How I’d Make D&D Next

I’ve debated quite a bit if a post like this, where I make a hypothetical stab at designing something that someone else is already making, is worthwhile. The fact that you’re reading this is proof I think this adds something to the discussion. If nothing else it’ll be interesting to look back and see in a year or so how close my ideas are. I’m filing this under Indies & More since it is in a way a response to those Legends & Lore columns, just like D&D Next is a response to those columns.

First off, what this is not: this is not designing my own version of D&D (though I am doing that). This is taking a stab at designing the game that fits the goals WotC has stated and the approaches they’ve made public. Let’s call those our Axioms and Theorems (axioms being the goals WotC has, theorems being the elements of design we already know from those goals).

Axioms and Theorems

  1. D&D as a brand must grow
  2. There are growth opportunities within the tabletop RPG market
    1. From 1 & 2 it follows that a new tabletop RPG edition of D&D is important.
  3. D&D is many things to many people (not just in minutiae, but in what the game is about)
  4. Each previous edition of D&D has catered to a different audience
    1. From 2, 3, & 4 it follows that a new version of D&D could succeed by bringing back in lapsed customers
  5. D&D must be one game (or at least one tabletop RPG)
    1. From 4.1 & 5, the one true D&D must also cater to people with very different tastes
    2. From 5.1 it follows that D&D should be a modular game that can be what each group wants it to be

This is at a high level what we know from WotC’s current statements. Before I start wildly designing from them, let’s look at a few ways to axioms could be different (this is the non-euclidean geometry section, essentially).

Axiom 2 is weak at the least. While there are plenty of people hating on 4E, I don’t know how much that effected sales, at least of the core books. Are those growth opportunities really big enough to be compatible with Axiom 1?

  1. To grow the tabletop RPG market we need flagship games of a different type

This doesn’t make a difference to any of the theorems, interestingly enough. Of course this hasn’t been their public statement, but it seems like a valid tack: if D&D could do more, would more people buy it? It actually subsumes part of axiom 3, in that it assumes there are other audiences than the current tabletop audience that can be reached by the right game.

Since this isn’t truly rigorous logic I’ve allowed myself some jumps here, but I’m not entirely sure 5.2 follows from 5.1. Couldn’t there be a new game that still caters to all people with different tastes, a non-modular one? To strengthen 5.2 I’ll introduce Axiom 6, which is entirely supposition on my part, and for which there are two candidates:

  1. Different types of play are better served by different rules (i.e. rules matter)
  1. More systems mean more opportunities for additional content

Both of these suggest why a modular system instead of one monolithic system. Which one is true for WotC (or if they’re even needed for 5.2) depends on your opinion of the company and the designers. Both may even have been used to different audiences. I’m a rules matter person, so I choose to think of this as a nod to the fact that GM, rules, and players are all responsible for the game produced.

The Design

Enough of that. Given those design goals, how would I make D&D? (I’m so glad you asked…)

First I’d look at what modules need to be able to do (i.e. what bits need to be completely changeable from group to group):

  • Some modules need to make small adjustments to specific areas, like lethality
  • Some modules need to add detail to certain types of play, like social situations or kingdom management
  • Some modules need to define setting-related stuff, like which races and classes are playable
  • Some modules need to set the playstyle, like action-adventure test-your-mettle, or character-driven story
  • Some modules need to define advancement

Given that list I can look at what doesn’t need to be modular and make that the “core game.” There’s not much left except for the very bones of a system for how to resolve an action or event, so that’ll be our “core.” (Note that to my mind just a resolution system doesn’t make a game, but that’s just semantics. The fact that everyone’s rolling the same dice in vaguely the same way may be enough for the marketing angle of “one game.”)

The actual details here don’t matter much, but I’ll throw them out anyway. The core game is:

Your character is six stats, each of which has a modifier associated with it. When the GM tells you to test your abilities they will tell you what number you need to roll, then you roll a d20 and add whatever modifier the GM says applies. If you meet or beat the target number you succeed, the GM will tell you how. Otherwise you fail, the GM will tell you how.

This is a ridiculously simple core, and one that says nothing about the actual important parts of play: when are your abilities tested? What does failure mean?

So far we have the most boring dice rolls ever (I refuse to even call it a game). This is the most I can see in common across the vast playstyles of D&D, mostly because something along these lines is common to the majority of RPGs. Look past the exact dice and modifiers, and remember that we don’t have the real meat yet, and this is 1st Edition to Primetime Adventures and everything in between.

