Player Character Intent

[Alex]

Hey there people! I am splitting this thread from this one, as I began to post something a little derailing there. I was trying to reply to Excommunicator's comment:

Okay, I understand where you are both coming from in terms of the roleplaying limitations idea, and I would mostly agree, except if there is an in-world explanation for why those powers are being taken away/restricted/limited/changed then I would accept that as part of the game, even if I preferred the alternative or thought it was a bad design decision.

But that is where it ends. The moment the designer starts working on a system that expects me to live according to rules or restrictions that they have not written into the setting and given a reasonable explanation for why the game is behaving that way then I don't accept it (e.g. acceptable would be omniscient god denying the powers due to the player having done something against god's wishes/principles or a government agent who actively sought out certain people to murder them when they were told not to do so, and having their gun taken away for a period of time). Of course, these things are reactions to undeniable facts in the game world and not intentions or motives being considered, and in the second example, if the game later prevented the character from picking up or using a gun found elsewhere (instead of providing consequences for doing so) then I would have a very big problem problem with that.

I also agree on all the disagreement with mechanics based generally around intent and motive. Such considerations never work, and should never be put in a game. Although, at times it is possible to eliminate all other possibilities in an isolated situation so that the intent behind a player action can be essentially known, e.g. if the king stays the whole game only in the the castle full of guards, even with a scripted assassination attempt later in the game, the game can know for certain if the player is behaving badly or with bad intent due to all other explanations being impossible (enemies simply can't get past the guards, there is no reasonable explanation for a sword being swung in the presence of the king, and the king would not be initiating any violence of his own). Those sorts of situations can have the possibilities ruled out, and when done properly it can quite easily approximate the motives of the player and react to them effectively.
From that basis we can then put in all sorts of consequences like charges for attempted assassination or expulsion from the castle or a death warrant, but only in-world explanations. Secretly take away 100xp every time such a situation is detected by the game and I would be very annoyed.

Bolding mine.

I am not sure I agree with what Excommunicator said, in particular, I think the bolded phrase is wrong. Let me rephrase it: "Game mechanics dependent on intent and motive never work, and should never be put in a game". First, I want to say I agree with this, as long as we are talking about player intent. I mean, designers obviously will assume some things about the player intent, such as assuming he intends to play an RPG of the kind he created. But that is where it ends. In fact, trying to psychoanalyze the player in the game sounds so hopeless it never went through my head when responding to Excommunicator. So, I won't be arguing about this, and ask that anyone wanting to argue about that do so in another thread.

So, what exactly am I arguing about? I am arguing that these might be true: "It is possible to let the player express to the game what he wants to be his character's thoughts are and make the game more immersive at the same time." and "By using what the player defined about his character feelings and objectives, the game can focus on that, giving the player an ability to customize the story that would be impossible (or at least very awkward) otherwise".

So, let's start not by these statements, but what I mean by character and player objectives. When a someone is playing an RPG, he might be in various different frames of mind. For example, he might be trying to exploit the game, making his choices according a what he believes will yield best benefits, he might be playing leisurely, without much pre-planning, taking choices that seem interesting. Also, he might be playing himself, taking the choices he sees himself doing in that situation, or maybe he is playing according to a pre-defined persona he came up. These mind frames will determine player intention (which we won't concern ourselves much in this post).

On the other hand, The PC's thoughts are something else... maybe. Usually, the game doesn't keep track of what the PCs thoughts and feelings might be, letting the player's imagination keep track of this, if he wants. Possibly, there is no such thing as the PC's thoughts, as the player doesn't bother thinking up what might be happening in his PC's head. Probably, the player has a mental picture of his PC, but a rather hazy one. This post is all about defining, inside the game, rather than the player's thoughts, what is going inside the PC's psyche and allowing that to have some kind of effect in the game. One important point here is that we shouldn't take away the player's ability to decide what his PC thinks, though we could add some restrictions or prohibit illogical behavior.

