What is Player Agency?

[OldTrees1]

I will go against most of the thread and say that player agency has nothing to do with freedom or lack thereof during a play session/quest/mission/however you call it. It's good when the game master is open to players' actions and allows them to evade combat through smart diplomacy, stealth or whatever for example - but that is not agency. That is simply having an adventure with more options and solutions available.

Player agency is about player and character goals - which are often (but not always!) the same thing. It's something more long term than having an option between going a left path to fight kobolds and a right path to fight goblins. These choices are ultimately meaningless.

It's about what the player wants from the game, and whether the GM allows the players to pursue these goals or not. The goals can be planned in advance but may as well appear ad hoc: I want to find my mentor's killer. I want to improve living conditions of the goblin hamlet we just visited. I want to explore difficult relationships between my character and their family. I want my character to suffer dramatically. If the GM knows of these goals and allows the players to chase after them, dropping in plot hooks and points or even creating whole adventures dedicated to pursue of individual player wants, then there is agency. If the goals are actively ignored or even stomped down - the family the player wanted to interact with killed for cheap drama, for example, then there is no player agency.

By definition this means player agency can't be there without a means of communication between players and the GM. The players need to know what they want from the game and the GM must be informed of it. If the players are just in for the ride, to have a good time dungeoneering or exploring the unknown with no personal stakes except "I want my character to survive this and get a lot of loot/exp"? There is no agency of course but that's also fine. If that's what everyone wants and they have fun, the lack of agency doesn't matter. Agency is a tool, a mean to an end, not a goal itself.

Also needless to say, not every player goal needs to be pursued and realized. Some may clash with the game's theme, be way above the characters' means to pursue or make other players uncomfortable and be just plain wrong.

Let's test your theory that this is different from rather than an emergent property of access to more frequent and greater scale meaningful choices.

Let's start by examining the minimum case:
A: Your character wants to find their mentor's killer. The DM coincidentally uses an extremely strict railroad to make the party face the BBEG who just happens to be your character's mentor's killer. The PC was faced with no meaningful choices, the player did not choose to have their PC face their PC's mentor's killer. But that desired event did happen. Agency or not? (If Agency, why so?)

B: Your character wants to find their mentor's killer. The DM knows of this and decides to use an extremely strict railroad to make the party face the BBEG who just happens to be your character's mentor's killer. The PC was faced with no meaningful choices, the player did influence the DM's railroad path to have their PC face their PC's mentor's killer. But that desired event did happen. Agency or not? (If Agency, why so?)

C: Your character wants to find their mentor's killer. The DM runs a campaign full of meaningful choices. Those choices do allow your character to make goals and pursue them BUT finding your mentor's killer either was not within the scope of those choices or your character did not succeed in finding a path that would work. Agency or not? (If not, why not?)

D: Your character wants to find their mentor's killer. The DM runs a campaign full of meaningful choices. As a result of those choices your PC ends up facing their mentor's killer. Agency or not? (If not, why not?)

F: Your character wants to find their mentor's killer. The DM runs a campaign full of meaningful choices BUT the characters also have enough scope in which to create their own opportunities. Following those self created opportunities leads your PC to face down their mentor's killer. Agency or not? (If not, why not?)

There are a few points I want to address:


The other thing I'm finding weird is the whole "agency is a tool, not a goal" line of thinking. It seems as incongruous as saying that logs are a tool to make a log cabin, and not the goal. Sure, I guess? But isn't the presence of the logs an instrinsic part of the goal in the first place? If the whole point of the house is that it is made from logs.... doesn't that make the logs an intrinsic and inseparable element of the goal? And isn't it possible for a game to have more than one goal?
I see no reason why I can't have Agency be a goal. For instance:
"I want this campaign to be shaped mostly by Player Agency," sounds like a perfectly reasonable goal to set as part of the campaign. So the Tool-Not-Goal angle seems... incomplete for the same reason my goal can be both "lose weight" AND "Do 50 situps per day," and arguing that situps are a tool and not a goal would be incomplete.

I believe that line of thinking arises once there is plenty.

I can build a log cabin, or a mansion, or a death star out of logs. If logs are the goal then I should always collect enough to make a death star. However if the logs are a tool for building the building I am after, then I can stop collecting at some point.

If I am crafting a standard lovecraftian horror game. I would not want to have the PCs start out with meaningful choices in which they decide if the inner planes should crash into sigil OR the upper and lower planes should jettison out into the Far Realms. The PCs do not need that scale of agency in that campaign AND that scale of agency is not quite appropriate for that campaign. But just as too much agency would mess up the campaign, I still need to make sure the campaign has enough (err on the side of too much) agency for the players to be able to enjoy the lovecraftian horror "try to survive a year longer this time" game they signed up for.

[Max_Killjoy]

Oh. This reminded me of one of the most frustrating instances of agency being illusory and verisimilitude-breaking I've experienced in recent memory.

In a PF module at GenCon a couple of years ago, we were chasing after our NPC ally who had been drawn off or captured (I forget which) by some BBEG of the module, trying to rescue him. We were given a choice after we'd been through a couple of major fights as to whether to rest up, or keep pressing on to try to catch up. We had no idea how far ahead he'd been taken, only that he was in trouble.

We chose to press on without rest. We arrived just in time to see him facing off with the BBEG, and we were just a few spells shy of being fast enough to kill off said BBEG before he killed the NPC ally.

I learned later, from others who'd done the same module, that that scene always starts off the same way: you always arrive with the NPC ally in exactly the same state of fighting the BBEG. Us not resting and pressing onwards did nothing to change how quickly we caught up, nor even the precise amount of "hurt" our ally was. But it did mean we lacked spell resources and HP that we otherwise would have had.

This was an illusory choice in the sense that either way, the situation in which we'd find ourselves was unchanged, so there's no benefit to rushing ahead. However, we still paid for having rushed ahead.

It's...not entirely lacking agency, as our actions had consequences, but because the consequences were only applied to our resources and not to our situation, we got worse than gypped.

This could have been a trade-off, a pair of offsetting calculated risks, the ultimate meaningful choice... and instead, meh.

[Segev]

This could have been a trade-off, a pair of offsetting calculated risks, the ultimate meaningful choice... and instead, meh.

Exactly. But a bit worse, because "choose wrong, and you only get punished."

I blame the DM running it, honestly. Not for having run the module as written, but for not telling us, "Look, the module assumes you rest here; you won't catch up any sooner if you don't." It's a slot at GenCon; I don't expect DMs to creatively interpret the module to account for all our choices. But at least don't pretend there's agency when there isn't and then penalize us for having exercised it without knowing there wasn't any.

[ImNotTrevor]

I believe that line of thinking arises once there is plenty.

I can build a log cabin, or a mansion, or a death star out of logs. If logs are the goal then I should always collect enough to make a death star. However if the logs are a tool for building the building I am after, then I can stop collecting at some point.

Yes, but having the logs is an intrinsic part of said construction. Diminishing it to just a tool makes it come across as optional. Call me ignorant, but I'm fairly sure the point of a log cabin is that it is made primarily of logs. Taking that away more or less makes the endeavour into something else entirely.

Hence why i say the stance is not Utterly Wrong, but is Incomplete. If the goal is to have a game with many potential outcomes then you damn well better have lots of player agency for the same reason you better have logs if you're building a log cabin. The apparent point of the "agency is a tool" standpoint is to backend justify the removal of agency.

If I am crafting a standard lovecraftian horror game. I would not want to have the PCs start out with meaningful choices in which they decide if the inner planes should crash into sigil OR the upper and lower planes should jettison out into the Far Realms.

Anything taken to a ridiculous extreme will be ridiculous and extreme, yes.
And yes, you don't need 86 quintillion logs to make a cabin. But you had better have them!

The PCs do not need that scale of agency in that campaign AND that scale of agency is not quite appropriate for that campaign. But just as too much agency would mess up the campaign, I still need to make sure the campaign has enough (err on the side of too much) agency for the players to be able to enjoy the lovecraftian horror "try to survive a year longer this time" game they signed up for.

As I said elsewhere, scale is a different question from "what is this thing."
Is Player Agency a tool? No. It is not a hammer.
It's construction material. Campaigns are made OF it, not made WITH it.
If that makes more sense.

[OldTrees1]

Hence why i say the stance is not Utterly Wrong, but is Incomplete. If the goal is to have a game with many potential outcomes then you damn well better have lots of player agency for the same reason you better have logs if you're building a log cabin. The apparent point of the "agency is a tool" standpoint is to backend justify the removal of agency.

As I said elsewhere, scale is a different question from "what is this thing."
Is Player Agency a tool? No. It is not a hammer.
It's construction material. Campaigns are made OF it, not made WITH it.
If that makes more sense.

