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

[PhoenixPhyre]

Yeah, though I think there is actually an empirical question buried in all of this rather than just a semantics one, which is whether or not there's an innate tendency for the influence of authorship or the influence of emergence to grow towards an extreme as one extends the campaign because of the way that iterative planning behaves. There's a sort of butterfly effect argument for one end of that existing - small amounts of emergence can ultimately lead to large deviations, if everything follows from the previous thing and generally stuff is interdependent (high-interaction regime). There's also a sort of zeitgeist of the alternate being true, that e.g. when people do speculative fiction or even just fanfiction, there's some river of history built out of larger scale realities that individual actions can't budge even through the lever of time.

So the empirical question is the degree to which these views or beliefs manifest in the action sorts of variabilities that exist when people run games? Does the distribution of impacts of particular player choices on the setting fall into a bimodal distribution across games people run, and is that just reflecting an underlying distribution of meta-beliefs about the nature of consequence, or does it arise from the actual mechanics of 'iterative planning'? Both? Neither?

That's I guess the thing that makes this interesting to me beyond the railroad/linear/sandbox/etc distinctions - feels like there's more potential for something here than just a way to badwrongfun at people.

I mean, depending on work, that could just be laziness.

if your a sci-fi author writing something in the vein of the Foundation novels, sure thats the intention. if your one of the many naruto fan fictions that make a change to naruto but some reason still does land of waves arc, its probably just laziness.

Personally, I'm more of the "history has momentum" position. Not insuperable momentum, but butterfly effects tend to damp out in both space and time. Change a character's hair color? Meh. No substantial change. Change a Naruto character's powerset in a significant way? Probably a bigger change. Change how chakra works? Huge change.

Effectively, changes tend to primarily affect their own level of detail unless they happen at a critical moment or at a critical point. 99 times out of a hundred, that penny on the railroad tracks mostly just gets flattened. But that one time...

I treat adventurers (specifically PCs) as catalysts of change. The setting exists, and the setting is mostly stable. At least with respect to the NPCs. But the PCs are an unpredictable force and we're playing to find out how things evolve because of them. But their changes are generally bounded--unless they move something deliberately or in a substantial way and set up a new stable equilibrium, the locals will generally go back to their old ways fairly soon.

Spoiler: digression

Show

This is important because I'm currently doing an arc where the PCs are involved with timey-wimey shenanigans. Involving three major timeline "bundles"[1] that have become shattered and intermingled in a local area. The PCs, it turns out, were actually instrumental in creating one of those timelines (by interfering with an attempted assassination). But before (subjectively) they did that, they'd already interacted with a detached fragment of that timeline bundle and brought it into their "primary" timeline. And that fragment was from their past but after the timelines split. So they must have split the timelines, yeah? It's fated, right? Not really. If they hadn't, someone would have. By interacting with the timelines, the fact that the timelines were split is "baked in" and has momentum. So the universe in order to maintain consistency, would demand someone do it.

And by doing so, they altered their own timeline via the fragments that had been shattered and brought in. Because the timeline that they sparked was slightly different than the one that someone else would have sparked. For example, one of the NPCs they'd interacted with in their main timeline was a horny skeeze of a human male, lusting after his (succubus) boss. After they made that change, the person filling that role is a (rather boring) triton female who has professional relationships with her (succubus????) boss. And a few other things changed. But their backstories are still conserved except for tiny little details.

[1] Every action splits a timeline, but most of them tend to converge into a few "bundles", with some fuzziness. Just like you can treat a piece of yarn as a single fiber for most macro-scale purposes, each "bundle" is really a myriad of timelines but in the macro-analysis all are basically the same.

This means that while their actions at the micro scale propagate, they don't do so endlessly. It's a damped system; ripples tend to die out. Each arc generally has 2-5 lasting changes that carry on to another arc. I don't know what those are going to be (or where the arcs will end or how they'll develop or what the next arc will be) before hand, but I know that once an arc is done I don't have to carry all the details along to the next one. Just a few "headline changes". And within an arc, events tend to develop fairly predictably and linearly--not because the players are forced into a path, but because they tend to stick with something until it has come to some sort of conclusion. If I played with more chaotic players (as opposed to characters, all the characters are always firmly on the chaotic end) that might be different.

/rambling

[Stonehead]

You clearly misunderstood what I was getting across there. The world is much bigger than the players.

