Subclasses, and What Came Before

This is DMing with Charisma’s 300th post. I was planning on an important, sweeping post describing how my last campaign affected me and how it will affect all of my campaigns going forward from the standpoint of the table experience and the game’s social contract. But instead, let’s talk about something I do like about 5E: subclasses.

Traditionally, a character’s class affects his place in the game more strongly than his other choices. His ability scores affect how good he is at things, and his feats affect specific corner cases, and race can be very relevant if you’re a terrible person, but a class lays out most of what a character does. It determines whether he’s good with certain kinds of equipment, whether he can cast magic, how easy it is to grow in other areas like skills, what things he can do that nobody else can, and often his role within the party. Obviously it’s not as simple as “all fighters have good AC and swing swords and can’t use spells”, but as players we do get a certain picture in mind when we think “fighter”. It’s exactly why I ask my players to describe their characters without using class names, because we have those mental shortcuts telling us how each class works in the game.

Classes are a fundamental part of D&D. You can’t play without them and you can’t ignore their place in the rules*. But you can take the concept of classes and supplement them, and each edition handles that in its own way.

3E liked prestige classes, and I liked prestige classes in 3E. A prestige class functions as a class except a character can’t use it unless they meet some prerequisites, usually ones a character can’t meet until he’s at least 5th-level. Prestige classes are much more specific than a normal base class and they usually have unique abilities relevant to their role in the campaign setting. Once a character meets the prerequisites, whenever he gains a level he can advance in the prestige class instead of his base class, up until the prestige class runs out of levels.

For example, consider the arcane archer. In 3.5E, the arcane archer was a prestige class a character could only take if he was an elf or half-elf, he had a high enough Base Attack Bonus, he had three specific feats, and he could cast minor arcane spells. There were several ways to meet these; the most common was six levels of ranger and one level of wizard, but a character could substitute fighter for ranger to get extra armor and feats, or substitute sorcerer for wizard if he didn’t want high Intelligence, or be a bard the whole time and take the prestige class later but avoid multiclassing. For being an arcane archer, he gets several abilities that enhance his arrows, like the ability to shoot through walls or use an arrow as the origin point of a spell.

Prestige classes were neat, but they weren’t perfect. To me their biggest issue was that they usurped the role of classes. Players didn’t want to be a boring fighter or sorcerer, not when they could be a could be a fighter who mastered exotic weapons or a sorcerer who sprouted wings and got cleric spells. Prestige classes defined characters, and classes were just the temp jobs they took on the way to the career they wanted. Add in that prestige classes weren’t always well-balanced (see the above “sorcerer with cleric spells”), there were far too many to reasonably parse (the 3.5E Dungeon Master’s Guide alone had sixteen, and several later books each added anywhere from ten to twenty-five), and each class was supposed to come with a unique role in a campaign’s setting, and the cognitive bloat became too high to manage. Prestige classes tried to make characters special, but as The Incredibles taught us, when everybody is special, nobody is.

3.5E also had a thing called substitution levels, where a character could switch out specific powers from specific levels of their base class to gain other specific powers in service of a theme. For example, a L6 rogue could give up one point of trap sense for the ability to detect nearby planar breaches. It was cute, but the designers didn’t commit to it enough for it to be worth exploring.

Pathfinder does a much better version of this idea with archetypes, which are like substitution levels except you have to take all the substitutions for a given archetype instead of cherry-picking which substitutions you want. You can have several rogues who share some core rogue abilities, but they all feel different. This one’s terrible with traps but gains traits related to his time in organized crime, while this one can’t sneak attack but has the best skill ranks in the game, and so on. You don’t play a rogue for a little while on the way to becoming an assassin, you’re always an assassin-type rogue, and there aren’t any hoops you have to jump through to do it.

But archetypes manage to have the same bloat problems prestige classes did without the fun character-building aspects to excuse them. The rogue has sixty-five archetypes and they’re not all mutually exclusive. It is possible to calculate how many possible combinations of rogue there are but I do not have the patience for it. If you want to find rules to fit the rogue you have in mind, you still have to comb through several books or dozens of web pages looking for the perfect match. When you’re done, you could end up with something that doesn’t even look like a rogue any more, which sort of defeats the point of using archetypes at all. But archetypes are class-exclusive, so you can’t even use the vast array of options the game has. While your arcane archers could be ranger/wizards, fighter/sorcerers, bards, or several other things, your rogue archetypes are only available to rogues, so you can’t find a clever way in. And each archetype still occupies a unique place in the setting, so archetypes didn’t solve the narrative problem.

