In my History of the Classes column on Tribality.com, I mentioned the idea of running a single-class 5e campaign, permitting only the Paladin class. Rather than leave it there, though, this post explores the potential of each class for a single-class campaign and similar deviations from "normal" party breakdowns. This article mostly assumes that the reader is a DM looking for a new and unusual campaign concept. I realize that most of us have had an idea like these before, but maybe I can put a new spin on it... or maybe you can teach me something in the Comments.
First, some terminology - if there's official or widely-accepted terms for these ideas, I don't know about them, so here's what I'll be using.
Single-class Campaign: Every PC has the same character class, though they have differing subclasses and other features. I might describe a campaign allowing only Paladins as a Single-class Paladins game.
A single-class campaign may be Open or Locked; Open signifies that you can multi-class freely, while Locked signifies that that one class is the only class any PC may take in the campaign at all.
Limited-class Campaign: Most PCs have the same class, but up to one party member at a time can be of a differing class. For example, a Limited-class Barbarians Campaign might find strong thematic justification to allow one player to bring in a Druid, even when the choices are otherwise Locked to Barbarians.
Single-background Campaign is pretty self-explanatory. Players can pursue the variations within a background freely. Because of how class and background sometimes touch on nearly-identical thematic notes, sometimes you can accomplish the same thing with single-background that you might otherwise set out to do with single-class, so you've got to consider whether what you care about is "only" thematic elements, or whether the mechanics-intensive classes are critical to getting at what interests you about the campaign idea. I definitely do not recommend pairing single-class and single-background in the same campaign.
A Dual-class Campaign limits players to related classes - say, Fighter and Paladin, or Sorcerer and Wizard. I don't have individual terms for all of the different ways you could classify these - a campaign could be limited to Clerics and Druids (thematically hanging on characters who behave as priests for their communities), or Clerics and Paladins (thematically hanging on characters in service to, say, a monotheistic faith), or Clerics and Warlocks (thematically, characters who are bound into service of a Greater Entity).
Without further ado, the classes, and a few pointers on how to hang a campaign on campaigns all about them.
The Barbarian game sounds great. You have a bunch of characters from the same tribe or group of tribes, playing out a conflict with enemy tribes or a civilizing empire. I mean, this is the premise of Hillfolk. In 5e, though, the Barbarian has but two subclasses, the Berserker and the Totem Warrior, so minor differences in build and personality take on greater significance. The Totem Warrior has a fair amount of internal variation, so a party of Barbarians might have a scout, a leader, and a tank within Totem Warrior, plus a Berserker or two. The absence of healing would be a substantial problem, so either the DM takes this into account with encounter design, or runs the campaign as an Open single-class campaign, and a few PCs pick up levels in Druid. Thematically, this is really friendly to the Totem Warrior - bear totem warriors who can turn into bears? Sure, why not?
This core concept also works quite well as a single-background campaign - everyone is an Outlander. It might call for a rewrite of the Outlander bonds, flaws, and ideals, though, as they're predicated to some extent on Outlander-as-outsider, where in this campaign they are the norm. This approach misses out on the interesting element of each character having a mechanically-relevent totem (for Totem Warrior barbarians) or... whatever you decided to do with the Berserker in that narrowed environment. Just thinking about this makes me want to write up a new barbarian subclass - I've been batting around concepts like "demon-scarred" or "death-scarred." Not sure where I'm going with those yet, but I'll probably do something with them eventually.
The Uthgardt tribes or the Rashemi of Forgotten Realms make especially good foundations for the barbarian campaign. It might be most appropriate to those cultures to keep the caster classes as NPC roles, as even the "friendly" spellcasters of those cultures have a mystique that can breed distrust, but a dual-class campaign with Valor bards, Druids, or Nature/Tempest Clerics for the Uthgardt, and Wizards as witches to present Rashemen, could be pretty excellent.
The Bards game has two things going for it, and one thing potentially against it. To its benefit, bards are incredibly good at cooperation - their teamwork-related class features and spells are amazing, and since they can't use these abilities on themselves, there's not much danger of overlap. Secondly, everyone can easily imagine a troupe of performers having wild and crazy adventures, right?
On the other hand, there are only two Colleges for bards in the Player's Handbook, and neither of them have any significant mechanical variation (as compared to the Totem Warrior barbarian or the Battle Master fighter, say). What they do have going for them is that they're quite different gameplay styles. Given that their spellcasting has Spells Known rather than spellbooks or automatically learning all spells of a given level, two bards in a party might find spell selection to be a major differentiating factor. Oh, and skill selection! Having completely open options for skill proficiency and adding Expertise on top of that actually might be enough, as long as skill use is an appropriately central part of gameplay.
