System Typing

[Ben Lehman]

I have an idea.
Now, as I begin to lay this idea out, I think people are going to instantly relate it to another set of three categories which have to do with RPG theory.  I'd just like to suggest that there are lots of subtleties going on here and, although such a relationship may prove -- in the end -- fruitful, it is best to consider this idea on its own merits first, because I don't think that these categories have a 1-1 GNS correspondence.

There are three types of RPG systems.  These systems are, in fact, *entirely different things*, and I think a lot of confusion and anger is based around mistaking one type for another, and thus making very unpleasant blanket statements like "Game Balance is an Illusion," "the Heart of Any Good Game System Is a Fortune Mechanic," "The System Cannot Apportion Credibility to Itself," or any other untrue sort of thing.

These three types are classified along the lines of how they relate to the shared imagined space.

I will call the first class of systems Physical Systems, or perhaps Natural Law Systems.  These systems are those in which the system provides resolution roughly equivalent to the natural laws of the world.  How fast does this truck go?  What is the chance of this bullet hitting?  Etc.  Note that these systems are not necessarily "realistic," realism being a subset of these systems.  For instance, Teenagers From Outer Space and Marvel Superheroes (FASERIP version) are two games which have this type of system and are hardly realistic at all.  Other examples are Silhouette, GURPS, Nobilis, etc.

The important thing about Natural Law Systems is what the system does -- it literally describes nearly every action or occurence within the game world, or at least has mechanics which can be extrapolated for that.

The Second class of systems I will call, perhaps, Metaphysical Systems, or perhaps Entertainment Systems, to avoid confusion with the presence of a game metaphysic or "non-realistic" setting content.  Actually, they really need a better name.  The point is that the system structure does not dictate how objects in the shared imagined space interact with each other, but rather how players at the table interact with each other and the shared imagined space.  These systems are often considerably more focused, in terms of mechanics, than a Natural Law System, and they tend to employ considerably more in the way of metagame elements.  Examples of this sort of system include Dogs in the Vinyard, Universalis, the Pool, Basic Dungeons and Dragons, and so on.

The important thing about Metaphysical Systems is that they dictate what is going on at an out of game level, and literally control the shared imagined space largely by controlling the interactions between players.

The Third class (which is possibly two classes) I call Mixed or Ambiguous systems.  Mixed systems take the two different types together, often (in the bad cases) somewhat ambiguous about what goes where.  The most common type of mixed system is tacking a "hero point" or "dramatic bonus" system onto a Natural Law Chassis.  Examples of Mixed Systems include Sorcerer, The Riddle of Steel, d6 Star Wars, 3rd Edition D&D, and so forth.  Ambiguous systems are rather similar to Mixed Systems, but are completely ambiguous as to whether their resolution is in terms of players at the table, making decisions about the shared imagined space, or about the interactions of the shared imagined space with itself.  The only system that comes to mind in this regard is Amber Diceless, which is blatantly nonspecific about say, whether Benedict is a good-sword fighter or whether he is symbolically dominant in the sphere physical conflict.

Any thoughts on these distinctions?  Is there some other type along this distinction?  How do these related to GNS?  Has someone else made this point in the past, and I missed it?

And, most importantly:  What are the design considerations for each type of system, and what to the following things mean in the context of each...

1) Scene vs. Task Resolution
2) Whether or not the System can be apportioned Credibility
3) The presence, or absence, of Fortune mechanics.
4) Game balance.

Thanks--
--Ben

[Eero Tuovinen]

I find at first glance that to my mind the Natural Law systems aren't really roleplaying game systems in any strong sense. They are just a bunch of guidelines about the explored setting for the GM to apply, more akin to physical models used in engineering.

Now, this opinion requires that you agree on the following: any die roll mechanics, when they are rolled, turn structures and such are actually a mixed in Entertainment system, or pieces of one. To say it in another way, the only pure NL systems to my mind are not games at all, but different kinds of simulation systems.

The above holds true with a great majority of systems; there are actually a couple of games which genuinely think that their dice mechanics simulate something. These are usually characterised by having a zillion different rules for different situations of die rolling. But for any game which uses just a couple of different rolls the conflict systems and such are really not a part of the NL system, but of the tagged on Entertainment system.

