Monday, May 10, 2021

Ability Modifiers in Immortal Rules

Superhero-like bare-chested deity by Elmore

Occasionally I gripe about the slightly unfortunate pattern of standard ability score modifiers chosen by Tom Moldvay in the B/X rules. I mean: On the one hand it's kind of fun (you can get a +3 bonus but only at the perfect score of 18). It sort of has the feeling of reflecting a bell-shaped distribution; but in so doing it actually over-amplifies the effect. (Example gripes part way through our recent Wandering DMs show on 50 Years of Chainmail). Moreover, at some point for a variety of reasons you're going to want to expand the possible ability scores to higher numbers, and then the pattern doesn't scale well. As a case study I refer to the ability modifier table from Frank Mentzer's Immortal Rules, the final boxed set in his BECMI D&D line (1986), which expands the possible ability scores from 0 to 100:

As a someone slightly spectrum-gazing, that makes me really unhappy. Note you have that 18 score with a span of just 1 pip for the +3 bonus. Then the ranges re-expand to spans of 2, 3, 4, 5, 6, 7, 8, 9 (in the category of 54-62), and then start contracting again, to spans of 8, 7, 6, 5, 5, 4, 2, 1, and 1 again (the last two for the 99 and 100 scores, of course). Here it is in chart form:

A bimodal distribution

Urgh, that bimodal distribution makes me feel like I've swallowed something very bitter.

I do think a better solution is to pick a standard class-width which allows you to predictably extrapolate to whatever point you might desire. This could be the 2-pips per bonus bump as seen in D&D from 3rd Edition (and echoing the way skill bonuses worked all the way back to 1E OA), or it could be the 3-pip ranges that I use in OED (which in my analysis is the best reflection of OD&D bonuses -- also has the statistical nicety that it reflects the z-score of a given ability versus the 3d6 population distribution, but don't worry about that too much).

(Side note: The Immortal Rules boxed set was the only D&D product I ever got which resulted in me returning it to the store. Partly for novelties like this, and that it's largely a totally different game; PCs simply discard whatever traits they had previously, etc.)

Can you think of any better way of expanding the Moldvay B/X ability modifier table than what Mentzer came up with here?


  1. If I remember correctly, Aaron Alston's "Wrath of the Immortals" (which essentially rewrote Mentzer's "I" set) altered the progression scale of the ability scores.

    However, it's been a while since I've read the book and it is currently in a place I can't get to (a cabinet in an office closet buried behind a ton of other "stuff") so I cannot check it at the moment.

    I prefer my B/X ability scores to represent the limits of human potential, with anything greater/lesser being "unnumbered" (a character with ogre strength simply doubles damage, for example, a la the Gauntlets of Ogre Power description). Giving special attributes or bonuses (a high intelligence being immune to illusion, etc.) for superhuman ability is preferable to being too "numerically focused."

    That being said, I have written/designed two B/X based games that have abilities that go "above 18:"

    - my Realms of Chaos inspired B/X supplement COMES CHAOS has the potential for a PC to increase STR beyond 18; the progression is +4 for 20-21, +5 for 22-23, and +6 for 24 (the absolute maximum). STR 19+ allows a character to wield battle axes and 2H swords with one hand (still losing initiative, as per B/X rules), and gives a chance to break magically held doors equal to 1 in 6 per point above 18.

    - my Shadowrun inspired game, CRY DARK FUTURE, has the potential for some abilities to go beyond 18, especially with cybernetic enhancement. Adjustments are made every 2 pips, and only on even numbers (so +4 at 20, +5 at 22, etc.); however, only a high CON or STR provide bonuses for abilities over 18 (i.e. even if a character has a DEX 20...which really shouldn't happen...the max bonus is +3). In play, this worked fine, and well-emulated the genre with beefy cyborgs and vat-grown muscle jocks.

    (both of these books are yet unpublished as I'm still waiting on artists' contributions. Covid times, you know?)

    Anyway...that's my two cents.

    1. That's all cool and reasonable! I totally grok the 3-18 range as "normal human" distribution, that's fine. And your experience of kind of wanting to keep that range but always stepping into a compelling reason to expand it is very much on my theme here. I like what you did in both cases!

      Now I need to go see the Allston Immortals version, I think I also had that in my hands at one point, then forgot about it for this post.

    2. Found it: I hadn't recollected that came out *after* the Allston's Rules Cyclopedia! I think that's contributing to my blind-spot: I tend to think of the RC was the final capstone in that line.

      The expanded ability score table is different, but actually suffers the same problem: the bonus ranges expand and shrink in much the same way. (Although it's a 2-pip range for 18-19, 20-21, 22-23, so it's not quite as sharp at that particular point.)

  2. I've toyed with having each point above 18 raise the bonus by +1 (19 = +4, 20 = +5, etc.). It worked well enough, and matches a bit better with AD&D's scale than the Immortals system.

    My main gripe with D&D ability score bonuses is that they can be made to scale upwards quite easily, but scaling them downwards is a pain. You only get two numbers to work with below 3 until you hit 0, and then it's into the negatives.

    1. +1 per point above 18 (and -1 per point below 3) is also what I've used for a while now. The bottom end doesn't bother me; I play such that a zero stat is death 🙂

  3. For ease of use I think the linear progression is best, I prefer 1/3 over 1/2, but that's at least partly out of a desire to differentiate between Old school and New school. I really appreciate Mentzer's attempts at expanding D&D to it's logical conclusion (and giving a canonical, by the book 'win condition' for D&D) but I think that 5e is wise in soft capping 'mortal' stats at 20 and hard capping 'demigod' stats at 30 (for +10). As I recall most of the Immortal stuff deals with power points anyway, so why go so high?