Now, on to the first phrase. I believe that "It is possible to let the player express to the game what he wants to be his character's thoughts are and make the game more immersive at the same time". The most obvious implementation of this would be simply let the player pick his PC's intentions and plans when he makes a choice. Examples of games that already did this were Planescape: Torment (usually by allowing the PC to either lie or not in dialogue) and Alter Ego (where the player can choose his character's mood for each action he takes). While this worked well enough in these games, I don't know if it is desirable or realistic to use this kind of input however. Planescape worked because it was solely concerned with alignment changes and Alter Ego worked because of the granularity of its actions.

Instead, I think it would be more worthwhile to use other, more indirect ways of determining the PC's mind. Basically, I am talking of putting the PC on the spotlight and letting the player choose his thoughts. I think that doing this would not only be more entertaining and immersive, but also would allow the game to focus on what it wants to know about the PC. Funnily enough, Planescape: Torment already did this too. For example:

Spoiler

When Fall From Grace (I think) interrupts TNO to ask him what are his thoughts about his own situation.

or when

Spoiler

The Nameless One must choose a regret to enter the fortress of regrets.

I thought that Planescape was a better game for those moments, that letting me choose what was in TNO's head and declare it in some way made the narrative better. My thoughts are that it would be worthwhile to provide similar situations in an RPG (whether it is a dream sequence, a Hamlet style soliloquy, an argument with the PC's own conscience, etc) in order to get the PC's stance toward important points in the story. Heck, I think this would be worth it even if the information simply wasn't used (like is the case in PS:T).

So, now all is left to argue s how this information could be used, or "By using what the player defined about his character feelings and objectives, the game can focus on that, giving the player an ability to customize the story that would be impossible (or at least very awkward) otherwise". There are many P&P games that make use of this kind of information. The way to get and use this information varies from game to game, but one almost universal aspect it that the GM should take this into consideration when creating his scenarios. If a player puts in his character sheet that his character is bitter, and he is acting bitter towards someone specific, it is a clear indication that the player wants to explore that side of his PC with the character he is acting bitter towards to. In a P&P game, the GM is then responsible for creating situations that allow such explorations.

In a CRPG, we are much more limited. Even if we have some system to generate a non-linear story, the specific interesting interactions need to be though of beforehand. However, while we can't allow the player as much freedom as in a P&P game, we can still allow some! If a game is simply designed to take into consideration a few possibilities about the PC's mind at each point, for example changing its plot a little to better fit the PC, I think we can get something very good already!

By the way, if you have any doubts, feel free to ask. If you people are interested, I can write an extended example of how I see this being used, though not this weekend. I wrote this because I think I wasn't able to express myself clearly in the other thread, so this is here. I really hope this is more readable.

 

[galsiah]

It's an interesting notion, which could work quite well I guess. However, I don't think all your examples work - at least the first PS:T one doesn't. Whenever some NPC asks the PC a question, there are thousands of reasons the PC might want to lie. Even where the designer can't think of a clear, pragmatic reason for the PC to lie in a given situation, he can't assume that the PC is telling the truth. Some characters can perfectly reasonably lie just because they feel like it. [[EDIT: although I suppose you could apply Truth/Lie modifiers to statements - but then a load of lies are unlikely to tell you much about PC intent...]]

Whatever mechanism you used to determine PC thinking/intent/motive/outlook, it'd have to be lie-resistant. Even if you're going the "dream-sequence / soliloquy / argument-with-PC's-conscience" route, I think you'd need to have it be a clearly defined system within which the player understood that his statements were assumed to be true (i.e. that they were player statements about the PC, not PC statements about the PC).
To take the "argument-with-PC's-conscience" case, it's relatively common for people to lie-to/mislead themselves, or to rationalize emotional decisions they're aware might not really make sense. If such a mechanism were used in game to establish PC intent/thought/..., you'd need to be sure to present it so that it didn't come over as a rationalization mini-game.

I guess it'd work best if separated from the core gameplay - a kind of ancillary character-development system. It needs to be clear that it's the player who's being asked the question on behalf of his PC (so it doesn't make sense to lie). As soon as you're asking the PC the question, lying is always going to be a factor - as it should be.

 

[Alex]

Galsiah! I thought you didn't post in the Codex anymore... Anyway, it is really good to see you around again.