Your construction material metaphor is quite apt. You need to have some in order for there to be anything and eventually you can have more than enough for your purposes. The metaphor even covers the needed quantity varying between various campaigns (horror, stereotypical, sandbox, etc).

[NichG]

The point of a log cabin is to provide shelter from the elements using locally available materials, not to be made of logs. If your locally available material is slabs of stone, saying 'log cabin or nothing, since there are no logs I'd like to freeze to death please' is silly.

The reason people are saying 'player agency is a tool' is because somewhere along the way, people stopped thinking about 'why is agency fun?' and instead got so pinned on the idea of agency itself that they substituted it as a replacement for fun. Saying 'agency is a tool' gets to the point that its a means, not an end. The distinction is important because the point of view of agency as an end says 'there (definitionally) cannot be a case where sacrificing agency for something else is justifiable', whereas agency as a means engages with the possibility that there are in fact cases where more or less agency is called for, or where different kinds of agency serve different ends, or so on.

Talking about it as a tool focuses the discussion on the 'why' rather than the 'what'. Not 'what is a theoretical definition of agency that no one manages to successfully attack' but rather 'Why do we care? What is good or enjoyable about experiencing agency? What happens when players have agency, and when they don't?'. Those are far more pragmatically useful discussions.

[ImNotTrevor]

The point of a log cabin is to provide shelter from the elements using locally available materials, not to be made of logs. If your locally available material is slabs of stone, saying 'log cabin or nothing, since there are no logs I'd like to freeze to death please' is silly.

People still build log cabins currently, in places you'd not expect to find them. But this here is missing the point of the metaphor.

The reason people are saying 'player agency is a tool' is because somewhere along the way, people stopped thinking about 'why is agency fun?' and instead got so pinned on the idea of agency itself that they substituted it as a replacement for fun. Saying 'agency is a tool' gets to the point that its a means, not an end. The distinction is important because the point of view of agency as an end says 'there (definitionally) cannot be a case where sacrificing agency for something else is justifiable', whereas agency as a means engages with the possibility that there are in fact cases where more or less agency is called for, or where different kinds of agency serve different ends, or so on.

Except I neither adopt the Tool stance nor do I have a problem with the concept of a sliding scale of Agency. As far as I can tell this theoretical position you're arguing against is not present.

Talking about it as a tool focuses the discussion on the 'why' rather than the 'what'. Not 'what is a theoretical definition of agency that no one manages to successfully attack' but rather 'Why do we care? What is good or enjoyable about experiencing agency? What happens when players have agency, and when they don't?'. Those are far more pragmatically useful discussions.

Since the topic is "what is player agency" as opposed to "why is player agency" I think we're not exactly out of bounds to dig into things like what people perceive player agency to be. Your perception is different, clearly. What I'm wondering is how you got to the conclusion that only from your particular perspective can those questions be asked? I see no reason to hold the Tool stance as somehow more enlightened or accurate. And as I've motioned, I view it not like a hammer, but like wood. A hammer is a means to an end.
Wood is an intrinsic part of what the end will be. The hammer can go somewhere else after. The wood cannot, for it becomes the completed work.

Player Agency becomes a part of what the campaign ends up being. It's makeup, its function, will inevitably affect what you end up with. It's a building block.
And of course, you can have more or less wood in a given building. Even none. Same with Player Agency.

[Max_Killjoy]

The point of a log cabin is to provide shelter from the elements using locally available materials, not to be made of logs. If your locally available material is slabs of stone, saying 'log cabin or nothing, since there are no logs I'd like to freeze to death please' is silly.

The reason people are saying 'player agency is a tool' is because somewhere along the way, people stopped thinking about 'why is agency fun?' and instead got so pinned on the idea of agency itself that they substituted it as a replacement for fun. Saying 'agency is a tool' gets to the point that its a means, not an end. The distinction is important because the point of view of agency as an end says 'there (definitionally) cannot be a case where sacrificing agency for something else is justifiable', whereas agency as a means engages with the possibility that there are in fact cases where more or less agency is called for, or where different kinds of agency serve different ends, or so on.

Talking about it as a tool focuses the discussion on the 'why' rather than the 'what'. Not 'what is a theoretical definition of agency that no one manages to successfully attack' but rather 'Why do we care? What is good or enjoyable about experiencing agency? What happens when players have agency, and when they don't?'. Those are far more pragmatically useful discussions.

Talking about it as a tool or a means to achieve something else, rather than a good thing in and of itself, also gives leave for some to push illusionism, for-your-own-good-ism, and so on.

After all, if something else is the actual point, and agency is just a way of make that something else happen, then anything else that can supposedly do as good a job or better of making that something else happen, allows them to favor the "better tool" and ignore player agency as the "inferior tool".

[NichG]

Except I neither adopt the Tool stance nor do I have a problem with the concept of a sliding scale of Agency. As far as I can tell this theoretical position you're arguing against is not present.

Concretely, the position you adopted was 'agency is not optional' (or specifically, you complained that the tool stance suggested that agency was optional, which you disagreed with).

There are genres of tabletop games where agency isn't the main point, and even ones where there is no agency when it comes to the ultimate outcome of the game. You see it primarily in horror-related genres. An example related to me recently is a game called '10 candles' where essentially each game is an apocalpytic scenario oneshot where by the end, all of the characters have died, and nothing can actually change that.

I'll grant its not my taste in games, but its a thing that someone might endeavor to do in a tabletop setting, and evidently there are enough people who enjoy that kind of thing that they find it worthwhile to pursue as well.

Similarly, tournament dungeoncrawl modules are very low agency but are that way for a purpose - it enables the kind of comparative, competitive play that they are targeting.

Agency-as-tool means that we don't have to accept a premise that 'agency is ultimately the point of the game' in order to discuss it, which allows us to discuss agency in a much wider array of games than if we're forced to assume that particular stance.

[ImNotTrevor]

Concretely, the position you adopted was 'agency is not optional' (or specifically, you complained that the tool stance suggested that agency was optional, which you disagreed with).

Yeah, I'm noticing that to continue this line you have to pick out one sentence and ignore everything I've said contrary to that one sentence to clarify the position, such as the several paragraphs I've written since which explicitly state that having different scales of agency is not a problem. (I think it should be greater than 0, since a 0 agency rpg is the oldest kind. They are called Novels.)

To clarify what I mean by optional is the idea that it is perfectly acceptable to just arbitrarily stop using the tool based on whims, or to specifically attempt to build things that don't use that tool at all.

There are genres of tabletop games where agency isn't the main point, and even ones where there is no agency when it comes to the ultimate outcome of the game. You see it primarily in horror-related genres. An example related to me recently is a game called '10 candles' where essentially each game is an apocalpytic scenario oneshot where by the end, all of the characters have died, and nothing can actually change that.

I'll grant its not my taste in games, but its a thing that someone might endeavor to do in a tabletop setting, and evidently there are enough people who enjoy that kind of thing that they find it worthwhile to pursue as well.

Similarly, tournament dungeoncrawl modules are very low agency but are that way for a purpose - it enables the kind of comparative, competitive play that they are targeting.

Agency-as-tool means that we don't have to accept a premise that 'agency is ultimately the point of the game' in order to discuss it, which allows us to discuss agency in a much wider array of games than if we're forced to assume that particular stance.

You're once again missing what I'm saying. I'm not saying agency is the end goal. I also don't accept the premise that the Tool position is the sole position from which meaningful discussion can occur. I'm gonna need you to at least make an attempt to understand what I'm getting at.

Secondly, both of your examples still maintain a degree of Player Agency. Even if the ending is inevitable, the players have a large amount of play room to move about in.

In tournament play, the player agency is focused away from the narrative layer, into the mechanical layer, where it moves freely.

So yeah, these examples simply involve different breeds of wood, as I veiw it.

[NichG]

Yeah, I'm noticing that to continue this line you have to pick out one sentence and ignore everything I've said contrary to that one sentence to clarify the position, such as the several paragraphs I've written since which explicitly state that having different scales of agency is not a problem. (I think it should be greater than 0, since a 0 agency rpg is the oldest kind. They are called Novels.)

To clarify what I mean by optional is the idea that it is perfectly acceptable to just arbitrarily stop using the tool based on whims, or to specifically attempt to build things that don't use that tool at all.

I would say that it is perfectly acceptable to specifically attempt to build games that don't make use of agency at all. I want to distinguish between statements like 'I would not like that game' and statements like 'that game is logically impossible' or 'it is inherently wrong to attempt to do that'.

