In The question of Race, or nature versus nurture I looked at races in Dungeons and Dragons, starting with BECMI and AD&D, and then moving onto 5e, with a view to developing cultural traits which could be swapped about for a character with a certain genetics raised by a different race.
In The Question of Race part 2 – comparisons I looked at the standard races from the Players Handbook, with reference to Tasha’s Cauldron of Everything to try to compare them and start to build up culture packs…and ended up concluding that the balance across genetics and culture for different races was too different to allow sensible balanced swappable culture packs. I was also struck by how imbalanced the different races appeared, and so I switched my attention to working through all the playable races to come up with ratings for each. This led to the following posts:
- The question of race part 3 – Volo’s Guide to Monsters compares the playable races from Volo’s Guide to Monsters
- The question of race part 4 – Ravnica, Theros and Ravenloft compares the playable races from Guildmaster’s Guide to Ravnica, Mythic Odysseys of Theros and Van Richten’s Guide to Ravenloft.
- The question of race part 5 – Eberron looks at the races from Eberron: Rising from the Last War/Wayfinder’s Guide to Eberron
- The question of race part 6 – Wildemount looks at the playable races from Explorer’s Guide to Wildemount.
Now that I have worked through all the races, it seems appropriate to look back. As I have rated more of races I have found my ideas shifting a bit. I have also had discussions in a couple of communities about my ratings, with various people disagreeing with my choices for scoring. So I thought it worth explaining the thinking behind my scoring, discussing some of the objections I have received, and looking back at the scores to see whether, knowing as much as I now do (having gone through the whole process) I want to tweak the scores I have assigned.
Initial rationale
As a reminder, my original goal was to split the racial attributes between genetic and cultural depending on whether Tasha’s Guide allows them to be swapped about, and then come up with “racial culture” bundles which were equivalent in benefit which could be taken to reflect an upbringing different to the one associated with the genetic race of the character. For this I was aiming for a fairly simple rating scheme which would be easy to calculate and apply.
I came up with the following guidelines:
- Ability score increase (cultural): +1 point per increase
- Proficiency (cultural): +2 points for skill, +1 points for double proficiency specific cases, +1 if light weapon, light armour or tool, +2 if martial weapons, medium or heavy armour
- Disadvantage (cultural): -2
- Spellcasting (cultural): +1 for cantrip only, +2 if additional at #3 & #5
- Language (cultural): +1 point per language
- Darkvision (genetic): +2 for 60′, +4 for 120′
- Resistance (genetic): +2 per attribute
- Speed (genetic): +1 per 5′ over 30’/-1 per 5′ under 30′
- Other special attributes to be scored similarly
My rationale was as follows:
- +1 to an ability score has a 50% chance of +1 to rolls against checks using that ability; +2 is required to get a +1 bonus to the roll. On the other hand it can affect multiple different skills, saving throws and checks.
- Proficiency in a skill gives a +2 to rolls against that skill at level 1, increasing to +6 at the highest levels, but it is situation-specific. There is an argument that different skills should be valued differently given the amount they get used, but treating them all the same keeps the rating simple.
- Given Proficiency doubles the bonus from an ability score to +2, but is more specific, it seems reasonable to rate a Proficiency equivalent to a +2 on a ability.
This sort of set a baseline of +1 for minor features, and +2 for features with more benefit.
- I considered light weapon/armour proficiency and tool proficiency of less impact than proficiency with martial weapons and medium or heavy armour, so +1 for light weapons/armour and tools and +2 for the heavier hitters.
- Advantage seems larger, so +2, which makes disadvantage -2.
- Darkvision can obviously have quite an impact in the right situations – it allows the character to function in darkness, so I gave it +2, and similarly damage resistance.
- Languages seemed less impact, but incremental for each language, so +1 per language.
- It felt like there should be a bit of a penalty for moving slower, so I gave dwarves and halflings a -1 for their 25’ movement. This suggested a rule of thumb of +1 per 5’ of movement above 30’ and -1 per 5’ of movement slower. It’s a bit of a benefit/drawback, particularly in a chase, but not a huge one.
- Spellcasting: some classes get a cantrip at level 1, and other spells at levels 3 and 5. Some just get the cantrip. So +1 for the cantrip and +2 if there are other spells as well.
- The trickiest thing to rate was the Variant Human Feats, and this is what has been most controversial in the discussions about the series. I skimmed through the feats, and I picked them up as roughly maybe an ability +1 and/or a proficiency, so I rated them as a larger feature, i.e. a +2. In retrospect they are probably worth more than that.
I decided not to rate age ranges or differences in size, since realistically they’re not going to affect most campaigns, and to completely ignore suggested features like alignment which are ultimately a player choice.