It's an interesting notion, which could work quite well I guess. However, I don't think all your examples work - at least the first PS:T one doesn't.

I was hoping to sneak that one by... :D You are, of course, completely right about that example not being strictly about the player exposing his PCs motives. In fact, I think that dialog might alter TNO's influence on FFG. However, it can be used by the player to show his PC's thoughts and I think it complements the game's narrative well when used so.

Whenever some NPC asks the PC a question, there are thousands of reasons the PC might want to lie. Even where the designer can't think of a clear, pragmatic reason for the PC to lie in a given situation, he can't assume that the PC is telling the truth. Some characters can perfectly reasonably lie just because they feel like it. [[EDIT: although I suppose you could apply Truth/Lie modifiers to statements - but then a load of lies are unlikely to tell you much about PC intent...]]

I agree completely. The player must know he needs to choose truly. Maybe preface all answers with [truth] without the matching [lie], or maybe something different (different fonts, different color, etc).

Whatever mechanism you used to determine PC thinking/intent/motive/outlook, it'd have to be lie-resistant. Even if you're going the "dream-sequence / soliloquy / argument-with-PC's-conscience" route, I think you'd need to have it be a clearly defined system within which the player understood that his statements were assumed to be true (i.e. that they were player statements about the PC, not PC statements about the PC).
To take the "argument-with-PC's-conscience" case, it's relatively common for people to lie-to/mislead themselves, or to rationalize emotional decisions they're aware might not really make sense. If such a mechanism were used in game to establish PC intent/thought/..., you'd need to be sure to present it so that it didn't come over as a rationalization mini-game.

I guess it'd work best if separated from the core gameplay - a kind of ancillary character-development system. It needs to be clear that it's the player who's being asked the question on behalf of his PC (so it doesn't make sense to lie). As soon as you're asking the PC the question, lying is always going to be a factor - as it should be.

I know where you are coming from but, if at all possible, I would prefer to never address the player directly. I think the narrative will be strengthened if the answer to these questions is given as narrative itself, like foreshadowing or tension creation or some such technique. We might have these separated, but the narration would be weakened if the player was allowed to "lie" when foreshadowing his own character. We might have such scenes have their outcome pre-determined by what the player choose to answer about his character's psyche, but then the player would feel as if losing control.

 

[Excommunicator]

Alex

I am not sure I agree with what Excommunicator said, in particular, I think the bolded phrase is wrong. Let me rephrase it: "Game mechanics dependent on intent and motive never work, and should never be put in a game". First, I want to say I agree with this, as long as we are talking about player intent. I mean, designers obviously will assume some things about the player intent, such as assuming he intends to play an RPG of the kind he created. But that is where it ends. In fact, trying to psychoanalyze the player in the game sounds so hopeless it never went through my head when responding to Excommunicator. So, I won't be arguing about this, and ask that anyone wanting to argue about that do so in another thread.

Firstly some quick comments on this part. The statement you pulled out was talking about player intent specifically as you guessed, and note the "mechanics dependent on intent and motive" by which I mean mechanics that actually try to measure and react to player's thinking patterns, and not simply generalisations by the designer in building the game, which are perfectly fair when trying to craft a game for a particular audience/section of the society. Either way, I think we are in agreement here that mechanical design around player intent is ultimately an unfulfilling pathway.


As for character "intent" that is a much more debatable and sticky issue, and not one which can be pulled apart easily.

Firstly, given that roleplaying games are (idealistically speaking) inherently a decision making system and subsequently a simulation, intended to react "realistically" to the decisions of the player (via actions by the character, specifically), it means that when you try to look at intent from any direction, you are going to find yourself in a dilemma of multiple intents and motives.

The dilemma essentially is between two things where on the one hand, the ultimate goal of the game is to allow the player to make decisions within the game context, but on the other hand, as a role-playing game the player is necessarily "taking the helm" of another person, physically and mentally, where all interactions need to occur between the world and the character as a kind of interface to the player. The problem lies in the fact that by assuming the mental identity of another individual, you need to take on the mental traits of that individual (including morals, personality, and past experiences leading to the formation of those traits) in order for that simulation of decision making to properly make sense, and at some point you are going to find that there is a contradiction of player identity and character identity because the player can only make the decision through a filter of his own psychology, regardless of how hard he tries to understand the motives of his character and play them out as best he can.

