Ability Score Improvements have been a terrible addition to D&D

This is going to be one of them rants I fear. 

It relates to my previous meditation on the heft of levels across various editions and my recent contemplation on ability checks in B/X, specifically my desire have the unmodified numbers mean something in and of themselves, rather than something purely to derive other numbers from that do have mechanical relevance. In a way, this posts is like a concluding remark on the heft of levels in TSR vs WotC D&D.

To summarise, if the mechanical relevance of ability scores are almost always somewhere on a scale of -3/+5, why do we bother with rolling 3-18 instead of just using the derived numbers to begin with? Why has that never changed? And why do I have a firm impression that there'd be a great outcry if it were ever changed in a future edition?

And it occurred to me that ability scores do have a relevance in the unmodified form, one that has remained across all editions - They are the formative narrative components of the character sheet. In this way, I'd argue, they carry even greater relevance than any mechanical bonuses derived from them. 

They tell the player, and everyone else allowed to see those numbers, something about what that character is like, how they are most likely to solve problems and what kind of challenges they most likely struggle with compared to other characters. Only class and level comes close to defining what a character is like in comparison to those ability scores.

And this is where 3e+ messed all that up by introducing ability score improvements at various levels.

Three player characters that are just as much defined by their ability scores as their class and level.

From a game engine perspective, I can see the sense in it. Ability scores in WotC D&D give more mechanical benefits than in TSR D&D and are far more tied into your general likelihood of success in various tasks. So it makes sense that as you level up, you get a chance to raise those scores, especially the ones you really benefit from in your class. You need them increased to keep up with the game, basically.

The result however, is an entropic principle that is deplorable: As levels go up, all characters become the same. I don't need to look at the character sheets of 5e fighters to know that by 6th level, most of them will have STR or DEX maxed out and by 8th level, all of them will. And then CON, once their main ability score is maxed out. 

Their ability scores no longer tell you anything about that character other than they lived long enough to become high level. As you level up, characters inevitably end up becoming more and more similar. Introducing point buy for optimising 'builds' only amplifies this entropic tendency, ensuring that even 1st level characters end up looking alike and even numbers start showing up far more frequently since those are the even giving bonuses.

It's the reason, I discovered, why my eyes glaze over when I look at NPC stat sheet in WotC D&D. Everything there is basically just a function of class and level, the scores there only to service other numbers. It tells me only how to roll the dice, but little about the character itself.

In contrast, I enjoy studying an [N]PC stat sheet in TSR D&D. Their ability scores, in conjunction with class and level of course, tell me a lot about that character. If they have been successful against the odds, or if solve their problems in other ways than you might typically expect from that class. It makes them interesting, every character unique, in ways that are utterly diluted in WotC D&D due to point buy+ability score increases. 

Another thing, besides point-buy, that amplifies this is the increased mechanical relevance of ability scores in WotC, in terms of derived numbers for other mechanics. 

Before we proceed, I am aware that this may seem like an odd complaint in light of my recent post seeking more relevance for ability scores. To that I would say I sought that relevance for the score itself. Not to have more derived numbers tied into it affecting other sub-systems. Anyhoo....

There is something almost hallowed about that 3-18 number.
It conveys an intuitive impression that can be grokked by just about anyone. At a glance, a STR 15 character is clearly different to a STR 13 one, even though in B/X the mechanical benefits are the same.
It conveys something quite clearly in the fiction of the game. As a narrative modifier it has just the right level of granularity. The score means something in and of itself, even if it speaks more to the fiction than the rules.

The more mechanical benefits you tie into derived numbers based on this however, especially derived numbers, the less significant the original ability score become in and of itself.

I look at a STR 18 character in B/X and what I see is a rare specimen that would stand out in any crowd, someone strongly defined by their physical prowess, perhaps even singularly so. 

I look at a STR 18 character in 5e and what I see is a quite strong character, +4 to hit and dmg and Strength Saving Throws.

In TSR D&D, the score tells a story of the character that just does not happen in the same way in 3e+. There are other things in the foreground of what that score means.  

My advice: Don't use point buy. Avoid ability score increases. You might start to find your characters far more meaningful after that.


  1. Continuing my comments from last post...

    I like ability score increases (ASIs) for several reasons.

    I agree with your concern about homogeneity. This can be avoided by giving away random ASIs, for example.

    I like fighters improving strength because it is the only way they get to hit harder in B/X. And, well, people DO hit harder with training.

    (Despite HP being abstract, think of a fighter throwing a spear at a dragon, for example).

    I like ASIs if you use ability checks at all: adventurers should become better at wilderness exploration, climbing ropes or whatever you use abilities for.

    MUs should get smarter/wiser as they advance, etc.

    If you don't want ASIs, there are ways around it, of course - like the "ability throws" we discussed in the previous post, of giving high-level fighters a bonus to damage (like ACKS).

    ASIs are just an easier method to do this.

    1. Random ASIs strike me as a tad aggravating tbh. "Your wizard levels up and you gain.... +2 STR."

      As for PCs getting better with experience, they do. At their class stuff. If fighters should hit harder (besides hitting better and more often) then give them a damage bonus.
      If wizards should get smarter as they level (which I really don't think I agree with), put a class feature in there to reflect it somehow.

      Overall getting better - Well, I don't think that should necessarily be the case. But when I do, I think 5e's proficiency bonus does a splendid job of reflecting that (I always liked Fighting Fantasy's singular Ability score as a reflection of overall heroic prowess and this goes some way towards that) or Action Throws as discussed.

      I don't think ASIs are an easier method. Ability scores are tied into a very strong narrative component and ASIs tamper with that to the point of nullifying it, which is not straightforward at all to my mind.

