Why do RPGs have rules?

[Thomas Shey]

On the bringing to a halt, to me one design choice is whether one wants to retain fail without setbacks. A benefit is that a player can attempt things where the cost of failure is simply not achieving the thing, rather than invoking additional badness. (Punished for trying.) On the other hand, fail without setbacks can feel static: it only obliquely drives momentum to fiction. Not necessarily in a bad way - for instance it can put it back on players to switch plans.

Anyway, I wondered what you take is on that?

Well, honestly, for general purpose usage I'm not sold that either fail-with-setbacks or neutral-fail are virtues. To use the Chill example again, a fail will give you motion forward, but minimal; it doesn't set you back but it doesn't advance your progress as much as a success or a critical will. Even a fumble is more of a mixed bag than an unmitigated disaster. For some sort of very simulationist oriented kinds of campaigns this is probably not a good model, and its used where it is simply because Chill is a monster-hunting game and there's always a risk of stalling out otherwise (its a different take on the principals used in Gumshoe in a way), but I came to very much appreciate it in play, and I suspect something like it would be virtuous in most types of campaigns.

 

[clearstream]

I think it is obvious why this sort of RPGing is experienced by many participants as centring the GM's conception of the fictional situation.

Under most game texts players are active here, describing what their characters do in a situation that they grasp. Under some texts they will say what rule their act falls within the scope of. In many cases game texts will set the target for success.

None of that is at odds with failure without setback. Here a GM need say nothing at all, but as referee encourage the upholding and carrying through of the lusory attitude. Perhaps explaining the lusory means - pointing to the rule.

The concern of yours that I have quoted does not turn on inclusion or otherwise of failure without setback.

 

[clearstream]

I used that example simply because it was clear. I do think unless the GM has thought through what consequences there are before the roll is made there's going to be a temptation to put his thumb on the scale; how well he'll resist that temptation various from GM to GM. Personally, I prefer that in most cases everyone know the general potential output states before it even comes up, which is one of the reasons I'm not a fan of games that put too much ad-hoc decision making in the GM's hands as a default rather than an occasional necessity.

Likewise. I suppose we could have different ideas of what counts as sufficient. In fact, I think that can vary within the flow of play.

 

[Thomas Shey]

What strikes me is the contrast between task resolution and intent resolution regardless of non-binary outcome. That is, in 5e the GM adjudicates the task at hand and states what the outcome is, positive, negative, or perhaps optionally a mixed outcome. In a lot of narrative play the player describes an intent or goal, or such is manifest in the situation, and then describes how they achieve it, or part of it, while the GM describes any problems or setbacks.

I haven't seen 5e played in the latter fashion.

Well, I'm not going to speak on that as I'm not particularly a fan of intent resolution, so my commentary would not be unbiased.

 

[Thomas Shey]

Likewise. I suppose we could have different ideas of what counts as sufficient. In fact, I think that can vary within the flow of play.

I was probably vaguer in that post than was useful. What I meant by that is that if a game is going to have, say, four results possible in a given task output, the player should have a fairly good idea what those results will be given other elements of the situation. There are going to be cases where they won't (because data about the situation is obscured for reasons in-game) or where the situation is sufficiently outside the normal play events that the only way they'll know is if the GM tells them (and under that circumstance, I'd much prefer he did), but to use a simple example, if someone is jumping across a small chasm, I'd much prefer that the standard mechanics tell me what the outcome possibilities are there and what category they land in.

 

[Lanefan]

You are saying that partial successes reduce the likelihood of a GM having to "manipulate things either behind the scenes or directly in order to keep things moving forward".

Your statement already assumes a bunch of things about gameplay namely:

• That there is a thing that needs to be moved forward.
• That the GM determines whether that thing is being moved or needs to get moving.

These situations can and do arise if (when) the players aren't driving the bus and things have ground to a relative standstill. And IME not all players are good bus-drivers.

But, a flexible system can equally-well handle either the GM, the players, or both being the drivers of the story.

• That the GM may at any point chooses to or chooses not to manipulate things behind the scenes.

