Who’s Afraid Of The OSR?

A couple of weeks ago Admiral Rabalias asked

I saw some thread saying most storygamers also play OSR. Well, I don’t – but that’s mostly because in my mind OSR is synonymous with D&D clones. I’m just not that into OD&D or hacks thereof. Is there more to OSR than I realise? Anyone want to sell me on it?

Aside from some derailing around US-centric forum politics and the etymology of “Story Games”, there were a few interesting points raised:

  • Do D&D clones truly interorperate? Does everyone play D&D the same way?
  • What attracts Storygamers to the OSR and vice versa?
  • Why don’t my friends like D&D?

I. On Interoperability

The original G+ post is public, so I don’t see a problem with quoting some of it. Let’s start with the ideal of compatibility:

Adams Tower:

there’s an ideal of compatibility, that exists within the OSR, that does not exist within all RPG games. The intercompatibility exists because OSR game designers, and even more so OSR adventure and supplement designers, make an effort to be compatible with multiple OSR systems.

Ron Edwards:

I completely agree with you that there’s an ideal. I’m saying that’s exactly what it is, and that it functions at the level of subculture solidarity, not at the level of interacting game mechanics/experiences.

This point leads us to actual mechanical interoperability:

Ron Edwards:

I suggest that the systemic differences among many OSR games, even the retroclones, are so profound that they exceed the community ideal of compatibility, which then must be papered over by claims of some kind of homogeneity. The practice of mix-and-matching among the texts for your own table makes sense to me, but the claim that this is possible because the games are inherently compatible in system or fictional content is not.

Josh Fox:

is not OSR one of those things Wittgenstein would call a family resemblance? Sure, they may not be homogeneous, but you can recognise one when you see one, no? Or are you actually saying they’re so different that the term is literally just defined by someone saying it’s OSR?

Ron Edwards:

I don’t buy the “family resemblance.” I’m talking about hard-core mechanics, rewards, integration of parts, and play if you actually play them instead of having them sit on the table while you do something else.

I don’t feel confident to comment on Wittgenstein’s Familienähnlichkeit as it applies to the OSR, though there does appear to be some cross-communication here — certainly I think Josh’s point is upheld precisely because the OSR is an Ideal (q.v.).

Nevertheless the comment about actual integration of moving parts, intent and rewards is a good one. I think (as I said in the thread) the OSR games are like Linux distributions: they reflect the operability ideals of the designers, they’re essentially a diffuse package of commands that the distribution maintainer curates and forces to operate together according to what’s makes sense to them (though of course the OSR has no identifiable kernel maintained under a centralised authority).

II. D&D’s just this game, you know?

Several of my friends “don’t like D&D”. This could be for a number of reasons:

  1. It’s not what they grew up with. Some of us started on urban fantasy, others with alternatives to D&D (which in the UK included GW’s printing of RQ3, as well as WFRP).
  2. They don’t like “generic fantasy”, whatever that means — it could be the dungeon crawling that’s a turn-off, or it could be the sub-Tolkien melange of races and monsters in an overdeveloped faux medieval setting.
  3. They don’t like the system. It betrays a lot of the expectations we’ve built up around skills-driven RPGs (Storyteller system, GURPS, etc.). Often specifically they don’t like levels.

We excuse shortcomings in a lot of games we play on the grounds that the system is secondary to the group and the GM — yet some people have an overwhelming negative reaction to D&D when they’re simply indifferent to those other games. This reminds me of a throwaway comment from a podcast (I think it was from this Walking Eye podcast but I can’t be sure) that “people can’t discuss D&D and not lose their shit”. Quite likely that same fantatical culture that makes fans go to war over their favourite edition is also what turns other people off.

D&D is just another game, and can be turned to whatever campaign you care to run — but at the same time, while problems 1 and 2 can be overcome by the right social group problem 3 is the real killer. This is the “system matters” caveat, but it’s less to do with the suitability of the system in the abstract and more about the expectations players have. Vancian Casting, Armour Class, all of these just seem alien compared to the modern alternatives.

The good news is that OSR is like Linux, in that you can identify the parts you don’t like and strip them out, change their context or even make them interact differently.

III. The Moving Parts

RPG design in the last 20 years has trained us to expect point buy, skills at the forefront, homogeneous competency systems (e.g. to-hit, skills, magic, saving throws rolled into one system). I’m thinking about Storyteller, GURPS, ORE, Unisystem, etc — non-traditional designs (Apocalypse World)

Compared to those, the OSR model seems a bit arbitrary with lots of sub-systems that don’t interoperate much.

