Thursday, February 25, 2016

OSR-Style Challenges: "Rulings Not Rules" is Insufficient

I was a Pathfinder guy who got inspired by blogs, and then spent some time trying to figure out what the fuck the OSR was.  I read stuff, like Matt Finch's Quick Primer for Old School Gaming, which is where I first heard about Rulings Not Rules.

The idea is that the OSR encourages a sort of innovative, ad-hoc gameplay where players are always innovating and solving problems with outside-of-the-box solutions.  They're thinking with their heads, not their character sheets.

But saying Rulings Not Rules is merely a description of the system, which is only a small chunk of what actually contributes to gameplay.

What Contributes to Gameplay

  1. The system.
  2. The adventure.
  3. The DM.
  4. The players.
The DM and the players could be bundled together, because the DM operates by a social contract, sort of like a charter.

Anyway, it's not enough to have a system that allows for rulings and improvisation.  If you want OSR-style gameplay, you need to encourage/allow it at all levels.

System Level

Like Mr. Finch says, this is about getting players to stop thinking with their character sheets.  (This is why skill lists are potentially so poisonous--players thinking about solutions sometimes start and end by looking at their skill list as if it were a list of permissions.)

(1) And to do that, you need an incomplete system.  You need to have room for rulings, and that means that there have to be gaps between the rules.  To put it another way, if I wrote up a game system that included two pages of rules on how to attack tiny animals in your stomach, I've codified the acceptable options and excluded more esoteric solutions.  (I've also complicated the game by introducing a fiddly and highly-situational subset of rules.)

If a players is familiar with the game system, they'll think back to what they know about the rules as a first resort.  Only when they've exhausted everything they remember from the How To Attack Tiny Animals In Your Stomach page, will they start to innovate.

For an example of a more complete skill system, all you need to do is look at the skill descriptions from 3.5th edition.  The more complete a ruleset is, the more tempting (and valid) it is to say "well, it's not covered in the rules, so you can't do it".  Or worse "this is covered in the rules, and if we add up all the situational modifiers, you will do so at a -14 penalty even though I personally agree that this task shouldn't be that difficult".

This is why I like running games without Perception checks, Find Trap checks, and social skills.  It leaves more room for player innovation.

(2) You also need a system that supports rulings.  There are two parts to this.

First, the system cannot have too much interdependence between the moving pieces.  Some mechanics are isolated (XP) while others touch on many other mechanics (Ability Scores).  The more interconnected a mechanic is, the more knock-on effects you'll have when you modify it.  If you want to just make a quick ruling and get on with your game, you usually want to make sure that your quick ruling won't have any unforeseen consequences.  For example: 

DM: "Alright, you all manage to tread water for 18 hours, but the act was so exhausting that you've all lost all of your healing surges."

Player: "Wait, I can't use any of my class abilities without any healing surges.  In fact, I turn into a pumpkin without any healing surges."

DM: "Well, we've already established that exhaustion drains healing surges.  That's been a house rule for months.  And treading water for 18 hours is definitely exhausting."

Player: "But that was before I picked up the Pumpkin King prestige class.  It's unfair now."

A little caricatured, but you get the point.

Second, the system needs to have simple ways to adjudicate rulings.

My first resort is to just ask a player to roll under the most relevant ability score. 
  • It's fast.  
  • The player already knows what I'm talking about because their ability scores are written down right at the top of their character sheet.  It's an associated mechanic, so it makes intuitive sense.
If it doesn't seem tied to any particular aspect of their character (i.e. it's entirely luck-based or dependent on some external variable that the character has no control over), I usually just ask for an X-in-6 roll, which I make up on the spot.  I like using a six-sider for these, because even a 1-in-6 chance is likely enough that it'll happen every once in a while.

For more extended efforts, I like some variation of "you need to get X successes before some other limit is hit".  But I only use these extended rulings very rarely.

Bad rulings are ones that are slow or confusing.  But the worse rulings are the ones that are ultimately unsatisfying, in the sense that they don't give results (or chances of results) that mesh with the player's expectations of how the world works.  If your make some rulings, and the consequence is that halflings are more intimidating than orcs, that's (probably) a bad ruling.  If you make some rulings, and the consequence is that even the most untrained peasant has a 90% chance to track anything, that's a bad ruling.  Just as players use common sense to come up with stuff that requires a ruling, use common sense to make rulings (as opposed to precedent or some other analogous rule you saw somewhere else).

Adventure Level

You also need to give players problems that are best solved through innovation.

If you give them a problem with three orcs in it, they're probably going to solve it through an already-established method: initiative and attack rolls.

