Ability Scores
I’ve been working on a homebrewed D&D-esque hack intended for running classic & OSR adventure modules in my preferred style. I’m going to be posting some design notes where I force myself to clarify my thoughts, starting with ability scores. ___
Ability scores in the past
AFAIK all versions of D&D1 use the classic six scores to represent a character’s innate strengths and weaknesses:
- Strength/STR (self-explanatory), usually used for melee attacks and physical efforts
- Dexterity/DEX (bodily agility and manual dexterity), usually added to armor class and rolls to reflexively avoid traps
- Constitution/CON (physical fortitude and hardiness), usually improves hit points and resistance to things like cold or poison
- Intelligence/INT (memory, logic), sometimes used for spellcasting, sometimes improves skill points
- Wisdom/WIS (perception, insight, common sense), usually used for divine/cleric magic, often used for skill checks to perceive things or detect lies
- Charisma/CHA (interpersonal likeability and force-of-will), sometimes used for spellcasting, usually used for the ever-controversial social skills
Most OSR retroclones keep the original six, but a few ‘nu-SR’ offerings use a reduced list. Into The Odd uses strength, dexterity, and willpower, as does its close descendent Cairn. Interestingly, Electric Bastionland (by the author of Into The Odd) swaps out willpower for charisma on the grounds that charisma is easier to translate into a mental picture or role-playing cue. MÖRK BORG has agility, presence, strength, and toughness.2
The DIY & Dragons blog provides a helpful meta-framework for all this by arguing that there are actually eight implicit abilities derived from three dichotomies: physical vs. mental, force vs. grace, and attack vs. defense. The OG six result from merging physical grace attack and defense into dexterity and mental force attack and mental grace attack into Charisma. The ITO reduction amounts to merging all the mental ones together, condensing the physical to force attack/defense and grace attack/defense. I think that the force vs. grace dichotomy is iffy in the mental realm, but I like the schema overall.
One of my goals for my homebrew is to land in ‘mid-crunch’ territory with enough material to support longer campaigns but as much streamlining as possible. I want to use ability scores, and I want them to play nicely with other systems like skills, combat, and encumbrance.
With no desire to rehash 40+ years of nerd debates about abilities, here are the salient points on my mind as I work on my system:
- Not all of the classic abilities are orthogonal - that is, there are some combinations of high and low scores that seem very unlikely (e.g. high strength and low constitution or high intelligence and low wisdom). This isn’t inherently a bad thing - these edge cases create space for very unique and memorable characters, but they can complicate how other mechanics interact with the ability scores — should a character’s capacity to carry loot be tied to strength or constitution?3
- The mental ability scores of a character can butt up against the mental qualities of their player, hence ever-present questions like ‘can a stupid player properly play a high INT character?’ or ‘if I role-play a really convincing argument with the NPC can it override my low Charisma roll?’. Obviously, clear demarcations of what the scores represent are needed. Erring on one side leans in a ‘story-game’ direction where the task of the player is just to fill in the details that lead to the roll’s result. Erring on the other side begs the question of why we have mental ability scores at all and potentially turns them into ‘dump stats’.
- It seems like a key reason why mental ability scores keep showing up is magic.4 Making magic equally available to everyone is unsatisfying, as is positing a ‘magical aptitude’ variable that isn’t tethered to any observable mental quality.4
- The number of abilities affects degrees of freedom for differentiating characters. The more abilities, the higher the odds of having at least one outlier and/or having a unique niche in an adventuring party. Conversely, as abilities get more and more specific, the odds that an outlier ever matters in play decrease.
- Ability scores are implicitly hard to change, so what is designated as an ability score contains fundamental assumptions about the game. I’m not interested in straying too far from D&D, but I’m aware that there are games where ‘Luck’, ‘Nature’, or ‘Wealth’ are treated as ability scores.
My approach
First, my goals:
- Reasonably compatible with D&D and my favorite D&D-derived games (BFRPG, WWN)
- Minimal redundancy - when another system (e.g. skills, encumbrance, magic) calls for an ability modifier, it should be obvious which to use.
- Easy to use as role-playing cues
- Cool names that are still easy to parse
- Clear limits to mental skills to keep the focus on role-playing
- Valid uses for each ability (no ‘dump stats’)
Second, I have already settled on using 3d6 to generate abilities with the same modifier spread as Worlds Without Number, to be used in conjunction with WWN-style 2d6 skill rolls and d20 attack rolls, and GLOG-style roll-under saves. So the numbers and mechanics are not up for adjustment - just the abilities themselves.
