Can we engineer a hypothetical RPG setup that's social yet essentially like solo? Let's find out!

edited August 2014 in Story Games
Edit: I changed the title, and have repurposed the post to be a bit different. The thought experiment below was meant to get social RPG'ing to feel as close as possible to the solitaire mode of this activity, and by proxy see if people thought that solitaire rpg is possible.

Since I was thinking the hypothetical setup and the solo experience were essentially equivalent, then something could be learned about whether it is meaningful to draw the line of what is an rpg on a pre-condition that it be social.

I hope to keep modifying this experiment to see if it gets to a state where it is a better social equivalent to solo gaming. Feedback and suggestions welcome, of course.

Essentially, we can build something up and then viciously try to tear it down, and then try to build it up stronger by changing something. This would continue as a cycle until it can't be torn down or it can't be built stronger. In the latter case, that would be when the changes would make the setup not social in the sense of the most common denominator definition of the word.

So, this is the latest set of rules for the setup:

First the assumptions:

1. Assume the players have unlimited patience, unlimited time, unlimited lifespan. Or, alternatively, that they have the ability to think and communicate faster than the speed of light or something.
2. Anything not explicitly prohibited, is permitted.
3. Rules might be adjusted in response to #2.
4. For now, assume just one GM and one player. (If more than one player, my intuition is that they must communicate through the GM using the yes/no rule.)

The yes/no rule restated and expanded:

1. The GM can only answer yes or no to anything the players ask.
2. Corollary: The players can only get information out of the GM by the use of any type of questions that can be answered by a yes or no. The cororally means that the players can very well write a long statement and ask at the end, "Is this true?"
3. Any idea the GM has must satisfy at least one of the following conditions:
3.a. It can be explained in verbal terms the player can understand
3.b. It can be extrapolated from other ideas that can be put in verbal terms the player can understand
3.c. It can be represented non-verbally in a way the player can access and can show the GM.
3.d. It has the possibility of being perceived directly in some way by the player, who will then be able to put it in terms of #1,#2, or #3
3.e. It has the possibility of being approximated by another idea that satisfies #3 or #4, which then can be put in terms of #3.

4. Players must ask the GM if they need to roll for any action their PC takes no matter how trivial. The only exception being internal thoughts, provided there isn't some context preventing that.
5. Players have the right to author facts of the world as long as it is constrained by circumstantial rules or tied up to a resource system.
6. If there are multiple players/PCs, they must be in separate rooms and must communicate through the GM. All GM communication restrictions apply.

-----------
Bookmarks and reminders of things I want to visit, or revisit later


Aslf's comments:
http://story-games.com/forums/discussion/comment/436167#Comment_436167

Dreamer: I'm not very math oriented so AI concepts tend to escape me when it gets to the nitty gritty. The closest thing I've found to a book I can understand is The Hitch-Hiker's Guide To AI. I've always been curious as to whether knowledge representation techniques could help with P&P solo rpgs. The old random tables standby works great, but maybe there are different approaches.
Paul_T's comments:
http://story-games.com/forums/discussion/comment/436187#Comment_436187

Dreamer: Not sure I want to think of things as automating, rather than assuming a larger load of responsibility and creativity while maintaining a sense of "pushback". The crazy thought: since in the hypothetical setup, the GM using dice results in an illusion, part of the feel must come down to what being aware of another human being does to us psychologically. If I created a web service that acted as a Mythic like Oracle, but also had human members secretly GM'ing with the same restrictions as the Oracle, could this illusion be maintained? You would log in, and you wouldn't know if you had a live GM answering your binary questions, and throwing you keywords from a limited list (like Mythic's). I wonder how that would feel.

Some more crazy musings:


For me, part of what makes the yes/no question method in solo games open to abuse is that you can ask anything and there's always a chance of receiving a yes answer. However, this power is also nice in that you get the possibility to control the vision of the world as a player. How fun would this be with a live GM? Superfluous, given conflict mechanics and narration rights mechanics?

-----------
I posted this on RPGNet, but I've slightly modified it here to make it clearer. I wonder if some of you would indulge me and carry out this thought experiment for me:

Imagine you have a regular RPG group that sits together physically. It can be any RPG you like.

The twist is that for this session you have agreed to some very weird restrictions:

1. A person will always be in charge of controlling the world and NPCs.

2. You have all agreed, hoever, that you will only roleplay and communicate by writing stuff down. This can be a single notebook, networked laptops that you brought to the table, etc a single laptop, etc. (It must be a single recording object)

3. The GM can read everything the players write. The players can read what the other players write only on their turn (as if using only one notebook/device).

4. *Players* (as opposed to the person GMing) can only write things in character, or as a player narrating what her character does or thinks, etc.

5. The only OOC communication allowed is when you write a question to the GM. That question has to be in yes/no format. You can ask whatever you want, but these are meant to learn about the GM controlled world.

6. When acting as a GM, the person's only contribution will be to answer those yes or no to player questions.

7. Players can, however, interpret those yes/no answers and extrapolate content into the world [the can also extrapolate from previously established fiction]. These additions are just as true as if the GM had answered 'yes' to a question. It's up to the player to decide when she wants to stop extrapolating and hit the GM for another question (or end her turn).


I'm not really concerned with whether the activity would be fun or interesting, or even something anyone would play, but I'm concerned with knowing:

However un-fun, or bad, does this still feel like role playing to you? If not, why? And what, at a minimum, would you need to change to make this be an RPG?

Comments

  • Yeah...to me, it definitely all sounds like an RPG, the sort of thing that a bunch of introverts might play in a coffee shop.
  • It is an RPG, but I find that such a generic term as to be almost useless: everyone tends to have a very different view of what you're talking about when you say RPG and all those views are correct because the term itself covers such a broad range of activities.
  • edited July 2014
    Yeah, RPG as a term is like a big tent. Still, I've seen folks draw some very rigid lines: taking on a role, shared imaginative space, etc.