Now the modules. The first and most important module would be the playstyle one. Of course playstyle is a term that probably wouldn’t fly with a broad audience, so I might call it the Focus, Adventure, or Advancement module.

Each playstyle module contains:

  • rules for what earns rewards (but not how to spend them)
  • rules for the GM’s basic responsibilities
  • rules for what success and failure mean

For example, the Sandbox playstyle sets rewards based on gaining knowledge about the world and the events going on, and sets the GM’s basic responsibilities as creating a living world and presenting it authentically. The Challenge playstyle sets rewards for beating enemies, the tougher the better, and gives the GM the responsibility to create challenging encounters that may not be winnable.

At this point, with core+module, this is more or less an actual game. The broad categories of  “GM responsibilities” and “success and failure” are actually huge deals. They fill in some of the key pieces from the core. In particular the GM’s responsibilities and the meaning of success and failure should establish what counts as a “test of abilities.”

The interesting bit here is that if groups were actually willing to do this kind of pick-and-choose playstyle they’ve basically been given a way to talk about creative agenda without sounding like navel-gazing theory snobs. If it could gain traction with actual players that’d be kinda cool.

At this point the group, the GM, the setting designer, or whoever is making the calls on which modules to use has said what the game’s about, now they need to choose what the players do (or at least what’s important for them to do).

This is where the detail modules come in. The detail modules would have sub-systems the give mechanical detail to certain situations while still allowing the call on which of those situations matter to be made by the playstyle module.

For example, the combat detail module describes combat a lot like modern D&D: attacks, HP, etc. The decision of when a combat matters is still down to the playstyle. The playstyle might establish that it’s only a test of ability if the character’s beliefs are on the line, which isn’t overidden by the detail module. Instead, if a test of ability would represent a combat, you use the combat detail system.

Anything not covered by a detail system is covered by the core game (+ playstyle). If we don’t use the combat detail module and our characters end up testing their ability in a fight, we just make one roll, success or failure, as the core game describes.

With core + playstyle + detail we now know what the game’s about and what the characters do. The next module choice describes the types of characters and the world around them: the setting module.

Setting modules define classic stuff like class and race, how many of each a character gets, etc. Of course how those are actually represented depends on the detail modules in place. The core of each character, as described in the core game, would need to be just stats. Based on the detail modules in play, different parts of the classes come into play, but the way they come into play still must be general.

As an example, let’s use arcane magic. In a game where there’s no arcane magic there should be no classes that primarily grant arcane magic and classes that grant arcane magic as a secondary ability should ignore it. If arcane magic exists but isn’t using a detail module, then it’s just like a stat: some modifier used when testing an ability. If there is a detail module that detail module has to describe how to translate that arcane magic stat into the right level of detail. Maybe for each point of arcane magic you know one spell castable once per day, or something like that.

Of course there can be different detail modules that use that some information. One detail module for arcane magic is vancian, another is skill-based (so your arcane magic stat turns into a set of points to spend on more specific skills), etc.

With core + playstyle + detail + setting we have enough to start playing, but how characters grow is down to the advancement module.

The advancement module takes the rewards allocated by the playstyle module and turns them into concrete additions to the characters. Of course the advancement module can’t know what detail or setting modules are in use, so it has to be generic. The advancement module would have to hand over information to the character in a standardized form: levels, new abilities, and advancing abilities.

Levels set a power cap in terms of what any individual ability can do. In the core game setting modules this means a cap to modifier, which can filter out into detail modules in different ways (no 2nd level spells until you’re level X, for example). New abilities are how many new abilities you can unlock, advancing abilities is how much you can increase any ability you currently have. There might even be catch-all for other bonuses, call them boons. This would account for feats and such.

The idea being that the advancement module you choose sets what advancing a character means. Does it mean a dramatic increase in power? Or is it a slow broadening of ability? Or do you just pick up specific new tricks?

Core + playstyle + detail + setting + advancement is everything D&D usually has. The last type of module, modifiers, are like the optional rules that get scattered through most editions of D&D.

Modifiers tie to one or more specific modules and just tweak them a bit. Combat modifiers might make certain combat detail modules more lethal. Magic modifiers might change up the schools of magic. Modifiers allow the design of two modules that are mostly the same, but with a few differences, without rewriting the whole thing. Instead of grid-based combat and deadly grid-based combat being two entirely different modules one is just a modifier to the other.

That’s it. Every group chooses 1 playstyle module, 0+ detail modules, 1+ setting modules, 1 advancement module, and 0+ modifier modules, and they’ve effectively made their own D&D, but under the umbrella of The One True D&D. No more edition wars, we’re all buying the same product. We may war over modules (“Grid-based non-deadly combat? Why?”) but at least we’re all still buying WotC products.

