The OSR Onion

— Boëtius Adamsz. Bolswert, early 17th century (foreground)
— Étienne Léopold Trouvelot, late 19th century (background)

Summary

The OSR may be best understood as a layered approach to design, much like Powered by the Apocalypse (PbtA). However, while the shared foundation of PbtA games is the structured conversation, the shared foundation of OSR games is a specific way of writing, refereeing, and playing adventures.

What the OSR Means to Me

There’s disagreement about what the acronym OSR even stands for—let alone how to define it. Broadly speaking, people define the OSR by one of two opposing criteria:

• The compatibility camp argues that in order to qualify as OSR, a game’s rules must be compatible with old D&D modules.
• The principles camp argues that a game’s support for certain principles, not its specific mechanics, determine whether or not it’s OSR.

I stand squarely in the principles camp, and principles are the focus of this piece. If you’re a firm believer in compatibility, either read on with curiosity or don’t; my aim isn’t to convert you.

What are OSR principles? There’s no definitive list, but here are my favorite sources:

• The Principia Apocrypha by Steven Lumpkin, Ben Milton, and David Perry, illustrated by Evlyn Moreau
• “The ICI Doctrine” by Chris McDowall
• “7 Maxims of the OSR” by Gus L.
• “Basic OSR Principles” by Yochai Gal

Because there’s overlap among these sources, and because I’ll reference specific principles in this post, I’ve done my best to compile a list with abbreviations indicating which sources express a variation of each principle:

• Prep situations, not plots (PA, 7M, BOP)
• Present interesting choices (PA, ICI, BOP)
• Leave space for and reward player questions (PA, ICI, 7M, BOP)
• Don’t gate information behind die rolls (PA, ICI, 7M, BOP)
• Telegraph danger (PA, ICI, BOP)
• Present a responsive, dynamic world (PA, ICI, BOP)
• Combat is a fail state / Combat as war, not sport (PA, 7M, BOP)
• Embrace high lethality / impact (PA, ICI, 7M, BOP)
• Create challenges with no obvious solution (PA, BOP)
• Reward ingenuity / Good plans don’t require die rolls (PA, 7M, BOP)
• Don’t fudge dice (PA, 7M)
• Embrace the oracular power of dice (PA, BOP)
• The answer isn’t on your character sheet (PA, 7M)
• Good items are unique tools (PA, 7M)
• Rulings, not rules (PA, 7M, BOP)
• Role-playing, not roll-playing (PA, 7M, BOP)
• Strict time records must be kept (PA, 7M)

Notice how many of these principles concern a way of writing, refereeing, and playing adventures. Several principles even caution against an overreliance on dice and “rules.”¹

Blorbpocalypse Now!

Vincent Baker wrote a fantastic series of blog posts on designing PbtA games. You should absolutely read it even if you somehow have no interest in PbtA or story games. In part one, Baker discusses the relationship between fictional things and real things in Apocalypse World using a rollercoaster diagram. He writes: “Apocalypse World’s philosophy is: use the real things, the dice and stats and so on, to give momentum to the fictional things.” Later, in a Q&A post, he revisits this idea: “PbtA focuses on the players’ interactions and their relationship with the real-world play material.”

I find his writing about the PbtA design philosophy riveting, even as it clarifies for my why I prefer the OSR style of play. Generally, the more a role-playing game focuses on fictional things, and the less it focuses on real things, the more I enjoy it.

If a game doesn’t rely on real-world play material for structure, then what can it rely on? It turns out Baker has already considered this. In the same Q&A quoted above, he writes:

Blorb is a disciplined, principled approach to old-school GMing. It emphasizes the reality of the game world, and provides fruitful constraints on the GM’s improvisation in play. […] I see it as a very strong system of support for GMing in play, not a set of restrictions to labor under, and at the same time a provocative spec for design.

Take a moment to read “Blorb Principles” if you haven’t already. While there’s tension between the Blorb’s focus on diegesis and non-diegetic OSR mechanics such as experience points, there’s also clear overlap between OSR and Blorb principles. Like the OSR, Blorb rejects plot in favor of emergent narrative, disdains the quantum ogre, and focuses on specific material conditions in the fiction. “Prep is central” to Blorb, and prepared material is the highest tier of truth. Beyond prep, Blorb principles encourage the referee to rely on general rules or mechanics such as random tables. Sounds pretty OSR to me.

In “Pleasures of the OSR: Secrecy and Discovery,” Ben Laurence makes several points that align closely with Blorb principles. He writes:

One crucial point is this: in the ordinary sense of the term, one can only discover something that is already there. If you make something up, that’s not discovering it. Another important point is that we need to distinguish between the character (PC) discovering something in the world of the game and the player of that character discovering something about the world of the game. Those are two very different things. It is possible for the character to discover something while the player does not discover anything. One case is where the player already knows what his character discovers. Another case is where in the “fiction” of the game the character comes to discover something, but the player does not because the “fact” was made up on the spot, either by the player or by the DM.

