Maybe it's "Authored" vs. "Emergent"

[clash]

I think the real problem with this discussion is that several different metrics are being proposed to delineate between authored and emergent.

1. Dm supplied story vs player generated story
2. Event driven play vs location driven play
3. Low player agency vs high player agency

I've probably missed some and I don't think they correlate with each other. A "sandbox" might be player generated story, location driven play and high player agency but it also might be event driven as others have demonstrated. Or it could be low player agency if you can't actually effect the world in a meaningful way. On the other hand you can have a dm supplied story with very high level of player agency if the story is written as you go based on the decisions of the players. Where's does emergent fall on these metrics? Where does authored? Are they strictly the first metric? Does the first metric even make sense, or as suggested are all stories really supplied by the dm and the real difference is planned or improvised?

I think we need to narrow down the list of metrics so we can talk about a single metric at a time.

[NichG]

It's not that I don't understand the idea of the idea of a discrete binary separation, it's that I don't think it accurately reflects DMing styles, for the reasons I listed above (DMing style is a human trait, the games I've experienced don't seem to fit, and the criteria for each are not mutually exclusive). I understand the concept, using what I think is a much cleaner metaphor, one textbook isn't exactly more true than another, it's that it contains a higher proportion of true statements than another.

The reason I don't find that convincing, is that these scenarios are being talked about as if they were more or less emergent than one another. To use a real world example again, in the campaign where the players decided to evacuate their home world, that scene, or that decision was emergent gameplay, because the players choice drastically affected the game world in a way the DM did not prepare for. The DM did however prepare an expected outcome, which makes it more authored than if he had not. This does not seem to be two different elements of the same campaign, it was one decision. It was one scene that was in some ways emergent, and in others authored.

I've made my arguments for why I believe it to not be a discrete category, and if you don't find them convincing, that's totally fair. (I never was good at debate in high school) I haven't really seen any actual arguments on the other side though, only the repeated assertion that it's true.

(I hope this doesn't come across as hostile, I am interested in your reasons, I'm not trying to insist you don't have any)

I think the supermicro, micro, macro, supermacro distinctions were helpful here. In that context, I take the claim about these things as being a binary dichotomy (true or not) as being a claim that e.g. 'consistently applying a sufficiently emergent approach (not necessarily a pure approach) at small scales results in things appearing more and more emergent as you go to large scales, whereas consistently applying a sufficiently authored approach (not necessarily a pure approach) at small scales results in things appearing more and more authored as you go to large scales'. There are certainly physical systems that work this way, so at least from a perspective of 'can this even make sense?', I think its possible for authored vs emergent to become binary at the super-macro scale in practice, even if those campaigns will have mixes of authored and emergent stances or nuanced interpretation of those stances at, say, the micro level. Is it actually the case? I wouldn't want to conclude one way or another at this point.

For example, lets say that whenever gameplay moves forwards, there's some 'body of past decisions that mattered' which gets carried forward. Any time there's an 'authored' step or sequence, you're either preventing certain choices from entering this pool, or you may be squeezing some of this information through a filter, resulting in some of the information potentially being discarded as 'now made irrelevant'. Any time there's an 'emergent' step, there's the possibility that some of the information in that body of past decisions gets replicated - there's the fact that the bard had an offhand conversation with a slumming noble in a tavern at some point in which he espoused an anarchist philosophy (1 thing recording that information), and now that the party is in that noble's lands the noble remembers and reacts differently leading to additional events that 'record' that previous information. So a game can include a mix of those authored steps and emergent steps, or even single steps that combine aspects of both. That's the 'continuum' part.

At the same time, we could look at the statistical tendency at the super-macro level. If on average more records become irrelevant than the number of copies that get created, then eventually old events will be forgotten. This is the 'river of history' view that PhoenixPhyre was taking - on large enough scales, the consequences of individual events, choices, people get averaged out. On the other hand, if the consequences of choices replicate faster than they become irrelevant, then even after an infinitely long campaign, some offhand thing that someone did in session 1 could still matter. That would be the 'chaos butterfly' viewpoint in a sense, that were you to play out the entire campaign changing only that one thing in session 1, ultimately they could become arbitrarily different stories.

So the question of whether there's a binary is basically 'is there in practice a phase transition between the river of history view and the chaos butterfly view, or do campaigns actually fall in between?'. If each step you take is drawn from the same statistical distribution and applies in the same way across all 'copies of past decisions' regardless of what they are, then its hard to avoid it becoming a phase transition since in one limit campaign trajectories converge on each-other exponentially, and in the other limit they diverge exponentially, and all of the continuum of 'how much emergence?' does is it changes the time constant of the exponential. exp(-0.001t) and exp(0.001t) become infinitely different when t->infinity.

On the other hand, there's no reason why a GM has to make statistically i.i.d. decisions about what past things should matter or not. You could have some 'privileged decisions' for example which are guaranteed to always matter and are never erased no matter what, which would mean that e.g. some particular session 1 choice still matters at t->infinity, but on average session 1 choices don't matter as t->infinity. You could have more exotic stuff as well, especially if you take into account players as active elements in keeping information about past choices alive rather than just relying on the GM for that.

So, I'd want to see the data. Because I can come up with various ways to be in the middle, but if people don't actually tend to GM like that it doesn't matter if some theoretical GM could do it just to spite the idea of there being a dichotomy.

[PhoenixPhyre]

For example, lets say that whenever gameplay moves forwards, there's some 'body of past decisions that mattered' which gets carried forward. Any time there's an 'authored' step or sequence, you're either preventing certain choices from entering this pool, or you may be squeezing some of this information through a filter, resulting in some of the information potentially being discarded as 'now made irrelevant'. Any time there's an 'emergent' step, there's the possibility that some of the information in that body of past decisions gets replicated - there's the fact that the bard had an offhand conversation with a slumming noble in a tavern at some point in which he espoused an anarchist philosophy (1 thing recording that information), and now that the party is in that noble's lands the noble remembers and reacts differently leading to additional events that 'record' that previous information. So a game can include a mix of those authored steps and emergent steps, or even single steps that combine aspects of both. That's the 'continuum' part.

Every game prevents certain choices from entering the pool of "what gets carried forward" or filters it out as irrelevant. Every game whatsoever. Most even do so actively as things go--your character is acting at a different frequency than the player is. And you're not getting the full, unfiltered stream from them. The whole point of abstractions is that they abstract information. Which necessarily filters some out as irrelevant. Because the information has to pass from the fictional world up through words (which are lossy and low-bandwidth) to players and then back downward. Information is lost[1] at every step of the way, no matter whether you're authoring things or letting things emerge.

This is not a distinction relevant here--even the most emergent game can't keep track of all or even more than a tiny fraction of the choices made. Even deciding which choices are salient is a filtering choice. And that must be done in order for the game to time-evolve. An authored game, as long as it's not authored in stone, in advance, can track those offhand conversations just as well as an "emergent" one. Or even better, because a fully emergent game requires that the facts surrounding the slumming noble and his lands to have been generated (which may not actually be conducive to anything interesting later on), while the authored one can look back at that and plunk a suitably-configured "noble's lands" somewhere along the path that the party has decided to take, as a "plot hook" (optional or not).

Personally, I find that when both sides take responsibility for moving things onward, collaboratively, accepting that by doing so you are authoring material, players' choices matter even more meaningfully. Because you can really explore how those changes affect things. Where that gets utterly lost in the noise most of the time in a less-consciously[2] authored scenario, just because there's so many other things going on (in a good one, anyway). In a semi-authored collaborative game, the people involved can decide that X decision is the one they want to focus on and work together to explore it. Whereas in a fully emergent one, players can be at cross-purposes, the players can try to do something that just doesn't fit the world (causing dissonance), or just do other things and that possibly-interesting consequence just gets lost in the shuffle.

[1] and that's a necessary and important fact.
[2] everyone is constantly authoring things at least in the very short-term. Every act involves making a (conditional) plan for the future. Players and DM. If the next PC does X, the monster will do Y. If the players ask for X, he'll respond Y. Waiting until your turn to figure out what you're going to do is widely considered a bad thing, because it slows things down tremendously.

At the same time, we could look at the statistical tendency at the super-macro level. If on average more records become irrelevant than the number of copies that get created, then eventually old events will be forgotten. This is the 'river of history' view that PhoenixPhyre was taking - on large enough scales, the consequences of individual events, choices, people get averaged out. On the other hand, if the consequences of choices replicate faster than they become irrelevant, then even after an infinitely long campaign, some offhand thing that someone did in session 1 could still matter. That would be the 'chaos butterfly' viewpoint in a sense, that were you to play out the entire campaign changing only that one thing in session 1, ultimately they could become arbitrarily different stories.

If that's what you got from it, one of the two of us miscommunicated. Generally, in many cases the consequences of individual events do average out. But not always. Just like generally, objects in motion remain in motion. But PCs can be and usually are "outside forces" that change the momentum of the system. The point was that, as I said above, every single game involves "averaging out" some events. No matter what. Period. Inevitably. The only question remains which events get averaged out and which can cause divergences that are noticeable above the noise.

Consider the following cases:
1. A character has brown hair. In a parallel universe, the same character dyes his hair slightly blonder. The character never does anything else notable. Does it make any sense that this would cause an arbitrarily large divergence in those universes? No.
2. A character burps while out on the road, with no one else around but the PCs. Does that cause a divergence in timelines of arbitrary magnitude? No.

Basically, there are a constant multitude of events going on. But the vast majority of those "microstates" end up in the same equilibrium macrostate, to use the statistical mechanics analogy. Most historical events are over-determined--there is no single but-for cause. If one person hadn't pulled that trigger (to start the war in the fantasy kingdom), someone else, somewhere close would have. Because the situation was that unstable. PCs, by nature of the beast, tend to be near critical points. So more of their actions tend to cause divergences than your average person. But that doesn't mean all their actions cause divergences.

Basically, the universe, because it's such a huge system, tends to be very near the thermodynamic equilibrium (even though that's a dynamic, evolving equilibrium for many reasons). And those equilibria tend (but are not always) to be relatively stable. Chaotic systems behave this way as well--chaos is deterministic and even predictable within certain bounds. Take your classic period-doubling experiment. Below certain values of the critical parameter, you get one and only one solution (a single period). Between slightly larger values, you get two and only two values. Which one you're in for any initial state is difficult to tell, but you have a 50% chance of being in either one. And a 0% chance of being in any other state. For slightly larger values, you get 4 possible periods. Again, which one follows from any given starting state is basically impossible to determine with more than 25% accuracy, but you can guarantee you're in one of those states and no other. Etc. And you can actually predict with exactness the values of the parameter at which each period-doubling occurs and what the possible states are. So given the value of the parameter, you can with 100% accuracy predict the set of possible states the system can be in. But you can't predict (better than chance) which individual state the system is in.