You're once again missing what I'm saying. I'm not saying agency is the end goal. I also don't accept the premise that the Tool position is the sole position from which meaningful discussion can occur. I'm gonna need you to at least make an attempt to understand what I'm getting at.

Secondly, both of your examples still maintain a degree of Player Agency. Even if the ending is inevitable, the players have a large amount of play room to move about in.

In tournament play, the player agency is focused away from the narrative layer, into the mechanical layer, where it moves freely.

So yeah, these examples simply involve different breeds of wood, as I veiw it.

If we take 10 candles, the purpose of the game essentially comes down to a form of self-exploration - how do you find meaning in a situation where nothing you do can possibly matter and you're going to die no matter what? The agency, such as it might exist in transient forms, is basically counter-productive to creating the state of mind with which to explore that question. While agency might exist in such a game, seeing agency as the primary ingredient which determines the structure of the game misses the point.

If we look at the tournament modules, they're similar to games like chess and Go as mentioned up-thread in that past a certain point, increasing your expertise in the game actually means decreasing your agency. If what you want in a tournament module is to achieve the best possible score, eventually there is only a narrow optimal path of actions that leads there. While agency might exist in such a game, seeing agency as the primary ingredient which determines the structure of the game misses the point.

There are games where agency is the actual heart and soul of the game. God-games as are sometimes played on these forums are an example of that, where the game is built to encourage open-ended creativity in a shared context - the outcome of the game is what players choose to create. Similarly, there's a subclass of transformative sandbox games - sandbox games where the campaign ends up being about changing the world in some kind of player-driven way - that are primarily agency-based. On the other hand, a traditional sandbox game is not necessarily so - while again agency may exist in places, it's more about the feeling of freedom than it is about agency as more often than not decisions in a sandbox tend to be uninformed due to conservation of detail.

If we look at the stories people tell after the fact about old campaigns, sometimes they're about moments of agency, but as often as not they're about zany unexpected stuff that happened, unplanned interactions, or other assorted moments of awesome that actually involve the (sudden, surprising) absence of agency more than its presence. Even the way in which dice generate tension and story is in direct opposition to agency - they represent an agreement between all at the table to abandon a portion of their agency to the decisions of a random process.

[Quertus]

Challenge accepted. There are many nuances that we cannot really touch upon with Darth Ultron but can with each other. Let's be comprehensive in our definition here.

First: I think we all agree that a necessary condition for a person to be a player in a game is for them to be capable of interaction with the game. If nothing the person does has any impact at all, then we should consider the person and "game" in question as independent and unrelated. I, a person, can be described as playing a game of chess with a normal board and pieces. I cannot be described as playing a game of chess if the pieces are all glued in place (no interaction possible --> no playing possible).

Second: Even in games with only 1 kind of interaction (cite one-button games here), the player is faced with choices of which of their interactions to use or to abstain from interacting at that time.

Third: Now that we have the background covered. It is time to discuss Meaningful Choices. I will note that the meaning of Meaningful can be subjective, especially as it applies to a specific case. So this is a good place for us to dissect further (both in the minimum case and in the ideal case).

To my understanding a choice needs the following characteristics for it to qualify as a Meaningful Choice in the context of enabling the possibility of Player Agency:
A: The Players need to be aware of the choice they face and be sufficiently informed about the choice. This does not require they know everything or even understand a majority of the situation. Sufficiently informed is defined by the following criteria.
B: The choice needs to have multiple outcomes and those outcomes need to be the result of differences between the options. To be sufficiently informed, the players need to know enough of those relevant differences that their limited information could be used to map the differences between the options to the difference outcomes.*

*Obviously there can be relevant differences that the players are not exposed to and those differences will also impact the outcome. However that is merely reminding us that many meaningful choices are contained within the context of a larger choice (the knowable and unknowable details of the choice).

Fourth. Now we have the concept of meaningful choices. In practice, meaningful choices are how a player is able to intentionally impact the shape of the game. This can range from really small scale (choosing to save person A vs person B) to large scale (reshaping the socio-economic structure of the material plane through a long chain of actions and interactions). It can range from infrequent to frequent. This is Player Agency.

Different campaigns will choose different shapes and sizes of Player Agency. It makes little sense to allow plane warping player agency in a campaign with the lovecraftian horror motif of insignificant & depowered. But such a game would still have Player Agency in the shape of the investigations made and the attempted plans to survive/stay sane.

This is also why it is so important to include player agency in a RPG. Without player agency, the people are not actually interacting with the game.

Wow. You must have a lot of experience with legalise (how does one spell legal ease?). Based on my (revised) definition, below, do you feel we differ on the "meaningful" bit of your definition?

The point of a log cabin is to provide shelter from the elements using locally available materials, not to be made of logs. If your locally available material is slabs of stone, saying 'log cabin or nothing, since there are no logs I'd like to freeze to death please' is silly.

The reason people are saying 'player agency is a tool' is because somewhere along the way, people stopped thinking about 'why is agency fun?' and instead got so pinned on the idea of agency itself that they substituted it as a replacement for fun. Saying 'agency is a tool' gets to the point that its a means, not an end. The distinction is important because the point of view of agency as an end says 'there (definitionally) cannot be a case where sacrificing agency for something else is justifiable', whereas agency as a means engages with the possibility that there are in fact cases where more or less agency is called for, or where different kinds of agency serve different ends, or so on.

Talking about it as a tool focuses the discussion on the 'why' rather than the 'what'. Not 'what is a theoretical definition of agency that no one manages to successfully attack' but rather 'Why do we care? What is good or enjoyable about experiencing agency? What happens when players have agency, and when they don't?'. Those are far more pragmatically useful discussions.

Just like "balance", agency is not a synonym for fun. But, if we cannot stick to the topic of defining what agency is, we will continue talking past each other in talking about your more pragmatic "why".

That having been said, once we think we are in agreement as to "what", asking "why" and noting areas of dissonance is a great test of the extent to which we have failed at the "what" step. Thus, I highly encourage asking "why" as part of asking "what". But, just as one does not generally start with the roof when building a log cabin, starting with "why" is generally rather counterproductive to conversational efficiency.

On the other hand, a traditional sandbox game is not necessarily so - while again agency may exist in places, it's more about the feeling of freedom than it is about agency as more often than not decisions in a sandbox tend to be uninformed due to conservation of detail.

If we look at the stories people tell after the fact about old campaigns, sometimes they're about moments of agency, but as often as not they're about zany unexpected stuff that happened, unplanned interactions, or other assorted moments of awesome that actually involve the (sudden, surprising) absence of agency more than its presence. Even the way in which dice generate tension and story is in direct opposition to agency - they represent an agreement between all at the table to abandon a portion of their agency to the decisions of a random process.

I can only conclude that the phrase "Player Agency" doesn't mean the same thing to you that it does to me.

So, I suppose it behooves me to explain my definition. Well, my revised definition, seeing as how I mentioned up-thread that I hadn't originally included the first step of knowledge.

So, to me, Player Agency involves the players getting to make choices for their characters that have a measurable, logical impact on the game. To make such choices, the players need to have knowledge to make informed choices, the characters need to have capabilities to perform actions, and the world needs to resolve in a logical way for the outcome of those choices to be predictable.

Now, here's where it gets tricky. I define agency as everything within the bounds of the characters capabilities. For example, if I'm playing a "normal" D&D character, it does not curtail my agency to not let me jump to the moon, because jumping to the moon is not within my capabilities. A game where my character cannot jump to the moon can still have 100% Player Agency.

Further, knowledge need not be perfect. However, knowledge and world consistency needs to be such that, when an event does not have the expected outcome, it is therefore obvious that our knowledge was inaccurate or incomplete. Everything should respond in a predictable way, and, when it does not, that must be indicative that there is something going on here to investigate.

Lastly, obtaining missing information must be possible. That may involve moving to OOC discussion (especially if the players and GM aren't on the same page / wavelength), and can even involve a retcon if the PCs should have known information of which the players were unaware.

So, to pull an example from another thread, if, in the bizarre game world, everyone knows that, contrary to the real world, barn fires are actually salvageable situations, then the PCs need to have that information to make an informed decision about what to do about a barn fire in order to be considered to have agency under my definition.

Let me put forth another bizarre world. The players say the PCs look around; suddenly, the sun grows much brighter and hotter. The PCs look for shelter. As they race for a convenient log cabin, the sun grows even brighter, and lightning and hail begin descending from the sky. They decide to rest in the cabin, and note that time itself begins to break down. The players express their anger at the GM, and volcanos begin erupting. As the players walk out, the GM narrates how the world breaks apart, and everyone dies.

In this game, the outcome (everyone dies due to forces outside their control) was predetermined. Which forces killed them was determined by the players' stated actions. How do the various definitions and metrics of Player Agency rate this example?