Where the players are making choices and engaging with the world a variety of things happen that are related to what the players are doing.

A few hundred miles away, or one plane over, there's a skirmish between two forces on a shared border that happens despite what the players are doing - whose result is a die roll which then has an outcome which may or may not impact the players in the near term, but which has potential to present its result to them for a new situation X, Y, or Z in the medium to longer term ... or, they never go that way and it just happens off screen when their choices take them elsewhere.
It happened despite the players because the world keeps existing regardless of whether or not the players live or die, succeed or fail, retire or keep adventuring.
Yes, that is a play style, one I've seen and used since I first started playing RPGs.
You calling it "bad DMing" is simply wrong.

Maybe it was just word choice, but "despite" usually means "in spite of". Where the events happen even though the players made choices to prevent them. Like the scenario where the DM preps his boss fight on the top of the wizard's tower, so no matter how hard the players try to avoid it, they are going to end up on the wizard tower for the finale. I know it's subjective and all that, but that hypothetical I'm comfortable with calling bad DMing.

Your example is perfectly fine. I would probably say "independently" of the players, not "despite" them, but that's just semantics. Clearly I misunderstood your point.

Funny, as I read these discussions I see people who do not like the core assumption of GM based TTRPGs. That GM run TTRPGs are run on GM fiat by design.

They have fallen for the the Illusion of Choice so hard, that they think that games with a GM are somehow controlled by player choice, RNG, Story tropes, Rules, charts, etc. They are not. The GM is the only ones with power to shape the game.

Some GM based TTRPGs make the hand of GM Fiat more visible than others, while others try to obscure it or reign it in mechanically. However, the core of all of these games is still the GM making choices about things, including how much to let the players make choices.

I can only speculate that this deep seated desire to reduce GM Fiat as the primary design of GM based TTRPGs is because:

1. They had bad GMs that made the GM Fiat feel heavy and constricting on the game)

OR

2. They had really good GMs who made the GM Fiat feel weightless and invisible on the game


To me the distinction is not Railroad vs Sandbox, Authored or Emergent; it is how heavy the hand of the GM should be during game play. That is a spectrum and not easy to gauge, unless you are playing at the table with the group, and is subject to guess and check feedback loops.

This is really interesting to me. It seems like other people are describing the types of games they play, and you're saying those types of games can't exist. And you're saying the reason they think they exist, is because they haven't played in enough types of games.

If the DM "decides" to let the players have control over their own characters, and the players come up with some novel actions, it seems like quite a stretch to say those actions happened by DM fiat.

If someone at the table asks "Why did Galahad chase after those goblins?", then "DM fiat" seems like a really bad answer to that question. A good answer would involve the player talking about their character's motivations.

You could probably technically say a character's actions are a distant consequence of DM fiat, but that's like saying I made a sandwich because of the 4 fundamental forces of physics. It's just not a useful way to talk about anything.

[Cheesegear]

However, the core of all of these games is still the GM making choices about things, including how much to let the players make choices.

My favourite is when people try to twist the definition of 'Negating player choice' to not mean exactly what it says.

'Well that's not [the bad thing], that's just DMing.'

Okay, we agree. If [the bad thing] is just DMing, then all DMs do [the bad thing], because [the bad thing] and DMing are inextricably linked under their definition of DMing. Crazy.

Noooo.....

To me the distinction is [...] how heavy the hand of the GM should be during game play.

As in the last several threads, I'm going to agree here, too.

If all DMs do [the bad thing] (as I believe that they do):
- How often do they do it?
- For what purpose?

Like I said, in a sarcastic point; 'You start in [arbitrary location].' is already a loaded sentence. The very first sentence of almost any game already in the DM's favour;
I've never let my players choose where they start the game. Literally never. I strongly doubt any DM has, and if they have, I don't believe they do it regularly.

[Tanarii]

It's certainly an interesting view that the start conditions could indeed be considered a Railroad that leads from "agreeing to join and play in a game with start condition A" through "creating a character that would arrive at start condition A somehow" to "start condition A". Most people probably discount it because it's part and parcel of signing up for that game. (If it wasn't, the requirements to be allowed to sign up were not clear.) And getting off that "Railroad" is generally much less painful than one later on. Because usually you never get on it in the first place, because you just don't sign up for the game. (Unless it was misrepresented, at which point you might have wasted your time creating a character when you get off and walk out of the first session.)