Chaos is not "any change in initial conditions makes an arbitrarily large divergence asymptotically". All it means is that arbitrarily small differences can, sometimes, under the right conditions, cause unpredictable (based on the knowledge of the initial conditions) changes in the behavior of the system." Most chaotic systems get entrained very easily and settle down to relatively stable states. The only thing is you can't predict which (of the possible) states any given set of initial conditions + perturbations will lead to. Chaotic =/= unstable equilibrium. Chaotic == hard to predict on long timescales.

[OldTrees1]

I think the real problem with this discussion is that several different metrics are being proposed to delineate between authored and emergent.

1. Dm supplied story vs player generated story
2. Event driven play vs location driven play
3. Low player agency vs high player agency

I've probably missed some and I don't think they correlate with each other. A "sandbox" might be player generated story, location driven play and high player agency but it also might be event driven as others have demonstrated. Or it could be low player agency if you can't actually effect the world in a meaningful way. On the other hand you can have a dm supplied story with very high level of player agency if the story is written as you go based on the decisions of the players. Where's does emergent fall on these metrics? Where does authored? Are they strictly the first metric? Does the first metric even make sense, or as suggested are all stories really supplied by the dm and the real difference is planned or improvised?

I think we need to narrow down the list of metrics so we can talk about a single metric at a time.

I think Authored vs Emergent was isolating something similar to the 1st metric. The 2nd and 3rd are symptoms of common approaches to both directions (with exceptions as you mentioned).

If a GM creates a linear branching story in advance of the players making those choices, then it would be an authored style. The depth of that tree can vary (and that might be useful for differentiating degrees) but in the end the GM is creating a choice node with determined options and then placed choice nodes for those determined options. This minimum model could be described as a tree of depth 2 (root node and its leaves).

If a GM waits for the players to create choices and create what responses they want to consider for those choices, then it would be an emergent style. The GM would still vet/adjudicate the options with respect to the verisimilitude of the setting. It would be a stretch to call this a tree of choice nodes except in hindsight. Once the choices are created and the options picked there will be an inevitable chain of "players made their 1st choice and chose to do __, then their 2nd choice and chose to do __, then their 3rd choice and chose to do __".

You get the difference between players being presented a network of paths vs carving their own path.

Agency:
When a GM's prep is specific to each option, then the prep scales poorly with high number of options. When a GM preps independent of the number of options, that prep scales better with the number of options available. So prima facie, we would expect more agency from the model that prepares to adjudicate any reasonable response to any choice than from the model that presents a choice with determined options that lead to the next choices. However this is only prima facie, other variables can impact the result. If a GM creates a linear branching tree of choice nodes, but only uses a short tree (depth 2-3 for example) that they extend between sessions then they can tailor each tree to create greater agency than they would provide with a long tree.

Event vs Location:
With the players creating the choices and their responses to the choices they raised, their movements can be unpredictable. It becomes more efficient to create content with a larger 4D footprint than content with a smaller 4D footprint when the players are in control of their 4D path. You have a bigger volume to fill with content, and bigger content helps. Event driven play usually has events with a smaller 4D footprint than the locations from Location driven play. This is not always true and sometimes the different is small. Consider a pirate that is raiding costal villages. In a game where the players are creating choices and their responses, how would you prep for this detail of the world?

Personally I would do some location based prep for each village (and the pirate ship, pirate cave, and costal waters) but would also do some even based prep for the pirate raiding these places at these times if not impacted by player choices. That way I would be prepared for the players if they do/do not go to these locations and if they do/do not arrive at the time of the events. I would expect the players might encounter a village after the raid and then decide to try to encounter the next raid (but why rely on guesses?).



I think this Authored vs Emergent distinction is about whether the GM create choice nodes with prepared leaf nodes OR if the Players create choices and then respond to those choices. In the former the GM knows the possible outcomes before the players know the choice. In the latter the GM knows the choice and the player's response after the players raise the choice and discuss their response.

[Stonehead]

The point that I'm trying to make it is that if you're the one fetching the juice, you ought to be able to tell if it was made by the juice machine or Steve, even if it makes no difference to the person who drinks the juice.

Applied to games, if you are the game master, you ought to be able to tell which elements of your game you authored and which emerged. On the level of entire games, the amounts and ratios of different elements can vary wildly, but on the level of moment-to-moment decision making, it really shouldn't be hard for a game master to identify which kind of process they are using. That's what I called Cheesegear out on: they were acting as if what a game master should do to achieve emergent game is obscure, yet immediately after they described two diametrically opposed methods of responding to unexpected player action.

I disagree with the juice analogy accurately representing the situation of DM style, for reasons I listed before.

The criteria described by him may have been diametrically opposed, but the criteria other people have used definitely have not been. A DM preparing a "default path", and a DM allowing player choices to change the direction of the story are not diametrically opposed.

Classifying entire games according to authored/emergent dichtomy may be fruitless; it's much more useful when it comes to looking at what a game master is doing to produce individual game elements.

See, this is the point I was talking about a few pages earlier. If classifying games like this is fruitless, then having a preference of one type of game doesn't make much sense. The counter argument to that was that the differences in GM style will bubble up to have different experiences at the table. But if that's true, then the games themselves can be categorized like this.

I think the real problem with this discussion is that several different metrics are being proposed to delineate between authored and emergent.

1. Dm supplied story vs player generated story
2. Event driven play vs location driven play
3. Low player agency vs high player agency

I've probably missed some and I don't think they correlate with each other. A "sandbox" might be player generated story, location driven play and high player agency but it also might be event driven as others have demonstrated. Or it could be low player agency if you can't actually effect the world in a meaningful way. On the other hand you can have a dm supplied story with very high level of player agency if the story is written as you go based on the decisions of the players. Where's does emergent fall on these metrics? Where does authored? Are they strictly the first metric? Does the first metric even make sense, or as suggested are all stories really supplied by the dm and the real difference is planned or improvised?

I think we need to narrow down the list of metrics so we can talk about a single metric at a time.

I think most people are using metric 1, but it does get complicated because there are others in this thread basically using completely different metrics. If I had to guess, I would say the three probably correlate some, but not strongly enough to be lumped together as the same concept.

So the question of whether there's a binary is basically 'is there in practice a phase transition between the river of history view and the chaos butterfly view, or do campaigns actually fall in between?'. If each step you take is drawn from the same statistical distribution and applies in the same way across all 'copies of past decisions' regardless of what they are, then its hard to avoid it becoming a phase transition since in one limit campaign trajectories converge on each-other exponentially, and in the other limit they diverge exponentially, and all of the continuum of 'how much emergence?' does is it changes the time constant of the exponential. exp(-0.001t) and exp(0.001t) become infinitely different when t->infinity.

On the other hand, there's no reason why a GM has to make statistically i.i.d. decisions about what past things should matter or not. You could have some 'privileged decisions' for example which are guaranteed to always matter and are never erased no matter what, which would mean that e.g. some particular session 1 choice still matters at t->infinity, but on average session 1 choices don't matter as t->infinity. You could have more exotic stuff as well, especially if you take into account players as active elements in keeping information about past choices alive rather than just relying on the GM for that.

So, I'd want to see the data. Because I can come up with various ways to be in the middle, but if people don't actually tend to GM like that it doesn't matter if some theoretical GM could do it just to spite the idea of there being a dichotomy.

This is the most convincing argument I've seen in favor of a binary. I'd have to see some actual data, just like you said, but the idea that in the space of all possible games, there are two big attractors around these two conceptual styles does seem possible. It doesn't describe the games I've been a part of, as a player or DM, but I've only played with a few dozen people over the course of my life. Hardly a representative sample. At the same time though, I've never gotten the impression from reading on the internet that I do anything particularly unique while DMing.

[NichG]

Every game prevents certain choices from entering the pool of "what gets carried forward" or filters it out as irrelevant. Every game whatsoever. Most even do so actively as things go--your character is acting at a different frequency than the player is. And you're not getting the full, unfiltered stream from them. The whole point of abstractions is that they abstract information. Which necessarily filters some out as irrelevant. Because the information has to pass from the fictional world up through words (which are lossy and low-bandwidth) to players and then back downward. Information is lost[1] at every step of the way, no matter whether you're authoring things or letting things emerge.

This is not a distinction relevant here--even the most emergent game can't keep track of all or even more than a tiny fraction of the choices made. Even deciding which choices are salient is a filtering choice. And that must be done in order for the game to time-evolve. An authored game, as long as it's not authored in stone, in advance, can track those offhand conversations just as well as an "emergent" one. Or even better, because a fully emergent game requires that the facts surrounding the slumming noble and his lands to have been generated (which may not actually be conducive to anything interesting later on), while the authored one can look back at that and plunk a suitably-configured "noble's lands" somewhere along the path that the party has decided to take, as a "plot hook" (optional or not).

Personally, I find that when both sides take responsibility for moving things onward, collaboratively, accepting that by doing so you are authoring material, players' choices matter even more meaningfully. Because you can really explore how those changes affect things. Where that gets utterly lost in the noise most of the time in a less-consciously[2] authored scenario, just because there's so many other things going on (in a good one, anyway). In a semi-authored collaborative game, the people involved can decide that X decision is the one they want to focus on and work together to explore it. Whereas in a fully emergent one, players can be at cross-purposes, the players can try to do something that just doesn't fit the world (causing dissonance), or just do other things and that possibly-interesting consequence just gets lost in the shuffle.

[1] and that's a necessary and important fact.
[2] everyone is constantly authoring things at least in the very short-term. Every act involves making a (conditional) plan for the future. Players and DM. If the next PC does X, the monster will do Y. If the players ask for X, he'll respond Y. Waiting until your turn to figure out what you're going to do is widely considered a bad thing, because it slows things down tremendously.



If that's what you got from it, one of the two of us miscommunicated. Generally, in many cases the consequences of individual events do average out. But not always. Just like generally, objects in motion remain in motion. But PCs can be and usually are "outside forces" that change the momentum of the system. The point was that, as I said above, every single game involves "averaging out" some events. No matter what. Period. Inevitably. The only question remains which events get averaged out and which can cause divergences that are noticeable above the noise.

Consider the following cases:
1. A character has brown hair. In a parallel universe, the same character dyes his hair slightly blonder. The character never does anything else notable. Does it make any sense that this would cause an arbitrarily large divergence in those universes? No.
2. A character burps while out on the road, with no one else around but the PCs. Does that cause a divergence in timelines of arbitrary magnitude? No.

