In Defense of Ability Scores
Don’t Have a Sacred Cow, Man
A remastered Pathfinder has rounded up and slaughtered all remaining sacred cattle of D&D. I have no beef with this approach and instead applaud any attempts to trim the fat. One cut made by the Paizo people that is forehead-slappingly obvious is once and for all ending the tyranny of ability scores (for Strength, Dexterity and so forth) in favor of ability modifiers.
Ability scores, as opposed to ability modifiers, have been vestigial for a while. In case you only follow this blog for its very occasional but highly practical advice regarding doppelgängers but are otherwise unfamiliar with Dungeons & Dragons, ability scores and ability modifiers are basically two sets of numbers describing the basic qualities of a character that do the same thing and are derivative of each other. More specifically, ability scores are traditionally determined by rolling 3d6 (though many less traditional methods have supplanted this method in popularity) for a nice bell curve distribution between 3 and 18, with 3 indicating an abysmally low ability and 18 denoting a nearly superhuman ability.
Ability scores predate D&D, although the abilities themselves have not always stayed the same. The array of Strength, Dexterity, Constitution, Intelligence, Wisdom, and Charisma are now all but ubiquitous (although it need not be). But the earliest character sheets from 1971 show the abilities in Dave Arneson’s games were “Brains, Looks, Credibility, Sex [specifically as in sexual prowess, oddly], Health, Strength, Courage, and Cunning.” It was in the hands of one E.G. Gygax that these became the classic six attributes we love today with their more highfalutin names. As will be discussed in more depth below, these early attributes were used for roll-under style tests, without the need for ability modifiers, but mostly impacted speed of progression depending on your class, among other things like how many heirlings your character could employ (highly unrealistic because I’ve never met an employer who was likable).
Twas the 3rd edition of D&D, produced by Wizards of the Coast after their conquest of TSR, that changed the game to emphasize a universal mechanic, the d20+modifier system. In the 2nd edition, game mechanics proliferated like species of a primordial rain forest–thieves used a d100 test for their skills, initiative involving rolling a d10, attacks and saves involved rolling d20s, etc. As a reaction to this fecund but perhaps unwieldy set of systems, the 3rd edition introduced unified skill systems, unified saves, unified level progression and most importantly for our purposes, a unified resolution mechanic where you rolled a d20, added relevant modifiers to the roll, and attempted to beat a certain numeric threshold. This unified mechanic has spread widely throughout RPGs and has remained a fixture of orthodox D&D. But with this change came the issue of how to deal with the ability scores?
The primary use of ability scores in 3rd and later editions of D&D have been to determine the more important ability modifiers. The modifier equals the score minus 10 and divided by 2, to produce a range between -5 and +5, which would be added to relevant d20 rolls. While this innovation is a bit chunky for the basic ability test, it did prove versatile in allowing ability modifiers to modify all sorts of thing: your Dexterity modifier affects your initiative rolls and Armor Class, your Strength modifier affects your melee damage, etc. The modifier became the number that mattered, the score was just how you got there. But as rolling dice to determine your ability scores gained less popularity for the playstyles preferred by 3rd and later edition D&D players in comparison to point-buy and similar options, even these 3 to 18 ranges became increasingly irrelevant.
As the wisdom score began to increasingly resemble the wisdom teeth, it is no surprise that game designers would look to ability scores with their pliers in hand. Cutting the ability score is an obvious solution for Pathfinder (so obvious that I’m sure the Wizards of the Coast employee chained to the typewriter for the 6th edition, upon hearing the news from Paizo, uttered “fuck, fuck, shit, fuck!”) but it isn’t the only path. Some lineages of fantasy games never embraced the dual score-cum-modifier regime, hewing to the old ways. And, frankly, even if you aren’t a sucker for nostalgia, the ability score has a lot to offer as a mechanic that the mighty ability modifier simply can’t replicate. From the origins of the hobby, people were putting these bell curves to work in interesting ways and that has only continued into the old school renaissance and beyond.
Sacred Wagyū
One of the most popular alternatives to the d20+modifier(s) core mechanic is the elegant “roll under” mechanic. This method elevates the need for any addition. Instead, the player simply rolls a d20 and compares it to their relevant ability score. If the result exceeds their score, they fail. If it is lower, they succeed. Thus a character with a higher score is more likely to succeed. I first came across this method after many years of using the 3.5 and 5th edition d20 resolution mechanic so its simplicity was a revelation to me. But it needn’t be simply so simple. It is a chassis and strong enough to build from.