    I am sort of curious as to which lines this might violate for those who would declare it as not fitting the RPG category.
  • I always found it hard to run a PbF or PBeM game, but a procedure this looks like one step towards solving that.
  • So, lest i come across to anyone here as dishonest, let me disclose that this thought experiment came about as an attempt to troubleshoot a not super common, but not unheard of, problem in the solo RPG experience. It seems to me that it is essentially like a bare bones solo RPG experience with so called GM emulators. In fact, i think that the GM in question could secretly be rolling dice at a 50/50 chance for every single question, and the players could very well not catch on (except for maybe thinking the GM sucks if the imaginary happenings get silly).

    The rule that i actually think could blow the experiment is the player restrictions, in that as written right now, the players are allowed to write back and forth to each other in character or about how their characters interact towards the other.

    The reason I didn't want to disclose what the experiment was about upfront is that I thought and still think it is the best approach to getting an opinion without conscious or unconscious biases. I figured that if a bit of misdirection was good in that regard for psychological surveys or experiments, it was good enough for this.

    Thanks for the input so far.
  • edited July 2014
    I figured that if a bit of misdirection was good in that regard for psychological surveys or experiments, it was good enough for this.
    I've been following this concept with interest, and I totally agree. Prodding mechanisms can be prone to certain types of social backfire, and therefore a didactic approach must be taken if these backfires are to be avoided. This is as true in game design as it is in sociological inventory. That's part of the Art.
  • This is basically Callisto, by the way.
  • This is basically Callisto, by the way.
    I really should check this one out. You're the second person to mention it.

  • I've noticed that as I think more about the setup, I seem to see ways in which I can get it to be much more closer to one of the approaches to solo gaming. The current one, as I mentioned, has a couple of wrinkles such as the players being able to IC roleplay with each other.

    I will probably keep updating the setup and post it in new posts, rather than edit the original post above. Hope that those interested don't mind too much.

    I actually think I'll edit the topic if I can to something more collaboratively.
  • edited July 2014
    So, I think that it is possible to get almost all information one needs about an imagined world inside the GM's head using yes/no questions (like 20 questions but without the limit). Just imagine you're the GM, and for every little detail of the world that you would describe to a player, pose a yes/no question to yourself instead.

    The only exception to that would be anything that the GM is imagining which the player has never seen or heard of. For example, imagine that your GM is an alien from Uranus (sorry, I'm feeling even more juvenile today). It's probable that there are life forms on Uranus that as an earthling you have never even heard of. Possibly they might have features that are unique to Uranus out of all creation, so there's no way you would ever be able to imagine them.

    To make it a less juvenile example, imagine that someone from a tropical country who's never even heard of snow. It just doesn't exist in his or her world and I think it would be hard (impossible?) to arrive at it extrapolating from pure imagination, so that would be one thing this player would never be able to guess. As such, this would never enter the player's imagined space via his yes/no questions, no matter the number or how exhaustive the list of questions.

    At the same time, in practical terms (and practical is a funny word to use in this hypothetical setup), I think most players would almost never run into such a wall since they're generally well versed in their settings and tropes-- so for most use cases, the yes/no setup are covered.

    Hope this makes sense to anyone reading.

  • The only exception to that would be anything that the GM is imagining which the player has never seen or heard of.
    You know, it's funny, but to me this is the primary role of the GM - the one I enjoy the most.

    For everything else, like you say, you could just ask yes or no. But the interesting stuff, the stuff I value the GM for? It's in there.

  • edited July 2014
    Hi @Paul_T!

    For sure. I love those moments where a GM surprises you in a good way.

    Your comment forced me to go back to the statement you quoted, and re-examine it. I assumed that the use cases the yes/no method covered were enough, but given your comment, now I need to find a way to defend it. The way I’m envisioning this exercise is that I try to build something, then try to tear it down viciously, then try to build it up again covering those objections, and repeat the process.

    I think I also need to explain some conceits that I never explained:

    In general, if something isn’t specified, I think we can treat it as something that is up to the imaginary persons involved. So, for example, a GM could lie in his answers, but we can’t assume she will. A player could ask illogical questions that paint him into a corner, but we can’t assume he will— taking the real life example of solo players, they tend not to ask an Oracle to confirm a conclusion that is the only logical one in terms of the setting.

    I also think that some real world physical rules and such don’t apply here. Assume the players have unlimited patience, unlimited time, unlimited lifespan. Or, alternatively, that they have the ability to think and communicate faster than the speed of light or something. Basically, this should imply that players are able to literally ask a gazillion questions just to get one concept or word right and they don’t mind. That’s why the exercise is theoretical.


    Anyway, your post made me think, and I think I may have a solution. I’ve been trying to find a succinct way to elucidate it, and completely failed in doing that. Here it goes anyway:

    I think that if we take all constraints out, the challenge is trying to get a concept across using one of the five senses. For the tropical person who’s never seen snow, the GM can show a couple of pictures and with a bit of explanation, it would probably suffice. The GM would only run into trouble if he can’t find the words or images to get the point across. If we put in back the constraint of words only (writing or otherwise), it’s could potentially get more difficult to find the words or analogies to explain to the player what the concept of snow is-- if what the player knows cannot approximate the concept of snow, or cannot be used to extrapolate from it.

    I really doubt this can happen except in cases where maybe you need some voodoo math to explain something, like perceiving more than 3 dimensions (and only if it was critical for the player to actually be able to accurately perceive what it would be like). For the other stuff, I just kept thinking that if man was able to imagine himself on the moon and actually getting there, we can probably be able to extrapolate from what we do know into something completely new. It’s still imperfect since you never really know what something is like until you actually experience it (I think). Again, this is without the constraint of yes/no questions.

    (Another example I tried to think up: imagine your mind by some accident of the universe somehow got transferred to an incorporeal dimension where entities cannot see, touch, or hear each other or anything. It’s basically a void with no matter. They can only hear each other’s voice or their own, in the way you hear your own voice in your mind. How do you explain the concept of snow to such an entity? Maybe there’s no way.)