Basically, there are a constant multitude of events going on. But the vast majority of those "microstates" end up in the same equilibrium macrostate, to use the statistical mechanics analogy. Most historical events are over-determined--there is no single but-for cause. If one person hadn't pulled that trigger (to start the war in the fantasy kingdom), someone else, somewhere close would have. Because the situation was that unstable. PCs, by nature of the beast, tend to be near critical points. So more of their actions tend to cause divergences than your average person. But that doesn't mean all their actions cause divergences.

Basically, the universe, because it's such a huge system, tends to be very near the thermodynamic equilibrium (even though that's a dynamic, evolving equilibrium for many reasons). And those equilibria tend (but are not always) to be relatively stable. Chaotic systems behave this way as well--chaos is deterministic and even predictable within certain bounds. Take your classic period-doubling experiment. Below certain values of the critical parameter, you get one and only one solution (a single period). Between slightly larger values, you get two and only two values. Which one you're in for any initial state is difficult to tell, but you have a 50% chance of being in either one. And a 0% chance of being in any other state. For slightly larger values, you get 4 possible periods. Again, which one follows from any given starting state is basically impossible to determine with more than 25% accuracy, but you can guarantee you're in one of those states and no other. Etc. And you can actually predict with exactness the values of the parameter at which each period-doubling occurs and what the possible states are. So given the value of the parameter, you can with 100% accuracy predict the set of possible states the system can be in. But you can't predict (better than chance) which individual state the system is in.

Chaos is not "any change in initial conditions makes an arbitrarily large divergence asymptotically". All it means is that arbitrarily small differences can, sometimes, under the right conditions, cause unpredictable (based on the knowledge of the initial conditions) changes in the behavior of the system." Most chaotic systems get entrained very easily and settle down to relatively stable states. The only thing is you can't predict which (of the possible) states any given set of initial conditions + perturbations will lead to. Chaotic =/= unstable equilibrium. Chaotic == hard to predict on long timescales.

Think of a directed percolation model. In directed percolation, you've got for example a grid of nodes that looks like a chess board rotated 45 degrees such that for each node, you could say go forward and to the left, or forward and to the right. Now, turn off some fraction of those links with a probability 'p'. In this version, imagine that the thing being propagated along the link is some impact of a choice made by a player at some particular past point in time - the node is '1' if history has been changed compared to what the GM imagined the history to be, or '0' if history is as the GM imagined it. The '1' values propagate to all connected nodes, whereas a node with no upstream '1' takes a value of zero.

Here 'p' is in a simplified sense the fraction of authored vs emergent moments in the game.

There is a threshold value of 'p' such that for a grid that is infinite in both space and time, below that 'p' all the nodes will become zero, and above that 'p' there will be some non-zero probability of finding nodes that still carry a 1. That doesn't mean that all nodes will carry a 1 in the latter case, or that if we were to make a multi-color version all colors would still not be extinct as x,t -> infinity, but if you're below that threshold 'p' then with probability 1, all of the nodes will become zero. For the square grid I described in 2d, it turns out that p has to be about 0.645 to have a non-zero chance of an influence infinitely far in the past still mattering to the present.

That's an example of a model that leads to the binary outcome even with a continuous range of styles.

[PhoenixPhyre]

Think of a directed percolation model. In directed percolation, you've got for example a grid of nodes that looks like a chess board rotated 45 degrees such that for each node, you could say go forward and to the left, or forward and to the right. Now, turn off some fraction of those links with a probability 'p'. In this version, imagine that the thing being propagated along the link is some impact of a choice made by a player at some particular past point in time - the node is '1' if history has been changed compared to what the GM imagined the history to be, or '0' if history is as the GM imagined it. The '1' values propagate to all connected nodes, whereas a node with no upstream '1' takes a value of zero.

Here 'p' is in a simplified sense the fraction of authored vs emergent moments in the game.

There is a threshold value of 'p' such that for a grid that is infinite in both space and time, below that 'p' all the nodes will become zero, and above that 'p' there will be some non-zero probability of finding nodes that still carry a 1. That doesn't mean that all nodes will carry a 1 in the latter case, or that if we were to make a multi-color version all colors would still not be extinct as x,t -> infinity, but if you're below that threshold 'p' then with probability 1, all of the nodes will become zero. For the square grid I described in 2d, it turns out that p has to be about 0.645 to have a non-zero chance of an influence infinitely far in the past still mattering to the present.

That's an example of a model that leads to the binary outcome even with a continuous range of styles.

Defining "p" as the "amount of authoring" is really stealing a base here. It's defining the conclusion (that this model represents authoring) as a premise. I could just as well say that "p" represents the natural damping effects of the environment (see below) and reach a very very different conclusion with the same model. And I'd say that that's a much more natural probability factor. Each event is carried over with probability p (which is really not a constant but an event-and-environment-dependent factor for each node) that depends on the rest of the system and the events.

If you're going to interpret it as the fraction of authoring...well...each event at the appropriate scale is either authored or not. So p = 0 or p = 1 for an individual node. Not some constant for the whole thing. Because events are propagated or not based on the individual event, not some global average value.

Sure, you can build such a model. But that doesn't mean that model means anything in this context. Or is operable. No game can use all the actions make by the characters infinitely. Every game will filter out low-impact events in some meaningful way (forcing signals below a certain value to zero).

And actually, that model shows how things can die off (ie be "averaged out"). If the replication value is low enough, any starting signal goes away eventually. The reverse is not true--if you have a value above the critical value, you can still have a starting fact die out.

Games naturally have damping factors built in. Lossy information flow. Because words and people are inherently lossy. Information inevitably will leak and be dropped. And things tend to return to the status quo ante (tend, not are forced to). Because the world's history is feedback-regulated. Even loud transients will tend to damp out, unless they're at a critical point that changes the behavior of the whole system. But even then, generally any loud signal at those points will tend to cause the behavior to change. And many microstates (ie different choices) tend to converge on the same asymptotic macrostate.

Very few signals in lossy environments are self-sustaining and infinitely propagating. That's rather the definition of a lossy environment. And a TTRPG (and even the in-universe environment) is rather the epitome of a lossy environment. For lots and lots and lots of reasons. The basic game abstractions cause massive information loss/filtering. And do so differently in different games. And they also impose "distortions" on the mindsets of the people involved.

TL;DR--information hiding/filtering is not a characteristic of authored games. It's an inherent characteristic of all lossy environments, which is everything humans do (especially via language). Because humans (and especially language) are lossy mechanisms. Heck, most of what people think they perceive isn't actually being actively perceived--it's munged together lossily from a combination of perception, memory, preconceptions, etc. And memory is even more so--you can have uncaused events because someone remembered that something had happened that had never actually done so. The classic example is a person waking up from a dream in which their partner had been unfaithful and being mad at their partner for something that never happened. So not only can you have choices die off, you can have events spring out of literal vapor.

[NichG]

Defining "p" as the "amount of authoring" is really stealing a base here. It's defining the conclusion (that this model represents authoring) as a premise. I could just as well say that "p" represents the natural damping effects of the environment (see below) and reach a very very different conclusion with the same model. And I'd say that that's a much more natural probability factor. Each event is carried over with probability p (which is really not a constant but an event-and-environment-dependent factor for each node) that depends on the rest of the system and the events.

I'm mostly arguing for possibility here, not that this particular conclusion will end up being true of real games, because a lot of the recent discussion in this thread has been of the 'how could it be that this is binary, when I can name continuous examples?'. I do not know whether actual games fall into these attractors.

Edit: Maybe it makes more sense if I say that instead of thinking of 'p' as 'degree of emergence', if you make a hard (irrevocable) preemptive decision about aspects of what a future state will be, you are preventing there from being a link into that node due to things which happen between now and then. And if you make a preemptive decision to nudge things towards a particular future state, you are in turn reducing the influence of things which happen between now and then even if you aren't eliminating the possibility of that influence. So for a different distribution of GM-ing techniques, there will be different degrees of what you call 'lossiness', and the point of the directed percolation model is that there can be a threshold level of lossiness where the macro feel changes suddenly, even if at the micro level the approaches seem to differ only by tiny matters of degree or detail.

If you're going to interpret it as the fraction of authoring...well...each event at the appropriate scale is either authored or not. So p = 0 or p = 1 for an individual node. Not some constant for the whole thing. Because events are propagated or not based on the individual event, not some global average value.

Even if whether a link is present or absent has a p=0 or 1, there will be some average number of links that end up being present or absent when you look at the thing as a whole. The important point is that small, systematic differences in how likely someone is to allow a signal to pass can correspond to big differences in whether or not the signal gets through in the end when you accumulate those decisions over a long time.

Sure, you can build such a model. But that doesn't mean that model means anything in this context. Or is operable. No game can use all the actions make by the characters infinitely. Every game will filter out low-impact events in some meaningful way (forcing signals below a certain value to zero).

And actually, that model shows how things can die off (ie be "averaged out"). If the replication value is low enough, any starting signal goes away eventually. The reverse is not true--if you have a value above the critical value, you can still have a starting fact die out.

And someone can therefore reasonably express a preference, for example, for playing with a GM, table culture, style of approach, sense of verisimilitude which favor things not dying out, versus a GM, table culture, style of approach, sense of verisimilitude which will guarantee things will die out.

If someone says e.g. 'I think the realistic survival rate of decisions at this level of detail is basically zero', someone else can say 'well okay, but that aside, I would really like to play in a game where that survival rate is at least 1 in 10'. It's also fine for someone to say 'I don't really care about the survival rate of things like my character's hair color, can I trade that for a more coherent story?'.

[Quertus]

Lot's of people prefer action and drama to lore and character exploration.

Even in a game of pure combat murderhobos, one could argue that the Dragon (or Princess) should be an interesting combat element outside the 15 seconds while the Princess is abducting the Dragon.

Yes, I usually mean it more the way you took it, but it's true in a war game context, too.

Of course.

But you can acknowledge that it is a spectrum, right?

If you come upon someone during a dramatic event, that does greatly enhance both the excitement of the situation and one's ability to influence it, right?

Um... that's... historically, the absolute lowest point on my "ability to influence" graph.

If I know about the event ahead of time, I can plan.

If I'm there, right then, I lose the element of surprise, and I'm a known variable in the equation going forward.

If I know about the event after the fact, I am not known to be involved by the relevant players, and can act with information gained through investigation, often without the parties knowing I'm involved.

So... actually... "showing up right as the Princess abducts the dragon", gotta say, greatly reduces my ability to influence the final outcome.

Nonetheless, I will concur more broadly that "positioning" in space-time matters.

In most fiction there is at least one unlikely coincidence to serve as an inciting incident. In real life, most people who change history appear to do so by being in the right place at the right time.

Contrast "inciting event" to my "time and again" / repeatedly / whatever I actually said.