Roll under is so obvious that it’s no surprise that it has early origins in the hobby. While its first official appearance in the annals of TSR in the Basic/Expert Rules of D&D in 1981, archeological evidence suggests this method dates back a decade prior. Game archeologists have unearthed character sheets belonging to early gamers (homo ludarius), believed by evolutionary scholars to be ancestors of all species of gamer living today. These circa 1971 character sheets, predating D&D (in published corpus but not in spirit) itself indicate that they were largely using the roll under method (albeit 2d6, not d20 under 3d6) for ability and skill checks. If these findings are to be believed, this is what Gary Gygax was gesturing at when in original D&D he spouted half-rules like “Strength will also aid in opening traps and so on.” Thank you, Gary, that’s very helpful. The method for determining whether an ability was in fact helpful was used at the time but just first articulated in B/X a decade after the hobby’s earliest days.
Despite its elegance, something feels amiss in basic roll under. It lacks the granularity of difficulty. Surely the odds of climbing a cliff-face should be meaningfully different than opening a pickle jar, yet both are mere rolls under your Strength score. The aforementioned 1981 rules proposes as a solution that referees can assign a modifier to the roll to account for difficulty but this feels clunky. If you’re going to throw in addition and subtraction, you might as well just use the d20+modifier(s) method. Luckily, we have had 40+ years to roll under our collective Intelligence scores to figure out a better way.
The blackjack method of ability checks, which may also be characterized as “roll high, under” is the most elegant evolution of the roll under check. In the blackjack method, the player still rolls a d20 and attempts to roll under their ability score, but now they also try to roll over a difficulty value set by the referee based on the difficulty of the task. This allows for more nuance and precision in your run-of-the-mill roll under system and doesn't involve any math beyond just comparing numbers. This innovation, as applied to the core mechanic for a D&D-like may rightfully be attributed to Errant. The author, Ava Islam, attributes her eureka to Whitehack, which uses a nearly identical method but just for attack rolls. Ava took this elegant method and applied it to the innermost rule of Errant.
My own (work-in-progress) system, Prismatic Wasteland, uses a modified blackjack method that allows for a gradient of success. I have long preferred systems that allowed partial successes rather than just binary failure, as partial successes are a useful vehicle for my style of referee-ing where I love to complicate the lives of player characters in ways that drive the game forward. If I wanted degrees of success and degrees of difficulty with the d20+modifier method, I would need to have so many different success thresholds that my head would quickly start to spin while running the game. Instead, I just put the elegant blackjack rolls to work: if the player rolls below their ability score and above the difficulty value (blackjack success), it is a full success. But if the player only rolls below their ability score but not above the difficulty value (roll under style success), it is only a partial success. Volia, added nuance with no additional complication. And that’s not all I’ve done with the core system! I didn’t hear no bell! I also wanted relevant skills to improve the player character’s odds of success and for how helpful being skilled is increase as the player character levels up. So I added the concept of a proficiency die, which goes from d4 to d10 as they level up. When the player character makes an ability score test where one of their skills is relevant and helpful, they roll both the d20 and the proficiency die and take the higher result to determine whether they succeed. I’ve packed so much into the blackjack mechanic but you still never need to add two numbers together, and you certainly never need to subtract 10 and divide by 2.
Ability scores don’t have to just be about ability checks. When a number can pull double or even triple duty, that is when you really start cooking. That is why the ability bonus is such an elite mechanic–you have checks but it can be applied to multiple other aspects of the game. For a game to use ability scores to their fullest in an equally elite way, they need to at least pull double duty. In Into the Odd (and many of its many, many descendants, including Cairn), the ability scores are both a yardstick for determining character aptitude but also are a resource that can be damaged: once the player character is out of HP, they start taking damage to their Strength ability score. This is basically like the “Grit versus Flesh” delineation for Hit Points, but much easier to use because you just have to track something, Strength, that you’re already tracking anyway. Similarly, some (less common) effects damage characters Willpower or Dexterity ability scores.
But we aren’t even done innovating with ability scores! In Prismatic Wasteland, which allows for ability score damage, the ability scores are also a resource beyond working as hit point alternatives. For instance, in combat a character can spend their Dexterity, reducing the score (not the maximum value, which the tests are based off of, just the current value–it works like hit points where you can rest to restore ability scores back to their maximum value) to dodge out of the way of an attack. How much Dexterity they spend depends on their armor, with heavier armor being more costly to dodge. The spellcasting system also involves spending Intelligence (typically, there are instances where the other mental ability, Charisma, may be used) based on the power of the spell cast. Why develop a separate spell point system when ability scores are right there? A character’s ancestry and feats also allow them to accomplish certain actions by spending ability scores. For instance, lizardfolk and elves can spend Strength to regenerate a severed limb, and halflings can spend Intelligence to become invisible for a short period.
And I doubt that is the final frontier for ability scores in TTRPGs. As the D&D-descended TTRPGs potentially split between ability modifier games (post-3e D&D, Pathfinder, etc.) and ability score games (Into the Odd, White Hack, Errant and many other P/OSR games that are inspired more by original D&D than WotC D&D), I expect both will continue to push the bounds of what these simple little mechanics can do.
