Reflections on RPG Design: Dungeons & Dragons

An opinionated exercise in Heartbreaker design

(Plus an equally opinionated examination of what different versions of D&D got right and wrong)

A few years ago, I set out on purpose to make a D&D Heartbreaker.

I had some very specific goals:

• Build a setting wherein recognizable versions of standard D&D tropes, corny or not, were justified and added richness to the setting, rather than simply being inherited spandrels of Gygax and Arneson’s quirky sense of humor.
  

• Build a system woven from different ideas from different editions of D&D, including roads-not-taken, based on how well they come together into a coherent, easy-to-use, and fun system for new players and grognards.
  

• Provide space for a culture of OSR-like play that provides friendly on-boarding for people more familiar with 3rd edition, 4th edition, 5th edition, or Pathfinder.
  

It’s still a work in progress. But, a ton of playtesting has left me with some pretty strong opinions about D&D systems, at least, and I want to spell them out here, both to justify the choices that fed into my heartbreaker, and to “plant a flag” - you don’t have to agree with anything I’m saying here, but if you do, you might like what I’m pitching.

So, let’s start with an important first question:

What exactly is Dungeons and Dragons?

By which I mean: at what point does a game stop being D&D, and start being its own thing? What counts as a “heartbreaker”? Is Paizo’s Pathfinder a heartbreaker? Is Kevin Siembieda’s Palladium a heartbreaker? Is Call of Cthulhu a heartbreaker? Is GURPS a heartbreaker? What about Worlds Without Number? What about Mörk Börg?

We can even take this in the other direction:

Is B/X a heartbreaker? Is AD&D a heartbreaker?

Is 4E a heartbreaker? Is it even really “Dungeons and Dragons”?

Older Millenials might remember that this was a legitimate question when 4th edition first came out, and this question is what led to the Pathfinder RPG existing at all.

The “Spirit of the Game”

Obviously, everyone is going to have their own opinion here. The success of this blog is going to depend, partially, on whether the opinions I present to you resonate with your own. So I can be bold, but I still have to be careful.

Nevertheless, it’s time to plant that flag:

I think that a game counts as a ‘Dungeons and Dragons’ variant, if it has the following features:

• six Abilities (Strength, Constitution, Dexterity, Intelligence, Wisdom, and Charisma), which typically range from 3 to 18 (centered on 10-11), often determined by rolling 3d6 — with some purists insisting on “3d6, in order, no modifiers.”

• classes and levels as the primary character advancement mechanic

• a d20 as a primary resolution mechanic in combat

• multiple additional dice of different sizes (d4, d6, d8, d10, d12), which get used in various ways

• hit points and damage dice as a primary “combat resilience” mechanic

• spell-like powers which can be prepared on a daily schedule by at least some subset of classes.

I assert that a D&D variant can break at most two of these rules and still “count” as Dungeons and Dragons. An RPG that breaks one, and bends one more, is almost certainly a D&D clone; an RPG that breaks three or more definitely feels like it’s doing its own thing.

So, a natural question to ask next is, which version of Dungeons and Dragons is the most Dungeon and Dragonny?

If your answer is “B/X”, then we just became best friends. If your answer is “AD&D second edition”, then I salute you as a man of honor, and will play at your table as a respectful guest of a foreign king.

If your answer is “D&D third edition” or later, then you still have my respect, but I would prefer you keep it on your side of the table.

Gygax vs. Arneson

One thing that became clear to me, as I examined different variants of D&D, is that Gygax and Arneson actually had different styles as game designers, and different senses of what they were even aiming for. And while Gygax is to be commended for carrying the torch as far as he did, I kinda like Arneson’s vibe better.

Gygax, especially towards the end, was desperately trying to establish a career for himself as the D&D guy. While I can respect that as a hustle, it’s fundamentally incompatible with making the best possible product.

Wizards of the Coast inherited this trend, and doubled down on it to produce D&D’s third edition. D&D3E was deliberately trying to create a mini game, or “metagame”, of build optimization during character creation. (This is a matter of public record.) I assert that obsessive build optimization has since metastasized into everything wrong about D&D.

That isn’t to say that build optimization is bad! Just that it’s grown to be most of what D&D is about these days, and that’s kinda sad.

Building a Heartbreaker

So, when I decided that I wanted to make my own heartbreaker, I started with B/X.

More specifically, I started with the B/X retro-clone Lamentations of the Flame Princess, which takes B/X, refines the Thief skills in an elegant way, and then proceeds to do enough drugs to kill Hunter S Thompson before crafting its setting and vibe.

As I continued, I played through several sessions of each edition of D&D — first (B/X), second (AD&D), third (actually 3.5), fourth, and fifth, and after each session asked myself what did this edition do better than any of the others?

These were my findings:

What B/X Got Right

B/X has a lot going for it. The travel → explore → encounter play loop is as tight as you could ever hope for. The classes are small, simple, and easy to understand. Each class has a schtick, a wheelhouse, that is sufficiently relevant to some part of this loop that the class has a reason to be in the party. The Ability modifiers are small enough that an 18(+3) is noticeable but not utterly overwhelming, and a 3(-3) is a significant handicap without utterly ruining a character. Anything more complicated than “I hit them with my axe” falls under “rulings, not rules”.

All in all, B/X is close to a perfect system for an experienced DM and players. The only reason to bolt anything more onto it, is if you’re a new DM that appreciates some guardrails and handholds. The only clunky things about B/X are its Armor Class system and its Thief skills system, both of which can be patched (admirably so by LotFP and various other retroclones). Currently, LotFP has become something like abandonware, but Old School Essentials seems to have admirably picked up the slack. OSE, at the moment, is the go-to system for “I just want to run a simple D&D game.”

What AD&D2E Got Right

AD&D goes in entirely the opposite direction from B/X. Everything has a subsystem. They’re all quirky and unique. It’s clear that if there isn’t a subsystem for something, your table can just create one and start using it. AD&D2E is essentially a houserule incubator - a hothouse for inspiring DMs and players to create new subsystems.

Consolidating these subsystems into a coherent and well-playtested whole is a Sisyphean challenge, which so far (to my reckoning) only ACKS comes close to getting right. If you want to run multiple long-running campaigns with multiple DMs in a shared world, where the player’s actions in one campaign can have coherent and measurable impact on the other campaigns, ACKS would be the way to do it.

What 3E Got Right

3E was the first real commercial attempt to get AD&D’s subsystem proliferation under control, and produce an AD&D variant that people could reasonably run at cons and at home without needing fifty different supplements and a 300 page binder of notes.

It succeeded as well as it possibly could have, given that it was saddled with a second purpose of make Wizards of the Coast’s purchase of TSR look economically justified. If you recall that WotC’s previous experience was with the Magic: the Gathering card game, a lot of why 3E wound up the way it did starts to make sense.

Now, there is an adequately good game hidden inside D&D3E, begging to come out. Certainly not a great game, but an adequate one. Paizo’s Pathfinder does probably as good of a job as you could hope for in finding it and presenting it to players and DMs in a coherent fashion, and their setting is top-notch.

What 4E Got Right

Okay, full disclosure: I loved 4th edition. I loved everything about what they were trying to do. I thought that the “gamification” aspects were exactly what 3E needed, and everyone’s complaints about them essentially boiled down to “this is too on-the-nose”. 4E is what 3E committed itself to being. 4E did it right. And in typical Accelerationist fashion, “give the people what they want, good and hard” caused people to wake up and realize that maybe they didn’t want to want what they wanted.

In other words, 4E started the Old-School Renaissance.

That said, 4E — as a system — has a lot going for it!

• Skills became much more streamlined, with multiple 3E skills combined into a single 4E skill. 4E gets the balance between “too many skills” and “Jesus Christ, just use abilities already” pretty close to ‘right’.

• At-Will / Per-Encounter / Daily powers is fucking brilliant. Categorizing powers this way, and standardizing across class boundaries how many of each type a character should have at each level, was exactly right. Everyone who complained about how “all the classes are the same now” really didn’t understand the game they were already playing — the game that 3E had lured them into. “Linear fighter / quadratic wizard” was finally patched, and everyone hated it. Philistines.

• Giving each class fewer powers meant that players didn’t have to keep track of a ridiculous number of different prepared spell slots or daily class feature uses. “One encounter power per 4 levels” + “One daily power per 4 levels” + “one at-will power per 2 levels” is actually just about right.

• Class roles were amazing for explaining to new players how to choose a class. WotC realized that online games were eating their lunch, and went ahead and adapted. Everyone snarking at them about this was seethe and cope from an audience that had failed to keep its own market relevant. Again, OSR revitalized that market, but it couldn’t have done so if 4E hadn’t planted a giant neon sign on top of the problem.

That said! 4E was by far the biggest break from “classic D&D” to date. The full commitment to grid-based combat, the class powers system, the roles - 4E very, very clearly wanted to be a tactical wargame.

It still fits most of the criteria for being a D&D version. The only bullet point that it misses is “spell-like powers prepared on a daily schedule”, but it even does that — it just bends the idea into a very different shape than D&D veterans are used to.

So, yeah. 4E is a D&D (or more accurately, an AD&D). Deal with it.

What 5E Got Right

5E builds on some of 4E’s simplifications in an incredibly novel way.