Imagine the player as a normal law-abiding 30-yr old guy with an IT job living in our favourite US of A, assuming the decision making prerogative of an opportunistic thief who has no troubles committing non-violent crimes that lead to an increase in his own personal wealth. We already know that the 30yr old IT worker is going to go to some lengths to rationalise his decisions through the mind of the character he is playing because he is having fun, and normally this would be implicit in whatever the actions he takes.
Now, after a particular bountiful stealing spree, what happens when you try to consider the character's intent and motive on stealing a purse from a sickly old man against the intent and motive of the player in telling the character to do it? The IT worker is a law abiding man himself, and it is his simplistic understanding that the thief will steal the money since it is a non-violent crime, therefore fitting the moral criteria he has set out for his character. On the other hand, the thief actually lives in that world, and he knows what life is like there. He knows stealing the money from the sickly man will almost certainly lead to his suffering and death, because there are no such things as welfare or human rights in these fantasy worlds, and when your money is no longer there, you are basically going to starve and die.

What does the game do?

If the game considers actions only, then we see that the thief character has pick pocketed the old man, and the game can react to this by having the old man found dead and diseased out in the street a few days later. It doesn't matter the intent of the character's in the result, because the character and/or the player actualised their intent in deciding to pick pocket the old man (whatever their motive from a sea of countless motives), and the results of that action have been realised now, in the dead man, the investigations into his death and the mourning family. As far as the game is concerned, this process is finished, and the player will be left trying to piece together the consequences in his mind with whatever his reasoning was at the time for doing it.
If the character has not pick pocketed the old man, you can still implement whatever story/quest related functions he has, and by the action of not pick pocketing him/killing him, you open your actions up to the rewards of those quests, regardless of why you did it.

[Let me just jump in here for the LARP-hunters: This isn't a case of simply having a generic NPC that has money on his person who dies if it is taken away. If that was the case, and you invented some background for the character of your own to support your motives, then you would be LARPing. In this particular situation, we have an old man who you can speak to to understand his personality and social situations, that you can talk with other people about, that has other characters who are family, law enforcement who will investigate for him, who has an explanation for where he got his money, why he is sick, he has quests to offer, other means to obtain the money via, and most importantly, consequences according to what happens to him. He is a fleshed out element of the world]

With that out of the way, now if you try to consider intent and motive, how exactly can you come to a reasonably accurate let alone meaningful conclusion about what intent is coming into play here? Did the player make the decision to steal from the old man because he knows he was just playing a game and the man wasn't actually real? Did he try to rationalise the behaviour through the mind of the non-violent thief by deciding that stealing it would be OK because it wouldn't be a violent activity? Did the character have proper agency here, and despite knowing what would happen to the old man, decide to take the chance that his family might support him? Did the character simply not care about this particular old man because of being witness to the old man's past indiscretions?

How might the game properly react to the intent of the player/character?

Let us even assume there was even a kind of motive confirmation process like you mentioned before, where the player is talking to another fellow thief about a sickly old man he has seen who happens to carry his life savings with him wherever. Perhaps you then allow the player to input a response to the character's thief friend along the lines of:

1. "Let me do this one friend, as I think I can manage this without harming the old man"
2. "I imagine without that money the poor sod will keel over and die before long, so methinks a more worthy target would be better suited"
3. "I hear he was embroiled in laundering back when the town was full of corruption, so taking the money from him is simply returning the favour"
4. "Well, no doubt his family would support him in the event that the money happened to go missing..."