Just like the Barbarians game, this works equally well as a single-background campaign, with everyone as an Entertainer... or Charlatan. As with the Outlander, the Entertainer background has a ton of internal variation. If you're trying to choose between single-class and single-background, the decision comes down to whether you want a traditional scope of mechanical variation, with defenders, strikers, leaders, and controllers, or if you want to see how D&D behaves with much less mechanical variation. They're compelling approaches that scratch different itches.
This is the easy one. There are a hojillion (technical term) different domains to pick from, and their playstyles cover melee brute, ranged damage, sneaky bastard (if a little shy on damage output), and (obviously) so. much. healing. In some settings, though, it would be weird to have clerics of competing gods working together as the entirety of a team. In other settings, it's just a pantheon coming together to accomplish broader goals, or the militant arm of an Ecumenical Council. Thematically, adding in paladins to make a dual-class campaign is perfectly straightforward and reasonable.
Looking back to 4e's meta-stories, it would be super cool to set such a group against the resurgent Primordials, or to have them be the mortal servants who bind the Primordials for the first time. If tinkering with mechanics is your thing, look for a way to incorporate 4e's Invoker into 5e, either as a new class, or as a set of options within the Cleric class. 4e's Invokers are specifically about slapping down the Primordials, and they do it in interesting ways.
There's not really an On The Other Hand for this one. I guess the team's main drawback is narrowness of skill selection, but even that is pretty trivial with the Knowledge domain and the backgrounds available. There might be some shortfalls in damage output, but as long as the fights don't have extreme time pressure, an all-cleric team has a trivially easy time outlasting the opposition.
A single-background campaign using Acolytes works really well too. An organized religion as a consistent source of missions that all of the PCs care about makes life ridiculously easy on the DM.
Funny enough, the 2e Complete Druid's Handbook goes into quite a bit of detail on how to make a Druids campaign work. It's not the most intuitive thing in the world, except that druids do have these Circles that are presumably in-world organizations as well as builds. I don't know how much the Circle of the Land options provide meaningfully different gameplay, but between the healing, defensive, and offensive spells of Land casters and the different brute or sneaky options of Moon shapeshifters, there's a ton of combat potential.
I've already mentioned how well a Druid-Barbarian dual-class campaign works, or Druid-Cleric. Druid-Ranger also sells itself. What about Druid-Sorcerer? You could interpret the sorcerer's Wild Magic bloodline not as "chaotic" (though it is), but as "when magic gets in your blood, this is the natural result." Or something. I'm pretty sure there's some fertile ground there, at least.
The single-background variant on this is probably Outlander again, though that's less compelling for paralleling druids than barbarians or rangers. At minimum, I find it interesting how much thematic heavy lifting the Outlander background does in this post. If you really wanted to get away from Outlander, you could run something vaguely similar to the Druids campaign as "the Hermitage." That background implies individual solitude, but small groups are still technically correct.
As far as coming up with campaign concepts goes, this is one of the two easiest classes in the game. Unless an action adventure story is explicitly fantasy and includes spellcasting protagonists, or is explicitly roguish or heist-based, a team made entirely of fighters is the default.
Lucius Vorenus and Titus Pullo Kill Everyone (colloquially known as Rome)
There's also very strong customization within the fighter class - the greatest number of Ability Score Improvements/feats, a ton of options within Battle Master so that there might be no Maneuver overlap until high level, lots of fighting style options that are pretty meaningful to how the character plays, and then spell selection for the Eldritch Knight. Thanks to Second Wind and strong defensive options, the Fighters campaign even has great survivability without magical healing. The Battle Master has some solid leadership options as well. If you wanted to run strictly non-magical D&D, a Fighters campaign limited to Champions and Battle Masters should be a lot of fun, though you'd have to keep a close eye on how the game developed over time.
It's not really a coincidence, either. One of the first times I ever heard the name "Mike Mearls" was for a game called Iron Heroes, which is all about a team of fighters (with some roguish elements) murdering people with weapons. Technically, there is a spellcasting class in the game. It's meant to be the enemy most of the time, because Iron Heroes is all about some Swords Versus Sorcery. Which brings me to the dual-class campaign I'd suggest - the one with the sorcery. Now, there are a lot of other games more tailored to pure sword-and-sorcery than D&D, since D&D has to cover essentially all fantasy bases, but it still handles that subgenre quite well.
Soldier is the obvious choice for a single-background variant of the Fighters campaign, though Folk Heroes, Noble Knights, Pirates, and Sailors could also be great single-background campaigns to cover the same territory as the list above. Also, basically any class makes sense as a dual-class campaign with fighters - it's cutting fighters out that might occasionally require justification.
You probably heard about Avatar: the Last Airbender, but just in case you didn't, it's one of the ideal iterations of the Monks campaign. It's even restricted to a single subclass... that just happens to be a high-customization. A team of ninjas carrying out assassinations against the tyrannical and corrupt samurai also pretty much explains itself. Depending on your outlook, the Monks campaign may just be the ideal setup for every wuxia flick ever made.