    With the yes/no constraint, I still think it’s possible to arrive at those same concepts, provided you don’t have to know the actual terminology for them. The main difference is that a player has to do an exhaustive search to finally arrive at a point to which the GM can say ‘yes’ (‘yes, that’s it’). He could exhaustively search for pictures to show the GM, for example. Or, he could do a painfully exhaustive search for all the necessary ingredients that make snow. “Is it something white?”, [A million ‘no’ answers in between], “Is it something powdery?”, [A billion ‘no’ answers in between], “Is it something much more colder than the temperature we are experiencing now?”.

    That player, relying on his own words only, might never be able to guess the name is actually ‘snow’, but he will be able to iteratively cobble together a description of snow using the vocabulary he does know, each time asking the GM whether the description so far matches what he is thinking of. The stop point would be when something extra is added to the description and it is no longer snow, so the player now knows to go back to his last description and imagine snow without knowing the correct term. Like, if we were to approximate the different words eskimos have for snow using English to get an Eskimo GM to understand what we mean. Actually, what I mean is if the Eskimo GM had to explain their concept of some type of snow to us without using their language, but using English. We wouldn't know the correct Eskimo term, but we could describe the concept.

    *****

    So I tried to come up with an algorithm of what a normal GM process for describing something might look like, and one for the reverse situation of the player trying to extract it from the GM with yes/no questions
    Normal GM process: (IS == Imaginary Space)

    For each GM description
    ````If player able to render correct concept in IS
    ````````player makes decision

    ````Else
    ```````` Do
    ````````````GM searches for synonymous descriptions oranalogies player can extrapolate from
    ````````Until player can render correct concept in IS, OR synonymous Descriptions and analogies are exhausted

    ````````If player renders correct concept in IS
    ````````````Player makes decision
    ````````Else
    ````````````???? (What happens?)
    End For

    ----------
    Yes/no question process for player. I don’t have an explicit stop point where a description cobbled together goes too far, but it’s implied in the “until” condition. If he can render it, he’s gone past the correct description and stepped back one step.

    For each description in GM’s mind

    ````Do
    ````````Player asks yes/no question
    ````````If player guess is correct Then
    ````````````GM answers yes
    ````````Else If player guess is synonymous
    ````````````GM answers yes
    ````````Else If player tests correct analogy ("is it like X?")
    ````````````GM answers yes
    ````````Else
    ````````````GM answers no
    ````Until player can render correct concept in IS, or all guesses, synonymous guesses and analogies are exhausted

    ````If player rendered correct concept in IS
    ````````Player makes decision
    ````Else
    ````````???? (What happens??)
    End For
    ---------

    Now I need to go to town on trying to find arguments against this.
  • With you so far, @Dreamer. Nice to see some procedural pseudocode up in here. Not to mention the philosophical thought-experiment (Helen Keller, anyone?)
  • edited July 2014
    Thanks @aslf! Glad you're keeping an eye on this thread.

    I think I found a way to break the assumption that any information can be extracted with binary questions. I had three ways to undermine the claim, but I found how to get around  them. The last one seems too solid to counter argue away.

    He first examples were proper names one has never heard of: "Pedro z$567##%", if for some reason the GM named someone that way. You could always construct it symbol by symbol with the same algorithm, so it's covered.

    That is, provided you knew the proper alphabet, so then came proper names in some alphabet you don't know. You could always try vocalizing sounds and constructing a string of sounds that sounds close enough. I think that's covered if the GM isn't ultra anal about it. Even if he was, I think I can let that case slide.

    Then came the one that threw me for a loop for a while: someone who's never seen light in their lives so theybcant even draw on similar concepts. But  the GM needs to get that across, while GMing  in the pitch black dark with them. You could never get that across with words only, I think, but you could always come prepared with a flash light. Let there be light, and there you go. How do you extract that with binary questions? 

    I finally figured it out when I realized there is no reason the GM can't also bring a light in that setup too, but the rule says no contributions other than answering yes or no...so thats the  case that I think shows one absolute limit to the binary question rule. I can't think of any way of getting around it. if anyone can, I love to hear something.

    The only way to handle it would be to change the rule to allow the GM to bring a prop and let the player discover it with questions.  But I can't do that because GM emulators don't bring props to our games...cheap mofos.

    All that being said, I feel this category of problem is a real edge case , so I'm inclined to say that the claim that the  binary question rule can get almost any info the player needs is at least reasonably robust (by which I really mean to say its pretty darn robust, but I'm biased). 

    I think I can move on to other rules, though I have a feeling this one is the most important, so I may just move to what I think is the equivalency to solo Rpgs.

    However, if anyone can think of other categories that this binary question rule can't cover, let's hear them too. I am too biased to see my blind spots.
  • edited July 2014
    I have to thank you for that Hellen Keller reference. Suddenly it hit me: In a normal GM situation, to get a concept across, you MUST be able to either do one of these:

    1. Explain it in language terms the player can understand
    2. Find other concepts, in language terms the player can understand, to extrapolate from
    3. Failing all that: have a representation of the concept the player can perceive,
    4. or have the player perceive the concept directly.
    5. or have the player directly perceive an approximation of the concept that can then be put in terms of #2 (verbal extrapolation)

    ThisThe act of showing a representation or concrete instance of the idea is what goes in the "?????" in the first algorithm for GM to player communication.

    The yes/no rule when I thought about it some more is simply making the player responsible for finding all these things and communicating them to the GM for a yes or no answer. So. one way to fix the experiment, without changing the yes/no rule, is to add the assumptions above but from the player's perspective.

    Any idea the GM has must satisfy at least one of the following conditions:

    1. It can be explained in verbal terms the player can understand
    2. It can be extrapolated from other ideas that can be put in verbal terms the player can understand
    3. It can be represented non-verbally in a way the player can access and can show the GM.
    4. It has the possibility of being perceived directly in some way by the player, who will then be able to put it in terms of #1,#2, or #3
    5. It has the possibility of being approximated by another idea that satisfies #3 or #4, which then can be put in terms of #3.

    All of these things can be found by doing an exhaustive search for them.

    All of this means that, for the example of explaining light, we've narrowed the activity to only players who haven't lived all their lives in a pitch-black, hermetically sealed tank that they can never leave. I think these conditions cover this kind of category having to do with direct perception.

