D&D 5E Fivethirtyeight Article About D&D Race and Class Combos

[Yaarel]

Players seem to strongly want overall character concepts that combine a specific race and specific class together into a single comprehensive idea and flavor. It is vital that D&D makes sure that these specific race-class combos synergize optimally.



The popular classes and races are an abstraction. On average, the most popular classes are Fighter, Rogue, Wizard, and Barbarian. And the most popular races are Human, Elf, Half-Elf, and Dwarf. But the raw data displays the specific incidences. So the most popular *characters* are as follows. (Here the top-20 characters cluster together according to their decimal magnitudes, approximating the inverse Golden Ratio.)

Top 20 Characters

Human Fighter

Elf Ranger
Elf Wizard
Human Wizard
Human Rogue

Human Cleric
Human Paladin
Elf Rogue
Dwarf Cleric
Tiefling Warlock
Dwarf Fighter

Human Monk
Half-Elf Bard
Halfling Rogue
Elf Druid
Goliath Barbarian
Human Ranger
Human Warlock
Half-Orc Barbarian
Dragonborn Paladin



So, the Human Fighter is the most popular character that players play. It is at a magnitude all its own. (Would love to know about Variant Human and Battle Master Fighter!)

Then come the Elf Ranger and Elf Wizard. These are top priority race-class combos. Both of these are extremely important to make sure that the mechanics for the Ranger and Wizard classes and their respective Elf cultures, all have extremely good mechanics and synergize excellently. It is important to get a Ranger class (or several spin-off classes) that most players are happy with − especially Elf players. Already, the Wood Elf synergizes with the Ranger. It is also important that the High Elf synergizes with the Wizard.

The fact the Elf Ranger is so popular, while the Human Ranger is significantly less popular, is one of the indications that players are choosing for the sake of overall character concept and flavor.

Human Fighter, Human Wizard, and Human Rogue seem good indications for popular classes, but also typify the flavor of Human culture!

Then come the Cleric − presumably because of the perceived need for a dedicated healer − and the Paladin holy warrior. Altho these two classes to slightly less well on average when abstracting the most popular classes, they make a good showing here in overall character concepts.

The flavor of the Paladin is decisively Human. Yet the remixed fusion of the Knight-v-Dragon archetype seems to find interest and traction in the form of a Dragonborn Paladin.

The Elf Rogue makes an appearance − perhaps as much because of the flavor of physical grace and elusiveness, as much as for its Dexterity mechanics. But the most popular choices of Ranger and Wizard evidence the prevailing flavor of the Elf is innately magical.

The Wood Elf Ranger class seems to have inherited all of the traditions of the early D&D High Elf being the ‘Fighter/Magic-User’. The Eldrich Knight seems unable to serve in this capacity. Thus, in the attempts to improve the Ranger class, one of its options must have the Wood Elf in mind to synergize with it. This kind of Ranger must be a magical warrior, with strong gishy overtly magical offense spells, as well as woodsy wilderness flavor.

By contrast, the Eldrich Knight Fighter seems less significant for the Elf.

The Elf flavor is either Ranger gish or Wizard full caster. These are the vibrant archetypes for the Elf cultures.

Dwarf Cleric and Dwarf Fighter. In that order. These are virtually the only concepts that typify the D&D Dwarf cultures. Probably each should dominate one Dwarf culture. Wisdom Cleric prevailing among the Hill Dwarf, and Strength Fighter prevailing among the Mountain Dwarf. Make sure these class-culture combos synergize excellently.

The fact Tiefling Warlock enjoys significantly more popularity than Human Warlock, suggests most players want an overall character concept that combos both race and class. A comprehensive concept.

Half-Elf Bard. Half-Elf only appears in the top 20 characters as a Bard. This is currently the only place where the Fey flavor can happen for Charisma as charm, magic, beauty, and art. Similarly, the Elf of the Feywild is primarily a Charismatic Bard culture.

The only time Halfling makes an appearance is as the childlike Halfling-Rogue combo.

The only time a Goliath makes an appearance is as the rugged Goliath-Barbarian.

The only time a Half-Orc makes an appearance is as the savage Half-Orc-Barbarian.

The only time a Dragonborn makes an appearance in the top 20 is as an ironically noble Dragonborn-Paladin.

The only time the Druid class makes an appearance is because it happens to be part of the comprehensive Elf Druid combo concept, a woodsy full caster. Nevertheless, even more so, Wood Elf flavor of the Elf is moreso a magical warrior gish, a Ranger. The High Elf flavor is moreso the full caster, a Wizard. The two types that prevail for the Elf are the Wood Elf Ranger and the High Elf Wizard. The game works better when both of these concepts are mechanically optimal.



