I don't think I can play D&D 5e again. I definitely can't commit to a 5e group. I strongly dislike this game. That said, the point of this post is not to be an angry rant but an intelligent essay about the modern D&D class design.
Modern Classes Aren't Distinct
Classes overlap so much that niche protection is lost. For example, players describe some classes as martials. These include, at least, the fighter, barbarian, and paladin. Barbarians more hit points, and they can rage and take half damage from everything. Paladins can wear as much hit points and armor as a fighter, but they can also smite for bonus damage at will and cast spells. So, when choosing a class, what is your incentive to choose fighter over barbarian or paladin? Everyone criticizes the fighter for falling short or being boring by comparison.
The rogue is your premier utility class. At first level and again at sixth level, a class feature called expertise allows them to choose two skills to double their proficiency. With the right races, backgrounds, and feats, the rogue can step on the toes of any other class. They can have better survival skills than the ranger, greater religion knowledge than the cleric, superior arcana knowledge than the wizard, etc. Why be a ranger when the rogue can track better than you and when their sneak attack damage can outdo yours?
Casters is a word that describes the classes that are primarily magic users. In classic D&D, magic was far more limited. In modern D&D, casters have access to at-will damaging spells (cantrips) that allow them to do as much damage as a martial character but from range. They scale up with level so that they don't fall behind characters who get multi-attacks. They can do consistent damage at no resource cost and less risk. So, why be a martial?
If any class can do anything or trivialize another class, then classes stop meaning anything. Why have classes?
Restore Niche Protection
This post is a companion post to a class design philosophy that I wrote here. To summarize one of my points, there are three pure classes: martial, magic, and utility (or skill). Classically they are the fighter, mage, and thief. The fighter should be the best at hitting, the mage should be the best at magic, and the thief should be the best at skills.
Other classes should either be variant or sub-classes of these three, or they should be combination classes. In each case, they trade one thing for another thing.
A ranger for example should not be its own class; it should be a variant of another class. Were it a variant of fighter, they trade heavy armor for wilderness survival. Were it a variant of thief, they trade skills suited for urban or dungeon environments for wilderness survival.
In either option, the ranger's niche as the party's expert in wilderness survival is protected, the fighter still gets the best AC, and the rogue's role as the expert in urban and dungeon themed utility skills is protected. These classes remain distinct and they retain something that gives them unique value.
The cleric is the first combination class. A cleric was conceived of something as an in between of a fighter and a mage. The combination class gets some things from both classes, and for balancing, gives up some things from both classes. Clerics get some but not all of the equipment of the fighter, and they get some magic but not as much as a mage. For balancing, clerics still can't hit as well as the fighter, they don't have as many hit points as a fighter, and they don't have as many spell slots as a mage.
In the presence of the cleric, the fighter is still the best at hitting and being hit, and the mage is still the best at magic.
Collapse Redundant Classes
You don't need new classes when existing classes can be customized or re-flavored.
We don't need a paladin if you can be a religious fighter or a martial priest. We don't need a barbarian if you can be a fighter from a different culture. You don't need a monk if you can modify the fighter into an effective brawler / pugilist.
You don't need a druid if you can be a nature cleric.
We don't need a bard if you can pick any class and play it like a musical, charismatic character. We don't need rangers if we can have a nature rogue.
You don't need Wizards, Sorcerers, and Warlocks if you can just make a single mage class and choose the source of your magic as being either academic, a blood-trait, or granted by a pact. You don't need an artificer if you can be a mage with a crafting skill.
There's a game called Five Torches Deep that collapses classes almost exactly like this. There's another game called Olde Swords Reign that offers four classes and lets you modify them with feats to resemble the cut classes. This customizability is anti-bloat.
Remove Skills as an Independent System
Most RPGs use either a class system or a skill system, not both. The pros of a skill system are customizability. The pros of a class system are simplicity. The pros of combining them are the simplicity of a class system with added customization but the cons are an overly complex character creation process that can lead to classes losing their niche.
This ability of the players to mix and match skills to different classes that do not describe the identity or role of their class has contributed to the ranger becoming a confused mess.