But it's not one that's inherent to all TTRPGs. There are quite a few where the players effectively define the start condition, within parameters set out by the game. And even if there are parameters, they still get agency to choose among them, so there is no Railroad to get to the start condition.

An example of such a game is Mutant Year Zero. You get to design your Ark, you pick where on the map it goes, and you even get to define who the Bosses and other relevant NPCs are with your characters relationships to them, and where they live in the Ark. The Parameters are you're playing mutants in the Zone who live in an Ark with N amount of 'build points'. This is quite clearly not a Railroad to a start condition.

[NichG]

Like I said, in a sarcastic point; 'You start in [arbitrary location].' is already a loaded sentence. The very first sentence of almost any game already in the DM's favour;
I've never let my players choose where they start the game. Literally never. I strongly doubt any DM has, and if they have, I don't believe they do it regularly.

I had a DM basically let each player define the entire universe their character came from and how that universe ended... Next thing I'm running, the players as a group are building the entire setting using Microscope...

The world is wide.

[PhoenixPhyre]

I've never let my players choose where they start the game. Literally never. I strongly doubt any DM has, and if they have, I don't believe they do it regularly.

My general policy for new players to my world is

"Here is the world at <year>. Here are a selection of starter quest blurbs, designed to get you comfortable with the world and doing something. Which one do we want to pursue?" This may even be before characters are created. Or afterward. And all I have are blurbs--the actual details of the quests and even more than a one-to-two sentence description of the immediate area are built after I know who the characters are and some of what they want. And they don't even really have to pursue that first quest very far--just really the first session.

For people who have played before in the settings (like one of my current campaigns), it's

"ok here's the whole world map. Here are the (one sentence) descriptions of what I know about each general area. Where do we want to play next? What kinds of themes are we interested in? High-flying? More grounded? What kinds of creatures?" My current party decided to pick an area I knew approximately zero about, and then we built characters. But we did this well in advance of finishing the first campaign, so I did world-building (high-level planning, more "what's there" than anything) well in advance.

Both of those are basically "let[ting] my players choose where they start the game". And I've done that since...well...the 3rd or 4th campaign I've ever done in this world. Now at campaign 16. And I play basically stock D&D 5e (mechanically, the world is very non-standard).

[Cheesegear]

It's certainly an interesting view that the start conditions could indeed be considered a Railroad that leads from "agreeing to join and play in a game with start condition A" through "creating a character that would arrive at start condition A somehow" to "start condition A". Most people probably discount it because it's part and parcel of signing up for that game.

That's kind of Easy e's point.

Part of signing up to a GM-moderated game, is accepting that the GM holds all the power, and that every arbitrary decision that a GM ever makes, is 'negating player choices'. So that in effect, if you frame [the bad thing] as 'negating player choices', you have to accept that all GMs - even the ones on this forum that claim that they don't - make arbitrary decisions based on nothing, without player input or affectation, all the time.

DM: You come across a band of goblins.
Players: No we don't. We came across some friendly Commoners sitting in a field. That's what we choose. So **** your Goblins.

Now again, the situation describes some hostile players. But the point is very real. DMs make decisions outside the players' control all the time. Anything the DM plans, is outside the player's control. The players can react to a scenario in any way they choose...But they are forced to react to that scenario. When do the players get to create the scenario? They don't.

This is why framing it as a player agency issue is so...Egregious. 'Cause if you really, really think about it, as Easy e has, players only have as much agency as the DM gives them, and quite often they have no agency at all:

'Well, the DM has to plan, right? Otherwise how does anything get done? That way lies Entropy Dragons.'
Correct. The DM does have to plan. The DM does have to make several decisions outside the players' agency. All the time. No, like...All the time.
'But that's just like, DMing, man.'

But you framed the issue around 'negating player choice/agency.' Which encompasses so many things that yes, a DM does - even the 'good' ones.

An example of such a game is Mutant Year Zero. You get to design your Ark, you pick where on the map it goes, and you even get to define who the Bosses and other relevant NPCs are with your characters relationships to them, and where they live in the Ark. The Parameters are you're playing mutants in the Zone who live in an Ark with N amount of 'build points'. This is quite clearly not a Railroad to a start condition.

I have never heard of such a thing, and yes that is the exact counter to my point. It may even be the exception that proves the rule?