Yes, "let's play the characters who happen to be on the Titanic", "Let's play the characters who happen to find the One Ring", "let's play the characters who happen to invent gunpowder"? Any one of those is a perfectly valid start condition for a campaign.

"Let's play the characters who happen to invent gunpowder while on the Titanic, and, due to a gunpowder accident, happen to blow a hole in the ship and sink it right when it's over the spot where the One Ring lies?" Um... maybe? That's piling on a lot of coincidence to our start conditions.

"Now that we're half-way through the campaign, let's happen to... [invent gunpowder / be on the Titanic / find the One Ring]"? This, I believe, is an "Authored vs Emergent" question. Allowing such mid-game coincidence allows a level of player-Authored content that I usually do not prefer, but could be fine at many tables (and is probably a hallmark of some systems).

"Let's play the characters who happen to invent Gunpowder, happen to be on the Titanic, and happen to find the One Ring, all in separate story arcs"? That's the "time and again" that I don't enjoy, that really breaks my sense of verisimilitude.

As Stonehead said in another thread, you can either be The Chosen One or you can be Bilbo.

As I have said in previous threads; the greatest heroes will need to be competent, driven, brave, clever, and lucky. Because if they aren't, someone else will be, and they will be left behind. As I have said before, it always hurts my immersion and ability to take the world seriously when a group of PCs who aren't played especially smart or heroic are rewarded and treated like heroes for coming up with solutions that any group of average folk could also think up and accomplish.

In an authored story, the GM (or the mechanics of the game I suppose) can help weigh the scale in the PCs favor so it doesn't break disbelief quite so hard.

I suppose you could also say the players are natural prodigies who are just better than everyone else, but isn't there 4-6 once in a generation prodigies all being in the same time and the same place and happening to have similar goals like a medieval Justice League also a sort of GM FIAT?

You need some plausible in universe explanation for why the PC's have Gandalf's ear or how they can come up with the idea for a Macguffin that nobody else could, lest the scenario is still running on GM FIAT.

"Let's play the nobility / representitives of the major (good) powers of Middle Earth, here to discuss McGuffin of Evil."

"Cool. Imma angel wizard totes OP thingy, and I say we keep the McGuffin here with my bestest bud."

"Yeah, weighing in as your 'bestest bud', the GM just passed me a note, reminding me that, ever since you brought that bloody thing here, the BBEG has become uncomfortably interested in my 'hidden' fortress (that's kinda been in the same spot for thousands of years, and everybody who's anybody knows where it is, thus all the visitors you invited actually showing up), and, given that your OP build lost against the BBEG last time you met, there's no way I can hold him off. Sorry. You need a new plan."

"Ooh, ooh, over here! I've spent lots of time on architecture and siege defense, and I'm playing a representative of the strongest (good) stronghold in the world. I say we bring the party to my house, and watch as my siege engines rock their world. Also, would that McGuffin artifact would boost my stats up to anywhere near your levels?"

Didn't feel too much like Fiat to me that they should meet.

And I have trouble imagining a game where the players wouldn't be allowed to do such things.

I feel like a lot of the confusion in this thread comes from the fact that most people play games that are pretty close to the middle of the authored / emergent divide, and see anything that deviates significantly looks like a ridiculous straw-man.

Now, of course, the main plot is still happening. Unless the players and the DM have a talk about it, they will still need to factor Voldimort into all of their plans.

For example, in my current campaign the earth mage and the time mage have talked about pooling their powers to avert the volcanic eruption that destroyed the Old Empire a century before. This is a great and extremely ambitious plan, except that the main plot of the campaign is about someone in the far future (a PC from a previous campaign in fact) becoming omniscient and pruning any time lines where they do not come to power, so those two are directly at odds and the conflict needs to be resolved somehow either in or out of character.

Which is something that my more emergent campaigns have always kind of failed at; the players are really bad at taking a hint or reading clues, so the main plot always comes crashing down on them unawares and it looks like a screw job out of nowhere coming to spoil their fun.

What makes that "Authored", in your mind? And why do you conflate "sufficient / insufficient information to know what's going on" with "Authored vs Emergent"?

. .. ... OK, "Authored" had a "pre-planning" advantage, which... OK, I can see you could maybe instead of planning complex set pieces and balanced fights and whatever, you could instead plan out exactly how you'll feed the party information, and pre-verify with your Evil-overlord mandated 5-year-old advisor substitute(s) that said information is necessary and sufficient for your sane players to see the Plot coming a mile away.

And, if you had planned for the old shopkeep in mudville to mention fact X in session 3, then get killed in goblin raids in session 4, but the PCs have the reigns, and don't make it to mudville until session 5, you're at a loss for how to get them fact X?

Is that why you conflate "Authored" with "Informed PCs"?

My games do tend to run more toward the linear side, and everyone in my group prefers it that way.

Getting lost and bored is a much bigger risk than feeling railroaded imo.

Maybe if we were still in school when we could stay up all night and game all weekend, but right now nobody really has the time or inclination to design entire sandbox worlds and simply free form RP for hours and see if something fun emerges.

But yeah, those are totally goals that my players could engage with, and I would absolutely work with them to generate "authored" plots that give them the opportunity to do it and to explore the consequences and ramifications after the fact.

The goblin army is on the march, the evil princess is abducting dragons, some unknown threat is burning down temples of McGuffin, and somebody just introduced hula-dancing to the campaign world, but your players just sit there bored for hours?

Ok, if it isn't a new character, then we just go back to square one.

I am unable to process this statement. Huh? "not a new character" and "back to square one" don't belong in the same sentence, in my book.

For one brief moment of clarity, I understood what we were talking about here, decided it wasn't important (because, having lost context, I wasn't talking about the right topic anymore?)... and then forgot what I had just understood. So, unless my moment of clarity was wrong, just ignore this bit.

[Devils_Advocate]

Bit weird to discuss how much impact individual actions have on the overall course of the history of the universe. How much impact our actions have on what happens in our own lives seems a lot more relevant. The game is about the player characters, what they do, and what happens to them.

How big an impact the PCs have on the setting is orthogonal to the authored/emergent axis. An authored scenario could totally thrust the PCs into the role of determining the fate of a region by supporting one of two diametrically opposed factions. And, if anything, an emergent campaign is more likely to be about the pursuit of personal goals without large-scale consequence. I hope most of you agree that it's generally less of a plot contrivance for main characters not to be in a position to determine the fate of the world.

[NichG]

Bit weird to discuss how much impact individual actions have on the overall course of the history of the universe. How much impact our actions have on what happens in our own lives seems a lot more relevant. The game is about the player characters, what they do, and what happens to them.

How big an impact the PCs have on the setting is orthogonal to the authored/emergent axis. An authored scenario could totally thrust the PCs into the role of determining the fate of a region by supporting one of two diametrically opposed factions. And, if anything, an emergent campaign is more likely to be about the pursuit of personal goals without large-scale consequence. I hope most of you agree that it's generally less of a plot contrivance for main characters not to be in a position to determine the fate of the world.

A lot of these distinctions deal in counterfactuals. In 99 of 100 ways that a campaign plays out, the players may never choose to create characters or pursue goals that change the fate of the world. But the idea that, if they decided 'this is what we're interested in doing with this opportunity', there actually is a possibility of pursuing that direction is important. It creates a feeling of openness that makes bothering to engage in lateral thinking, to consider goals and relationships and details and such important rather than just fluff and window dressing.

If you know (as in, the GM outright tells you) - 'no matter what you do, this kingdom will fall to a barbarian horde in 4 in-game years' - then that implies that there will be avenues that could have been open in a real world which are being artificially closed, and in turn that can make a lot of things not feel worthwhile to bother with anymore, even if that's far away from what you might have wanted to try to play out. 'So you mean, even if we warned the kingdom ...? So you mean, we can't actually gather a mercenary band big enough to fight the horde given 4 whole years? So you mean, we can't actually get any allies or support no matter what we do? So you mean, we can't even assassinate the barbarians' leaders?'. It doesn't mean that e.g. the game where the players decide to be a bunch of fishermen trying to find the Lake King would ever go there, but it doesn't have to. If that kind of way of thinking is 'how the GM looks at the world', then maybe its not worth bringing up that idea of draining the lake, or making the in-character decision to give up fishing and go spend the time you have left with your family, or ...

[Devils_Advocate]

It seems that I expressed myself poorly. I didn't mean that it's more plausible for there to exist no possible series of actions whatsoever for a character to, e.g., destroy the world. Rather that, under normal circumstances, the odds of success are gonna be pretty low on that one. The character can try, and maaaaybe succeed. But, y'know, probably not.

I was thinking of scenarios where the likelihood of, e.g, the world being destroyed, strongly depends on the protagonists' decisions. Probably not even because they worked especially hard to be in that position, but because they were in the right place at the right time to stumble across the right plot coupons or whatever. I seem to recall that there's some webcomic with this sort of plot. ;)

In an absolute sense, we're all in a position to determine influence the fate of the world, and that's not a trivial point. The PCs should also be in that position in that sense. That's not what I meant.

[NichG]

It seems that I expressed myself poorly. I didn't mean that it's more plausible for there to exist no possible series of actions whatsoever for a character to, e.g., destroy the world. Rather that, under normal circumstances, the odds of success are gonna be pretty low on that one. The character can try, and maaaaybe succeed. But, y'know, probably not.

I was thinking of scenarios where the likelihood of, e.g, the world being destroyed, strongly depends on the protagonists' decisions. Probably not even because they worked especially hard to be in that position, but because they were in the right place at the right time to stumble across the right plot coupons or whatever. I seem to recall that there's some webcomic with this sort of plot. ;)

In an absolute sense, we're all in a position to determine influence the fate of the world, and that's not a trivial point. The PCs should also be in that position in that sense. That's not what I meant.

Well, putting the 'should' aside...

I take the authored vs emergent stance thing to really be about 'when is something decided?', not about the specific decision. So if something is legitimately difficult and the party tries to achieve it and fails, that could either be an authored or emergent outcome. But if the outcome (either success or failure) is decided first and the method is filled in, that's authored stance, whereas if people at the table just make decisions in the moment and it turns out that the conclusion is success or failure, that's emergent stance.

So the point of preferring the emergent stance over, say, the 'soft authored' stance of pre-deciding but being open to recalculation would be the idea that if someone has formed an opinion about whether something is plausible or not before seeing the actual details of what happens, that prior bias can push something that might be unlikely to succeed into something that actually has a zero chance to succeed because they might be abstracting away details which could matter based on the pre-analysis of the situation. So it's something like saying 'I don't want you to form an opinion about whether or not my plan might succeed until we've actually tried playing it out', except writ large and not just with respect to specific likely-to-succed or unlikely-to-succeed plans, but also subtle things like NPC attitudes and who does what when. The sort of mindset where, because you didn't plan anything in advance, when you think 'who would react to this situation?' a minor NPC who had a brief chat with a PC 10 minutes ago might end up being part of the action even though you wouldn't have thought that NPC relevant when prepping the week before.

