Vincent Baker on mechanics, system and fiction in RPGs

clearstream said:

but is silent on "improv problems" like "given we could say anything, how do we know what we should say?" How do we know that the word on a character sheet successfully bedded itself in (arrowed to) the kinds of things we wanted to say distinctive to this TTRPG, and will successfully return a well-connected system position?

Click to expand...

This isn't a shortcoming of the methodology. Its a shortcoming of the fact that I deliberately chose not to model this in the example I gave.

Keep in mind the point of these machination diagrams is to abstract gameplay so we can directly examine whats happening.

Thats why for example when we look at Pac-Man, its more useful to look at this:



To examine what the gameplay is generally doing. In this case, this models the overall goal of Pac-Man and what players do.

This isn't Pac-Man though. This is Pac-Man:



And in fact this diagram, if it was a full replication of Pac-Man would actually be replicated 256 times in sequence, representing the 256 levels in Pac Man that individually change the parameters of the basic gameplay. (Ie less/more dots, different power up times, etc)

An RPG is naturally a lot more complex than Pac-Man is. If one wants to fix an issue with Pac Man, trying to do so on a Macro scale wouldn't really work. After all, trying to fix or adjust all the different parts of that diagram simultaneously and carrying it through all 256 stages, while also playtesting, just isn't an efficient design method.

Instead, its better to zoom in to the micro scale and adjust there, as those changes can then be individually applied and tested for a desired effect. Then it becomes relatively easy to carry through the change throughout the rest of the game.

So when it comes to RPGs what you are hung up on is trying to design in the macro. Thats why the terminology Baker uses is obtuse and why nobody is actually getting any sort of design method out of this. After all, no one has actually shown up and demonstrated what I asked pemerton to do. If this works, one should be able to cleanly and concisely lay out the methodology and demonstrate its use.

clearstream said:

For example, I can take the diagram to be such a zoomed-out description. Is that what you are getting at?

Click to expand...

Zoomed in. The abstraction of gameplay serves to provide clarity for the gameplay itself, so that it is readily observable and malleable. Being able to go in and make micro-scale tweaks is the ideal, as those are more practical than trying to wrap one's head around an esoteric macro level problem.

For example, rather than losing the plot trying to make sense of the Position stuff in the OP, we can look at the Fiction as a "game state". This State changes as Actions are engaged, and the new State that results may or may not influence the available Actions.

In an Improv game, the initial game state is established by Method X. Method X can be a lot of things. In theater it relies on suggestions from an audience interacting with the players interpretations.

In a solo RPG, it relies on suggestions from Oracles and/or GM Emulators interacting with the Players interpretations.

In a Trad RPG, it relies on the GM directly establishing a State of their choice, that the Players then interact with with little to no independent player interpretation, though collaborative interpretation is often included.

In AW, the GM and Players establish the initial State collaboratively, though theres options for other mixes, including one that mirrors Trad play.

From here, the Players through whatever means ("The Conversation" as PBTA games use) then interact with the Game State, and are permitted effectively unlimited Actions. There is no strict limit on what can be done mechanically. (To Do It, Do It)

Some Actions, however, Trigger a feedback response. What counts as a trigger is typically established alongside the initial game state, and the game rules itself typically establish many of these. (Moves being the most common in PBTA. Yes,And et al are simpler forms of it; prescriptive abilities and certain rules in trad games a more complex version)

The feedback then gets resolved, and integrates with the previous game state to form a new one. This then repeats until the game is considered over.

Broke down this way, it should be really apparent why I keep saying what we're talking about is an improv game.

Emberashh said:

So when it comes to RPGs what you are hung up on is trying to design in the macro. Thats why the terminology Baker uses is obtuse and why nobody is actually getting any sort of design method out of this. After all, no one has actually shown up and demonstrated what I asked pemerton to do. If this works, one should be able to cleanly and concisely lay out the methodology and demonstrate its use.

Click to expand...

In this blog, Baker explains how to apply his design language. Linked is the first of a nine-part series.

Emberashh said:

For example, rather than losing the plot trying to make sense of the Position stuff in the OP, we can look at the Fiction as a "game state". This State changes as Actions are engaged, and the new State that results may or may not influence the available Actions.

Click to expand...

Baker worked through problems like these in diagrammatic form a few years prior to the publication of Adams' book. Over those same years Ernest had been publishing his own series of videogame design articles in Gamasutra. I printed and read many and for awhile kept a folder of them.

