Wizard Subclass Design

Wizard subclasses are difficult to design. The 2014 PHB covers all the schools of magic, the lens D&D traditionally uses to describe a wizard’s specialization. There’s a reason why supplements after the core rules rarely added more wizard subclasses. The key themes present in D&D were already covered.

Theming Wizards

A good subclass has a good theme, but that’s a tricky design goal to hit with a wizard. Most of a wizard’s power comes from their spells. Naturally, we would expect a subclass to alter the spells that a wizard uses. The existing subclasses often do that. The evocation wizard can unleash area of effect spells without harming their allies, a clear upgrade compared to other wizards.

While some games have a robust system of keywords, tags, and other identifiers to sort spells and special abilities, D&D goes for a minimalist approach. There is no easy way to sort spells by elemental magic (air, earth, fire, water), their net effect on a target (mind control, telekinesis), and so on. Spells are sorted by schools of magic, and those themes are already covered by subclasses. What’s a designer to do?

Going back and trying to add a new system for organizing spells is a bad idea. The game has hundreds of spells, none of which were written to fit into whatever new scheme you try to invent. You might find a way to map them all to a set of effects, but as new spells come into the game your players have no easy way to fit them into your scheme.

There are two paths forward. The first one is how I used to design wizard subclasses, and I think it can work but is tricky to pull. The second one is how I’m trying to tackle them in the future. We’ll see if it works!

Approach One: Spell-less Theming

Under this approach, you don’t worry about how a wizard’s spells interact with a subclass. A wizard is free to take whichever spells they want, and their subclass features sit alongside them. The war mage from Xanathar’s Guide to Everything takes this approach and, as you can judge by the amount of war mage commentary on the Internet, no one uses it.

Contrast that with the bladesinger, a subclass that sees quite a lot of play. Why do we see a divergence between these these two subclasses?

Let’s start with the war mage. It focuses on bolstering a wizard’s defense, which is not an exciting promise. Players tend to prefer active, exciting things for their characters. The abjuration wizard also focuses on defense, making this subclass feel a bit redundant.

Looking at its features, at 2nd level the war mage gains a bonus to initiative and the ability to gain a bonus to AC or saving throws. However, the defensive boost shuts down the character’s spellcasting on their next turn. That’s a huge hit. Do I really want to survive a round, while also essentially dropping out of a fight on my next turn?

At 6th level, you gain the ability to increase your damage against a target by half your wizard level, provided that you used dispel magic or counterspell to end a magical effect. First, we’re already veering off the subclass’s theme. Second, that’s a very narrow requirement.

At 10th level you gain an AC and save bonus while concentrating on a spell, and at 14th level you gain some splash damage if you use your 2nd level ability to gain a defense bonus at the cost of doing something impactful on your next turn.

As you can see, this subclass varies from broadly useful but restraining effects, to a weird reliance on specific spells.

In contrast, the bladesinger gives a clear upgrade that pushes the wizard into a warrior-mage archetype. It grants armor, access to weapons, and a core mechanic – Bladesong – that allows a wizard to buff their AC to the point that they can wade into close combat.

At higher levels, the bladesinger gains an extra attack that allows cantrip usage, the ability to spend spell slots to reduce incoming damage, and a bonus to weapon damage equal to their Intelligence bonus.

Those are all notable, fun benefits that create a distinct wizard. It’s wild to think that these two subclasses aim for the general area.

To my mind, the bladesinger’s Bladesong ability is the best model for wizard subclasses that don’t want to interact with the spell list. This design reverses the approach used in the PHB. Rather than create subclasses that interact with specific spells, this approach creates a major new class feature that the player can then use as a guidepost to select spells. The player can consider which spells are good to use with Bladesong and go from there.

Approach Two: New Schools

Since every spell already needs a school of magic, we can consider creating a new school of magic to accommodate a wizard subclass. This approach potentially gives us the best of both worlds. Since we’re designing the new spells alongside the subclass features, we have perfect control over how the two interact with each other. The downside is that we (potentially) have a significant design overhead to grapple with. We might need to design a lot of spells to support a subclass.

Under this approach, you need a clear idea of how your wizard is different from other ones. The new spells you create must justify their existence when compared to the hundreds of spells already in the game. Are they doing something unique? Is there something notable or interesting about their design?

Having designed a lot of spells for various editions of D&D over the years, the chance of finding a new spell outside of the core that people actually want to use is fairly low. In my experience, there are only a handful of spells that have migrated from an expansion to the game’s core. Most of the spells that have entered the game since the 1980s fill specific niches that represent overall design trends, such as low-level area of effect spells (thunder wave) or more liberal healing (healing word, revivify).

I’d recommend making a clear, obvious tie between your subclass and its new spells, to the point that I’d restrict the new spells to that subclass only. With that approach, you can have very explicit ties that make a spell nonfunctional for a character who does not have your subclass.

That might sound overly restrictive, but the wizard already has robust subclass support based on its spell list. I think it’s OK to limit spells in this manner, as it helps your new subclass stand out in a crowded field.

For instance, let’s say you want to make a wizard who manipulates time. That subclass could grant a special action, let’s call it a time step, that emulates the mage shifting time and traveling through it. The class’s unique spells could all be cast using a time step rather than an action or bonus action.

Even better, you now have perfect control over what a character can do with this new action you’ve added. Since you’re designing the only spells usable with a time step, you can control how they work, what they effect, and how a character uses them.

Bonus Topic: Spell Tags

D&D’s spell system works great for describing the magic needed for the characters in the PHB, but it is tricky to extend out to new design. The lack of tags beyond school, which to me is too abstract to extend to new concepts, makes it difficult to introduce whole new types of casters.

Looking back, I wish we had the time and foresight to add a series of simple tags to spells, such as elemental energy source, to make it easier to extend the game with future design. Tags are a great tool, because simply having that space set aside for descriptors gives you space to work with.

The challenge with this approach is that it requires you to potentially confuse your players. The PHB spells could add tags like fire, thunder, and so on to spells, but if none of the subclasses in that book use them then players might think that the book is missing content. If you want to support each tag, then you are potentially looking at producing a ton of new content.

With a new game, this is an easier problem to navigate, but D&D comes with a lot of assumptions and expectations that sometimes make the better design path difficult to implement on a practical level.