[Frozen_Feet]

The discontinuity of argument here is because you're focusing on a specific type of choice ("things which affect the final outcome") and not the total number of choices.

Take Chess - the start of the game and the end game have only few possible choices. Majority of choices exists during midgame.

Does the lack of agency at the start, or lack of agency at the end, mean that the whole game lacks agency? The answer should be "obviously not".

Same principle applies to RPGs such as the one talked of above. Yeah, ultimately everyone dies. Does that mean players have no impact before the ultimate situation? I'm fairly sure the answer is "no".

[NichG]

Just like "balance", agency is not a synonym for fun. But, if we cannot stick to the topic of defining what agency is, we will continue talking past each other in talking about your more pragmatic "why".

That having been said, once we think we are in agreement as to "what", asking "why" and noting areas of dissonance is a great test of the extent to which we have failed at the "what" step. Thus, I highly encourage asking "why" as part of asking "what". But, just as one does not generally start with the roof when building a log cabin, starting with "why" is generally rather counterproductive to conversational efficiency.

The problem is that discussing 'what' independent of 'why' leaves one without any guide as to what actually constitutes a useful definition. Often bids for different definitions or ways of framing the discussion end up being driven by unspoken whys, so talking about it explicitly gets that out in the open. For example, Max_Killjoy's response that "Talking about it as a tool or a means to achieve something else, rather than a good thing in and of itself, also gives leave for some to push illusionism, for-your-own-good-ism, and so on." is a bid of that sort - it implicitly suggests that the reason to favor one way of framing over another is to marginalize a specific set of DM behaviors. As part of that is the implicit call that 'agency' must have a strict positive connotation, which I think is unhelpful. I had the same reaction to the 'agency as building material' objection to the 'agency is a tool' comment - its more about 'what I want my DM to believe' rather than 'what is it actually?'.

On the other hand, Frozen_Feet gave a very objective coverage of the term, specifically going as far to say that what the players feel from it doesn't matter, it's just the number of meaningful moves that they could take at any given moment.

For me, out of the ones discussed so far, I'd probably go more with PhoenixPhyre's definition.

To justify that, I have to go to the 'why' of things. It's the 'why' where, likely, we might differ in what we're actually trying to accomplish by discussing agency in the first place. I find the 'agency as building material' or 'agency is what the campaign is made of' kind of thing to be too focused on trying to establish what constitutes good gaming. Frozen_Feet's objective thing has the problem, for me, that while you can probably count it most cleanly of all the definitions, it introduces an extra layer where you have to constantly take into account that what matters for the actual game is some other thing (the players' current estimate or belief about their agency, or some-such) and I don't think that the objectivity of it is worth the extra fuss. I'd lean more towards it in a context such as measuring the agency of an artificial agent or something like that (I'd probably use the information theoretic definition of 'empowerment' rather than agency, for what its worth), but for tabletop gaming I think it hides the important psychological aspects.

If I'm going to use terms like 'agency' its because I would like those terms to ultimately be somewhat predictive. I want to be able to investigate statements such as 'when players apply their agency towards obtaining more agency, it feels like such and such', so that ultimately I have a better idea how games as a whole function - from the point of view of player, DM, and designer. On the other hand, I'm decidedly uninterested in using definitions to push for certain standards as to what constitutes good gaming. So I pushed back against what seemed to be a bid to include those normative elements into the discussion.

I can only conclude that the phrase "Player Agency" doesn't mean the same thing to you that it does to me.

So, I suppose it behooves me to explain my definition. Well, my revised definition, seeing as how I mentioned up-thread that I hadn't originally included the first step of knowledge.

So, to me, Player Agency involves the players getting to make choices for their characters that have a measurable, logical impact on the game. To make such choices, the players need to have knowledge to make informed choices, the characters need to have capabilities to perform actions, and the world needs to resolve in a logical way for the outcome of those choices to be predictable.

Now, here's where it gets tricky. I define agency as everything within the bounds of the characters capabilities. For example, if I'm playing a "normal" D&D character, it does not curtail my agency to not let me jump to the moon, because jumping to the moon is not within my capabilities. A game where my character cannot jump to the moon can still have 100% Player Agency.

Further, knowledge need not be perfect. However, knowledge and world consistency needs to be such that, when an event does not have the expected outcome, it is therefore obvious that our knowledge was inaccurate or incomplete. Everything should respond in a predictable way, and, when it does not, that must be indicative that there is something going on here to investigate.

Lastly, obtaining missing information must be possible. That may involve moving to OOC discussion (especially if the players and GM aren't on the same page / wavelength), and can even involve a retcon if the PCs should have known information of which the players were unaware.

For example, there are two things going on in your definitions here. First you lay out what you think agency is. So far so good, I'm roughly on board with that. But then you go on to dictate a bunch of prescriptions about what 'should' or 'must' happen or be possible, as well as make safe harbors for certain ways agency might be denied that might still not be unreasonable in a game.

I would say, stop after the first block (everything before 'it gets tricky'). The rest of it is not 'what is agency?', it's 'what do I expect of DMs with regards to my agency in their games?'. You can ask the second question, but I would like for the two to not become tangled.

[PhoenixPhyre]

I think there are several different scales of player agency being conflated here, much to the detriment of the discussion.

Large-scale agency is the ability of players to determine what the (next part of the) campaign is about. This includes deciding on a game system. In most games I've been in, this is strongest either at session 0 or between story arcs. It's also a collective agency--no single player (including the DM) should have full control over this. This is the strategy analogue--do we handle problem X or do we pursue goal Y. Once that has been decided, the players and the DM must deal with the consequences, which include more limited agency in this regard. Unilateral changes (e.g. DM deciding that the dungeon-crawling campaign they all agreed on would be better as a high-intrigue political campaign) are a violation of agency, specifically the consequences part. The consequences no longer follow from the choice. Choosing to run a module or a game like 12 candles is a use of agency that has consequences that limit further uses of large-scale agency. As long as this choice is not coerced or made under false pretenses ("This 12 candles game is a total sandbox!"), no violation of agency occurs because the later limits are the consequences of an earlier choice. In fact, this is a great example of agency in action. People knowingly made a choice that had meaningful consequences.

Small-scale agency is what is most commonly dealt with in-game, on-camera. Can my character make a knowing, meaningful choice at time T that affects the situation at time T'? Examples where a violation of this short-term agency makes anything better for anyone are much more rare. OOC lying (knowledge violation, whether by omission or commission) and tactical railroading (choice violation) have never worked in my experience. Fudging or ret-coning (consequences violation) has been needed, but only as the lesser of two evils due to a mistake on the DM or player's part whose consequences were unsupportable within the game's framework.

Note that voluntarily surrendered agency is not an agency violation. A player can choose to hand off the choice to someone or something else (the DM, another player, or the dice). However, they must do so knowingly, willingly, and most importantly be willing to accept the consequences of doing so. Actions (uses of agency) can also limit further agency without being a violation of agency. If you get yourself (in real life) arrested and imprisoned, you don't have as much agency as before. But that's a consequence of your actions.

Agency is also not unbounded. A person (character or not) can't claim agency to make choices that are outside his or her sphere of rightful influence. This includes the actions of others, setting details once play begins (unless that authority has been granted by the DM or by a system rule), etc. While playing D&D, I can't claim that rolling a 1 is a hit unless I have an ability/feat/etc that grants that power. That's a consequence of the choice of system. And so forth. Not being able to jump to the moon is not a violation of agency.

[Quertus]

The problem is that discussing 'what' independent of 'why' leaves one without any guide as to what actually constitutes a useful definition. Often bids for different definitions or ways of framing the discussion end up being driven by unspoken whys, so talking about it explicitly gets that out in the open. For example, Max_Killjoy's response that "Talking about it as a tool or a means to achieve something else, rather than a good thing in and of itself, also gives leave for some to push illusionism, for-your-own-good-ism, and so on." is a bid of that sort - it implicitly suggests that the reason to favor one way of framing over another is to marginalize a specific set of DM behaviors. As part of that is the implicit call that 'agency' must have a strict positive connotation, which I think is unhelpful. I had the same reaction to the 'agency as building material' objection to the 'agency is a tool' comment - its more about 'what I want my DM to believe' rather than 'what is it actually?'.

On the other hand, Frozen_Feet gave a very objective coverage of the term, specifically going as far to say that what the players feel from it doesn't matter, it's just the number of meaningful moves that they could take at any given moment.

For me, out of the ones discussed so far, I'd probably go more with PhoenixPhyre's definition.