I don't have any special arguments for my opinion roght now, I just see it as natural. Anything at all in the game that can be removed by manipulating the game world (like classes in the new D&D) is really just formalized language conserning the game world. When the game world changes, these will change too. How can there be actual game rules connected to such phenomena? They aren't game rules, they are the things themselves. The game rules are always between the players, and the simulations are part of the world they explore. This is typified in a great majority of cases where the GM "cheats" or breaks the rules. Typically such a case happens when the common vision about a given game world thing clashes with how the rules concerning it work. In those cases the GM doesn't really change the rules, but the phenomenon explored.

[BPetroff93]

Hold the phone for a second! Doesn't this sound similar to GNS design applied to game rules?  I mean, natural law systems sounds just like a Sim design priority.  Metaphysical could be either Gamist or Narritist design priorities, depending on which main road the group takes.  A Mixed system is a Hybrid CA design.

[lumpley]

Well Ben, I agree with you...

Except that I'd say that the distinction exists at the particular rule level, not the ruleset level - so it's a particular rule that's Natural Law or Entertainment or ambiguous, not a whole game.

I'd also say that Natural Law rules are, y'know, a kind of Entertainment rule, just as Eero says.  They control what goes on between the players by controlling in-game causality.

Can I call Entertainment rules "Credibility rules" instead?

I've been thinking about this very subject a lot, especially about the balance between Natural Law rules and Credibility rules in any given game.  For a while I was taking a pretty hard line: a game can exist, be playable, be fun, with no Natural Law rules a'tall, no point where the Credibility rules refer to the things and events in-game.  I held up my Nighttime Animals game as an example to myself.

(Here's how the Nighttime Animals resolution works, for those of you who don't know: you say what your character does, like "I run along the power lines over the river."  I say what the danger is, like "the danger is that the wind will knock you off and you'll fall in."  I hold out a coin.  If you hold out a bigger coin, the danger doesn't come true; if you hold out a smaller coin, it does.  We swap coins.  There's no relationship at all between your character's capabilities and the coins you start with, noplace where the coins follow any kind of in-game Natural Law.)

But now I think the opposite.  Every RPG's gotta have some Natural Law rulage, and the RPG with no "naked" Cred rules is the one that's possible.  The Natural Law rule in the Nighttime Animals is hiding in the danger: When I say what the danger is, I have to follow in-game causality.  When the danger comes true, the consequences have to follow in-game causality.  I can't introduce nonsense dangers or have them come true nonsensically without breaking the game.

So Ben.  Is there something at the System level that you're seeing and I'm not, that lets us characterize a whole ruleset as Natural Law vs Credibility, or are you cool with applying the distinction to individual rules instead?  And does it make sense to you to say that all rules are Credibility rules, and Natural Law rules have an added characteristic, which is that they're about in-game causality?

Oh and I agree ab-so-lutely that this has no direct relationship with GNS.

-Vincent

[John Kim]

Hold the phone for a second! Doesn't this sound similar to GNS design applied to game rules?  I mean, natural law systems sounds just like a Sim design priority.  Metaphysical could be either Gamist or Narritist design priorities, depending on which main road the group takes.  A Mixed system is a Hybrid CA design.

That is a common association which people make, but in GNS as it is currently defined, it has nothing to do with the GNS definitions.  i.e. A system can be purely "Metaphysical" and still be GNS Simulationist.  Conversely, there is nothing about a "Natural Law" system which makes it inherently GNS Simulationist.  

This is instead similar to the http://www.darkshire.net/~jhkim/rpg/theory/threefold/">Threefold Model split.  The definition of "Natural Law" is similar to rgfa Threefold Simulationism, though not the same.  

Now, this opinion requires that you agree on the following: any die roll mechanics, when they are rolled, turn structures and such are actually a mixed in Entertainment system, or pieces of one. To say it in another way, the only pure NL systems to my mind are not games at all, but different kinds of simulation systems.  

The above holds true with a great majority of systems; there are actually a couple of games which genuinely think that their dice mechanics simulate something. These are usually characterised by having a zillion different rules for different situations of die rolling. But for any game which uses just a couple of different rolls the conflict systems and such are really not a part of the NL system, but of the tagged on Entertainment system.  

If I'm not mistaken, this is the concept that "A simulation has to be incredibly complicated and accurate or it isn't a simulation."  Which is nonsense.  All simulations have inaccuracies as well as deliberate simplifications, even state-of-the-art scientific simulations.  Indeed, I would say that over-complexity often makes a system less accurate.  