Yeah, old-school as I am, I still think this represents bad GMing.

 

[AbdulAlhazred]

First off, apologies to all if this is ground already covered - I was called in on page 81 here and am not about to go over the previous 80 to find out what I've missed...

A hard rule (a.k.a. an actual rule) is explicit, either allowing or disallowing (or both) something(s) specific to happen in the play of a game.

"A king can move up to one square in any direction" is a hard rule: it allows a king to move one square and disallows it from moving any further. "Constitution 16 gives a bonus 3 hit points per character level" is a hard rule: it allows exactly that bonus and disallows any greater or lesser bonus. "A player on the attacking team may not precede the puck across the defensive team's blue line" is a hard rule: it disallows certain player positionings on the ice in certain situations.

Contrast this with "the objective of the game of hockey is to win via scoring more goals than the opposing team" which, while in the rulebook, neither specifically allows nor disallows anything. It's simply a guideline or suggestion as to how to play and not a rule at all until harder-coded into actual rules that define the win condition, define what a goal is and how it is scored, define the length of a game, and so on. And even then that guideline can still be ignored; a badly-overmatched team, for example, might take a different approach and instead of playing to try to win will instead play purely defensively, and try to lose by as small a margin as it can.

"Play to find out" falls into this same category - it neither specifically allows nor disallows anything and thus is not a rule: it's a guideline.

In hockey (and most other sports), each league has its own minor variants on an otherwise fairly-consistent set of rules; but in the end the league sets the (hard-coded) rules. Chess doesn't often even have that much variability, nor do most boardgames, card games, etc.

TTRPGs, however, are a different animal. Here, while the publisher might want to play the role of the league and set the hard rules*, there's these annoying things called GMs and players who - in the fine well-established tradition of RPG rules-kitbashing - want to make the game their own by taking those hard rules and in some cases putting them through a blender. And so, the role of the "league" falls on the GM (and players, maybe) at each table; to - to some degree - set the hard rules they're willing to play by and then play the game.

And some publishers realize this, and so rather than hard-coding lots of rules they give guidelines and suggestions, backed up by a far lesser amount of hard-coded rules. These guidelines are by nature a bit fuzzy, and that fuzziness makes them harder to change to any extent without (ntentionally or otherwise) changing the underlying foundation of the game as designed.

* - worth noting that most of these hard rules are dealing with quantifications and-or abstractions of those parts of the fiction that cannot be roleplayed in meat-space.

I think the above implies a far greater amount of "flaky GM whim" than I'm getting at. That said, IMO someone - be it the publisher, the GM, the table as a whole, or whoever - has to take on the role of "the league" and set the hard rules. And if all the publisher gives you is guidelines and expects them to be taken as hard rules, that doesn't seem to provide much help with the nitty-gritty of sorting those fictional abstractions; instead it shuts that responsibility on to some combination of the GM and players and asks them to figure it out for themselves.

And sure, if you've a table of agreeable not-competitive people willing to help with that figuring-out process on an ongoing basis this set-up could be great. My experience, however, is that a) some people tend to be more stubborn and-or competitive than that, and b) some - generally the more casual types - aren't willing to help with that figure-it-out process and would prefer the game do it for them.

Game mechanics are usually a fairly basic input-->processing-->output sequence. Most of the time the players provide the input(1), the game does the processing(2a) and gives output(2b), and the GM interprets that output and adds it to the fiction(3).

1. Player input: Action declaration "We search the room carefully, looking for any sign the princess was ever here."
2a. Processing: (meat-space game mechanics occur e.g. rolling of dice, checking of notes, or whatever the system in use asks for)
2b. Output: (meat-space game mechanics determine a result - let's say success in this case - which must be honoured in step 3)
3. Interpretation: GM narration "Tolbert, you find a few long blonde hairs caught in the window sash that roughly match what you'd expect to be the princess'; and Jerelle, you notice a stain on the floor - could be spilled tea - that can't be more than a few days old."