Updates from later books
As I worked through the other sourcebooks, other features started to emerge, and I rated them in a slightly (but not very) methodical manner.
- We start to get different breathing – Lizardfolk able to hold their breath underwater for 15 minutes, Reborn not needing to breathe, Triton amphibious. This seems major, so +2.
- We get different methods of movement, such as the Tabaxi climbing speed and the Triton/Lizardfolk swimming speed. I’ve rated this at +1 for 20’, +2 for 30’ and +3 for the Aarakocra 50’.
- Immunity definitely seems worth more than resistance, so the first +4 after 120’ darkvision.
- Long reach and not slowing down with heavy loads I both rated at +1.
- We start to get features which are once per short or long rest, such as the Firbolg hidden step or the Goliath Stone’s Endurance. I’ve typically rated these as +1 due to the limitation of effectively once per encounter.
- We start to get natural weapons. These substitute for unarmed attacks, so turn that 1+STR into (typically) d4+STR or d6+STR. Given that’s less than most weapons, this only seems worth +1.
- Similarly natural armour; I’ve rated this +1 for the Lizardfolk (AC13) and Loxodon (AC12), and +2 for the Tortle (AC17). Which is possibly inconsistent with proficiency giving +2…
Ultimately, however objective I try to be there’s going to be a degree of subjectivity…
From around the web
As the series continued, I started to get feedback, and unsurprisingly I’m not the first to have thought of this. After all, it was all inspired by a post by The Angry GM – Why race isn’t broken in Pathfinder, and how to fix it. There he does actually come up with balanced splits for the main races between genetics and culture (for Pathfinder, not 5e).
Then there’s Bruce Heard’s recent series developing classes for the BECMI ruleset. You will remember that originally elf, dwarf and halfling were actually classes in those rules rather than races which could have classes layered on top. Bruce has continued this with his gnome and half-orc, although he has also added a paladin and a (Calidar) ranger among others. Now BECMI already has a built-in mechanism for balance – the more highly-featured the class, the more experience it takes to go up a level. So thieves and clerics go galloping ahead with 1200 and 1500 as the basic increment for level 2, magic users toil behind with 2500, and elves (both magic and less squishy) need a full 4000XP for level 2.
So Bruce has worked out a spreadsheet to allow him to calculate the relevant experience thresholds based on the features of the class. This is fairly coarse-grained. You rank the class on ten attributes on a 3-5 step scale, and it adds XP to the threshold appropriately. The attributes are hit die (the size of die for rolling hit points, where their combat abilities fall, how good their saving throws are, whether they have cleric spells or magic user spells (and if so whether full-blown or limited), whether they have unique class or racial abilities or special skills, and what limits (if any) they have on weapons and armour. The lowest level in each scale costs 0XP and it goes up from there.
Detect Balance, a 5e homebrew race guide
But neither of those address D&D 5e. What does is this Google spreadsheet, apparently originally created in 2016 but updated only yesterday (as I write in July 2021). This is incredibly detailed with scores ranging from 1 to 20 for different attributes, and seems to be quite widely used for home-brew races. In this Reddit thread, jwbjerk comments about how unbalanced the different Volo races are, so they’ve obviously been struck the same way as me. In looking at the spreadsheet, though, I do find myself disagreeing with some of the ratings and in particular with how fast things can scale. Because the numbers can get large, it feels a bit like the values coming out are precise but not necessarily balanced. It does also try to allow for synergies and in particular areas where races get features which wouldn’t combine well in practice.
It starts off with a much wider scoring range, from +1 to something not likely to make much difference, through +2 for a sometimes useful feature and +4 to a frequently useful or powerful feature to +8 for an unusually powerful feature. This feels like too much of a swing. And then looking down there are things which are worth 12, 16 or even 20. I don’t see that one feature can be 8-16 times as useful as another.
Things I notice looking at this:
- They rate ability score increases as more valuable (relatively) than I do – effectively +4 per point.
- Conversely, they rate languages as almost useless – no points for the first language, 1 point for two or three languages or for a choice of language. Only telepathy scores highly (at +6).
- Proficiency – tool proficiency is +1, i.e. almost worthless, skill proficiency is +2, or +3 if you get to choose, and +5 if you can choose two. Expertise is +1 on a rare subset, and +2 on a general subset. When combined with proficiency it becomes +6
- Spells range from +2 for a cantrip or a level 1 spell once per day to +4 for a level 2 spell immediately once per day.
- Resistances are +2 for rare attacks up to +4 for common attacks, or +10 for the standard melee damages (bludgeoning, piercing, slashing), but vulnerabilities are considered much more of a drawback than the benefit from resistances, starting at -8 for a rare vulnerability.