      Proficiency bonus/Action Throws on the other hand, reflects the fiction of these assumptions which makes it more elegant.

    2. I agree 5e's proficiency bonus did a good job at that, and OTOH that ability scores in 5e look very samey across characters.

      There is more to be said about random ASIs (e.g., maturing/aging, raising INT could be part of a possible random list for MUS that included getting more spells and a smaller chance of raising WIS, etc.), but I don't really use them TBH.

      I see your point about ability scores, but I don't see many ways around changing them with time.

      I think of the Dragon #36 where Gygax shows Conan changing ability scores up and down as he ages, which is also a cool distinction to have (although somewhat exaggerated here - his stats are all 15-18 by 70 years old!).

      But, ultimately, I think it is a matter of taste. For me it is expect that a fighter eventually becomes stronger, etc.

    3. I am not opposed to Ability score increases at all costs. Maybe you found a ring of wishes, drew the right card from a deck of many things, extracted a favor from a god or obtained enlightenment from the Codex of Infinite Planes.

      As long as it tells a story about your character, that's cool.

      But formulaic progression of them by level - no, the blandness is too much for me.

  2. In 5e I tend to only use the basic rules. I like the increases rather than feats which I don't use. (And they are not in the basic rules). I getcha though. I also never use point buy. In osr games I allow a reroll of one stat upon levelling.

  3. The 13 vs 15 Str in B/X example is a good one, however it only holds in systems that use a "roll under" method for task resolution. If they both only grant a +1 to rolls, then the implied fiction (wow, 15 Str PC is stronger) falls apart in play, particularly with how swingy a d20 roll inherently is.

  4. I agree with the example of 13 vs 15 STR, but can't the same thing be applied to a 5e character? An 18 STR character stands out either way right? It's just that they also get some firm mechanical bonuses, whereas for a B/X character the difference between 13/15 is just for show.

    At least, that's what it seems like, I like the idea of the ability scores themselves being defining, and want to see a way through to that.

    1. I thought about this some more and I think my point about mechanical significance of ability scores is perhaps a bit overwrought.

      I think what it boils down to for me is that 5e character with ST18 is that it is more likely to tell me he's a 4th level fighter than anything else. And as such does not stand out in any way from all the other 4th+ lvl fighters you'd meet.

  5. It's been a looooong time since I've played 3e, but didn't those ability score increases only happen at +1 every 4 levels?

    1. Yes, so not quite as egregrious as 5e in this regard.

    2. I should add, Yora's recent post on 3e vs 3.5

      makes me want to revisit 3e at some point to see if there are more merits to it than I recall.

  6. IMHO you're describing an issue with combat design and levelling:

    1. As characters level, they have to gain ability scores because their combat prowess is dependent on it. It's a very 80s action hero aesthetic to say that a build like Arnold Schwarzenegger is best for swinging a sword (reality: it's not).

    2. As characters level, they have ability score increases dumped on them so often that they begin to homogenize, and thus the meaningful innate differences in characters (my barbarian is a huge muscular brute) fades away in service of balancing combat (my gnome warrior has to increase in strength as well so he can deal an equal amount of damage with his massive warhammer that is twice as tall as he is).

    So if you want to combat to meaningfully change and characters to gain power, you need to dump stats on them as they level. And if you don't want your game to be imbalanced, homogeneity ensues.

    I like to detach combat damage from ability scores. Ability scores still impact combat, but swinging a weapon is not a strength issue. Casting magic isn't the same thing as doing calculus problems in your head. Those artificial limits are goofy, and they feel goofy to players.

    Check out Warhammer's approach. Weapon skill is its own stat and players roll it against each other when they entre melee combat. So you could increase weapon skill without inventing a gnome waddling around with muscles so big they can't touch their nose.

  7. As a counterpoint... (I should preface, assume a B/X D&D context here) if you *don't* use point-buy, but *do* use ASIs, I'd argue you end up with something that has some of the best of both worlds. Characters are still distinct from one another - yes, fighters may all have STR 18 by level 10 or whatever, but the fighter who also rolled a 15 Wisdom is still quite distinct from the one who rolled a 15 Intelligence (or heck, let's say 5 Wisdom). Those narrative cues are still there in such a system.

    On the other hand, by still allowing some limited ASIs you're also helping with one of the issues that often frustrates players - the idea of playing a character who is "stuck" being permanently less effective than they could otherwise be. These also give some relevance to ability scores between the "breakpoints" for bonuses - a 16 STR and 17 STR fighter have no mechanical distinction in a game without ASIs, but in a game with them the 17 STR fighter will reach that coveted +3 bonus an entire level earlier than the former.

    In my game, I've paired limited ASIs with an injury table that allows ability scores to go down. The end result is ability scores feel more dynamic overall. Maybe you took a pretty bad injury that dropped your CON down to 8. When you next level up, there's an interesting choice between bumping it back up to 9, or continuing to pump STR. I've also made my ASIs less automatic than in WotC D&D. To attempt to raise a score, you have to roll a d20 *over* the current value of that score. This also means the "all fighters will max strength" flaw is less of an issue, because you're actually arguably better off shoring up some other weaknesses (maybe your DEX or WIS) than going straight for the maxed prime req.

    Anyway - I enjoyed the post! Hope some of what I said is food for thought, at least!


Post a Comment

Popular posts from this blog

Fantasy Map Review IV: Forgotten Realms

Fantasy Map Review II: Greyhawk

Fantasy Map Review III: Dragonlance

Review: Five Torches Deep

Comparison: Five Torches Deep vs Into the Unknown