The Hermit is the closest thing to a single-background campaign that parallels the Monks, but it doesn't get across most of the feel. The monk has a long list of highly specific abilities, and other classes don't emulate that feel because they're intended not to do so.
Oh, and if you're building a dual-class campaign with monks as one of the classes? For God's sake, make the other class the rogue, for their superior reflexes. Or, slightly more seriously, wait for a full psionics release, and pair the monk with the psion to run either pseudo-Star Wars or The Matrix.
Finally I get to the class that I started this whole post to write about, and the one I most want to play. Everyone knows I've written a bazillion articles about the Paladin class over in Tribality.com, right? Good. So now that I have carpal tunnel syndrome from writing "paladin" so damn many times, I want to expand a bit on some remarks I made in the sixth article, covering 5e. I think the Paladin is one of the most compelling options for a single-class campaign, because the structure of the Oaths means that there's built-in roleplaying conflict that is all about honor and duty. Even two or three Devotion paladins dedicated to championing slightly different causes could have enough friction to generate light and heat.
Yet there's no internal customization on any of the paladin subclasses, so it's got to come from emphasizing commonalities and ethos conflicts, and from fighting styles and how they spend the power that they have. Do they hold back on smiting so they can heal more damage? Do they go for more damage or more conditions with their smites? This could easily be enough to support a Pendragon Great Campaign style of game. The DM needs a skill I've never fully developed: challenging players on specific aspects of their code, whether it's chivalry or something else.
Digression: This is something I've always wanted to do better. I have a hard time looking at an itemized list of virtues or elements of an oath and extrapolating a compelling antagonist and a challenge from it. If that's something you do well, you're cordially invited to talk about it in the comments.
The Noble Knights single-background campaign retains several of the compelling points of the Paladins game, thanks to Traits, Ideals, Bonds, and Flaws. A rules module that let characters of other classes take Oaths for more limited effects is on my personal docket, and it would be a great help for this type of campaign.
For dual-class campaigns:
Add the barbarian for Arthur and his Knights vs. the Saxons
Add the bard so that someone can sing about the deeds of the paladins... and hang onto their lore for them
Add the cleric for a campaign all about the Crusades or the Church
Add the druid for Merlin to go with all of those knights
Add the fighter for a less-holy-than-thou contrast
Add the monk for the other kind of ascetic? I dunno, this doesn't seem ideal to me
Add the ranger for Percival and Pellinore
Add the rogue for someone to be a commoner and do some of the things that knightly honor cannot bend so far as to perform. Fun contrast, at least.
Add the sorcerer... um... for a party mascot? I dunno, this one doesn't speak to me as much unless you use my Royal Sorcery bloodline (which, of course, I am thrilled for you to do).
Add the warlock for a more different Merlin.
Add the wizard for the "most traditional" (ehh...) Merlin.
My go-to campaign model here would have to be Last of the Mohicans. If I had read that book or any of the others in the Leatherstocking series, I'm sure I'd be pointing to the books for this, but... not so much. Anyway, the main problem with a Rangers campaign is not getting the concepts to work - it's that the class has some design problems. The Hunter has enough internal variation (specifically similar to the Totem Warrior barbarian), but needs better base damage output. The Beast Master is simply lacking, unless you pursue fan-created adjustments.
I would definitely recommend an all-Outlander campaign above a Rangers campaign. There are, by the way, some really excellent baseball and/or hockey jokes withering on the vine here, because I can't phrase them just right. Something about how the idea of the Rangers campaign stinks on ice. Humor writing is hard, y'all. Also, I stopped paying any attention to hockey when the Knights left Atlanta. Shut up, I loved the Knights.
The other weird thing about a Rangers campaign is that their Natural Explorer ability is carefully designed to make sure you don't need a party of rangers to do well in the wilderness, just one person with the right terrain choice. There are more terrains than two rangers could cover, but three rangers get into overlap and someone's going to wind up extraneous. Their spells known also represent niches that team members can fill, but it's not that long of a list.
Rogues are second only to fighters in the annals of Easiest Single-Class Campaigns. The only problem is that there are much better systems to use for a Rogues campaign than D&D. Don't get me wrong, D&D does an okay job here... but Leverage and Blades in the Dark spring to mind as purpose-built systems for this. Everyone loves a good heist story, amirite? There are plenty of other campaign models too - spies, pirates, mafioso, other mafioso, assassins with some sort of code of behavior... the only downside is that D&D's rogue class really needs other classes for survivability, unless combat is retuned to suit them.
I'll also seize this chance to mention the two rogue archetypes I've created. The Mastermind and the Investigator might make a Rogues campaign easier to run by broadening the things rogues can do well.
