What is Player Agency?

[Mordaedil]

{Laughing} A mere 32 pages? That one thread? I've seen a lot more than that from him. The fact I still try to have a serious conversations with him probably makes me insane, but I've never claimed otherwise.

I mean, stubborness is a weakness in face of a wall. I can certainly relate, but I think I've kind of got an excuse and I've learned to kinda steer away from discussions that go nowhere with people nowadays.

I did read the thread, and I've read his posts in other threads, but it's all kind of nicely collected in that thread, all of his spectrum. I mean, it's pretty one-note, even if it does surprise me how hard it is for that person to put a stake anywhere.

[Darth Ultron]

You don't need to affect the world in order to effect the game. That's a misunderstanding you seem to have. If I got a choice between ally myself with either a noble or a criminal, that choice might not affect the world but certainly will effect the game (in the sense that it will, for example, lead to different adventures).

Now see this is where I see very little difference or just slight flavor difference. There is an war between the orcs and elves, what side the characters join does not really matter much. The war will go onward no matter what side the characters choose. If the characters are present for a big turning point, they can act and maybe change the course of the whole war or even end it, but it does not matter what side they are on. After all, even if they did pick the elf side...they can still do (or try) *anything*.

And a lot of the time it won't effect the overall plot anyway. You have the noble baron fighting the criminal guild. The baron is good and the criminals bad. Both more then welcome the help of the ''unknown outsider'' characters. And both intend to use and then get rid of the characters. OR it is possible both intend to thank the characters and give them a ton of treasure. OR they both might just ''cut them loose''. Or one might do one thing and the other might do another. It is impossible to just ''guess'' what might happen, even in vague terms.

So ''one'' choice like this, won't matter at all for the future of the game: anything can happen. But then it is just *one* choice, and if you make it and forget about it...then things can spin in any direction. Though if you keep the choice more active, and keep and eye on your allies, you might have much more say in what happens.

Like:

Game A: The characters pick the criminal side (''crime is cool") and forget about it and just have fun being bad. It is no surprise to anyone, except the players and characters, that the criminals attempt to kill the characters once they have no more use.

Game B: The characters pick the Good Baron side (''Good is cool'') and forget about it and just have fun being good guys. The good baron is more then happy to keep his hands clean and let the characters do the dirty work, and once it is done...arrest the characters (''well, it does seem you all have been using deadly weapons to commit murder without a written charter from the baron. So just surrender and go to jail, you will have your chance to sort this out in court. Don't worry I appointed Judge Jones myself and he has been a life long friend, I'm sure we will clear this up in no time.")

Game C is the same choice as A, but the characters keep watch on the criminals that they never trust...so they don't get caught by surprise and are ready for the attack.

Game D is the same choice as B, but the characters again don't trust the baron, and are ready if he tries anything (''well, we found these finical scrolls of yours baron and unless you want us to release them to the public, you will let us go.")

So it is not really about that *one* choice....it is more about a *choice of gameplay*. It is not so much the guessing of what might happen in the games future...but more just being ready for it. It's not ''Player Agency choices matter'' it is more ''it matters how you play the game''.

[Lorsa]

Fortunately, the site has eaten several longer posts, and I've had time to ponder my position. I realize now why I define agency the way I do: because I am (and, this, my definition is) only concerned with the GM removing it.

This is true for me as well, although I might have a broader definition of what counts as the GM removing agency.

It makes everything I've posted on the topic make sense: I don't feel loss of Agency that I impose by having a personality, I don't care about the existence or loss of Agency that is beyond the character's capabilities, etc. The only agency I care about is that the character be allowed to do anything that the character realistically could do.

I agree that I don't feel a loss of agency for having a personality or limited character capabilities. I can, however, feel a loss of agency if the scenario or adventure design is set up in such a way that with the personality or capabilities I do have, there is only one practical choice. In effect, what I want to avoid is scenarios with intersections cluttered with big signs that say "go here!".

Another thing I've thought about is that more capable characters can do more, and, thus, have more agency. So, if Agency was the only good, then in D&D, we should clearly all start at epic level, and things like E6 should be an abomination.

I think kyoryu answered this question very well. It depends what the game is about, but even low level characters can effect the outcome of events.

Regarding the bridge over lava... IMO, the true test of Agency is whether I can craft wax wings, use a hot air balloon (note that both of those should probably result in lava death), jump across, build my own bridge, use water to cool the lava, or otherwise do anything else that is within the character's capabilities.

If I had been the GM my question would be "what do you do to find the raw materials" and "what skills do you have that would make sure you character can build these things". If you want to try and defeat an obstacle in some creative way, you are quite welcome to try. It won't always work, but that's to be expected.

I was in a game where the entire party ate poison. Only one PC made their save. The conscious PC fought off the invisible assassin (using a strategic withdraw to the kitchen for flour, IIRC) while the rest of us slept.

The GM had us all roll 2d6 for the number of hours we slept. IIRC, despite the party being large enough to fit in a Tolken novel, everyone else rolled in the 8-11 range. So, the GM began narrating, "11 hours later, when you all wake up...". I immediately interrupted with, "no, 3 hours later, when my character wakes up, he ties everyone to their horses, and we leave town." Because my character had no intention of putting innocent townsfolk at risk.

Now, the GM had already told us that those 8 hours in town would have been uneventful. The GM has several options at this point. He could deny my action, saying he's already moved on. He could alter reality, saying that my character stays KO'd for 11 hours in order to preserve his reality. He could allow my action, even though it's pointless and gives the party "what if" information that they wouldn't otherwise have, and have nothing bad happen during those 8 hours, either. Or he could allow my action, then have something bad happen while we're on the road.