This on its own doesn't seem controversial. So where's the controversy? Is it the specifics around 2a? Is it lack of honouring the result in 2b? Is it that the GM gets to do the interpretation in step 3?

IMHO there's nothing terribly controversial in terms of a specific type of play. Maybe there's a question like "do all styles of RPG play use rules in the same way?" That sounds like possibly the direction the discussion could be going in, I'm not sure...

 

[clearstream]

This seems to imply there could be no first RPG - but there was! And there's no reason in purely abstract principle why it couldn't have been AW!

Sure, and that would bring in "extend".

Furthermore, as Baker puts it in AW, "Apocalypse World divvies the conversation up in a strict and pretty traditional way." He is not overriding pre-existing expectations ("tradition"). He is affirming them!

That's really beside the point. Making it a rule gives it force it would otherwise lack. Norms are probabilistic: there may be in a population in which behaviour or expectations etc converge to a norm, a distribution. Rules forcefully narrow that distribution. In the example, it’s plausibly the case that not all players have the understanding Baker is affirming.

The players’ job is to say what their characters say and undertake to do, first and exclusively; to say what their characters think, feel and remember, also exclusively; and to answer your questions about their characters’ lives and surroundings.​

I've underlined the parts of that rule that set out the who. And I've italicised the bits that constrain the what. Players are under no other constraints as to content, beyond a general one to cohere in what they say with the already-established fiction. The GM is of course under a wide range of constraints when it comes to exercising the permissions and fulfilling the obligations that pertain to their introduction of new shared fiction.

So just to be clear, you're now telling us that this post - which uses the phrases "description" and "consequence", and talks about a "matching" problem with ambiguity in the gam text, and gives as illustrations PbtA player-side moves and rolling dice to determin degree of success - is also talking about rules like "If you are a player, you are permitted to say this thing about your character."

Yes. As a rule it supersedes, including extending, preexisting norms.

That rules has no "matching" problem (there is no problem working out who is a player). The consequence is not a "fitting" one (I had taken "fitting" to mean something like "fits with the established fiction", consistent with your own extended history of posting about "saying what follows".)

This has been addressed upthread. The matching problem is not part of that definition of a rule, it is entailed by that definition of a rule.

[EDIT I can perhaps see a way to better put this, more clearly locating the criteria for a match in the rule.]

Such as? Like the suggestion on the AD&D character sheet that you might want to draw a picture of your PC?

If you like, yes. It's normative in encouraging a common behaviour.

A rule that you must draw a picture of your PC will have greater force. It will override "might want".

 

[Lanefan]

I was probably vaguer in that post than was useful. What I meant by that is that if a game is going to have, say, four results possible in a given task output, the player should have a fairly good idea what those results will be given other elements of the situation. There are going to be cases where they won't (because data about the situation is obscured for reasons in-game) or where the situation is sufficiently outside the normal play events that the only way they'll know is if the GM tells them (and under that circumstance, I'd much prefer he did), but to use a simple example, if someone is jumping across a small chasm, I'd much prefer that the standard mechanics tell me what the outcome possibilities are there and what category they land in.

Using this example in particular, the possible outcomes are pretty obvious before game mechanics even get involved:

--- you make the jump without problem
--- you make the jump but don't stick the landing
--- you mostly make the jump but end up partly hanging off the far edge
--- you face-plant into the far side but can grab something before you fall
--- you fall to your doom
--- you abort at the last minute and don't even try the jump.

Now, how closely you-as-player should know the odds of each result ideally would mirror your character's knowledge, but even then only in general rather than hard-coded numerical terms.

So yeah, if it's a 5-foot-wide chasm your odds of jumping it arte pretty damn good, if not perfect; if it's a 15-foot chasm that's a different story.

 

[Thomas Shey]

Using this example in particular, the possible outcomes are pretty obvious before game mechanics even get involved:

--- you make the jump without problem
--- you make the jump but don't stick the landing
--- you mostly make the jump but end up partly hanging off the far edge
--- you face-plant into the far side but can grab something before you fall
--- you fall to your doom
--- you abort at the last minute and don't even try the jump.

