In the ever-growing marketplace of role-playing games, designers of role-playing games have an incentive to keep their rules simple enough to learn in 10 minutes with no nasty surprises, or to reuse rules that players may already know.1 (Ideally, both.)
To that end designers have created many systems meant to be reused for different settings and even different genres. I’ve seen some of these endure for decades, others flourish briefly then burn out like any other fad, and still others fall quickly into obscurity and irrelevance despite trumpeting how “universal” they are. I’ve long been curious about why some of these systems succeed and others fail. While market forces, business decisions, and luck play a part, I’m convinced some designs are more flexible, adaptable, and general (or “generic”) than others. Below I’ll try to tease out a few traits that I think make a difference.
Descent With Modification:
As I learned in the world of software, for something to be reusable it first must be usable. Systems that extend beyond one genre or more than a few similar settings began by serving one genre well. Hero System started as the superhero game Champions, Cypher System began with Numenera, Basic Roleplaying (BRP) began with RuneQuest, the modern Fate System with Spirit of the Century, “Powered by the Apocalypse” games with Apocalypse World, and so forth. Even GURPS was a massively over-elaborated version of Steve Jackson’s first RPG The Fantasy Trip (with plenty of inspiration from Champions).
These systems vary widely in many ways, most notably in how much survives the transition from one genre and setting to the next. Cypher System is essentially the same as Numenera with expanded ability lists, types, and other options. On the other hand, Fria Ligan’s various “Year Zero Engine”2 games share relatively few elements, e.g. the d6 pool mechanic, some attribute and skill names, and Critical Injury tables (more or less). GURPS is always GURPS; each “Powered by GURPS” game uses a (mostly) strict subset of the full rules, omitting parts not appropriate to the setting or genre. In comparison “Powered By the Apocalypse” games share few if any details, save the 2d6 mechanic3 and the ideas of Moves and character booklets.
Two Common Strategies:
For a long time I’ve divided multi-genre and multi-setting systems into two rough strategies:
Endoskeletal: the rules are a scaffold on which GMs and players hang their character concepts.
Exoskeletal: the rules form a shell within which GMs and players build their characters.
(Honestly, I’m not entirely happy with those two terms. I thought about switching to “tool kit” vs “rule set” but they didn’t quite click. So for now I’ll stick with endoskeletal and exoskeletal.)
Endoskeletal games can potentially change everything but a few “central” ideas, whatever they are, but because they’re made to be tinkered with, and have relatively few “necessary” rules, they propagate more easily. In a BRP game one can change the basic characteristics, the names and definitions of skills, and all the rules around magic, yet it will typically work pretty well. (Well, reinventing magic sometimes breaks.) They’re somewhat self-balancing … or rather they provide tools GMs can use to decide specific cases. The Fate System includes extensive notes on how and when to hack it.
Exoskeletal games can be used with few if any modifications, but at a cost. GURPS and Hero System rely on point costs and formulas to keep characters – PCs and NPCs – roughly equal in power, and even those have limits. Character generation in these systems often feels like doing one’s taxes.
Some might argue, but I consider the Cypher System from Monte Cook Games more of an exoskeletal system. Granted, the system reduces virtually every in-game action not allowed or denied by GM fiat into a single die roll. (Or sequence of die rolls, for combat and a few rarer situations.) However, it attempts to cover every genre through genre-independent “types” and genre-dependent “descriptors” and “foci”. Characters customize their “type” by choosing from lists of abilities, categorized by “tier”, but there’s little indication why an ability belongs to a specific tier. Cypher System avoids heavy math, but at least the first edition of the Cypher System Rulebook had only a few ways to swap or create abilities for each character “type”. (As of this writing, I haven’t read the 2nd edition yet.) Even if the core mechanic is ultra-light, players must work within the complex player-facing rules to create characters (with the GM’s guidance) and to leverage those characters’ abilities to tilt the odds in their favor.
Note, though, that “endoskeletal” and “exoskeletal” aren’t rigid categories, but philosophies or tendencies. Every system has a core architecture without which the game falls apart, and parts that can be added, removed, or changed with few if any side effects. Even GURPS, despite its claim of completeness, still has room for new or changed skills, advantages, and disadvantages, and as GURPS Thaumatology proves radically new magic systems. On the other side, Chaosium’s “Basic Roleplaying System” can change a lot – witness the changes to Call of Cthulhu 7th Edition – but change enough features – e.g. replacing d100 with d20, opposed Traits, simplified characteristics and combat rules, rules for playing generations of characters – it becomes something else with similar design decisions, e.g. King Arthur Pendragon. Savage Worlds is somewhere in the middle: the core book is complete in and of itself for common genres, but most setting-specific books add custom Edges and other special rules. Like all abstractions and analogies, the endoskeletal/exoskeletal framework has limits. The real question is which rules are truly load-bearing, how many can be modified with impunity, and how easily can a specific change to a specific subsystem break the game.
In any case, the market today favors endoskeletal rules, e.g. Fate System, Powered by the Apocalypse games, the family of “Storytelling” games from White Wolf and Onyx League, and the “Year Zero” system from Free League Publishing. They’re typically low-crunch, light weight, and easily hacked to produce new games with familiar mechanics. But the endoskeletal philosophy is hardly new: one could also cite Basic Roleplaying from Chaosium, based on RuneQuest which was first published in 1978 and tweaked for Call of Cthulhu, a whole host of mostly Moorcockian games now out of print, and finally (via Mongoose’s reimplementation) OpenQuest, Renaissance, and Mythras (after yet another total rewrite).
Exoskeletal rules aren’t as common anymore. GURPS and Hero System are the archetypal examples. These systems have their die-hard fans, who can turn (metaphorical) dials and switches to approach a specific genre. But given their current rarity, and the complexity inherent in trying to cover every eventuality, I’d have to conclude this approach is far less successful.