How do different people's definitions of Player Agency rate those responses?

Personally I think the GM should allow your action. Not doing so is a clear violation of agency under most definitions I think.

Whether or not the GM has something bad happen during those 8 hours (which now take place outside of town) is not really relevant to the question of your agency. Both are fine in my opinion.

Personally, I would ask you if you plan on continuing on the road until everyone has woken up, or if you are going to hide in some forest glen or whatever just outside. For the case of you continuing on the road, I would probably roll a die where low numbers mean nothing happens, high numbers mean something interesting (read problematic) happen and medium numbers is just "something happens".

So for medium numbers you might have a short encounter with a farmer traveling in the opposite direction and being a bit worried if something happened to your group and why you have so many people tied to horses. High numbers would mean a group of guards ask the same questions with the belief that you are, in fact, a kidnapper, perhaps demanding you surrender to them until the others have woken up and your story can be corroborated.

In any case, the GM was far too haste with assuming everyone would just stay put. Negating your wish to leave town is an act of agency violation.

[Segev]

The notion that a war is won on "big turning points" may be dramatic, but it's generally one of the less-believable tropes in fantasy warfare fiction. Wars are won and lost on accumulations of victories and defeats, on thousands of logistical problems and morale triumphs. The PCs can shift the course of a war by being PCs involved in dozens of encounters. They might protect the outpost that would otherwise have fallen for lack of a team of spec ops worthies being present. They might take out the lieutenant who would otherwise have risen through the ranks on the other side as a tactical mastermind. Some of these things will be obvious victories they won that helped shift the tide of war; others may be less obvious.

The PCs don't have to single-handedly win the war to impact it. They just have to have pivotal impact on individual encounters and battles within it. And they will tend to have higher impact than foot soldiers just by virtue of being a spec ops team. A lot of spec ops missions look similar to shadowruns or dungeon crawls, and that's what PCs tend to be good at.

[Lorsa]

Now see this is where I see very little difference or just slight flavor difference. There is an war between the orcs and elves, what side the characters join does not really matter much. The war will go onward no matter what side the characters choose. If the characters are present for a big turning point, they can act and maybe change the course of the whole war or even end it, but it does not matter what side they are on. After all, even if they did pick the elf side...they can still do (or try) *anything*.

First off, if the players can't change the war in any way, then the game isn't really about the war, it is about what happens during the war (or something else). Read kyoryu's post for more description about this.

Secondly the game should look very different depending on which of these sides you do missions for. If you do missions for Orcs, you might end up raiding civilian communities, raping women and gather slaves. If you do missions for the Elves, there might be more skirmishes against their patrols or small outposts and involve missions to rescue slaves.

I mean, just the questions regarding rape will be different. In one case, there might be a question of whether or not the characters can stand being passive or active participants, whereas in the other they might deal with issues as to what to do with half-orc babies or pregnant women wanting to do dangerous abortions, let alone deal with trauma. The choices the players will face during the game will (or should) depend strongly on which side they choose.

Also, if the main plot of the game is the war, then the characters really should be able to determine the outcome of it. Otherwise it's a really crappy railroading game. And if the players can determine the outcome of the war, they can effect the world. I mean seriously, you can't tell me the world looks identical in the two cases of Orcs winning a bit war vs. Elves winning a big war. That should have a clear impact on the lives of the people living in that general area.

And a lot of the time it won't effect the overall plot anyway. You have the noble baron fighting the criminal guild. The baron is good and the criminals bad. Both more then welcome the help of the ''unknown outsider'' characters. And both intend to use and then get rid of the characters. OR it is possible both intend to thank the characters and give them a ton of treasure. OR they both might just ''cut them loose''. Or one might do one thing and the other might do another. It is impossible to just ''guess'' what might happen, even in vague terms.

You make several errors here.

A good baron wouldn't use and then get rid of characters. And even if so, the characters should be able to find some historical information to shed light on the various characters they are planning on "jumping into bed with". Which means it's not a choice in the dark, but one slightly illuminated.

If you set up a scenario where both choices are functionally identical, you are in effect railroading the players and denying them any form of agency. That's on YOU, it doesn't make the idea of player agency useless or players unable to effect the game in meaningful ways in what you consider "normal games". I mean seriously, the type of adventures the players will do for a criminal should be very different from the ones they do for a good baron. So will the choices they will face during those adventures. If they're not different, then you're running a very bad game.

So ''one'' choice like this, won't matter at all for the future of the game: anything can happen. But then it is just *one* choice, and if you make it and forget about it...then things can spin in any direction. Though if you keep the choice more active, and keep and eye on your allies, you might have much more say in what happens.

You're right that the game isn't only about one choice. It is about many choices piling up and making the game (and the world), different.

Like:

Game A: The characters pick the criminal side (''crime is cool") and forget about it and just have fun being bad. It is no surprise to anyone, except the players and characters, that the criminals attempt to kill the characters once they have no more use.

Game B: The characters pick the Good Baron side (''Good is cool'') and forget about it and just have fun being good guys. The good baron is more then happy to keep his hands clean and let the characters do the dirty work, and once it is done...arrest the characters (''well, it does seem you all have been using deadly weapons to commit murder without a written charter from the baron. So just surrender and go to jail, you will have your chance to sort this out in court. Don't worry I appointed Judge Jones myself and he has been a life long friend, I'm sure we will clear this up in no time.")

Game C is the same choice as A, but the characters keep watch on the criminals that they never trust...so they don't get caught by surprise and are ready for the attack.

