts to 15-90, which is convenient for percentile based systems, and there are commonly known ways, do convert 3-18 to the -4 to +4 range as used in Fudge and FATE.

Now put this together … How about this, just set up a matrix with the difference of magical aptitude and spell complexity:

Spell Complexity

This already seems to be somewhat usable to me.

For example, if a magic user with magical aptitude of 12 wanted to cast a spell of complexity level 2, he’d need to roll a 10 or lower on 1d20 or 3d6 … depending on the base rules in action. On a success, the spell would work as expected, on a failure it would fail with some narrative consequence told by the GM.

In a system with vancian magic spell slots, this success roll could indicate, whether the mage managed to learn the spell, and is able to use it safely from then on.

It’s obvious that a character with low magical aptitude and little experience will never manage to work high complexity spells. Or only under certain circumstances, yet to be defined. Seems alright to me.

Now, since those odds seem somewhat depressing, lets add some modifiers:


Measured on an abstract scale from 1 to 10, experience could be used as a direct modifier. The experience number could represent years of study in a story telling game, or be a derivative of character level in a more classical approach.

As an example, a Mage with aptitude 12 and 5 years of experience would need to roll a 12 or less in order to work a complexity level 5 spell (base number 7 and +5 for experience).


A magic user might invest more or less time into working a spell. Modifiers could look like this, with a basic time to work a spell being hours:


In essence, this would make it possible for a high experience mage to cast a spell within seconds, while his apprentice might take month or even years to master the complicated ritual.

The time to cast could be further quantified by rolling a d6, so 1-6 seconds for a combat spell, or 1-6 days for some necromantic ritual.


The magic users might invest a fraction or multiple of the base cost into working the spell. Again, think mana, fatigue points, gold … what ever your system uses. Modifiers could look like this:

Cost multiplier1/101/41/21210100

You might think of yet other possible modifiers, but I think this basically covers it.

The actual Spells

I think in order to approach a wide range of spells in a generic way, I’d try a syntax based approach like in GURPS improvised magic, Maze Rats or Magic in the Moment, over on the Papers and Pencils blog: combinations of generic magical nouns and verbs like conjure, banish, hasten, heal, harm, fire, water, demon, monster and so on.

The actual number of magical words going into a spell could be used as yet another modifier. The more words, the higher the penalty ….

As I said, it’s a work in progress. I’ll try to model some commonly known spells with this system and see how it goes. Maybe start with good old light, heal light wounds, sleep and fireball. The players will certainly want those 😉