    And in fact, it just occurred to me, assuming Hellen Keller's world was pitch black like that, how do you as a GM with absolutely no limits on what you can do, get across the concept of what light is? Sounds, you could still show a very distant approximation (let her feel the vibrations of a boom box). For light, you would need to wait for a cure.

    Now that I think about it, as a parent, her story makes me feel sad. I'm gonna go hug my kid now.
  • Thanks for the interesting conversation. It's a fun read, and I also enjoyed the pseudocode, as AsIf calls it. :)

    I'm curious, though, what is the goal here?

    The reason I ask is because many people come here looking to design a game.

    But you say things like:

    "Assume the players have unlimited patience, unlimited time, unlimited lifespan. Or, alternatively, that they have the ability to think and communicate faster than the speed of light or something."

    Is this just a thought experiment? Because an assumption like this seems (to me) to rule out any actual possible game design.
  • edited July 2014
    Hi  @Paul_T !

    It is a thought experiment trying to find out if there is anything in the social roleplaying spectrum that is equivalent to some mode of solo rpg gaming that I've experienced. I am obsessed with it for the moment as a sort of brain teaser, but I also feel I may get insights into the activity and what direction I can take with it.

    It started from thinking about those times I've either felt or heard others complain that all or some of their sessions of solo gaming felt like "just writing" (sometimes "just writing with dice"). It's not an endemic complaint from what I've observed, but it's not super rare. Yet, there are people who have never made that comparison and play solo all the time just fine, and feel they are roleplaying.

    I constructed the first version of the experiment as a challenge to myself and those like me, because the implied point of comparison is social roleplaying. It's meant to ask, "if said social activity is roleplaying, and it's equivalent to how you play solo, what exactly is causing you to feel that solo feels different?" I think it will help to find out if it's something to do with awareness, or if it's the medium of writing that many soloists tend to default to (so they wouldn't like pbf formats), or perhaps an unacknowledged preference for social RPGs that excludes the other mode-- or maybe something  can't even think of yet. 

    As I was doing the thought experiment by myself, however, I came to realize that if the hypothetical social situation fits the definition of roleplaying, then the solo analog would too. I.e. if the GM secretly starts rolling dice for a 50/50  chance of answering yes or no, without the player being the wiser, then from the player's perspective, is there a difference at all?  

    if that is not true, if solo RPGs conclusively do not exist, I can stop wasting my time trying to focus on the taking on a role part and focus on the other aspects. I would then try to find ways of making the activity a fun story-writing or story reading  experience. Maybe even incorporating aspects of the RPGs i love that are amenable to making being an author a fun game (furry fanfic here i come!!). That alone would make this exercise worth it for me. :)

    On the other hand, if solo RPGs  really are a thing, then I can still approach my problems with an assumption that I can roleplay in some form solo, and it may be possible to play around with that form to make it more useable &  more fun without losing what makes it an rpg experience.

    The yes/no rule, for example, is super onerous if you want constant interaction with the GM. Yet I know that many games nowadays give players the power to author some of the world. What effect could those approaches have on the hypothetical setup if we added them?   Can I add them without breaking the illusion that the GM is contributing to the SIS (as opposed to the oracle he's heating the player with)?  What implications does that then have for solo RPGs, where you can't trick yourself like that? Would those approaches  still help make for a more fun experience? Etc. 

    In that same vein, I could also look for parallels in social RPGs to fix my problems in solitaire. For example,  I may need to resist the urge of trying to imitate William Gibson's talent during actual play, and learn to go with the the flow of my sucky writing. After all, social rpg sessions tend to move fast and people are not trying to write like Nabokov as they play. 

    But I'm getting way ahead of myself now. I don't want to reach any conclusions, no matter how obvious,  until I reason my way to the final conclusion: which is that if the player can't tell the GM is cheating by using an oracle, then the player is roleplaying solo since a couple of dice are essentially dictating what enters the fiction-- as long as the player accepts it. I feel this is true, but only if he same activity when the GM is not cheating with an oracle is roleplaying. That's why I tried (and so far, I think succeeded) in proving that the yes/no question rule allows the player to know all the info needed to take on a role and act in the fictional world. I still have at least one more objection I want to think about before I move on.

    By the way, I posted an early version of this on a solo group, and I had  good feedback, but hardly anyone tried to challenge my conclusion.  That's why I also posted it in places where i know there is a  mix of opinions, because someone is bound to push back and test my logic. 

    Some perspectives might be changed (mine/others'), or it might end in agreeing to disagree. I've already learned some things, I feel. :)  cheers!
  • I should warn that this is an interpretation of the Lumpley Principle that was mentioned to me on RPGNet by someone. I guess one must sooner or later reckon with that principle and the Czege oen, if exploring this sort of question. I am not sure if this person's intepretation is correct or not, so what follows may very well be the attempted destruction of a strawman argument.

    So, the objection is that the yes/no rule "reduces the GM to a mechanic", thus the GM cannot contribute to the SIS. Ergo the activity is not a roleplaying game. I guess that depends on whether being able to shape the SIS is a contribution or not. This opinion appears to assert it is not.

    That doesn't feel intuitively true to me. Let's ignore the fact that people can play 1 player games, and focus on the veto power. The GM veto is essentially a deletion, since you can't express something without imagining it first, and you can't reject it consciously, when it is put to you, unless imagine it first. For example:

    Do not imagine a pink elephant walking down the street as I say this.

    Were you able to avoid having that image flash into your brain? It's been asserted that most people can't.

    In light of that, one would have to consider deletions from the imagined space not to be contributions. As an analogy, that would mean deletions like the following, which create new words, would not be considered a contribution:

    (Cribbed from a forum...joink!!)


    Delete a letter...to always make a real english word each time until you are left with only a 1 letter word!
    V. clever (I'll post the answer if no-one gets it!).

    STARTLING
    STARLING
    STARING
    STRING
    STING
    SING
    SIN
    IN
    I

    To me, the GM in this setup is like a wood craftsman taking away features and superfluous details the player presents to shape the GM's vision. Her veto is her knife.