Game D is the same choice as B, but the characters again don't trust the baron, and are ready if he tries anything (''well, we found these finical scrolls of yours baron and unless you want us to release them to the public, you will let us go.")

So it is not really about that *one* choice....it is more about a *choice of gameplay*. It is not so much the guessing of what might happen in the games future...but more just being ready for it. It's not ''Player Agency choices matter'' it is more ''it matters how you play the game''.

Characters choosing to be vigilant or not is part of player agency. It's a choice, and, AS YOU YOURSELF PUT IT, it will effect the game in quite drastic ways (with them being dead or alive).

Secondly, as I said before, making two choices functionally identical when there is really absolutely no reason whatsoever that they should be, is just poor DMing.

[Darth Ultron]

Secondly the game should look very different depending on which of these sides you do missions for. If you do missions for Orcs, you might end up raiding civilian communities, raping women and gather slaves. If you do missions for the Elves, there might be more skirmishes against their patrols or small outposts and involve missions to rescue slaves.

This is only true if your in a Simple Game. So just to be Clear: A Simple Game is Black and White, like a cartoon or anything made by Disney or made for kidz. In such a game the evil bad orcs will only do evil bad thing as they are evil bad monsters. Always. The good nice elves will only do good nice things as they are good nice paragons. Always. In a Simple Game, the players choice is just about all ways right and things go exactly the way the player thinks they might, with only a very rare and very small chance of anything else happening.

Now, note there is nothing wrong with this sort of game.

In the complex game: anything can happen. The evil bad orcs ''only kill warriors with weapons and never, ever hurt innocents'', the good nice elves will ''slaughter some innocents for a greater good'', and so on to infinity. In a Complex Game the players choice might be right and things might go the way the player thinks about half of the time, but more often then not: anything can happen.

So, this puts a huge restriction on Player Agency: it only is possible in a Simple Game.

Also, if the main plot of the game is the war, then the characters really should be able to determine the outcome of it. Otherwise it's a really crappy railroading game. And if the players can determine the outcome of the war, they can effect the world. I mean seriously, you can't tell me the world looks identical in the two cases of Orcs winning a bit war vs. Elves winning a big war. That should have a clear impact on the lives of the people living in that general area.

This gets back to Game Zero stuff. If before the game, it is agreed by everyone that the game will be a tailor made war plot for the whole games ''main plot''...then it is. However if the war is an event that just happens in the game play, then it won't be a ''main plot''.

And ''looking different'' is exactly my point. If the Orcs or Elves win the war, the land will look different....but how much will it be different? Is the world so different if the bar at the end of the street is the Bloody Axe or the Golden Tree? A lot of things can look and even feel different, but they won't be all that different really.

A good baron wouldn't use and then get rid of characters. And even if so, the characters should be able to find some historical information to shed light on the various characters they are planning on "jumping into bed with". Which means it's not a choice in the dark, but one slightly illuminated.

Of course, as I noted, the Good Baron will only be an Absolute Beacon of Pure Good in a Simple Game. And it is true the characters can get information before making any choice or descision, but this is all on the players to do this.

If you set up a scenario where both choices are functionally identical, you are in effect railroading the players and denying them any form of agency. That's on YOU, it doesn't make the idea of player agency useless or players unable to effect the game in meaningful ways in what you consider "normal games". I mean seriously, the type of adventures the players will do for a criminal should be very different from the ones they do for a good baron. So will the choices they will face during those adventures. If they're not different, then you're running a very bad game.

Though your saying here the DM must do things that only the players will like. The ''same outcome'' from two choices is very common: if you ally yourself with crime group A or group B...both will not too surprisingly betray you.

And, again, your talking only about a Simple Game.

Secondly, as I said before, making two choices functionally identical when there is really absolutely no reason whatsoever that they should be, is just poor DMing.

The thing is: this is just how Reality is...even game Reality.

[Max_Killjoy]

If you set up a scenario where both choices are functionally identical, you are in effect railroading the players and denying them any form of agency. That's on YOU, it doesn't make the idea of player agency useless or players unable to effect the game in meaningful ways in what you consider "normal games". I mean seriously, the type of adventures the players will do for a criminal should be very different from the ones they do for a good baron. So will the choices they will face during those adventures. If they're not different, then you're running a very bad game.

Though your saying here the DM must do things that only the players will like.

????

It's like a master class in non sequitur...

[Quertus]

In the complex game: anything can happen. The evil bad orcs ''only kill warriors with weapons and never, ever hurt innocents'', the good nice elves will ''slaughter some innocents for a greater good'', and so on to infinity. In a Complex Game the players choice might be right and things might go the way the player thinks about half of the time, but more often then not: anything can happen.

That's what you mean by simple vs complex game?! That's what you meant by "Anything can happen"?!

DU, honest question, are you significantly worse at explaining things in writing than you are in person? Because, on the forums, you don't use your words to mean what everyone else means by them, and your explanations rarely actually get to the point of explaining what you mean.

As a GM, you are the eyes and ears of the characters. You are the filter through which the players experience the world. If you are as incomprehensible IRL as you are here, I strongly encourage you to work on your communication skills. Your game will presumably be much more enjoyable if your players understand it.

[Max_Killjoy]

That's what you mean by simple vs complex game?! That's what you meant by "Anything can happen"?!

DU, honest question, are you significantly worse at explaining things in writing than you are in person? Because, on the forums, you don't use your words to mean what everyone else means by them, and your explanations rarely actually get to the point of explaining what you mean.

As a GM, you are the eyes and ears of the characters. You are the filter through which the players experience the world. If you are as incomprehensible IRL as you are here, I strongly encourage you to work on your communication skills. Your game will presumably be much more enjoyable if your players understand it.

