This July marks seven years since the release of D&D 5e, though it feels like even longer than that since I’m still running the same campaign I started in D&D Next, with several original PCs still involved. That’s quite a respectable lifespan for a WotC edition of D&D, considering that Sixth Edition is an unsubstantiated rumor among fans, not a near-future reality. (The runaway growth in sales has a lot to do with that. Oddly, WotC doesn’t want to kill migratory waterfowl that lay eggs of uncommon metallic content.)
What I’m here to talk about is classes and subclasses, and where the Player’s Handbook is aging gracefully – or not. We’ve had two major releases of official subclasses since 2014, in the form of Xanathar’s Guide to Everything and Tasha’s Cauldron of Everything. Each setting book also has its own subclasses and other player-facing content; those subclasses are often picked up (sometimes with tweaks) by later Everything books.
Power creep has always been an aspect of ongoing content release, whether we’re talking about any edition of D&D or White Wolf’s Clanbooks, Kithbooks, and so on. On the business side, power creep looks great in the short term and awful in the long term – it keeps fans buying your new books to keep up with the wacky crap their buddies bring to the table, but each of those books sells fewer copies than the book before it and before long you’re writing a new edition.
That’s the short version of why there have only been two major player-facing content releases in seven years – 5e wants to stretch that cycle out as long as possible. It’s working out well for them so far – XGTE and TCOE have been wildly successful sellers and the Player’s Handbook is still a best-seller. Apparently 2020 led to a 35% growth in revenue for D&D. Hard to see how they’ll follow that, since there shouldn’t be another Tasha-like book this year
Anyway. Let’s see how each of the Player’s Handbook classes and subclasses stacks up. Obviously, the artificer won’t be significantly appearing here, except to say that its design is very unlike the other half-caster classes in the game. I’m going to throw the word “playstyle” around a LOT in this post – and no class receives as sharply diverging of playstyles from its subclasses as the artificer does.
Only two barbarian subclasses in the PH, though one of them does us the favor of being a la carte.
I think the Berserker’s Frenzied Rage feature is showing its age, not in damage output necessarily – three greataxe attacks per round at 5th level is still incredible – but as a 3rd-level feature that you only get once per long rest (until greater restoration or other ways to purge exhaustion are on the table, and GR is expensive) and that carries a drawback. I’m mostly sure they just wouldn’t design it like this anymore. Also, Intimidating Presence seems like it should be combat-friendly, but it really isn’t until you have Persistent Rage five levels later. Retaliation is, of course, still phenomenally powerful. So this one’s okay overall, but would get written differently now.
The Totem Warrior, like the Hunter ranger, has you pick one feature from a set of several at three different points in your progression. That 3rd-level Bear totem feature is still one of the most powerful options ever put into the game, and it only gets better as you move into the late-game and more energy damage gets tossed around. If the subclass were getting designed today, I expect the 10th-level feature might still include a commune with nature ritual, but there would probably also be a combat boost of some sort.
I stand by my group’s criticism of the barbarian class as a whole, though – so much of what the class does is just passive benefits stacked on top of standard attacks while you are raging. The class could use another tactical (not necessarily build-based) choice point. The Path of the Beast and the Path of Wild Magic both address this with an additional chosen or random effect when you pop your rage.
Only Lore and Valor here.
The College of Lore holds up pretty well, I think – three extra skills is big, Cutting Words is pretty solid, extra Magical Secrets can’t go wrong… Peerless Skill isn’t bad by any means, but it hasn’t aged well as a 14th-level feature, now that the College of Eloquence is over there doing all that… existing. Overall, this is still a good, very straightforward bard experience. Less change to the default bard gameplay loop than just about any other subclass.
The College of Valor has needed a little bit of a tune-up for ages, and I don’t think it’s a stretch to look at the College of Swords and see an effort at that – but Swords is still offering something different from the core promise of Valor. Valor sets out to be a sword-and-spell pseudo-warlord. That’s a “how could you NOT want to play this?” combo, and honestly, all it needs is a way to give you Battle Magic a lot earlier. If this were getting made now, there is a 0% chance that you’d see both Extra Attack and Battle Magic – good odds of dumping Extra Attack, moving Battle Magic to 6th, and creating something all-new for 14th that actually brings the concept together.