[Tanarii]

No, that's not a valid point. The DM can generate the content procedurally depending on player choices on where to go and what to do. The DM can place the content and let the players desire where to go within it, give them rumors or let them research info on where to go, and let them scout it out.

Neither of those are actively negating player agency. Players not getting to design the content is not negating player agency.

Negating player agency is negating their active decisions so they must encounter specific content.

There's a valid point to be made that forced starting conditions may negate their ability to make decisions about encountering starting content. But it's very easy for a DM not to do that once the game starts.

[Cheesegear]

The DM can generate the content...

The DM can place the content...

The DM creates the content. End.

DM: As you're walking through the Forest, you come across a band of Goblins.
Player: Wrong. I came across a bear.
DM: ...But there are Goblins here.
Player: This sounds like a railroad. I want to fight a Bear, not Goblins. Please respect my agency and remove the Goblins from the forest. I didn't choose for them to be there. I chose a Bear.

That's not how anything works. You and I both know that.

[Satinavian]

T
DM: You come across a band of goblins.
Players: No we don't. We came across some friendly Commoners sitting in a field.

Yes, there are games out there that work exactly this way.

It is not my preferred style, but it sure does exist.

[Telok]

The DM creates the content. End.

DM: As you're walking through the Forest, you come across a band of Goblins.
Player: Wrong. I came across a bear.
DM: ...But there are Goblins here.
Player: This sounds like a railroad. I want to fight a Bear, not Goblins. Please respect my agency and remove the Goblins from the forest. I didn't choose for them to be there. I chose a Bear.

That's not how anything works. You and I both know that.

Heh, Amber Diceless. The DM drops goblins, the PCs walk past a tree, the PCs get the bear they want. Tho honestly, "goblins" in that game could be anything from D&D chaff mobs to seriously powerful fey. But if they aren't Amberites or from the Courts of Chaos then they can't do anything about the PCs stepping to the next shadow with a bear.

[Composer99]

Easy e, I have to say the way you frame the issue is not entirely correct, and is in any event really rather reductionist to the point of not being all that useful for discussion.

Who gets to "have a say" in what happens in an RPG is some combination of GM, system, or players, with the precise balance depending on the system. Certainly many GM-mediated games, such as D&D, assume an overabundance of "GM says", but even with a GM that's not the case in other games. But even in D&D there is a lot of "system says what happens" going on. For instance, most players would rightly balk if a DM continually overrides, say, the procedure for making attacks (which is a "system says" part of the game) when no agreement for such behaviour was part of the social contract, you might say, for that table and campaign.

(I will here note that there is no actual rule in the 5e DMG - that I have found upon recent re-reading, that is - that says the DM can simply disregard, ignore, or break any of the non-optional/variant "system says" rules. DMs are expected to follow the rules for attack rolls, for resolving spell effects, and so on. The closest the DMG gets to this is suggesting that rolling behind a screen allows DMs to fudge if they feel the need to.)

What is more, there is to my mind a pretty obvious difference between a DM who uses some combination of deceit, compulsion, or manipulation to deny the players whatever say they may normally have in establishing what happens during gameplay and a DM who doesn't - even in a heavily GM/DM-mediated game, such as D&D, where the rules more or less restrict the "players say" element to "the players say what their characters think, feel, speak, do, or attempt to do".

Likewise, by my reckoning there's a pretty obvious difference between a DM/GM who (pick one):
(1) Doesn't know what happens next (at a given scale of gameplay or metagame) and relies on some combination of player decision, "system says" (mechanics such as, say, reaction rolls), and GM decision (whether based on fictional positioning, metanarrative, or metagame factors, etc. (*)) to help figure it out.
(2) Knows what's going to happen next, whether to some level of detail or just in broad strokes, and shapes or adjusts the happening based on the players' decisions (as well as "system says" mechanics and GM-facing considerations)
(3) Knows what's going to happen next and uses some combination of deceit, manipulation, or compulsion to prevent the players' decisions from having any impact on the outcome. (**)

If the discussion begins and ends with your assertion, "The GM is the only ones with power to shape the game", these distinctions are all smeared together into "GM Fiat Über Alles", which is unhelpful to say the least.

(*) Metagame factors could include considerations such as "Our 'fighty' and 'drama' players are starting to get bored as we haven't had a combat in over an hour. Better throw something in there to spice things up!" or "Geez, it's 10:30, better wrap things up!"