However, I agree that RPG systems aren't purely "Natural Law" (or "in-game-world") -- because such a system wouldn't have any provision for who decides what in the real world.  i.e. There's no provision for someone being the player of a particular character, or someone being "GM".  However, it is perfectly valid to have a purely in-game-world rule that, say, a spell has a 50% chance of success.  Further, probabilities can be expressed as target numbers on 3d6 just as surely as percentages or any other measure of probability.  A "Natural Law" rule can also appeal to judgement calls as long as the judgement is on the basis of in-game causes.  

Threefold Simulationism can include more than "Natural Law" rules.  It is defined as not allowing in-game results to be explicitly based on meta-game issues.  But it can have purely meta-game rules, like who decides what, for example.

[Jason Lee]

Though I don't think you can make a direct correlation to GNS, I think you can to the whole model.

DISCLAIMER:  I'm ignoring Sim, because I do that.

Super-Duper Jargon Engine Switch On!!!  Go Go!!!  (Insert transformation noise)

It seems to me that natural law techniques are concerned with maintaining Fidelity of individual exploration elements, whereas metaphysical techniques are concerned with preserving creative agenda.

Taken in this context, I agree that game systems are always a mix, though the mix may vary.

If you buy my take on this, then that puts game balance in the category of metaphysical, as it is primarily concerned with creative agenda (a player need for a level playing field).  It then follows that metaphysical mechanics would include player control mechanics, conflict resolution, and possibly scene framing.

[BPetroff93]

I was afraid of everyone saying that.  I wasn't so much saying this IS GNS, but rather this seems to be how GNS design priorites influence system design.

[Eero Tuovinen]

If I'm not mistaken, this is the concept that "A simulation has to be incredibly complicated and accurate or it isn't a simulation."  Which is nonsense.  All simulations have inaccuracies as well as deliberate simplifications, even state-of-the-art scientific simulations.  Indeed, I would say that over-complexity often makes a system less accurate.  

Actually, what I was getting at here was that almost no roleplaying game claims that the system simulates the world explored. It's tacitly assumed that the world is a much more complex place than the rules make them. This is a main reason for "GM is always right" in many games; after all, if our rules are not perfect, we need the GM to override them sometimes. It's entirely possible for a game to claim that it's rules are the paramount definition of the game world, but it's rare; usually the paramount thing are sensible real world physics and such.

I'm familiar with the three-fold discussions, and do not hold any opinion similar to the quoted fallacy. Instead I was pointing out that for any Natural Law system there has to be a system bit of the actual game assigning that Natural Law system some relevant role in the game. If it just sits there with no claim to authority, it's not part of the game. There has to be some rules about how "this and this simulation is used to decide what's going on, except when the GM says so". The point is that this is essentially the same as assigning that authority to one of the players or a pure die roll without any in-game meaning whatever. There is no special role for the Natural Law systems, they are needed as an infodumb about the game world. Any associated game rules are Entertainment systems. This is essentially what lumpley said much better.

An illustrating example, Call of Cthulhu's skill rules. They are conseivably a simulation, but consider: I can interpret the rules as assigning success rights to characters in certain situations instead. My character with 90% skill in swimming is almost assured success in any task involving swimming. There need not be any Natural Law fact in place here, it's all game rules in abstract. It just so happens that the same 90% has a double role in the Natural Law system of the game, indicating that my character has swimming experience. This is incidental to the game rules themselves, we might as well decide after ten successes in the skill that my character is a good swimmer, so he must have some background. The latter way avoids using the Natural Law system concordant with the game rules, while the former uses the system to define in-world meanings to my character's skills.

So I'm still holding to the original point, that there is no Natural Law systems as Ben conceived them. Or rather they exist, but don't have any direct connection to roleplaying game rules.

I'm agreeing with lumpley in what he said. A great analysis.

[BPetroff93]

Are you saying that a Natural Law system must be attempting to simulate "reality" or just the reality of the game world?

[BPetroff93]

Furthermore, it is also my understanding, from reading the original post, that "system" is being used to describe the entire game's system, not a subsystem such as combat, magic, or reward.  Is this correct?

[FredGarber]

Maybe they fit better somewhere else in the Big Picture?

Are there rules to describe In-Game conflict resolution versus Meta-Game conflict resolution?