Don’t get me wrong, you can absolutely still have fun as a Valor bard or any of the subclasses that I don’t roundly praise. This isn’t me giving you permission to have fun with something. We’re talking power balance here and that’s all. Valor got in its own way on Day 1, and the cleaner designs of later releases have overshadowed it in power.
Oh dear. This is a lot to cover: Death (yes, we’re doing Death), Knowledge, Life, Light, Nature, Tempest, Trickery, and War. Some of these are still fine!
So the Death Domain. Well, Reaper aged real well, thanks to XGTE’s toll the dead. It’s definitely not on the same page with Divine Strike at 8th level, though. Tasha’s replacement for Divine Strike looks pretty great here, so sure, this has aged just fine.
The Knowledge Domain was always a specialty interest rather than a power move – amazing at social interaction and exploration because you’re proficient in approximately everything and Read Thoughts is great, but near-zero combat application. I love this domain, I have played it before and would play it again, but it’s not even charted on the same graph of power level as other domains. It requires exactly as much self-preparation as it ever did, but it’s good at what it does.
The similarities of function between the Life Domain and TCOE’s Peace domain are part of what pushed me to even write this. I don’t think there’s much of an available interpretation where Peace isn’t overall more powerful than Life, because Emboldening Bond is just that amazing. I feel like Supreme Healing would not do this if it were written today, especially since the big top-end heals that you’re picking up in the late game can’t benefit from this feature. (Also the example here is… odd? d6s are specifically never offered as a healing die?)
The Light Domain is 5e’s original fire cleric and it has aged very well. The biggest difference from 2021 design is that Warding Flare would scale by proficiency bonus rather than Wisdom, probably, and Improved Flare might very well just get folded into Warding Flare right off the bat. Cleric-with-fireball is a concept that should command attention even now.
Conversely, the Nature Domain never worked. It needed a second look in terms of general usability from the day the PH manuscript went to layout. Why would you be hyper-specialized to control animals and plants for short spans of time? What does heavy armor have to do with this playstyle? Just… go play a particularly religious druid or something.
The Tempest Domain is the earliest in a series of lightning-themed subclasses, official and otherwise. I will never understand how this subclass doesn’t gain resistance to lightning or thunder damage at any point. There’s nothing wrong with any of the features, but they would all be written pretty differently today. You’d see an incredibly tight game loop around dealing lightning damage, possibly as a peripheral effect of all other spellcasting (cf. Order Domain).
The Trickery Domain was hard to figure out how to use well back in 2014, and they’ve gotten very good at making it clear how you’re supposed to use a subclass’s tools. Good clear usage for this domain always needed to shove you toward a Wis/Dex build with proficiency in shortswords or rapiers, or with a Sneak Attack-like feature to capitalize on Invoke Duplicity and Cloak of Shadows. Blessing of the Trickster not letting you buff your own Stealth roll is a needless restriction – as Jester Lavorre shows us, Trickery should be a little more selfish as a playstyle than Grave or Life or whatever. I’m willing to believe that it’s possible to put together a very effective Trickery build, but we’re trying not to do hardcore system mastery as a prereq here!
The War Domain is only okay. A lot of what it does, Order now does better while also folding in other stuff. Guided Strike and War God’s Blessing shouldn’t be separate features – that just feels like a waste of a new feature. It’s a shame that Saint of Forge and Fire in the Forge Domain is completely better than Avatar of War. Skip it for Order, or go play a paladin.
As you see, even in the early going, cleric domain had a mix of the great and the completely skippable. Order, Peace, and Twilight are nothing like these in design style.
Again, just two subclasses, though Circle of the Land is hypothetically really broad in concept.
The Circle of the Land looks dated now, though its classic-druid-from-earlier-editions status still gets it some note. Compare the grab-bag of spells and features here to the focused experience of XGTE or TCOE druids. (But of course it looks a bit rough – the druid subclasses of XGTE and TCOE are some of the best material in their respective books.) Natural Recovery is always going to be nice to have, but at the time you get it, we’re talking about one extra 1st-level spell per long rest as your cool playstyle. That… doesn’t fly. The subclass just isn’t here to influence your actions in a meaningful, round-by-round way, since it’s mostly passive defenses.