Now, how closely you-as-player should know the odds of each result ideally would mirror your character's knowledge, but even then only in general rather than hard-coded numerical terms.

So yeah, if it's a 5-foot-wide chasm your odds of jumping it arte pretty damn good, if not perfect; if it's a 15-foot chasm that's a different story.

Your second to last clause is where I disagree, because this is the kind of area where I want both the player and the GM/system to be on the same page, and the only reliable way to do that is, in the end of the days, knowing the numbers. Natural language just won't cut it because they can often mean very different things to different people.

 

[AbdulAlhazred]

"Play to find out" falls into this same category - it neither specifically allows nor disallows anything and thus is not a rule: it's a guideline.

I agree, but its "all of a piece" so to speak. Any one bit is fairly insignificant, or looks like nothing but guidance, until you put the entire thing together and then we see that "Play to find out" as an agenda has a huge impact on, and colors, all the things it is packaged along with. My character, Meda, has a motivating personal history, her father is missing. Will she find him? Play to find out! Will she abandon the search when it starts to bear fruit in order to satisfy her curiosity, or maybe to save Stonetop, or her friends? Play to find out! The GM doesn't know, the players don't know, the PCs sure don't know. But I know that, in keeping with that agenda item, the GM will say things that are unwelcome to Meda and put pressure of the sorts I just speculated about, or others, on that point. We will find out, is Meda a faithful devoted daughter? Is she going to put the village ahead of finding him?

And note, I GUARANTEE you, he's not just going to randomly show back up on his own, nor is he going to be revealed to be dead, etc. I mean, those things are technically possible IF they comport with the filling of the character's lives with adventure, etc. in some way, but there will be no easy victories, any such reappearance is sure to come with serious strings attached. I may well be asked to help decide what those are, but you can be sure the GM will have a big say too.

 

[innerdude]

Can you say how assigning authority is not who gets to say what?

"Who gets to say what" is slightly different from "who or what decides if what was just said is now true in the fictional game state."

Consider a game where other players were authorized to make action declarations for your character. (Not saying it's a game you or I would play, but that it's entirely possible. I've seen games like this, where you declare the action taken by other participants. The FFG card game Citadel being a partial example.)

Authority to make declarations doesn't change the fact that someone still has to determine if the declarations themselves are valid---i.e., have meaningfully changed the fictional state, with or without processing the declaration through a rules paradigm.

 

[clearstream]

Under most game texts players are active here, describing what their characters do in a situation that they grasp. Under some texts they will say what rule their act falls within the scope of. In many cases game texts will set the target for success.

None of that is at odds with failure without setback. Here a GM need say nothing at all, but as referee encourage the upholding and carrying through of the lusory attitude. Perhaps explaining the lusory means - pointing to the rule.

The concern of yours that I have quoted does not turn on inclusion or otherwise of failure without setback.

This said, flat failure can act to derail story much as called out in @andreszarta's third paragraph in #1646. Not because it imposes GM's version of events, but because no one's version is made to prevail.

In that sense, @pemerton's concerns about the player experience are right. Even though I would separate it out from whether the play is that in which GM's decides the fiction. Acknowledged that historically it has very often accompanied that sort of play... it's actually deprioritising of stories players have in mind that it drives. Making it unsuitable for play prioritising player character stories.

How is failure-with-setback any different? In many cases it's exactly where GM does add twists to the fiction. If that is not to also derail player stories there must be a principle relating the setbacks to such stories. Flat failure obviously lacks the resources to relate to anything. In the face of flat failure, stories crumble. GM stories are not typically subject to die rolls so wind up prevailing just by default... which is suitable for modes in which that's intended.

In this case I change my mind on

I think it is obvious why this sort of RPGing is experienced by many participants as centring the GM's conception of the fictional situation.

and am in agreement that it could be experienced exactly that way by many participants. One route away from that is to let the chips fall as they may on all sides, i.e. ensure that any stories on GM's side are equally subject to flat failure. The BBEG rolls to see if they draw the army of darkness to their side.... nope, that fails. That kind of roll can play a part in play prioritising gamism and simulationism.

