Thursday, December 17, 2015

Ability Score Minimums in B/X (Part 2)

I've been looking at ability score minimums for humans in an earlier post, suggesting that we can impose them without too much trouble if we so choose. Today I want to look at the existing ability score minimums for demi-humans briefly. If you check the B/X Basic Rulebook, pages B9-B10, you'll find the following:

  • Dwarves require a constitution score of 9 or higher.
  • Elves require an intelligence score of 9 or higher.
  • Halflings require a constitution score and a dexterity score of 9 or higher.

What's the effect of this? Well, if we roll up a million characters using 3d6 in order, we find that about 74% of those could be dwarves or elves and about 55% could be halflings. Let's not worry too much about the details here, what's important is that without futher consideration, a player is least likely to roll up a halfling.

If we go a little further and for each of our one million characters pick an actual class uniformly at random (out of all the classes a character qualifies for), we find (approximately) the following:

  • 17% Clerics
  • 17% Fighters
  • 17% Magic-Users
  • 17% Thieves
  • 12% Dwarves
  • 12% Elves
  • 8% Halflings

I don't know about you, but this doesn't sit right with me when it comes to the demographics implied by B/X. I always thought that elves should be rarer than dwarves which in turn should be rarer than halflings. The only thing that does work out as expected is that humans are the most populous: we get about 68% humans and 32% demi-humans.

Curious side note: Before I ever ran these simulations, I came up with a house rule to randomly determine race. I did this because I got tired of parties that were mostly demi-human. (I am with Gary that demi-humans should be exotic, not commonplace.) The rule I "guesstimated" was this:


In other words, 70% humans and 30% demi-humans seemed mostly reasonable to me. Now I'd be hard-pressed to say why exactly. I seem to recall fiddling with the d20 numbers for a while and this seemed to be the simplest way to make sure that elves are rarest, followed by dwarves, followed by halflings. So it may just have been an accident. (End of curious side note.)

If we wanted to tweak the resulting population by only modifying the required minimums, where should we start? If we added strength as a minimum to dwarves and elves all we'd achieve would be that each of those populations will be about the same as the halflings. Here's the (approximate) breakdown:

  • 18% Clerics
  • 18% Fighters
  • 18% Magic-Users
  • 18% Thieves
  • 9% Dwarves
  • 9% Elves
  • 9% Halflings

Not really what we'd want. But how about now tweaking the actual numbers? Let's leave halflings at strength and dexterity 9+ but move dwarves to strength 9+ and constitution 12+ and elves to strength 9+ and intelligence 14+. What do we get (approximately)?

  • 21% Clerics
  • 21% Fighters
  • 21% Magic-Users
  • 21% Thieves
  • 10% Halflings
  • 5% Dwarves
  • 2% Elves

That's more like it, at least it's closer to what I would expect. So if you feel like I do about what the population mix should be like, I'd suggest going to these stricter requirements for dwarves and elves. (Or you could do intelligence 13+ for elves, then you'd get slightly more of them, roughly 3%, but still not too many.)

Keep in mind that we've dealt only with minimum requirements here. It turns out that B/X has many more mechanics (prime requisite adjustment, XP adjustment, etc.) that impact the actual demographics implied by the rules. But that's for another post...


  1. Hmm...I'm going to have to disagree with you.

    I personally find the current breakdown of *adventurers* (68% human, 32% demihuman with halflings to be the smallest portion) just fine.

    Consider first the idea that ANY human can be an adventurer. Doesn't matter what they're ability scores are, if they've got the ambition to "get off the farm," they can and will. Of course not all are going to be good at their chosen profession (hence the PR penalties to earned XP), but that's okay...they still have that human potential for greatness.

    Demihumans just aren't the same. Sure they have some neat racial abilities but it's an exceptional elf who leaves the forest, or dwarf who comes down from the mountain, or halfling who leaves their peaceful garden. Halflings (based on hobbits) are the LEAST probable adventurers...a human is more than eight times as likely to adopt an adventuring career as the peace-loving Shire folk. That dwarves and elves share the same proportion makes sense...while there may be fewer elves, they are more prone to wandering the wide world than the dwarves who enjoy the security of their caves and clans.

    One-third of the (B/X) adventuring population ends up being demihuman because they begin the game more competent than their human counterparts (there may be humans who set out to be adventurers that never made it to "veteran" status, consider the entry for Normal Humans in Moldvay, and how many NH men-at-arms/torch-bearers actually make it back alive). That there are twice as many human adventurers as all other (PC) races combined seems appropriate to me.

    Every B/X campaign I've run has started with a lot of demihumans. As the campaign goes on and characters progress, I find the PCs that die are more often replaced by human characters till there's very few (if any) demihumans remaining.

