By default, D&D uses Vancian magic, and has done since its earliest days, more or less. Vancian magic is where spells are prepared in advance and can only be used a finite number of times. It is sometimes known as ‘fire and forget’ magic, or, more disparagingly, as ‘utility belt’ magic. The term ‘Vancian’ comes from fantasy writer Jack Vance, and this is how spellcasting works in Vance’s novels. (Fun fact: the name ‘Vecna’ is an anagram of Vance.)
Vancian magic has its place. It helps to create game balance. It has been part of the game’s fabric for a long time. But it can also feel a little artificial and unintuitive. It locks you into thinking about magic in a particular way.
In this article, I present a variant system where spellcasting is more unpredictable but also, potentially, more organic. You don’t have to use it. It isn’t thoroughly playtested. But it sits as an alternative to the usual D&D method of tracking spell slots and spells per day.
How it works
With this variant system, a character who has the Spellcasting feature makes a mana check instead of using spell slots to fuel spells.
Instead of gaining a number of spell slots each day to cast your spells, you must make a DC 10 mana check after you cast a spell. By default, a mana check is 1d20 plus your proficiency bonus (but does not include your spellcasting ability modifier). For easy spells, you roll with advantage. For hard spells, you roll with disadvantage. The difficulty of a spell is determined by the table below: Spell Difficulty per Level. If you succeed on your mana check, nothing happens, and you can keep casting spells. If you fail, you can no longer cast spells of that difficulty.
By default, this is a system for ‘full’ casters: bards, clerics, druids, sorcerers, and wizards. I have tried an alternative system with ‘half casters’ like paladins and rangers, but I think it needs a bit more work.
As an optional ‘burnout’ rule, you might allow spellcasters to take on a level of exhaustion to keep casting spells. A spellcaster can also cast spells at a higher level, as normal, but doing so may make the mana check more difficult.
Spells of 6th level and higher are particularly taxing to cast. A spellcaster can only cast one spell per day at each level, and no mana check is required. Cantrips, on the other hand, can be cast at will, as usual.
This system can be applied to monsters that cast spells using spell slots, but you will need to work out the spellcaster’s proficiency bonus based on their spellcasting level. A flameskull, for example, casts spells as 5th-level spellcaster, so its mana roll would be +3 (the same as a 5th-level wizard). To be honest, I’m not sure it’s worth the effort, since most NPC spellcasters are dead within three rounds anyway, but the option is there if you want it.
Example
Let’s take a Mialee, 7th-level wizard as an example. 1st-level and 2nd-level spells are ‘easy’ for her: she can roll her mana check with advantage. 4th-level spells are ‘hard’: she rolls with disadvantage. For 3rd-level spells, she makes a mana check without advantage or disadvantage.
In the first round of combat, Mialee casts magic missile, a 1st-level spell. Mialee’s player rolls damage as normal and then makes her DC 10 mana check. This is 1d20 plus her proficiency bonus, which is +3, but she rolls with advantage as 1st-level spells are ‘easy’ for her. She rolls 20: easily enough to keep casting spells.
On her next turn, she casts fireball, a 3rd-level spell. Her enemies make saving throws as normal, and she rolls damage. She then makes another mana check, this time without advantage (fireball is neither ‘easy’ nor ‘hard’). She rolls a 3: even with her +3 proficiency bonus, that’s not enough to succeed. If Mialee wants to cast fireball again, she would need to cast it at a higher level, which would be a ‘hard’ mana check.
In desperation, Mialee decides to do this on her third turn. She casts fireball at 4th level. After rolling damage, she makes her mana check, this time with disadvantage. She gets a 10! It looks like she can keep casting fireballs – for now.
In play
As stated at the outset, I haven’t playtested these mechanics extensively. However, when we tried these rules as part of my regular Monday-night game, they seemed to work well. And the maths checks out, too (I think). On average, you you should be able to cast the same number of spells per day as a normal spellcaster of your level, give or take.
The biggest difference you might see with this mechanic is that spellcasting becomes a bit more of a gamble. Spellcasting is less reliable than before: you might find yourself out of spells in the middle of a boss fight, for example. But it swings both ways. You might find yourself suddenly powered up by being able to cast far more spells than you normally would.
Try it out. Let me know in the comments below how you get on.
To subscribe, click here. You can unsubscribe any time. You can find me on Facebook at scrollforinitiative, Twitter at scrollforinit, and Instagram at scrollforinitiative. And if you like what I do, you can support me on Patreon or buy me a coffee here.
At early levels, I think it works pretty well. After 4 levels, it has some probably undesired effects. For example, you probably never want to cast a level 1 spell (or at least without upcasting, if applicable) from the moment you hit level 5 onward, since they are both equally risky. With a PB of +3, you need to hit a 7 or more at advantage, which ends up being a 91% pass rate, which ends up being 8 or more easy slots half the time. A full caster normally has 4 firsts and 3 seconds. Most of the time, they will effectively have at least 7-8 2nd level slots. Then at level 9, the same happens with third level spells. It essentially adds free upcasting at 5 and 9.
The “standard” and “hard” are fine and math out about correctly at any given time, but having a range for easy does make upcasting the lowest level slots pretty much free for no trade-off.