In essence, do you let the chaotic stuff that's hard to predict be the thing that matters, or do you present a world where the big things are all convergent and even if the details become random through chaos, the fates of nations and the like are all conserved quantities? I don't think there's an objectively right answer when we're talking about a game - I see no reason why someone couldn't present coherent fiction from either perspective.

[Talakeal]

Bit weird to discuss how much impact individual actions have on the overall course of the history of the universe. How much impact our actions have on what happens in our own lives seems a lot more relevant. The game is about the player characters, what they do, and what happens to them.

How big an impact the PCs have on the setting is orthogonal to the authored/emergent axis. An authored scenario could totally thrust the PCs into the role of determining the fate of a region by supporting one of two diametrically opposed factions. And, if anything, an emergent campaign is more likely to be about the pursuit of personal goals without large-scale consequence. I hope most of you agree that it's generally less of a plot contrivance for main characters not to be in a position to determine the fate of the world.

This is more or less what I have been saying this whole time.

It seems that I expressed myself poorly. I didn't mean that it's more plausible for there to exist no possible series of actions whatsoever for a character to, e.g., destroy the world. Rather that, under normal circumstances, the odds of success are gonna be pretty low on that one. The character can try, and maaaaybe succeed. But, y'know, probably not.

I was thinking of scenarios where the likelihood of, e.g, the world being destroyed, strongly depends on the protagonists' decisions. Probably not even because they worked especially hard to be in that position, but because they were in the right place at the right time to stumble across the right plot coupons or whatever. I seem to recall that there's some webcomic with this sort of plot. ;)

In an absolute sense, we're all in a position to determine influence the fate of the world, and that's not a trivial point. The PCs should also be in that position in that sense. That's not what I meant.

One thing about the whole railroading spectrum(s) that I think most people miss is that it goes both ways.

Some DMs fudge to deny the players agency, but many DMs also fudge to give the players more agency, usually by way of narrative contrivance or the "rule of kewl". A totally fair impartial GM who runs the world as a perfect simulation according to the setting and the games rules means that the players are going to need to actually stand out on their own merits (or get incredibly lucky) to accomplish their goals.

Um... that's... historically, the absolute lowest point on my "ability to influence" graph.

If I know about the event ahead of time, I can plan.

If I'm there, right then, I lose the element of surprise, and I'm a known variable in the equation going forward.

If I know about the event after the fact, I am not known to be involved by the relevant players, and can act with information gained through investigation, often without the parties knowing I'm involved.

So... actually... "showing up right as the Princess abducts the dragon", gotta say, greatly reduces my ability to influence the final outcome.

Nonetheless, I will concur more broadly that "positioning" in space-time matters.

Well... that does track with your 'self made man' narrative.

But I personally think that opportunity and coincidence are pretty big factors in outcome. Especially if you have no reason to know / care about the events you need to influence to get your way.

"Let's play the characters who happen to invent Gunpowder, happen to be on the Titanic, and happen to find the One Ring, all in separate story arcs"? That's the "time and again" that I don't enjoy, that really breaks my sense of verisimilitude.

That's just serialized fiction though.

Yeah, its kind of unrealistic that something interesting happens every week, but its a heck of a lot better than the alternative.

Heck, just surviving adventures weeks on end is pretty unrealistic even in a totally emergent game.

"Let's play the nobility / representitives of the major (good) powers of Middle Earth, here to discuss McGuffin of Evil."

"Cool. Imma angel wizard totes OP thingy, and I say we keep the McGuffin here with my bestest bud."

"Yeah, weighing in as your 'bestest bud', the GM just passed me a note, reminding me that, ever since you brought that bloody thing here, the BBEG has become uncomfortably interested in my 'hidden' fortress (that's kinda been in the same spot for thousands of years, and everybody who's anybody knows where it is, thus all the visitors you invited actually showing up), and, given that your OP build lost against the BBEG last time you met, there's no way I can hold him off. Sorry. You need a new plan."

"Ooh, ooh, over here! I've spent lots of time on architecture and siege defense, and I'm playing a representative of the strongest (good) stronghold in the world. I say we bring the party to my house, and watch as my siege engines rock their world. Also, would that McGuffin artifact would boost my stats up to anywhere near your levels?"

Didn't feel too much like Fiat to me that they should meet.

True... but on the other hand... those are not level 1 characters. And it is a pretty weird coincidence that the heirs to every kingdom in Middle Earth just happen to all be amongst the greatest warriors of their age. But then, this is a world that runs on the logic of an epic sage, and which actively has an all powerful god pulling the strings behind the scenes. Does your game world of choice have a similar set up?

What makes that "Authored", in your mind? And why do you conflate "sufficient / insufficient information to know what's going on" with "Authored vs Emergent"?

. .. ... OK, "Authored" had a "pre-planning" advantage, which... OK, I can see you could maybe instead of planning complex set pieces and balanced fights and whatever, you could instead plan out exactly how you'll feed the party information, and pre-verify with your Evil-overlord mandated 5-year-old advisor substitute(s) that said information is necessary and sufficient for your sane players to see the Plot coming a mile away.

And, if you had planned for the old shopkeep in mudville to mention fact X in session 3, then get killed in goblin raids in session 4, but the PCs have the reigns, and don't make it to mudville until session 5, you're at a loss for how to get them fact X?

Is that why you conflate "Authored" with "Informed PCs"?



The goblin army is on the march, the evil princess is abducting dragons, some unknown threat is burning down temples of McGuffin, and somebody just introduced hula-dancing to the campaign world, but your players just sit there bored for hours?

The boredom is a result of lack of drive as well as lack of agreement. Most players are not proactive, that takes too much energy, and those players who are proactive get shouted down by groups who don't share their interests / goals / morality / what have you.

For example, in a previous campaign I had a scenario where we spent the entire session diddling our thumbs because in the face of an unstoppable rampaging horde one player wanted to die fighting, one player wanted to run away, and one player wanted to make an alliance with a lich; but each also objected to the other two plans so we had deadlock.

What I want more talking about though is that I have players who are paranoid and oblivious. So we will have a situation like:

Me: You hear rumors that an undead horde has destroyed town A.
Players: That's nice, we come up with a plan to take over town E.
Me: You hear rumors that the undead horde has destroyed town B.
Players: That's nice, we continue our plan to take over town E.
Me: You hear rumors that the undead horde has destroyed town C.
Players: That's nice, we continue our plan to take over town E.
Me: You hear rumors that the undead horde has destroyed town D.
Players: That's nice, we continue our plan to take over town E.
Me: The undead horde arrives in town E and begins destroying it!
Players: What? You just didn't want us taking over a town, so you pulled an undead horde out of your butt to screw us over! RAGE!

I am unable to process this statement. Huh? "not a new character" and "back to square one" don't belong in the same sentence, in my book.[/S]

For one brief moment of clarity, I understood what we were talking about here, decided it wasn't important (because, having lost context, I wasn't talking about the right topic anymore?)... and then forgot what I had just understood. So, unless my moment of clarity was wrong, just ignore this bit.

My point is that while yes, level 20 PCs can shake the world by virtue of their personal power, they have to get their first. That takes a long time, and most campaigns end at 20, if they get there at all. So if you are relying on high level power to get what you want out of the game, you are going to spend a lot of time playing a game you don't want for the possibility of a small time playing one you do.

[Satinavian]

Some DMs fudge to deny the players agency, but many DMs also fudge to give the players more agency, usually by way of narrative contrivance or the "rule of kewl". A totally fair impartial GM who runs the world as a perfect simulation according to the setting and the games rules means that the players are going to need to actually stand out on their own merits (or get incredibly lucky) to accomplish their goals.

That is true.
But instead of fudging it, it would be very much possible to make the characters just so powerful that their actions are more relevant than those of all the NPC bystanders. That is usually enough to stand out, you don't need to rely on the players being particularly cunning and lucky nor do you need to rely on contrievances if you just let them play the movers and shakers from the beginning.

[Talakeal]

That is true.
But instead of fudging it, it would be very much possible to make the characters just so powerful that their actions are more relevant than those of all the NPC bystanders. That is usually enough to stand out, you don't need to rely on the players being particularly cunning and lucky nor do you need to rely on contrivances if you just let them play the movers and shakers from the beginning.

That was, if I read correctly, also Quertus' suggestion.

As I said though, I don't really find the idea of a group of once in a generation prodigies all living in the same time and place and happening to have close enough goals, allegiances, and outlooks to form a lasting party to be any less unlikely than a group of more normal people being in the right place at the right time to stumble into fantastic circumstances, but ymmv.

[Satinavian]

That was, if I read correctly, also Quertus' suggestion.

As I said though, I don't really find the idea of a group of once in a generation prodigies all living in the same time and place and happening to have close enough goals, allegiances, and outlooks to form a lasting party to be any less unlikely than a group of more normal people being in the right place at the right time to stumble into fantastic circumstances, but ymmv.

If they are powerful enough to stand out, they will automatically be driven to interact.

Also depending on setting, power comes in many forms. It could innate personal power like in a superhero genre and the superhoroes of the area interacting when a threat comes that is too big for each one is pretty normal. Same works for fantasy and rare magic/superhuman abilities, except it is even easier for the general populance to ask the powerful members of the society to do something about the problem.
But it could also be political power. If the PCs are part of the nobility, it is extremely likely they got to know each other and formed relations even when growing up.


I mean, i have no problem playing normal people stumbling into stuff way beyond their paygrade. But i like that more for a survival scenario, i dislike to have those normel people presented with McGuffings and circumstances to solve the problem to make up for their innate normalness. When i play normal people, i want to play normal people and that quite often means just trying to run away/ call in the powerful NPCs to solve the problem.

[Devils_Advocate]

Well, putting the 'should' aside...

Ah, right... "Realistically would"? "Verisimilitudonously would"? Eh, you get the idea, hopefully. Expectant "should", not moral "should".

In essence, do you let the chaotic stuff that's hard to predict be the thing that matters, or do you present a world where the big things are all convergent and even if the details become random through chaos, the fates of nations and the like are all conserved quantities? I don't think there's an objectively right answer when we're talking about a game - I see no reason why someone couldn't present coherent fiction from either perspective.

One's credulity is strained roughly in proportion to how easy it is to destroy the world multiplied by how long the world has existed. Even if it can only be destroyed by deliberate action, it seems unlikely that there haven't been at least a few people who want to destroy the world. Beyond some point, things do start to look a bit incoherent. "A just and loving creator would never make such a thing. A cruel, malevolent one would simply have made the air out of acid. And it wouldn't have evolved on its own, since there's no natural advantage to living in a world poised on the brink of annihilation!"