Physical abilities
I decided to combine strength and constitution into a single score (‘Physique’) due to the collinearity discussed above. The high-con low-strength edge case is easy to imagine (ultra-skinny marathon runner), but the high-strength low-con side is trickier. It’s true that there are strong people who are physically unhealthy, but I'd argue they aren't really fragile in the way that low CON scores usually represent.5
In my ~lived experience~ lifting weights as a hobby, efforts to improve my STR have all benefitted what I imagine to be my CON-derived attributes. This also simplifies ambiguities around things like carry capacity as mentioned above.
I kept dexterity as-is but renamed it to ‘Grace’. There’s now a nice symmetry where Physique and Grace both have an offensive (melee vs. ranged) and a defensive (HP vs AC) use.
Thus:
Physique (PHY): governs melee attacks and damage, hit points, feats of strength and endurance
Grace (GRA): governs ranged attacks, armor class, reflex saves, feats of agility and dexterity
Mental abilities
Of the three classic mental abilities, I regard wisdom as the most ripe for the chopping block:
- Abstracting ‘common sense’ with a die roll steps hardest on the toes of the basic OSR philosophy.
- Same for wisdom as perception - I want to design dungeons where detection of traps and secrets is more deterministic (e.g. a function of time invested searching or a reward for exploring other areas).
- I suck at role-playing NPCs, and so despite how un-OSR it is I want a reliable way to run abstracted negotiations - thus charisma stays. And I like having an intelligence stat to smooth out the ever-present question of ‘would I know this?’ — intelligence stays too.
So WIS is out. But intelligence and charisma won’t escape untouched - I decided to name them after two of the modes of persuasion in Aristotle’s Rhetoric for a bit of arcane flourish.[^8] Intelligence thus becomes Logos, specifically denoting a character’s mastery of their language-world and social knowledge-systems (i.e. lore, magic, languages, and etiquette) rather than anything resembling wit, common sense, or even logic. Charisma becomes Ethos and is more about having an authoritative performance of the self rather than likability, glibness, or charm.
Thus:
Logos (LOG): governs magical spells to manipulate the elements and play with time and space, mastery of language, capacity to memorize lore and past education, and adeptness at detail-oriented tasks like surgery and lockpicking. NOT a replacement for a player’s common sense or logic.
Ethos (ETH): governs magical spells to create illusions, dominate others, and summon entities, body language, confidence, and interpersonal acuity, ego. May be used to bias an NPC’s attitude but NOT a replacement for a player’s use of argument.
Final notes
Returning to the DIY & Dragons post, my four abilities essentially keep the physical/mental and force/grace dichotomies and collapse attack and defense together. Using four abilities sits nicely in my ‘crunch goldilocks zone’ and should hopefully make characters slightly more legible to newbies (not withstanding the ridiculous names). Hopefully, the orthogonality of the abilities will cut down on ambiguous situations where the appropriate modifier is unclear.
Lastly, as a fan of using tarot cards as writing/GMing aids, there’s a convenient mapping of PHY → Pentacles, GRA → Cups, LOG → Swords, and ETH → Wands. More on this later, perhaps.
From DIY & Dragons: “It feels worth pointing out that in the original version of D&D, Strength, Intelligence, and Wisdom didn't represent what they do now. They only things they modified were the XP you received from playing certain classes, which means that they were more like measures of Fighting-Man-ness, Magic-User-ness, and Cleric-ness, respectively, than they were like the abilities familiar to us using those same names today. And that means, when you take those three away, that in playing OD&D, you're left with only 3 real ability scores, Dexterity, Constitution, and Charisma, which is just what you get in I2TO and Numenera.”↩
I haven’t played MÖRK BORG so I don’t want to knock it too hard, but I really don’t like this split. Agility/Presence and Strength/Toughness seem like they’d be stepping on each others’ toes a lot.↩
Every time a non-obvious question like this comes up at the table, the rulebook has to come out or the GM has to process the ambiguity and make an on-the-fly ruling. Either is taxing. I have a loose theory of ‘tight abilities, loose skills’ that I hope to expand on later.↩
I suspect this is why they made it into Into The Odd and its descendants but I can’t say for sure.↩
If CON was just a measure of cardiovascular endurance it might work, but constitution in D&D is usually used to derive things that have little to do with cardio. The ‘fat powerlifter’ stereotype would, I think, still be modeled in D&D as having high CON.↩