Tuesday, January 10, 2017

Meditating on Vancian Magic

The tomes which held Turjan’s sorcery lay on the long table of black steel or were thrust helter-skelter into shelves. These were volumes compiled by many wizards of the past, untidy folios collected by the Sage, leather-bound librams setting forth the syllables of a hundred powerful spells, so cogent that Turjan’s brain could know but four at a time.

Turjan found a musty portfolio, turned the heavy pages to the spell the Sage had shown him, the Call to the Violent Cloud. He stared down at the characters and they burned with an urgent power, pressing off the page as if frantic to leave the dark solitude of the book.

Turjan closed the book, forcing the spell back into oblivion. He robed himself with a short cape, tucked a blade into his belt, fitted the amulet holding Laccodel’s Rune to his wrist. Then he sat down and from a journal chose the spells he would take with him. What dangers he might meet he could not know, so he selected three spells of general application: the Excellent Prismatic Spray, Phandal’s Mantle of Stealth, and the Spell of the Slow Hour.

— Turjan of Miir, Jack Vance

I have always really liked the idea of Vancian magic, but have never been sold on how D&D handles it. Don’t get me wrong, from a purely mechanical point of view, D&D does a good job of simulating the way Jack Vance describes how Turian of Miir uses magic. Unfortunately, it doesn’t have the feel. If someone had never read Jack Vance and had no idea of how he describes the spell as an almost living thing that frantically tries to leap off the page of a spell book, there is nothing in D&D that shows players that this is what the mechanic of memorizing spells is supposed to represent.

My other issue with spell casters in D&D is that I am someone who much prefers the utility spell over the combat spell; however, D&D almost dictates that a magic-user will always take Sleep over Read Languages. On an average dungeon delve, a player can almost be guaranteed to be able to use a combat spell, but that utility spell may or may not ever be useful. I find this boring, in a way.

To be honest, when I play a spell caster, it is almost always a cleric or a bard because I can get away with having utility spells more often because these classes can pull their weight in a fight sans spells. I have only ever played a straight-up magic-user once and a straight-up illusionist once. I much preferred the latter experience because of the sheer frustration about having spells like Enlarge in my spell book and having really good ways to use it, but having to go through the whole “I gotta rest, re-memorize spells and come back tomorrow” only to have the party try some other way to solve the problem. At least with the Illusionist, the very nature of illusion magic requires creativity and problem solving skills with every spell cast. I found it much more satisfying to take out a pack of gnolls with Phantasmal Force by “moving” a pit and getting them to fall in than I ever was shooting lightning bolts or fire balls.

When I got into the nitty-gritty of the classes in 5e, I really got excited bout the potential of the the Warlock as a utility/problem solving kind of spell caster. They have enough combat punch with Eldritch Blast to justify using utility spells. Couple that with the Pact Boon Book of Shadows and the ability to learn some Cantrips off any spell list and the potential to collect and use ritual spells from any spell list and you’ve got a spell caster I’d love to try and play some day.

Unfortunately, whereas the Warlock does a good job of making utility spells justifiable and does a good job of making magic feel dangerous, it largely abandons the Vancian model of magic. Therefore, in terms of trying to make a BX magic-user class that does Vancian magic “right,” the Warlock is inspirational, but not exactly what I am looking for.

I recently managed to get my hands on a second edition copy of Runequest. I have always wanted to love Runequest, but my experience of it was always tainted by the Avalon Hill edition, which is just badly written. I could never get my head around that magic system. The second edition, though, is a transitional version that bridges its OD&D roots and the Chaosium BRP system it would become. Here is a version of Runequest I finally get and its magic system is something I finally understand. Unfortunately, it, too, turns away from the Vancian model; however, it presents with a mechanic that makes magic feel really dangerous and could be coupled with a Vancian model. Certain types of spell casters capture spirits and bind them into items in order to help them cast spells. The danger is in the combat to capture the spirit: if the spirit wins, the caster is possessed and the PC becomes an NPC.

This got me thinking about how to incorporate this level of danger into a magic-user class while maintaining a Vancian model and allowing players fuller access to the utility spells that litter the spell lists. Normally, the limiting factors of spell use in FRPGs are one or more of the following:

  • memorization
  • number of spells known
  • power/mana points
  • class level

Of these, the only one that is expressly Vancian is memorization. Therefore, if I am to move towards a magic user that satisfies my desires for an arcane spell caster like the Warlock does but remain true to the idea of Vancian magic, there needs to be a different limiting factor than one presented.

If one assumes that arcane spells are more of a living thing — a kind of semi-sentient energy being — and that spell books and arcane magic are a way of coaxing these beings to do what a spell caster wants, then there is a limiting factor available that is already suggested by the game: money.