• bounded accuracy and advantage are, together, a brilliant solution to the “how does a first-level adventurer keep up with a 14th-level party” problem.

• six different Ability-based saving throws feels like the correct endpoint of D&D’s evolution from OD&D’s five confusing saves, down to 3E and 4E’s three over-generalized saves, and then out the other side. Essentially, “ability checks” and “saving throws” have been merged into a single mechanic, which slots in neatly with the rest of the system’s unified (d20+Ability modifier+Proficiency modifier) resolution process.

• At-Will / Short Rest / Long Rest is as good of an obfuscation of 4E’s At-Will/Encounter/Daily system as you’re going to get. The fact that you need to obfuscate it is a sad commentary on players’ unwillingness to face their own shortcomings as a market demographic, but here we are.

• “Proficiency dice”, a house rule from 5E’s Unearthed Arcana, is a great way to keep failure relevant even at higher levels, and I find it sad that it didn’t make it into the official rules as the default way of handling proficiency.

Of course, 5E utterly fails to break away from the 2E/3E/4E progression of the “combat optimization mini-game”. In fact, 5E seems to be even more about big set-piece encounters than 4E. For someone who looks forward to doing stuff in the ‘travel’ and ‘explore’ parts of the travel →explore →encounter loop, this is kinda disappointing.

Okay, it’s really disappointing.

And yes, a good DM can use 5E to emulate old-school playstyles. But I’d far prefer more native support for it.

Frustrations With Class Levels

Pretty much no edition of D&D feels like it gets class levels right. 20 class levels is simply too many. 30 class levels is ridiculous. Even B/X’s 14 class levels feels wrong — the deliberate discontinuity between level 9 and 10, when hit points stop increasing by hit die and instead get a smaller flat bonus, is right around where the leveling system “breaks down”.

The “E6” houserule/hack for 3E is a pretty great homebrew solution to this problem, but it goes just a hair too far. Six class levels simply isn’t enough. You need more “breathing room”. 4E does a “tier” thing, with levels 11-20 and 21-30 each in a different “phase of the game” which acknowledge the superhumanity of the player characters, which helps a bit but also locks a lot of the more fun class powers behind a wall that you have to ‘earn’ — which also cheapens the first ten levels by making them just a speed bump to get to the “good stuff”.

But when you look at B/X, E6, and 4E’s “Heroic tier”, you start to see a human-level game that deserves fleshing out in its own right. And that game seems, at its most intuitive, to want to fill out 10 class levels.

Putting it All Together

So, now we have a bunch of strongly held opinions, which we are ready to distill into a coherent set of houserule (or a “heartbreaker” if we decide to try to publish it). So, where do we want to start?

Well, we have six bullet points that define “how D&D a game is”. So we should start with a framework that is small, simple, and already adequately satisfies each of those six bullet points.

3D6, In Order

Committing to an OD&D way-of-life is one of those “constrained optimization” choices that promises lots of creative expression down the line, so let’s go all-in: we’ll build our little custom hot-rod on a B/X chassis and see what happens.

Abilities

We’ve already decided that one of the core features of D&D is that it uses six canonical Ability scores, converts those scores into Modifiers through some formula or lookup table, and then applies them to a die roll — and that at least some of those dice rolls (especially the ones in combat) should use a d20 as their primary source of randomization.

I started building my Heartbreaker as a B/X mod, because B/X “gets most of the basics right”. And “3d6, in order” is metal as fuck.

So, we’ve got six Ability scores:

• Strength

• Constitution

• Dexterity

• Intelligence

• Wisdom

• Charisma

Each Ability score is determined by rolling 3d6, creating a number from 3 to 18. Then each of those Ability scores is compared to a lookup table, to determine the corresponding modifier for that Ability. 3E/4E/5E’s lookup modifier is the easiest to remember, since it’s just “divide the Ability score by 2 (round down), then subtract 5”. But this gets a bit swingy at the top and bottom - we’d rather cap our Modifiers to a smaller range than 5E’s +5 top-end, which can overwhelm smaller die sizes.

Most modern OSR clones have a nice modifier progression system, taken from B/X, that looks like this:

• Score of 3 = Modifier of -3

• Score of 4 or 5 = Modifier of -2

• Score of 6, 7, or 8 = Modifier of -1

• Score of 9, 10, 11, or 12 = Modifier of +0

• Score of 13, 14, or 15 = Modifier of +1

• Score of 16 or 17 = Modifier of +2

• Score of 18 = Modifier of +3

This is almost as easy to remember as 3E/4E/5E’s — it has a nice rhythm to it. There’s four score values in the middle that each get no modifier at all. Then three on either side of that, which get a -1/+1. Then another two that get -2/+2, then a single extreme score on either end that gets -3/+3.

We can work with this.

Now, in various re-implementations of B/X, your ability scores limit which classes you can take in different ways. We can limit this as simply as we’d like, so let’s go with the following rule:

• Each class will have a Prime Requisite Ability. A player character must have at least a +1 modifier in that Ability score, to begin training in that class.

Since it’s possible to roll shit for Abilities with “3d6, in order”, most variants of B/X let you “cash in” Ability scores you don’t need, in exchange for the one you want. But we’d rather have things a bit more random, to encourage players to try out different classes and step out of their comfort zone more often. So, how’s this sound:

• If you roll up a character, and none of your Ability scores has a +1 or better modifier, then start over by rolling a single d6 to pick an Ability at random. That Ability scores starts at 13 (+1), and all your other Ability scores start at 10 (+0).

So, that takes care of Abilities!

Proficiencies

Having a unified set of task resolution mechanics is probably as good a place as any to begin designing the details of a D&D clone. Well, the good news is, we have five different editions of D&D that do this, each in their own unique ways. So we can start by just picking one, importing as many of its mechanics in as we want to use, and then build from there.

The first choice I made, after importing the “d20+Ability modifier+proficiency” attack throw/saving throw standard that gets carried across all editions of D&D, was using the d6 Thief skill system from LotFP, Dolmenwood, and various other retroclones. As I played with it, it became clear that this was a perfect place to use the 5E proficiency die alternate rule from UA.

To recap, “proficiency die” means that, instead of having a static proficiency bonus that starts at, say, +1 and keeps going up each level without bound, your proficiency bonus is always a die - starting at +d4 when you first gain the proficiency - and increasing your proficiency means increasing the die size.

Since the proficiency die is never less than 66% or more than 200% of a d6’s range, it seems simple and natural to just replace B/X’s d6 with 5E’s proficiency die, as the primary “skill test” mechanic.

So, a “skill test” is always an Ability modifier + a Proficiency die, vs. a chance of failure. Chance of Failure is set by the referee, and starts at 2+, going up by 2 each time. A ‘1’ on the die is always a failure, regardless of Ability modifier or chance of failure.

It’s clear that Proficiency dice are a workable system for progressively increasing bonuses, so we may as well use them anywhere. This settles us on three distinct sets of Proficiency:

• Skill proficiencies, which use a d6+ability modifier

• Defense proficiencies, which use a d20+ability modifier

• Weapon proficiencies, which use a d20+ability modifier

The next question is, how many different Skills, Defenses, and Weapon proficiencies do we want?

I went back-and-forth on “different proficiencies for different weapon groups”, so at one time there was a Blades proficiency, an Axes proficiency, a Bludgeons proficiency, and so on. Eventually this condensed down to two (Melee and Ranged), at which point it became clear that I could simply do a single ‘Attack’ proficiency and call it good.

Which means that distinguishing between which weapons your Attack proficiency did or didn’t apply to, needed to be called something different than proficiency. I settled on the term familiarity, and realized that this could work for things other than weapons or weapon groups - languages, for example, are familiarities. Knowledges are familiarities. The ability to use a particular kind of tool is a familiarity.

So, if you have familiarity with something, this allows you to apply one of your proficiencies to any use of the thing you are familiar with, without penalty. Without familiarity, you can only do basic actions or wield basic tools.

This immediately helped me narrow down the proficiency list. One thing that struck me as I did so, was how various saving throws seemed pretty closely related to various skills. This led to the consolidation of the following 5E skills and saving throws into six defensive proficiencies:

• Athletics → Athletics (Strength saving throw)

• Survival → Fortitude (Constitution saving throw)

• Acrobatics → Reflex (Dexterity saving throw)

• Investigation → Deduction (Intelligence saving throw)

• Insight + Perception → Perception (Wisdom saving throw)

• Intimidation → Willpower (Charisma saving throw)

One thing that 4E did well, is having “passive saving throw values”. So each defense proficiency, in addition to having a Proficiency die, will also have a Defense Value (DV) — a static number that attacks which target this Defense can use as a target, just like Armor Class. This number will be equal to 10 + the ability modifier + half the proficiency die maximum — so someone with a Perception of +d6 and a Wisdom of +2 would have a 10+3+2 = 15 Perception DV. (5E would call this “passive Perception”. We’re just extending it so there’s also a passive Fortitude, passive Willpower, passive Deduction, passive Reflex, and passive Athletics. This will come in useful.)