Now you have two main approaches here, in terms of the game considering the motives behind what is being said, and not just the words spoken (reacting to words spoken is what I would always advocate in such a situation):
1. You either go by designer crafted interpretations, and make the game process the situation according to whatever designer-chosen consequence applies to the player's chosen dialogue response, meaning that the game will assume the tone, intent, motive, morals and way of speaking on behalf of the character, with all other potential motives and approaches being completely insignificant and impossible for the player to put forth (a staple of Bioware RPGs, and many, many others), with the consequence being that the game explores and reacts in a more detailed way to the particular motives available for choosing (remember, it has to be done from a list), thereby alienating all non-explicit motives from the scenario, or,

2. You try to psychoanalyse and understand the potential motives through each option, and react to those motives accordingly, with various secondary confirmation scenarios and approximations, and estimations on the moral and psychological nature of both player and character:

1. Does he feel compassion enough to be lying so that the other thief doesn't take the money? Has he got the intent to actually steal from him in a purely non-violent way? Is he just putting on a nice-guy facade so he has an alibi after he brutally tortures and kills the man?
2. Does he feel compassion enough to leave the man be? Does he feel that the target is not going to put up an entertaining enough bit of resistance to be worth the time? Is he lying about his opinion of the old man? Is he trying to lie to the thief friend because he doesn't like him?...
3. Is he really justifying the act because he truly believes this? Is he merely making this up to seem principled to his friend?....
4. Is he saying this intending to steal? not intending to steal? just making conversation? hinting so that the other guy might perform the act so that he can miraculously save the old man at the right time?

As you can see, even with a reasonable selection of options like this, the intent of character, player, whatever is nowhere to be found in a mechanical process like this, it is only a selection of things which could potentially represent the motive. Perhaps most annoyingly of all is the fact that once you start considering the motive and intent, you can then only allow the player/character to have intent along the lines of the ones you have implemented in the scenario. Do you really want to go down the path of trying to incorporate complex situations that exclude most forms or motive/intent (let's face it, the number of potential motives behind the above dialogue options, would be infinity) just so that you can then include various consequences that flesh out the narrative of those motives?

Basically, these design approaches are antithetical to RPGs because they hinder the decision making process for the player, even though they might incorporate narrative exposition for certain motives. They are story-telling devices because they are a case of choosing which pre-made adventure you want, not making the decisions to form your own adventure. Additionally, once the game has interpreted and made a decision as to the characters/players motives, is it then the game's responsibility to keep track of contradictions to these motives/intents/morals/behaviours? Does the game change the label of intent and motive on the character's actions every time they do something which the game sees as a new motive? If so, what meaning can the motive have if the game continually changes it, or more annoyingly, then tells the player what motive they must have in future situations, even though the game cant possibly understand the motive in the first place?


tl;dr - No. Get out of the workshop. You aren't welcome here.

 

[JarlFrank]

I do approve of the [Lie] and [Truth] options in PST as far as player intent goes. But beyond that... why should it have any consequence, except for the player's lawful/chaotic alignment in PST since it uses the D&D rules, for the game what the player's intent was? Until you reveal your intent by either

a) doing what you said, therefore proving that you said the truth.

b) not doing what you said by, say, selling the quest item you promised to bring back, therefore proving that you lied.

there is no logical consequence that you lying or saying the truth could have on the world around you (if we assume that NPCs do not have the ability to sense your motive, but this is something I'd only implement for lies about things that already have happened - as in, "I've not done this" instead of "Yes I will do this").

Any consequences it has on the player's behaviour are down to the player's own discretion and should not be enforced by rules. What if he lies, but then decides to do what he promised anyway, after getting the quest item? Nothing should prevent him from doing that. In a D&D system like PST, this could make your character even more chaotic because he broke a lie, but beyond that I don't see how it could have any meaningful consequences on either the player character or the world and people around him.

Yes, the game could include consequences for liars. If you lie often and then also keep the lie (as in, do not do what you promised) people will trust you less if they hear about your reputation. If you're always truthful to what you say, you'd be trusted more and people might approach you and give you quests that a liar would never get.

But here, again, this would be a consequence of the actual actions following the player's intention. The intention in and by itself has no consequence whatsoever on anything because it only exists within the player's - or the player character's, if you will - head.

This way, your thief example is easily debunked: if your character says "Let me do it, I can steal the money without harming the man." but you then go on and torture and kill him for the money, the intention becomes clear: you just wanted to get the job and lied about how you would do it. This could provoke dialogue responses by the other thief the next time you meet him.