    1. First I need to say that I actually did a "happy dance" when I saw you commenting on my lowly blog! :-)

      I think it's awesome that we can all find "our world" in the rules somewhere. You think halflings are hobbits and therefore they are reluctant to wander the world. I take the view that all the races in B/X have about the same percentage of folks who want to see more of the world, so it really just comes down to population. I don't think there's a "right" or "wrong" here.

      Having said that I should mention that this post deals with "3d6 in order" and nothing else. It turns out that the numbers change quite a bit once you throw those other B/X mechanics I mention toward the end in the mix.

      For example, if you simply filter for XP adjustment (instead of making a uniformly random choice between all possible classes) you'll get something like 22% clerics, 20% magic-users, 17% thieves, 14% fighters, 13% halflings, 8% dwarves, 6% elves. Note that this is not me playing around with stuff, it's just the "rational" assumption that a player would pick the class (and in B/X therefore race) that maximizes XP adjustment (random choice if there are several that do so). So beyond rolling ability scores, other properties of B/X actually lead to a distribution with more halflings than dwarves, and more dwarves than elves. Gets even more complex if you add other things like prime requisite adjustments.

      It's interesting that your experience has been that players gravitate to human characters over time. Sadly that's not my experience. Case in point, my weekly AD&D 2e game that has been running for almost 3 years now: An elf, two half-elves, a dwarf, and a recently arrived human. But the entire "core" is demi-human. I have fun playing in that game, but I find the party composition horrifying because really none of us do a good job at playing those "alien" races in an effective way.

    2. @ Peter:

      I'm so glad I have the power to encourage dancing outside my own children!

      My bias regarding B/X populations probably come more from the composition of the "Known World" found in the Cook/Marsh set (what BECMI folks call Mystara) than from the game's Tolkien roots (which, for the record, I think are modeled best by the original level restrictions found in least for dwarves and halflings). But you're right...there's no right answer.

      2E is a differen animal from B/X and one I don't have any long-term experience with...just a hand-full of one-offs, really. When I played AD&D there were a number of half-elves and halflings that kept pace with humans, but humans still outweighed the non-humans in straight numbers.

      I will say that, with regard to actual B/X play, I have NEVER found XP requirements to be a determining factor in which class was chosen. Ever. The 3D6 in order/adjust for PR was the usual factor...players choosing classes based on their characters' strengths. Thus, I saw more humans because it wasn't often a player rolled (for example) a high enough INT and STR to warrant a shot at elves. Dwarves and human fighters were fairly split with CON requirements being a determining factor. AD&D players, on the other hand, were far more concerned about racial level restrictions when it came to choosing actual class-race combos (thus, for example, the high density of halfling thieves). Long time AD&D players (like my brother and myself) might also have some leftover bias when returning to B/X (choosing human characters with an eye towards future limitations), but I haven't seen anyone snub a class type due to an XP one is choosing a fighter over a dwarf because of the 10% higher XP requirement, nor a thief over a magic-user because the former only needs half the XP to level. I'd guess (I don't know) that they're counting on the DM to provide enough XP that it doesn't matter.

      As far as playing up the "alien-ness," of demihumans, that's really on the DM running the game (and the setting being used), not the players. However, it sure is a lot easier when you've got a race-as-class edition like B/X...the more versatile and limit-free the demihumans, the more humanlike they become to the point where they simply carry racial stereotypes (see Dragonlance's "grumpy dwarves," "arrogant elves," and klepto kender as an example). In D20 and later editions there really is little to distinguish between them...race simply becomes a set of mechanical adjustments to add-on. Unless (of course) you really have a DM who is pressing the issue with setting constraints.

    3. Just because I don't have much time (grading!) I want to focus on your statement that XP adjustment or XP progression is not a primary factor for class selection. I wonder if you can characterize what is instead more fully.

      You say "player's choosing classes based on the character's strengths" and you seem to imply that ability scores (and maybe related modifiers) are the big thing. That's actually what I am most fighting with in my programming right now. For example, is it "rational" to max out a fighter's strength if that would mean losing a wisdom bonus (which helps with saving throws)? If the criterion is "best modifier in prime requisite" well, then it's obvious: max it out. If the criterion is "highest overall modifiers" then it's not.

      It may very well be the case that there simply is not "most rational" strategy so I might be "tilting and windmills" or whatever the expression is. Just curious what you think most players do.

  2. How about:
    17d6+5 inches to determine physical height of human.
    Optional (if Stats qualify as non human):
    4'8"-5'8" Is elf
    3'8"-4'4" Is dwarf
    2'10"-3'2" Is halfling

    1. You know... That isn't half bad! :-)

    2. Verne troyer (Mini-me) was shorter than a Halfling. Can you imagine being told "I'm sorry, you're too short to be a Halfling." At the casting call for a D&D movie?