Because of potential mechanical limitations with a Rogues campaign, I'd recommend a single-background Criminals game slightly more, or a dual-background Criminals and Urchins game. In the broader sense of fantasy and historical adventure fiction, all it takes to be a rogue is a certain skeptical attitude toward property rights and, if possible, an avoidance of overt "heroism."
Rogues are also great for any dual-class campaign, as they are non-magical contrasts for magical characters and sneaky contrasts for in-your-face warrior types. Depending on the campaign, you might want to cut (or require) Arcane Trickster.
I have a really hard time seeing a Sorcerers campaign working. In part that's because there are only two Bloodlines in the Player's Handbook, but chiefly it's because the sorcerer spell list encourages a pretty narrow band of gameplay approaches: pretty much blasting and mind control. They lack the vast versatility of the wizard, both in utility effects and day-to-day spell options. I dunno, it seems analogous to a superhero team made up entirely of blasters. Expand the sorcerer's options to include a Brick, and maybe you're getting somewhere. (Shield helps with this an awful lot.) With no ability to heal damage and a d6 Hit Die, a party of sorcerers would be incredibly fragile - even with the help of the Draconic bloodline.
There's also no clear background parallel of the sorcerer - the class's theme comes exclusively from its powers, much like the monk. It's great for a dual-class campaign, or as the single outlier in an otherwise single-class campaign.
Probably the easiest way to make the Sorcerers campaign work on a story level is to bring them together as the distant descendants of a group of (allied or warring) dragons, with each player taking a different draconic ancestor. Those draconic ancestors become central NPCs (present or absent) in the campaign's story.
A coven of warlocks is a story that kind of tells itself. With three Pact options and three Patron options, there are nominally nine different warlocks in the Player's Handbook, but that number stretches to "larger than would ever be needed" once you take invocations into account. As long as you have a good reason for warlocks of different Patrons to work together, this campaign model works great. If everyone has the same Patron, there's some mechanical sameness that may grow tedious, so give serious consideration to alternate Patron abilities.
As with the sorcerer, there's not an obvious background to parallel a Warlocks campaign, though Acolyte, Charlatan, and Sage all touch on parts of it. Once again, the class's theme is derived from its powers, not from a social role that they fulfill. (Outsider, exile, and oath-breaker are the opposite of social roles. If the warlock has a position in society, it is likely a cover identity.)
Warlock is a phenomenal choice for the second class in a campaign, because their magic is so thematically rich. It's a great way to add a "all magic has elements of horror" note to a campaign, especially if the other class you're allowing in is a non-spellcasting class. Barbarians and warlocks sound like an especially cool Dark Fantasy pairing to me.
A Wizards campaign using D&D is totally reasonable, what with eight different schools of magic, a massive spell list, and the like. On the other hand, if you're going to go that far, you should at least read Ars Magica and Mage:the Edition of Your Preference and the many other games written specifically to play cabals of wizards. There's a lot to learn about good campaign models for games all about wizards, and those two systems (historically connected as they are) have a lot to teach.
While I'm here, I might as well mention that you could do a pretty nice parallel of the five Paths of Mage: the Awakening in D&D, matching a class to each Path:
Acanthus -- Bard, or Diviner Wizard
Mastigos -- Warlock (little on the nose, maybe), or Conjurer/Enchanter Wizard
Moros -- Necromancer/Transmuter Wizard
Obrimos -- Cleric...ish, or Evoker/Illusionist Wizard
Thyrsus -- Druid (thanks to D&D's history of no arcane healing and no particular history of animism, there's not a great Wizard school matchup)
Mechanically, a Wizards campaign has serious problems, but they're wizards, so they can probably outmaneuver those problems with proper preparation. No defenders? Use the abjurer's spells, or conjure a bunch of... whatever you've got. No healers? That's more of a problem, but letting one player at a time play a bard rather than a wizard is probably still close enough to protect party theme. Or, since there's no cleric in the party who needs niche protection, introduce cure wounds as a wizard spell. Whatever works.
For a single-background campaign, Sage is really the only choice here, and it's a weak fit. You're probably better off starting everyone with the Magic Initiate feat for free, with or without the Sage background, or requiring the spellcasting subclasses for fighters and rogues. D&D doesn't offer a clean fit here, because it wasn't purpose-built for that.
The twelve classes of 5e fit together quite well, but you can say some really cool things about a setting or individual campaign if you restrict the classes available. Even completely arbitrary boundaries can be interesting if the story explores or invents reasons for them. There are players who will be turned off by this, of course; every class is someone's absolute least favorite. If you wind up with that kind of objection, then either change the campaign plan or plan to go forward without that player.
If you're a player and the DM suggests something like this, you're under no obligation to like it, but you can probably suck it up and give the campaign a try (without sandbagging to make sure you're miserable, because that makes you the douchebag).