Classes: Breadth vs. Depth
Been thinking about how a D&D-type game approaches classes can be roughly plotted on two axes: class breadth & class depth.
Having high 'class breadth' means that there are many classes tooled to specific character concepts and implies that classes are meant to mechanically represent the diversity and differentiation of characters. Low class-breadth games use classes to convey broad archetypes (if they have them at all) and rely on mechanics outside of character class to differentiate characters.
The high end of the 'class depth' scale represents games where a character's class grants them lots of unique mechanics - special abilities, powers, feats, spells, etc. At the lower end, a class might only define a hit die and attack bonus progression.
For example (from games I've played in some form):
- 5e D&D is high on both axes. There are many classes and archetypes within the classes, and each class has 20 levels of special abilities, feats, etc.
- B/X & OD&D have lowish breadth (with only a handful of classes) and low depth (the classes provide a few unique powers but otherwise just cause numbers to go up).
- The GLOG has staggering breadth (with classes like Cube Owner or Surly Gnome) but low depth - each class typically has a handful of powers.
- Worlds Without Number has high class depth with its skills & foci but can take class breadth in either direction depending on if the Deluxe classes are used or only the Warrior/Expert/Mage base classes. I'd argue that the latter case is a rare-ish example of a low-breadth / high-depth game, where you get lots of 'build options' to customize within three very broad archetypes.
- Knave, Cairn, and Into The Odd are low on both axes.
Applying this schema has led to a couple vaguely useful insights:
As a game element, classes set a lower bound on how unique your character can - high-breadth games allow for unique characters from the get-go, high-depth games allow players to specialize into unique roles. In either case, the class is a list of what you are guaranteed to get if you do the bare minimum by showing up and earning experience. Accordingly, they constrain narrative coherency; RAW a 5e character can become an archimage simply by killing monsters. This cheapens things, but it also saves the GM having to hand-place spellbooks for the players to find. In a game like Cairn, you aren't guaranteed anything, and every special ability you do acquire comes from a specific event in the fiction.
Breadth gets problematic in the middle regions of the axis. At the low end, classes draw hard boundaries around commonly accepted archetypes (e.g. fighter vs. mage). In the fiction we assume that the palace guards, the wandering hedge knight, and the bandit chief all fall under the fighter archetype, their differences accounted for by diegetic elements (their equipment, special training they may have recieved, etc.). At the highest end (as in GLOG), the palace guard, wandering hedge knight, and bandit chief could all get their own class; the list of classes incorporates traits that a low-breadth system might treat as contingent. And in the middle, things get wonky - why does 5e have both a nature cleric and a druid class? What, really, is the difference between a wizard and a sorcerer? How powerful must your patron be to graduate from warlock to cleric?
For Archons & Armigers, I tried to stay fairly low on both axes. I want the 'Armiger' class to cover pretty much any martial archetype, for example, and I want characters to fit on notecards (so no list of class features or feats).