(**) Out of these three, (1) strikes me as mostly lining up with kyoryu's definition of emergence/emergent gameplay, (2) strikes me as mostly lining up with kyoryu's definition of authored gameplay, and (3) is what I personally would call railroading (and the term also applies to the DM/GM who uses the same techniques to deny the players the "players say" part of getting to say what happens in the game).

[KorvinStarmast]

Next thing I'm running, the players as a group are building the entire setting using Microscope.

I have Microscope, but I've not got an in person group to try that ... and based on my preferences, I'd really like to have the players invested in the world by having helped to create it.

I have never heard of such a thing, and yes that is the exact counter to my point. It may even be the exception that proves the rule?

The original play style of the wilderness adventure lent itself to procedurally generated content that both the DM and players would not have prepared ahead of time. (Dungeons not as much, but if you really did roll random encounters based on dungeon level for wandering monsters, you once again have content and encounters that are an unknown to both the players and the DM when they occur.

We had loads of fun doing both.

I tend now to curate lists of "wandering monsters/NPCs who are logically in the area because of {X} reason" because we have found that this lends more coherence to the setting we are playing in. (And not all of those encounters will necessarily be 'level appropriate' even though I am running 5e).

[Tanarii]

Player: This sounds like a railroad. I want to fight a Bear, not Goblins. Please respect my agency and remove the Goblins from the forest. I didn't choose for them to be there. I chose a Bear.

The problem here is you (and this made up player) don't understand what player agency is, and as a result the difference between limited player agency (not railroading) and negated player agency (railroading).

Just to start with, player agency is about making decisions for what their characters attempt to do within the fantasy environment (aka roleplaying), not defining the fantasy environment itself. And that's before we get into limited options vs active negation of choices made.

[clash]

Having followed this discussion I like the terms authored and emergent but I would like to propose a third term to be used in conjunction with those two.

Divergent: An authored game where player choices have the ability to change any part of the pre-written story.

When I dm I know at a super macro and macro level what the story is going to be. When the campaign starts. I know at the start of a session what the story is going to be at a micro level as well, but having the story pre written doesn't prevent player choices from changing it at any level. The players often diverge from the authored narrative at the micro level and given the right choices could easily derail the story at a macro and super macro level. At which point the rest of the session will be improv and the story is re-authored at a macro and super macro level in response.

90% of the time they follow along with the authored content but that doesn't prevent them from diverging from it at any level.

So I purpose authored, emergent and divergent.

Spoiler: example

Show

As an example to divergent in the campaign I am currently running, if the players follow the authored content they will reach a point where the "villains" of the story try to present their case to recruit the players. The villains have been their enemies for a while now and the most likely case is that the players will reflect their proposal and fight to restore the imperial family to the throne after the way concludes. However, should they choose to accept the proposal I will reauthor the major points of the campaign based on an alternate direction where they fight to overthrow the imperial family and the end state of the setting is entirely different as a result. They could also choose to join niether side and go to an adjacent country at which point I would reauthor the entire story going forward to present them with a list of challenges and an interesting narrative in the new country. It is most definitely authored but the players can diverge from the authored content at any point

[Stonehead]

Like I said, in a sarcastic point; 'You start in [arbitrary location].' is already a loaded sentence. The very first sentence of almost any game already in the DM's favour;
I've never let my players choose where they start the game. Literally never. I strongly doubt any DM has, and if they have, I don't believe they do it regularly.

The DM creates the content. End.

DM: As you're walking through the Forest, you come across a band of Goblins.
Player: Wrong. I came across a bear.
DM: ...But there are Goblins here.
Player: This sounds like a railroad. I want to fight a Bear, not Goblins. Please respect my agency and remove the Goblins from the forest. I didn't choose for them to be there. I chose a Bear.

Has anyone claimed the players have agency over anything other than their character? Was this argument really about encounters or something? The players have control over their characters and the GM has control over everything else. When people say "player agency" they mean agency over their character. When people say the players have agency over the world, they mean their character's actions affect the world in a reasonably causal way. I've never seen anyone claim that the players have control over what monsters live in the cave, so that seems like a really weird point to make.

I can make bad examples too,
GM: Galahad steals gold from the beggar.
Player: But that's my character, he wouldn't do that.
GM: Yes he would, I'm the GM. Player agency doesn't exist. Everything is GM fiat.
Just as ridiculous from the other side.

Now, you might say that's not a fair example, because that wasn't your point, but the other side's point wasn't "Players should have agency over what random encounters they meet".