 

[clearstream]

"Who gets to say what" is slightly different from "who or what decides if what was just said is now true in the fictional game state."

Baker

if all your formal rules do is structure your group's ongoing agreement about what happens in the game, they are a) interchangeable with any other rpg rules out there, and b) probably a waste of your attention. Live negotiation and honest collaboration are almost certainly better. . .

My take away from what was quoted in the OP is that "who or what decides if what was just said is now true in the fictional game state" isn't a nothing, but it's also not the focus. The focus is securing that "what was just said" should include "the unwelcome and the unwanted".

I believe our quibble here arises from a parsing of my "who gets to say what" as deconstructible. So yes, Baker focuses on "who", and he focuses on "what" (that being the unwelcome and the unwanted.) Even deconstructed, if there were a sole focus it should be on the latter, not the former, unless it is also read that the focus is on the who-and-the-what. I didn't anticipate that construal. As I said, of the who and the what, I took away from the OP a focus on the what.

 

[pemerton]

It is possible to design rules so that they generate a certain sort of phenomenon, although no one is obliged, in applying the rule, to have regard to the phenomenon in question. A classic discussion of this possibility is Rawls's Two Concepts of Rules.

Here is a rule that addresses who has to say what in a RPG (it is not, on its own, a complete statement of rules for a RPG):

If a player declares an action for their PC, and then makes a roll to determine the outcome of that action declaration, and the roll fails, the GM must say something about what happens next, and the thing that the GM says must clearly defeat or set back the goal the player was hoping the action would achieve for their PC.​

(Some readers may recognise that this is a rule from the RPG Burning Wheel.)

This rule does not speak about "the unwelcome and unwanted". It talks about some particular player hopes, and the relationship to them of fiction that a different participant is obliged to narrate. Nevertheless, following this rule, in conjunction with some other appropriate rules, will make the unwanted and unexpected a part of play.

 

[pemerton]

Can you say how assigning authority is not who gets to say what?

Assigning authority is necessary to establish who gets to say what. It is not sufficient. This is why Apocalypse World spends only a few sentences on assigning authority to the MC (over everything that is not assigned to the players - ie everything except (i) what PCs say, undertake to do, think and feel, and (ii) answers to the questions the MC poses to the players), but spends many many pages on what the MC is obliged to say when they exercise the authority that they have been given (this is elaborated in detail in discussions of player-side moves, the agenda, the principles, the GM-side moves, the discussion of fronts, and probably other places too that I'm not remembering at the moment).

EDIT:

As per the OP, Baker says

Some very good designers consider the assignment of authority to be the point of rpg design. I do not.

As a designer, it's my job to make as sure as possible that the game won't break down into moment-to-moment negotiations about raw assent despite the game's rules and the players' upfront commitment to them. But the brute assignment of authority is NOT how to accomplish that.

When my games assign authority they do so in strict service to what I consider the real point: setting expectations and granting permission.​

Why are games prone to breaking down into moment-to-moment negotiations, despite upfront commitment to rules? Because the rules are adopted purely voluntarily, and hence can be set aside at any time!

And "brute" assignments of authority - you get to say what your PC tries to do, while you other get to say what happens in the setting - won't avoid that risk. If anything, as I noted upthread, they are prone to exacerbate it: there's a reason why conflicts over GM authority are a recurrent feature of RPGs like many approaches to D&D that deal with authority primarily in this fashion.

By setting expectations about how authority will be used, where those expectations are acceptable, rules help ensure that they will be followed. And one way to set expectations is to grant permissions - ie permissions to say things that might not be said in a purely negotiated environment. We can see this in BW: If I succeed on my roll, intent and task become part of the fiction; but if I fail, you - the GM - get to narrate something about how my goals for my PC are defeated or set back.

The GM is conferred a clear permission to narrate horrible things about what befalls the player's PC, but that permission is gated behind a mechanism which also generates an expectation about when the player gets to establish truths about the fiction that are good for their PC. This in turn makes the player accept that when the GM gets their turn, what they say is OK.