The Circle of the Moon is one I can’t be all that objective about, then or now, because I just… don’t like the way Wild Shaping works. It’s the separate hit point pool thing. Anyway, any new collection of monsters that comes out and includes beasts below CR 6 might be more power for this subclass – but the subclass has power-scaling issues throughout, because shutting off your spellcasting to grow some claws and ruin your AC isn’t consistently worth it, and becomes sharply less so as you level. Compared to the later druid circles? This is also a pass.
This is one of the few situations where it looks like the subclasses are getting active maintenance going forward, thanks to the new fighting styles (a boon to all fighters and Champions in particular) and maneuvers found in Tasha’s.
Champions are not really for me but… between the new fighting styles and the on-crit features of Crusher/Piercer/Slasher, man, the Champion is in better shape now than it was in 2014, without changing a word of the subclass. Still completely viable.
Battle Masters are obviously the best and the new maneuvers just keep that going. Sorry, my objectivity here has f’ed off to the pub. Battle Masters are great, because spending a die for bonus damage and extra stuff is so strong. Also, you now have another way to turn a feat slot into more Combat Superiority and maneuvers, so that’s pretty great.
Eldritch Knight started off in a weak position, but the new cantrips of the Sword Coast Adventurer’s Guide that have been updated in TCOE are a huge help. I haven’t gone through all of the newer official spells to see what best fits an EK’s needs. It’s probably fine now? Maybe a little on the lower end of power, but not out of the running.
There are three Monastic Traditions in the PH, and one of them is the watchword for subclasses that don’t really… work.
The Way of the Open Hand is still competitive, just because Open Hand Technique is so good. Wholeness of Body wasn’t ever a major reason to be in this subclass, Tranquility even less so. Quivering Palm is great, any day of the week.
The Way of Shadow has always been an odd duck. It sets up one promise but delivers something all-but-unrelated. Would you like to play a ninja? Well, here’s an illusionist and teleporter. That’s the same, yeah? It’s incredibly good at infiltration and exfiltration, but it has nothing going on defensively, Cloak of Shadows is awfully expensive as an action, and its first round-over-round damage increase is Opportunist at 17th. It’s hard to see this as even in the same family of subclasses as Mercy or Astral Self – it would be written completely differently now. Probably a good bit of “when you use Patient Defense” features.
The Way of Four Elements Total Landscaping was bad from the start – it’s just too expensive for what it does. TCOE’s Ki-Fueled Attack optional feature at 3rd level looks like it was pretty specifically designed to help them (and Shadow monks) out. It’s their version of War Magic, basically. Is it enough…? I’m not breaking out the spreadsheet to run the numbers, but I feel fairly confident that XGTE/TCOE monk subclasses outshine it significantly. Also, Sun Soul is pretty explicitly out here presenting the Way of Only One Element – still feeling less decisively powerful than you might hope, but better.
Monk subclasses on the whole are a great demonstration of the shift from “here’s some stuff you can do” to “this is a twist on your class’s gameplay routines” in 5e design.
The three paladin Oaths stand up fairly well, and need to add one word in a couple of places to move fully into 2021 design. Part of that is that all paladin subclass design is in a weird place, as I discuss briefly in the last four paragraphs of my Tasha’s breakdown article. The super-short version is that each subclass is locked into a strict formula, more so than any other class.
The Oath of Devotion is just as strong now as it was then, and that’s a lot. An immunity aura to charmed that stacks with your immunity aura to frightened? Well… yeah, that’s incredible. Passive protection from evil and good is just incredible in the late game. The only flaw here is that CD: Sacred Weapon and Holy Nimbus need the word “bonus” inserted in front of “action.” After the PH came out, the designers more fully embraced the idea that most fights are 3-4 rounds, so spending one of your actions self-buffing is a whole problem. If you look at TCOE oaths, they’ve moved toward bonus actions in all of their CDs and 20th-level transformation. I had misremembered that Invincible Conqueror, the Conquest transformation, is an action rather than bonus action, but Glory and Watchers make my point for me.