You’re quite right – it was something we picked up on in playtesting. I’m not sure what the solution is, frankly! I suppose it’s part of the gamble of the system: sure, you can upcast 1st- and 2nd-level spells for free while the going is good, but you can lose everything in a moment. It didn’t feel game-breaking when we tested it, but I totally see your point.
One fairly simple solution might be rely on D6 checks, and have the depletion on a per-spell level basis rather than type of check.
Recovery on 6+ would result in an expected 1.2 casts per rest
Recovery on 5+ would result in an expected 1.5 casts per rest
Recovery on 4+ would result in an expected 2 casts per rest
Recovery on 3+ would result in an expected 3 casts per rest
Recovery on 2+ would result in an expected 6 casts per rest
There isn’t a clean way to do expected 4 casts besides 4+ with a reroll, but that makes the progression less simple. However, the current chart only goes to 3 except for 1st level spells, so it might not be all that important. Could either translate the existing chart, or render it into an approximately equivalent formula. The recovery rules could be roughly stated as:
– Recovery occurs on a 4+ on a D6
— Add (Caster level – 2xSpell level) to the roll (this is -1 on the level the spell slot is first gained)
— +1 to the roll if the spell slot is level 1, 2 or 3
— Net modifiers cap at +1
This ends up being slightly generous on the top end (1.5 casts instead of 1 for level 4 spells at 7, 1.5 casts instead of 1 for level 5 spells at 9) but also tops out at 3 level 1 spells. Also transitions to 3 level 5 spells at caster 11; probably easiest to say the progression stops at 10 (i.e. caster level is capped at 10) and special case a bonus to the roll at level 18:
— Caster level for the check is capped at 10
— At caster level 18, +1 to the roll (still capped, so only affects level 5 slot)
The most interesting thing about a mechanic defined like this is that it allows a means of dialing spellcasting up and down in a way that could add more tactical decisions. Consider a magical ‘tide’ where the recovery roll gets a bonus or malus depending on the specific turn, or a weaker variant of spell mastery that made it easier to recover slots when casting mastered spells, or a way to express opposition schools (though 5th edition has been good about describing differences by addition rather than subtraction).
Arcane recovery is somewhat difficult to model with this system, but the easiest approximate solution appears to be to just allow the caster to get a re-roll (per day) on a depleted slot to recover it during a short rest. This appears to work out to roughly the right number of spell levels, though it needs to scale up after level 10 if it needs to match the arcane recovery scaling (one additional reroll somewhere after 11 seems to match). It could also just be kept the way it is (i.e. allow for bonus uses that don’t roll after a slot depletes) but that probably defeats the purposes of the new mechanic in the first place. Main issue with the reroll is that it’d feel pretty bad missing the roll (since that generally means you were unlucky to start with). Alternatively it could be just flat out recovering a slot, but probably would have to be capped to something like spell level = casting level/6 (each new slot is worth ~3 casts, and arcane recovery is worth half the caster level in spell levels).
Sorcery points for slots probably can be kept, though, since they can be spent as they are used so don’t have to have a record of the new slots (most likely change the rule to allow/require this – have it expire on the turn they are created and maybe make it a once-per-round free action instead of a bonus action).
Interesting system, that really gives some flexibility in approaching encounters and the use of magic in rp outside of encounters. Any system that includes checks after casting though becomes more challenging for new players. Adding a table to manage those checks even more so. Did you test making a straight check of base dc + spell level? With the check being made against spell casting modifier (or arcana/religion/nature/performance depending on caster background. You could add a feat that would add proficiency to spell casting ability check.)
Would be a much simpler conceptual system, scale with up casting and require more player thought on stat investment. Without testing it seems like more opportunities for failure, but you could balance by making failure spell specific not level specific. You could also allow casters to sacrifice hit die on short rests to recover spells (also in your system you could do spell levels say 1 hit dice per spell level)
You’re quite right that checks after casting could become challenging for new players. I must say, though, in my experience, Vancian magic is challenging for new players because it’s so different to how magic works in most other fantasy stories. (And then there’s the whole issue of ‘5th level wizards cast 3rd level spells’!)
I did look at some kind of DC + spell level, but it ended up making 4th- and 5th-level spells a little too easy to cast. After all, each spell level would only be 5% harder to cast than the level below it. I was also wary of using any kind of existing skill check as I could see it being exploited in the name of optimization (eg, characters taking levels in rogue just to benefit from expertise). If someone manages to cook something up, though, I’d be pleased to see it.
I like the idea of allowing casters to sacrifice hit dice on short rests to recover spells!
Its certainly interesting, but I’m wondering what the mechanical difference between wizards and sorcerers becomes. If this is doing away with prepared spells, then doesn’t a sorcerer just end up being a wizard with fewer spells known? If its only removing spell slots then you get wizards with potentially fewer spells (on a bad roll) than a stock sorcerer.
That’s a fair point, for sure. Maybe one could maintain Vancian magic for wizards to give them a ‘bookish’ feel and use the mana check for sorcerers? I know some tables already use spell points from the DMG to differentiate sorcerers and wizards, and I can see why. They can become too similar in flavour otherwise.
Allow wizards to build Hard spells using two Medium spell checks over two turns or similar representation of “academic” magic with props and formulae and safeguards, vs sorcery’s raw rush. Use sorcery points to reroll failed checks to tweak balance if this option ends up uneven.
I know you didn’t mention them but I feel this would really mess with Warlocks and how their entire class works.
I did mention them briefly! And yes, I agree.