As saving throws and skills consolidated, I eventually arrived at ten “basic skills”, which don’t fall under the umbrella of any of the defensive proficiencies, and which could be refined further by Familiarities into any level of detail that a player or DM would want:

• Finesse (Dexterity)

• Stealth (Dexterity)

• Arcana (Intelligence)

• Crafting (Intelligence)

• Lore (Intelligence)

• Animal Ken (Wisdom)

• Medicine (Wisdom)

• Nature (Wisdom)

• Charm (Charisma)

• Wit (Charisma)

So ten skill proficiencies, plus six defensive proficiencies, plus one attack proficiency. Simple, and short enough to all fit easily on a character sheet with room for a portrait and various other notes.

And we’ve decided that we want to fit all our class powers into 10 class levels, so the only thing left is to build our character classes.

Advantage/Disadvantage and Other Bonuses

If B/X gets the core game right, 5E gets situational bonuses right. 5E’s advantage/disadvantage mechanic is super fast and easy, and their “class powers usually just add a die to a d20 roll” mechanic is equally great.

In fact, we can combine this with the previous proficiency mechanic in a brilliant and fast way - class powers almost always add a Proficiency die from one proficiency as a bonus to another another proficiency’s dice roll. With this rubric for designing class powers, we should probably begin building some actual classes.

Warriors, Experts, and Magic-Users

Going all the way back to B/X’s roots, the earliest role split was between Fighters and Magic-Users, with the Thief showing up early in the game’s development as a third “skilled expert” option that wasn’t either. Then someone wanted to play a vampire-hunter, and the Priest/Cleric was created as a variant of the Magic-User that could do defensive and healing magic instead of primarily offensive or utility magic.

So, as a nod to the nostalgia of early-game weirdness, let’s define our three roles as the Warrior, the Expert, and the Magic User, and build our class system frame around this split.

Warriors

Warriors all need a set of class powers that let them hold their own in a fight. The simplest and most straightforward thing to give them, right at level one, is a big boost to their Attack Proficiency. So, let’s start Warriors off at +d6 Attack bonus.

Next up, Warriors should probably have a good boost to their physical Defenses. So we want them to have at least a +d4 in each of Athletics, Reflex, and Fortitude.

We also want them to be able to do Feats of Strength, but the Athletics proficiency already represents “Feats of Strength” type stuff, so it seems like we don’t really need to do anything else here. And we want them to have good combat mobility, but again — Athletics and Reflex already seem to represent “combat mobility” adequately, so there isn’t anything else left to do here, either.

The only other things we want a good Warrior to do, is use all sorts of weapons and armor (which we already decided would be handled by a binary Familiarity system), and be able to pull off cool combat tricks/maneuvers. So really, building a Warrior class just means figuring out what kinds of cool combat tricks/maneuvers we want, how we want them to work mechanically, and at what levels we want a particular Warrior class to earn them.

About this time, we should also start thinking about “class requirements”. We’ve already decided that we’re using Ability modifiers as prerequisites for each class. So what should that look like, in practice?

Well, for warriors, who rely most on their physical Abilities, we seem to have a pretty great opportunity to build one Warrior class per physical Ability Prime Requisite — so one Dex warrior, one Con warrior, and one Str warrior.

Looking across all editions for inspiration, let’s try building something like a Monk as our Dex warrior, something like a Paladin as our Con warrior, and something like a Barbarian as our Str warrior. But we want each of them to work as pure warriors (no magic or quasi-magical abilities); we can use some kind of dual-classing system if we want a character to “cross the boundary” between Warrior, Expert, or Magic-User.

As we build our not-a-Monk, not-a-Paladin, and not-a-Barbarian, certain patterns start to emerge, which will define the “skeletal background” of the Warrior role - shared class powers and shared level milestones, which we can expect any Warrior class to express. We wound up settling on the following “generic Warrior progression”:

• Level 1: “Weapon Mastery”, a simple feat that lets Warriors add their Attack proficiency to their weapon damage, and get an additional bonus effect from their weapon whenever their Attack proficiency die rolls a ‘6’ or better. Just this one power establishes Warriors as the “hit stuff hard with my weapon” guys.

• Level 2: “Superior Athlete”: A feat that simply bumps the Warrior’s Athletics, Reflex, and Fortitude proficiencies each by 1 die size.

• Level 4: “Warrior Training”: a feat that increases the Warrior’s Attack proficiency from +d6 to +d8, increases their Prime Requisite ability score by +1 (to a maximum of 18), and lets them choose another ability score to raise by +1.

• Level 5: “Extra Attack”: A feat that lets Warriors make two full weapon attacks per turn, or continue through (“cleave”) after slaughtering a low-hd monster and apply the leftover damage as a new attack on another low-hd monster, for as many times as there continues to be leftover damage.

• Level 7: “Improved Critical”: A feat that increases the Warrior’s Attack proficiency from +d8 to +d10, and lets them automatically get a Critical Hit with an attack roll if they hit and their Attack proficiency die rolls a ‘10’ or higher.

• Level 8: “Ability Training”: A feat that increases the Warrior’s Prime Requisite score by +1, and lets them choose another ability score to raise by +1.

• Level 9: “Extra Attack”: Now the Warrior can make 3 attacks per turn, and can “cleave” low-hd targets with each of those attacks.

• Level 10: “Warrior Training”: increases the Warrior’s Attack proficiency from +d10 to +d12, increases their Prime Requisite ability score by +1, and lets them choose another ability score to raise by +1.

This leaves us two gaps — one at level 3, and one at level 6. So, if we also get a unique class feat at level 1, this gives us three levels to gain unique class feats: level 1 (to establish this class as unique from other Warriors), level 3, and level 6.

At this point, we decide to borrow from 4th Edition, and establish a standard for all classes going forward: each class will get an “encounter power” pool at level 3, which gives them a set of feats that they can perform by spending points from that pool (much like 5E’s Monk, with their “ki pool” or “discipline pool”). Having a single pool, rather than having each encounter power refresh separately, gives the player a lot more tactical flexibility in how they expend their per-fight resources. Also, it feels more natural — rather than a bunch of arbitrary powers that you can use once and then magically stop working, you have something like a “stamina bar” that you can burn through in a fight, and once it runs out you’re too tired to do all your Cool Tricks anymore.

So, in 4E terms, each Warrior class will get some “at-will” combat powers at level 1, and then some “per-encounter” combat powers at level 3, and some more advanced “per-encounter” combat powers at level 6.

So, riffing off the Con/Str/Dex split:

• the Dex-based not-a-Monk should get mobility, speed, and dodge powers

• The Con-based not-a-Paladin should get defensive powers that can also protect allies

• The Str-based not-a-Barbarian should get a simple “I hit things harder” button to mash

At this point, we should probably name these guys. Let’s call the Dex guy the “Martial Artist”, the Con guy the “Knight”, and the Str guy the “Champion”.

Okay! So, we’ve got three classes — one dodges good, one blocks and tanks, and one hits hard. But how do we want them to do it?

Well, we have this whole other mechanic laying around - Skill and Defense Proficiencies - so why not start using them?

Step one: we need to establish a rate of skill progression for Warriors. So, let’s start by giving each Warrior class the Athletics proficiency at +d4 at level 1, one other physical defense proficiency at level 1 (based on their Prime Requisite), and then a third defense proficiency based on their “schtick”.

So:

• Martial Artists get Athletics, Reflex, and Perception.

• Knights get Athletics, Fortitude, and Willpower.

• Champions go all-in with Athletics, Reflex, and Fortitude.

And then we remember that these proficiencies are dice, so we can use them as extra bonuses for various class feats. A Martial Artist’s unarmed strikes can do damage equal to their Athletics proficiency die. A Knight can roll their Fortitude to add it to an adjacent ally’s AC or saving throw, by physically body-blocking the attack. A Champion can add their Athletics die to the damage roll of each melee weapon attack they make.

Now we’re starting to see the whole game take shape.

So, Warriors should have some customization as they level up. And we can keep using the proficiency system to do this! At level 1, each Warrior gets their starting proficiencies; then at level 2, they get a boost to Athletics, Reflex, and Fortitude. Past that, at level 3, 4, 5, 6, 7, 8, 9, and 10, we can give each Warrior a choice - increase the die size of any one Skill or Saving Throw, as long as it’s below a certain threshold.

We want this threshold to be lower than the one we’re eventually going to give to the Expert, because skills should be the Expert’s whole thing. But we also want the threshold to be high enough that the Warrior doesn’t feel completely hobbled. And, their physical saving throws need to be at least as good as the Expert’s.

So, a new rule emerges: a Warrior can pick one skill or saving throw at level 3, and at each level after 3, and either gain it at +d4 if they didn’t already have it, or increase it by one die size if it’s below a threshold based on their Warrior level: +d4 at levels 1-4, +d6 at levels 5-9, and +d8 at level 10. But physical defenses (Athletics, Reflex, and Fortitude) need to break these limits, and are already breaking these limits at level 2 (because of the Superior Athlete feat), so we just add a subclause to that “Superior Athlete” feat: A warrior’s Athletics, Reflex and Fortitude proficiencies are limited by their Attack proficiency, rather than their level cap.

Well, that was easy!

Now we just need to figure out the grab-bag of powers that Knights, Champions, and Martial-Artists each get at levels 1, 3, and 6. We want to hew to the following heuristics:

• Champions should hit hard and be really simple.