He says that, and then turns around to make the claim that railroading results in a complex game, while not-railroading results in a simple game.

Problem for that position is, railroading does result in a game in which only one thing can happen... a simple game.

While a game adjudicated from the context and circumstances and characters at hand can result in a game where many many different things might happen... a complex game (and not just in the sense that he tries to impose on the term).

[Cluedrew]

Funny, I feel we actually answered the main question of "What is Player Agency" on page one. Define agency (ability to make meaning full choices which means using knowledge and having consequences) and then apply it to a player in a game. That is to say player agency is agency a player has. The same can be said of GM agency or player^ agency.

So I think both the player part and the agency part have been covered. Is there anything left besides stress testing the definition by smashing it against Darth Ultron? If there is we should probably cover that before getting to the stress testing.

^ Marks uses of the word player that exclude the GM, who is after all playing the game and is a player in the general sense.

[PhoenixPhyre]

Funny, I feel we actually answered the main question of "What is Player Agency" on page one. Define agency (ability to make meaning full choices which means using knowledge and having consequences) and then apply it to a player in a game. That is to say player agency is agency a player has. The same can be said of GM agency or player^ agency.

So I think both the player part and the agency part have been covered. Is there anything left besides stress testing the definition by smashing it against Darth Ultron? If there is we should probably cover that before getting to the stress testing.

^ Marks uses of the word player that exclude the GM, who is after all playing the game and is a player in the general sense.

I agree that the definition is well covered. I'd be interested in seeing discussion of common pitfalls, especially unconscious ones. What do DMs do to diminish player^ agency without intending to? What can I as a DM recognize and avoid to be better at encouraging agency?

I'm not interested in bashing things against DU. It's just noise to me.

[Cluedrew]

Spoiler: On player^: Because I started going on and on off topic.

Show

Someone else actually copied that? Umm... maybe I should put a bit more thought into that if other people might use it. I mean for one it changes the current common usage of player. I mean it is more accurate, have a short hand for "players minus GM" while players means players, but the convention of players excluding the GM has been around so long I don't think I can uproot it. Although I was trying to call that implicate bias into focus by jarring that standard originally, which may or may not have helped my point, but is a bad reason for a general standard.

You could have two: players+ explicitly includes the GM while players- explicitly excludes the GM. Players on its own would be ambiguous and depend on context as it usually does. Or just add one for players including the GM, players^ or all players if we want to get away from symbols, and accept the slightly inaccurate but history use of players as not including all players.

I like to be sure I am saying what I mean.

On Agency Pitfalls: Sending players in blind is a big thing I've had some issues with before. Even if discovery is an intentional part of the game it is easy to actually leave them with so little knowledge that they can't even ask the right questions, completely cutting out the knowledge part of the equation. Discovery can make for some interesting stories, which is why I think people try for it, but translating it into a game doesn't seem to work as well.

The other is I think of right now is placing too many arrows forward. At a certain point there is an obligation to follow them, even if the characters wouldn't. And even if the characters would to that, because of a single path of least resistance, you get a situation where 99% of people would do that. So you only can interestingly talk about the remaining 1%, which tends to be a Chaotic Neutral Rogue who throws the campaign out the window.

[Quertus]

On Agency Pitfalls: Sending players in blind is a big thing I've had some issues with before. Even if discovery is an intentional part of the game it is easy to actually leave them with so little knowledge that they can't even ask the right questions, completely cutting out the knowledge part of the equation. Discovery can make for some interesting stories, which is why I think people try for it, but translating it into a game doesn't seem to work as well.

The other is I think of right now is placing too many arrows forward. At a certain point there is an obligation to follow them, even if the characters wouldn't. And even if the characters would to that, because of a single path of least resistance, you get a situation where 99% of people would do that. So you only can interestingly talk about the remaining 1%, which tends to be a Chaotic Neutral Rogue who throws the campaign out the window.

I find that Discovery is not a problem so long as it is rooted in an understandable base. Going in blind is fine, so long as the GM builds up trust that there is a reason for everything. Calvin ball discovery isn't worthwhile.

Where I have personally run into problems, however, is with players who don't seem to know how to ask questions / how to investigate beyond rolling something on their character sheet, or those who just aren't interested in discovery. I have neither the skills not the inclination to run a game that would solve the former, although I'm usually fine with the latter.

Similarly, I've had great experiences with "too many arrows forward". What problems do you see with a high option, high agency sandbox?

[OldTrees1]

This is only true if your in a Simple Game. So just to be Clear: A Simple Game is Black and White, like a cartoon or anything made by Disney or made for kidz. In such a game the evil bad orcs will only do evil bad thing as they are evil bad monsters. Always. The good nice elves will only do good nice things as they are good nice paragons. Always. In a Simple Game, the players choice is just about all ways right and things go exactly the way the player thinks they might, with only a very rare and very small chance of anything else happening.

Now, note there is nothing wrong with this sort of game.

In the complex game: anything can happen. The evil bad orcs ''only kill warriors with weapons and never, ever hurt innocents'', the good nice elves will ''slaughter some innocents for a greater good'', and so on to infinity. In a Complex Game the players choice might be right and things might go the way the player thinks about half of the time, but more often then not: anything can happen.

So, this puts a huge restriction on Player Agency: it only is possible in a Simple Game.

Thank you for proving that Player Agency creates complex games.

In a complex game the players may be faced with a meaningful choice between joining the disciplined cruel orcs or the zealous benevolent elves, such a choice would have different outcomes depending on which option the players choose to take.