The Oath of the Ancients is arguably the strongest of the PH Oaths, and… nothing has changed there, it’s still staggeringly good. Aura of Warding is flat-out amazing, then as now. Nature’s Wrath probably is good enough to justify taking an action rather than bonus action; less so for Elder Champion. Don’t get me wrong, Elder Champion is great, it just needs to be a bonus action to use.
Oh, and you should be able to reset Holy Nimbus and Elder Champion by burning a 5th-level spell slot, as in Glory and Watchers. Still a trivial modification.
The Oath of Vengeance is striking as one of the few paladin Oaths not to get an aura at 7th, but it’s a 5e-ificiation of a 4e avenger power. Looking at hunter’s mark, I think my main takeaway about Vengeance is that they need a default access to Perception and/or Survival, to get stronger use out of the things that differentiate hunter’s mark from divine favor. This one is basically fine, but you probably need a way to get a second Vow of Enmity per short rest.
(More idiosyncratically – I still wish this subclass carried robust support for playing a light- or no-armor zealot, instead of needing to go to the Zealot barbarian for that. I remain a huge fan of the 4e avenger aesthetic.)
The Oathbreaker is overpowered in a way that is unfriendly to what players usually need to do (Aura of Hate). It stands up very well next to the Oath of Conquest, and probably outshines TCOE Oaths. That’s… basically fine, it’s not really meant for protagonist PCs anyway.
In my notes, I wrote over all of this in red ink “how to start a fight on the internet.”
The Hunter ranger is another of those a la carte subclasses that the PH loves but WotC hasn’t done again. I kind of miss it as an approach, but the Hunter is not the world’s greatest example of why. Hunter’s Prey is sort of a mess because the rest of the class tries to get you not to use great weapons, but they fit in with Giant Killer and Horde Breaker so well, and neither of them are all that great with two-weapon fighting. Colossus Slayer is great too, of course; its 1/turn limit is likewise super hostile to two-weapon fighting. Defensive Tactics is okay, could probably use a little bit of a boost to feel less situational. Multiattack needs to give you both options rather than one or the other, since you couldn’t use both with the same weapon style anyway. The Superior Hunter’s Defense options are all good, though making Hunters wait until 15th level to get the features of 5th or 7th-level rogues is another of those “why do you hate rangers” moments (fucking Vanish, man – long live Nature’s Veil).
The Beast Master gets a massive facelift in TCOE, with a total rework of beast companion stats. It’s a lot gentler on your action economy and it demands far less system mastery to understand how to play this thing well. Having been all but redesigned from the ground up, I don’t think I can logically say that this would be designed all that differently if it were made today. Thanks to the action-economy shifts, two-weapon fighting is off the table here, but Bestial Fury is great scaling support for your beast’s damage output alongside your own. You’re making four attacks per round, is what I’m saying. I… think you can make your beast attack a total of three times per round, and make one yourself, or two by it and two from you? Parsing the new rules isn’t awful, but there are parts I might be getting wrong.
Edited to Add: I got the Beast Master attack routine dead wrong. Quoting from an upcoming Tasha’s breakdown for Tribality:
Your attack routine: at 3rd, you attack once, your beast attacks once as a bonus action. At 5th, you attack once, beast attacks once, you take a bonus action, or you attack twice, beast attacks once. At 11th, you attack once, beast attacks twice, you take a bonus action, or you attack twice, beast attacks twice. (Side note, this functionally protects your ability to cast a spell without sacrificing your beast’s turn.)
Three rogue subclasses, no waiting.
The Thief isn’t all that combat-oriented until the incredibly powerful Thief’s Reflexes at 17th level, but it’s very strong on exploration and niche utility features. It’s rewarding for people who love being tricky in combat and aren’t as focused on peak damage output – compare this to the incredible damage boost of Wails from the Grave. Much like the Knowledge Domain, you’re probably best served to think of this subclass’s power on a completely different graph.
Randomly – I didn’t think about it before right now, but it’s interesting that the fighter and rogue subclasses are presented out of alphabetical order. Fighter goes C/B/E, rogue goes T/As/Ar. For these classes and no others, they had a default in mind and placed it first.
