I touched briefly on access to magic in the Power Variety post, but I had set aside writing about it at length for this post.
For those who don't know, the term "Vancian magic" comes from the original mechanics of spellcasting in D&D having been inspired by Jack Vance, particularly his Dying Earth series. Spells are powered by complex formulae which effectively implant a packet of knowledge/energy into the caster’s mind when prepared and which is lost when the spell is released.
In my opinion, the main impact this has in play (when compared with mechanics using spell points or magic-at-will) is that spellcasters retain a high degree of strategic flexibility while having constraints on their tactical flexibility. What I mean by that is a spellcaster has the full breadth of their spell arsenal at their disposal when it comes to answering the question of "what do I want to use on this adventure?", but that breadth reduces to just their (remaining) prepared spells when it comes to answering "what do I want to cast at this particular moment?". Thus, when considering what, if any, adjustments should be made for spellcasting in DM+1 play, it's important not to veer too far out of these bounds if you're concerned with maintaining a game experience compatible with regular D&D (not wanting to maintain that is valid in itself but outside the scope of this series).
By extension, the main drawback of sticking with Vancian magic in DM+1 play is the limited quantity of spell slots available per day, particularly at low levels. By standard D&D rules (from OD&D rules through to 2nd edition, excluding specialist wizards), a first-level magic-user has only one spell slot. Being able to cast just a few spells per day can undermine just how “wizardly” a wizard feels even as part of a group, let alone in a DM+1 game.
Looking back at what I'd said in Power Variety (which was mostly from the perspective of non-spellcaster PCs running into an obstacle requiring magic to overcome), I'd presented three options: change nothing for the PC, provide some limited access to magic, and be less restrained on providing magic items. Let me go over how each of those options interacts with the magic use of spellcasting PCs.
Changing nothing for the PC is actually a huge decision, because it maintains the greatest amount of discrepancy between classes. This can be good. If the PC is having a DM+1 adventure as a break from being part of a larger group that they will return to (as Matt Colville talks about doing), you probably shouldn't be changing how that PC plays just based on how many allies they have. Also, the player is communicating some part of their desires when they choose a certain class, regardless of whether they're basing the choice on narrative ideas or game mechanics. Avoiding muddling that class's playstyle can be a way of honoring the player's choice. Lastly, the constraint of not being able to do something can help to spur creativity.
Of course, like any meaningful choice, it's not without flaws. This approach puts the greatest burden on the GM to consider the fairness of the PC's obstacles. I'm not talking about that in some rigorously detailed way but rather at the high level of "does the PC have the potential resources to overcome this challenge?". It's entirely unfair to run Night Below if the PC won't have access to wizard spells (either directly or via items/allies), ideally including being able to pick what to cast, since there's a core puzzle related to that. It's mostly unfair to run L2: The Assassin's Knot without spellcasters, unless the GM is very generous about social manipulation or possibly if the player is a detective, because the timeline makes it highly unlikely for even a group to solve the mystery without magic, let alone a single player. A seemingly-insurmountable obstacle is only interesting if there are ways of handling or circumventing it that the player has (or has the potential to have) access to; otherwise, it's just a wall with another name.
It should be obvious, but this approach also does nothing about the previously-mentioned drawback of the PC having so few spells available per day that they may not “feel” like a spellcaster in actual play. I’ll mention some ideas for addressing this later in this post, but I’d advise against using them if the DM+1 game is a subset of a greater campaign for the aforementioned reason of mechanics consistency.
To a lesser degree, this also limits the scope of narrative possibilities. There are plenty of examples in inspirational media of either fighters with a little magic (e.g. Geralt of Rivia) or spellcasters with noteworthy martial ability (e.g. Elric of Melniboné) going on adventures that require both sets of skills. While sticking with a strict division of class abilities doesn't rule out playing similar adventures categorically (even if only through multiclassing), it does require more effort to make them work.
One compromise is to allow the PC ways of gaining limited access to more magic. As mentioned previously, this comes up in other media, typically because of a special ancestral bloodline or making deals with supernatural creatures (and I'll give WotC credit for bringing both archetypes into official D&D as sorcerers and warlocks, respectively). I've also had it come up as a result of either consuming parts of a supernatural creature or the PC investing resources into studying under a teacher. In all cases, I've found there are two keys to making this work well: keep it simple and keep it limited.
Keeping it simple means to minimize both interactions with and references to other mechanical elements. A character who brokers a deal with Explicitica Defilus might gain the "ability, by locking gazes, to influence another person's mind to obey a request once per day (save negates)". That's it; no need to grant/use spell slots or refer to formal spell text(*). When adding mechanics to a game, it’s always important to give some thought to how they’ll mesh with existing mechanics, and the easiest way to avoid unintended consequences is to isolate the new rules as much as possible.
(*) Keying the ability to an existing spell (e.g. the Seelie queen’s blessing grants Purify Food and Water once per day) isn’t terrible, but it’s something I try to avoid. This is mostly part of my general belief that minimizing look-ups/reference is good practice to reduce disruption in play, admittedly.