As you can see, intentions become clear as soon as you do an action, not as soon as you say something. If you say something and do exactly what you said later, then your intention was to do what you said. If you do the opposite - your intention was to lie in order to get the quest or whatever.

tl;dr:
Intention does not matter beyond what you say and what you do. What the player (or the character) thinks about it doesn't affect the world in any way since it only exists in the char's head.

 

[zeitgeist]

Player intent is impossible to guess accurately and shouldn't be taken into account.

Character intent, on the other hand, would be a neat mechanic to implement consistently throughout the entire game (through something more complex than a handful of [Lie] dialogue options like it's been implemented in most games where it's even been featured). However...

As you can see, intentions become clear as soon as you do an action, not as soon as you say something. If you say something and do exactly what you said later, then your intention was to do what you said. If you do the opposite - your intention was to lie in order to get the quest or whatever.

...this is a pretty good point. If a game has a well developed system of NPC interaction, goals/quests/missions/whatever, and a world that actually reacts to the player character's actions beyond quest completed TRUE/FALSE, is broadcasting intent to the game preemptively really needed for anything except tracking the inner alignment, which no one in the game (except perhaps a deity or a supernatural entity or something) would see?

I mean, let's say your character says "I'll do quest X" to an impoverished peasant. Then he doesn't. Does it really matter to the peasant why exactly your character didn't do this? Maybe you lied out of sheer malice. Maybe you forgot. Maybe you just couldn't do it for objective reasons. Maybe someone offered you something in return for not doing it. But for all practical purposes, your character broke his promise, and the peasant NPC will have to explain this to himself according to his inner set of values, his relationship to your character, and other variables that don't really have that much to do with what you actually intended to do. One possible consequence for this could be that the NPC marks you as having certain levels of unreliability and/or malice (in his personal "what I think of various NPCs/PCs in the gameworld database) and spreads this information to the NPCs he interacts with in the world (who might believe him to some extent, might believe him fully, or might not believe him at all depending on his relation to the other NPC and his social standing etc.)

Anyway, to do any of this properly, you'd have to track more variables than any RPG made to date has tracked. [Lie] dialogue options or a mood/intent setting you choose during a conversation are probably the only feasible abstractions/simplifications of this system.

 

[Shemar]

Anyway, to do any of this properly, you'd have to track more variables than any RPG made to date has tracked. [Lie] dialogue options or a mood/intent setting you choose during a conversation are probably the only feasible abstractions/simplifications of this system.

This. My characters are almost always the diplomatic rogue type, yet somehow I end up being the paragon of goodness, just because I am smart enough to tell people what they want to hear instead of being a jerk to them. It would take a whole lot more opions than truth/lie, for each of all possible dialogue options, to even begin to properly measure my character's intent.

 

[Excommunicator]

This way, your thief example is easily debunked: if your character says "Let me do it, I can steal the money without harming the man." but you then go on and torture and kill him for the money, the intention becomes clear: you just wanted to get the job and lied about how you would do it. This could provoke dialogue responses by the other thief the next time you meet him.

I understand you are in agreement of the principle of reacting to action instead of reasoning behind the action, but I felt the need to respond to this part here. My example was not just about intent, but motive as well. Your suggestion certainly doesn't "debunk" the problem, it is a very simplified version that only determines that at one point the character decided he would not kill the man.

Your "solution" implies the character's morality is known; I stated the character's morality according to the 30yr old RPGer, but the game wouldn't know this, and therefore you can't consider it in your solution. As far as this example is concerned, you are a machine who only knows that at one point the character has said that he can steal from the man without harming him.

You don't know why the character said what he did at the time of the conversation, and you certainly don't know if he initially lied, or actually meant it at the time.

And let us be honest here, I know I made an offhand mention of "killed and tortured" as part of my example, but considering the situation as a "killed and tortured" murder is a theoretical cop-out, which avoids one of the biggest problems in games and the real world alike - the differences between meaning no harm, meaning actual harm, and meaning death. For the sake of the integrity of the discussion, let us not pretend that you won't be confronted with many situations in which the character can be killed accidentally by the player character (or other characters, as any good simulation should allow).

Either way, I understand you have the right idea about the main approach, it simply needed to be shown how much of a complex issue this really is.