• Knights should be defensive and tanky.

• Martial-Artists should be mobile and dodgy.

Let’s start with the Champion. I’m gonna come out and say something maybe counterintuitive, considering everything else I’ve been saying: if B/X gets core mechanics and gameplay right, 4E and 5E get class powers right. And of the two, 5E is more current, better-organized, and less “stuck up its own ass” in terms of how to present those powers. (Don’t get me wrong, the 4E format for class powers was brilliant — it’s just a lot harder to use that format for anything that isn’t fully committed to 4E-as-a-bit).

So. Level one power, to go along with Weapon Mastery, is going to rip straight from 5E: Reckless Attack. A Champion can choose to grant advantage to anyone who would attack him until the start of his next turn; in exchange, he also gains advantage attacking anyone he wants until the start of his next turn.

His level 3 power can also be pulled straight from 5E: Rage. A Champion gets a pool of Rage (starting at 1 point at level 3, increasing to 2 points at level 4, 3 points at level 7, and 4 points at level 10) which they can spend whenever they use their Reckless Attack feat. If they do, they add their Athletics proficiency die as a damage bonus to each attack they make until the start of their next turn. This is the “I win harder” button I was talking about earlier.

To make Rage useful outside of combat, we can also throw in another bonus - whenever a Champion Rages, they add their Attack die to any Athletics, Reflex, or Fortitude saving throw that they make during that turn; and they can do so in an encounter or during exploration.

The level 6 power should be an amplification of the level 3 power, so let’s pull in another useful mechanic from 4E/5E - the Champion can now concentrate whenever they successfully hit someone with an attack while Raging. If they do, they can maintain the effects of their Rage through to the next round, without having to spend another point of Rage.

So, that’s the Champion, all wrapped up in a bow: Strength prime requisite, Reckless Attack as their level 1 “fighting style”, Rage as their level 3 “encounter power”, and Relentless Rage as their level 6 “encounter power enhancement”.

Now, onto the Knight.

For our level 1 “fighting style”, we’ll go with Guardian - essentially making it a generic form of 5E’s “Protector fighting style”. Since we already made the Champion as simple as possible, we can afford to give the Guardian a few different “maneuvers” at level 1, to make them something like 5E’s “battle master” fighter. We’ll stick to defensive maneuvers, though, to keep them all in-theme for a tanky Warrior:

• Guarding Strike will let the Knight make an opportunity attack vs. anyone in reach who attacks an adjacent ally.

• Interrupting Strike will let the Knight instantly end an enemy’s turn by hitting them with an opportunity attack.

• Shield Block will let the Knight add their Attack proficiency to an adjacent ally’s Armor Class.

Actually, at this point, we should probably pause and start thinking about action economy.

Say what you like about 4E and 5E (and believe you me, I will), but they do a pretty good job with their combat simulator, especially with the action economy. “Bonus Actions” feel a little more fiddly than we want to track with a B/X chassis, but action and reaction both feel like good, meaty mechanics to hang a whole combat system on.

So, combat happens in six-second turns; everybody makes a d20 Initiative throw at the start of combat to determine turn order, and then when your turn comes up, you get an action to perform, and a reaction that you can save until the start of your next turn to do something with. Attacking is an action, opportunity attacks are a reaction. Using a skill is usually an action, rolling a saving throw is usually either a reaction or “free”, depending on how much effort the saving throw costs.

Since we’re bringing up Initiative, one thing Pathfinder 4E does that I really like is make Initiative into a Perception roll. So, an Initiative roll is a Perception saving throw.

Okay, this is dead simple, and still gives us a lot of options. Let’s build some for the Knight to use.

• Guarding Strike is obviously a Reaction, since it’s just an enhancement of Opportunity Attacks.

• Interrupting Strike, same.

• Shield Block sure feels like a reaction, so let’s make it a reaction too.

So now we have three different ways for the Knight to use their reaction at-will, at level 1. Each is defensive, and each helps their allies. This is starting to look good!

At level 3, the Knight needs an “encounter pool”. The Champion’s pool was called “Rage”; the knight’s should be something more knightly. How about “Resolve”?

So, Knights have a Resolve pool. Just like they have more at-will options than Champion, we can afford to give them a few per-encounter options at level 3. Cribbing from the 4E and 5E fighter powers, let’s go with:

• Mark Foe, to let them “pull aggro” as the kids say these days.

• Power Strike, so they don’t feel completely outclassed by the Champion’s damage output

• Second Wind, because this is a Con-based fighter.

Now we just have to figure out the exact mechanics we want for each of them.

Champions get Rage, which lets them add their Athletics to the damage of every attack they make for the whole round, and Reckless Attack, which lets them add advantage to the accuracy of every attack they make for the whole round. Power Strike is going to enhance a single attack for the same resource cost, so it can afford to be at least that powerful. Let’s go with:

• Power Strike — a Knight can spend one Resolve to retroactively add their Athletics proficiency die to the accuracy and damage of a single attack.

This feels right tactically; it costs an encounter power point, so it shouldn’t also cost the Knight their reaction. It spends a whole power point to augment a single attack, so it’s justified in having a better accuracy bump than Reckless Attack. And making it retroactive means that the Knight only needs to spend it if they need to. We can justify the whole mechanic in-game as “taking extra effort to exploit an opening”, so it doesn’t feel like wooish fate-manipulation bullshit.

So, speaking of avoiding wooish fate-manipulation bullshit, how should marks work? Let’s try this:

• Mark Foe - a Knight can spend one Resolve as they attack an opponent to mark them. While the Knight maintains concentration and attacks no other targets, they have advantage on all opportunity attacks vs. that creature, and they have disadvantage to any attack that doesn’t include a target that marked them while they are within reach of your weapon. As their action, the target can attempt to make a Perception saving throw vs. your Willpower DV to break the mark.

This has some potential! First of all, the mechanics kind of suggest what’s actually going on - you attack someone, intimidate the hell out of them, and get them to decide that you are the active threat. And you continue to focus on them, so that if they decide that you aren’t the active threat, you can rapidly correct their misunderstanding. The fact that this requires you to be within opportunity attack range makes the whole thing a very flavorful, in-character mechanic, without stacking on too many fiddly riders or conditions. I think we’re done, here!

So all that’s left is Second Wind. We want to use our proficiencies, and we’ve already used Willpower for Mark Foe, and Athletics for Power Strike. That leaves Fortitude, which seems pretty well-suited for Second Wind! So let’s try:

• Second Wind — whenever you make a Fortitude saving throw, you may spend one Resolve to regain hit points equal to your Fortitude proficiency die result + your Constitution modifier. You may use this feat when another effect causes you to make a Fortitude saving throw, or your action to spend one Resolve and make a Fortitude saving throw deliberately to regain hit points.

This is a nice, simple decision point — any time you make a Fortitude saving throw, you can use the roll result to convert one Resolve into some free self-healing. If the enemy doesn’t make you roll Fortitude, then you can just use your action to do it yourself.

So, that’s a good stack of three level-3 encounter powers. Now we need a good level-6 enhancement to at least one of them.

Of the three, Second Wind seems the most Con-related, and seems the most tanky, so let’s go with that one. We can riff from the 5E SRD again, for inspiration:

• Indomitable — whenever you use your second wind, you may add your attack die to the saving throw result and the number of hit points regained. You may also spend one Resolve to add your attack die to any other saving throw that you make, or to any of your passive DVs when another creature makes an attack or saving throw against them.

There we are! Simple, on-theme, and incredibly tanky.

Now, the Martial-Artist.

The Champion was the simple class, and the Knight was the defensively tactical class, so the Martial Artist should be the offensive/maneuver tactical class. Also, we can afford to make this one the “high complexity” member of the set, since we’re basically replicating the Monk.

Our level 1 “fighting style” feat, Martial Arts, should at least be on-par with 3E and 5E’s Monk stuff, while using our “proficiencies for everything” rubric. So how about:

• Martial Arts — a Monk gains Familiarity with all exotic weapons, and their unarmed strikes become an exotic melee weapon attack that uses their combined Strength+Dexterity modifiers for accuracy, and their Strength+Athletics proficiency as the damage die.

So far, so good; we don't even need to give multiple at-will powers, here. This is pretty good just on its own. (This also means we can save more “complexity budget” for the level 3 encounter power stuff!)

So, for level 3, we need a better word than ‘ki’ to reflect their equivalent to Resolve or Rage. How about “Focus”?

• Focus — at level 3, a Martial Artist gains a Focus pool, which they can use to perform the following maneuvers:

• Flurry of Blows — spend one Focus after you hit with a melee or unarmed attack to immediately make another attack.

• Nerve Strike — spend one Focus when you hit with a melee or unarmed attack to stun or slow the target.

• Uncanny Dodge — spend one Focus as you make a Reflex saving throw against an area effect, to move a number of paces equal to half your Reflex proficiency die result. If this moves you out of the area, it doesn’t affect you.

Since we just came up with “stunning”, let’s figure out how this mechanic works. Here’s a simple idea: when a “stunning” attack hits, it subtracts its damage both from the target’s hit points, and from their initiative. If this reduces a target’s Initiative score to zero, they lose their next turn, and must make a new Perception saving throw at Initiative zero (with the result becoming their new Initiative score for the next round).