In a complex game the players may be faced with making choices under imperfect information. In such cases only the informed part of the choice is a meaningful choice, but said part does exist. If you had to cross a mountain range, you might be faced with a choice to go over the mountain through a pass, or under the mountain through a dwarven mine. You might know that the pass would be faster but will soon become impassible. You expect you could make it through the pass if you packed light. So you make your choice based upon if you think your gear is worth more than the time you would save by going through the pass rather than the mines. You knew nothing about the Worgs at the pass or the Balrog in the dwarven mines. Despite there being a blind choice entangled with the meaningful choice, there was a meaningful choice and there was player agency.


Unless "complex game" is just your current keyword for "the DM randomly decides the Players will stick to the script".

[jayem]

Funny, I feel we actually answered the main question of "What is Player Agency" on page one. Define agency (ability to make meaning full choices which means using knowledge and having consequences) and then apply it to a player in a game. That is to say player agency is agency a player has. The same can be said of GM agency or player^ agency.

So I think both the player part and the agency part have been covered. Is there anything left besides stress testing the definition by smashing it against Darth Ultron? If there is we should probably cover that before getting to the stress testing.

^ Marks uses of the word player that exclude the GM, who is after all playing the game and is a player in the general sense.

We established several locations of potential Agency.
Short-term & Mid-term & Long-term
Small-scale & Mid-scale & Large-scale (I'm using this to distinguish between Village and World events)
In-character & As-player & Meta-game (these are what PP called small scale/large scale)

I don't think we've established whether (partial/full) agency in each of them is needed or even (always/ever) good. Perhaps we should?
But it showed that a low agency game required (ironically) not just a choice by the game-controller but multiple ones.

I think (thanks, actually to various responses one of DU's comments) we found that perfect information stops providing agency. And probably ought to discuss the various factors that affect how we actually balance that?

[jayem]

This is only true if your in a Simple Game. So just to be Clear: A Simple Game is Black and White, like a cartoon or anything made by Disney or made for kidz. In such a game the evil bad orcs will only do evil bad thing as they are evil bad monsters. Always. The good nice elves will only do good nice things as they are good nice paragons. Always. In a Simple Game, the players choice is just about all ways right and things go exactly the way the player thinks they might, with only a very rare and very small chance of anything else happening.

A Black and White game is indeed a simple game, there we agree.
The interactions with one interesting person say (angelic/good/bad/evil) (lawful/...chaotic) (int ... ext) (ignorant ... clever) (wise ...foolish) (hasty ... cautious) (posh ... ) (poor ... wealthy) (cowardly ... brave) (strong ... weak)(pragmatic..dogmatic) may be 1000 times more complex than 10 black and white ones.
However a 10 person B/W game is 1000 times more complex than the RGB=888 one you describe being in favour of. That is not a complex game, it is a hyper simple game.

[PhoenixPhyre]

We established several locations of potential Agency.
Short-term & Mid-term & Long-term
Small-scale & Mid-scale & Large-scale (I'm using this to distinguish between Village and World events)
In-character & As-player & Meta-game (these are what PP called small scale/large scale)

I don't think we've established whether (partial/full) agency in each of them is needed or even (always/ever) good. Perhaps we should?
But it showed that a low agency game required (ironically) not just a choice by the game-controller but multiple ones.

I think (thanks, actually to various responses one of DU's comments) we found that perfect information stops providing agency. And probably ought to discuss the various factors that affect how we actually balance that?

The bold part reminds me of related theological comments, but I'll refrain due to forum rules. In game terms, the important agency-preserving information is proximal, true, and character accessible.

Proximality is the idea that the information should concern first-order consequences. "If you take action X, the result will be (conditionally) Y." This requires not giving information about 2nd and further order consequences. Give information about the action they're doing now so that they can make an informed choice. Don't elaborate on the longer-term/larger-level consequences unless the characters would be able to deduce them*.

Truth requires that the information be valid if presented as 3rd-person narration. Questionable facts should be flagged as such in some consistent manner. Remember that the characters are there and would have much more information than can be presented in words.

Character accessibility requires that the information be the sort of thing that the characters could know. Giving too much meta knowledge can impact agency as it makes some choices obvious. It does mean that the DM should err on the side of making things character accessible even if it may not be--the click of a door latch is very soft and can be lost in ambient noise. But telling the players about it gives them important information about the state of the world.

Note--these thoughts are scattered because I'm too busy debugging an application...perils of multi-tasking.

[Darth Ultron]

DU, honest question, are you significantly worse at explaining things in writing than you are in person? Because, on the forums, you don't use your words to mean what everyone else means by them, and your explanations rarely actually get to the point of explaining what you mean.

The thing is that in person I can say more, but I can only type so much. And I don't have the Book of Words that ''everyone'' has.

He says that, and then turns around to make the claim that railroading results in a complex game, while not-railroading results in a simple game.

Well, I was not doing that ''trap''....that is what you do. I'd point out you'd need Railroading in any detailed game with a plot, simple or complex.

Funny, I feel we actually answered the main question of "What is Player Agency" on page one.

I have a point to make and questions...

So Player Agency is when a player feels that their decisions and input have an impact on the events, outcome, and tone of the game; and make meaningful decisions.

I guess some threw ''freedom'' in there, but that is a bit vague to be meaningless. And any time a player might say they are denied ''freedom'', they would also be saying at best the game should not make sense or at worst the game should all ways go their way.

And ''Consequences'' is tossed in there, but that is a bit vague to be meaningless. Other then the very basic ''if your characters rob a bank the cops will come after them'', the players won't really know the ''Consequences'' or not...unless the DM tells them.