To justify that, I have to go to the 'why' of things. It's the 'why' where, likely, we might differ in what we're actually trying to accomplish by discussing agency in the first place. I find the 'agency as building material' or 'agency is what the campaign is made of' kind of thing to be too focused on trying to establish what constitutes good gaming. Frozen_Feet's objective thing has the problem, for me, that while you can probably count it most cleanly of all the definitions, it introduces an extra layer where you have to constantly take into account that what matters for the actual game is some other thing (the players' current estimate or belief about their agency, or some-such) and I don't think that the objectivity of it is worth the extra fuss. I'd lean more towards it in a context such as measuring the agency of an artificial agent or something like that (I'd probably use the information theoretic definition of 'empowerment' rather than agency, for what its worth), but for tabletop gaming I think it hides the important psychological aspects.

If I'm going to use terms like 'agency' its because I would like those terms to ultimately be somewhat predictive. I want to be able to investigate statements such as 'when players apply their agency towards obtaining more agency, it feels like such and such', so that ultimately I have a better idea how games as a whole function - from the point of view of player, DM, and designer. On the other hand, I'm decidedly uninterested in using definitions to push for certain standards as to what constitutes good gaming. So I pushed back against what seemed to be a bid to include those normative elements into the discussion.



For example, there are two things going on in your definitions here. First you lay out what you think agency is. So far so good, I'm roughly on board with that. But then you go on to dictate a bunch of prescriptions about what 'should' or 'must' happen or be possible, as well as make safe harbors for certain ways agency might be denied that might still not be unreasonable in a game.

I would say, stop after the first block (everything before 'it gets tricky'). The rest of it is not 'what is agency?', it's 'what do I expect of DMs with regards to my agency in their games?'. You can ask the second question, but I would like for the two to not become tangled.

... Despite making the thread, I can't really keep up with it. So I'm just skimming, and poking at large areas of dissonance.

I fully agree with the your desire to remove notions of positive connotations, and to smoke out hidden agendas. I just disagreed with your methods, in part because they felt like what you were arguing against.

When watching a movie, I almost always have zero agency to affect the outcome. And this is not a bad thing. Lack of agency is not inherently bad.

So, regarding your disagreement with the existence of the second half of my definition, how would you suggest I go about expressing the, to me, utterly integral and indispensable notion that my definition of Player Agency is strictly bounded by the character's capabilities?

I think there are several different scales of player agency being conflated here, much to the detriment of the discussion.

Large-scale agency is the ability of players to determine what the (next part of the) campaign is about. This includes deciding on a game system. In most games I've been in, this is strongest either at session 0 or between story arcs. It's also a collective agency--no single player (including the DM) should have full control over this. This is the strategy analogue--do we handle problem X or do we pursue goal Y. Once that has been decided, the players and the DM must deal with the consequences, which include more limited agency in this regard. Unilateral changes (e.g. DM deciding that the dungeon-crawling campaign they all agreed on would be better as a high-intrigue political campaign) are a violation of agency, specifically the consequences part. The consequences no longer follow from the choice. Choosing to run a module or a game like 12 candles is a use of agency that has consequences that limit further uses of large-scale agency. As long as this choice is not coerced or made under false pretenses ("This 12 candles game is a total sandbox!"), no violation of agency occurs because the later limits are the consequences of an earlier choice. In fact, this is a great example of agency in action. People knowingly made a choice that had meaningful consequences.

Small-scale agency is what is most commonly dealt with in-game, on-camera. Can my character make a knowing, meaningful choice at time T that affects the situation at time T'? Examples where a violation of this short-term agency makes anything better for anyone are much more rare. OOC lying (knowledge violation, whether by omission or commission) and tactical railroading (choice violation) have never worked in my experience. Fudging or ret-coning (consequences violation) has been needed, but only as the lesser of two evils due to a mistake on the DM or player's part whose consequences were unsupportable within the game's framework.

Note that voluntarily surrendered agency is not an agency violation. A player can choose to hand off the choice to someone or something else (the DM, another player, or the dice). However, they must do so knowingly, willingly, and most importantly be willing to accept the consequences of doing so. Actions (uses of agency) can also limit further agency without being a violation of agency. If you get yourself (in real life) arrested and imprisoned, you don't have as much agency as before. But that's a consequence of your actions.

Agency is also not unbounded. A person (character or not) can't claim agency to make choices that are outside his or her sphere of rightful influence. This includes the actions of others, setting details once play begins (unless that authority has been granted by the DM or by a system rule), etc. While playing D&D, I can't claim that rolling a 1 is a hit unless I have an ability/feat/etc that grants that power. That's a consequence of the choice of system. And so forth. Not being able to jump to the moon is not a violation of agency.

I don't consider what you call large-scale agency to be a part of Player Agency. But I do consider "take an action at time T that has an effect at time T+X, not just at time T" to fall under Player Agency. And that includes very large-scale, campaign-changing consequences.

[NichG]

... Despite making the thread, I can't really keep up with it. So I'm just skimming, and poking at large areas of dissonance.

I fully agree with the your desire to remove notions of positive connotations, and to smoke out hidden agendas. I just disagreed with your methods, in part because they felt like what you were arguing against.

When watching a movie, I almost always have zero agency to affect the outcome. And this is not a bad thing. Lack of agency is not inherently bad.

So, regarding your disagreement with the existence of the second half of my definition, how would you suggest I go about expressing the, to me, utterly integral and indispensable notion that my definition of Player Agency is strictly bounded by the character's capabilities?

I guess I'd say it's better as a logical conclusion about the practical limitations of agency in any given situation rather than being stated as part of the definition of agency itself. If for example I were to say 'agency is what can be intentionally brought about', then if the character is the bottleneck through which actions are taken then things that the character cannot achieve will naturally not be part of that player's agency without requiring any additional caveats. The word 'intentionally' would cover the knowledge aspect, and the word 'can' covers the limits imposed by e.g. character capabilities.

[Max_Killjoy]

I think there are several different scales of player agency being conflated here, much to the detriment of the discussion.

Large-scale agency is the ability of players to determine what the (next part of the) campaign is about. This includes deciding on a game system. In most games I've been in, this is strongest either at session 0 or between story arcs. It's also a collective agency--no single player (including the DM) should have full control over this. This is the strategy analogue--do we handle problem X or do we pursue goal Y. Once that has been decided, the players and the DM must deal with the consequences, which include more limited agency in this regard. Unilateral changes (e.g. DM deciding that the dungeon-crawling campaign they all agreed on would be better as a high-intrigue political campaign) are a violation of agency, specifically the consequences part. The consequences no longer follow from the choice. Choosing to run a module or a game like 12 candles is a use of agency that has consequences that limit further uses of large-scale agency. As long as this choice is not coerced or made under false pretenses ("This 12 candles game is a total sandbox!"), no violation of agency occurs because the later limits are the consequences of an earlier choice. In fact, this is a great example of agency in action. People knowingly made a choice that had meaningful consequences.

Well said.

Player agency includes the informed choice to sit down at the table and take part in the game that players (including the GM) choose to play, including the setting, rules, and campaign guidelines.

Small-scale agency is what is most commonly dealt with in-game, on-camera. Can my character make a knowing, meaningful choice at time T that affects the situation at time T'? Examples where a violation of this short-term agency makes anything better for anyone are much more rare. OOC lying (knowledge violation, whether by omission or commission) and tactical railroading (choice violation) have never worked in my experience. Fudging or ret-coning (consequences violation) has been needed, but only as the lesser of two evils due to a mistake on the DM or player's part whose consequences were unsupportable within the game's framework.

Note that voluntarily surrendered agency is not an agency violation. A player can choose to hand off the choice to someone or something else (the DM, another player, or the dice). However, they must do so knowingly, willingly, and most importantly be willing to accept the consequences of doing so. Actions (uses of agency) can also limit further agency without being a violation of agency. If you get yourself (in real life) arrested and imprisoned, you don't have as much agency as before. But that's a consequence of your actions.

Agency is also not unbounded. A person (character or not) can't claim agency to make choices that are outside his or her sphere of rightful influence. This includes the actions of others, setting details once play begins (unless that authority has been granted by the DM or by a system rule), etc. While playing D&D, I can't claim that rolling a 1 is a hit unless I have an ability/feat/etc that grants that power. That's a consequence of the choice of system. And so forth. Not being able to jump to the moon is not a violation of agency.

Yes -- having agreed to a set of boundaries at the large scale on one's agency does not remove all agency, and one's character being bound by the fictional reality or by the consequences of said character's prior choices and actions does not remove all agency.