To emphasize; player agency means control over their characters, whose actions impact the world.

[PhoenixPhyre]

Having followed this discussion I like the terms authored and emergent but I would like to propose a third term to be used in conjunction with those two.

Divergent: An authored game where player choices have the ability to change any part of the pre-written story.

When I dm I know at a super macro and macro level what the story is going to be. When the campaign starts. I know at the start of a session what the story is going to be at a micro level as well, but having the story pre written doesn't prevent player choices from changing it at any level. The players often diverge from the authored narrative at the micro level and given the right choices could easily derail the story at a macro and super macro level. At which point the rest of the session will be improv and the story is re-authored at a macro and super macro level in response.

90% of the time they follow along with the authored content but that doesn't prevent them from diverging from it at any level.

So I purpose authored, emergent and divergent.

Spoiler: example

Show

As an example to divergent in the campaign I am currently running, if the players follow the authored content they will reach a point where the "villains" of the story try to present their case to recruit the players. The villains have been their enemies for a while now and the most likely case is that the players will reflect their proposal and fight to restore the imperial family to the throne after the way concludes. However, should they choose to accept the proposal I will reauthor the major points of the campaign based on an alternate direction where they fight to overthrow the imperial family and the end state of the setting is entirely different as a result. They could also choose to join niether side and go to an adjacent country at which point I would reauthor the entire story going forward to present them with a list of challenges and an interesting narrative in the new country. It is most definitely authored but the players can diverge from the authored content at any point

That's basically what I was getting at with my rambling. The story is authored (at some selection of levels), but that authorship can be thrown away and rebuilt at any point due to facts on the ground (ie the actions of the party and their consequences). So instead of being authored (ie set in stone), it's projected. There's a strong "if they follow the line I think they will", with allowance for limited variation within that while still staying "on the predicted path." Whether they slaughter every person in <group A> or merely cause them all to flee may matter in the short term (one to two session) but will only matter for the large scale under certain specific circumstances. But in those specific circumstances (which may be as frequent as you and the table desire), that difference in action can radically change the upcoming sessions out to asymptotic time.

[KorvinStarmast]

To emphasize; player agency means control over their characters, whose actions impact the world.

Have people been reading my sig? (The one from greenstone)

[Easy e]

If the DM negates player agency to make sure they get to those things, it's railroading.
If the DM hides the negation of agency by presenting an illusionary choice, it's illusionism and railroading.
If the DM hides the negation of agency by presenting an illusionary choice between three different woods, but in reality specifically the first wood travelled to will have an Ogre encounter and the second will have the macguffin, it's a quantum ogre and illusionism and railroading.

That's what the terms railroading, illusionism and quantum ogre actually means.

So, the version of the game Icefractal laid out is objectively bad, because it is railroading, illusionism, and quantum ogre? Or am I misinterpreting your post?

This one:

#3: The GM has a definite plot in mind, with a beginning, certain milestones, and a pre-defined ending. However, the game is run in a very improv style, and between those fixed points almost anything can happen. The GM uses subtle railroading / quantum ogres to make all the infinite possible paths converge on those fixed points, but does almost no actual "authoring" of material - the entire campaign notes fit on a single page.

Maybe you should play a game without GM for a while to challenge your "core assumptions"

I do all the time, granted those are mostly board game and wargames. Hence why I limited my discussion to TT RPGs with GMs.

However, much of the same can be said of wargames with a GM as TT RPGs with a GM, but that is way off topic.

[JNAProductions]

So, the version of the game Icefractal laid out is objectively bad, because it is railroading, illusionism, and quantum ogre? Or am I misinterpreting your post?

This one?

I'm not sure I'm getting the classification here. What would you call these?

#1: The game is a sandbox - a setting that players can travel to any part of, ally with or oppose almost any group, take on mercenary jobs or not, help people or not, seek out treasures or not. However, there's an agreement that at the end of a session, the players decide what they're going to do during the next session, and they stick to that. And that's because the GM will then prep the appropriate material between sessions.
So the players have a lot of freedom on a macro-level, but during a session most things that happen will be pre-written by the GM rather than improv.

#2: The GM is using a huge existing piece of material, like Ptolus or a mega-dungeon. Within that, the players can do whatever they want to, but they can't decide to leave the city entirely (well they can, but it means the end of the campaign).

