S is for Spell Selection

This is my 6th post in my Building a Better Villain thread that runs throughout April A to Z

Just a warning to non-gamers, this post is a bit crunchy (heavy on gamespeak and rules.)


Spell-casting in D&D is an odd beast. Up until 4th edition, the game used a system commonly known as Vancian magic* With 4th edition, the game scrambled it’s spellcasting, shifting to a more Magic the Gathering-style system of powers with set recharge rates, combined with ritual
(for magic-users, these were still called “spells”, but were no longer unique from the new powers given to fighters or other non-magic classes).

For this post, I would like to focus on managing spells in the older Vancian system. Vancian magic basically treats spells like ammo. You have a specific number of spells you can cast each day, and for wizards and clerics, they must be prepared in advance at the beginning of that day. This means that if you want to be able to cast 3 fireball spells, you need to memorize the spell 3 times. A friend of mine described it as fixing the magical patterns in the wizard’s brain. When the spell is cast, the pattern burns itself out in the process.

Most players try to select spells to cover a broad range of potential situations. Since they don’t  know what the GM has in store, this is a wise way to go about it. For a GM, however, taking this same approach might not be the best choice. Players spend a lot of time with their characters, they get to know the spells they have at hand, and become familiar with their mechanics. The GM, by contrast, may roleplay several NPC spellcasters during the course of a single game session, and several more the next game. Becoming familiar with diverse spell sets for each of these NPCs quickly becomes cumbersome for any busy GM who hasn’t had 20 years to memorize the Player’s Handbook by rote.

As such, I recommend taking a different approach with NPC spellcasters. Don’t try to optimize them to meet any possible challenge. Optimize them instead to a.) present a particular challenge or asset to the PCs (depending on their role as support or adversary) and b.) give the magic user a distinct style (e.g. pyromancer, summoner or mentalist), characterized by fewer spells with more of the same spell prepared. This gives you fewer spell descriptions to memorize and adds a distinct feel to the NPC. I would try to give them a couple protective spells, a couple attack spells and a few interesting effect spells they can throw into an encounter. In general, a caster will use more attack spells during combat than defensive spells, as the latter typically get spent while buffing up in the early rounds. Therefore, it is important to have at least 1 low-level attack that can be cast a bunch of times, while fewer self-defense spells might suffice.

Here’s an example:
A 10th level wizard will likely be able to cast at least 23 spells each day. Rather than giving them 23 different spells, do something like this:

10th level necromancer
0 level (4 spells) Detect magic x2, Touch of fatigue x21 level (5 spells) Ray of Enfeeblement x3 (low level debuff) Shield x2 (defense)2 level (5 spells) Melf’s Acid Arrow x5 (low level attack)3 level (4 spells) Vampiric Touch x4 (stronger debuff/healing)4 level (3 spells) Animate Dead x3 (summoning)5 level (2 spells) Cloudkill x2 (powerful area of effect)

With the above example, you only need to be familiar with 8 spells instead of 23. It still has a mix of direct damage attacks, area attacks, defense and other interesting effect spells, but should be much less cumbersome to run. The above necromancer also focuses on spells that either drain an enemy’s life force, or summon undead to assist the character. You could also throw in a couple extra spells that have simple mechanics or that you are already familiar with if certain types of spells are missing (e.g. mage armor to further beef up defense)

*named for its similarity with the system found in Jack Vance’s Dying Earth series.