Here are some good examples of OSR-style problems.
  • Get over this moat.  It's full of crocodiles.  (I think I first heard this example from Zak. S, and it's been stuck in my head ever since.)
  • There's a circle of mushrooms with a girl inside it.  Everything inside the circle of mushrooms will do everything in their power to get more people inside the circle (no save).  The girl is already their thrall.
  • There's a tiny octopus inside your stomach and it's biting you.
  • There's a bowl built into the ground.  It's lined with gold but full of acid.  (From ASE.)
  • There's a smooth glass sphere, 100' high, with an opening at the top.  It doesn't roll easily.  Inside is something you want.  (From some LotFP product, I forget which one.)
  • The bad guy cannot be hurt by any weapon forged by mortal hands.
  • This glass sphere (3' in diameter) is filled with gems and horrible undead snakes.
  • Pretty much all of the dungeons that +Chris McDowall writes.  He's like a laser pointer when it comes to writing interesting problems.
Writing a good OSR-style problem is tougher than it sounds.  It needs to be something that. . .
  • has no easy solution.
  • has many difficult solutions.
  • requires no special tools (e.g. unique spells, plot devices).
  • can be solved with common sense (as opposed to system knowledge or setting lore).
  • isn't solvable through some ability someone has on their character sheet.  Or at least, it isn't preferentially solvable.  I'm okay with players attacking the sphinx (a risky undertaking) if they can't figure out the riddle, because risky-but-obvious can be a solution, too.
The fun thing about OSR-style problems is that they often require rulings-not-rules.  (Try to solve the tiny octopus in 4e with RAW.)  So there's a benefit in having a system that's easily hackable.

But at the same time, OSR-style problems aren't dependent on system.  You could plop them into any system and then players will still have to innovate to solve them (and probably have a lot of fun in the process).

It's also important to give your players OSR-style tools.  (This is an idea I've half-articulated before.)

The anti-examples of this are going to be things like a sword +1, or a cloak that gives you +4 to stealth.  Anything that gives you a numeric bonus is not an OSR-style tool.  Anything that gives you a known, established ability is not an OSR-style tool (like a potion of healing).

These are tools that allow for innovative problem solving.  They stretch the brain.  Good examples include:
  • Immovable Rod.
  • Polyjuice Potion.
  • Ring of Cadaverous-but-Reversible Sleep.
  • Love Potion.
  • Psychic Paper.
  • Sovereign Glue.
  • Cursed Wand of Enlarge, only enlarges one part of an object.
  • Bag of Infinite Rats.
  • Some of these items.
  • And some of these items, too.
I especially like to make these types of items single-use or limited-use.  It prevents the item from becoming a known solution to an established problem (which is pretty much the antithesis of OSR-style problem solving).

DM Level

There's two things you gotta do.

First, talk to your players like adults.  Tell them that this game will have problems that aren't obviously solvable, and that some of these problems will have solutions that aren't on the character sheet. 

Actually, if you're dealing with complete newbies to tabletop RPGs, the less stuff that is on their character sheet, the better.  A level-0 funnel can help get new players thinking about common-sense solutions to problems.  Adding skills to the game after one or two sessions can also be a big help (if you ever add skills at all).

Second, you need to reward creativity when you see it.  When players ask you if something is possible, say yes.  (Or "yes, but".)  When you are devising a ruling for some ridiculous player shenanigans, lean in the player's favor.

I'm not advocating that you should allow stupid ideas to succeed, but solving an OSR-style problem is usually going to involve some kludgery, so be lenient when deciding how likely crocodiles are to eat a bomb disguised as a pig.

Player Level

I'm writing this article with the assumption that everyone enjoys the same types of game that I do.  This is not always true (unfortunately).  Talk to your DM and each other about your expectations.  Give feedback.

When it actually comes to solving these problems, I can't really help.  It's just you and your brain.  Here a few pointers, though.

  • Think about all of the resources at your disposal, including resources in other rooms.  
  • It helps to take notes.
  • Make the hireling do it.  
  • See if any of your magic items can do cool stuff if used in combination--sometimes the answer is spread across multiple peoples' inventories.  
  • Take it to someone who knows more about it.  
  • If it looks like it might do something horrible, pick it up on the way out.  
  • Come back later with the right tool.  
  • Experiment, experiment, experiment.  
  • Before you do anything, ask the DM lots of questions.  
  • Before you touch the dangerous parts, learn as much as you can about the non-dangerous parts.


  1. Sometimes skill checks in old school play can be "last resort". Ie. After the thief uses his magnets, ball bearings, talc powder, etc. to check for traps and doesn't find them, THEN she uses a find traps roll.