The Products

What would a product look like in this modular world?

Each product would need to be heavily themed. Since what mechanics get used is down to which systems are used you can’t sell people a book based on the fact that all the rules will be useful for their game, you have to sell it based on the interest in the theme of the book.

So, a Martial Power style book might have:

  • A new combat detail system
  • Modifiers for certain classes (or types of classes) based on martial options
  • An advancement system that prioritizes combat prowess
  • Modifiers for deadlier or more lenient combat
  • One or more settings, where a setting is a list of what modules to use together for a certain game
    • For example, a mercenary setting, made up of the challenge playstyle, combat-ready setting modules (fighty-y classes), details for combat and negotiation, modifier for deadly combat, etc.

The book really has to appeal based on “ooh, I want to play that, no matter the rule” than “oh, this has a useful rules update.” It might lead to the option of smaller books, since a setting could easily be established using a specific combination of existing modules with maybe some custom modifiers.

WotC could then open up the system, much like 3E, but with a clearer back-and-forth. The modular system gives other companies a clearer way to position their books as complimentary to WotC’s, and even an entirely new game would still be providing more modules that are then part of the overall ecosystem.

3E OGL products could only go so far before they’re not compatible with D&D and that line wasn’t always clear. The system wasn’t modularized so an alternate skill system would entirely break the skill-based magic of another book. So long as each designer abides by the division of modules inconsistencies can at least cleanly be recognized (“oh, this setting module requires this detail module” for example).

Mutants and Masterminds is an example here. In the 3E ecosystem M&M’s contribution was mostly to mindshare: I know I was excited by the idea of playing superheroes with a system I already “knew.” The problem being that in truth M&M shares only a few concepts with d20 at the lowest level (d20+something vs. DC, a few others) so none of that cool content is really part of the larger ecosystem; Mutants and Masterminds didn’t give anything back to the overall design except for the marketing power of  “d20 can do superheroes!”

With a modular system M&M could be a larger part of the design. Powers as ranked skills would be a detail module that converts the basic ranks defined by setting (class) modules into specific powers. The book would also contain playstyles for supers, setting modules for supers, and so on. It might even have other detail modules that depend on the powers detail module. Now all those have greater compatibility with the world at large: I can choose to describe my fantasy adventurer’s abilities using superpowers, or use a superhero playstyle with my wild west rough riders.

Adventures would be largely presented in the fiction. Most modern adventures devote huge swaths to stats and encounters which would now be largely dependent on the modules in play. Instead adventures would be mostly about the in-game situations. Specific modules can only be leveraged within an adventure if the adventure says clearly that it depends on them.

That means that an adventure that doesn’t say “requires one of these detail modules” has to assume that all combat can and will be resolved with the core system (i.e. one dice roll). That’s a challenge to designers certainly, but I think it’s a good one. Adventure design really has to be about what the adventure means within the world, not about what rules it invokes.

Of course most adventures would only work well with certain playstyles. This makes playstyles the least-extensible point, which is probably just as it should be. There are many ways to pursue the same creative agenda (which playstyle basically is) while focusing on different details, but only so many entirely different playstyles will gain enough traction to make sense to a company like WotC.

The Links Between Modules

How modules work together is important, something that I’ve slightly glossed over. If you’re a software person, this may all seem like a flashback to a design class (if you had one) or spec review. Sorry.

As I’ve mentioned above each module links with other modules of other types in certain ways. The playstyle module has to describe rewards, which the advancement module has to be able to consume. In fact the rewards produced by any playstyle module need to be consumable by any  advancement module.

Of course there can be declared incompatibilities, or declared synergies. Maybe each advancement module has a “best with” section listing other modules it works well with, to guide the GM. Maybe there are even some edge case combinations that just flat don’t work (the more of these there are, the less the system is modular).

The entire design, being modular, comes down to the designers’ ability to smartly design the interfaces between modules. In my design above I went into some detail on the interface between advancement and setting: each class element of the setting has to be able to consume the output of any advancement, which means the output of advancement has to be standardized. I did this with Level, new ability, increased ability, and boon, but there are any number of workable solutions.

How interfaces are defined becomes the most important part of the design.

Versioning interfaces is a software problem as well, but I think with D&D there’s a way around it: series of modules.