It's certainly possible to create diagrams representing the claims in the OP in order to explore and succinctly communicate those ideas. Vincent Baker has performed that sort of exploration, and as documented on his blog used the analysis to construct the PbtA TTRPG design language. Like Harper, Baker iterated over several published TTRPGs. His approach is to my mind hard to fault. Harper's work draws on Baker's analysis (and if anything applies it more rigorously.)

Emberashh said:

In an Improv game, the initial game state is established by Method X. Method X can be a lot of things. In theater it relies on suggestions from an audience interacting with the players interpretations.

In a solo RPG, it relies on suggestions from Oracles and/or GM Emulators interacting with the Players interpretations.

Click to expand...

It's kind of a digression, but I think the specific problem solved by Oracles - for instance in Ironsworn as you discussed earlier - is to allow players to explore their own (and each other's) imagination by propelling them in directions they would not otherwise have considered.

Emberashh said:

In a Trad RPG, it relies on the GM directly establishing a State of their choice, that the Players then interact with with little to no independent player interpretation, though collaborative interpretation is often included.

In AW, the GM and Players establish the initial State collaboratively, though theres options for other mixes, including one that mirrors Trad play.

From here, the Players through whatever means ("The Conversation" as PBTA games use) then interact with the Game State, and are permitted effectively unlimited Actions. There is no strict limit on what can be done mechanically. (To Do It, Do It)

Some Actions, however, Trigger a feedback response. What counts as a trigger is typically established alongside the initial game state, and the game rules itself typically establish many of these. (Moves being the most common in PBTA. Yes,And et al are simpler forms of it; prescriptive abilities and certain rules in trad games a more complex version)

The feedback then gets resolved, and integrates with the previous game state to form a new one. This then repeats until the game is considered over.

Broke down this way, it should be really apparent why I keep saying what we're talking about is an improv game.

Click to expand...

So that's the high-level description. We can easily see that there is linkage between system and fiction. One then has to work through the sorts of linkages available, and draw conclusions to guide future design. Applied to the problem space, with enough iterations, one lands at something like Vincent Baker's clouds and cubes diagrams.

Anyway, I believe I have lost track of what you were arguing for (or against)?

clearstream said:

Anyway, I believe I have lost track of what you were arguing for (or against)?

Click to expand...

At this point this is just debate for debates sake. You're not open to the idea, and there isn't going to be some come to jesus moment where you find a new god to adhere to, as thats not what Im trying to do. Ergo, this is pointless.

clearstream said:

In this blog, Baker explains how to apply his design language. Linked is the first of a nine-part series.


Baker worked through problems like these in diagrammatic form a few years prior to the publication of Adams' book. Over those same years Ernest had been publishing his own series of videogame design articles in Gamasutra. I printed and read many and for awhile kept a folder of them.

It's certainly possible to create diagrams representing the claims in the OP in order to explore and succinctly communicate those ideas. Vincent Baker has performed that sort of exploration, and as documented on his blog used the analysis to construct the PbtA TTRPG design language. Like Harper, Baker iterated over several published TTRPGs. His approach is to my mind hard to fault. Harper's work draws on Baker's analysis (and if anything applies it more rigorously.)


It's kind of a digression, but I think the specific problem solved by Oracles - for instance in Ironsworn as you discussed earlier - is to allow players to explore their own (and each other's) imagination by propelling them in directions they would not otherwise have considered.


So that's the high-level description. We can easily see that there is linkage between system and fiction. One then has to work through the sorts of linkages available, and draw conclusions to guide future design. Applied to the problem space, with enough iterations, one lands at something like Vincent Baker's clouds and cubes diagrams.

Anyway, I believe I have lost track of what you were arguing for (or against)?

Click to expand...

I came across an interesting bit on emergent storytelling in a sidebar in chapter 11:


You know what games are solving this problem? Tabletop role-playing games. The work of Baker, Harper, and others seems complementary to the work of Adams and Dormans. It’s not clear why it has to be either-or.

kenada said:

The work of Baker, Harper, and others seems complementary to the work of Adams and Dormans.

Click to expand...

Baker Et al are focused on building genre emulation mechanics. Genre Emulation is only one means of storytelling in games.

The assertion that genre emulation = (TT)RPGs (as was seen more than a few times in these topics), isn't accurate and is in turn very exclusionary to other methods.

Plus, Baker et al's overall method for genre emulation isn't the only way to do it, and the third person authorial stance it takes wouldn't be compatible with most video games until some sort of proto-Holodeck type game is possible.