So everyone seems to be talking like there is Path A where the game world moves along down that path. Then, as soon as a character takes any big action the path suddenly becomes the different path B. This makes the player happy as they have done something meaningful and made a new path.

But how does this work in an RPG again? I see three ways for a game like D&D with unequal players and a DM:

1.The DM has a plan, story, plot or other framework that is strong and stable so a player can't overly change or alter anything most of the time with a single action of a character. Things in the game reality are set and not subject to quick changes on whims. To really change anything would take some effort by the player, time, and several actions by the character....and ultimately whatever happens is decided by the DM.

2.The DM has some vague idea of a plan, or some notes or something of substance that is a least a bit firm so a player can't overly change or alter anything most of the time with a single action of a character. To really change anything would take some effort by the player, time, and several actions by the character.....but only from the handful of vaguely set things. But as most things in the game reality are not set, there is nothing for for the player to effect or change. And ultimately whatever happens is decided by the DM.

3.The DM has nothing. The game is pure improv. As everything in the game reality is not set, there is nothing for for the player to effect or change. And ultimately whatever happens is decided by the DM.

Set Up-The characters are in a city with a dragon moving around it attacking folks. The players want to get out of the city and avoid the dragon.

So first we have the Lack of Information Trap: If the players fool themselves into thinking the dragon is north of the city they might try and make a run out the south road OR just try and randomly/clueless make a run for it.....then anything can and might happen. But this lack of information is all on the players. Then we get:

1.The Dragon Siege of Wayhaven adventure. So the DM has a whole detailed adventure written here. So there is a whole story and plot as to why the dragon is even there. The dragon is looking for it's stolen egg, and has spies and agents inside the city working for it. The DM has details on everything. The adventure plot here is, of course, is to ''find the dragon egg, give it to the dragon and save the city'', so the players by just ''having their characters run away like sad cowards'' is not following that plot...but that does not matter for this example. To try and make a good escape from the dragon the players would need to take the effort, time and many actions of the players to do so, based on all the details of the adventure.....and ultimately whatever happens is decided by the DM.

2.The Dragon Encounter. The DM here only has vague ideas and some handful of notes. The DM has a vague idea that ''it is a cool idea to have the dragon looking for something'', but has no idea what it is, and the DM has a couple of dragon thug guys to encounter, but every little else about them. The DM has the vague idea that the characters might find or use something in the city to stop the dragon, but has not committed to any details. To try and make a good escape from the dragon the players would need to take the effort, time and many actions of the players to do so, based on nothing except the vague ideas the DM has.....and ultimately whatever happens is decided by the DM.

3.The Game. The DM randomly picks a dragon to simply be there. DM thinks maybe ''there is a reason'' and some dragon cult like people might be fun, but again has not details on anything. The non-plot here is more ''the characters react to the dragon'', and the DM just can't wait to react to whatever the players choose to do...but the DM really has no idea what that reaction might be. To try and make a good escape from the dragon the players would need to take the effort, time and many actions of the players to do so, based on absolutely nothing.....and ultimately whatever happens is decided by the DM.

As the page show: everyone hates #1, thinks #2 is a good ''middle ground'' and thinks #3 is the Best Game Ever.

But it is odd as:

Well #1 allows the players to use the game information and detail to potentially change and alter the things in the game in meaningful ways based on using that information and detail........and ultimately whatever happens is decided by the DM.

And #2 allows the players to use the bits and pieces of the vague information and ideas to potentially change and alter the few things in the game in meaningful ways based on using that vague information and detail........and ultimately whatever happens is decided by the DM.

And #3 is the players can't change and alter the things in the game in meaningful ways as there is nothing of substance to change or alter....and ultimately whatever happens is decided by the DM.

So why is #3 so great if it, in fact, has no Player Agency at all? And if the answer is that the Improv DM will just ''fill in the blanks'' with whatever the the Players want and the way the Players want, I'd ask why that is a good thing and why can not the other two DMs do that; assuming any of the DMs wanted a lame Player controlled game?

[OldTrees1]

And #3 is the players can't change and alter the things in the game in meaningful ways as there is nothing of substance to change or alter....and ultimately whatever happens is decided by the DM.

So why is #3 so great if it, in fact, has no Player Agency at all? And if the answer is that the Improv DM will just ''fill in the blanks'' with whatever the the Players want and the way the Players want, I'd ask why that is a good thing and why can not the other two DMs do that; assuming any of the DMs wanted a lame Player controlled game?

If you can't comprehend an improv game, of course you will fail to comprehend why it has player agency. But that speaks only about your failing and not about improv games or player agency themselves.

[Lumipon]

assuming any of the DMs wanted a lame Player controlled game?

Why would a player-controlled game be any more lame than a GM controlled one? This seems like a loaded statement that requires some unpacking.

[Quertus]

I think (thanks, actually to various responses one of DU's comments) we found that perfect information stops providing agency. And probably ought to discuss the various factors that affect how we actually balance that?

Apologies, I missed this. How is a player not able to impact a world about which he has perfect information?

The thing is that in person I can say more, but I can only type so much. And I don't have the Book of Words that ''everyone'' has.

Fair enough.

Two points: one, you seem quite resistant to learning our definitions - why is that; two, it's nice to see that you can tell when "everyone" is being used incorrectly.

I have a point to make and questions...

So Player Agency is when a player feels that their decisions and input have an impact on the events, outcome, and tone of the game; and make meaningful decisions.

Drop the word "feels". Yes, that's arguably the part I care about, but I'd still be wrong to define Player Agency based off feelings.