Many players are dissatisfied by the 5e ranger. In classic D&D, the ranger was distinguished by tracking. No one else had tracking. In modern D&D, anyone can have tracking and some classes, inappropriately, can be built to do it better. The 5e ranger has lost its niche. To compensate, the designers of 5e made changes to the powers of the ranger for game balance, but the changes distort the identity of the class.
Now the ranger is another class that can do some hitting, some magic, and some skills, and we already have classes who can do that! It's not special.
For simplicity, all classes should have their own built-in skills. You don't need a line item on your character sheet to tell you that the cleric is proficient with religion, that the wizard is proficient with arcana, or that the bard can sing and dance. You should have an understanding that when you pick thief, you can pick locks and sneak.
Stop Balancing Classes Against Each Other
I want you to consider two different jobs: doctors and cashiers. These are very different and distinct. Now I want you to take on the asinine task of trying to balance these two jobs so that they are equally attractive careers.
Balancing fantasy classes so that they are equally attractive or fair to play is similarly inane. When you tweak one, you have to re-evaluate the other eleven or twelve. It's just a load of trouble, and in spite of the trouble there are some classes (or sub-classes) that stand out as having more value or less.
This gets even more cumbersome whenever you add new subclasses to the game.
Blade Singers are wizards that can have incredibly high AC. Why would you be a regular wizard? Gloom Stalkers are rangers that can automatically become invisible in the dark to opponents who rely on darkvision. Why would you choose another ranger subclass?
Eventually, the core sub-classes get overshadowed by a few choice supplemental sub-classes. Eventually, the over-abundance of supplemental sub-classes makes the imbalance even more pronounced and problematic.
Not everyone needs to be able to do consistent damage. Not everyone needs to have a variety of different actions. Not everyone needs to have something to do in any given situation. Classes should be unequal. That's the point. Balance is the enemy of distinctiveness.
Make Class Features Rare, Not Constant
I made seventh level character for a one shot. It had fifteen features between race, background, and class. I want you to consider the word count. If each feature is one sentence of fifteen words (a very conservative estimate), that's 225 words. That's a paragraph on yours and everyone's character sheet.
All that to remind me of the special things I can do. And it wasn't even a spell-caster.
Classic D&D did not have a new feature (sometimes two?) at every level, and it did not front-load your character with a multitude of features at first level.
This is complexity that does not necessarily improve the game. It adds to your cognitive load, the number of things that your mind must keep track of. It makes the game harder for new players to learn. It can slow the games down when someone wants to figure out what they can do and how it works.
Features should be meaningful, not bloated, not constant. A hit die of d10 and a line item that states "can use all weapons, all armor, and shields" defines a fighter. A future line item that adds multi-attack makes them feel more powerful.
A hit die of d12 and a line item that states "rage adds a bonus to attack and damage rolls" defines a barbarian. Then trickle in the other features. Adding an unarmored or lightly armored AC bonus to barbarian after a few levels rather than immediately makes that feature earned, special. It shows growth.
Fewer features would create stronger identity. Some features feel like they were added for the sake of giving the class more stuff.
The bard's jack of all trade feature lets you add half your proficiency bonus, rounded down, to skill rolls in which you aren't proficient. At first level, this means +1 on a d20. It doesn't improve to +2 until nineth level. Is +1 to your least favorite skills worth of a whole feature? The lore college subclass adds their proficiency to three more skills which trivializes this feature further.
The ninth level fighter's indomitable feature comes up rarely, and they make it more complicated than it needs to be by adding a resource management component of giving you a limited number of uses per day.
A fifteenth level monk, levels that players often do not reach, is immune to the effects of aging and old age. How common is that?
A first level ranger's natural explore feature lists six different benefits and is applicable during hex crawls, but the game notoriously under-supports hex crawl style play. They can't get lost which takes out half the challenge of hex crawls, and they aren't slowed by difficult terrain which makes difficult terrain a redundant portion of the GM's game prep and world design.
In Lamentations of the Flame Princess, only the fighter's to-hit bonus improves. Only the thief's skills improve. Only the mage can do magic. Then there's the race-as-class options for Elf, Halfling, and Dwarf which are like combination classes. These are class abilities with clear value and identity.