We see quite a bit of a false dichotomy, however, asserting that player agency can only be total or non-existent, so that players are either allowed to do anything they want, or have no agency at all -- and oddly enough, we see it on opposite fringes of the gaming community, espoused both by those who insist that players are along for a ride completely controlled by the GM, and those who insist that everything in the game must come from (and only from) the players during the course of play.

Or we see a bit of rhetorical trickery that goes something like this: "If player agency is a good thing, then more of it is obviously better, and absolute agency is absolutely the best... right?" This is either done to "prove" that unbound player agency is the "best" form of game, or to "prove" that player agency isn't a good thing. Which is cute, but ignores the basic fact that it's possible to have too much of a good thing. Water is good, but too much will drown you. Air is good, but too much in too small a space is dangerous. Food is good, but too much will make you sick. Etc.

I fully agree with the your desire to remove notions of positive connotations, and to smoke out hidden agendas. I just disagreed with your methods, in part because they felt like what you were arguing against.

When watching a movie, I almost always have zero agency to affect the outcome. And this is not a bad thing. Lack of agency is not inherently bad.

Why would we want to remove the positive connotations?

Player agency is a core part of what makes an RPG an RPG.

[PhoenixPhyre]

I don't consider what you call large-scale agency to be a part of Player Agency. But I do consider "take an action at time T that has an effect at time T+X, not just at time T" to fall under Player Agency. And that includes very large-scale, campaign-changing consequences.

Why not? Players (including the DM) make informed, meaningful choices well before characters even enter picture. Key decisions include choice of system, choice of setting, character construction choices (playing Quertus results in different game-play than does playing Arnus), choice of story (play a module or not? If so, which module? These have large effects in play.). In fact, there is only one choice that cannot (barring a Saw "I want to play a game" scenario) be taken away, and it's a large-scale choice. The choice to play or not to play.

More broadly, defining agency in terms of characters obscures one key idea--responsibility. Characters have no agency, and bear no responsibility for their choices. Only players do. Hiding behind "It's what my character would do" to excuse antagonistic or disruptive play is trying to shift the blame to the character. But the player made those choices, not the character. What that statement really means is "I want to be disruptive, but I don't want to suffer the consequences of disruption." It's a denial of your own agency, and is false. The character is who you say he is--decisions to grow/change are on your shoulders, not the character's.

For example, you could decide that Quertus is tired of being tactically inept and that he's going to put his enormous intelligence to work to solve that problem. That's a large-scale agency choice--you're choosing his direction independently of the exact circumstances. And it's entirely up to you (at least in D&D)--there are no mechanical or other binding restraints on character personality and growth.

[jayem]

Having 100% PA mean 100% character capability seems reasonable enough. We can always call Fate a 200% player agency variant which kind of shows somethings up.

I'm not sure I'd count PhoenixPhyre "long term agency" as long term agency, I'd reserve that for long term in game consequences of in game actions (particularly character actions, but it is PA rather than PCA), and give that a different name (I agree it is important-and a type of agency).

I do like the "meaningful" "choice" definition..

[PhoenixPhyre]

Having 100% PA mean 100% character capability seems reasonable enough. We can always call Fate a 200% player agency variant which kind of shows somethings up.

I'm not sure I'd count PhoenixPhyre "long term agency" as long term agency, I'd reserve that for long term in game consequences of in game actions (particularly character actions, but it is PA rather than PCA), and give that a different name (I agree it is important-and a type of agency).

I do like the "meaningful" "choice" definition..

Large-scale, not long-term. It's about granularity and scope. Large-scale agency has to do with what kind of game you're playing or goals and attitudes of a character, not actions in-character. Maybe calling them "Out-of-Character Agency" and "In-character agency" would be clearer?

With large-scale agency, I, the player, am making choices as myself, for myself. With small-scale agency, I, the player, am making choices for a character (or characters in the DM's case) in a specific situation. Both are player agency, but they're very different types of things.

[jayem]

Maybe calling them "Out-of-Character Agency" and "In-character agency" would be clearer?

Possibly, I made the jump (large-scale=big-consequences=long-time) when reading the titles as to what to expect.

[edit-though that I suppose has a conflict with Fate style agency. in any case you've described what you mean quite clearly]

[OldTrees1]

Wow. You must have a lot of experience with legalise (how does one spell legal ease?). Based on my (revised) definition, below, do you feel we differ on the "meaningful" bit of your definition?

So, I suppose it behooves me to explain my definition. Well, my revised definition, seeing as how I mentioned up-thread that I hadn't originally included the first step of knowledge.

So, to me, Player Agency involves the players getting to make choices for their characters that have a measurable, logical impact on the game. To make such choices, the players need to have knowledge to make informed choices, the characters need to have capabilities to perform actions, and the world needs to resolve in a logical way for the outcome of those choices to be predictable.

Now, here's where it gets tricky. I define agency as everything within the bounds of the characters capabilities. For example, if I'm playing a "normal" D&D character, it does not curtail my agency to not let me jump to the moon, because jumping to the moon is not within my capabilities. A game where my character cannot jump to the moon can still have 100% Player Agency.

Further, knowledge need not be perfect. However, knowledge and world consistency needs to be such that, when an event does not have the expected outcome, it is therefore obvious that our knowledge was inaccurate or incomplete. Everything should respond in a predictable way, and, when it does not, that must be indicative that there is something going on here to investigate.

Lastly, obtaining missing information must be possible. That may involve moving to OOC discussion (especially if the players and GM aren't on the same page / wavelength), and can even involve a retcon if the PCs should have known information of which the players were unaware.

So, to pull an example from another thread, if, in the bizarre game world, everyone knows that, contrary to the real world, barn fires are actually salvageable situations, then the PCs need to have that information to make an informed decision about what to do about a barn fire in order to be considered to have agency under my definition.

Let me put forth another bizarre world. The players say the PCs look around; suddenly, the sun grows much brighter and hotter. The PCs look for shelter. As they race for a convenient log cabin, the sun grows even brighter, and lightning and hail begin descending from the sky. They decide to rest in the cabin, and note that time itself begins to break down. The players express their anger at the GM, and volcanos begin erupting. As the players walk out, the GM narrates how the world breaks apart, and everyone dies.

In this game, the outcome (everyone dies due to forces outside their control) was predetermined. Which forces killed them was determined by the players' stated actions. How do the various definitions and metrics of Player Agency rate this example?

You define Agency as the ability to have impact and then start detailing what is necessary for that impact.
I defined Agency as the ability to have impact arising from being faced with meaningful choices and I defined Meaningful Choices in a manner that I expect the ability for a player to impact the world through them is an emergent property of the kind of choice.

We did differ on whether the characters capabilities were included in or treated as bounds for the definition of Agency. Since I prefer to run Sandbox games, I see that the characters abilities are limits on their ability to impact the world. This usually leads to be starting such campaigns at a higher level than I otherwise would so that the Players have a larger ability to impact the world. This intentional adjustment for the sake of increasing the ability to cause impact, leads me towards considering the limits of character capabilities as an artificial ceiling on the Agency in that game but not on the scale in general. Aka a character that can jump to the moon does have more potential agency than the identical character that cannot (all that is required is for the DM to make going to the moon have impact on the game). But a character does not need total Agency under that definition, because limitations on character capabilities have value in campaign design.

I think you went from describing to prescribing when you talked about missing information MUST be obtainable. For a definition of Agency I think we can leave it as the choice or part of a choice that depended on that missing information was not a meaningful choice but the rest might have been. I can play a Monty Hall puzzle and have some agency (informed impactful choice to always switch) despite not having all the information needed to know the result.

Large-scale, not long-term. It's about granularity and scope. Large-scale agency has to do with what kind of game you're playing or goals and attitudes of a character, not actions in-character. Maybe calling them "Out-of-Character Agency" and "In-character agency" would be clearer?

With large-scale agency, I, the player, am making choices as myself, for myself. With small-scale agency, I, the player, am making choices for a character (or characters in the DM's case) in a specific situation. Both are player agency, but they're very different types of things.

If the DM frames questions about what the next campaign would be about in the form of meaningful choices the PCs face, is that Large Scale or Short Scale in your terminology?

What about PCs that construct situations where they get meaningful choices about how to shape the world? Imagine a Rogue that is Guildmaster of all the thieves guilds on the material plane.

[PhoenixPhyre]

If the DM frames questions about what the next campaign would be about in the form of meaningful choices the PCs face, is that Large Scale or Short Scale in your terminology?

What about PCs that construct situations where they get meaningful choices about how to shape the world? Imagine a Rogue that is Guildmaster of all the thieves guilds on the material plane.

The fundamental difference (as I put in my second response) is one of who the decisions are being made for.