4E fixed this with a chainsaw rather than a scalpel, by eliminating most of the choices from classes at all. It very nearly banned multiclassing, hard-coded almost every class feature, and forced characters into paragon paths (like prestige classes) at L11 and epic destinies (like super-prestige classes) at L21. When players could a class feature choose from a list of options, like deciding a wizard’s favored implement, those features didn’t functionally change the character; whether a wizard uses an orb or a wand, she’s always a controller, always uses Intelligence, always has the same skill options, and always picks from the same power list. In fact, she probably picks from a reduced power list because some powers are designed for a single implement, meaning they’re less powerful for any other wizards. Most classes have this problem, and if there’s anything 4E can’t stand it’s making a non-optimal choice. Further, before Essentials came out the class features only affected what happened at L1, so the features themselves had no growth. The bloat is mostly gone (except for paragon paths, of which there are a billion), but it took the creativity of class choice with it. And we still have the narrative and balance issues.

To solve all these problems, 5E needed a system that gave players balanced options (but not too many options), made those options matter (but not matter so much that they made the class itself irrelevant), didn’t shove dozens of new groups into the campaign setting, and allowed the freedom to design characters instead of merely members of a class. Its subclass system actually mostly succeeds.

In 5E every member of a given class gets their class features and they must also choose a subclass that gives them other features at certain levels. In this respect it’s like 4E with Essentials, except the choices matter without overwriting the class itself. All clerics have similar features, like spell casting and turning undead and an L8 feature that increases their damage, but their automatic spells, their minor features and the nature of their bonus damage change with their subclass. Classes themselves are broader than in 4E so they can function in different roles, which means we aren’t tethered to the designers’ opinions of what a fighter should want to do. The bloat isn’t as bad because the flow of new content is much, much lower than in past editions, so I’m hoping the long-term growth is limited enough to be manageable. I can’t comment on balance issues because I don’t yet know enough about 5E to judge it at all levels, but so far I see it somewhere between 3E’s problems (let players do everything) and 4E’s suffocating solution (never let players do anything).

What I like best about subclasses is how vague they are. They aren’t things like “knight of the Gray Guard” or “alienist who deals with the Far Realms specifically” or, so help me, “Fochlucan lyrist.” The fighter’s subclasses are the champion (guy what does lots of damage), the battle master (guy what has neat tricks) and the eldritch knight (guy what casts wizard magic). Is the champion a gladiator, a mindless brute, a grizzled veteran, or a serene warrior who gains power from meditation? The subclass doesn’t care. It gives you rules and a brief flavor text to the effect of “the champion hits enemies real hard” and lets you figure it out. The characters available with a given subclass far exceed the characters available with any prestige class or paragon path even without an ounce of reskinning, and I love seeing that out of the official Core rules.

I’m really excited about looking into custom subclass design. While the design space for prestige classes and archetypes is too broad and the space for paragon paths is too narrow, subclasses feel just right. You know exactly when a subclass gets its abilities and those abilities drive character action more than in 4E. A subclass could have as few as three abilities, which is an excellently low cognitive barrier, and the Core books show how widely they can vary even within the same class. The official Unearthed Arcana series has been putting out new subclasses sporadically to push the limits of the system even harder. I figure it’s just a matter of time until a player asks if it’s alright to play a subclass they wrote, and honestly I can’t wait.

I wholeheartedly endorse the subclass system. No matter how I look at it, it’s good for players. I do notice that the player experience for 5E is fun and straightforward while the DMing experience is limited and frustrating, and I also notice that most of the books released for 5E are adventure modules designed to limit the role, impact, and creativity of the DM. I’m trying not to read into that.

* — Technically, there are variant systems that let you play D&D without classes. The ones I’ve seen range from “catastrophically bad” all the way up to “ill-advised and possibly physically harmful.”

This entry was posted in D&D 3.5, D&D 3rd Edition, D&D 4th Edition, Gaming Systems, Pathfinder and tagged . Bookmark the permalink.

One Response to Subclasses, and What Came Before

  1. Blake says:

    I enjoyed reading this.vNice return to form.

Leave a Reply

Your email address will not be published. Required fields are marked *