  2. Ooookaaayyy. Just a couple of tiny problems I have with these concepts.
    1) You wrote a bunch of fiddly rules about how to deal with tiny octopus in your stomach... That's exactly what you said not to do. Although the rule set is several lines, not several pages, you're still showing how practically difficult it is to avoid creating a bunch of situational sub-rules instead of just relying on rulings.

    2) The OTHER thing skills and skill checks do is allow you to handwave certain things in the game that either a) you don't want to bother spending time on, or b) you don't know enough about to play out. Like, I don't give a FUUUUUUCKKKKKK how the thief finds the trap. I'm interested in whether or not he did. So. Skill roll. I also don't care exactly how the Ranger tracks the owlbear through the forest, I want to know if he lost it's trail or not.

    Fairly generic skill list allow me to do all this quickly and easily, with a bit more granularity that "roll wisdom". The issue with using just stats is that everyone can do that... and if everyone is super, then no-one is.

    1. 1) I didn't create rules. I wrote down some ideas and said "these might be valid solutions", depending on the exact context. Chugging a bottle of whiskey might be sufficient. A shot of whiskey? A bottle of wine? An ounce of hashish oil? The exact mechanism to resolve this check is still up to the DM.

      What I did NOT do is create a big table of situational modifiers (-1 to the octopus's Constitution check for every ounce of whiskey drank or every 5 ounces of wine, -2 for every gram of hashish oil, stomach weasels get +4 to resist whiskey, etc). The actual rulings are still very much in the DM's domain (not mine to advise).

      Different DMs with rule it different ways. Some might make it a roll. Others might just say, "you drink enough whiskey to kill the octopus, but now you're shithouse drunk". It's up to the DM, not the system.

      And it serves a second purpose: it's a proof of concept that I'm not giving the player an unsolvable problem. When designing problems with non-obvious solutions, there is a temptation to give them things that they cannot solve: a crushing ceiling trap with no way out, a trail that they have no way of following. By writing down one or three *possible* solutions, I have reaffirmed that the players have ways to solve it. But those are just proof that solutions exist--there are many, many other ways the players could solve it. Maybe spinning in circles to get the octopus dizzy and then vomiting. Maybe turning your blood into tiny snakes.

    2. 2) Alright, we have different goals and expectations for gameplay. We could say that you're taking the gamist side, and I'm taking the simulationist side, but that's reductionistic and not very helpful. Or that your gameplay comes from character design and mine comes from environment description.

      I enjoy exactly the kind of game where it matters where the secret switch is hidden. I find it satisfying for a player to say "I look behind the moose head." and then find the secret switch. This is true whether I am a player or a DM. Because for me, that's where the game is.

      It's not a question of granularity or specificity-on-the-character sheet, it's a question of using common sense to interact with the game world, instead of numbers on a character sheet (some of which are more abstract than others).

      I would never use a Wis check to see if a player finds a trap or a Dex check to see if they disarm it. If it's a pit trap, you find it by poking the ground in front of it and disarm it by jumping over it (or walking around it).

      Trapfinding is easy to talk through--you don't have to abstract it down to a roll. You and the player can discuss things with natural language and common sense.

      Tracking an owlbear involves too many details spread across too much time--you pretty much have to abstract it down to a roll (or rolls).

      When I put a trap in a room, I want to test player ingenuity, not the numbers on their character sheet. (I wrote another essay about this over here:

      But that's just the kind of game that I enjoy, and the kind of game that I consider OSR-ish.

      There's nothing wrong with the game you described. Your game involves building a strong character sheet (or a strong party composition, with someone who has a good number attached to trapfinding and someone with a good number attached to tracking). But it's a different design goal than I have.

    3. Sorry I didn't get back to this sooner. I broadly agree with the concept of creating interesting challenges that characters can solve without referring to their character sheet. But.

      The thing is, the fundamental foundation of a role-playing game is that characters are represented by a set of core numbers. HP, AC, stats, saves, skills. They are the baseline that everything is built off of.

      You can (and definitely should) have lots of things that happen in-game that don't interact with these numbers, since the great innovation of the rpg is to move beyond the core numbers in a way that tabletop wargames don't, but ultimately, for something to fundamentally impact a character, it impacts those numbers.

      The octopus impacts those numbers via HP loss. Which is the gold standard of interaction with your numbers, but other effects in the game interact with other numbers. Stat loss, xp loss, bonuses to AC or saves. This represents a toolbox available to the DM, which they use to make up rules for interaction with those numbers.