Are the players choosing what type of game they want to play? Large-scale (or Out of character) agency. This can include making decisions that strongly restrict the agency later ("the entire game will take place in a single room with no way out or in" is an exercise of OOC agency). Characters may not have even been created yet. Another example is this--I gave a couple of groups a choice from a set of campaign seeds. These included a brief description of what environments the characters (which had completed a tutorial mission at this point) would face, what enemies, and what the over-arching goal of the campaign would be. This is OOC agency--the characters had no way of knowing these details, but the players could choose which they preferred (believed that they would find more fun).

Are the players choosing actions for a restricted set of characters in a scenario/set of circumstances? That's small-scale (IC agency). 90% of the time, this is the dominant play loop of D&D-like games.

1) DM narrates a situation.
2) Players decide what their character attempts. Players and DM resolve the attempt.
3) GOTO 1 (updating the situation with the changes due to the attempt).

Of course, the two influence each other--neither stands alone and this is not a binary, it's a set of continua. Here are four exceptional cases that showcase the differences. All start with "A group of friends gather to play a game."

1) High OOC, low IC agency: The participants set initial parameters (build a setting, choose a system, design characters, etc) and feed that information into a glorified version of Conway's Game of Life (a zero-external-input simulation engine). Their only input was at the configuration stage, but those actions mattered.

2) Low OOC, high IC agency: They randomly (as in, using a randomizer) pick a system and a setting with pre-built characters (including personality and goals). They then play a game where each player strives to follow the pre-built personalities and goals in whatever way they feel best fits the character. Their decisions shape the outcome.

3) Low OOC, low IC agency: An external DM arrives with a pre-generated, highly-railroaded module. Their only choice is to play the pre-generated characters through the scripted plot, or not play at all. They can make choices, but they don't change anything significant.

4) High OOC, high IC agency: They play a free-form, DM-less (high narrative control) game in a self-defined setting. The only rules are consensus. Actions both OOC and IC matter strongly at every step.

Of course, most TTRPGs are not at any of these extremes (except maybe DU's games). D&D tends to limit setting/rules-related OOC agency while giving strong character-related OOC agency (defining build, personality and goals for a character). Running modules restricts IC agency, but since that's knowingly agreed on in advance, it's not a harmful agency violation in my opinion.

[OldTrees1]

The fundamental difference (as I put in my second response) is one of who the decisions are being made for. -snip-

Crystal clear.

Yes, I think that is a good distinction to remember.

[Quertus]

I guess my issue with certain definitions of Player Agency is that the choice between playing Checkers and playing Chess has nothing to do with the degree to which Chess gives you agency.

Further, as a player, you absolutely have agency at all phases of Chess - you absolutely have the option to make a suboptimal move, just like you have the option to pick a level in Commoner or take an attack of opportunity on your ally in D&D, or play with Muck Dwellers or 1000 islands + 4 Sword of the Ages in MtG. Now, does the fact that certain moves are suboptimal trap options say something about the option to pick them, with regards to agency? Currently, unlike with Illusions or Inevitable Ogres, I'm coming down on the side of "no", but I may be wrong.

Let me flip that around. Playing Quertus, or Armus, or any other character I'd care to play, they have a personality. Role-playing that personality, and their non-omniscient perspective on reality, greatly curtails my options. Yet I don't feel a loss of Agency. Why not? (I don't have an answer for this. Have I lost Agency without realizing it? Or should Player Agency be defined in some way such that options curtailed by role-playing do not result in a loss of Agency?)

I guess I'd say it's better as a logical conclusion about the practical limitations of agency in any given situation rather than being stated as part of the definition of agency itself. If for example I were to say 'agency is what can be intentionally brought about', then if the character is the bottleneck through which actions are taken then things that the character cannot achieve will naturally not be part of that player's agency without requiring any additional caveats. The word 'intentionally' would cover the knowledge aspect, and the word 'can' covers the limits imposed by e.g. character capabilities.

Hmmm... What if we consider those paragraphs to be clarifying the meanings of my previous words? They are a test: if you read them to be opposed to each other, then I haven't been clear? Most software developers can't test their own code - I can. I think this easy in terms of safeguards and fall conditions.

But, you're right - I should work harder to make sure that the first half of the definition stands alone, and clearly supports the second (interpreted) half.

Why would we want to remove the positive connotations?

Player agency is a core part of what makes an RPG an RPG.

Because I am even more biased than you in believing that Player Agency is the bestest thing ever, and essential to a good RPG.

Because a definition of Player Agency should be able to stand in its own, independent of a characterization of its value to an RPG.

Because there are those who contend that there are times when Agency is suboptimal, and if Player Agency is conflated with being better, it makes for a more difficult conversation.

Or, on a more personal note, because you dislike supernatural mind control based on your previous bad experiences with it under a railroading GM. Whereas me, I haven't had horrible experiences with it, so I don't dislike it. And my definition of Player Agency doesn't call out supernatural mind control as removing Agency (as, to me, mind control merely changes the list of what is within the character's capabilities). And I want to be able to have this discussion about what is Player Agency without crossing wires about what is good and bad.

And because, if possible, I want to avoid a scenario like, "if the players don't like it, it's railroading". Where someone contends that, for me, mind control isn't a loss of Player Agency, but, for you, it is.

Maybe calling them "Out-of-Character Agency" and "In-character agency" would be clearer?

I agree with this nomenclature.

Why not? Players (including the DM) make informed, meaningful choices well before characters even enter picture. Key decisions include choice of system, choice of setting, character construction choices (playing Quertus results in different game-play than does playing Arnus), choice of story (play a module or not? If so, which module? These have large effects in play.). In fact, there is only one choice that cannot (barring a Saw "I want to play a game" scenario) be taken away, and it's a large-scale choice. The choice to play or not to play.

Why do I not include "OOC agency" in my definition of Player Agency? That's a very good question, and it's questions like this that make me glad I made this thread.

Hmmm... As I said above, the presence or absence of the choice between playing Chess and playing Checkers tells us nothing about how much Agency one has while playing Chess.

I can't say that I don't care about OOC Agency - I would very much rather play a character that I know that I'll enjoy than have to take a 20 creating new characters until one fits.

But I also care about how my gaming snacks tastes, what dice I use, what my mini looks like, how my character sheet is laid out, how well lit the room is, what house rules we are using, how many 5-year-olds are convenient for demonstrating points. But caring about a thing does not make it related to Agency. So, what does?

My personal definition going into this thread revolved (and still revolves) around the character's capabilities. As such, OOC Agency does not fall within my personal definition of Player Agency. To pull an example from another thread, to me, getting to create entire countries for RPGs is as relevant to Player Agency as getting to name the princess in Grandpa's (railroading) story. That is to say, not at all, because it has nothing to do with the character having an impact on the game.

But, afaict, it's entirely circular logic. So I'm perfectly content to have two separate labels for IC and OOC Player Agency, and determining ways to differentiate between them.

More broadly, defining agency in terms of characters obscures one key idea--responsibility. Characters have no agency, and bear no responsibility for their choices. Only players do. Hiding behind "It's what my character would do" to excuse antagonistic or disruptive play is trying to shift the blame to the character. But the player made those choices, not the character. What that statement really means is "I want to be disruptive, but I don't want to suffer the consequences of disruption." It's a denial of your own agency, and is false. The character is who you say he is--decisions to grow/change are on your shoulders, not the character's.

For example, you could decide that Quertus is tired of being tactically inept and that he's going to put his enormous intelligence to work to solve that problem. That's a large-scale agency choice--you're choosing his direction independently of the exact circumstances. And it's entirely up to you (at least in D&D)--there are no mechanical or other binding restraints on character personality and growth.

... This is a much larger conversation. But the quick of it is, I believe in a much more narrow range of likely behaviors than I suspect you do.

The first step to fixing a problem is realizing that there is one. Quertus has no concept that he is tactically inept. He had received decades of reinforcement of his current behavior. While he knows he's not an Imperial War Wizard, and that he lacks their training and understanding of battlefield tactics, he does not have the perspective to recognize just how bad his choices are. Further, he knows enough to have saved over 100 worlds, and the only person I can remember trying to teach Quertus tactics somehow managed to be worse at it than Quertus (they recommended buffing the enemy). Quertus has no desire to learn "advanced" combat tactics - all he wants is to retire!

So Quertus has nothing to build from to recognize the problem and develop a desire to improve.

On the larger issue of "it's what my character would do", I think the focus is wrong. WWJD (etc) are a thing precisely because those letters have meaning - to deny that should be clearly wrong minded.

It takes a much more careful analysis of the elements to determine exactly what caused any given problem, and what can reasonably be done to fix it, than to just blithely assume that characterization of a character is meaningless.