Let’s consider the building blocks of D&D-like retroclones:

  • Attributes
  • Armour Class
  • Hit Points
  • Attack Bonus
  • Damage
  • Saving Throws
  • Levels
  • Experience points
  • Spells, Skills, Feats

They flow together like this:


This is a functional mix, but a bit random in places. For example:

  • Attributes look like they’re central to the character, but really they’re a gateway to the class you want (whether it’s implied or explicit, “fighter” is synonymous with a high STR). They give a one-time bonus and then they’re overtaken by the character class benefits. Yes, sometimes they’re used for freeforming skill checks, but that’s about as relevant as they get. And players tend not to like this random gatekeeping anyway — leading to point-buy and UA’s absurd Method V character generation.
  • Saving Throws are just weird and arbitrary. They’re for negotiating hazards like spells and monster effects and traps, but not hazards like getting thumped with a mace — that’s what AC is for, no rolling there. The interpretations vary across different OSR games, and very few of them are geared to be used solely on GM judgement (Kevin Crawfords games are one exception). Mostly they just mesh with some very specific spell effects — that’s a big chunk of the character sheet to be held on a maybe.
  • Of course HP, AC, Damage and To-Hit are throwbacks from the original wargame — a lot of mechanics devoted to a very specific context. No wonder players think D&D is only about killing monsters…
  • …and finally, there’s the big messy bucket that is Spells, Skills, and Feats. There are two kinds of object in this bucket: some are very specific, self-contained rules with save conditions and consequences of use, and the rest are totally freeform and negotiated for whatever situation the player chooses to use them in.

So yes, pretty incoherent. It’s not surprising that a lot of people have looked at these elements and hacked the way they work together. That said, most of the OSR types take the base framework of stats, hit dice etc. and plaster on a skill and/or magic system of their choice and declare it their niche in the OSR.

I find it most useful to consider three areas:

  1. Statements of Nature. These are the starting points for the character, before the player chooses the class. In the OSR framework they’re attributes, and frankly if that’s all the attributes do, why not just skip them? There are a lot more useful things that can go in this area like statements of ambition, backstory and so on.
  2. Statements of Competency. This is where all the loose statements of competency go — from the specific but contextual Feats and Spells to the broad and qualitative Skills system (whatever that is). I put the Saving Throws in here too.
  3. Derived stats for Procedures. These are all about achieving a core activity, and the core activity is, make no mistake, fighting.



Now we know where everything goes and how the player views it, the hacking can begin. Some approaches:

  • Make the skills and magic system operate on the same currency (e.g. GURPS)
  • Ditch Attributes in favour of specific bonuses (True20/Blue Rose)
  • Make skills, not levels directly affect combat bonuses (most skill-based RPGs, e.g. RuneQuest)
  • Make AC and Damage interact (RuneQuest)
  • Make Saving Throws and Damage interact (True20/Blue Rose)
  • Make Saving Throws part of combat and only players roll the dice (Cinematic Unisystem)
  • Ditch levels in favour of incremental class gains for XP (pretty much every game that isn’t a D&D clone)
  • Roll different dice, or otherwise change the scale of dice effects (e.g. Scarlet Heroes’ handling of damage)


A lot of these changes are pretty disruptive; it’s fair to say that the end product you get will not be D&D, and that’s a good thing. Personally I like the idea of D&D’s isolated but promenent sub-systems that direct the players to a certain kind of game — e.g. I’ve always liked the way LotFP explicityly rewards gaining treasure and killing monsters in that order, and nothing else. I strongly feel that Beyond the Wall needs its Old School combat features to focus on the dangerous nature of going beyond the village; at the same time it’s lacking some internal community mechanics, and that’s a layer I intend to consider shortly.

Closing Remarks

There’s probably something to be said about what drives people to write Fantasy Heartbreakers, but I’m running out of steam. For now let’s sum up:

  • D&D’s system is a messy collection of systems that don’t work together seamlessly.
  • More to the point, the various interpretations of D&D vary in subtle ways that require you to think on the intent of the designer or hacker, and depending on how transparent the designer has made their thoughts this may be easy or difficult.
  • D&D’s different subsystems are simple enough to understand individually, and once you do that you can partition them to do different jobs in your own game — and this is clearly what some OSR games do.
  • There are good reasons to keep your flavour of OSR D&D looking like D&D, namely interoperability and familiarity with the various moving parts.

That’s it for now. I’ll write about my own hacks for Beyond the Wall some other time.


  1. Mo

    Now I don’t have any problem with D&D, I used to play absolutely loads of it, and when I was a kid I did try to improve it – as in the multi-classed experience handling, dexterity classes and so on in Hezbollah! But I suppose my conceptual problem with the OSR is that I can’t see why anyone would bother spending the time and energy on that sort of thing once they’d experienced the wide variety of other types of game that are around these days. If I wanted to run a “generic fantasy” type thing now I’d either just use some familiar version of the original and be happy to put up with its flaws, or else use something completely different like I dunno, Hillfolk, rather than devote myself to polishing a turdgilding a lily.

    But I suppose that is just the same unedifying question as ‘why does anyone bother doing [any thing that I can’t personally see the point of]’. 🙂

  2. Pingback: My Reply to “Who’s Afraid Of The OSR?” | Basic Fantasy RPG Blog
  3. Pingback: Who’s Afraid of the OSR, part 2: clarifications |