The excessive quantity of features built-in to your races, class, and background, in addition to the tight math of bounded accuracy, are why magic items are de-emphasized in modern D&D. Your characters are already really powerful, why design a game around characters with who can attune to more than three magic items? I think 5e has lost an opportunity to tether important special powers to magic weapons and equipment which would give magic items considerably more importance to the players and the game.
Use Prestige Classes Properly
A prestige class is not a class you have access to from character creation. They're classes that you earn through play. In essence, it's a class you promote into through play. Think about a soldier without a title being knighted or an acolyte being promoted to a priest. There should be a narrative component.
Prestige classes describe the setting. If anyone can just choose a prestige class, what's so prestigious about them? If any fighter can reach level three and choose to become an eldritch knight, that describes a setting where fighters can easily gain magic. That describes a setting where magic is common and ordinary.
There are no restrictions on the number of paladins, arcane tricksters, or assassins in the world, and the requirements aren't very strict if present at all. I would think assassins should be extremely rare since poison crafting and poison use should be a forbidden, rare, and specialized skill. There are also no rules about paladins and open assassins being in the same party.
All players can choose to be a noble or sage via a background option, and the game does not sit the GM down and explain "it's your setting, you can impose sensible restrictions." Nobility is given, not earned. Prestige is an expectation.
Classes Describe the Setting
Classes options and flavor should describe the setting, but 5e keeps this loose. Consider some of these classes based on 5e rules.
Barbarians are not required to originate from a low-tech culture, foreign culture, or an underdeveloped culture.
Bards' sub-classes describe colleges. Are these physical or figurative colleges?
Clerics are not required to be agents of a god or gods, yet they have access to divine powers. Therefore, divinity is trivial.
Similarly, Paladins are not required to be religious or belong to a knightly order. They're just freelance fighters who wield divine powers.
Personally, I think divine spell-casters should always have a religious component.
Classes shouldn't be casual choices. From the players point of view, classes are just options from a catalog, but from the character's point of view, where did their features come from?
Your class should ground your character in the world. If you're a fighter, this should suggest a history in a military or some strict martial training regimen. Were you a peasant who became a part of a town militia or a kingdom's army? Or are you the son of a noble brought up as a knight-in-training?
Some class choices make backgrounds redundant. If you're a thief subclass of rogue, this should necessarily mean a criminal background. If you're a ranger, this means you spent extensive time on a frontier or borderland, and so the outlander background is implied. For clerics, the acolyte background is implied. 5e permits thieves with an acolyte background and suggests that a person who was simultaneously a criminal and also an acolyte could exist, and that one player could benefit from both. What world allows pocket-picking priests-in-training?
5e does not give guidance on who can be what, and why you would impose race-class restrictions. Dwarves can be wizards, halflings can be barbarians, war-forged and other non-organic creatures that do not reasonably have ki can be monks.
Classically, dwarves were restricted to fighters. This described dwarven culture. Dwarves did not become thieves and they did not become magic-users. If they became clerics, they didn't adventure; that wasn't their role in dwarven society.
Classically, elves were restricted to either fighter or magic-user, and they oddly could be both. This described a culture of people who either became warriors or mages, and sometimes dabbled in both.
And so 5e character creation describes a very ill-defined world where everything and anything goes. Player expectations are difficult to manage. Player experiences vary wildly.
Make Multi-Classing Diegetic
Diegetic means "happens organically within the world or narrative". You can't multiclass because the player wants to; you can multiclass because the character wants to and pursued it in play.
By default, if multi-classing is allowed, the player expectations are that anything goes.
Multiclassing should require a character-driven goal, otherwise it feels arbitrary. Why did the warlock become paladin today? Because the powers synergize really well? Is the warlock's patron OK with this? Is the paladin's deity OK with this? How about any NPC mentors?
Consider the absurdity of this scenario. The other adventurers are taking a rest. The warlock approaches with new a new paladin class. What's the first joke that comes to mind? How about "how come we weren't invited to the ceremony?" or, "why didn't you tell us you found religion!" or simply, "congratulations."
Multi-classing should require time, training, mentorship, or joining an organization. New class features should not be instant.