For example, what's described here as a Natural Law system is a system in which the allocation of points on my sheet determine what I can do well.  My GURPS character cannot ninja-climb up the wall with any good chance of success because he has a low climb skill.   He can cook very well, because that is where I spent some points.  I can spend my reward points to increase my character's ability to climb, cook, or swim. But I can't really spend my reward points (XP) to affect things outside my character.  These games only let me spend the reward for challenges on things that affect my character (with a few exceptions.)

But what he describes as a Metaphysical system is more like the Pool, where how I can affect the world is more limited by the Social Contract that formed my initial Character description.  I can spend my reward points either now (MoV) to control the game world as a Player. In other words, I might be playing an inept medieval wizard, but use my MoV to describe a magic sword that might aid my fellow player who is playing a swordsman.

(Note that taking the die in the Pool is like delaying my MoV from this challenge until some time when I feel that it is more important to make a MoV and can risk more dice.  Note that spending dice on Traits increases the chance that I have a MoV, but doesn't really increase how effective I am.  I could use all of my MoV to craft an bumbling fool detective character who muddles through the adventure and "failing" to succeed at anything, but I can give the rest of the party "my" clues through my MoV.)

In summary, I think the question simplifies to which "conflict resolution" do the rules tend to describe. Is it conflict between the players (and their differing ideas outside the Shared Imaginary Space) or conflicts between the characters (that is, conflicts inside the SIS).

[John Kim]

If I'm not mistaken, this is the concept that "A simulation has to be incredibly complicated and accurate or it isn't a simulation."  Which is nonsense.  All simulations have inaccuracies as well as deliberate simplifications, even state-of-the-art scientific simulations.  

Actually, what I was getting at here was that almost no roleplaying game claims that the system simulates the world explored. It's tacitly assumed that the world is a much more complex place than the rules make them.  This is a main reason for "GM is always right" in many games; after all, if our rules are not perfect, we need the GM to override them sometimes.  

But isn't this precisely claiming that a simulation by definition has to be perfectly accurate and detailed?  I mean, when I run a high-energy physics particle simulation program, it is still true that the real detector is more complex than what is in my simulation.  Using your logic, since I assume that the real detector is more complex than my program, therefore I cannot claim that my program simulates the detector.  That's exactly the claim I am debunking.  

My counter would be that an approximation is still a simulation.  Moreover, it is possible for rules to be designed with GM moderation in mind -- indeed, this is generally true of tabletop RPGs.  Let's consider a game set in the real world, and we want a mechanic for, say, jumping.  Let me consider three cases:
1) An impossibly-complex rules mechanic which gives a good answer for 100% of the cases.
2) A complex rules mechanic which gives a good answer for 99.5% of the cases, taking into account all different types of species, scales, and terrains.  
3) A simple rules mechanic which gives a good answer for most cases (maybe 95% -- works pretty well for humans and similar mammals), but gives bad results and should be overruled by the GM in some uncommon cases.  

I'm saying that #1 doesn't exist, and that both #2 and #3 are valid choices for a simulation.  #3 is not ceasing to be a simulation -- it is just a less accurate simulation which assumes human input.  Simulations are not required to be purely digital mathematical equations.  For example, http://home.freeuk.com/henridecat/">free kriegspiel is a simulation which relies on human judgement.  

Now, I agree that for something to be a true "game" it needs more than just in-game simulation.  But on the other hand, one can make a decent case that typical RPGs are not "games" anyway -- since they lack winners and losers and victory conditions.

[Eero Tuovinen]

But isn't this precisely claiming that a simulation by definition has to be perfectly accurate and detailed?

Really, believe me, I believe you. I agree on that count. What you are writing about (using Ben's parlance) is the Natural Law system. It does not need to be very real to the game world simulated. Some times it's more, sometimes it's less, and even a child should know that you cannot "simulate" anything perfectly without using as much detail as the thing to be simulated. Thus to simulate a world you'll need a wolrd, at least, as your medium.

That, however, is not the thing I'm arguing here. I'm talking about the relationship of the NL system to the Entertainment system, not about the world's relation to the NL system.