Okay; now we just need the level 6 maneuvers! Since we didn’t give the Martial Artist any level 1 maneuvers, we can afford to stuff their level 6 class feat with goodies, and even make one or two of them at-will. Let’s try this:

• Deflect Attack — you can use your reaction to dodge or parry any attack, retroactively adding your Reflex or Attack proficiency die to your AC.

Wait, hold up. This is a great mechanic in general. I think I want it to just become a basic thing that anyone can do! So, now we have a new set of general encounter reactions that anyone can perform:

• Dodge — If you are proficient in Reflex, you may use your reaction after an attack throw is made that targets your AC, but before it hits, to retroactively add your Reflex proficiency die to your AC vs. this attack.

• Parry — If you are proficient in weapon Attacks and wielding a weapon or shield that can parry, you may use your reaction after an attack throw is made that targets your AC, but before it hits, to retroactively add your Attack proficiency die to your AC vs. this attack.

Okay, back to the Martial-Artist! Now that we’ve defined Dodge and Parry as generic reactions, a new level-3 option for Focus comes to mind:

• Focused Defense — spend one Focus if you do not have your reaction, to immediately regain your reaction.

And now we can build Deflect Attack for real, informed by the previous combat mechanics:

• Deflect Attack — you may use your reaction after a melee or ranged attack throw is made that targets your AC, but before it hits, to retroactively add your Reflex and Attack proficiency dice together to your AC vs. this attack. If the attack still hits, subtract your Reflex and Attack proficiency die results from the attack’s damage. You do not need to spend Focus on this maneuver. If you successfully deflect an arrow, bolt, thrown weapon, or other physical projectile, or reduce its damage to zero, you may catch the projectile in your hand.

• Deflect Energy — you may use your reaction and spend one Focus after an elemental or magic attack throw is made that targets any of your DVs or forces any saving throw, to retroactively add your Attack proficiency die to that DV or saving throw result.

• Redirect Attack — Whenever you successfully deflect a melee or ranged attack, you may use your Focus to redirect the attack to another adjacent creature if it was a melee attack, or to any creature or space within 5 paces of you if it was a ranged attack. Roll a d20 and add it to your Dexterity modifier and attack and Reflex die results from your Deflect Attack roll to determine the attack throw vs. the new target.

• Leaping Dodge — whenever you deflect an area attack that requires a saving throw, you may choose to use your Reflex proficiency instead of any other defensive proficiency. If you do, you may use the higher of your Reflex or Attack proficiency die to determine the number of paces your Uncanny Dodge feat can move you.

• Danger Sense — Whenever you perform the focused defense maneuver, you also become intuitively aware of each attack or hazard that targets or threatens you until the beginning of your next turn. You may attempt to dodge any attack or area eect that you would otherwise be unaware of or unable to dodge. If you do, you use your Reflex instead of any other saving throw proficiency, and ignore any disadvantages.

There we go! A complex, powerful, tactical, but resource-heavy Martial Artist class. That closes out our Warriors.

Experts

The base Expert “class skeleton” doesn’t have to follow the same pattern as the Warrior class skeleton, but it’d help if it at least rhymed. So, let’s start with:

• A level 1 feat to establish “this is what Experts will be about”

• A level 2 defensive or mobility feat

• A level 3 encounter power feat

• A level 4 “bump my prime requisite and another ability score” feat

• A level 5 “simple numeric increase to my base schtick” feat

• A level 6 enhancement to the encounter power feat

• A level 7 “critical hit enhancer” equivalent

• A level 8 “bump my prime requisite and another ability score” feat

• A level 9 “simple numeric increase to my base schtick” feat

• A level 10 “bump my prime requisite and another ability score” feat

So, for level 1, let’s be on-the-nose and call the base Expert feat “Expertise”.

Expertise should give us something like:

• Four starting skills instead of three

• At least one starting skills at d6, instead of d4

• a d4 attack bonus - less than the Warrior, but still useful

• Faster skill progression

Put it all together, and we have:

• Expertise — an Expert picks one of their starting skills or defenses, and increases that proficiency’s die size from d4 to d6. Whenever they gain a level, they choose two skills or defenses to increase, rather than one. Their proficiency cap is d6 at levels 1-3, d8 at levels 4-6, d10 at levels 7-9, and d12 at level 10.

• Reliable Talent — at level 3, an Expert gains a Reliable Talent pool, which they can spend whenever they roll a skill test to replace their proficiency die with its highest possible value, or whenever they roll a saving throw to replace the d20 part of that roll with the corresponding Ability score.

• Expert Training — at level 4, an Expert’s proficiency cap raises from d6 to d8, their Prime Requisite ability score increases by +1 (to a maximum of 18), and they choose another ability score to raise by +1.

• Superior Expertise — at level 7, an Expert’s proficiency cap raises from d8 to d10. Whenever they roll a ‘10’ or higher on a Proficiency die, that roll becomes a critical success if the total roll succeeds. If the proficiency die contributed to an attack’s accuracy or damage, the attack is a critical hit.

• Ability Training — at level 8, their Prime Requisite ability score increases by +1 (to a maximum of 18), and they choose another ability score to raise by +1.

• Expert Training — at level 4, an Expert’s proficiency cap raises from d10 to d12, their Prime Requisite ability score increases by +1 (to a maximum of 18), and they choose another ability score to raise by +1.

So, the Expert and Warrior aren’t exactly the same, but they rhyme. This is looking good so far.

Like the Warrior, each Expert class needs a level 1 feat that will define that particular class’s “schtick”. Unlike the warrior, the Expert’s level 3 feat is locked in by the “skeleton” — all Experts get Reliable Talent. But! the level 2, 5, 6, and 9 feats are wide open.

We’ll try to mirror the Warrior’s “Extra Attack” feat with the Expert’s level 5 and 9 feats, but we’ll do that with their individual level 1 class feat, rather than with the Expertise feat. And at level 6, we’ll want to provide a “bag of tricks” that enhances Reliable Talent in ways appropriate to that particular class. And level 2 can be anything — an enhancement to their level 1 feat, or some new thing — as long as it’s roughly on-par with Superior Athlete.

Let’s start with the old B/X standby, the Thief. Thieves need to be sneaky, treacherous, and nimble. They’re obviously a Dexterity prime requisite class. So, they’ll need the Reflex defense, the Finesse skill, and the Stealth skill. Now the only question is, do we want to give them Athletics or Perception to round them off? Let’s go with Perception to start with; they can choose to train in Athletics as they level up, but a Thief needs to be able to notice traps and get the jump on enemies. For their first-level class feat, let’s go with the old standby:

• Sneak Attack — a Thief can attack an enemy that is unaware of their presence with any ranged or light melee weapon, using their Stealth proficiency instead of their Attack proficiency to make the attack. If they hit, they add their Stealth proficiency die result, plus one Finesse proficiency die, to the damage. At level 5, this increases to two Finesse proficiency dice; at level 9, this increases to three Finesse proficiency dice. For each proficiency die that rolls a ‘6’ or higher in this attack, the thief may choose one of the following options:

  • Wound — the attack inflicts one wound.

  • Stun — the attack inflicts stunning damage.

  • Hobble — the victim is hobbled, reducing their speed to one pace and preventing them from dashing until they make a Fortitude saving throw to shake it off.

  • Disarm — the victim drops their weapon, or one item on their person. If the thief is within reach of the victim, the thief may grab this item instead of having the victim drop it.

So, there’s our first, fifth, and ninth-level feats taken care of! Now we just need 2nd and 6th level feats.

For second level, let’s take a riff from 5th edition:

• Evasion — A Thief can hide and still perform another action in combat. Additionally, whenever they use their reaction to successfully dodge an attack or successfully make a Reflex saving throw, they immediately regain their reaction.

Now, we need some sixth-level “Thief’s tricks”, that they can spend Reliable talent on in ways that other Experts can’t. We already have Danger Sense and Uncanny Dodge from the Martial-Artist, both of which seem pretty appropriately Thief-flavored as well (this makes sense, since they’re both Dex classes). So in addition to those, let’s do:

• Danger Sense - while you are conscious and unencumbered, you may spend your Reliable Talent to attempt to dodge an attack or area effect that you would otherwise be unaware of or unable to dodge. If you do, you use your Reflex instead of any other saving throw proficiency, and ignore any disadvantages.

• Uncanny Dodge - If you successfully dodge an area attack, you can also move or jump a number of paces equal to half your Reflex die result before the attack resolves. If this moves you out of the area or fully behind cover, the attack doesn’t aect you at all.

• Hide in Shadows - whenever you use your Reliable Talent to perform the Hide action, creatures have disadvantage on Perception tests to spot you for as long as you maintain concentration and move less than half your speed each round. If any creature would already have disadvantage to notice you, they automatically fail. You cannot use this class feature if you are encumbered, wearing medium or heavy armor, or clothing that makes you exceptionally visible or audible.

• Trap Mastery - whenever you use your Reliable Talent to successfully detect a trap or ambush, you may also apply it to your first attempt to avoid or disarm that trap or ambush without further cost.

Okay! That’s a wrap for the Thief.