This is extremely different from (eg) how @Micah Sweet, @Maxperson and you have characterised "trad" or "immersionist" play in this thread.

 

[clearstream]

If a player declares an action for their PC, and then makes a roll to determine the outcome of that action declaration, and the roll fails, the GM must say something about what happens next, and the thing that the GM says must clearly defeat or set back the goal the player was hoping the action would achieve for their PC.

Just to check, do you mean "must clearly not defeat or set back"?

I ask because "defeat" could be taken to imply a flat negative, which I think BW rules out.

 

[clearstream]

Assigning authority is necessary to establish who gets to say what. It is not sufficient. This is why Apocalypse World spends only a few sentences on assigning authority to the MC (over everything that is not assigned to the players - ie everything except (i) what PCs say, undertake to do, think and feel, and (ii) answers to the questions the MC poses to the players), but spends many many pages on what the MC is obliged to say when they exercise the authority that they have been given (this is elaborated in detail in discussions of player-side moves, the agenda, the principles, the GM-side moves, the discussion of fronts, and probably other places too that I'm not remembering at the moment).

Right, so my meaning then was that in the text you quoted in your OP, it seemed to me Baker is directing focus more toward the what than the who. Does that track for you?

 

[pemerton]

Just to check, do you mean "must clearly not defeat or set back"?

I ask because "defeat" could be taken to imply a flat negative, which I think BW rules out.

It must defeat the players' goal. For instance, if their goal is to get safely through the door, that must be defeated. Failure of task may or may not be part of this.

Nothing happens probably doesn't count as saying something about what happens next ("nothing" is something of a degenerate case) and it also doesn't defeat a goal. Hence it is not an acceptable narration.

 

[clearstream]

It is possible to design rules so that they generate a certain sort of phenomenon, although no one is obliged, in applying the rule, to have regard to the phenomenon in question. A classic discussion of this possibility is Rawls's Two Concepts of Rules.

Here is a rule that addresses who has to say what in a RPG (it is not, on its own, a complete statement of rules for a RPG):

If a player declares an action for their PC, and then makes a roll to determine the outcome of that action declaration, and the roll fails, the GM must say something about what happens next, and the thing that the GM says must clearly defeat or set back the goal the player was hoping the action would achieve for their PC.​

(Some readers may recognise that this is a rule from the RPG Burning Wheel.)

This rule does not speak about "the unwelcome and unwanted". It talks about some particular player hopes, and the relationship to them of fiction that a different participant is obliged to narrate. Nevertheless, following this rule, in conjunction with some other appropriate rules, will make the unwanted and unexpected a part of play.

I see this rule as deconstructible, like this

If a player declares an action for their PC, and then makes a roll to determine the outcome of that action declaration, and the roll fails,
So this fits what I have loosely called "description" (I acknowledge a burden to find a better term.) What did we hear and see? It should be called attention to that at least one other rule is implied here. I don't think that is of importance to the discussion at hand. If it is, we can pick it up later.

the GM must say something about what happens next,
Where the description matches the rule it invokes it: functionally, GM must now say something about what happens next

and the thing that the GM says must clearly defeat or set back the goal the player was hoping the action would achieve for their PC.
The rule will fail if the description does not include the goal the player is hoping to achieve. Therefore it applies that criterion retroactively to secure that the description contains a goal. You can see how that could go in play. With that in place, the rule supplies an explicit statement relating to fitting consequences: if whatever GM says does not defeat or setback the goal, then they have failed to choose a consequence that fits.

I can make the assumption that defeat or setback is unwanted, given that the rule solicited player goal and one assumes goals are things that players want, while defeat of goals is something they don't want. If that assumption is a good one, then this rule alone is an appropriate lusory means that can secure the unwelcome and the unwanted. Suppose that defeat did not debar further attempts? In that case the unwanted will come down to inefficient means (I have to spend an hour when I wanted to spend a minute, or perhaps must pay the costs all over again.)

There's more that can be said, but this seems like a good place to pause and take stock.