I guess some threw ''freedom'' in there, but that is a bit vague to be meaningless. And any time a player might say they are denied ''freedom'', they would also be saying at best the game should not make sense or at worst the game should all ways go their way.

Personally, I argue for the game making sense in order to have agency - actions to have logical consequences, for wood to float and wounds to hurt and barns to burn.

And ''Consequences'' is tossed in there, but that is a bit vague to be meaningless. Other then the very basic ''if your characters rob a bank the cops will come after them'', the players won't really know the ''Consequences'' or not...unless the DM tells them.

Wood floats, wounds hurt, barns burn. Unless you're going completely off the reservation, most actions have logical, predictable consequences.

So everyone seems to be talking like there is Path A where the game world moves along down that path. Then, as soon as a character takes any big action the path suddenly becomes the different path B. This makes the player happy as they have done something meaningful and made a new path.

Kind of, yeah. "This is how things would have played out if not for the PCs. This is how things played out with the PCs. Are they different? If so, then the PCs mattered."

.But how does this work in an RPG again? I see three ways for a game like D&D with unequal players and a DM:

1.The DM has a plan, story, plot or other framework that is strong and stable so a player can't overly change or alter anything most of the time with a single action of a character. Things in the game reality are set and not subject to quick changes on whims. To really change anything would take some effort by the player, time, and several actions by the character....and ultimately whatever happens is decided by the DM.

2.The DM has some vague idea of a plan, or some notes or something of substance that is a least a bit firm so a player can't overly change or alter anything most of the time with a single action of a character. To really change anything would take some effort by the player, time, and several actions by the character.....but only from the handful of vaguely set things. But as most things in the game reality are not set, there is nothing for for the player to effect or change. And ultimately whatever happens is decided by the DM.

3.The DM has nothing. The game is pure improv. As everything in the game reality is not set, there is nothing for for the player to effect or change. And ultimately whatever happens is decided by the DM.

If that is what you see, then you see less than nothing. That is, what you see has less value than if you were completely blind.

Now, I'm half blinded by my biases, but here's what I see (we'll even use your words) (I expect others will tell me how wrong I am):

The GM has a "stable" plot, where the PCs have no agency to change anything by virtue of being powerless pawns compared to the events unfolding.

The GM has an unstable plot that he'd really like to be stable, so he railroads away the player's actions whenever they go off the rails, denying them true agency.

The GM has a good, unstable plot that the PCs actually have agency to affect, by virtue of having sufficient knowledge, capabilities, etc, and the GM bloody lets their abilities have their logical effect, and lets them affect the story. This is what most of us (or, at least, what I) call a good / healthy / normal game.

[jayem]

Apologies, I missed this. How is a player not able to impact a world about which he has perfect information?

Basically all the choices become one binary choice (Do I impact the world?).

Ignoring chance:
While one choice leaves a small "% of outcomes " matching the desired requirements and the other a "large %". Then it doesn't matter which you choose, because you can always choose the % afterwards. When you get to the point where one choice has no chance and one chance has a small "%", then it's a no brainer (you could have decided to 'lose').

With chance (either by die, imperfect information, or other [unpredictable] actors actions), knowledge of the consequences are not complete. And to the extent that it is, there will be a calculatable ideal strategy that maximizes the odds of success for simple objectives (but now for complex objectives you have to hold them in balance).

(And yes PP, that discussion would be interesting, and the limits you suggest seem like a good start.)

The example of what "Perfect Information" would be is if DU's strawman were made flesh (in a non-fate style game).
"The DM tells the players everything about the game...every single detail. Then the players have their characters act on all that information, but they sort of pretend the characters ''don't know''. This would be like Player Bob knows the pit trap is at Square 1-A, then he role plays character Zorg ''pretending'' not to know that, but ultimately Player Bob makes the informed meaningful decision and choice of if or not character Zorg falls into the pit trap or not."
Max and Pleh, made comments about what would be an appropriate level of information, and demonstrated that the alternative need not be no-info.

[NichG]

Basically all the choices become one binary choice (Do I impact the world?).

This would only be the case if there's also only a single goal that you could have.

If there are multiple possible goals, you still have the choice between those goals - e.g. how would you like the world to look?

For example, mucking around in a paint program is deterministic and perfect information, but there's a lot more to it than 'make a picture' or 'don't make a picture'.

[OldTrees1]

Basically all the choices become one binary choice (Do I impact the world?).

Ignoring chance:
While one choice leaves a small "% of outcomes " matching the desired requirements and the other a "large %". Then it doesn't matter which you choose, because you can always choose the % afterwards. When you get to the point where one choice has no chance and one chance has a small "%", then it's a no brainer (you could have decided to 'lose').

With chance (either by die, imperfect information, or other [unpredictable] actors actions), knowledge of the consequences are not complete. And to the extent that it is, there will be a calculatable ideal strategy that maximizes the odds of success for simple objectives (but now for complex objectives you have to hold them in balance).

(And yes PP, that discussion would be interesting, and the limits you suggest seem like a good start.)

The example of what "Perfect Information" would be is if DU's strawman were made flesh (in a non-fate style game).
"The DM tells the players everything about the game...every single detail. Then the players have their characters act on all that information, but they sort of pretend the characters ''don't know''. This would be like Player Bob knows the pit trap is at Square 1-A, then he role plays character Zorg ''pretending'' not to know that, but ultimately Player Bob makes the informed meaningful decision and choice of if or not character Zorg falls into the pit trap or not."
Max and Pleh, made comments about what would be an appropriate level of information, and demonstrated that the alternative need not be no-info.

Each move in a game of chess is made with perfect information(see the definition of perfect games on wikipedia) but still has plenty of incomparable options per choice such that the agency experienced is not a matter of "do I affect the game or not".