Keeping it limited means that it shouldn't cause a major change in what the PC does, neither at a mechanical level nor at a narrative/roleplay level (unless it was intended, e.g. a priest falling from grace by swearing an infernal oath, but that’s outside the scope of DM+1 basics). This can achieved by limiting power, reducing accessibility, or adding casting time/component requirements. The new magic should be an accessory to what the PC can do on their own, an extra tool in their arsenal rather than a defining trait, and it shouldn’t break the general paradigm of spellcasting providing great strategic flexibility but constrained tactical flexibility.
On a related note, I find that being more generous with magic items (whether consumables like potions, scrolls, wands, etc. or more permanent like a helm of teleportation or a flying broom) is generally less of an issue with spellcasting PCs. Remember, by choosing to play a spellcaster, the player is signaling that they want their character to feel magical, and inspirational media tends to have far more examples of mages and mystics with troves of weird concoctions or items enchanted for specialty/utility purposes than seeing the same with mundane warriors. This still needs care to avoid overshadowing the PC’s own capabilities, lest it turn into a crutch like Jafar’s staff, and it isn’t something I’d go to if the DM+1 game is a subset of a greater campaign, but for a dedicated DM+1 game, I’ve never had this be a problem. At worst, an overused magic item will tend to draw attention from NPCs who’d want it for themselves, after all.
Now, what about the drawback of very limited access to spells at low levels? There are a few ways I’ve thought of to expand access without abandoning the Vancian structure, though I haven’t used most of them in actual play, and I have come to reject the ones that I did try in favor of the aforementioned approaches. Even so, I want to include them for the sake of completion. Consider the following as ideas for further contemplation rather than practiced advice.
The most straightforward is to increase the availability of spell slots, either by increasing the total number or by decreasing how long it takes to restore them. While this is simple, I’m not a fan of it since it undermines the tactical restraints that I consider a core benefit of Vancian magic. Still, something like having one extra slot per spell level can help the character feel more magical without causing major changes in the play experience.
Somewhat more complicated is having a chance to not expend the spell slot, typically by succeeding on a saving throw or ability check. My concern here is that it’d be prone to either having no effective impact or undermining the resource management aspect of Vancian magic with little middle ground. Saving throws are particularly troublesome since, in most systems, they tend to fail at low levels and succeed at high levels, so they won’t really address the limited spell slot issue while it’s important and will instead tend to make spell slot tracking into tedious busywork at a stage when they’re so abundant that the hassle isn’t justified. Ultimately, I see either of these as adding extra rolls to play without giving a net benefit in play.
Synthesizing those two ideas leads to the system proposed in the Lamentations of the Flame Princess supplement Vaginas Are Magic: the character's spell slots represent the spells they can cast safely, and trying to cast more beyond that point requires saving throws to avoid losing control. I like this more than the previous ideas since it gives the player an option to push their luck while keeping their odds of success low enough that it should remain as a relatively desperate resort until the character reaches a level where they have enough spell slots to not run out regularly (assuming typical progressions for spell slots and saving throws). I’d be inclined to include the attempted spell’s level as a penalty to the saving throw so that the choice retains not-insignificant risk even at high levels, but otherwise, this seems like a reasonable compromise.
Lastly, I want to mention two ways of beefing up Vancian magic without altering the structure: adding secondary effects to prepared spells and making all spells scalable.
Adding secondary effects is about giving spellcasters at-will invocations (with or without some associated roll) based on their prepared spells. This is something that I’ve touched on previously with elf heartspell mutations. The idea is to give a way for the character to express their strategic choice of which spells to prepare without requiring them to expend the spell. I’ve used this within the specific context of heartspells and have been pleased with how it’s worked out. For expanding the idea to work with all spells, I’d avoid defining exact effects for each spell. Instead, I’d take it as a freeform quality of magic as a whole, treating the minor effects similar to cantrips in Beyond the Wall, and give the choice of either having a magical mishap or expending the spell slot to avoid a mishap in the event of a failed ability check.
While certain D&D spells (like the iconic Fire Ball) have always had effects that scale with the caster’s level, the trend really took off with AD&D, and there are various homebrews out there that have tried to apply that approach to all spells, reducing them to first-level equivalents with effect-scaling formulae. This seems interesting, but I haven’t used it because I just don’t like the idea of spellcasters being able to conjure extradimensional creatures or teleport or bring back the dead straight from their starting baseline of capability. I might be more open to it in specific campaign contexts, but as a general approach, it’d risk throwing out the logistical/experiential evolution in how the game feels based on the (character) level of play.
That concludes my basic advice on Vancian magic in DM+1 play and indeed on DM+1 play as a whole. As I’ve said before, I’m willing to talk more about certain topics (either in comments or with further posts) on request, but as with all things TTRPG, all the talk in the world doesn’t really matter unless it’s backed up with actual play experience. I hope that I’ve shown some insights on how to transition effectively from group play to DM+1 play, and I hope that I’ve sown some seeds for ways of getting more out of DM+1 play than you would’ve expected. Now, go out and try it for yourself!