      Because regardless of what you present here, you ARE making up rules. You are making them up all the time, on-the-fly. Every time you create a trap or a spell that has a mechanic associated with it, you're making up rules. Not making rulings. The rules might be really simple, and they might only last for the session or the encounter or the lifespan of the monster, but they are rules, and to make them up you use the toolbox of numbers and effects that you have available.

      Your policy here is just to push the rule-making off to the DM. And different DM's do it differently. Some just make it a CON roll. One makes up the rule that a bottle of brandy kills the octopus, but the character is now drunk. There is no such thing as a ruling. The DM is making up a rule and adjudicating its results.

      And this is the big issue I have. There is a practical tipping point where the toolbox is so sparse that the DM is constantly relying on these ad-hoc rules that are created and forgotten constantly. The DM becomes more vulnerable to "charisma bias" where the persuasive character gets away with all sorts of bullshit. It's a big part of why the OSR has a bad rap, reputationally speaking.

      This is not to say that it isn't possible to go wrong is the other direction, too. 3e (and Pathfinder) is a sterling example of how the codified rules can leave you just as lost as the trackless emptiness of no rules. If there is a rule associated with every interaction, you end up needing spreadsheets or actuarial tables to play the game. The Wyld or the Weaver, to steal iconography from Werewolf. No wonder the Wyrm went nuts and started killing everybody.

    4. Now, let's talk about 2.

      We really don't have different goals and expectations of gameplay. I'm just as happy to have the trap disarmed by the switch behind the moose head.

      This is all ground that's been covered exhaustively in other forums, but anyways.

      You are fundamentally miss-attributing what skills do in-game. They are mechanics for providing information, not for skipping things, although they can be used for that, if you want.

      If a character wants to play a skilled thief, I'm not going to ask him to learn all the different mechanisms that might potentially trap a door. He can check a door and the surrounding area and receive information about it. Specialist information that allows him to make informed decisions. In natural language with common sense. Likewise a "disarm traps" roll doesn't necessarily just turn off the trap. It gives specialist information about how such a trap could be disarmed, leaving the player to still make a decision about how to proceed.

      As a DM, it helps me to have options where I can give characters more information as required, and a mechanism for them to get specialist information that they as players might not be aware of, but that their characters would have a strong chance of knowing.

      It really comes down to what you consider "too many details spread across too much time." I don't enjoy too many details or spending too much time on things, but whether the things you describe are actual components of OSR-type games is pretty open to interpretation. My feeling is that much of what people consider OSR-esque is actually current interpretations, applied retroactively.

    5. Jeremy:

      You said "the fundamental foundation of a role-playing game is that characters are represented by a set of core numbers". I disagree. The fundamental foundation is that the player assumes a role, and that does not require numbers.

      Like, imagine the interaction that happens when a player picks up a key and uses it to unlock a door. Or when a player makes a promise to a king. No numbers are being exchanged. These things require no numbers, and yet they still feel like roleplaying.

      Strip an RPG of all roles and context (my yellow cone does 25 points of damage to your green cube) and it no longer feels like an RPG.

      Strip an RPG of all numbers and it still feels like an RPG. I remember playing a really simple RPG during a road trip. No character sheets, no numbers. Ambiguous things were settled by a coin flip (whether the next car to pass us had an even or odd number as their license plate). It was a little bit like the computer game Myst, which I had recently played. One player could talk to dead people and the other one could make homunculi. The players would inquire about a task and I would tell them if it was impossible, automatic, or dependent on a coin flip. It still felt like an RPG.

      And the octopus doesn't have to affect numbers. It could just be "you are charmed and can no longer willingly harm anything until you get rid of the octopus". That also feels OSR-as-fuck to me.

      Incorporating numbers into things just gives the DM and the players more established characters and objects to work with. Everyone can agree to the same expectations, and less needs to be assumed. It makes the world more concrete.


    6. And yes, the definition of a ruling is "a rule that the DM just made up". That's not a bad thing. Some of the rules that professional game designers write are arrived at after years of playtesting; others are just pulled out of their ass, like any good ruling. That doesn't mean that rulings are bad rules, just that they were invented quickly.

      A good rule/ruling has to be (a) fair and (b) fulfill genre expectations (e.g. strong guys should be better at strong guy stuff). And that's ALL.

      And it is often desireable to push that off onto the DM, because some things are best handled locally (rulings) rather than globally (rules). The octopus, for example, has too many unusual solutions. For the adventure writer to list all of them, and describe their chances of success would be tedious, bloated, and arbitrary (since they'd probably be ad-hoc rulings as well).