Reduce The Number of Classes to the Three Basics
Again, I already wrote about this at length in my older post here. To summarize, there are only three pure classes: martial, magic, and skill. All other classes are a variant, a flavor, a re-skin, a sub-class, a combination or hybrid class, a prestige class, or a cultural expression.
What Modern D&D Gets Wrong
The emphasis on the specifically defined classes and even more specifically defined sub-class is not real character customization. There are a dozen classes when fewer would do. Every class gets its own subclasses. These are not options that enrich player choice, they restrict it.
I proposed three to four broadly defined classes with open options to modify them with special powers or abilities from a catalog or by DIY. It's flexible. It's open customizability.
Compare that to modern D&D's current thirteen, more narrowly defined classes. They present closed-ended character growth paths. They force an early decision to choose a sub-class which presents an even more closed-ended path. In essence, 5e offers you only two customization options: one choice for your class, and one choice for your sub-class.
I don't even like all the subclasses. Many of them have their own flavor, but I want to flavor my character myself.
Sub-classes are an illusion of choice and bloat!
There's also way, way too much magic! Almost every class has access to magic by default. Too many spells can resolve too many problems too easily. This makes preparing and running D&D more challenging, more effortful, and more time-consuming, especially for a new GM, because you just can't anticipate what the players will do with their 100+ spells. Typically, the players use magic to trivialize everything you prepare that would be interesting or challenging.
Sometimes less is more.
What OSR / NSR Gets Right
I have referenced a couple other games that allow for alternative ways or simpler ways to make characters. What many of them have in common is that they are simple, they work, and they're still fun.
If you want to play Dungeons and Dragons B/X for example, you can make a simple character that's easy to play in only a few minutes. You can teach the game to new players quickly and easily!
Simpler games are also more adaptable to other settings and genres.
A lot of the OSR games are compatible, so adventures from one OSR game work for another! That means you have more options for campaigns and adventures.
Fewer rules, simpler rules mean less rules talk. Rules talk is a detriment to immersion and pacing. 3e D&D has been described as thirty minutes of fun packed into three hours.
Numbers are smaller, and so combats don't take an hour or two to resolve.
The rules-lite nature of OSR / NSR games makes them easier for a GM to prep and run, and so these games do not have a GM shortage. It's easier to train new GMs.
OSR / NSR games are more affordable. Many are free!
Why this Matters
D&D is the entry point into the TTRPG hobby. It has brand recognition and many prospective gamers want to play the shiniest, newest game. What they don't know is that it's also more difficult to learn, to play and to run than they think. People, especially GMs, are easily overwhelmed and burnt out. Groups fall apart under the weight of it all.
The funny thing is, I bet if 5e had to openly compete with its earlier editions, it wouldn't look so shiny.
5e has a problem with overpowered player characters. This is because there are so many player options that the game designers can't anticipate all the combinations of powers and how they might create an exploit that players can use to excel; Sometimes in cheap and unfulfilling ways.
If you're a player and your character is consistently being outclassed by an overpowered player character, this can affect your ability to have fun. You may have less to contribute. You may feel like your character doesn't belong or isn't needed. You may lose confidence in your character. You may ask why this other amazing character would have you along for their adventure.
Another problem with overpowered player characters is that if you want a game with some challenge, that overpowered character might spoil challenges consistently. Don't worry, I bet the next obstacle will be interesting.
Overpowered player characters also make prep harder for a GM, and they spoil the fun of running a game for the GM.
Niche protection combats overpowered characters by creating clear characters roles for players to understand. Simpler games limit the number of possible exploits for broken characters and make it easier to correct.
Reduced player options help to manage player expectations. They help you fit the characters into the world, and they help your players understand the kind of game they're playing.
The complexity of the game leads people to believe that one has to be a professional game designer to design anything, and this diminishes people's willingness or interest in the DIY aspect of the hobby. It's a paper-pencil game. The users should be able to modify it.
Conclusion
Modern class design can be fixed by remembering why classes exist: to make distinct characters. Everything else is noise. Classes need a niche. Classes shouldn't overlap. They should describe the world. They shouldn't be balanced into sameness. They shouldn't be bloated. They don't need an independent skill sub-system.
5e D&D doesn't have a class system, it has a catalog.