In sum, the evidence demonstrates how players are choosing specific race-class combos, rather than races alone or classes alone. Together these comprehensive combos are what define a ‘type’, the archetype that players want to play. Players choose this because the specific combo is what creates a flavor.

Advanced players will occasionally want to play ‘against type’. But each type itself − as an optimal synergistic combo − must work well in the first place.

 

[Alexemplar]

I'm pretty sure his point was not that Hercules is a Barbarian and Aragorn is a Paladin. It was that each of them (and every single other one) demonstrate characteristics of more than one D&D class.

I.e. they can't be Fighters because of the idea that a "Warrior who doesn't have any kind of supernatural/magical aid or items or distinguishing exploration/social/cultural tricks" can't be modeled with a Fighter.

 

[Mephista]

The flavor of the Paladin is decisively Human. Yet the remixed fusion of the Knight-v-Dragon archetype seems to find interest and traction in the form of a Dragonborn Paladin.

Dragonborn paladins of Bahamut first made their debut in 3e, and carried over into 4e. Spellscales, a companion race to dragonborn in the same book, in 3e were the dragon sorcerers. A mix of paladin and sorcerer for dragonborn has been a thing for two editions now, and people remember it, I'd say. They're the two iconic classes, much like is cleric/fighter for dwarves, and informs their culture.

Its not ironically noble, its the iconic image of the race.

 

[Imaro]

I.e. they can't be Fighters because of the idea that a "Warrior who doesn't have any kind of supernatural/magical aid or items or distinguishing exploration/social/cultural tricks" can't be modeled with a Fighter.

You want a fighter who heals like Aragorn... take the Healer feat as a bonus feat. Want to be really strong like Hercules... raise strength, seek out a god for a boon or find a magic item to increase your strength beyond normal means... Hercules wasn't born with super strength he achieved it because he nursed on god's milk (He nursed from Hera)...

 

[Tony Vargas]

There is a tendency among all people to overestimate the prominence of whatever subgroup they happen to be in.

Interestingly, when viewing a group of people, there's also a tendency to over-estimate the overall prevalence of the 'others' in that group.

... and so it goes with D&D. If you really enjoy system mastery, you tend to believe that this is the natural state of things. OTOH, if you don't value system mastery as much, you also tend to believe this is the natural state of things.

But... but.... no, we're D&D-playing-nerds, we're a rarefied, exclusive, intellectual elite! ;P
Seriously, though, whatever the individual bias about system-mastery or Role-not-Roll or pick-your-GNS-letter or whatever, the underlying system is the same starting point in each case. The system master, thespian, and old-schooler may have different ideas about whether a character 'should' have a 20 or what it means - but it's a +5 bonus for each of them, regardless of that bias.

*One of the reasons I enjoy looking at numbers like this, and running polls and what-not on enworld is that it gives me a little insight into the great variances of opinion that we TTRPGers have. And I find the most joy not in the numbers that agree with me, but rather in the ones that challenge what I want to be true; for example, I am not a big fan of the Tieflings and Dragonborn, yet I accept that they are very popular races.

They're relatively popular - moreso than you might expect from the general prejudice against all things 4e. Perhaps, in part, because the Tiefling is way older than that, and the Dragonborn a fair stand-in for the Munchkin-beloved Half-Dragon.

My favourite aspect of this statistics is that the MOST popular class is ONLY TWICE as popular as the LEAST popular class. It might sound like a huge difference, but it is not. This tells me that all 5e classes are good solid design.

The arguably worst-designed class is also in the middle of the pack. I think it's more likely that the popularity of the class has nothing at all to do with the quality of the design. If quality of design were a high priority for you, you'd likely be looking hard at 13A and indie games after suffering through a few sessions of D&D. ;P

(I love D&D, honestly, but I'm not delusional about it. It's designs have often been indifferent. On the rare occasion it produces something functional or elegant, it's generally ruined by association with the rest of the current version of game and/or its environment.)