The Assassin has always depended on whether or not the DM lets surprise happen. I think we all know that it’s not a reliable producer. Advantage against creatures that haven’t acted yet is more consistently helpful, but from 3rd level to 8th, almost everything this subclass does for you is only on the first round, and only if you’re lucky. Infiltration Expertise and Impostor rely on aspects of the Assassin fiction that are minimally helpful to most D&D campaigns, but if you have a DM and a party who are willing to lean incredibly hard on them, they do things. Death Strike at 17th is the one really good feature… that still only works with surprise. In short, this was bad on Day 1 and it hasn’t gotten better.
The Arcane Trickster is basically fine still? New spells in XGTE and TCOE have sustained it to some degree, though there are rules errata and tweaks that undermine some of those things in ways I don’t completely understand.
Two sorcerous origins, both incredibly bad. (This is another reminder that if you like them, you are not bad for liking them. Bad here is just a power-level judgment, not a judgment on you.)
Draconic Sorcery is fine in concept? It would just be designed differently now. Elemental Affinity shouldn’t cost a sorcery point or have a limited duration (compare it to Psychic Defenses), and some of those damage types are not well-represented across all levels of spells. Draconic Presence is nowhere near good enough for the pride-of-place it receives – by 18th level, a whole lot of what you’re encountering is likely to be immune to charmed or frightened. Bold statement here, but: DMs shouldn’t need to redesign encounters just to make sure players can use their top-end features. Extra hit points and free mage armor are great, but something more active at 1st level would really help.
Wild Magic hasn’t stopped being a great way to accidentally wipe the whole party at 1st level. If this were designed today, the table would be a d8, d10, or d12 table rather than d100, most or all of the effects would be beneficial to the party, and you’d choose when to have the wild magic stuff happen (cf. Wild Magic barbarian). The fact that it’d be nothing like 2e wild magic is… not a drawback.
The three warlock patrons are another great example of how much things have changed in 5e design. The issue is that they don’t have inform the gameplay loop of the warlock really at all. When they created these, they weren’t thinking in those terms. The fact that any gameplay loop needs to support both eldritch blast spam and Pact of the Blade weapon attacks just adds another wrinkle.
The Archfey Patron’s first feature is a big part of the problem – it’s an AoE charm/frighten, which sounds great except that the duration is too short to accomplish much. The features are things you want to be able to do, just… Misty Escape and Dark Delirium (at 6th and 14th) are the only things you’ll want to use in common situations. I took my own run at a variant Archfey patron, but it doesn’t meet the criteria of my critiques now either. In short – if this were written now, the 1st-level feature would be something you use often, and at least one later feature would hook into it directly.
The Fiend Patron starts with an on-kill effect. If you’re not familiar with this conversation: on-kill effects are a problem because you create a situation where your friends wind up kill-stealing from you (and most of them don’t have an on-kill effect of their own), and PCs don’t have a way to pace out a fight to help the characters who do have those features use them more effectively. The rest of the features, again, don’t really interrelate at all. They’re just extra stuff in isolation.
The Great Old One Patron is the real problem kid of the bunch, because Awakened Mind is perilously close to a non-feature. Telepathy is cool and all, but as your only feature before 6th level.. oof. Imposing disadvantage on an incoming attack once per short rest is no one’s idea of a cool-enough 6th-level feature, though the benefit when you cause a miss is a nice extra bit. Thought Shield is great. Create Thrall… in addition to being kinda icky, is also not going to be useful in the midst of Tier 3+ adventures nearly often enough, because it’s humanoid-only. In a 2021 rewrite, I doubt that any of these features make it to release.
The problem, of course, is that the story of these three Patrons is still massively compelling – even more so in Descent into Avernus, Rime of the Frostmaiden, or my forthcoming adventure in Candlekeep Mysteries. Their mechanics have been utterly overshadowed by Hexblade and Celestial, to say nothing of the Genie and the Fathomless, but as a DM, I absolutely want my players to engage with fey, fiend, and aberration plot from a warlock’s perspective.
