Fitting within limitations is a fun thought-experiment, when it comes to design. Figuring out which classes fit with which dice thematically is one thing, but so is fitting together dice mechanics with class abilities. Something I did in preparation for Beta 4 was to work out some math (which I have talked about briefly in other places) pertaining to "roll and compare" mechanics.
What I figured out was a set of unique mechanics for a couple of the class dice, that generally would be applied to "support class"-style abilities.
- d4: When an unsuccessful d20 roll is made, roll this class die; if the d20 result does not exceed the d4 result, treat the roll as a critical success.
- d8: When an unsuccessful d20 roll is made, roll this class die; if the d20 result equals the d8 result, treat the roll as a critical success. Otherwise, combine the rolls together.
- d12: When a d20 roll is made, roll this class die and use either result; if the d20 result equals the d12 result, treat the roll as a critical success.
Effectively what each of these mechanics do, is to turn rolls that would normally "miss" into successes or critical successes; mostly they do not turn successes into critical successes.
Now, while you can use other dice to produce similar outcomes, my experimentation found that d6 generally ended up working as either "same as the d4 but stronger" or "same as the d8, but clunkier." The same held true of the d10, when compared to the d8 or d12. This sort of works out ok, because generally the intent was to use d6 and d10 as damage-stacking classes, rather than support classes; this math just serves to outline how that kind of "ergonomic" thinking actual dovetails logically into the mechanics (i.e. every set of polyhedral dice comes with two d10s; every boardgame in history uses one or more d6s, so most households have extras they can dip into.)
Upon this, we can layer the use of class die as an effective cap on the range of results. So, for example, the Monk as a d4 class can roll their class die and make a number of attacks equal to the result; this means they will be doing 1-4 attacks on their turn, which we can modify further (with mechanics like Expertise or Advantage) but still will never exceed that range. The Beta 3 version of the Mystic had a similar mechanic, but using d10s with Disadvantage; this ends up a bit clunkier, and with a higher (if statistically less likely) top range. Also, the d4 "support" mechanic specifically is a straight increase in crit chance rather than "hit" chance, which makes it distinct from the other support mechanics.
Likewise, d12 as a class die translates to higher HP and more Reserves, thus informing that these classes should generally be inclined towards the middle of the fray, with perhaps fewer options for outright damage mitigation. So, it was decided back in Beta 3 that d12 classes' niche should be in melee, or possibly built around burning those resources (HP or Reserves) to fuel more-powerful effects. A unique thing about the d12 is that it's the only class die that dips into the "hit" range (aside from the d10, which ends up a little weird, being that it can only "hit" on one number.)
With d8, we end up with something that's sort of "middle of the road." It works well as a support class, if we use the class die mechanic introduced above, but it also works well (I think) for a class that is meant to dabble between melee and range (as the d8 Ranger class has done, since the beginnings of The Next Project) but leans more on the "core" dice mechanics. The d8 is interesting in that its range covers most of the "miss" range on the d20, but none of the "hit" range.
So if d10 and d6 are meant to mostly be damage-stacking classes, this gives us two "flavours" for that sort of design space; one that can take a few hits and mix it up in melee, and one that probably can't/shouldn't. The existing Sorcerer (d6) archetype and Paladin (d10) class from previous versions illustrate this the best, I think.
Taken on the whole, the various mechanics paint a roadmap to how any future classes should be built:
- d4: multi-attacking classes / high crit support
- d6: ranged damage-stacking
- d8: "skirmishing" style combatant / good hit support
- d10: melee damage-stacking
- d12: melee combatant / minor hit support
Building upon the foundations established in previous posts, we can fiddle with this a little bit, by letting d4 or d6 classes use 2 class dice for certain things. However, I've found that mixing this sort of "doubling" with "roll and compare" mechanics is a bit clunky; I want to keep those things separated, for the most part.