    Another way I look at it is by imagining a GM who is paralyzed, and only able to communicate by blinking yes or no ( or ring a bell like Tio Salamanca from Breaking Bad). If this person were a GM, sometimes (most?) it would be faster perhaps to point at each letter and build a description that way. However, what if at some point based on the context of the setting, the player is able to guess ahead of time what the GM is trying to get at? Does that suddenly not count as the GM's contribution even though it's the GM's vision? Should the player just wait for the GM to spell it out so that it counts as the GM's contribution? The same could be asked about a player's contributions, if the situation were reversed.

    Thus, I find this position intuitively hard to accept, as it's hard for me to view someone who is imposing their will and a specific vision as a non contributor. This one perhaps comes to a matter of opinion, so a person who agrees to disagree with me on this point, will not find the experiment useful past this point. Thanks for coming along so far. :)
  • I'd argue that deletions from the SIS are reinforcements of what belongs in the SIS. This is the "constructive denial" interpretation of Right to Dream play.

    Realize that SIS is pretty close in meaning to "the fiction" but I think it also must include some meta constructs like "the stuff we all know doesn't belong in the fiction, now that we've talked about it." Big-E Exploration in the Big Model includes little-e exploration of all the components of play. While that includes the usual things like Setting, Character, and Situation, it also must include System. Exploration of System includes pieces of "constructive denial."
  • edited August 2014
    Automation of the GM will get better as time passes, and eventually it will happen that a mechanical turk GM can pass a modified "Turing Test". Ultimately it should be impossible for a player to tell whether a newly-inserted element was indicated by a random die roll, triggered by a mechanical rule, or simply made up on the spot (just as it is in a well-run trad GMd game).

    But personally I believe the hardest part (of creating a system that could pass that test convincingly) has less to do with allowances/rejections per se and more to do with connecting datapoints via spontaneous intuitive leaps, (i.e. linking previously-disparate elements together in surprising ways while remaining loyal to the tone/genre and the symbolism of datapoints already established). Then secondarily, the logic behind the suppression/withholding of such data until a properly dramatic moment for their revelation occurs.

    I know I'm getting ahead of you, but there's something more to think about.
  • edited August 2014
    Thanks folks! I can always count on this forum to give me food for thought.

    @AsIf

    I'm now creating "bookmarks" on my OP to mark things I want to re-visit. There are a few things I was thinking about which I wanted to save for later, but I've forgotten them. I may get them back. I wasn't really aiming for AI initially, as opposed to the psychological effect just having a person in front of you might have as it regards rpgs. Yet, AI and solo rpgs is just too juicy to ignore, so I hope we get to talk about it a bit. :)

    @Adam_Dray

    Thanks for that, man! I found this link, and it's interesting reading: http://indie-rpgs.com/archive/index.php?topic=17334.msg188019#msg188019. Though not totally unfamiliar with the Big Model, I'm not really conversant in it, so I can't really provide too much intelligent conversation in those terms, but I think it cemented my feelings on what I said above.

    I think the 'yes' is probably more 'active' in the act of reinforcing than the veto is. The veto, as a repudiation, can also be a force that shapes the future the course of the fiction while, perhaps, at the same time ratifying what came before. It's the 'pushback', albeit more indirect and passive, that MatrixGamer mentioned a while back, on another post, as being one of the things he needed for the activity to feel satisfying. It actually feels a bit more 'active' than the 'yes', though I view them both as sort of passive and indirect.

    One feels like "OK, proceed as you were doing anyway", while the other one has this feeling of either "stop right there" or "change course." The veto is more like a an obstacle that forces the flow of the river to change and move around it, while the stamp of approval is like...the water analogies escape me. :)
  • Constructive Denial is a stamp of approval in the negative!

    In a highly functional Right to Dream group, the "veto" you're talking about is probably as exciting to the group as when someone Gets It Right. Everyone is in agreement about it, either way, and that solidifies everyone's understanding of the SIS.

    "No, man, elves don't hate dwarves in this world at ALL. They get along fantastically." "Because of their shared hatred of halflings, right?" "Yeah, man. *Everyone* hates the halflings!" "Fucking halflings." "Yeah."
  • Let me ask a hypothetical question:

    Would it be at all interesting and/or possible to "automate" a player this way?

    What if you were GMing, and you flipped coins on yes/no questions to see what the player did?

    If it's different, why?
  • edited August 2014
    @Adam_Dray

    I get it. If the group knows where the fiction is going anyway,and there are no forseeable disagreements, the rejection reinforces what they had in mind anyway.

    But let's not get completely out of context as to the situation I'm describing when vetoes are applied, which is a situation where the GM is explicitly limited to yes/no answers, and the player is limited to guesses if he wants to know what the GM has in mind. The vetoes could apply to disagreements, but they are also a guiding force, since the player is just trying to guess what the GM is thinking so he can get the information he needs to act. The vetoes act like obstacles placed at different exists so that the player's imagination is forced to go in the direction the GM wants, until the exact piece of information is arrived at so that they can be on the same page.

    This is not like mind control either, but like binary guiding sign-posts that shape the course of the fictional stuff.

    Edit; Maybe 'veto' is an unfortunate word choice. Denial or just keeping it as 'no' would have been more accurate.

    @Paul_T

    I've actually seen people play like that, even going so far as to invent player personalities that are meta to the PCs themselves. Some of them have run parties of PCs through old D&D modules. I don't know if it'd be interesting for me, but it seems to be for them. Edit: Mind you not in an exact yes/no only setup. (I'd like to explore this later as well)

    I'm also not sure I think of it as automation anymore. It implies too much about a pair of dice 'thinking'...ay, no me gusta. I think of it a bit like what it really is: player assuming more responsibility and creativity, yet maintaining that sense of 'pushback'. A crazy thought occurred to me, but I want to save this line of thinking for later. (I "bookmarked" your post with a link on my OP for later)
  • Ok. Interesting!