It’s a landmark post, and I highly recommend reading the whole thing if you haven’t already.

Anyway, my takeaway from Blorb, Laurence, and a several OSR principles is this:

In order to support discovery and fair, objective challenge, you must have prepared material. There has to be a there there. In other words, you need an adventure.

The OSR Onion

Let’s return to part one of Vincent Baker’s series on PbtA design. “Apocalypse World is designed in concentric layers,” he writes, “Like an onion.” There’s a great diagram you should check out. Baker continues:

A crucial feature of Apocalypse World’s design is that these layers are designed to collapse gracefully inward:

• Forget the peripheral harm moves? That’s cool. You’re missing out, but the rules for harm have got you covered.
• Forget the rules for harm? that’s cool. You’re missing out, but the basic moves have got you covered. Just describe the splattering blood and let the moves handle the rest.

That doesn’t sound all that different from “Rulings, not rules,” does it? I’ve often heard OSR referees say something to the effect of, “If you forget a rule, don’t stop play to look it up. Just think about what’s happening in the fiction, make a ruling, and move on.” Quite a few OSR principles indicate that the OSR features concentric layers. (More on that soon). Here’s my version of an onion diagram showing the OSR’s layers:

The Adventure layer contains all of the Blorby stuff like:

• Pulling the lever causes the water in the flooded corridor to recede.
• This 20×20 foot room has a tapestry depicting the death of Pongo the Pompous; behind the tapestry is a secret tunnel.
• The high priestess of the mudlings desires the destruction of the dam.

I call the central layer the Adventure layer to draw attention to the centrality of adventures and prep to the OSR. However, it also contains things beyond what’s included in an adventure’s text² or a referee’s notes, such as:

• Milena is carrying an angry gremlin in a burlap sack.
• Fredo has been turned to stone.
• Robin has earned the blessing of Fiddlehead of the Linnelmede Forest.

The Procedures layer contains “the rules for using the rules,” as Gus L. puts it:

• “When the characters subject a retainer to particularly dangerous conditions, roll a morale check.” — D&D Basic Rules, page 21
• “Moving quickly or without caution may increase the chance of encountering a wandering monster, springing a trap, or triggering a roll on the Dungeon Events table.” — Cairn 2e Player’s Guide, page 70
• “For tasks with no inherent danger and no time pressure, PCs with appropriate skills and tools should generally succeed automatically.” — Dolmenwood Player’s Book, page 139

The Mechanics layer contains specific rules that involve less interpretation and tend to change a lot from system to system:

• “Roll a d20 + ability score + modifiers (in +5/-5 increments from advantages, disadvantages, and careers).” — Knave 2e, inner side of front cover
• “A smart (philistine) hero can hack out 1d6 + Level percent of a treasure’s value in one turn.” — Ultraviolet Grasslands, page 149
• “Scars: If an attack takes you to exactly 0hp, take a permanent Scar. Look up the amount of Damage caused on the Scars Table below.” — Electric Bastionland, page 9

The Character layer contains rules that differentiate characters from each other. These rules not only change from system to system, they sometimes get omitted entirely:

• Experience points
• Classes
• Specific ability scores, saving throw types, spells, etc.

My point is not that PbtA and the OSR share a similar approach to system design, but rather that the OSR also has layers. I wouldn’t necessarily say that the OSR’s layers “collapse gracefully inward,” nor that they’re a product of intentional design. If anything, they may be incidental to the OSR’s status as a sort of folk D&D. But that’s a topic for another time.

Principles Point the Way

The ordering of these layers isn’t arbitrary; it’s something I arrived at through a close study of OSR principles. Many have a clear directionality to them. They “point” from one layer to another—and always in the same direction: inward.

Let’s look at some examples, starting with principles that point from the outermost layer to the innermost layer, like this:

The principles “The answer isn’t on your character sheet” and “Role-playing, not roll-playing” encourage players to focus on the fictional situation in the adventure rather than skills, class features, and ability scores.

Another reason that Character is the outermost layer is that its contents are non-essential. Some OSR games don’t have classes, experience points, skills, the Detect Evil spell, or even the original six ability scores, and they function perfectly well without.

“Rulings, not rules” points from the Mechanics layer to the Adventure layer, like this:

This principle acknowledges that not every fictional situation that crops up will be covered by a specific rule, so the referee should use their judgement. And, as Gus L. writes, “Both players and referees should not appeal to or demand citations to the rules over the referee’s judgment about events.” In other words, the fiction takes precedence over the rules.

A handful of OSR principles are procedures in and of themselves; they instruct the referee in how and when to apply specific rules. To me it’s quite clear that procedures inherently affect play more than mechanics do and therefore occupy a more central layer. For excellent examples of this, check out some of the examples in Matt Finch’s A Quick Primer For Old School Gaming. Even two games that both have extensive skill lists, such as D&D 5e and Mothership might play completely differently due to procedures that govern how those skills apply to play.³

Two of these procedural principles also point from the Mechanics layer to the Adventure layer: “Don’t gate information behind die rolls” and “Good plans don’t require die rolls.”