The obvious next Expert to build is the Bard. Bards are obviously the Charisma expert, so giving them Charm and Wit is a no-brainer. They also need Lore and Finesse, which rounds out their four starting skills. They don’t start with any saving throw proficiencies, so they might want to start upping them when they level up.

Starting the Bard off right seems obvious:

• Inspiration — You may use your reaction after an ally makes a proficiency test, attack throw, or saving throw to oer some quick words of encouragement, retroactively adding your Charm die + Charisma modifier to their result. The ally must be within 5 paces and must be able to see or hear you.

Their second level feat also seems obvious:

• Vicious Mockery — At bard level 2, you may use your reaction just before a creature makes an attack throw, saving throw, or proficiency test. Roll your Wit die + your Charisma modifier, then subtract half the result from their roll. The target must be within 5 paces and must be able to see or hear you.

And then we need to enhance Inspiration at level 5:

• Inspiring Presence — At bard level 5, you may use your Reliable Talent as your action to grant advantage to each attack, saving throw, or proficiency test made by any ally within 5 paces that can see or hear you for as long as you maintain concentration. You must continue to use your action for each turn that you wish to use this eect. As long as you incorporate dancing, singing, or playing a musical instrument as part of this feat, you may use your Finesse proficiency rather than your Fortitude whenever you are forced to make a concentration check to maintain its eects.

And then, we need a pack of “Bard’s Tricks” at level 6:

• Inspire Competence — whenever you use your Inspiration or Inspiring Presence on an ally’s roll, you may use your Reliable Talent on that ally’s proficiency die as if it were one of your own proficiency dice.

• Jack of All Trades — whenever you attempt a skill proficiency test that you do not have proficiency in, you may use your Reliable Talent to roll your Lore die and add the die result to that skill’s natural ability modifier as if it was your proficiency die for that skill.

• Silvered Tongue — whenever you use your Reliable Talent on a proficiency test as part of a social contest, your target has disadvantage on their opposing roll.

• Taunt — whenever you hit a creature that can see or hear you with any attack, or when a creature misses you with a melee attack, you may use your Reliable Talent to cause that creature to suer disadvantage to all attacks that don’t target you until the start of your next turn, and any attack that they do make provokes an opportunity attack from any of your allies within reach.

• Enthralling Performance — when you use your Reliable Talent on a proficiency test as part of a song, dance, oration, or other performance, each creature with an Intelligence modifier of -3 or greater that witnesses your performance must make a Willpower contest vs. your proficiency test result or be beguiled by you. For as long as you maintain concentration on your performance, these creatures have disadvantage to all rolls, and their attitude towards you is treated as Friendly if it would normally be worse. You may choose to exclude specific creatures from this effect.

Finally, we need a level 9 enhancement to Inspiration:

• Font of Inspiration — At bard level 9, whenever an ally succeeds on a roll aided by your Inspiration, you regain your reaction.

And now we’ve got a Bard!

Next up, the Ranger!

So, time to do a little reflection. The Thief is our Dexterity expert, and the Bard is our Charisma expert. Experts are skill-monkeys. Strength and Constitution don’t have any skills attached. Should Rangers be Wisdom-based?

They’d get Animal Ken and Nature as skills, and Perception as a defense, so this is off to a good start. Swapping out Medicine for Stealth seems too on-brand for a Ranger to turn down. So now, we just need their class powers.

What if we just reskinned all the Thief class powers to be more… Ranger-y? Let’s try it:

• Track Quarry — At level 1, Ranger may use their action to choose any creature that they can see as their quarry; they may also make a Perception or Animal Ken test to attempt to track any quarry whose footprints or other evidence they can identify. All rolls made to track, attack, or defend against a chosen quarry have advantage for as long as they maintain concentration on the task.
  

  Slay Quarry - When you hit a quarry with an attack, you add a quarry die to your damage roll, which is a proficiency die from your Perception skill. You add a second quarry die to the roll at ranger level 5, and a third at ranger level 9.
  When you hit a quarry with an attack, if any quarry die rolls a ‘6’ or higher you may choose one of the following options:

  • Wound - each quarry die that rolls ‘6’ or higher deals a wound in addition to its hit point damage, if the target tracks wounds.

  • Stun - the target reduces its current initiative by the total damage dealt.

  • Hobble - the target is slowed, preventing them from dashing and reducing their speed to one pace per round until they make a Fortitude saving throw to shake it off.

Okay, this is looking pretty good! Now for their level 2 power, how do we reskin Evasion to be Ranger-flavored?

Wait, instead of just reskinning Evasion, how about:

• Favored Environment - At ranger level 2, you gain a number of special Expertise Familiarity based on your Nature proficiency, and gain more as your Nature proficiency increases (two favored environments at +d4, three at +d6, four at +d8, five at +d10, and six at +d12). Each favored environment is a type of natural terrain, such as forest, tundra, desert, ocean, plains, swampland, and so on. You may alternately choose a single urban location such as a hometown and its surroundings, or a large city. All skill rolls and saving throws related to your favored environments are made with advantage. You may climb, jump, and swim freely while in a favored environment without using your action, ignoring difficult terrain as you move.

Okay, this is looking good so far. Now we need the level 6 “Ranger’s tricks” and we’re done! Let’s try:

• Camouflage - as an exploration action, you may use your Reliable Talent to construct camouflage for yourself or one other creature while in a favored environment. Whenever a camouflaged creature performs the Hide action while within that environment, they add your Nature die to their Stealth roll, and other creatures have disadvantage on Perception tests to spot them for as long as the camouflaged creature maintains concentration and moves less than half their speed each round. If any creature would already have disadvantage to notice a camouflaged creature, they automatically fail.

• Sense Surroundings - you may use your Reliable Talent to guard, evade, or hide while in any favored environment without using your action. You may only choose one of these options per round.

• Primal Rituals - while you are in a favored environment, you may use your Reliable Talent as an exploration action to duplicate the eects of the first-circle Druid rituals Brew, Call Beasts, Cleansing Waters, Detect Poison and Disease, or Purify Food and Drink, or use two Reliable Talent to duplicate the effects of the second-circle Druid rituals Brew, Traceless Passage, Find Animal or Plant, or Beast Bond.

• Primal Senses - while you are in a favored environment, you may use your Reliable Talent to become aware of an attack or situation that would otherwise surprise you, or to add your Nature proficiency die to any Deduction or Perception saving throw. Similarly, while you are tracking a quarry, you may use your Reliable Talent to become aware of an attack or surprise from your quarry, or add your Animal Ken proficiency die to any Deduction or Perception saving throw against them.

Okay, obviously we’re going to need to flesh out some kind of Druid soon, but that was on the to-do list anyways.

So! we now have three Experts, to match our three Warriors. But there’s a fourth Ability score full of Skills — Intelligence. What should an Intelligence-based expert look like?

(Side-eyes Eberron)

Alchemist it is!

Alchemists get Arcana and Crafting, obviously. They should probably also get Medicine if they’re going to be brewing up poisons and elixirs, and Finesse if they’re going to be building little steampunk contraptions. That’s four skills, so let’s get started on class feats!

• Alchemy — You know how to quickly brew volatile versions of common potions, elixirs and poisons, even in field conditions. You have a notebook with six alchemical formulae and magic item blueprints - three common, two uncommon, and one rare; whenever you come across a notebook of another artisan that contains new formulae or blueprints, you may copy them into your notebook as your action during an exploration round. You may spend one exploration round brewing a single common formula from your notebook, even in field conditions. Any formula that you brew with this ability is volatile, losing its potency by the next sunrise or sunset. While brewing, you are automatically surprised if you are interrupted by a random encounter.
  
  Volatile brews require supervision to maintain their potency, even for this long; you can only maintain a total number of volatile brews depending on your Arcana proficiency (one at +d4 or +d6, two at +d8 or +d10, or three at +d12). You may hand volatile brews to companions, so long as you can inspect and maintain each of them for a few seconds once every few exploration rounds.

• Invention - At alchemist level 2, you may begin working on experimental inventions which duplicate common magic items but require daily maintenance. You can maintain a total number of inventions depending on your Crafting proficiency (one at +d4 or +d6, two at +d8 or +d10, or three at +d12). You must devote at least one hour to their maintenance each sunrise and sunset, or they stop working; you can also reconfigure one or more of them to operate as dierent magic items during this time, if you have appropriate blueprints.

• Uncommon Formulae - At alchemist level 5, one of your maintained inventions and one of your maintained volatile brews may be each Uncommon. If you have all the necessary tools and ingredients, you can now modify one of your inventions into a permanent magic item much more quickly than one could normally construct such a device. A Common item takes one day of uninterrupted labor, while an Uncommon item takes one week.

Artisan’s Tricks - At alchemist level 6, you learn new ways to use your expertise. You may use your Reliable Talent to perform the following tricks:

• Recharge Magic Item - you may use your Reliable Talent on your action to attempt to recharge any invention or magic item by infusing it with the appropriate materia, rolling your Finesse proficiency vs. a CoF or DV based on the item’s Rarity – Easy for a Common item, Average for an Uncommon item, Hard for a Rare item, and Extreme for a Legendary item. You gain advantage on this roll if you have the item’s blueprints.