What it’d actually be is a way of marking which “version” of an interface the module supports. Maybe advancement modules don’t start with a good way of communicating feat-style advancement to a class (and classes don’t have feat-like improvements), but later WotC wants to add that. They simply start a new series (S series, let’s call them) where all the classes have feat-like advancement options and all the advancement modules describe how to gain feats. They could even update the old A series basic advancement options with modifiers to make them S-compatible. Now each module can mark compatibility within a series, or with everything but a certain series. It’s not perfect, but it does allow the kind of versioning that’s needed.

Using the Mutants and Masterminds example, M&M would have some modules that fit existing interfaces and then maybe have some that depend on those. For example, a powers detail module that does M&M-style powers, then detail and modifier modules that do superpowered chase scenes or adapt the core combat module for powers.

I also love the idea that, in D&D Next, “series of modules” would be completely relevant but completely different from the old adventures. Nice nod to the past.

The Last Edition

This entire system of versioned interfaces means that nothing ever really has to be abandoned. Instead of a dramatic shift to a new edition, publishers can define a new set of interfaces and slowly shift support to them.

Example time: we have our A series playstyles, the 2 or 3 from the core book. WotC eventually decides that playstyles need to also set difficulties, which had been part of detail modules. This changes the interface between playstyle and detail modules, which means a new series, let’s call it B series. Instead of doing a whole-hog edition swap they can start with this new B series in a large expansion book, self-contained. Then some later small expansions support both A and B series to some degree. Finally B series becomes the default, integrated into a new printing of the core, and A series doesn’t get much if any new support.

This is more or less what WotC does with Magic to great effect. Instead of saying that this new edition of Magic entirely supplants the old one, they maintain official “formats” that include only selected sets but casual play usually includes whatever sets and editions the players care to bring.

Similarly each setting could be defined by a set of modules which can slowly change over time, maybe with the evolution of the setting, or with a focus on different areas. Forgotten Realms: Wild Areas uses the A series and focuses on combat. Forgotten Realms: Lords and Ladies uses B series and focuses on intrigue.

Since series are not games (again this is kind of a trick of semantics) there’s no edition wars or all-together stop of support. Each series would (hopefully) not completely redefine every interface so even though A and B series playstyles and details are incompatible, their setting and advancement aren’t. This way groups can slowly migrate as they see fit and get long term support in some areas (but not all). My A series game can make use of selected parts of B-series books, and of course by buying B-series books I’m acumulating B-series content, so eventually shifting over to being a B-series group is fairly painless.


Of course as presented here this seems wildly complex, but hopefully part of that is because I’m presenting this to you as a game designer, no as a potential player. (We’re all game designers, of course, but more on that in a later post.) The way this is presented in a core book would be much simpler. Instead of this huge number of combinations there would be different levels of brew-your-own.

At the top are 3 or 4 ready-to-play combinations reflecting broad kinds of D&D play. These are presented as canonical game styles, and are heavily encouraged. None of the 3E-style “the first thing to do is completely customize this game.” Instead these core combinations are what are assumed by most adventures, and most adventures support all of them.

Below that are the recommended combinations. Instead of all the choices being made for you, most of the choices are made for you. They might even be presented as variants of the ready-to-play combinations. “To customize the Exploration play-set, you can choose from these options.” This is for the group that wants something just a bit different, or the group that feels like it’s not play without customization.

Finally there’s the free-for-all, choose whatever modules you like. This is presented as an advanced option, with the statement that you should play one of the pre-configs first to see what you like. There’s no guarantee that they’ll have any real synergy, but they will at least work on the basic level of handing off the right information between modules.

Enough Already

That’s my vision of D&D next based on the comments made: a modular “game” where the modules can actually swap out huge swaths of what the game is about, what the players do, and so on.

This is radically different from 3E where, sure, you could swap out a fair bit of how things work (details), all of the fiction (setting), and there were plenty of house rules (modifiers). First off 3E had no standard way of separating these concepts: a new class might be entirely dependent on the existing combat system, but not make that clear. 3E also had no clear way of swapping out what the game was about. It was certainly possible, but it required massive rewrites, taking d20 basically down to the core mechanic.

It’s not the simplest of ideas. In fact it’s a pretty complex systems design problem. But it’s a way of both saying “this one game is for everyone” and giving everyone a different game. I’m interested to see if my ideas are anywhere close to what WotC’s doing.

Not that this is the one true way, but if WotC hasn’t bought in to some of my proposed axioms, up top there, they may be saying that this is one game for everyone but not actually making a game that does that. It could be like GURPS: a system that claims universality but only supports one mode of play. With a different choice of axiom 6 the entire modular system could be about choosing between miniature combat and  zone combat and freeform combat, not about actually changing what the game can do.