The Elephant In The Room The Dragon In The Dungeon
And then, as usual, is D&D and the “d20 System”, which defies so many simple categories.
On the one hand, especially in its earliest versions, D&D was a hodgepodge of rules that DMs almost had to hack for their own use. With 3rd Edition came Open Game Content4 and the d20 System Reference Document (d20 SRD)5 which publishers could and did copy, change, and reuse for their own purposes.6 Most publishers tweaked classes, skills, feats, spells, and other bits to support a different setting (in roughly the same genre). In practice, it was easier to refer to sections of the Player’s Handbook7, and add heavily rewritten sections of OGC as needed for clarity.
Only a few publishers (including, occasionally Wizards of the Coast) overhauled the entire SRD for their product. For example, Mutants & Masterminds (Green Ronin), True20 (also Green Ronin), and Call of Cthulhu d20 (WotC) reduced or entirely eliminated classes, and streamlined other systems. d20 Modern (WotC) overhauled classes, magic, and other rules as a basis for games in historical, modern, and futuristic settings; d20 Modern itself was hacked and rehacked by other publishers. In contrast Pathfinder (Paizo) expanded 3.5 D&D enormously, mostly as a reaction to the incompatible D&D 4th Edition. The entire Old School Renaissance used OGC to recreate TSR-era editions, then hacked those even further. So one can’t say D&D isn’t hackable.
On the other hand it’s hard to hack D&D/d20 without breaking “the math”. Creating new classes in even the earliest editions was a black art, and the now explicit philosophy of “simple rules, multiple exceptions” leads to a game where those multiple exceptions are carefully balanced against each other. In D&D and its close derivatives, player characters must kill monsters and gain treasure in order to earn experience points and gain levels; thus PCs have to acquire enough magic items and other perks to kill those monsters, and encounters need to be balanced so that PCs can earn those points without undue risk of a Total Party Kill. Even supplements from D&D’s publishers can break the game in non-obvious ways. Like a character in a sitcom opening an overfull closet to get “just one thing”, modifying the d20 system to change one assumption can cause a huge mess.
On the third hand, maybe D&D shouldn’t be considered multi-genre at all, despite the dreams of the d20 SRD’s authors. One can change its classes, spells, feats, settings, and so forth, but there’s still an inevitable D&D-ish-ness that survives. Classes, levels, hit points, and saving throws evoke memories of dungeon crawls past – intentionally so – even in “clean-room” reimplementations of D&D like Stars Without Number or Shadow of the Demon Lord. Whether far future or grimdark pre-apocalypse, the change in genre feels mainly like a new skin on an old game engine.
Removing some of those familiar elements – e.g. hit points and classes – and adding a lot of new stuff – e.g. point-buy and powers – yields something like Mutants & Masterminds, which is barely recognizable as OGC except in isolated paragraphs. Call of Cthulhu d20 eschewed classes and leaned into skills and profession templates, then added Sanity rules, magic, and other elements from the Chaosium original. (It also kinda tanked.)
Is a system really multi-genre if, to fit it into a truly different genre, one has to remove or rewrite most of it?
RPG systems that adapt well to new genres and new styles seem to share the following traits:
They’re designed and implemented first for one common genre, before being adapted to others, rather being created a priori as a “generic system”.
They provide consistent tools and heuristics that a GM can use to handle new requirements or situations (“endoskeletal”), instead of trying to cover every possibility out of the box (“exoskeletal”).
They’re easy to tinker with, pull apart, and rebuild, with as few hidden assumptions and interdependent parts as possible.8
Ultimately players and most GMs care more about their experience at the table than the rules’ relationship to other games they’re not playing or the “math” that makes a game work. But a well-crafted system of rules, tailored to a specific genre and setting, typically yields a better experience than a hodgepodge of rules that are hard to remember, distract from play more than they enable it, and/or require the GM or Referee to modify them on the fly.9 Familiar, well-tested, reusable rules enhance the experience.
Marketing hype to the contrary, no one set of rules can cover all potential settings, genres, moods, and in-game experiences … although players and GMs for whatever reason may seek out only a few. ↩︎
As a result of their Forbidden Lands Kickstarter, Fria Ligan developed a Year Zero Engine SRD. The rules most closely resemble Mutant Year Zero and Forbidden Lands, with elements from Coriolis, Tales from the Loop, Alien, and the upcoming Vaesen. ↩︎
Except Kult which uses 2d10. ↩︎
Open Game Content wouldn’t exist without the Open Game License (OGL), which has had a much broader impact on the industry, but that’s outside this discussion. ↩︎
This started a trend in SRDs for other, completely unrelated systems like Fate, GUMSHOE, Mongoose’s version of RuneQuest, and a whole host of mostly forgotten RPGs trying to hit the big time. The pros and cons of “open source” RPG rules are another interesting topic that deserves its own article … which someone else has no doubt written. ↩︎
Publishers can produce D&D-like games without using a word of OGC, and have (e.g. Sine Nomine Publishing). But OGC’s mere existence meant that, unlike D&D’s previous owners TSR, Wizards couldn’t sue over D&D-compatible content … or, as it turned out, remixed D&D rules. ↩︎
To officially bill itself as “d20 System Compatible” under the d20 License, a product had to refer to the D&D 3.x Players Handbook Experience Table, intentionally excluded from OGC, rather than substitute their own. In only a few cases, e.g. Mike Mearls’s Iron Heroes, did this requirement become absurd. ↩︎
D&D-like class-and-level systems seem harder to tinker with than those that define characters with independent or weakly coupled skills and abilities. ↩︎
In a future essay I hope to expand on when and why System Sometimes Matters. ↩︎