• Reliable Activation - you may now use your Reliable Talent on any bonus die from any magic item that you have blueprints for, as if it was one of your skill proficiency dice, and may use your Reliable Talent on your attack proficiency when you attack with a potion or magic weapon.

• Reliable Brew - you may now use your Reliable Talent on any die roll for any brew that you imbibe, administer, or otherwise activate as if it was one of your skill proficiency dice. You may also use this trick on the Fortitude saving throw of anyone you administer multiple elixirs to, to help mitigate undesirable side-effects.

• Rare Formulae - At alchemist level 9, one of your maintained volatile brews may be a Rare formula; if you do not maintain a Rare formula, then each of your maintained volatile brews may be Uncommon. Likewise, one of your maintained inventions may be a Rare invention; if you do not maintain a Rare invention, then each of your maintained inventions may be Uncommon. Finally, you may now build rarer magic items much more quickly. If you have all the necessary tools and ingredients, a permanent Rare magic item now takes one month of uninterrupted labor to complete, and a permanent Legendary magic item takes one year.

Welp! That looks like all our Experts.

Magic-Users

Getting magic-users right is tricky. First, we need to decide how we’re going to divvy up our Magic-User classes. We have three Warriors, one for each physical ability. Why not three Magic-Users, one for each mental ability?

• Wizards should obviously be the Intelligence class, using Arcana to cast spells.

• Druids need Wisdom, using Nature to cast spells and Animal Ken to transform into wild shapes.

• That leaves Clerics as Charisma casters, using… Willpower? Willpower seems like a much better choice than Charm or Wit.

Magic-Users need to not be skill-monkeys or combat monsters. They start with their magic skill, and maybe one other skill or saving throw. And at every odd level, their magic skill should go up by one die size — leaving them with only one skill or defense proficiency to bump at every even level. We don’t even really need a proficiency cap, at this point, since there isn’t much they can do to surpass one.

At level 1, they get a Spellcasting feat, which lets them perform magic. At level 5, this bumps up to … I hate “spell levels” and “class levels” as different things, so let’s call it second-circle magic. At level 9, they get third-circle magic. So now they just need a cool at-will thing at level 2, a cool per-encounter thing at level 3, a per-encounter enhancement at level 6, and another enhancement of some kind at level 7.

We can also break with the mold here, and replace the level 10 ‘training’ feat with another cool power, sort of a capstone of magical ability.

Alright, let’s start with our Wizard! Wizards need Arcana, obviously. Giving them Deduction seems like an interesting choice, since it’s an Int-based defense skill and we haven’t given it to anyone else as a starting proficiency yet. Now they need some class powers, following the “skeleton” we laid out above:

• Arcane Spellcasting — You have a spellbook that contains three first-circle Arcane spells, two second-circle Arcane spells, and one third-circle Arcane spell. You may choose which spells you have written in your spellbook when you create your character, subject to approval from the referee. Whenever you come across a spell scroll or spellbook of another wizard that contains spells that you do not know, you may copy those spells into your spellbook as your action during an exploration round.
  Each sunrise or sunset, when magical energies flow most easily across the Veil, you can prepare arcane spells from your spellbook that you believe will be useful throughout the day. You can prepare a number of first-circle Arcane spells at this time based on your Arcana proficiency: two at +d4, three at +d6, four at +d8, five at +d10, and six at +d12. Each spell is prepared onto an arcane symbol taken from your spellbook, which is painted somewhere on your body with magic ink. Casting the spell causes the ink to ignite into a ghostly flame, then harmlessly burn o your body. Prepared spells retain their potency until they are used, or until the next sunrise or sunset.

• Cantrips — at wizard level 2, you learn to pull minor acts of magic from your prepared spells without expending them. Each arcane spell that you prepare has a cantrip eect, which you may perform at-will so long as the main spell remains prepared.

• School of Magic — At wizard level 3, your Arcana proficiency increases from d4 to d6. You now begin to focus your studies on one school of magic. Choose one: Aetheric Energies, Elemental Energies, Psychic Energies, Abjuration, Conjuration, Divination, or Transmutation, You may choose one such spell from each circle, pending approval from the referee, and add them to your spellbook. When you prepare your daily spells from your spellbook, one of them can be a second-circle spell from this school. Whenever you have fewer spells prepared than your maximum, you can use an exploration action to prepare a first-circle spell from your chosen school. If you do, you are automatically surprised if you are interrupted by a random encounter.

• Ability Training - At wizard levels 4 and 8, increase your Intelligence score and one other ability score by +1 each (to a maximum of 18).

• Second Circle Spellcasting - At wizard level 5, your Arcana proficiency increases from d4 to d6. When you prepare your daily spells from your spellbook, one of them can now be a second-circle arcane spell from any school.

• Arcane Recovery - at wizard level 6, whenever you have fewer spells prepared than your maximum, you may use an exploration action to prepare one first-circle spell from any school. If you do, you are automatically surprised if you are interrupted by a random encounter. If you currently have fewer than two second-circle spells prepared, you may instead prepare one second-circle spell that matches your chosen school of magic. Again, if you do, you are automatically surprised if you are interrupted by a random encounter.

• Arcane Power - At wizard level 7, your Arcana proficiency increases from d8 to d10. Whenever your Arcana die rolls a ‘10’ or greater while casting a spell, the eect is a critical success; whenever a spell damage die rolls a ‘10’, the spell inflicts a critical hit. When you prepare your daily spells from your spellbook, one of them can be a third-circle arcane spell from your chosen school. Additionally, you can now perform third-circle rituals from your domain, and you also perform lesser rituals from your chosen school more quickly: a second-circle ritual from your chosen school takes two exploration rounds (twenty minutes) to perform, while a first-circle ritual from your chosen school may be performed as a single exploration action.

• Third Circle Spellcasting — At wizard level 9, your Arcana proficiency increases from d10 to d12. When you prepare your daily spells from your spellbook, one of them may be a third-circle arcane spell from any school.

• School Mastery — at wizard level 10, when you cast a prepared first-circle spell from your chosen school, it stays prepared and may be cast again on subsequent turns.

That all looks good! Now, let’s see how much of that we can just duplicate for the Cleric, and how much we want to do differently for flavor. Clerics obviously need Willpower to cast their spells, and one other proficiency. Insight/Sense Motive would be ideal, and since that got absorbed into Perception, we can give that a try.

Now, for class powers:

• Divine Miracles — You have a book of prayers and miracles that contains three first-circle prayers, two second-circle prayers, and one third-circle prayer. You may choose which prayers you have written in your prayer book when you create your character, subject to approval from the referee. Whenever you come across a prayer scroll or prayer book of another cleric that contains prayers that you do not know, you may copy those prayers into your prayer book as your action during an exploration round.
  Each sunrise or sunset, when magical energies flow most easily across the Veil, you can prepare divine miracles from your prayer book that you believe will be useful throughout the day. You can prepare a number of first-circle miracles each day based on your Willpower proficiency: two at +d4, three at +d6, four at +d8, five at +d10, and six at +d12. Each prayer is prepared onto a prayer strip, which is a small piece of paper that axes to the target of your miracle, then burns with a ghostly light until it is consumed. Prepared miracles retain their potency until they are used, or until the next sunrise or sunset.

• Sense the Unholy — At cleric level 2, you can use your action to touch a creature or object and make a Perception test. On a success, you sense whether there is any infernal or undead energies present within the creature or object. On a critical success, you intuitively understand what kind of magic is present, similar to the identify magic spell.

• Faith — At cleric level 3, your Willpower proficiency increases from d4 to d6. You also choose a specific aspect of the Divine to focus your faith upon. So long as your actions stay true to those principles, you can use your Faith to perform miracles within that domain without having prepared them. It costs 1 Faith to perform a first-circle domain miracle, 2 Faith to perform a second-circle domain miracle, or 3 Faith to perform a third-circle domain miracle. You can also perform the miracles bless and rebuke the unholy, which have no domain.

• Second Circle Miracles — At cleric level 5, your Willpower proficiency increases from d6 to d8. When you prepare your daily miracles from your prayer book, two of them can be second-circle miracles.

• Moral Guidance — at cleric level 6, your contact with the Divine allows you to counsel others away from the darkness. Whenever you or an ally would perform an action that goes against both their inner moral code and the precepts of your deity, you may spend one Faith and say a few words of inspired counsel. If there is any unnatural influence on their actions, they may immediately make a saving throw to break that influence, with a bonus equal to your Willpower die + your Charisma modifier. If they succeed but the influence cannot be broken, they may still ignore it for their current action.

• Divine Power — At cleric level 7, your Willpower proficiency increases from d8 to d10. Whenever your Willpower die rolls a ‘10’ or greater while performing a miracle, the effect is a critical success; whenever a curse damage die rolls a ‘10’, the curse inflicts a critical hit. Additionally, you can now perform third-circle rituals from your domain, and whenever you perform any domain blessing or curse as a ritual during the hour before or after sunrise or sunset, the effects remain until the next sunrise or sunset without the need to maintain concentration. Performing a ritual at this time reduces the time you have to prepare prayer strips - and thus the number of prayer strips that you can prepare - by half, and only one may be a second-circle miracle.