The idea here is similar to the way scrolls work: the magic-user spends gold and time on getting the materials necessary to have a spell “in waiting” ready to cast without having to memorize it. Normally, this is relatively expensive (100gp and a day per spell level, for example) and, with some exceptions like Holmes, not an option for 1st level characters.

My thinking here is to couple the gold/time with spells known (as per INT in Holmes) as the primary limiting factor for magic-users. If they would be allowed to bind spells to disposable objects (like runes) for a relatively low price, then they could go into an adventure stocked up with all manner of spells. It would also incentivize adventuring as a primary occupation of magic-users. Right now, I am positing 10gp and about an hour for a 1st level spell. This cost would go up exponentially with spell levels, adding a zero for each level.Thus, a 6th level spell would require 1,000,000 gp worth of materials to bind.

There would also be an option for binding spells permanently to a magic-user. It would work like this: make a saving throw vs. magic and add the magic-user’s CHA bonus and subtract the spell level. If successful, the spell can be cast at-will by the magic-user. Failure, however, means that the magic-user has been possessed by the spell itself and is now a Chaotic spell-casting monster that will be hunted down and killed by Civilization.

The risk involved in binding spells this way could be mitigated by having other magic-users aid in the binding: they get to add their CHA bonuses to the roll. This, then, places a kind of cultural limiting factor on the practice. Whereas a Wizard’s Guild would have little issue binding a Detect Magic spell to a fellow magic-user, they might have something to say about a magic-user who was interested in binding a Lightning Bolt spell. This also implies that there would be magic-users out there that would seek alternative ways to mitigate these bindings in order to gain power. The whole idea of a magic-user just got a whole lot more dangerous: imagine an adventure hook where the Wizard’s Guild has put a bounty on the head of that magic-user who successfully bound his Lightning Bolt spell…

This way, I get my Warlock (bound utility spells) and my Vancian magic (prepared “burn” spells) all in the same class. Thoughts?


  1. Simply Brilliant! Though I'm a huge fan of the binding mechanic (and all its social implications) I'm hesitant to make spell casters pay for every spell. The fighter doesn't pay for every sword swing, and it would really frustrate me as a player. The first level spells aren't so bad, but spending a million gold for a one time use class feature is more akin to buying a magic item.

    But the very low prices still intrigue me. I definitely need to playtest this.

  2. The problem with making the arcane magic user “pay” to memorize his spells is that it places a real cost (in gp) on the routine exercise of his class attributes that is not required of other character classes
    i.e.., the thief doesn’t pay every time he picks a lock, the fighter doesn’t pay every time he swings a magic sword and the cleric/priest doesn’t pay every time he heals. Holmes dealt with the limitations of the Vancian system by adding scrolls:

    “This rule places great limitations on the magic-user’s power, but there are ways to partially overcome them. One is to have the spell written on a magic scroll. Scrolls are written in magic runes that fade from the page as they are read, so a scroll can only be used once. Magic-users (not clerics) may make a scroll of a spell they already “know” (i.e., have in their magic book (s) at a cost of 100 gold pieces and one week’s work for each spell of the first level, 200 gold pieces and two weeks work for a second level spell 300 gold pieces and three weeks work for a third level spell, etc.”

    This is a good place to start BEFORE one considers a spell-point system:


    “Another difficulty with a spell point system is low-level spells that scale with caster level. A 9th-level magic user who gets five magic missiles per casting could dish out 5d4 +6 points of damage per round 31 times in a day in a 1-spell-point-per-spell level system. There's almost no reason at all for him to even bother with fireball or lightning bolt, the wizard's traditional big guns. Two rounds of magic missiles, for 2 spell points, deals 10d4+10 damage, eclipsing a fireball's 9d6 for 3 spell points. “

    1. Yes, I mentioned Holmes in the post specifically. I also admit that D&D and the older rules do a good job with magic. I have never been a fan of point systems, I actually like how Vancian magic works, I would just like it to feel more Vancian.

      I quibble with the argument that other classes don't have to pay for routine class skills. Bow and crossbow users must pay for ammunition, thieves must pay for tools uno order to use certain abilities, with the training rules in place all characters must pay for their next level abilities and even in an Arneson-driven 1gp spent = 1xp all characters are paying gold for all their routine abilities.

      I am not denying there is a trade-off here, but it is one that I am not sure players would reject. I have seen 3e parties pool resources so that spell casters forge magic items. I don't see why the same wouldn't happen when the party needed a high level spell. In addition, with low level spells being so inexpensive, I wouldn't mind being able to have 5-10 spells available to me right off the bat as a 1st level character vs. just one even though I could re-memorize it the next day.

  3. Sounds interesting and dangerous. I like it.

    And really gets me thinking of plot hooks and adventures based around possessed spell casters, or those who committed atrocities to learn the spells they wanted. Add some prestige and awe to the truely powerful spell casters as well, letting them break the conventional "rules" by learning several spells.