Although not all of the moves in chess are examples of player agency. Forced moves do exist as a non choice (or a binary choice to keep playing or not), but those are outliers.

[jayem]

Each move in a game of chess is made with perfect information(see the definition of perfect [information] games on wikipedia) but still has plenty of incomparable options per choice such that the agency experienced is not a matter of "do I affect the game or not".

Although not all of the moves in chess are examples of player agency. Forced moves do exist as a non choice (or a binary choice to keep playing or not), but those are outliers.

I should have considered the technical definitions more. Though for most Perfect-information games (e.g. Tic-Tac-Toc/Connect 4), the outcome I described occurs.
Chess and Go avoid it because the sheer number of states (I was going to say options) means we haven't yet analyzed it well enough.
Perhaps I should have called it hyper-perfect information.

Regarding the multiple goals, if they are compatible they can be wrapped up in one (it's why I had desired requirements). If they are not, then you do have to choose, once. But the paint example shows it's not properly thought through* (It was initially bought up as one of the things to get sorted out).

*I guess that would have some interesting relation to improv games

[Quertus]

I should have considered the technical definitions more. Though for most Perfect-information games (e.g. Tic-Tac-Toc/Connect 4), the outcome I described occurs.
Chess and Go avoid it because the sheer number of states (I was going to say options) means we haven't yet analyzed it well enough.
Perhaps I should have called it hyper-perfect information.

Regarding the multiple goals, if they are compatible they can be wrapped up in one (it's why I had desired requirements). If they are not, then you do have to choose, once. But the paint example shows it's not properly thought through* (It was initially bought up as one of the things to get sorted out).

*I guess that would have some interesting relation to improv games

With enough time travel / wishes / immortality / alternate realities, all goals are compatible.

Most people find hyper-perfect information "games" to no longer be fun. I once had a co-worker comment that he couldn't understand why anyone would ever play Sudoku - once you memorize a few techniques, it's just application of those techniques, and no fun. I replied that that was exactly why I had never learned those techniques.

Truly advanced player skills (the likes of which are unlikely in humans for some games) can make all deterministic perfect-information games boring. Sure. But what does boredom have to do with Player Agency?

[OldTrees1]

I should have considered the technical definitions more. Though for most Perfect-information games (e.g. Tic-Tac-Toc/Connect 4), the outcome I described occurs.
Chess and Go avoid it because the sheer number of states (I was going to say options) means we haven't yet analyzed it well enough.
Perhaps I should have called it hyper-perfect information.

Regarding the multiple goals, if they are compatible they can be wrapped up in one (it's why I had desired requirements). If they are not, then you do have to choose, once. But the paint example shows it's not properly thought through* (It was initially bought up as one of the things to get sorted out).

*I guess that would have some interesting relation to improv games

1)
It is true that perfect information games are ideal exercises for game theory to attempt to solve. The solution to the prisoner's dilemma tournament (20 people each play 1 game of the prisoner's dilemma with each other person and total their outcomes) is a fine example.

However I would argue that RPGs (when they happen to be perfect information games), just like Chess and Go, avoid being a solved game due to the vast number of states.

2)
Regarding multiple goals. Having multiple goals being perfectly compatible is a rare thing. More often they will be partially compatible with a variety of different outcomes that result in different ratios of achievement. But as you said (before examining the paint example) that changes it from a binary choice to a single choice with multiple options. I would provide another counterargument to that perspective.

Imagine the choices that lead to those endings as a branching tree. You might play that tree by choosing a leaf and then the path to that leaf. I might play by, at each choice, discarding the branch that leads to the leaves I desire least of my remaining leaves. If we have the same goals we would end up at the same ending, but I enjoyed my agency during the entire game while you precommitted all those future meaningful choices with your first choice.

3)
I have done the math and solved Tic Tac Toe like so many people before me. While it always ends in a tied game(unless someone makes a mistake), I do have a prefered ending position. However, despite it being a perfect information game, I cannot choose and end state before the fourth move is made (159 followed by 2 or 6 ends the game in the same state). That means each game of Tic Tac Toe has at least 4 choices (excluding games lost by mistakes).

[jayem]

That seems like I'm not being consistent as to whether the DM's reactions/intentions are part of the information (among other things?). I definitely need to put more thought into it.

[OldTrees1]

That seems like I'm not being consistent as to whether the DM's reactions/intentions are part of the information (among other things?). I definitely need to put more thought into it.

I was thinking about things like when your choice gives an NPC a choice which then gives you another choice. I may be able to predict the NPC's choice if their payouts result in a easy choice (similar to how I can predict some, but not all, of your responses in tic tac toe). However there will be times when I don't know which option they will choose despite knowing their payouts.

Furthermore, we could be talking about a more complicated perfect game like chess where I know the NPC opponent's payouts for winning/losing both on the grand scale(losing a game) and the micro scale(losing a piece) but not in between(how much do they value this line of attack over that line of attack?).

It is becoming clear that you have to add a huge number of qualifiers ("hyper perfect") to approach a system that will always result in reducing all choices down to binary choices.

[Frozen_Feet]

I'm pretty sure that any game where possible choices can be enumerated in a closed list can be reduced into a a string of binary choices, and that this criteria is necessarily filled by every game that can or has been computerized using binary logic.

This is less usefull than it sounds, because the string of choices, if implemented this way, can be so long that humans playing it cannot reach an end state before dying of old age. It's also unnecessary because humans don't need to think using strict binary logic.

---

@Lorsa: I've been too busy to respond to you. I'll see if I can find time during the next week.