Introduction
What this post isn't
I'm not here to talk about how inherently evil races are problematic, or how dicey and eugenics-y stat changes can be. I'm not here to explain that monocultural fantasy races reflect how people treat ethnic or cultural groups they're not part of in real life. This is cohost; you knew that. You're probably tired of hearing it and a million people have said it better than I would.
I'm here because the D&Default way of handling it is boring.
Stats aren't interesting, and single abilities that aren't very impactful are barely flavor. Monolithic cultures by race are also boring as world-building goes, and they're extremely restrictive from a character/scenario design perspective.
The goal here to come up with a philosophy for actually constructing fantasy races for an RPG.
Usual caveats up front about my experience not being universal etc.
I am particularly tuned to the way that black people (the skin-color-based over-broad non-ethnicity) and Black people (the American cultural group) are treated, seeing as I am one. I know I will have blind spots to other types of experience.
Species Not Race
First things first. I hate that we use "race" as the term. "Species" is better and more flexible in just about every way, especially for playing characters who aren't all that human-like.
The species should all be meaningfully different in some serious physical and possibly in psychological ways; let's explore what that means.
Using Species In Practice
What kinds of traits to think about
I've already mentioned that I think stat changes aren't interesting. Species should play in the unique and weird parts of your system. Need for food (increased or decreased); size-class; inverted vision (pitch darkness becomes a blinding light and vice-versa); hiveminds that use the "swarm of creatures" template rather than the "single creature" template.
Generally, I think you should limit yourself to using exactly three such traits. Not only will this keep your mechanics consistent, but it makes sure that some species don't end up way overdeveloped in comparison to others. Anyone who's spend time looking at the D&D race list knows what I'm talking about. Three traits is also just enough to make sure that the impact of at least one of those traits is felt on just about every playstyle.
Psychology and culture
Being of human biology, your thinking is inherently rooted in humanity. Therefore, for each species, you should at try to think of one or two ways that they might think or act or hold preferences that would be completely alien to a human, and vice-versa. These should be beyond normal culture-shock; think the Vulcans from Star Trek and their Total Emotional Suppression. A friend of mine once played a Kobold who held fully different views of what it meant to form emotional connections (you don't form connections with just one friend, you form a general connection with a whole group).
It is, of course, totally ok to have species who are different from humans a little bit more subtly in regards to psychology. Playing a character who is somehow different in a fundamentally inhuman way is very difficult, and generally only possible for people who find that to be an interesting narrative space to explore. The reason I said "might think or act" above is that for some players and some settings, truly alien psychology just isn't going to be viable. When you can get away with it, lean into it fully; when you can't, try to keep it contained-but-present so that avenue of narrative exploration stays open for players who can/will take advantage of it.
Traditions and cultural customs should be tied to region, class, trade, religion, or all of the above. You should only ever make cultural statements about a species in a particular context; talk about the "the culture of the Bird-Men of The City That Breathes Sand", don't just talk about "the culture of the Bird-Men". Furthermore, most environments will be physically hospitable to more than one species, and assuming they can communicate at all they probably live together. Think about the civic infrastructure that needs to be in place: wide doors for minotaurs, tall-but-sturdy-and-climbable stools for kobolds.
Gender
Gender is a social construct and so you should be willing to go absolutely hog wild. There is much to be said on this topic but my one real note is that unless you only have meticulous role-players among your group, you probably shouldn't try to have players use more than one or two purely-fictional in-universe "neopronouns" (for lack of a better term). It's not that it's a bad idea in fiction generally; it's a good idea, it's just that players will tend to forget to use them/what they are altogether if there are more than a very small number.
The other thing to note is that humans in a universe with multiple species living in one society probably have very different ideas of gender than humans who only have one species to draw from. Exactly how would depend on what gender categories exist in the species/cultures they regularly interact with.
Language
Language is generally thought of as a cultural feature, but consider: a parrot can make a lot of noises that you can't, and you can make a lot of noises a parrot can't. I usually have characters who can more-or-less approximate each other's vocalizations, but having a character who is understood to be "Chewbacca-ing" (where the player says what they mean and this is understood to in-universe be a complex series of noises) can also work just as well. Telepathy is also available for creatures who don't have mouths or ears in a traditional sense.
As with all cultural features, language should not be tied directly to species in a monolithic way unless the population of a species is truly so small and concentrated that there is only room for one language.
People as monsters
It should go without saying that every species sapient enough to talk that isn't godlike in some way (like a dragon) should be playable. This will avoid the "beastification" of goblins, orcs, etc. Having the flexibility to basically just template bandits or pirates or whatever and then just give them various species when building individual encounters will also lend itself to more interesting design on all fronts.
Humans
A case study in species design
Humans are the species that most games make "the default". This is bad, not only because it centers humans as "normal" and the rest of the species as "alien", but because it makes humans narratively and mechanically uninteresting. Let's fix that by considering what the species traits for a human should be. What came to mind for me designing Project Theseus were the following:
-
Our bone structure is designed to throw things; therefore, all throwing weapons gain double range when wielded by a human.
-
Humans can sweat and are exhaustion hunters. Advantage to any check involving long-term physical endurance, such as distance running or staying awake for unnaturally long hours.
-
Humans can consume way more alcohol than other creatures can. Advantage to social checks when drinking if the other characters involved are drinking. Furthermore, humans have advantage on checks made to stretch rations or drink without becoming poisoned.
These are arguably situational abilities, but most human characters will find uses for some or all of them. They inform what humans are capable of without specifying what our culture is. When writing abilities for other species, these are the kinds of things you'll want to be considering.
Humans with hats
As much as I hate to admit it, some people are just going to want to play an elf or dwarf or halfling or some shit like that. This is where Humans With Hats come in. Since species are made of traits, the "hominids" can just be humans with a single trait swapped out.
Dwarves swap the throwing arms for the ability to operate in darkness; elves swap the alcohol tolerance for a lack of need to sleep; gnomes swap the exhaustion hunting for natural magic, etc.
This is honestly something I did mostly for compatibility with standard fantasy settings, but once you get a big pool of species traits, you can start slapping them together and figuring out what kind of species would have a need for that combination of things. It's fun, and also why I designed way too many kinds of bird-man.
Conclusion
This post is too long to be one post already so I'll stop here. I originally planned two other case-studies, but I'll save them for later because it's late and I am le tired.