Deciding what happened in the past based what makes sense for the setting necessarily involves some "backwards causality", because any past that doesn't lead to the present is invalid. Perhaps "anthropic reasoning" is more applicable than "backwards causality". It's a matter of adjudicating as impartially as possible what set of circumstances is in place. It's not a matter of predetermining outcomes; to the contrary, beyond some point it becomes implausibly difficult to destroy the world.

Basically... The GM having no preconceptions about likelihood of success is only compatible with tasks happening to be legitimately difficult. It's incompatible with some tasks needing to be difficult in principle in order for the setting to not fall apart. If there's not really much preventing kings from being assassinated and there are lots of powerful people who want to assassinate any given king, it follows that kings are getting assassinated left and right. And, hey, maybe they are! But if that's incompatible with what's already been established, then by implication kings have means in place to prevent assassinations, even if you don't know specifically what they are. And "I'll fill in that blank when and if it becomes relevant" has to be compatible with emergence, because it's impossible to detail everything that might be relevant beforehand if anything in the world might be relevant. Unless "You can't have an emergent game without fully detailing everything in the entire world" is just straight facts and not a strawman, and no actual game is emergent.

How much sense am I making?

A totally fair impartial GM who runs the world as a perfect simulation according to the setting and the games rules means that the players are going to need to actually stand out on their own merits (or get incredibly lucky) to accomplish their goals.

That highly depends on what their goals are. Granted, if you're staying true to the default genre in D&D, the PCs' pursuit of their goals tends to involve a series of deadly encounters one way or another. And anything where that seems like a reasonable approach pretty much by definition is also going to be difficult at best to accomplish via safer means. Still, some games live outside the default genre. Some games aren't even D&D.

Yeah, its kind of unrealistic that something interesting happens every week, but its a heck of a lot better than the alternative.

Something interesting happening every week doesn't require a string of unrelated implausibilities.

"Long ago, there was a vast, prosperous empire, powerful both martially and magically. But for reasons yet unknown, that empire collapsed, leading to a dark age of ignorance, violence, famine, and pestilence. After long centuries, we have begun to rebuild what has been lost, but there remains much work to do. The growing wealth of civilization increasingly attracts raiders from afar. Foul, sinister things lurk in the ruins of a past golden age, and those same ruins contain vast treasure and objects of great magical power. Ambitious lords plot conquest, ever seeking to turn the latest development into a personal advantage through some scheme."

That's a setting where it would be unlikely for interesting things not to happen on a regular basis. The whole "world of adventure" setup may itself seem implausible, but (a) there's a difference between constructing an interesting scenario before play even begins and continually contriving new stuff to be interesting (because the setting isn't interesting on its own, or there would be no need for this), and (b) it really isn't. There are significant periods of history with lots of significant events that people will later tell lots of stories about. I don't know how many works of fiction are set during World War II, but I'm estimating "a lot".

And even that is treating "interesting" as "exciting". There are lots of other ways that something can be interesting. Maybe there are hints about the nature of some mysterious phenomenon; that's interesting because it's intriguing. Or maybe something is of interest to the player characters just because it connects to something that one of them cares about.

But I think that we're living in a hella interesting period of history now, so anyone who wants to "escape the boring real world" is gonna have different standards. In particular, if a player craves constant novelty in the form of repeatedly encountering new and different things, but also doesn't want to have to put any work in to seek them out... that pretty much is going to require a string of contrivances.

[NichG]

Ah, right... "Realistically would"? "Verisimilitudonously would"? Eh, you get the idea, hopefully. Expectant "should", not moral "should".

One's credulity is strained roughly in proportion to how easy it is to destroy the world multiplied by how long the world has existed. Even if it can only be destroyed by deliberate action, it seems unlikely that there haven't been at least a few people who want to destroy the world. Beyond some point, things do start to look a bit incoherent. "A just and loving creator would never make such a thing. A cruel, malevolent one would simply have made the air out of acid. And it wouldn't have evolved on its own, since there's no natural advantage to living in a world poised on the brink of annihilation!"

Deciding what happened in the past based what makes sense for the setting necessarily involves some "backwards causality", because any past that doesn't lead to the present is invalid. Perhaps "anthropic reasoning" is more applicable than "backwards causality". It's a matter of adjudicating as impartially as possible what set of circumstances is in place. It's not a matter of predetermining outcomes; to the contrary, beyond some point it becomes implausibly difficult to destroy the world.

Basically... The GM having no preconceptions about likelihood of success is only compatible with tasks happening to be legitimately difficult. It's incompatible with some tasks needing to be difficult in principle in order for the setting to not fall apart. If there's not really much preventing kings from being assassinated and there are lots of powerful people who want to assassinate any given king, it follows that kings are getting assassinated left and right. And, hey, maybe they are! But if that's incompatible with what's already been established, then by implication kings have means in place to prevent assassinations, even if you don't know specifically what they are. And "I'll fill in that blank when and if it becomes relevant" has to be compatible with emergence, because it's impossible to detail everything that might be relevant beforehand if anything in the world might be relevant. Unless "You can't have an emergent game without fully detailing everything in the entire world" is just straight facts and not a strawman, and no actual game is emergent.

How much sense am I making?

The reason to put aside the 'should' is because ultimately we're talking about ways of expressing preferences. It's perfectly fine for someone to say 'you know what, all of this argumentation about how realistic the world is isn't really important to me - it doesn't impact my ability to enjoy the game'. You can say e.g. 'it's easier to ensure realism by taking an authored approach', but that's not the same as saying 'therefore, an authored approach is the only reasonable way to play', because not everyone values realism or verisimilitude above other considerations. Might the result of playing things out be that settings are often more precarious than 'makes sense' according to some logic? Sure. Will that actually matter to a given player? Ask them!

Similarly, different GMs have different skill sets, so one GM using all the tricks of authoring may still fail to produce a setting as realistic as another GM who runs from a fully emergent stance, but who for example has more experience running games, or who has studied more history and can recognize the potential pain points more quickly.

So I don't think the 'should' direction is at all productive to discuss here...

Anyhow, as far as 'deciding the past', the line that I drew for this (which no one so far has explicitly disagreed with, but I have to allow that they might...) is that decisions about things before or within the present are valid from an emergence stance, its specifically OOC decisions about things in the future (rather than e.g. NPCs attempting to pursue goals) that are not within the spirit of emergence. So you leave the entire setting blank and fill things in as they become relevant? Sure, that can be emergent stance. But if you fill something in a certain way because you OOC want something to happen in the future, that's authored stance.

[PhoenixPhyre]

But if you fill something in a certain way because you OOC want something to happen in the future, that's authored stance.

But what if you fill something in a certain way because you OOC want something to be able to happen in the future (ie become more probable or even possible at all). Without stating that it will happen (either in any particular game or even overall)? Note that this sort of thing happens all the time--deciding that elves are a playable race is deciding that someone might play an elf, with all that entails. I'd suggest very few people actually plan that things will happen, instead they plan in ways that increase the probability of certain paths being taken later on.

Or what if you set up some cosmological event. It's a fact of the setting. But that cosmological event has observable consequences (ie "events") that happen on some regular, repeating timescale. Something like "On the last day of the 10th month between sundown and sunrise, the dead walk the lands of the living" (aka halloween, in its original incarnation). You've set up something that will happen, no matter what the PCs do anything (short of blowing up the setting entirely or radically changing the cosmology, neither of which is necessarily in the scope of their agency anyway).

I ask, because I do a lot of worldbuilding based on principles like this. Seeing where the intersection of fictional physical laws and the things players decide to do causes fun stuff to happen. And exactly what happens (ie "what do they do with the fact that the dead walk on that night?") is completely undetermined. But there's just a cosmological fact.

Additionally, where I see this hard distinction breaking down utterly is in living settings (like the one I run). One group can make a change to the setting by their actions that ripples out and will affect any contemporaneous groups. Like the fact that one of my groups dethroned a god. The other group, once their world clock hits that point (the beginning of 251 AC) will be exposed to that event and its fallout. In this case, what was very emergent for one group will be utterly authored and set-in-stone for the other. They, like everyone else in the setting at that point, will have a front-row seat for this event (for divine shenanigan reasons it was broadcast to literally everyone alive, simultaneously). But it's a done deal--one of the core principles of the setting (OOC) is that I don't retcon one group's actions based on another group's actions (which takes some juggling to keep separate[1]).

[1] the times where there have been conflict, I've been forced to fork the canon timeline and rejoin once the actions and their reasonable consequences no longer are in conflict. Both happened and technically both are divergent timelines, but I've yet to have one where the conflict was big enough to ripple out significantly beyond a few sessions, so I can patch the events back in later. Note that the forking and patching happens in retrospect, after both campaigns are concluded (while I'm prepping for the next group of campaigns).

[NichG]

But what if you fill something in a certain way because you OOC want something to be able to happen in the future (ie become more probable or even possible at all). Without stating that it will happen (either in any particular game or even overall)? Note that this sort of thing happens all the time--deciding that elves are a playable race is deciding that someone might play an elf, with all that entails. I'd suggest very few people actually plan that things will happen, instead they plan in ways that increase the probability of certain paths being taken later on.

I'd still call it authored stance. Whether that's a sufficiently strong authored mindset to push the game as a whole over to the post-hoc 'authored' attractor that kyoryu proposed to exist would be a different question. Given that the model for that attractor is as a phase transition, any finite number of authored or emergent moments shouldn't be enough to decide that, only overall average trends in taking the authored or emergent stance with regards to the game (possibly weighted by the scale and magnitude of those decisions, etc). And that's only if that phase transition actually does check out in the end, which may not be the case.

Or what if you set up some cosmological event. It's a fact of the setting. But that cosmological event has observable consequences (ie "events") that happen on some regular, repeating timescale. Something like "On the last day of the 10th month between sundown and sunrise, the dead walk the lands of the living" (aka halloween, in its original incarnation). You've set up something that will happen, no matter what the PCs do anything (short of blowing up the setting entirely or radically changing the cosmology, neither of which is necessarily in the scope of their agency anyway).

Did you set it up 'in order to have it happen' - e.g. was this an intentional thing trying to bias the future a certain way - or did you make a choice about a past that ends up having future consequences, without regards to those future consequences? If because of events in the meantime, the consequences of that cosmological event were the reverse of what you actually thought of when making that choice, are you going to see that as a problem or disruption, or is it a neutral outcome to you?

I ask, because I do a lot of worldbuilding based on principles like this. Seeing where the intersection of fictional physical laws and the things players decide to do causes fun stuff to happen. And exactly what happens (ie "what do they do with the fact that the dead walk on that night?") is completely undetermined. But there's just a cosmological fact.