      And even if the designer wrote down five possible solutions, and mechanics for each one, players will still come up with new ways to solve the octopus thing. "The orc should grab the gnome by the ankles and whirl him around until he barfs up the octopus. What do I roll for that?" And that's great. That's the best part about an RPG, and that is the single characterstic that tabletop RPGs do better than computer RPGs. If you're missing that, you're missing the best feature of a tabletop RPG. That human-mediated flexibility.

      Does it really make a game less fun if the octopus problem is handled differently at one table versus another? Why must it be standardized across the country?

      And yes, the DM becomes more vulnerable to "charisma bias", but this is something shitty DMs do. The more a game depends on DM rulings, the more it depends on the DM. Good DMs make the game better, because they have the freedom to do so now. Bad DMs make the game worse, for the same reason. It is a dependency that computer games don't have.

      In a strict system (all rules, no rulings) like 4e, there is little room for the DM. A good DM in 4e feels very similar to a bad DM in 4e (at least when it comes to rules and rulings; there is still other space for good DMing, such as in descriptions and funny voices).

      In a looser system (some rules, many rulings) like LotFP, a lot more depends on the DM. But that's okay; the game has been dependent on players for a long time.

      If your DM is treating players preferentially, then that is shitty DMing. Talk to your DM about not being so shitty, or go get a new DM.

    7. Regarding #2:

      Give players all the information they need to make good decisions. Don't hide necessary information behind skill checks.

      "Oh, you failed your Knowledge roll. I guess I won't tell you about this interesting weakness that this monster has."

      "Oh, you failed your Perception roll. I guess you didn't notice that the blind orphan is now tailing you through the crowd, even though that's where the story is."

      "Oh, you failed your Disable Traps roll. I guess you have no idea how to open this lock without setting off the trap, so you'll just have to hope that whatever solution you come up with works (the player goes into this blind) or just take the damage (a form of HP tax)."

      And yes, the OSR is something that is nebulously defined, even on the best of days. Everyone has their own definition, and I have mine.

      But we do agree on one thing, though. "Much of [the OSR] is actually current interpretations." to which I say "yes!". It's not a recreation of actual games from the 70s/early 80s. (Many of those were kind of shitty, if the modules are at all representative.) The OSR is very much a new thing. It imports a lot of mechanics from the 70s (and so the rulebooks read very similarly), but the gameplay is pushed in a very different direction. So although the base mechanics are very similar, the instructions to the DM are very modern. Hence the "Renaissance".

    8. I know they've had enough of this conversation for years, but let me have an answer and some thoughts.

      Decisions, not rules
      Challenges, not pitfalls
      Skills give information, do not allow skipping challenges
      Use the idea of ​​opportunity cost
      Most of the time, do not test players for what they can do. Test them for how well informed they are

      Tests of knowledge, perception, disarmament, or charisma are not made by the player but by the master and are of the same type as reaction tests. They change either the information or the way the information is passed on to the players.

      So you are not tracking owlbear. It keeps the same distance of players, but when no one is seeing it, it comes close. That's a challenge. The skill test may inform players about this phenomenon.

      For traps, the skill test may inform how to disarm. But the disarmament still cost something. Like giving up some item, or occupying a character / hireling, or cost some time, at least. Everything can have an opportunity cost. Or consequences for the history. You have been tracked by some powerful mages in a ancient maze and notice that you have broken the closing mechanism of the library's door, and if the door close again, it can not open anymore. That kind of shit.

  3. Arnold, what do you think about thief/specialist skills and the Lamentations D6 skill system

    1. I like them except for the Find Traps/Doors thing.

  4. As a GM, how do you handle a situation where you have a player who doesn't want to figure out the riddle of the room, and wants to be able to let their character's skill set define whether or not they find the trap/secret door?

    When it comes to problems, I think the best piece of advice is simply to not prepare a solution. Whatever the players come up with to get that octopus out of their stomach is going to be 10 times better (or stupider) than anything I might prepare for!

  5. Ungeneralise skills. e.g.
    P: I check for traps, skill 2
    GM: How do you check? Your skill comes into play only when you give specifics.
    P: Ok, I look about the door frame and just lightly run my fingers over it.
    GM: Ok that's an easy check, so with your skill that allows you to be sure, after 1 minute, that there are no secret studs, movable switches, pressure plates or magical wards on the door frame.

  6. Sorry if this sounds abrupt, I don't want to argument.
    But @Patrick Mallah you're opposing 2 things that want to live together in peace: do not prepare VS whatever the players. Sketch a solution so there is one and you can give details and "hints" of sorts, but reward creative solutions.