#3: The GM has a definite plot in mind, with a beginning, certain milestones, and a pre-defined ending. However, the game is run in a very improv style, and between those fixed points almost anything can happen. The GM uses subtle railroading / quantum ogres to make all the infinite possible paths converge on those fixed points, but does almost no actual "authoring" of material - the entire campaign notes fit on a single page.

One is completely and totally fine-there may be a theoretical issue where the players finish content faster than expected and the DM doesn't have more material, but in that case, the session can just end a bit early, or turn into campsite RP.

Two is also totally fine, assuming the DM made clear with the pitch that the entire campaign takes place in the given setting.

Three is the one I would consider problematic. It won't by necessity be an unfun game, but unless the DM tells the players outright that they have a plot in mind and that the plot will be followed, it's an issue.

A good question to ask yourself on "Is this style of DMing okay?" would be "How would the players react if they found out what you're doing behind the scenes?"
For one, they already know what you're doing-the DM asks you at the end of a session what your plan is for the next one, and they prep it for next session. No lies, no problems.
For two, they already know that the DM is following the setting notes, whether it's premade or their own creation. They know that monsters, treasures, factions, all that is pre-set, and if they go somewhere, they'll encounter the same thing whether they go now or in four sessions. No lies, no problems.
For three, if the DM tells the players "I have a plot in mind, and we're gonna follow it. Here's the general gist of it... Sound fun?" then they know that the DM will lead them along the path that the plot follows. No lies, no problems.
For three, if the DM did NOT tell the players that it would be a linear game, but they make sure it stays linear by DM fiat, that's an issue. And the players would likely be upset if they found that out.

[Easy e]

Wait a sec, so an exchange like this:

Dm:"Ok, thats the current stuff mostly wrapped up. What do you lot want to do next?"
Pcs:"We want to go find a person who can make a thing and get them to make some stuff for us."
Dm:"Hmm. Ok, there are people around like that. We'll wrap up here tonight and I'll randomly generate some people, leads, etc., for next time."

That's the DM being the only one shaping the game?

Yes, because the GM is the one that decides that they wanted the player input in the first place, and it then acting to meet the players wants. The GM does not have to do that at all, and the game runs fine.

The exchange could have easily been:

Gm: "Ok that is the current stuff wrapped up. Next time, you find yourself in X."
PCs: "Oh, how did we get there?"
Gm: "You travelled via boat, and then walked 4 days and it was uneventful and your feet are sore. We will take it from X next time as you explore X."

The GM in the case you describe decided NOT to do it that way. It was a GM choice on how they wanted to shape the game, with PC input.

[BRC]

Yes, because the GM is the one that decides that they wanted the player input in the first place, and it then acting to meet the players wants. The GM does not have to do that at all, and the game runs fine.

The exchange could have easily been:

Gm: "Ok that is the current stuff wrapped up. Next time, you find yourself in X."
PCs: "Oh, how did we get there?"
Gm: "You travelled via boat, and then walked 4 days and it was uneventful and your feet are sore. We will take it from X next time as you explore X."

The GM in the case you describe decided NOT to do it that way. It was a GM choice on how they wanted to shape the game, with PC input.

Agency may be granted by the GM, but that doesn't make it Not Agency. If Player Input is genuinely driving the game, then the game is driven by player input, even if the GM isn't required to allow that to happen.

[JNAProductions]

Yes, because the GM is the one that decides that they wanted the player input in the first place, and it then acting to meet the players wants. The GM does not have to do that at all, and the game runs fine.

The exchange could have easily been:

Gm: "Ok that is the current stuff wrapped up. Next time, you find yourself in X."
PCs: "Oh, how did we get there?"
Gm: "You travelled via boat, and then walked 4 days and it was uneventful and your feet are sore. We will take it from X next time as you explore X."

The GM in the case you describe decided NOT to do it that way. It was a GM choice on how they wanted to shape the game, with PC input.

Sure, but if I was playing a game where the DM decided what my PC did on anything even approaching a regular basis, they'd find themselves down a player in short order.

[Tanarii]

So, the version of the game Icefractal laid out is objectively bad, because it is railroading, illusionism, and quantum ogre? Or am I misinterpreting your post?

No. I was just defining what was necessary for them to actually be railroading, illusionism, and quantum ogres.
They are terms that get mis-used a lot, usually far too broadly.