Emberashh said:

Baker Et al are focused on building genre emulation mechanics. Genre Emulation is only one means of storytelling in games.

Click to expand...

This seems to be conflating the games Baker designed with the design work he did to design them. It’s not clear how boxes and clouds, conflict resolution, etc are limited to genre emulation. I’ve certainly found them useful for the homebrew system I’m designing, which I described post #54 as being OSR-adjacent. It’s pretty open-ended, and story-telling isn’t even really its focus. We do get to learn about the characters, but the overall mechanical orientation is towards treating the world as a place where they live.

I would also cite Principia Apocarypha as an example in the OSR space of taking Baker’s approach to principles in Apocalypse World and extending it to another style of play. However, this doesn’t really seem to have anything to do with my point (see below).

Emberashh said:

The assertion that genre emulation = (TT)RPGs (as was seen more than a few times in these topics), isn't accurate and is in turn very exclusionary to other methods.

Click to expand...

I’m suggesting that tabletop role-playing games address the emergent storytelling problem identified in that sidebar. I mentioned Baker et al because it’s topical to this thread, but it doesn’t follow that I’m saying they offer the only solution. That doesn’t even make sense in the context of my next statement, which was to question why it has to be either-or. I’m saying it should be both or more; there’s no one source with all the answers.

Emberashh said:

Plus, Baker et al's overall method for genre emulation isn't the only way to do it, and the third person authorial stance it takes wouldn't be compatible with most video games until some sort of proto-Holodeck type game is possible.

Click to expand...

I almost included that point in my post, so I guess I’ll address it now: Yes. The book’s statement that games are media-independent is a nice way to think about them in the abstract (much like a spherical cow), but it doesn’t hold up in reality. Different media have different advantages and limitations. Sometimes games that are practical to implement in one medium are impossible (or at least impractical) to implement in others.

I don’t think designers should be limiting themselves to what works in all media. Sometimes that’s the right decision, but sometimes what’s right is to design for the target media. I think that’s especially true of something like tabletop role-playing games. Limiting ourselves to avoid leveraging the human element effectively foregoes what makes them unique compared to other games. Losing that would be a shame.

kenada said:

This seems to be conflating the games Baker designed with the design work he did to design them

Click to expand...

Im not sure I can reconcile one person telling me his games are evidence of his design work and another telling me that they're not related.

kenada said:

I’m suggesting that tabletop role-playing games address the emergent storytelling problem identified in that sidebar.

Click to expand...

Not all of them, and thats the reason why genre emulation was picked on. Genre Emulation mechanics are progression mechanics, and additionally are all far too complex for the resulting stories to be considered genuinely emergent.

Their more basic forms Yes,and et al would count, however.

Plus, as an aside, my personal theory is that emergent storytelling is already happening in games inherently. Mechanics in of themselves tell stories, and great webs of them tell even greater stories.

Games in general already have genres of their own, and as an artform all games convey meaning through interactivity, which comes from Mechanics. Genre Emulation works for gamifying genres in other art forms (and we even see examples of that style in video games already. Uncharted/Max Payne style games), but it fundamentally isn't emergent on its own.

As controversial as it is to say, as I no doubt will get naughty word for it, genre emulation as progression mechanics are just a form of railroading, and most genre emulation games are hybridizing with an emergent game form to hide this. Uncharted allows for emergence with open levels and free form gameplay, PBTA with the integrated Improv game.

kenada said:

Sometimes games that are practical to implement in one medium are impossible (or at least impractical) to implement in others.

Click to expand...

That comes down to mechanic types. As the book says and Ive reiterated, that statement of universality applies to Discrete mechanics only.

kenada said:

I don’t think designers should be limiting themselves to what works in all media.

Click to expand...

Nor do I, but thats not what I was communicating at any point. All Ive attempted to communicate is just being cognizant of what we're actually talking about so we can address whatever issues properly, without getting into a debate to establish we're speaking the same language first.

Emberashh said:

In AW, the GM and Players establish the initial State collaboratively, though theres options for other mixes, including one that mirrors Trad play.

Click to expand...

I notice here that you contrast the GM role and the player role, much as I have done in some other posts. So I infer from this that you are pretty clear about the meaning of my usage.

I also note that the initial state is not an arrangement of pieces, or of mathematical structures. It is an agreed fiction. Shared imagination.

Emberashh said:

From here, the Players through whatever means ("The Conversation" as PBTA games use) then interact with the Game State, and are permitted effectively unlimited Actions. There is no strict limit on what can be done mechanically. (To Do It, Do It)

Click to expand...

