Todd Kenreck: Designing subclasses in D&D can be very daunting. They need to not only be mechanically sound, but they also need to add to the story. Every subclass needs an element that can bring something to every adventure.
Jeremy Crawford: With a subclass that gets created for D&D, there are many different stages in the creation process. Balance is a piece of that process, but in some ways one of the least important pieces. We like to talk about the initial drafting phase of a piece of design, so in this case we're talking about subclasses, and then we talk about finishing the design.
Every design, at least the best designs, they start with a story. They start with something that excites you, an archetype, particularly when it comes to a subclass. This is usually going to be a type of person you're going to want to play, whether it is a wizard who's a diviner, or a cleric who worships a god of light, or it's a rogue who's very thieving sort of person, and then therefore is drawn toward the thief subclass. In each of these cases, we're talking about story archetypes, where you can imagine a person. You can often imagine a costume. You can imagine a personality. You can imagine the types of things this person would be able to do, in combat and out. That's the first step, what's the story.
What flows from there, then, are things like the text that describes that story. We sometimes internally will talk about the little story text that goes along with a subclass as sort of its mission statement. It says, "This is what, basically, this subclass is about. Here's a guide, basically, for anyone who's going to play this subclass, to help you role play this character, and also for you to imagine this character's place in the worlds of Dungeons & Dragons."
That story text, though, is also really crucial for us as we do the other parts of our design, because that mission statement for the design will help us, and specifically me, as the person who oversees the finalization of all game mechanics for D&D. I always go back to that story text, essentially. What tale are we trying to tell? And I need to make sure that as well as possible, the game mechanics we then create, and shepherd through our development process, reflect that story. Sometimes, we will end up making changes to the mechanics that push the story in a new direction, and then we go back, and ... Sort of talking on the practical side, we'll go back and actually change that story text so that it now reflects the new story that the subclass is telling.
We'll then, often in the initial creation process, which people have got to see a little glimpse of in Mike Mearls' Happy Fun Hour, on Twitch, where you get to see that sort of early concepting phase, of let's spitball something, let's come up with a story, and then let's brainstorm some of the little class features and other mechanical nuggets that could express that story. And it really is, in a way, you just let your imagination run wild, with some limits.
One of the big limits, even in that initial phase, is you're playing in kind of a playground that already exists. Any time you create a subclass for a particular class, you want to make sure you're familiar with the subclasses that already exist for that class, because that starts giving you a sense of the mechanical heft that a subclass has for a particular class, because the secret is not all subclasses are created equal, especially when you compare the subclasses of one class to the subclasses of another.
If you look at a cleric, divine domain for instance, and compare it to the barbarian's subclasses, the subclasses have different mechanical heft. And what I mean by that is they're pulling a different amount of weight, story-wise and in terms of the capability of that character in each of those classes. What you can do with a domain, even sometimes when it comes to something as concrete as damage output, might be heavier in the subclasses of one class than in another. Some classes, a lot of their mechanical heft, their damage output, their healing output, their abilities in an exploration context, or a social interaction context, sometimes a lot of that weight exists in the core class, whereas you'll come to another class where a lot of that weight actually exists in the subclasses of that class.
So that's why it's really important for you to understand when you're designing for a particular class, sort of where in the power spectrum do the subclasses of that class tend to fall relative to other classes. The last thing you want to do is design a path for the barbarian, to go back to my previous example, and give it a power level that's appropriate for a cleric domain. You want to make sure you're giving it a power level that's appropriate for a barbarian subclass. So that's sort of step one when you start. You go from story concept to start dealing with mechanics, when you're past the spitballing phase.
The next thing you want to do is you're going to want to make sure you're filling in any special buckets that a class might have in its subclasses. Clerics, again I like ... Clerics are a good one to use, because they have ... There's a unity in how cleric subclasses are designed. Cleric subclasses all need to deliver at least one channel divinity option, so you've got to make sure already, in your design, that you're hitting that checkbox. Cleric subclasses also need to give you either divine strike or potent spellcasting. So, you need to make sure when you're designing for a class that you're taking care of those sort of infrastructural things.