• Third Circle Miracles — At cleric level 9, your Willpower proficiency increases from d10 to d12. When you prepare your daily miracles from your prayer book, one of them can be a third-circle miracle.

• Divine Intervention — At cleric level 10, you can pray directly to the Divine for aid. Each exploration round spent in fervent prayer lets you make a Willpower test with a chance of failure based on your piety, the danger that your party finds itself in, and how aligned the request is with your Domain of faith. If you succeed, the referee will describe the effect.

This is starting to look good! The break at level 3 between Wizard and Cleric is interesting — Clerics get a full-on “encounter power pool” at level 3, while Wizards instead simply get the ability to re-prepare spells at roughly the rate that everyone else replenishes their encounter pool. Keeping wizards the “pure Vancian caster” seems important for a B/X clone, somehow.

Now, let’s figure out how we want to do Druids. I would love to make Druids a little different/special, as magic-users go — something like a cross between Dolmenwood’s Enchanter class, 3E/4E/5E’s Sorcerer class (Fey-aspected, of course), and the old-school “Nature-worshipping Divine caster” class.

Obviously, let’s give them Nature and Animal Ken as their starting skills. As for magic, let’s try on something a little weird:

• Primal Attunement — Each sunrise or sunset, you can prepare a number of primal attunements based on your Nature proficiency: two at +d4, three at +d6, four at +d8, five at +d10, and six at +d12. Each attunement is prepared into a small medicine bag full of sacred herbs and materia harvested from the wilds, and each attunement is aspected to that materia’s element when prepared (Stone, Water, Wood, Air, Flame, or Fey). Attunements retain their potency until they are used, or until the next sunrise or sunset. Druids can also sometimes befriend small sentient bits of animated materia, called Wisps, which serve the same purpose as attunements but also provide various additional benefits.
  If you are not wearing, touching, or carrying any iron or steel, you may use a prepared attunement to cast any first-circle Druidcraft that matches its element. Unlike a wizard or cleric, you do not need a book to remember your magics.

• Wild Shape — at druid level 2, you learn to transform into various animal forms. You learn the shape of one CR ½ natural animal at level 2, and can perform a ritual to learn to transform into any natural animal whose CR is less than your own. You may learn a total number of shapes based on your Animal Ken proficiency: two at +d4, three at +d6, four at +d8, five at +d10, and six at +d12. You cannot be wearing or touching any iron or steel as you transform, and transforming into or out of wild shape during combat uses your action. While in wild shape, you use your Animal Ken proficiency as your attack proficiency for any natural weapons that your animal form possesses, as well as your skill proficiency for any skill that the animal form natively possesses. You cannot cast spells while in wild shape, and you must maintain concentration to keep your form.

• Mana — at druid level 3, you may choose to increase either your Nature or your Animal Ken by one die size. You begin to tap into a wellspring of primal power called mana, which you may spend to retain an attunement when you cast a spell from it.

• Second Circle Druidcraft - At druid level 5, you may choose to increase either your Nature or your Animal Ken by one die size.
  You may now use an attunement and two mana to cast a second-circle primal spell that matches its element.

• Resist Glamour — At druid level 6, you have advantage on all saving throws vs. supernatural effects that would influence your emotions, beliefs or perceptions. You may also spend one mana whenever you would fail such a saving throw to retroactively add your Nature die to your roll.

• Primal Rejuvenation - At druid level 7, you may choose to increase either your Nature or your Animal Ken by one die size. You may now use your Mana while in wild shape to heal yourself. As your action or reaction, you spend one mana, then make an Animal Ken test and add your current form’s Constitution modifier to heal yourself for that many hit points. If you use this feat as your reaction in response to receiving damage, you do not need to make a Fortitude saving throw to avoid losing concentration on your wild shape form.

• Third Circle Druidcraft — At druid level 9, you may choose to increase either your Nature or your Animal Ken by one die size. You may now expend an attunement and three mana to cast a third-circle primal spell of that charge’s element.

• Fey Shape — At druid level 10, you can learn to transform into any species of natural animal or fey beast with CR less than your own, and can use your mana to power any special supernatural ability that a fey beast might have while you wear its form.

So there we have it! Three different magic-users, each with its own unique magic system, but still close enough to each other to make comparisons.

We can expand from here with some useful DM-facing procedural rules — how to utilize skill tests and saving throws during travel, how to utilize skill tests and saving throws in specific exploration situations, how to design traps and dungeons and hexcrawls, and obviously how to run combat, but all of that is just a matter of picking the edition that does it best and stripping it for parts. I’ve wound up with something like a simplification of 5E’s combat system, a simplification of Justin Alexander’s excellent dungeon-crawl procedures, and a straightforward 3-mile-per-hour hexcrawl loop. All of which take up, maybe, fifteen pages of rules? Tops?

Let’s take a look at a hexcrawl travel loop, utilizing the dice resolution mechanics, skills, and defense proficiencies presented here:

• Hexes and Leagues — Maps are typically divided into hexes for convenience, with each hex measuring a three-mile league; on a well-traveled road with no hazards or interruptions, a well-stocked adventuring party can expect to cross a single league in about an hour. Difficult terrain, heavy burdens, or bad weather might cause the party to become bogged down within a hex, requiring an Athletics test (or rarely a different proficiency test) vs. a chance of failure based on the terrain to make progress. Each member of the party must succeed for the group to be able to advance; the group may attempt clever solutions to adjust the CoF, grant themselves advantage, or aid the rolls of the less athletically-inclined members.

• Assign Duties — Before each four-hour shift of travel, the party should declare who is navigating, who is keeping watch, and who is attending to other duties.

• Navigation — The referee makes secret Nature and Lore tests on behalf of each navigator at the start of the shift, and notes the highest result; they will use this value to determine whether the party becomes lost.

• Tending to the Load — Depending on the terrain, certain members of the party will need to tend to the wagons, packs, and other loads, to keep them balanced in difficult terrain, or to keep them from getting soaked or otherwise damaged in inclement weather. Crafting skill tests are typically used for this purpose. Similarly, Animal Ken tests might need to be made, to keep animals on-track and motivated, to prevent them from being spooked, and so on. Anyone tending to the load can’t also be on watch or navigating.

• Stealth — The party may choose to employ stealth while traveling, halving their speed. For each league of travel, each character rolls a Stealth test; the lowest result among the party determines whether the group stumbles into a random encounter.

• Forced March — The party can attempt to force-march through a league, allowing them to attempt two league-crossings in one hour; each time they do so, each member of the party must succeed at a Fortitude test with a CoF set by the referee, or suer one level of fatigue. The CoF initially depends on the terrain and the encumbrance of each party member, and increases by 2 for each consecutive shift of travel without a full day of rest, or each further hour of forced march.

• Third Shift — In addition to forced march, a party may decide to attempt a third shi of travel, granting them an additional 4 hours of travel in a day. A third shift also requires a Fortitude test to avoid fatigue.

• Random Encounters — Each league of travel comes with the chance of a random encounter. The nature of those potential encounters, and the probability that they occur, is dierent from one league to another. The referee will typically have a random encounter table that lists the chance of encounter for a given area, and will roll for each league of travel; if an encounter occurs, travel stops as the party deals with the situation. If either party was moving stealthily, then each member of the other party who was keeping watch rolls a Perception test. If either side’s highest Perception result fails to beat the other side’s lowest Stealth result, they are surprised by the encounter.

• Line of Sight — Line of sight on perfectly flat terrain and on a clear day only extends for about a league. This means that, barring terrain or weather, the party can usually see midway into the next hex when standing in the middle of a hex, and can fully survey an adjacent hex when standing at its border. Of course, perfectly flat terrain and perfect weather are almost unheard of in the wilds. Fog and precipitation will obscure line of sight, preventing the party from seeing out of their hex; various terrain features will also be noted as blocking line of sight across a hex boundary. Surveying an adjacent hex grants the party advantage to their Nature test to cross that league; the referee is encouraged to be descriptive in detailing the landscape, and the party is encouraged to be descriptive in how they intend to overcome the challenges they see.

• Lookouts — On the other hand, various points on the map will be designated as lookouts. A lookout is an elevated vantage point that the party could reasonably get to and survey the local landscape. Weather permitting, a lookout allows the party to see one or even two leagues away in certain directions. A lookout will note both what directions it provides a view in, and the diculty of any Athletics or Nature test needed to ascend it. Detouring onto a lookout typically costs about an hour of travel. Artificial structures such as watchtowers also dot the landscape, which provide stairs or ladders to easily scale them, and serve as lookouts in all directions when the weather is clear.

As you can see, once you’ve got a sufficiently sensible and versatile core mechanic, a page or two describing a procedure (or maybe even just a flowchart) is all you need to communicate an entire pillar of the D&D game loop!

(4E made a noble attempt to do this with “Skill Challenges”, but was doomed from the start because they were trying to abstract away the whole point. They did this because a decade or more of )

After that, it’s just padding out your heartbreaker with all the gonzo setting stuff you can think of — monsters, magic items, potions and elixirs, spells, all that good stuff. Really, one of the advantages of B/X is that you can basically judge how well you’re doing by how much of the book is taken up by setting vs. rules — if you’ve got more than 50% of your page count dedicated to rules-and-procedures, you may want to go home and rethink your life choices.