- Movements are fairly normal, though flight is considered a massive advantage, and darkvision varies from 1 to 4 depending on range; blindsight is +8 unless blind beyond 30’.
- Advantage scale depending on how common the roll is, but again I feel it scales too far. Weapons training is considered fairly minor; similarly unarmed strikes, although a d8 rates pretty highly at +6.
- Breath varies from +1 – can hold breath for 15 mins – to +4 for doesn’t breath, and natural armour is only considered significant if +12 or above.
- Powerful build is a +2.
- And then we come onto “assorted racial features” and here is where it just looks like a random number generator. Can you really distinguish between feature to the level of 4 vs 5 vs 6 vs 7 vs 8? It’s precise, but it feels like spurious precision which gives the illusion of accuracy.
- And the feat – a massive +20! Are they really that powerful? Obviously they think they are, and that’s also the message I’ve received from some people looking at my series, but it seems a bit like “well, humans ought to be good, so let’s pick a big number to skew it”.
Time for review
Now I’ve gone through all the races, had feedback and seen Detect Balance, I think it is worth revisiting my weightings and reviewing the scores. But I think this post has already waited long enough, so that’s a topic for the next post in this series…
Here are the ratings as they stand at the moment, including the corresponding Detect Balance ratings. My ratings run from +5 (kobold) to +16, Detect Balance ratings run from 16 (kobold) to 47
| Race | Genetic | Cultural | Subrace | Overall | Detect Balance |
|---|---|---|---|---|---|
| Aarakocra | +5-1 | +5 | +10-1 = +9 | 41 | |
| Aasimar | +5 | +4 | +3-4 | +12-13 | 31-32 |
| Bugbear | +4 | +8 | +12 | 30 | |
| Centaur | +6-1 | +6 | +12-1 = +11 | 27 | |
| Changeling | +2 | +9 | +11 | 28 | |
| Dhampir | +10-2 | +4 | +14-2 = +12 | ||
| Dragonborn | +4 | +4 | – | +8 | 21 |
| Dragonborn – Draconblood | +4 | +5 | +9 | ||
| Dragonborn – Ravenite | +4 | +5 | +9 | ||
| Dwarf | +5-1 | +8 | +3 | +16-1 = +15 | 29-30 |
| Elf | +6 | +5 | +5 | +16 | 27-28 |
| Elf (Wildemount) | +6 | +5 | +5-7 | +16-18 | |
| Firbolg | +2 | +7 | +9 | 26 | |
| Genasi | +0 | +3 | +3-7 | +6-10 | 19-25 |
| Gnome | +4-1 | +1 | +3–5 | +9–11 – 1 = +8–10 | 23-24 |
| Goblin | +2 | +7 | +9 | 26 | |
| Goliath | +4 | +6 | +10 | 24 | |
| Half-elf | +6 | +10 | – | +16 | 29 |
| Halfling | +6-1 | +1 | +2–3 | +9–10 – 1 = +8–9 | |
| Half-orc | +6 | +6 | – | +12 | 24 |
| Hexblood | +4 | +6 | +10 | ||
| Hobgoblin | +2 | +7 | +9 | 25 | |
| Human | +0 | +7 | – | +7 | 17-33 |
| Kalashtar | +8 | +5 | +13 | 28 | |
| Kenku | +2-2 | +9 | +11-2 = +9 | 21 | |
| Kobold | +2-4 | +7 | +9-4 = +5 | 16 | |
| Leonin | +6 | +6 | +12 | ||
| Lizardfolk | +7 | +9 | +16 | 29 | |
| Loxodon | +5 | +6 | +11 | ||
| Minotaur | +4 | +6 | +10 | 20 | |
| Orc | +4 | +6-2 | +10-2 = +8 | 18 | |
| Extended Orc | +4 | +8 | +12 | ||
| Reborn | +10 | +4 | +14 | ||
| Satyr | +7 | +9 | +16 | ||
| Shifter | +6 | +0 | +5 | +11 | |
| Simic Hybrid | +6 | +4 | +10 | ||
| Tabaxi | +6 | +8 | +14 | 25 | |
| Tasha custom | +0 | +7 | +7 | ||
| Tiefling | +6 | +1 | – | +7 | 23 |
| Tortle | +6 | +6 | +12 | 28 | |
| Triton | +6 | +7 | +13 | 28 | |
| Vedalken | +2 | +12 | +14 | ||
| Warforged | +8 | +8 | +16 | 39 | |
| Yuan-Ti Pureblood | +8 | +8 | +16 | 47 |