Much like clerics, this is one of the hefty ones in the PH – the eight traditions each get their own subclass! (In my campaign, the eight Great Traditions formed the Society of the Wise and they spend a lot of time sneering at any other wizard subclasses as renegades.) All of them have a pretty good grasp of playstyle – it’s just a question of how the subclass pays you for casting spells of that school or makes normally-challenging parts of gameplay easy.
Abjuration is a tight and clear playstyle – cast abjuration spells, keep yourself and your pals alive. If I have any issue, it’s only that I’d like to see more abjuration reactions that protected allies. Call it a City of Heroes bubble defender if you want. (I’m also curious about whether an Abjurer with Sentinel and War Caster could work as a tank-mage.)
Conjuration is only okay at 2nd level – Minor Conjuration might be a marvelous solution to a bunch of problems, or it might do nothing. I know which of my players I most worry about taking this subclass, let me tell you. Benign Transposition grants survivability with cheap combat-range teleportation, or support through repositioning allies. The last two features lean into helping your summons stick around. All around, pretty great – it might do a little bit more if it were written now, but I don’t think it’s underwhelming as is.
Divination is, hands down, still one of the strongest wizard subclasses out there. Portent is that good at deciding to succeed, or to make someone else fail. Adding a choice point of “do I use this now?” to every d20 roll that goes by is sufficient to be a playstyle. Expert Divination and Third Eye are also saving you just an enormous number of spell slots for things you’re sure to want.
Enchantment was a disaster area in 2014, and I don’t think anything has changed. The problem is that immunity to the charmed condition is widespread, and every single feature here other than Enchantment Savant hangs on charms. It was really dismal before XGTE gave us charm monster. It’s still a subclass that might crap out on you at random. Mind control and all forms of crowd control are thorny areas for design at the best of times, though.
Evocation is one I don’t hear a lot about people playing, but Sculpt Spells looks great to me still. Potent Cantrip and Empowered Evocation are deliberately the simplest possible features – and for that matter Sculpt Spells is designed to let you not worry about your spell targeting. Of course they put the most straightforward of all wizards in the 5e Basic rules and the SRD – but that doesn’t mean it’s not a strong subclass. I feel like it might have more explosion in its features if it were written now, but I think it’s probably fine.
Illusion wants you to know that it’s not a trick. Tricks are things that 4e PCs do for residuum.
…what was I saying?
As is true of illusions in general and has been true in most editions of D&D, it’s an open question whether they’ll do anything. There’s a lot of room for DMs to make them feel awesome or almost useless in how enemies react to them. Improved Minor Illusion at 2nd and Malleable Illusions at 6th do nothing to correct that. It’s not until 10th that you get features that definitely do anything. Not really ideal.
Necromancy has always been troublesome. Check out this post that I wrote about the issue back in 2017, including some proposed fixes. The short version is that Grim Harvest is an on-kill effect that gives you something that ideal necromancer play means you never need, and Undead Thralls clashes with your allies’ features and general social mores to such a degree that it may be a non-feature too much of the time. Command Undead requires there to be, you know, undead to command, so it’s highly situational.
Transmutation is structured a good bit like Conjuration, with Minor Alchemy being either amazing or useless depending on your problem-solving cunning. Everything here could stand to be useful a bit more often, especially Transmuter’s Stone – it’s very good but engaging with this feature just doesn’t happen enough, especially considering that Shapechanger and Master Transmuter are also very situational uses.
It’s all a very mixed bag at this point. WotC’s design team as clearly kept learning about what works and what doesn’t in 5e’s mechanics and player community. There’s a strong emphasis now on subclasses touching your action-bonus action-reaction cycle, and that just wasn’t a thing early on.
If you’re a freelancer who writes a lot of subclasses, you must read Unearthed Arcana or Tasha’s Cauldron of Everything and understand these trends. I know TCOE isn’t everyone’s favorite – no book is everyone’s favorite – but if your subclass is measurably less appealing than official content, you’re starting with two strikes against you when players decide what they’re going to play. (Strike One is the strong bias of players and DMs toward official content.)
Thanks for coming with me on this 5,000-word journey, and thanks to Steve, Colin, Jon, Jillian, and others who chatted with me about this material.