'Authored' doesn't mean 'bad'. You're allowed to take the authored stance. You're allowed to mix them. None of these things are a moral position...

Additionally, where I see this hard distinction breaking down utterly is in living settings (like the one I run). One group can make a change to the setting by their actions that ripples out and will affect any contemporaneous groups. Like the fact that one of my groups dethroned a god. The other group, once their world clock hits that point (the beginning of 251 AC) will be exposed to that event and its fallout. In this case, what was very emergent for one group will be utterly authored and set-in-stone for the other. They, like everyone else in the setting at that point, will have a front-row seat for this event (for divine shenanigan reasons it was broadcast to literally everyone alive, simultaneously). But it's a done deal--one of the core principles of the setting (OOC) is that I don't retcon one group's actions based on another group's actions (which takes some juggling to keep separate[1]).

Running the same world for two groups at the same time, even when reality forces that to have a temporal order, is basically like running a game with time travel. So no shock that making it work might require an authored stance. Which, again, is fine. That's not a hard distinction between stances breaking down, because a kind of game not always being emergent doesn't somehow invalidate the term.

[PhoenixPhyre]

I'd still call it authored stance. Whether that's a sufficiently strong authored mindset to push the game as a whole over to the post-hoc 'authored' attractor that kyoryu proposed to exist would be a different question. Given that the model for that attractor is as a phase transition, any finite number of authored or emergent moments shouldn't be enough to decide that, only overall average trends in taking the authored or emergent stance with regards to the game (possibly weighted by the scale and magnitude of those decisions, etc). And that's only if that phase transition actually does check out in the end, which may not be the case.

Did you set it up 'in order to have it happen' - e.g. was this an intentional thing trying to bias the future a certain way - or did you make a choice about a past that ends up having future consequences, without regards to those future consequences? If because of events in the meantime, the consequences of that cosmological event were the reverse of what you actually thought of when making that choice, are you going to see that as a problem or disruption, or is it a neutral outcome to you?

'Authored' doesn't mean 'bad'. You're allowed to take the authored stance. You're allowed to mix them. None of these things are a moral position...

Running the same world for two groups at the same time, even when reality forces that to have a temporal order, is basically like running a game with time travel. So no shock that making it work might require an authored stance. Which, again, is fine. That's not a hard distinction between stances breaking down, because a kind of game not always being emergent doesn't somehow invalidate the term.

But all of that squeezes the available space for "emergent" gameplay down to a tiny tiny area. Which makes them very much not on equal terms. Of the entire phase space, emergent games (by this definition) occupy effectively a single point. Because all the normal things a DM can do, even from a worldbuilding perspective independent of the game itself (which describes most of what you described as authored, at least as far as my world is concerned) are authored.

Effectively, the only way I can see you could even conceivably come close to this definition of "emergent" is to not plan anything at all. Which was decried as a strawman.

Build a random table? You're planning for things that could happen, with the intent that (conditional on the roll), they will happen. Build a world with elves? You're stating that elves exist, which constrains the future. Heck, the system itself constrains the set of possible actions and outcomes. Often in huge ways. So basically you're always firmly authored unless you're freeform and ad-libbing everything. That's...not satisfying or useful as a definition.

I can't see the utility of this definition under the stated concerns. And I'd say that any definition that calls "stocking the world with elves" or "running multiple groups" authored departs so far from the underlying common-language meaning of those terms as to mean that they're causing more confusion than the value that they provide.

And I'm struggling to see the value they're providing at all anymore.

Edit: maybe a clearer way of saying that is that I don't think you can separate "building because you want something to be able to happen" and "just building things". Building a setting is an intentional act, and one that necessarily implies making choices about what can and cannot happen in the future. From the instant you start your metaphysics (or choose one) you've constrained the future and stated that some things will happen and some things will not happen. And every single choice constrains the future further. So unless you're literally just making everything up on the spot without any thought at all (which I hope is a strawman), you're making choices with an eye for the future. There's never enough information in any given choice to uniquely determine the future, so you're always choosing which of the possible events will happen. Based on subjective judgement. "Neutrality" is a lie.

[NichG]

But all of that squeezes the available space for "emergent" gameplay down to a tiny tiny area. Which makes them very much not on equal terms. Of the entire phase space, emergent games (by this definition) occupy effectively a single point. Because all the normal things a DM can do, even from a worldbuilding perspective independent of the game itself (which describes most of what you described as authored, at least as far as my world is concerned) are authored.

No? Only games which are 'always emergent stance no matter what' are squeezed down to a point. But the same would be true of 'always authored stance no matter what' games. When it comes to mindset, I don't think GMs necessarily fall into one or the other 100% of the time and I said as much up thread. If the binary exists, its something that exists at the limit of an infinitely long campaign and not decision by decision.

What isn't clear is whether there's some threshold where games that are run 'mostly but not entirely in the authored stance' become distinctly 'authored games' and where games that are run 'mostly but not entirely in the emergent stance' become distinctly 'emergent games'.

In effect, the 'stances' are like a thermodynamic variable like temperature. The question is whether there exists something like a 'high temperature phase' and 'low temperature phase' of campaigns, with some particular temperature as the tipping point.

I can't see the utility of this definition under the stated concerns. And I'd say that any definition that calls "stocking the world with elves" or "running multiple groups" authored departs so far from the underlying common-language meaning of those terms as to mean that they're causing more confusion than the value that they provide.

When we're talking about the stances, the reason 'why' matters. If you put elves in a world (make a decision about the past) its not necessarily authored stance. It's authored stance if you put elves in the world because you have in mind some future thing involving elves that you need there to be elves to enable. But if you say 'okay, I'm just deciding there are elves because I like elves', that'd be emergent stance.

Do you consider 'determining the future' to be within your remit? If and when so, authored stance.

Edit: maybe a clearer way of saying that is that I don't think you can separate "building because you want something to be able to happen" and "just building things". Building a setting is an intentional act, and one that necessarily implies making choices about what can and cannot happen in the future. From the instant you start your metaphysics (or choose one) you've constrained the future and stated that some things will happen and some things will not happen. And every single choice constrains the future further. So unless you're literally just making everything up on the spot without any thought at all (which I hope is a strawman), you're making choices with an eye for the future. There's never enough information in any given choice to uniquely determine the future, so you're always choosing which of the possible events will happen. Based on subjective judgement. "Neutrality" is a lie.

I mean, I think this is really the heart of it. I've heard Quertus express something like (paraphrased) 'the greatest sin a GM can commit is to want'. Everything you choose influences the future, sure, but that's fine. Once you start to believe 'it's okay for me to us GM power to make things I want to happen happen' then, for some people, that takes the game in a direction they don't like. It may be that you're incapable of running a game that simultaneously would feel satisfying to you, and would appeal to that set of players. But, so what? Why take it personally that some people like a style of gameplay that doesn't make sense to you?

[PhoenixPhyre]

I mean, I think this is really the heart of it. I've heard Quertus express something like (paraphrased) 'the greatest sin a GM can commit is to want'. Everything you choose influences the future, sure, but that's fine. Once you start to believe 'it's okay for me to us GM power to make things I want to happen happen' then, for some people, that takes the game in a direction they don't like. It may be that you're incapable of running a game that simultaneously would feel satisfying to you, and would appeal to that set of players. But, so what? Why take it personally that some people like a style of gameplay that doesn't make sense to you?

And I think that Quertus's statement is meaningless. Everyone wants. They can't help it. Neutrality-as-the-highest-virtue is an illusion, and a toxic one. One that destroys the entire power of a TTRPG over a CRPG. If you want a DM who doesn't want...go play a computer game. Seriously. Because that's the only way you can get close...and even then you're dealing with the programmers' wants.

And I'm not taking anything personally--I'm saying that the proposed distinction doesn't actually describe anything meaningful in this context. Because treating it as a dichotomy removes all but a vanishing fraction of games from consideration. If the only way to get an "emergent" game is to have a DM that doesn't want...then you're not playing any game system I know of. Because even the DM-less ones have players in that role, and players want. The game system itself, even the ones like PbtA, explicitly tell the DM to want. PbtA has "be fans of the characters" as a prime virtue. And that inherently puts a finger on the scale.

Which means that if someone comes up and says they want an emergent game (under this definition), my reaction must be "sorry, go play somewhere else. Because what you want is impossible within the confines of a TTRPG. Can't get there from here. Incompatible with the existence of a game system as a thing."

Effectively, I reject the definitions as providing actionable information, because they collapse under their own weight into linguistic point singularities.

As I've said before, I think that the only meaningful, useful definitions here are as broad (and fuzzy) labels for halves of the "how is responsibility for generating content shared" spectrum. And will inevitably be scope-bound--people can and do want different values (differing in large or small amounts) at different scopes of play. My players, for instance, want to be deeply involved in creating content at the meta[1] and micro levels, but explicitly want me to do most of the heavy lifting at the world-building and "orchestration" levels. They mostly want to be able to decide how they react to things (or how they act generally) and want those actions to matter in ways that leaving it up to a neutral arbiter don't support. They want their backstories to matter (which an emergent game by the definitions provided cannot do); but they also want mysteries and prophecies and unconnected events to unfold. They want collaboration on creation of content. They want things to be a dance combining the OOC and the IC. Which doesn't fit at all into the dichotomy definition. They want me to want. Some things. But not others.

I'll repeat--every action involves intent. You can't "just build something". Every single choice is a choice about what will (or can) be and what will not (or cannot) be in the future. There is no other option. And you can't even pawn it off on someone else by using someone else's content--that just shifts the burden of the wanting. Any definition that revolves around not wanting or not caring about what will happen is one that cannot be applied to any human effort.

[NichG]

And I think that Quertus's statement is meaningless. Everyone wants. They can't help it. Neutrality-as-the-highest-virtue is an illusion, and a toxic one. One that destroys the entire power of a TTRPG over a CRPG. If you want a DM who doesn't want...go play a computer game. Seriously. Because that's the only way you can get close...and even then you're dealing with the programmers' wants.

Or you play a GM-less game. Or you don't demand perfection, but you do demand to play with other people who view that bias as a negative to be reduced when discovered rather than as a positive to embrace. Where the value is being able to say to the GM 'hey, this setup feels like you're trying to get us to go to X' and the GM says 'oh, damn, I bet you're right, my bad, I'll try to do that less' as opposed to 'yes, that's intentional, you should go to X' even if that particular thing isn't retconned.

And I'm not taking anything personally--I'm saying that the proposed distinction doesn't actually describe anything meaningful in this context. Because treating it as a dichotomy removes all but a vanishing fraction of games from consideration. If the only way to get an "emergent" game is to have a DM that doesn't want...then you're not playing any game system I know of. Because even the DM-less ones have players in that role, and players want. The game system itself, even the ones like PbtA, explicitly tell the DM to want. PbtA has "be fans of the characters" as a prime virtue. And that inherently puts a finger on the scale.