I think you went from describing to prescribing when you talked about missing information MUST be obtainable. For a definition of Agency I think we can leave it as the choice or part of a choice that depended on that missing information was not a meaningful choice but the rest might have been. I can play a Monty Hall puzzle and have some agency (informed impactful choice to always switch) despite not having all the information needed to know the result.

Hmmm... I probably rolled a 1 on expression. Let me try again.

Science likes to believe that the real world is a predictable place, with rhyme and reason and repeatable phenomenon.

I believe I was attempting to contend that, for Player Agency to exist, similar principles must apply to the game world. As a test of such principles, I contended that "all information is learnable" would potentially suffice; further, that being able to learn it OOC was adequate. I am, however, currently questioning the accuracy of my contention.

Running modules restricts IC agency, but since that's knowingly agreed on in advance, it's not a harmful agency violation in my opinion.

Well, now, that's an interesting question: if you're playing an Adventure Path, you know that if you stray too far off the rails, you cannot continue, and you've lost the game. But, you technically still have the option to do so. So is it truly a loss of Player Agency?

[NichG]

Let me flip that around. Playing Quertus, or Armus, or any other character I'd care to play, they have a personality. Role-playing that personality, and their non-omniscient perspective on reality, greatly curtails my options. Yet I don't feel a loss of Agency. Why not? (I don't have an answer for this. Have I lost Agency without realizing it? Or should Player Agency be defined in some way such that options curtailed by role-playing do not result in a loss of Agency?)

Hmmm... What if we consider those paragraphs to be clarifying the meanings of my previous words? They are a test: if you read them to be opposed to each other, then I haven't been clear? Most software developers can't test their own code - I can. I think this easy in terms of safeguards and fall conditions.

There's again two separate things though. You're talking here about your feeling that agency was curtailed, not actually what agency you have. In other places you talk about things like '100% player agency'. That all speaks of a framework that wants to assume a default and then talk in terms of way that it is limited (but, notice, doing it that way makes it really hard to talk about ways that it might be increased, necessitating the awkward 200% agency thing when talking about games like FATE).

So if we talk about the 'why?', it makes me think - what could the 'why?' be behind wanting to talk about agency more in terms of how it is modified from some assumed baseline, than to talk about agency independent of that expectation?

If we're talking about your characters that you play, it only seems weird that you'd voluntarily lower your own agency if you assume a normative definition - e.g. 'agency is always good and you always want as much of it as possible up to the natural baseline', so the definition gets tangled. That's where this question of defining agency such that certain limits don't count as a loss comes from - it only makes sense if it's unacceptable to ever consider agency as something you don't want or need as much of as possible. But if agency is just a particular thing, then its not strange to sometimes want to trade off agency or spend agency or sacrifice agency or so on. So in that case, you can start to examine why some decreases of agency feel bad, but other decreases of agency don't, without the implicit assumption that it must be because in one case you aren't 'really' losing agency.

For example, if we go with 'agency is the outcomes which you can intentionally bring about', then choosing to use agency always means that your agency goes down - you're committing to a particular outcome, whereas before you made the decision you could have chosen any of those achievable outcomes. But it goes down in exchange for getting the outcome you want, which is in that moment more satisfying than keeping the breadth of options alive. So the experience of how interacting with agency feels can be more complicated than just 'more = good', and the discussion can be about exploring that psychology rather than exploring the meaning of agency itself.

[OldTrees1]

I guess my issue with certain definitions of Player Agency is that the choice between playing Checkers and playing Chess has nothing to do with the degree to which Chess gives you agency.

Further, as a player, you absolutely have agency at all phases of Chess - you absolutely have the option to make a suboptimal move, just like you have the option to pick a level in Commoner or take an attack of opportunity on your ally in D&D, or play with Muck Dwellers or 1000 islands + 4 Sword of the Ages in MtG. Now, does the fact that certain moves are suboptimal trap options say something about the option to pick them, with regards to agency? Currently, unlike with Illusions or Inevitable Ogres, I'm coming down on the side of "no", but I may be wrong.

Let me flip that around. Playing Quertus, or Armus, or any other character I'd care to play, they have a personality. Role-playing that personality, and their non-omniscient perspective on reality, greatly curtails my options. Yet I don't feel a loss of Agency. Why not? (I don't have an answer for this. Have I lost Agency without realizing it? Or should Player Agency be defined in some way such that options curtailed by role-playing do not result in a loss of Agency?)

Shifted example might bring clarity: There are many variations to chess. If you both give each other the choice between bughouse and vanilla chess, that does speak to the agency in the game you both are playing even if it does not speak to the agency within bughouse or vanilla chess.

Why do you feel like you have not lost agency going from checkers to RPGs despite the transition from a perfect information game to an imperfect information game? It is probably because RPGs give you more options even after removing most of the options it adds. Armus and Quertus literally could play a game of chess in an RPG but instead engage in a imperfect information game against a chessmaster BBEG.

Hmmm... I probably rolled a 1 on expression. Let me try again.

Science likes to believe that the real world is a predictable place, with rhyme and reason and repeatable phenomenon.

I believe I was attempting to contend that, for Player Agency to exist, similar principles must apply to the game world. As a test of such principles, I contended that "all information is learnable" would potentially suffice; further, that being able to learn it OOC was adequate. I am, however, currently questioning the accuracy of my contention.

While I usually run sandboxes with enough agency for the PCs to be major players, I tend to use Lovecraftian Horror as my examples. Go figure.

Imagine a horror game full of a mixture of blind (the choice has impact but the players are uninformed when choosing) and meaningful choices that the PCs run through as they try to survive. Would it detract from the meaningful choices if the blind choices did not have "all information is learnable"?

Say there was a fork in the hallway & the PCs ran down one direction. There was no information available to the PCs (blind choice). Later they see a creature and choose to engage to slink away based upon its description. The DM never reveals to the players what was down the other direction (other than that it was different). Did being unable to learn OOC about the other direction detract from the meaningful choice they had later?

Say they engaged the creature and chased it to another fork in the hallway. This time there is a slime trail leading towards the right. The PCs choose whether to follow the slime trail (unknowingly towards the science lab) or go left away from the creature (and unknowingly towards the library). If the DM never reveals the lab/library that was in the other direction, did that detract from the meaningful choice to follow or avoid the slime trail?

I do not think the lack of information matters as long as we include meaningful choices and the PCs being informed about the details of the choice as a necessary condition of it being meaningful.

[Floret]

I do not think the lack of information matters as long as we include meaningful choices and the PCs being informed about the details of the choice as a necessary condition of it being meaningful.

I'd absolutely say it does. Without any information about the situation (as in the two hallways) this is effectively akin to someone telling you to make a cointoss. Are you exercising your agency by doing it? Maybe a bit, you could refuse.
But are you exercising your agency by whether you get heads or tails? Not in a million years.

I agree that it feels like having more control than when someone else throws that coin for you (even though that's entirely illusionary), but the result is effectively random (to the one making the decision), and cannot be said to be determined by the decision anymore than the result of a dieroll is.

Thinking about it, this is another reason why I would put percieved agency as the thing more relevant to care about than actual agency* - yes, the decision lead to different results, but no matter if we classify that as agency or not, the important part is that it doesn't feel to the player as if they had any sort of meaningful control (even though they did).

*While noting that it is nigh impossible to create the feeling of agency without a great deal of actual agency being given.

[oxybe]

Let me flip that around. Playing Quertus, or Armus, or any other character I'd care to play, they have a personality. Role-playing that personality, and their non-omniscient perspective on reality, greatly curtails my options. Yet I don't feel a loss of Agency. Why not? (I don't have an answer for this. Have I lost Agency without realizing it? Or should Player Agency be defined in some way such that options curtailed by role-playing do not result in a loss of Agency?)

Every choice you make up until the game starts is usually about deciding where you want to apply or focus your agency on once gameplay starts, by making and weighing the opportunity costs for doing so. You can play Quertus, Armus or any other character, but by doing so you're restricting your ability to play those unchosen characters and their unique options.

One could largely that that lost "potential agency" is relegated to acceptable losses and you've largely moved on in your mind.

Similarly to how deciding to play a fantasy game restricts the types of characters you can make, and further the setting and genre may put more restrictions, think darksun and it's choice of races for example: If you accept to play in a game of Darksun, you've tossed playing Crunk the Half-Orc aside as well as any character with a beef VS orcs. Those are things you've already instantly accepted in the back of your mind as acceptable losses where your agency is concerned when you chose to play in that Darksun game.

If "no orcs" is deal breaker, you can always choose to apply your single greatest bit of agency as a player and just choose to not play in that game, same as if you don't care for games that feature heavy diplomacy or combat.