    It makes more sense (and sounds more appealing) when I think about it as "a sense of pushback" than "GM automation".
  • Folks,

    I put up a sort of concrete version of the experiment up at the G+ group Lone Wolf Roleplaying. It's not as constrained as what I have in mind here, so it's not exactly an analog. Players have a lot of freedom to do as they want (and I can't control people), so we'll see how it goes. I do have an idea that could force more constraints, but it would involve software. Anyway, I'm rambling.

    I'll get back to my inquiry soon. :)

    Paul,

    Here's a link to a post where the player made up personalities for other players while playing Microscope. Not my thing, but he seemed to do fine with it: http://www.risusmonkey.com/2011/03/solo-gaming-on-long-flight.html :)
  • edited August 2014
    It makes more sense (and sounds more appealing) when I think about it as "a sense of pushback" than "GM automation".
    I know the machine metaphor bugs you, Paul. But on my view, that's really all the GM ever was. A component of the system, who just happens to be human. Someone who could adjudicate the rules and direct the general scene without being totally focused on playing a particular character. In the earliest days of our hobby it became obvious that for certain types of game, such a role was necessary and desirable. If the rules could have done it themselves, they would have been written that way. Clearly the reason a human is required to do that job is because the rules themselves couldn't possibly handle the complexity of all the creative ideas that might arise to need adjudicating. I'm sure you're with me so far.

    But computers are good at handling complexity. And AI is getting better all the time. So...

    If there was a program that exhibited all the creative and intelligent traits a GM needs to have - and one day AI will get there - do you think you'd eschew using it? I'd bet not! I'm pretty sure you'd use the hell out of that program!
  • Yeah, we definitely have a total and complete disconnect here, Tod.

    I wouldn't use that program. Not unless it matched the intellectual, creative, and social aptitude of a human - but in that case I would probably consider it a human (or at least as intelligent, sentient, and deserving of free will as a human being).

    Let me ask you a similar question:

    If such a program could be designed to take on the players' roles, would you GM a game for several such "automatons"?

    (It may well be that we just look for entirely different things out of roleplaying, of course.)
  • edited August 2014
    Yes, I am absolutely assuming the AI of movies. The AI that can replace a creative human. My question assumes that.

    Yes, I would run AI Players for sure. I would use them for playtesting systems in development, at the least. Maybe for fleshing out small parties.

    But note that in the genealogy of the GM, it is the Player who is assumed to be the important party. The one for whom the game is written in the first place. The Player is like a "user" of the system, and a move that triggers a mechanic is like an internet client requesting a function output from a server. The human GM is in there basically because it isn't possible to do certain types of games without one. Yet.
  • Ah, I see. So, in this view, a human is just an advanced machine, as well.

    Fair enough!
  • edited August 2014
    So, I think I've covered every objection I've come across so far, or that I could think of myself. I can always revisit if necessary, but for now I'm moving on, and re-stating the yes/no rule as a set of rules.

    First the assumptions:

    1. Assume the players have unlimited patience, unlimited time, unlimited lifespan. Or, alternatively, that they have the ability to think and communicate faster than the speed of light or something.
    2. Anything not explicitly prohibited, is permitted.
    3. Rules might be adjusted in response to #2.
    4. For now, assume just one GM and one player. (If more than one player, my intuition is that they must communicate through the GM using the yes/no rule.)

    The yes/no rule restated and expanded:

    1. The GM can only answer yes or no to anything the players ask.
    2. Corollary: The players can only get information out of the GM by the use of any type of questions that can be answered by a yes or no. The cororally means that the players can very well write a long statement and ask at the end, "Is this true?"
    3. Any idea the GM has must satisfy at least one of the following conditions:
    a. It can be explained in verbal terms the player can understand
    b. It can be extrapolated from other ideas that can be put in verbal terms the player can understand
    c. It can be represented non-verbally in a way the player can access and can show the GM.
    d. It has the possibility of being perceived directly in some way by the player, who will then be able to put it in terms of #1,#2, or #3
    e. It has the possibility of being approximated by another idea that satisfies #3 or #4, which then can be put in terms of #3.

    Moving on from these first rules, I can now think about what other rules and assumptions are needed. I will focus now on what the Player(s) can do or cannot do. One good place to start is how do they decide when to make a roll, say for task resolution?

    The safest way to ensure the GM has complete control of this is to say that the player must ask the GM "Do I need to roll?" for every action, outside of stating what the PC is thinking. Like, "My PC will brush his teeth. Do I need to roll?" Well, normally you can just do it, but there's the implicit "yes" from the GM. The GM could tell you, "no, you must roll against dexterity to see if you can avoid poking your eye." He could just be a jerk, or maybe it could be the context that warrants it. Whatever the case, the GM is always making a judgement (IMO).

    The yes/no rule, like almost everything else, almost inverts this, so if the player isn't allowed to make any such judgements or assumptions, he must ask the GM whether he needs to roll for any given action. Otherwise, the GM and the player need to decide on some parameters, and some power must be given to the player to make judgements.

    I think there's a whole spectrum there, from no power at all (must ask for every single action), to all the power to judge (never has to ask). The most restrictive end of the spectrum still fits within the description of playing an RPG (IMO), since it essentially continues the yes/no rule-- knowing whether to roll or not is information you get from the GM after all. However,what about the other extreme end? Would that still fit the definition of playing an RPG? If not, where in the spectrum do you draw the line?

    A related issue pops up with narration or authoring rights as concern adding things to the world beyond just controlling your PC. The most restrictive extreme fits the definition of playing an RPG (a traditional one perhaps), but what about the other extreme of total freedom?

    Most of the newer games I'm familiar with, which allow players authoring rights, usually have some limitation around them like specific conditions in which that can be done, or it's tied to some resource like bennies. So maybe that's where the line should be drawn. As of now, I will assume a system or resource of some kind should be a requirement for any rule concerning player authoring (something which incidentally is missing in most solo play).

    So, provisional rules for the Player(s):

    4. Players must ask the GM if they need to roll for any action their PC takes no matter how trivial. The only exception being internal thoughts, provided there isn't some context preventing that.