Well, you're doing a lot of comparing your own GM-ing to this dichotomy like having elements that would count as authored which you think are important for the quality of your games would mean that the definitions can't possibly be right... That reads as taking it personally (or at least making it personal?) to me.

Like, my GM-ing style isn't purist emergent. There are times when my players just aren't doing anything, don't know what to do, and I introduce elements to the game for the purpose of getting them moving, creating play when they can't create it themselves. That's authored stance. But the terms are still useful to me so that if someone said they wanted an emergent game I could point out the differences and say 'is this okay with you, or too much?', or understand ways that I could make my style more emergent for them, or other such things. I can still understand and empathize with someone who wants those things. I can figure out what would be necessary to bridge those gaps. That's useful stuff to me. It's actionable stuff. I'm not going to get hung up on corner cases because the ideas of the dichotomy give me a direction to move in. And if someone said 'I want a more authored experience than what you normally offer' I'd know what direction to move in then as well.

As I've said before, I think that the only meaningful, useful definitions here are as broad (and fuzzy) labels for halves of the "how is responsibility for generating content shared" spectrum. And will inevitably be scope-bound--people can and do want different values (differing in large or small amounts) at different scopes of play. My players, for instance, want to be deeply involved in creating content at the meta[1] and micro levels, but explicitly want me to do most of the heavy lifting at the world-building and "orchestration" levels. They mostly want to be able to decide how they react to things (or how they act generally) and want those actions to matter in ways that leaving it up to a neutral arbiter don't support. They want their backstories to matter (which an emergent game by the definitions provided cannot do); but they also want mysteries and prophecies and unconnected events to unfold. They want collaboration on creation of content. They want things to be a dance combining the OOC and the IC. Which doesn't fit at all into the dichotomy definition. They want me to want. Some things. But not others.

Sure, nothing wrong with any of that. You have a style that suits your group, and you didn't need these concepts to get there.

I think the 'who has responsibility for what' question is a different one, by the way. Understanding authored vs emergent as being about the direction of causation, I can express that as a GM, I would prefer my players to take an emergent stance rather than an authored stance towards their characters. That kind of expression is sort of 'mu' in the 'who has responsibility' axis. The players still have responsibility for their characters, but I want them to execute that responsibility in response to what happens in the game, rather than in anticipation of it. And I'm not going to pitch a fit if it isn't perfect, but if a player were to ask me between two ways of doing things I think I could use that guide to reliably pick which I'd prefer they take. And from a game design point of view, I can author (in the sense of this thread) things to encourage players to approach their characters from an emergent perspective - for example, by having hidden stuff in the world that would totally up-end optimal builds and the like. So in that sense, its not at all about responsibility, its a separate axis and one that does have some resonance to me.

I'll repeat--every action involves intent. You can't "just build something". Every single choice is a choice about what will (or can) be and what will not (or cannot) be in the future. There is no other option. And you can't even pawn it off on someone else by using someone else's content--that just shifts the burden of the wanting. Any definition that revolves around not wanting or not caring about what will happen is one that cannot be applied to any human effort.

Hard disagree on this, but its such a fundamental worldview issue that I don't really see a way to discuss it that won't either take 20 pages or hit a wall...

I can go to a canvas and put a stroke of paint on the paper just because I want to take that action. It doesn't have to be 'in order to paint a flower'. Once I see the stroke, I can place another. It might become a flower, or a sunset, or a space station, or whatever.

I can also go to a canvas and put a stroke of paint on the paper in order to paint a flower.

Those two modes of operation feel fundamentally different to me in the act. From the perspective of, say, building AI, one is a 'policy network' or Braitenburg vehicle framing, the other would be a world model or value function (Q-learning, etc) approach. In the end, the same sequence of actions could derive from either architecture, but the counterfactuals, the way they generalize, etc are all different.

But I also know people who would say that no matter what you do, every single thing you do - including emotion-driven acts, reflex acts, etc - comes first from 'having a goal'. I think that framing is really limiting, but its like a choice of coordinate system. You can technically do everything in that coordinate system if you don't mind really messy assignment of 'whose goal is it?' - maybe its not your goal, maybe its evolution's goal or something.

[PhoenixPhyre]

Hard disagree on this, but its such a fundamental worldview issue that I don't really see a way to discuss it that won't either take 20 pages or hit a wall...

I can go to a canvas and put a stroke of paint on the paper just because I want to take that action. It doesn't have to be 'in order to paint a flower'. Once I see the stroke, I can place another. It might become a flower, or a sunset, or a space station, or whatever.

I can also go to a canvas and put a stroke of paint on the paper in order to paint a flower.

Those two modes of operation feel fundamentally different to me in the act. From the perspective of, say, building AI, one is a 'policy network' or Braitenburg vehicle framing, the other would be a world model or value function (Q-learning, etc) approach. In the end, the same sequence of actions could derive from either architecture, but the counterfactuals, the way they generalize, etc are all different.

But I also know people who would say that no matter what you do, every single thing you do - including emotion-driven acts, reflex acts, etc - comes first from 'having a goal'. I think that framing is really limiting, but its like a choice of coordinate system. You can technically do everything in that coordinate system if you don't mind really messy assignment of 'whose goal is it?' - maybe its not your goal, maybe its evolution's goal or something.

The thing is that "goals" or "wants" are not single valued entities. They're...more like weighted sets. Even purely random actions (by any living actor) are driven at least in part by wants. But there can be a multitude of such wants including contradictory ones. I can simultaneously want to be thinner AND want to eat that candy. And it's not even that one "wins out", they're both in there. But you can't get away from wants. They are so innate in the nature of being alive as to almost be definitional.

-----

Furthermore, speaking in terms of neutrality of operation doesn't work because a CRPG is utterly neutral in operation. But the outcomes are all fixed in place. Every action, no matter what, constrains the future. And every action implicates intent. Your choices as a DM are
* put things you think your players will like
* put things you think your players won't like
* put things you think your players will be apathetic (neither like nor dislike) toward.

That's the entire set. And no matter what, you're picking one of those three options. Whether you think you are or not. Or you'll be perceived as doing so.

-----

My big point is that the definitions you're providing don't match anything like what I experience. Because naively I'd say I'm dominantly emergent--I don't plan plots, I don't even plan events until they're revealed. And I go by what feels right in most situations, often discarding my notes. But by the definitions provided, I'm nearly 100% authored because I care a lot about how the world works and what makes sense in that context and because every choice that has more than one possibility is based around what I believe the players will engage with most. Which tells me that the terms don't map to anything like my lived experience.

[NichG]

The thing is that "goals" or "wants" are not single valued entities. They're...more like weighted sets. Even purely random actions (by any living actor) are driven at least in part by wants. But there can be a multitude of such wants including contradictory ones. I can simultaneously want to be thinner AND want to eat that candy. And it's not even that one "wins out", they're both in there. But you can't get away from wants. They are so innate in the nature of being alive as to almost be definitional.

-----

Furthermore, speaking in terms of neutrality of operation doesn't work because a CRPG is utterly neutral in operation. But the outcomes are all fixed in place. Every action, no matter what, constrains the future. And every action implicates intent. Your choices as a DM are
* put things you think your players will like
* put things you think your players won't like
* put things you think your players will be apathetic (neither like nor dislike) toward.

That's the entire set. And no matter what, you're picking one of those three options. Whether you think you are or not. Or you'll be perceived as doing so.

I feel like I've said it a dozen times now, but just because a choice impacts the future doesn't mean that that choice was made because of its impact.

The three modes you're describing are all emergent stance.

The authored stance version of those would be: put things that will bring about things you think your players will like, put things that will bring about things you think your players won't like, put things that will bring about things you think your players will be apathetic towards.

And where I think we'll never agree is, you also have choices like:

* don't think about your players, just put that thing you saw on TV last night

Not necessarily good choices, but choices which are logically possible. Now, it might turn out that your players like, dislike, or are apathetic towards that. But there's a difference in the 'why?' you put the thing there. Some people care about the 'why' more than the 'what'. The same exact campaign setting, constructed according to different motivations, may indicate an environment someone is willing to play in or is not willing to play in, even if the specific choices are all exactly identical. Because at the end of the day, you're playing with the person who made those choices and you're going to be trusting them to make more choices going forwards.

My big point is that the definitions you're providing don't match anything like what I experience. Because naively I'd say I'm dominantly emergent--I don't plan plots, I don't even plan events until they're revealed. And I go by what feels right in most situations, often discarding my notes. But by the definitions provided, I'm nearly 100% authored because I care a lot about how the world works and what makes sense in that context and because every choice that has more than one possibility is based around what I believe the players will engage with most. Which tells me that the terms don't map to anything like my lived experience.

I mean, nothing says you can't be dominantly emergent except for a few things which are specific pain points for you.

Like, if the party does something that threatens to reveal a giant hole in a kingdom's economic system, the emergent philosophy there would be 'I presented a system, the PCs found a hole in it, therefore there was a hole in it and for whatever reason the PCs were the first people to think of exploiting it'. If the PCs find that goblins are changing galleons for gold at a rate vulnerable to arbitrage, then arbitrage away. From an emergent stance, if you need the economic system to not have holes, you have to make it correctly from the start and then you're committed to the consequences of that choice. But maybe while for the most part you're pro-emergence, you recognize that you're not going to quickly spin out a flawless economic system for a fantasy kingdom after thinking for 30 seconds, and the disruption of that kind of thing is big enough that it outweighs how much (if at all) you care about emergence vs authored stances. So then you say 'okay normally I'd say do whatever and I'll run what happens, but for the economy just pretend like I made it ironclad so there are no easy get-rich-quick schemes, and if you try to sell something back to someone for more than you paid its not going to work no matter what the rules say you can do'.

It'd then be a question, to players who care about the authored/emergent distinction, does that cross a line for them? Well, it might, or it might not, but that's for them to decide, no?

[Quixotic1]

The level of agency players have at the campaign, adventure, scene and round levels.

The way in which the players transition from one scene to the next.

Who is driving the narrative and who is reacting to it.

--I feel like those are all very different things, but they all get lumped together when people draw a line in the sand.

"Sandbox" and "railroad". These terms are so broad, so muddied and so emotionally loaded that they are worse than useless.

I don't think coming up with new terms will really help. I think we'd need to break them apart and talk about their components. At least that way, we might be able to set aside the parts that make people froth at the mouth and have an actual discussion.

And that discussion better directly relate to how I can get better at running games or give me a better understanding of story structure and stuff like that, or I don't really get why anyone would have it.