Now, that said, D&D is an exceptions-based game. You could technically design a cleric subclass that doesn't give you divine strike or potent spellcasting, and gives you no channel divinity option, might give you something else, but your design better be aware of the fact that it's not giving you those things, and where necessary, give you something else to compensate. So you need to make sure when you're designing that you're very aware of the environment that you're designing in, and when it comes to subclass, the environment you're designing in is the class you're designing for.
Okay. So let's say we have a series of class features that you've spitballed, you've done your homework, you've made sure you understand how the class works that you're designing for, you've studied other subclasses for it. One of the things you're going to want to do is see, do you have any duplicated design? Did you design something that is actually really close to something that already exists?
Now, I'm going to say something now that might surprise people. Duplication is not in itself bad. Sometimes, duplication can actually be really good. However, subclass is a place where duplication, we have found through playtest feedback, can be a little dangerous, because people really like the character archetypes that subclasses represent to feel distinctive, and players will often get cranky if they feel like one subclass is stepping on the toes of another. So, you need to be careful in subclasses with duplication.
That said, it is fine ... Once you've kind of dived into the nuts and bolts of a particular feature, it's okay if it's reminiscent of another feature somewhere, if it's appropriate and you don't overdo it. Here's what I mean. You could have a single feature, and in fact, just yesterday I did some development work on something that was concepted in a Happy Fun Hour, and started to push it toward more of a final version. One of the new features I wrote for it, and it was for the ordered domain for the cleric, was a feature that is a little reminiscent of a wizard feature. That's fine. However, if I had sort of copied the entire shtick of another subclass, that would be pushing duplication too far. So it's okay to use tech that already exists in the game if it hits exactly the target you need.
You do need to be careful, though, about naming. Don't use the same name for something unless it is indeed meant to be the exact same game mechanical object. For instance, some classes share the feature evasion, and they are all called evasion. That's on purpose. Uncanny dodge appears in more than one place. That's on purpose, but only use the name more than once if you really mean it.
All right, you have your features. Then people start wondering, are they appropriate? This is where it's good, go back to your story. Make sure that all the features that you now have, some of which you might have gotten really excited about, because it's like, "Ooh, this is a really interesting game mechanic." Here's what I recommend. Always be very suspicious when you find yourself saying, about your own design or someone else's design that maybe you're helping them with, be suspicious when you find yourself actually saying, "That's a really interesting game mechanic."
Far better for your feeling to be, "This tells an exciting story," because what we have found is that when we're being kind of armchair game designers, and also armchair game fans, and I say this as a person who's a fan of a lot of games. I play a lot of video games. I play other role playing games. I play a lot of board games. I'm very interested by a lot of different game mechanics. You know, this is sort of my game nerd, nerds out on, "Ooh, that's a fascinating mix of elements."
Here's the thing. We have found, looking at the playtest feedback of hundreds of thousands of people, for 5th Edition, over the last ... Wow, it's already five years. What people want is an exciting story. Interesting doesn't necessarily always deliver the fun, and doesn't necessarily always deliver a compelling tale. So what you want to pause and think is does this class feature that you're giving a particular subclass, or even a base class, if let's say you're going wild and you're making a whole new character class, does it tell an interesting story, a fun story, a compelling story for this character, in the world?
What's going on in the world? That's often the question I pose to something I design, and when I pose ... a question I pose to the design that I'm evaluating by other people. What's happening in the world of Dungeons & Dragons when a person uses this ability? What tale is going to arise at the game table because of the use of this ability? Does this ability help me visualize this character in the movie in my mind that's going on of the potential D&D sessions? Does it help me visualize them better because of the wonderful storytelling that's hiding out in this design?
Now, not every class feature is going to be a stellar pillar of game design. I mentioned the cleric's divine strike ability, which is basically you deal some more damage. So not every feature has to deliver, you know, exciting storytelling all on its own, but it's a holistic thing. You want to look at the features, how do they all gel together? Are they all, together, contributing to a great story, and specifically a story that is true to the archetype that the subclass is expressing?
If the features pass the story test, and they actually look like they're going to be fun, even better if you actually try them out in play yourself. You know, have a quick encounter. If it's a combat ability, a quick social interaction, an exploration scene. Give it a try. Often, a quick try of the thing will tell you if it's going to work or not, if it's too complicated, if maybe it's flimsy and maybe could use some more meat on its bones. Those are all important things to assess.