My counter would be that an approximation is still a simulation.  Moreover, it is possible for rules to be designed with GM moderation in mind -- indeed, this is generally true of tabletop RPGs.  Let's consider a game set in the real world, and we want a mechanic for, say, jumping.  Let me consider three cases:
1) An impossibly-complex rules mechanic which gives a good answer for 100% of the cases.
2) A complex rules mechanic which gives a good answer for 99.5% of the cases, taking into account all different types of species, scales, and terrains.  
3) A simple rules mechanic which gives a good answer for most cases (maybe 95% -- works pretty well for humans and similar mammals), but gives bad results and should be overruled by the GM in some uncommon cases.  

Indeed so. Now, consider: the above rules systems are Entertainment systems with the added claim of giving out information about the game world. That's a fine way of doing things, sensible to human mind. Now, consider system number four:

4) A rules mechanic apportioning me credibility towards claims of jumping by my character.

You'll note that this could be any of the above Natural Law systems, but it could as well be something with no Natural Law behind it at all. My original point was that for any NL system to be actually relevant for roleplaying there has to be a nigh independent Entertainment system that applies the NL system for something. That system is something that says that "this here simulation is what were using to decide if your character can make the jump". Thus NL systems are not actually roleplaying games rules, but just rules about the game world, implemented for the game by an added layer of mechanics. This layer could be simply "this we use" or "this we use until GM says otherwise" or even "this we use, but you can override it with drama points".

I agree with lumpley in that we actually do need a kind of NL system for the games, but that system is not anything conveniently deemed game rules. The minimal expection for the NL system is world adherence in narration; the things claimed by the players have to be in harmony with the game world they happen in, as the world is only existing in relation to those actions. The actions are by definition in accordance with the NL system of whatever psychology controls the causal relations people keep to in narration. When we narrate a billiard ball first being hit and then moving we are applying this most basic NL system.

So, to recap, there are two kinds of rules systems, and roleplaying theory really concerns itself only with the Entertainment type and it's relation to the NL type. Every game has one of each in some form.

Now, I agree that for something to be a true "game" it needs more than just in-game simulation.  But on the other hand, one can make a decent case that typical RPGs are not "games" anyway -- since they lack winners and losers and victory conditions.

Well, that was in essence what I was claiming. There need be the Entertainment system that references the NL system.

[John Kim]

My original point was that for any NL system to be actually relevant for roleplaying there has to be a nigh independent Entertainment system that applies the NL system for something. That system is something that says that "this here simulation is what were using to decide if your character can make the jump". Thus NL systems are not actually roleplaying games rules, but just rules about the game world, implemented for the game by an added layer of mechanics. This layer could be simply "this we use" or "this we use until GM says otherwise" or even "this we use, but you can override it with drama points".

I agree with lumpley in that we actually do need a kind of NL system for the games, but that system is not anything conveniently deemed game rules.  

OK, I think I understand you here and we're saying nearly the same thing.  However, I have a quibble over semantics.  If there is a mechanic which says how many meters a Strength 17 character can jump, that is commonly called a "game rule".  You can make a case that theoretically it is a separate system from true instructions to the players, but you're deviating sharply with language which nearly all RPG players use.  

Now, I agree that for something to be a true "game" it needs more than just in-game simulation.  But on the other hand, one can make a decent case that typical RPGs are not "games" anyway -- since they lack winners and losers and victory conditions.

Well, that was in essence what I was claiming. There need be the Entertainment system that references the NL system.

Yeah, here we're agreed.  Note that the rgfa Threefold Simulationism allows that there is a meta-game level of rules -- i.e. it includes both "Natural Law" subsystems and pure-metagame "Entertainment" systems, but not systems which cross and designated non-in-game-caused in-game effects.

[Eero Tuovinen]

OK, I think I understand you here and we're saying nearly the same thing.  However, I have a quibble over semantics.  If there is a mechanic which says how many meters a Strength 17 character can jump, that is commonly called a "game rule".  You can make a case that theoretically it is a separate system from true instructions to the players, but you're deviating sharply with language which nearly all RPG players use.  

Certainly, but consider the thread in it's entirety. I'm not talking about my own parlance, I'm analyzing Ben's. He offered the distinction between the two types of systems, and I started arguing for them being really two different levels, both needed for the games.

In real life it's quite true that both of these are called game rules. That's not an useful fact for a thread concentrating on the difference between the types of rules, though. I chose to call only Entertainment systems game rules because there's not anything really gamey in the NL systems without the entertainment. They are just relatively simple notions about how such and such phenomenon can be codified with simple addition of integers.

Ben: any further thoughts on the matter? Seems we are all essentially agreeing here.