As to "objectively bad", no the "badness" is my opinion. They are negatively associated terms, that I personally think are bad as a Dm but also as a player, but that doesn't mean that every player inherently thinks that the thing happening in the definitions (as opposed to the terms themselves) are bad. And that's what really matters, not that the DM thinks they're okay, but that the players don't object. In the case of illusionism, object when they inevitably find out.

Also, just in case you're also unclear on what player agency is, because I didn't actually define it in my post you just quoted, this isn't a bad wording for it:

To emphasize; player agency means control over their characters, whose actions impact the world.

Player agency is over making decisions for their characters. Not the content.

The GM in the case you describe decided NOT to do it that way. It was a GM choice on how they wanted to shape the game, with PC input.

Agency may be granted by the GM, but that doesn't make it Not Agency. If Player Input is genuinely driving the game, then the game is driven by player input, even if the GM isn't required to allow that to happen.

I think saying the default state is DM-deny-agency because any DM in many game systems might try to exercise the ability to deny agency is not correct.

IMX any DM that tries to deny agency regularly often/usually results in them not being a DM any longer. Therefore they have failed to deny agency, and also to DM. So it can't be the default. Either that or they become convention DMs.

[Vahnavoi]

Having followed this discussion I like the terms authored and emergent but I would like to propose a third term to be used in conjunction with those two.

Divergent: An authored game where player choices have the ability to change any part of the pre-written story.

Divergence (and convergence) is undoubtedly a real quality - if you know what a game (or decision) tree is and know what branching is, even a partial map of a game's movespace will make the matter self-evident- but using it alongside authored and emergent doesn't add all that much information. The simple reason is that divergent (and convergent) branching can be either authored or emergent and a player of an imperfect information game has no reliable way to tell which is the case while playing a game. Only the game designer can reliably make that distinction.

[Easy e]

So we all ultimately agree, that PC agency is controlled by GM Fiat,

Do we also agree that all GMs decide how much they exert their GM Fiat powers during a game?

[KorvinStarmast]

So we all ultimately agree, that PC agency is controlled by GM Fiat,

I will cite 'the rule of so' and ask that you avoid attempting to speak for others, particularly when making such a broad, general statement. And since you framed it that way I will state that 'no, I do not agree with how you framed that.'

Do we also agree that all GMs decide how much they exert their GM Fiat powers during a game?

I think I do, mostly.

[Satinavian]

I do all the time, granted those are mostly board game and wargames. Hence why I limited my discussion to TT RPGs with GMs.

However, much of the same can be said of wargames with a GM as TT RPGs with a GM, but that is way off topic.

Should have been obvious this was about tabletop RPGs without GM.

To emphasize; player agency means control over their characters, whose actions impact the world.

That is the traditional and also most common way to to it. But it is not universal. There are e.g, many games out there where players get direct power over the gameworld, often coupled to metaressources. In those games that would also fall under player agency.

So we all ultimately agree, that PC agency is controlled by GM Fiat,

Do we also agree that all GMs decide how much they exert their GM Fiat powers during a game?

Nope.

The game only happens as long as everyone agrees who has what kind of power. So the GM controls player agency exactly as much as the players control GM agency Because if anyone tries to force their opinions about it through against the table, it only blows up in its face and the game is on halt.

[Tanarii]

So we all ultimately agree, that PC agency is controlled by GM Fiat,

Do we also agree that all GMs decide how much they exert their GM Fiat powers during a game?

In many TTRPG games, a GM can try to fiat denial of player agency without breaking the defined rules of the game for the GM. The fact that not all games include defined rules for the GM, let alone suggestions, which can make avoiding breaking them in the process of trying a lot easier.

The GM can decide if they want to try that. They risk losing players an no longer being a DM in the process of trying. The players ultimately have a say on if they're going to accept it too. Their say is to walk from the table. That's the Player Fiat on the same level as this particular kind of DM Fiat.

[Easy e]

Agency may be granted by the GM, but that doesn't make it Not Agency. If Player Input is genuinely driving the game, then the game is driven by player input, even if the GM isn't required to allow that to happen.

Player Input is the Illusionism!

After your "input" who creates the content and drives the story beats? I have played very Improv heavy games, where I work with the players to craft the plot, impacts, etc. Make no mistake though, as the GM using Rule 0, I get to say what we acted on and what got reeled back in; not the players.

To paraphrase The Matrix, "Would You Still Have Broken It If I Hadn't Said Anything?"