This is the fifth in my series of class-focused d20 Design Diaries. I suspect I only have a couple more posts to go on this topic, but we’ll see how the topics actually shake out (and what kind of feedback I get).

If you followed class design steps in the order I have written about them, there’s one big step left to actually creating your class, even after you settle on an appropriate and interesting concept, set up the right class progression tools, made sure you are following (or at least only breaking by intent rather than by accident) the game’s style and etiquette, and looked at how many options you want for each level of your class and how that impacts complexity.

You still need to design the actual class features, the special abilities you class gets that (at least mostly) others don’t.

I mean, technically you don’t HAVE to give a class features beyond it’s progressions. If you gave a Starfinder class 10 SP and HP/level, all good saving throws, 12 skill points + Int/level, any key ability score, all class skills and weapon and armor proficiencies (and Weapon Specialization as appropriate), and a full attack bonus, it would honestly probably be pretty balanced with no other class features at all.

It would also be boring and flavorless as heck. And I have no idea what concept you’d start with that would lead you to that design. but yes, it COULD be done.

And that does touch on an important element of designing interesting and balanced classes — the more useful things the class gets outside its class features, the less room you have to make its class features useful without making the class overpowered. A 5e barbarian has d12 hit dice, and 2 skill proficiencies (selecting from 6 options) and 5 weapon and armor proficiencies. A fighter has d10 hit dice, and 2 skill proficiencies (selecting from 8 options) and 6 weapon and armor proficiencies. A rogue has d8 hit dice, and 4 skill proficiencies (selecting from 11 options), one tool proficiency, and 2.5 weapon and armor proficiencies. It’s not hard to see that while their proficinecy starting points are different, when combined with their hit dice they all come out on a fairly even playing field, allowing their classes to have equally-useful class features.

One of the biggest and most impactful potential class features is spellcasting. Assuming you are building classes for a game that already has a full set of classes you can use as examples, it’s normally best to stick to the spell progression and acquisition schemes that already exist, unless you feel it’s a severely underdeveloped design space. (Classes with some number of spell-like abilities are a different matter than the spellcasting class feature we are discussing in this article.)

For example, first edition Pathfinder has both spontaneous and prepared spellcasting acquisition, as well as spell lists that go from 1st-4th level, 0-4th level, 0-6th level, and 0-9th level. However, every spontaneous class in Pathfinder with access to a 0-6th level spell list has the same base access to spells known and spell slots per day (though OTHER class features, such as domains or archetypes, can vary their total beyond the simple base). Starfinder, on the other hand, *only* has spontaneous spellcasters with access to 0-6th level spells. While adding a whole new spell progression or access to Pathfinder would likely muddle a crowded field, there’s easily room in Starfinder for class with reduced spell access (perhaps level 0-3 spells).

The more spell power a class has, the less room it has for any other options. For example, in all the most popular d20 games classes with the greatest spell access never have the highest Hit Point/health value of classes, or beginning proficiency with all types of armor. This has two significant impacts on their design. First, it means that they generally need to use some of their spell power to bring their defenses up to their best level and, even at that level, it’s generally not as good as the best defenses of the most defense-focused class. Secondly, it means they aren’t as durable without depending on their spells (and even then some classes with major spell access have very little in the way of healing or damage mitigation spells — a 1st edition Pathfinder cleric can heal themselves much more easily than a wizard).

Again, using other classes as benchmarks can be extremely useful for making your first stab at granting spellcasting to a class. In 5th edition D&D, paladins and rangers gain up to 5th level spells, clerics and wizards gain up to 10th level spells, and specific specializations of fighters and rogues get up to 4th level spells. Those benchmarks make it pretty easy to see what kinds of class features, both in terms of scope and utility, a class with each of those options can gain. For example, a great deal of the class features of sorcerers and wizards are focused on their spells–allowing them to be more flexible, used more often, or even just boosted in power. Paladins and rangers however, have very few spell-focused class features, with their class features more likely to actually give them entirely new abilities.

Even once you know how your spellcasting class is going to acquire spells and to what degree, there still another crucial question–what spell list do they use?

We’ll tackle that one next week.

PATREON

These Design Diaries are among the most popular of the things I wrote, but they are also the biggest, hardest, and most time-consuming to create. If you want to keep seeing them, I encourage you to join my Patreon. Just a few dollars a month can make the difference between me having the time to tackle these larger, in-depth design articles, and sticking to shorter, simpler topics.