  4. Our solution back in the day was, you could have a score higher than 18, but the modifier capped out a +3.
    So, you could be super strong/dexterous/smart, but could only leverage "+3" of it in combat/die rolls. So score above 18 only came into play for lifting, and comparing 2 sets of very strong/dexterous/charismatic folks.
    not sure how/why we settled on it, I guess it just made sense to us and since that is where the chart stopped and we didn't question it?

    1. Ah, interesting! That's an option I'd considered. A little like the AD&D magazine 18/% values for abilities other than strength (no mechanical benefit).

  5. I already use a slightly different table anyhow, which is just the d20 System table cut in half and rounded off, viz.:

    3 … −2
    4–7 … −1
    8–13 … ±0
    14–17 … +1
    18 … +2

    If I wanted to scale the scores above 18 (which I don't), the steps of four work well enough to approximate the table created by Mentzer and tweaked by Allston. An ability score of 100 (cut in half to get 50, subtract 5 to get 45, cut in half again and drop the fraction) would yield a modifier of +22, which isn't too far off from the table value of +20.

    1. That's not bad! Like I say, I think any fixed increment is better than shrinking/contracting ranges. Reasonable argument for 4-point steps.

  6. I always wanted to do an extrapolation based on the bonuses like maybe a + 4 for 19 and + 5 for 20 or something. Adding a bonus for each standard deviation makes sense, but I like having that + 3 for 18.

    So I pulled out my old casio graphing calculator and ran some equations. I ended up with a 3rd order polynomial: y = 3.0466*10^-3x^3 - 0.1023738x^2 + 1.36581462x + 6.1604784

    It gives a +4 for 19, +5 for for 20 and +6 for 21 and the reverse for 0, 1, and 2. If you take it all the way to 100 it would give a +2153. I know it's ridiculous, but it certainly does convey what godlike power really is :D

    1. LOL, thank you for that extrapolation!

    2. I'm actually considering using this, and adjusting the PP cost an immortal must pay to increase to an exponential scale so cost is more commensurate with bonus. Then immortals with high scores would be far rarer, there wouldn't be a strict cap as suggested above, but for all practical purposes I don't see any being going above a score of 30 which gives a bonus of +25

    3. Great argument for an exponential advancement system, actually.

  7. In our five year long OD&D campaign, I wanted to limit modifiers to basically +1 and much more seldomly -1 to make the majority of characters feel very much 3LBB like. However I also wanted to include a small chance for that extra excitement should one of the players roll really well in character creation (which in fact led to a Str 18/97 fighting-man ...).

    So here's my modifiers table to accomplish this:

    3 ... -3
    4 ... -2
    5 ... -1
    6-13 ... 0
    14-16 ... +1
    17 ... +2
    18 ... +3

    Form follows function I guess ;-)

  8. I have one PC with a 19 stat from an ioun stone. I used +4. I plan to use additional +1 for scores above that.

    We use 3d6 in order and I like the bell curve of the BX mods allowing a PC with a low score to be playable and those with high scores to not dominate. There is too much focus on ability scores and excellence.

    I've been tempted to go the to the ODD ways of very limited bonuses based on ability scores (to emphasize that skill>stats), but have stuck with the bell curve of BX. Once players bought into playing what the dice gave them, they changed their play style to match their limitations and strengths of the numbers. (e.g. the thief with the 6 CON and 13 CHA was the first to hire henchman, the 18 DEX 5 CON assassin focused on ending combats quickly).

    That was when I figured to stop fiddling.

  9. With your 3 pip range for ODD, do you allow players to increase their ability scores over time? So I don't know does a level 5 fighter gain bonus to strength?

  10. I use the following general Ability Score chart with my Labyrinth Lord games:

    Score (Modifier)
    0 (Dead)
    1 (-5)
    2 (-4)
    3 (-3)
    4-5 (-2)
    6-8 (-1)
    9-12 (+/-0)
    13-15 (+1)
    16-17 (+2)
    18 (+3, Ogre)
    19 (+4, Hill Giant)
    20 (+5, Stone Giant)
    21 (+6, Frost Giant)
    22 (+7, Fire Giant)
    23 (+8, Cloud Giant)
    24 (+9, Storm Giant)
    25 (+10, Titan) MAXIMUM MORTAL LEVEL

    Giants are included to show general Strength equivalents (and for potions/girdles of giant strength). Mortals do not gain the great base damage of the giants when they have that Strength as the damage is as much from the size and heft of the weapon as the strength of the wielder; they gain the listed bonus to hit and damage instead.

    Wishes and alter reality spell can increase Ability Scores. Anything less than 13 is raised to 13 with merely one wish. To raise an Ability Score thereafter costs a number of wishes equal to the new bonus; thus, 13 to 14 is one wish, 15 to 16 is two wishes, 19 to 20 is five wishes, and so forth. Various magical items and locations can bypass this, granting a full point at whatever level of ability. PCs do not gain in scores by rising through levels, only through great magic or (up to 18) via specialized and very costly training (in money and time).

    Monsters do not use the same system of Ability Scores as do PCs and human/demi-human NPCs, any more than they do for any other abilities or powers; I think making a unified system was a great error in 3E and beyond, codifying that which should not be codified, but that's just my opinion.

    1. That's nice! That ability chart is basically the same as seen in the AD&D line.

      It's a good point about how giant damage is from strength vs. mass, and whether that's separable or not. I've wrestled with that multiple times.

  11. I actually liked the bimodal distribution, because it effectively reflects the fact that there are really two populations (measured on the same scale): the mortals vs the immortals. And mixing the two under a distribution that is not a mixture wouldn't feel right.

    1. Yeah, actually... when I added the chart to this post a few days after I initially wrote it, that thought did jump into my head. A fair observation!