Yes and no. A lot of the spells used by either class are actually in tune with what could be defined in mundane abilities. If a Ranger uses Cure Light Wounds, you could interpret as magical or just simply as a nature-based healing ability ("here - put this herb on the wound to heal better)

That would be cool, but D&D has never actual gone there officially. (Ironically so, because EGG was letting his players do just that back in the primeval period - see if you can dig up the Giants in the Earth write-up of Myrlund, I think it was, a wizard re-skinned as an old-west inventor.)
Starting with AD&D the game on a more and more of rule-for-everything attitude, so there was less 'need' to have "counts as" variants, even the option of changing the cosmetic appearance of a magical spell was quantified in 2e with the spell-affecting-spell 'Sense Shifting.' The WotC era started to stray from that, even as 3e got more detailed than ever, it /did/ explicitly put the appearance of the character & it's gear in the hands of the player, so you could go pretty far afield - I once played a cleric who's caster's shield was described as a book, for a minor instance. You could mess around with race pretty dramatically. You just couldn't change the underlying mechanics. But, 4e took it too far - the player was free to re-write the 'fluff' line of his powers, and that was part of the huge push-back we got about magic no longer being magical - though, even in 4e, you couldn't actually change a spell into something non-magical, the keywords, like 'Arcane' were off-limits to casual re-skinning. 5e is in no small part a reaction to that (and many other things in 4e), and has very clear lines drawn about what's magic and what isn't and a few, somewhat important mechanics, hang on that distinction.

So if you want to heal non-magically with herbs, take the skill, take the feat - the spell makes you a caster.

When D&D was busy converting 4E to 5E, they integrated a lot of the various abilities and 'powers' for each Class into the collective spell lists. Pretty much every Class, baring the Barbarian (who still get some spirit-based abilities as options), can access 'spells' at some point.

Yep. Magical powers that in 4e were divied up by 'Source' in 5e are back to almost all being 'spells,' and all being explicitly magical. That means they can't be included in a character concept that isn't overtly magical, for good or ill. The good is that magic is 'really magical' again. The ill is that concepts that can't accommodate magical abilities build from a very limited set of blocks: Berserker, Campion, BM, Thief & Assassin.

And, yes, every 5e class /can/ use spells in one form or another - the Totem Barbarian only in the form of a few rituals, Monks by fueling them with Ki, everyone else with actual spellcasting. It's really more a case of a few classes - Monk, Barbarian, Fighter, Rogue having the option to not use spells in one or more sub-classes, and even fewer having the option to not use magic, at all - Barbarian, Fighter, & Rogue, only.

The prevalence of hard-coded magic, particularly spells, in 5e is one of the factors that channels players to those last three classes that are only soft-coded to use magic, that is, have magic abilities available in only one sub-class.

All three of those classes, even the narrow-concept Barbarian, are in the top 4 classes in the data, and the fighter, the class the designers have lamented making 'too generic,' is firmly in the #1 spot.

The dude had a holy sword, and "The hands of a king are the hands of as healer."

Point, the Autherian cycle did get deeply religious (even if there were uncomfortable bits, like Excalibur essentially coming from a sort of genius loci), and 'King's Magic' was certainly a thing. Those abilities don't remotely map to D&D spells, but, yeah, between Lay on Hands not technically being a spell, and burning slots exclusively to smite, a Paladin could do well. I'll relent on him - and Galahad/Percival was certainly a major inspiration for the Paladin, anyway. Most of the rest of the Knights of the Round Table, though, didn't go that far and D&D could only attempt to model them with fighters.

You're pretty much just revising any warrior type into Fighter, by creating unrealistic definitions of other classes and leaving Fighter without its own

I didn't do that, the system did. The system gives Rangers and Paladins spells, and Barbarians Rage, and makes Rogues dependent on SA in combat, the system gives Paladins & Barbarians very strongly-defined, relatively narrow, conceptual space.
The system left the Fighter comparatively generic in both ability and concept.

That makes them the default. It's far from ideal - the fighter often lacks abilities a given character /should/ have, but can't gain without accepting abilities it shouldn't - but there's no other way to parse it.

Aragorn is defined by his healing hands, you know.

It was an herb, and had more to do with his heritage than his 'class.' It wasn't a spell. Healer feat would probably be a closer fit in 5e. A custom Background, perhaps, though things he eventually did went beyond that - more like a 4e Epic Destiny, really. But casting spells every fight to shoot people better? Nah. Fighter - or spell-less Ranger were that an option.

This argument is basically boiling down to just "any warrior defaults to Fighter if it doesn't fit a narrow definition."

One of several narrow definitions, yes, exactly. Because the other classes are narrow definitions and 'you must cast spells' is a hard-coded part of most of those definitions.

I want to say it's almost trying to shame the game and those who play it because it's popular. Kind of silly actually.

