Ars Magica-style Casting

After Doug’s suggestion this afternoon in my post on spell casting options, I put a bit of thought into an Ars Magica-style casting system in Echelon.

I’ll ignore spontaneous casting for now because I don’t want to get into it, and rituals for the same reason.

Right now, pondering whether Ars Magica makes a good model for Echelon.  My first thought is that it is probably a viable approach, though I lack detail.

Simple Description of Ars Magica Rules

Ars Magica uses fifteen magic arts (I’ll spare you the dog latin).  Five are techniques (create, perceive, transform, destroy, control).  Ten are forms (animal, water, air, body, plant, fire, image/illusion, mind, earth, power/magic).  Casting a spell involves rolling a die and adding your scores in the required technique(s) and form(s) [some spells require more than one form, in which case you use the lower of the scores involved; I am not aware of any spells requiring more than one technique but did not examine closely] against a target number based on the ‘level’ of the spell.

Possible Approaches in Echelon

I’ve discussed before how I feel about rolling for spell casting, so I’ll ignore that bit for now.

I can probably use a similar structure to control spell access, though.  I am thinking of something like

• Use the D&D spell definitions, but add one to the level of each spell;
• Each art (technique or form) is a different talent and is worth one point per tier;
• Knowledge of a spell requires a combined total of art modifiers equal to the spell level (if you have ‘create 2’ and ‘fire 2’ you have ‘create fire 4’ and can know fireball; if a particular spell requires two forms you must use the lower of the two).

Given the number of arts required to fully master magic (15) and the top-tier slots available at any given time (no more than six, though up to eleven eight in the top two tiers) it is not possible to be equally proficient with all magic — some amount of selection or specialization is built into the system, which I like.

At the same time, very narrow mastery (single technique/form pair) is easily feasible and can be done with only two slots (a ‘warrior’ might learn to ‘create fire’ as well as a ‘wizard’ of the same level).  Two top-tier slots for such narrow  mastery is a pretty high buy-in, but I am not convinced it is out of line.  Given how easily this can be expanded by spending more slots, it might be an appropriate buy-in limitation to prevent “everyone from taking it”.

Interestingly, mastering a form requires six top-level slots (one for the form, five for techniques).  A ‘fire wizard’ might thus have create/perceive/transform/destroy/control fire, but no other arts.  A ‘transmuter’ (transform technique, all forms) would require eleven slots for the technique and all ten forms.

Incomplete Study

I am not certain how to handle cases where a character has only the technique or the form.  As written, a character does not need to have specific knowledge of both technique and form — ‘create 4’ or ‘fire 4’ would both be sufficient to allow knowledge of fireball.  This appears to be in line with the rules from Ars Magica.  The legendary fire warrior mentioned above (warrior with ‘legendary create’ and ‘legendary fire’) would have ‘create 6’ and ‘fire 6’, which could allow him to know all ‘create fire 12’ spells (obviously), plus ‘create * 6’ spells and all ‘* fire 6’ spells.  I’m not convinced this is right.  I’m also not convinced this is wrong.

An alternative would be to restrict it tightly, so you must have the techniques and forms required for the spell — at least a bit.  This superficially reduces the issue in that you no longer have a disconnect (you know ‘create’ and ‘fire’, so how are you transforming fire or creating water?), but only superficially because you can spend even minor talent slots on other techniques and forms to be better than the simpler rule.

Another alternative would be to require both techniques to equal or exceed the spell level (which would mean each tier could grant access to a new spell level, but only one).  This would really slow things down from D&D 3.x “new spell level every two character levels”, but might be worth considering in order to soften the power curve.

A third alternative might be to do the same but treat the techniques and forms as granting Training Bonuses.  The first talent (Technique: Create) gives just the Training Bonus, the second (Improved Technique: Create) gives the Training Bonus and a small technique-specific ability.  The technique and form Training Bonuses might then both be prerequisites, as above — it costs more talents to get the Training Bonus as high as possible as quickly as possible, but grants earlier access than just taking the one talent and a supplementary ability.  It would still take fifteen slots (three whole tiers, almost) to get access to all (technique, form) pairs, but a character could spend thirty slots (six whole tiers!) to try to get greater access to all spells.  I think specialization would be the order of the day here.

Spell Knowledge and Familiarity

I just realized, it might be worth considering two “degrees of knowledge”.  If the character meets the prerequisites for both technique and form (in whatever form identified above, either “both available and total high enough” — possibly with the rider of “technique and form each equal to or greater than half the spell level” — or “both available and each high enough”) the character might know the relevant spells.  If lacking sufficient knowledge (technique or method not high enough) the character may still be familiar with the spell.  In D&D terms this might mean “spell is considered on the ‘class list'” and thus be available for use with spell completion and spell trigger items (so the legendary fire warrior above — create 6, fire 6 — could cast up to 12th-level ‘create fire’ spells and use scrolls and wands containing ‘create 6’ or ‘fire 6’ spells safely.  It might even allow him to cast these spells directly, just not as well as his known spells.  Whether this means “additional time required”, “extra cost” (whatever that means right now), “skill check required”, or even “skill check with potential backfire”.

Options abound, here.

Closing Comments

So, the shape of this looks pretty good.  I’m somewhat concerned about the effort required to track all this, and I haven’t worked out casting capacity… but this feels pretty good otherwise.

Even within this “simple approach” there are a number of options and choices to be made, any of which can affect the power curve, and potentially the feel of the system.  As much as a decision would need to be made before implementation, that rebalancing mechanisms are already evident rather appeals to me.