At the risk of throwing out the baby with the bathwater, I think the solution to the "cantrip problem" is to pivot the design space off of attribute mechanics. I started to think about it more as, "I use a spell, therefore I add 1d6 to the pool, for this attack." Now, as has been stated clearly before, not all classes/subclasses will get access to cantrips -- so what if we make having that specific class feature be the source of the 1d6 that we add to the pool? This lets us more easily differentiate Clerics from Paladins, and Druids from Rangers. (Another big distinction between spellcasters in 5e is ritual casting; this class feature is what I would argue is the main thing separating Sorcerers and Warlocks from being as high on my list as Druids or Wizards, in particular.)
Now that we've established what attributes won't do, we need to figure out what they will do, with regards to cantrips. If two-handed weapons are assumed to be getting a +8 damage mod, and two-weapon fighting will get +3 damage to each attack (assuming our standard array of +4/+3/+2/+1/+0) then we probably need to figure out something comparable to that, for cantrip attacks.
Would a single-target attack simply add both mental stats, as damage? Is there an easy way to balance out a "number of attacks" stat with the assumption that a "cantrip damage" stat will now likely be your highest stat? The problem is that if the only thing that "cantrip casters" care about is their mental stats, then suddenly everyone's got a +4 in one and a +3 in the other, and there's no variety. Maybe we still need to encourage spellcasters to have a +2 finger-waggling stat, after all. Do we simply treat (ranged) cantrip attacks the same way as (melee) two-handed weapon attacks, and add double the damage mod? How many attacks does an archer need to make, in order to be able to compete with that? This will take a bit more thought in order to flesh out, it seems.
The other aspect of class features that I wanted to bring to the fore was the idea of "exceptions-based design." I touched on it a little bit (albeit ham-fistedly) in a recent post, where there are general rules, but then there are exceptions made to those rules. One example with the sequel mechanics is that the general rule (so far) is that we assume TWF will use the lower damage mod between STR and DEX. This would encourage a class like a Fighter to put high stats into both, in order to be good at TWF; it would also make it more expensive for spellcasting classes (who rely more on investment in mental stats) to also be good at TWF.
However, as an exception, we might allow the Ranger subclass to simply always use the higher of these two stats, meaning they don't have to invest in both -- or to simply override this rule and use a different stat, such as AGIL. If the assumption within the "lore" of the Ranger is that they are always scouting ahead and need to be able to react quickly, then incentivizing them to have a higher AGIL would reflect that in the mechanics, via having a higher initiative score. AGIL also being tied to stealth would likewise tie in to the presumption of Rangers as hunters and ambushers, stalking their prey from the shadows.
This also makes me ask, how should the Monk/Rogue class be designed? If the starting point is that Monks will probably get an INT bonus to their AC, my immediate thought actually isn't to slant Rogues towards CHA. Monk generally lends itself to being the multi-attack class, whereas Rogue is more of an "alpha strike" class, using sneak attack. If the general rule is that DEX (or AGIL?) is the "number of ranged weapon attacks" stat, the obvious exception to build in for Monks would be to allow them to also use this number for melee attacks (perhaps using their lowest physical stat as the damage modifier?) Or maybe it makes more sense to go with "highest physical stat" as number of attacks and "lowest physical stat" as their melee damage mod.
Likewise, if the Rogue's preferred damage mod stat is DEX, would it make more sense for their dice pool stat to also be DEX, or should it be AGIL? We definitely want Rogues to be the best at things like lockpicking, so leaning more towards DEX seems like the obvious move... but it really is a question of how to balance the two. Much like in 5e, this gives the Rogue the versatility to put any remaining stat points into whichever mental stat they choose, making it more of a decision around which skills to focus on, than purely a combat consideration.
Dovetailing into the Rogue idea, the mechanic of "finesse" weapons from 5e is something that I think could make an appearance, allowing either STR or DEX to be used for damage, when attacking with a one-handed weapon (and not doing TWF). To tie this back into the conceit of class features, perhaps the "one-handed weapon stat" could be a function of classes, rather than having a general rule. Or, say, the general rule is that STR is the default, but characters with "finesse weapon proficiency" can instead use DEX. So we would expect classes like Rogues or maybe Rangers to have this proficiency, but Barbarians likely would not. Maybe a benefit of being a Barbarian is that you could always use STR, when doing two-weapon fighting.
---
Slightly off topic, but bringing things back to Eldritch Horror... One of my complaints with the game is how some characters are more about their items than their intrinsic abilities. While this is also a complaint I've had about D&D, it's significantly worse in a game where world-events regularly cause you to lose possessions at random. I do appreciate the conceit of reusing existing in-game items as a way to equip the starting class features for characters, but this style of game in particularly causes the weakness of item-dependent characters to be laid bare. The classic D&D comparable is when the party is captured and stripped of their belongings, imprisoned, etc. This is why TNP and the sequel generally focus on treating everything as a class feature -- even things that could reasonably be represented as items. In this way, they are part of the character at all times, and not something that can be taken away from them; they are just always assumed to be part of what the characters can do.
...
This post is a little bit ahead of schedule (or "on time," as far as our most optimistic forecasting is concerned...) Next post will likely be up by November 19th, at the latest -- but possibly even by the end of next week, depending on how things go.