The Mechanics and Procedures layers are closely linked, and so too are the Procedures and Adventure layers. The principle that “Strict time records must be kept” illustrates this nicely. In “7 Maxims of the OSR,” Gus L. explains the meaning of this principle: “The mechanics and procedure of turnkeeping, dungeon navigation, encumbrance, supply depletion and random encounters matter as structural supports to location exploration” and “random encounter checks and encumbrance […] require turnkeeping to know when monsters threaten the party and when supplies are used.”

And these mechanics and procedures require an adventure to function. Encumbrance-based movement rates don’t matter if the size of the adventure location isn’t defined. One cannot track the time the party takes to traverse a space without knowing the distance traveled. Random encounter rolls require random encounter tables.

All Roads Lead to Adventure

Finally, remember how many OSR principles are specifically about a way of writing, refereeing, and playing adventures:

• Prep situations, not plots.
• Present interesting choices.
• Telegraph danger.
• Create challenges with no obvious solution.
• Good items are unique tools.

The list goes on. So many of these procedures are about adventures, and that’s because the adventure is the central concern of the OSR. Everything else—procedures, mechanics, character attributes—revolves around the adventure. Without adventure content to give it meaning, the procedures collapse. Players cannot experience discovery, overcome objective challenges, or make truly meaningful choices without concrete, immutable fiction.

On the other hand, if you strip everything else away and rely solely on the adventure text, players can still ask questions about the environment, pursue goals, discover secrets, and make meaningful choices. Yes, such a game will lack many of the OSR’s defining features, but a well-written adventure can still function as a game in its own right.

I’m not saying system doesn’t matter. But in the OSR, adventures matter more, and we should all give greater attention to adventure design. Because after all, adventure design is system design—just not in the way most commonly discussed in the context of role-playing games. In “Systems of Relation,” Jay Dragon writes:

I believe it’s accurate to describe House as existing in relationship to a system just as much as Dungeons & Dragons, but while Dungeons & Dragons is built on a system of mechanics, House […] is built on a system of relation.

[…] Once we free the system from the burden of being a purely mechanical description, we begin to discover what else is possible for that unshackled noun to describe — and it would be foolish to assume this is where it ends.

Adventures consist of systems of conflict, systems of environment, and many others. And people have been writing about designing such systems all along. “Landmark, Hidden, Secret” and the “Three Clue Rule,” for example, are both approaches to designing systems of secrecy and discovery.

Conclusion

A lot of what I’ve written amounts to “the OSR is about adventures,” which probably won’t be news to anyone reading this blog. For me, however, this layered view offers a way to understand the OSR as a consistent, coherent design movement. The OSR is primarily an approach to writing, refereeing, and playing adventures, and it always has been.

From this perspective, the evolution of the OSR appears perfectly natural; of course people iterated on mechanics and procedures over time. Does anyone really believe that the first role-playing game ever invented did everything perfectly and that design has not progressed in the past 50 years? Over time, our understanding of the specific play experience that the OSR offers has grown. As OSR principles become better defined and the art of adventure design matures, designers develop new mechanics and procedures to better support discovery, diegetic play, and impactful, informed decisions.

Finally, I want to touch on the OSR’s future. Here’s the one thing I hope you take away from this post:

OSR designers have done a lot of work to create and perfect mechanical systems. When it comes to other kinds of systems, however, there exists incredible potential for progress. Ty of the Mindstorm blog recently wrote “Pocket-Sized Powder Kegs,” an excellent exploration of systems of conflict and instability. I’d love to see more work like this, and it’s where I intend to focus my attention moving forward. Here are just a few questions I hope to explore:

• How can we design systems of relation to facilitate intrigue?
• How can we design systems of environment to evoke wonder?
• How can we design systems of care to explore the toll of violence?
• How can we design systems of secrecy to support discovery?
• How can we design systems of place to enhance verisimilitude?

We already have a wealth of mechanical systems. I believe the future of the OSR will involve perfecting the design of these other systems.

Footnotes

1. The word “rules” is used imprecisely. Rules can be diegetic or hypo-diegetic, social, or even external to the game entirely. But that’s not what people mean by “Rulings, not rules.” ↩︎
2. It wouldn’t be accurate to call this layer the “diegetic” layer because some diegetic elements, such as classes and spells, belong to the outermost layer. ↩︎
3. A lot of this also comes down to play culture. The 5e Dungeon Master’s Guide describes two mutually exclusive approaches to using dice (pages 236–237). The guide then suggests taking the middle ground: “By balancing the use of dice against deciding on success, you can encourage your players to strike a balance between relying on their bonuses and abilities and paying attention to the game and immersing themselves in its world.” That’s pretty solid advice, but I find that my interpretation of that passage is at odds with the consensus I’ve observed in the play culture at large. ↩︎