    5. Players have the right to author facts of the world as long as it is constrained by circumstantial rules or tied up to a resource system.

    I think I need some ideas on #4 when I get into thinking about how to build something more practical from this. Like, how to give the player some power to apply self-judgement based on context without completely leaving the GM behind. Would common sense judgement be enough, based on an honor system? Or would something more meaty be more beneficial (and maybe more fun)? That's kind of interesting to think about, IMO.
  • @Paul_T and @AsIf:

    The whole advanced AI topic is super interesting. As I've been thinking about all this, it seems to me that having a real human in front of you, reacting to you, is a significant portion of the fun for some people. So, would knowing that they're dealing with an artificial personality result in a decrease in fun for them? Meaning, if those people didn't view an AI as truly sentient (see chinese box experiment), but as a mechanism so complex that it is essentially a much better illusion, or much better version of "pushback", than what we now have with paper "oracles", would they still have as much fun with them as with another person?

    In my the Un-Solo RPG G+ I created recently, I'm trying to test just that (haven't had too many takers so far). I'm basically restricting "GMs" to giving me the output in the same format a p&p oracle would give me, without letting me know if they're making conscious answers or acting as a proxy to a p&p oracle.

    So far, the experience hasn't really felt much different than playing solo, since I'm assuming the same responsibilities as when I do it solo-- yet I am arguably playing a social roleplaying game. I think the MMV from person to person. I'm waiting for more people to jump in as Players with new games of their own. I think GM'ing them in this fashion is very low investment, and maybe fun to pass the time on the train: "yes, no, yes." Or when answering complex questions, picking two keywords from the Oracle's list to try and influence the Player in a broad directio (i.e. "What do I see?" Answer: "Solve, Evil").
  • edited August 2014
    So with those provisional rules in place, I think that we have a complete setup. I specified "in writing only" before to make it match what appears to be the typical solo output method during play. It certainly matches mine.

    However, that rule seems to have acted as a red herring for people in other forums, so I'll leave it out. I think writing can be left off for when it's needed to explore the output method itself. For now, it can be assumed the player and GM can verbalize their output.

    In the case of multiple players the rule is that they can't be in the same room and must communicate through the GM who will still have the same communication restrictions. All information can be gathered in that fashion, so that the information concerns other players makes no difference. If any disagreements on that point, do let me know, but for now, rule #6:

    6. If there are multiple players/PCs, they must be in separate rooms and must communicate through the GM. All GM communication restrictions apply.

    With that, I guess now I can pose the question, is this social activity, in this particular form, equivalent to playing an RPG?

    IMO, it is. Almost certainly not satisfying to normal humans, but that's an issue separate from the question (I can make arguments for that, but I think they're pretty obvious).

    The next step, provided a person answered the previous question in the affirmative is: if the GM, unbeknownst to the player rolls a d100 dice before answering, assigning 'yes' to anything <= 50, and 'no' to any result >= 51...is this still equivalent to playing an RPG? Why? Also, from the players' perspective, are they playing an RPG?

    My answers would be 'yes' and 'yes', because if the player cannot tell the difference between the two, they're effectively causing the same states of mind in them. Everything outside of that becomes a semantical or definitional argument, IMO.

    Assuming we are somehow on the same page regarding this conclusion, I think the next step would be to play around with rules #4 and #5. I think that within their spectrum, there are possible manifestations that could potentially make the activity no longer fit the definition an RPG. For example, the extreme end of player freedom would mean that the player no longer has to engage the GM at all. If they don't engage the GM, do they get any pushback at all? If not, are they roleplaying? What if the pushback was only limited to what strict logic would dictate? Would that be enough to call something roleplaying?

    P.S. I wanted to give more links to either actual play involving the emulation of PCs, or links where folks were talking about having done it. Except for one link, all of these are in the Mythic GME yahoo group which is private, though.


    These were the ones I was thinking about at first, but couldn't find them:

    http://monicatbox.blogspot.com/2010/01/gming-with-mythic-player-emulation.html
    https://groups.yahoo.com/neo/groups/Mythic_Role_Playing/conversations/messages/2902
    https://groups.yahoo.com/neo/groups/Mythic_Role_Playing/conversations/messages/3247

    The rest of these are more like kibitzing about doing this kind of gaming session:

    https://groups.yahoo.com/neo/groups/Mythic_Role_Playing/conversations/messages/3222
    https://groups.yahoo.com/neo/groups/Mythic_Role_Playing/conversations/messages/2704
    https://groups.yahoo.com/neo/groups/Mythic_Role_Playing/conversations/messages/2930

  • Ok, this is evolving, which is fun.

    My take so far is that *of course* playing with a "real" GM who is pretending to be a machine emulator is not all that different from playing with a machine emulator. As you point out, it would "feel" different if the actual person was sitting in front of you (and, for instance, rolling dice). It would feel more human.

    But, as best as I can imagine, it still wouldn't feel like playing in RPG, in the sense of what "Paul likes in RPGs". Which I think you agree with, given your statement below:
    Almost certainly not satisfying to normal humans [...]
    Still an interesting discussion, of course, and I applaud you for pursuing it so thoroughly.

    I wonder if you're familiar with Vincent Baker's theory/concept of 'the unwelcome'? Does that ring a bell for you at all?

    He considers it an issue of game design (which I don't disagree with), but I think in this context we can see that it also operates based on live, human input, and therefore is a feature of human play, as well. (I suppose a sufficiently-advanced computer could provide this, but would it be aesthetically and personally satisfying?)

    If my comments have no bearing on what you're looking for, here, then don't mind me!

  • If my comments have no bearing on what you're looking for, here, then don't mind me!
    Au contraire, man! You, AsIf and Adam have been very helpful. And some of the ideas you folks brought up are great. :)

    So, part of the reason for the thought experiment is that I wanted to explore why solo RPGs might not be RPGs at all beyond seemingly superficial reasons like "RPGs are social". So, there, we now have a social activity much like "solo RPGs" and if it's not an RPG, then we can explore the reasons (this may also help the folks who have that "just writing with dice" feeling sort out why)

    For example, if this social activity is not playing an RPG, then things that can be explored are:

    1. How much authoring power can you take away from a GM role before the activity stops being an RPG?
    2. Corollary: How much authoring power can you give the Player before the activity stops being an RPG?
    3. Likewise, imagining we wanted to get rid of the GM's implied 'yes', how much power to judge whether a situation requires a roll can you give a player before it stops being an RPG? How do you draw the line? Would that be a desireable thing at all?