The more so because D&D, even at it's height, has been enjoyed by only a tiny minority of people. It's positively 'elite' in that sense. More pistachio or roasted-garlic* than vanilla in the ice cream analogy....













* not as bad as you might think, if you ever go to the Gilroy Garlic Festival, give it a try.

 

[lowkey13]

*Deleted by user*

 

[Imaro]

The more so because D&D, even at it's height, has been enjoyed by only a tiny minority of people. It's positively 'elite' in that sense. More pistachio or roasted-garlic* than vanilla in the ice cream analogy....













* not as bad as you might think, if you ever go to the Gilroy Garlic Festival, give it a try.

You keep saying things like this but I'm pretty sure almost everybody understands what's meant is that it's popular within the context of roleplaying games and even editions of D&D...

 

[Tony Vargas]

what's meant is that it's popular within the context of roleplaying games and even editions of D&D...

Nod. It's just equally invalid reasoning in either context.

"Oh woe! D&D sucks because it's too popular!"

Invalid.

"Oh woe! D&D sucks because it's not popular at all!"

Invalid.

Because D&D is a dynamic game, with (hopefully?) a DM that adjusts the challenges based upon what the players want, and what they are doing.
So in the end, it doesn't matter whether your bonus is +3, or +5. It really, really doesn't. Because the difficulty can just be scaled against the higher bonus.

Yeah, relatively few DMs get to the point of dynamically adjusting it for different characters: You have a 12? Your DC is 15... oh, you have a 20, your DC is 19...

That's why the bias doesn't matter. Plus, I mean, in the long run, we're all dead anyway. So there's that.

Thank you John Maynard Keynes. ;P

 

[Lanefan]

Let's see!

People that hit stuff-

FIGHTER 12.7%
BARBARIAN 8.3%
RANGER 8.1%
PALADIN 8.1%

Total- 37.2% (Predicted 40%)

Fine so far...

People that pray and steal-

ROGUE 10.4%
CLERIC 8.3%


Total- 18.7% (Predicted 20%)

But here you've mis-read what I wrote. Cleric and Thief should add to 50% in total on a 20-30 breakdown, my problem was (and still is) I can't remember which one's supposed to be 20% and which one's supposed to be 30%.

Also Druid should go in with Cleric.

Thus what we get is:

ROGUE: 10.4%

Total 10.4% (predicted - either 20 or 30%)

CLERIC: 8.3%
DRUID: 5.8%

Total 14.1% (predicted - either 20 or 30%)

Interesting! (Admittedly, you can mess with the numbers a little, like putting Sorcerer in the Wizard category, etc. But still!)

About the only consistency is that Fighter-types are still around 40% and Wizards around 10% - everything between seems to have been thrown in a blender.

Lanefan

 

[Lanefan]

I guess I would look at this data in a different light and question just how close it is to being an accurate representation of played classes. As has been mentioned, just because a class is built in the character builder, doesn't mean it has been played. I would take it a step further.

I would suggest, which at least in my case is true, that the classes being made the most are often not played at all. For example, one of my current two active characters is an Eldritch Knight. To that end, I don't believe I have built a fighter in a builder in months. What I have built are various versions of the next characters that I look forward to playing in the future, but haven't played yet and may never get to play. So the characters being made are the ones that haven't been played yet. I've played and built one fighter, but I've built and theory crafted dozens of other characters.

The second thing I would consider is that, I've built more characters in concepts that I can't get to work then in ones that I have. Barbarian for example, I have never played, but I have a pretty straightforward concept and build that won't change much. I haven't spent much time on trying to redo it because it just works. In comparison, I've spent a tremendous amount of time building Rangers and Sorcerers in the hope of making one that I like. Unfortunately, I find both classes extremely underwhleming and have yet make a build I would like to play. That doesn't stop me from trying, however. In the end, I've built many, many more Rangers and Sorcerers than Barbarians, even though I would likely never play either a Ranger or a Sorcerer with the current ruleset.

So maybe what we need to do to augment the data is have lots of people here post the stats on what's been played in their game(s). Could be done via one of those wiki threads, I suppose, listing all the classes and just getting DMs to add numbers to a total - though I'm not sure how to prevent crossposts and multiple people trying to edit it at once. If I'm bored later today maybe I'll set this up...see if it's any use or not.

Of course, this could get messy once people start including homebrew classes and suchlike, though many of these can often be linked to (and thus parsed as) an existing class or multi combo.

Lan-"and this is one case where I can be a truly neutral voice"-efan