The means of talking to people and getting them to imagine things together is fundamental here. This is the key to RPG design: because it is adapting something humans have done for (one assumes) 100+ thousand years - that is, tell stories to other human beings to get them to imagine things - into a context where the storytelling is (i) shared in some fashion, and (ii) simultaneously constrained in some fashion, so as (iii) to permit a type of game play.

Identifying the method that will satisfy (ii) so as to yield (iii) is a technical challenge. The methods that Arneson identified (ie map-and-key, together with GM extrapolation free kriegsspiel-style) permit a limited range of imaginary situations to produce a particular sort of play ("Gygaxian skilled play"). Vincent Baker is both generalising, and identifying and articulating new methods.

Emberashh said:

Some Actions, however, Trigger a feedback response. What counts as a trigger is typically established alongside the initial game state, and the game rules itself typically establish many of these.

<snip>

The feedback then gets resolved, and integrates with the previous game state to form a new one. This then repeats until the game is considered over.

Click to expand...

The idea of actions and trigger again depends fundamentally on shared imagination - "events in the fiction". In this respect it is quite different from (eg) moving a piece in chess, or playing a card in bridge, or pressing a button on a keyboard.

And the feedback response, in AW, also depends on imagination - eg Announce future badness.

Your characterisation of the game completely elides all the actual complex details of play which someone had to invent, and which are far from obvious, in that the general mode of game was invented in the early 1970s, but a systematic understanding of the techniques used in AW was not developed until 30+ years later.

kenada said:

I came across an interesting bit on emergent storytelling in a sidebar in chapter 11:

Emergent storytelling, which we mentioned briefly in Chapter 10, is a research field that seeks to resolve the inconsistency between traditional gameplay experiences and traditional story experiences. A hypothetical emergent storytelling game would use gamelike emergent mechanics to create gameplay and an emergent progression system to generate dramatically interesting plot events without an author’s involvement. At the same time, it would somehow guarantee that the experience feels properly storylike, without repetition, randomness, reversals in time, or noncredible characters. Some efforts to create such a system have used artificial intelligence to search through possible future events in a plot the way a chess program searches for possible future moves in a chess game. Instead of trying to checkmate the king, the search algorithm tries to find an enjoyable plot. To date, no one has succeeded in building a full-game-sized emergent storytelling system. All the efforts thus far have produced only small prototypes.​

You know what games are solving this problem? Tabletop role-playing games. The work of Baker, Harper, and others seems complementary to the work of Adams and Dormans. It’s not clear why it has to be either-or.

Click to expand...

The problem they describe was solved some time between 1989 (Prince Valiant) and 1998 (Maelstrom Storytelling). I put forward a date and game range because there's also a bit of blurriness in what counts as a full solution.

These RPGs don't require an AI to do a chess-like search of future states. The technique that they use only requires a human to make decisions about imagined situation-changes based on the stakes and other parameters established in the framing and the action declaration.

Note that, with AW, Baker has identified a different though in some ways similar way of solving the problem - it relies less on an abstract notion of "stakes" and more on a finer-grained conception of how a story can evolve ("badness", "opportunities", "costs", "separation of protagonists", setbacks vs losses, etc).

pemerton said:

I notice here that you contrast the GM role and the player role, much as I have done in some other posts. So I infer from this that you are pretty clear about the meaning of my usage.

Click to expand...

I never said there wasn't a specific function the GM performs.

You're being incredibly uncharitable to make a cheap dig.

pemerton said:

I also note that the initial state is not an arrangement of pieces, or of mathematical structures.

Click to expand...

And no where did I say otherwise.

pemerton said:

The means of talking to people and getting them to imagine things together is fundamental here.

Click to expand...

Reminder:

Emberashh said:

The abstraction of gameplay serves to provide clarity for the gameplay itself, so that it is readily observable and malleable. Being able to go in and make micro-scale tweaks is the ideal, as those are more practical than trying to wrap one's head around an esoteric macro level problem.

Click to expand...

pemerton said:

In this respect it is quite different from (eg) moving a piece in chess, or playing a card in bridge, or pressing a button on a keyboard.

Click to expand...

Nothing I said says otherwise.

pemerton said:

Your characterisation of the game completely elides all the actual complex details of play which someone had to invent, and which are far from obvious, in that the general mode of game was invented in the early 1970s, but a systematic understanding of the techniques used in AW was not developed until 30+ years later.

Click to expand...

I really don't care to continue debating the merits of your chosen reality.