    They're interesting questions to me. May have a practical value too. I feel it in my bones. :)

    One thing I've taken away as well is something that I actually read in this forum (I think), that the activity we call solo roleplaying is actually closer to GM prep than anything. It also makes me question, how much does GM prep differ from "just writing"? Does it remain just writing until you're able to use that prep in a game?

    In the same vein, your question about flipping the roles in the experiment by taking on the role of the GM, now makes me wonder whether the flipside is inherently more fun and doable. After all, as a player in this sort of activity, you're taking on GM duties to the max, so why not go all the way? It may actually be less onerous to roll for each player's decision, or at least narrow it to the most logical. :)

    By the way, while playing on the experimental forum, I've not really noticed much of a difference between that and some modes of solo play. Still, knowing that there's a person playing with me, did give me a sense of urgency not to leave them hanging. YMMV, I guess!
  • Interesting! Thanks.
  • Role-playing (NOUN) 1chiefly Psychology the acting out or performance of a particular role, either consciously (as a technique in psychotherapy or training) or unconsciously, in accordance with the perceived expectations of society with regard to a person's behavior in a particular context. 2 participation in a role-playing game.

    It is not just rolling dice to write out a story. I take on a role (a character), or rather many roles, to entertain myself. If I happen to write out the results later as a story, so be it. Many people tell me that role-playing is a social activity. In my opinion you don't need others to role-play, just an oracle which drives the story and your imagination (and some randomizers lists, cards, dice, & tables help also). I have had plenty of fun filled sessions and was surprised through out them to the very end.

    In the final outcome it does not matter who or what other people's opinions are of your role-playing, solo or group. If you are having fun then you are not role-playing wrong. If it is a chore or writing exercise that bores you, go play a video game, watch a movie, read a book, or join a group of people to role-play.
  • hi @Virginian_John!

    It's not about caring about other's opinions, but more like wanting to see reasons that go beyond "RPGs must be social", both because it's interesting to me to work through the arguments and because knowing (for me) what they are closer to in kind will help me find ways to enjoy it better.

    I have now fully accepted that solo RPG as it's currently done, outside of the random dungeon delving/hexcrawling or CYOA experience, is most closely related to the GM side of things that the player's. The amount of content creation required of the solo gamer is what makes it so, IMO. I can't wait for sophisticated AI, but now I know where to look for parallels when looking for new ideas in solo rpg gaming.
  • edited August 2014
    @Dreamer, +1.

    @Virginian_John, totally yes, but I just want to point out that if you roleplay in public, you don't even need an oracle! Other people will provide pushback, and can be used to generate ideas. For instance, try acting like a detective, or a beggar, or a cold-as-ice assassin. Walk around playing that role. :-)
  • So, I think that other than topics I “bookmarked” for later, or variations of the experiment, I’m pretty much done with the main thought experiment. The final form is as follows:


    First the assumptions:

    1. Assume the players have unlimited patience, unlimited time, unlimited lifespan. Or, alternatively, that they have the ability to think and communicate faster than the speed of light or something.
    2. Anything not explicitly prohibited, is permitted.
    3. Rules might be adjusted in response to #2.
    4. For now, assume just one GM and one player. (If more than one player, my intuition is that they must communicate through the GM using the yes/no rule.)

    The yes/no rule restated and expanded:

    1. The GM can only answer yes or no to anything the players ask.
    2. Corollary: The players can only get information out of the GM by the use of any type of questions that can be answered by a yes or no. The cororally means that the players can very well write a long statement and ask at the end, "Is this true?"
    3. Any idea the GM has must satisfy at least one of the following conditions:
    3.a. It can be explained in verbal terms the player can understand
    3.b. It can be extrapolated from other ideas that can be put in verbal terms the player can understand
    3.c. It can be represented non-verbally in a way the player can access and can show the GM.
    3.d. It has the possibility of being perceived directly in some way by the player, who will then be able to put it in terms of #1,#2, or #3
    3.e. It has the possibility of being approximated by another idea that satisfies #3 or #4, which then can be put in terms of #3.

    4. Players must ask the GM if they need to roll for any action their PC takes no matter how trivial. The only exception being internal thoughts, provided there isn't some context preventing that.
    5. Players have the right to author facts of the world as long as it is constrained by circumstantial rules or tied up to a resource system.
    6. If there are multiple players/PCs, they must be in separate rooms and must communicate through the GM. All GM communication restrictions apply.

    The variations on it are basically playing around with the limits on (I forget if I touched upon them, so I’ll just summarize):

    1. How much authority to make a judgement on whether a roll is required do players have?
    2. How much narration/authoring authority do you give a player? Do you constrain it? If so, how much?
    3. Variations on the output method: written, verbal, etc (?)


    Related to these are these questions I already asked:

    1. How much authoring power can you take away from a GM role before the activity stops being an RPG?
    2. Corollary: How much authoring power can you give the Player before the activity stops being an RPG?
    3. Likewise, imagining we wanted to get rid of the GM's implied 'yes', how much power to judge whether a situation requires a roll can you give a player before it stops being an RPG? How do you draw the line? Would that be a desireable thing at all?

    I think it’d be interesting to hear people’s opinions on the last three. I wonder if there is a consensus, and whether fun RPG games can be made that play against those limits (whatever the consensus is).

    P.S. As I’ve already mentioned, I now view “solo role playing” as a GM-like activity, so I already have some ideas floating in my head. I have the skeleton of one possible game outlined. I don’t think it’s anything original, though, the combination of things *maybe* could be. If anything, it could be called a storywriting game. I don’t know if it fits within the scope of the site, but I’ll post it up for feedback.
Sign In or Register to comment.