Why does the party need to be balanced?

[ezekielraiden]

I'm not sure whether something is a game under a particular formal sense is all that relevant, at least to my goals. If my design goal is better satisfied by a non-game, that's not actually a reason to abandon either that direction of design, or the goal.
Okay. I appreciate that thread drift and stuff happens, but I guess I'm coming at this from, "How is this relevant to the thread topic?" We already answered the original question--"need" is too strong a word without any context--and had, as I understood it, moved on to the more useful, "What does 'balance' mean for the D&D context, and how necessary is it there?" Talking about abstract goals that may go entirely outside of gaming and game design isn't bad, but it seems a spinoff topic, at least to me.


Puzzle games may not be game theoretic games, but that doesn't prevent people from playing them...
Well, I'd argue that a "puzzle game" always has one of two features: either it is a puzzle generator, and thus open-ended rather than closed-ended like a proper singular puzzle is, or it is a significantly long puzzle sequence, so that it comes more closely to resemble the IPD (that is, play is about learning to solve puzzles quickly, or about building up a skillset to identify key information, etc.) In other words, anything that a typical person would call a "puzzle game" needs that iterative component; otherwise it would just be a stand-alone puzzle.


I can definitely see 'I want to design a group activity, the experience of which tries to be as rich and varied as real life while yet being distinct from it' as an interesting and worthy goal for a tabletop RPG. Even if the result ends up not technically being a game.
Well, I mean, then it's not really a goal for a tabletop RPG, which was sort of my point. I'm interested in talking about games, since this is the Roleplaying Games forum, and our focus topic is balance (with the more-refined "what is balance, and how much is worthwhile?" question).


In the context of your Prisoner's Dilemma vs IPD example, I would say that my description of Limit Break for example (the overpowered superhero game) is like me describing a single round of PD. We can agree on the events that happened and maybe even what it was like, but the next round - another campaign in that system with different players - is highly likely to be extremely different. That's because the system is intentionally incomplete in a way that completes itself using the psychology and interests of the players. One could strategize in the abstract, but without knowing how the other players are going to play those strategies would have a very short half life.

In the game I ran, every player chose highly abstract power themes with (conflicting) cosmological implications. Gameplay often revolved around literally editing the meanings of concepts in order to undermine crowd behaviors that had gained sentience.

Next game, someone could decide they want time travel powers, and whatever conclusions we drew might go up in flames.

The concept of the system itself doesn't even really work if the same player plays twice, because advancement is at least in large part player-side. It's like taking the same IQ test twice.
Then...yeah I guess I just don't consider that a game. If you play it twice and necessarily generate effectively identical outcomes....it's not a game. More like a really entertaining personality test or self-introspection questionnaire? Regardless, though, while it is a worthy subject of discussion, I just...don't really have anything to say about it in the context of "Roleplaying Games," balance, and TTRPGs in general.

I'd also argue that that's really not very life-like--a second go-round at life should at least in theory look pretty different, right? That's why belief systems with reincarnation don't expect you to be a carbon-copy of the person you were before, just similar. (Frex, the Dalai Lama.)

Edit:
To give a different example of something I consider extremely worthwhile, that is similar to all this, but that I am not interested in discussing here and now: Writing exercises. Have you ever used the "Exquisite Corpse (https://en.wikipedia.org/wiki/Exquisite_corpse)" writing prompt method? Some might call it a "game," but it would be rather pointless to try to come up with "strategy" for it--the whole point is to break anticipated patterns and force unconventional, unexpected thinking through random association or "telephone"-like lack of symmetry. Either the word-sequence is in a fixed pattern (like "The <adjective> <noun> <adverb> <present-tense verb> a/n <adjective> <noun>," so that a guaranteed-grammatical but possibly very surreal sentence results), OR you can only see the previous word (thus connotations/expectations can change sharply from one word to the next). Either way, the whole intent is to generate surprising things, and the only "victory" condition is being inspired to write more formally as a result.

I think it's wonderful that we have this method, and that it's always fruitful to discuss this and other koan-like "break your chains" methods...just not in a discussion about balance in RPG design.

[NichG]

Okay. I appreciate that thread drift and stuff happens, but I guess I'm coming at this from, "How is this relevant to the thread topic?" We already answered the original question--"need" is too strong a word without any context--and had, as I understood it, moved on to the more useful, "What does 'balance' mean for the D&D context, and how necessary is it there?" Talking about abstract goals that may go entirely outside of gaming and game design isn't bad, but it seems a spinoff topic, at least to me.


It's relevant to understanding the motivations of people who engage with tabletop RPGs, D&D included. I think a lot of the circular conflict in this thread between e.g. you and Quertus is that his motivations for engaging in this activity have little to do with the question of whether it's formally a game. So when e.g. you define balance as a game being converged in its mathematically testable design goals, if it turns out that that framework of discussion totally misses out on a set of goals that players or DMs might have when engaging with the activity, it's going to present a blind spot that prevents any kind of mutual understanding or convergence.

The point of this spinoff is that I'm trying to make the boundaries of that blind spot clearer by way of examples that more clearly belong to it, but otherwise have aspects in common with the kinds of design problems that one might go about solving in order to ensure that one's group has a good time (whether they're technically playing a game or not).

If there's a set of activities that people engage in in the context of tabletop RPGs, which tabletop RPGs can be designed to enhance or bring about, and which comprise a portion of the goals of at least a subset of players, then a framework which is forced to exclude them on the basis of definitional boundaries isn't going to be the end-all be-all of discussion. While 'balance' can be defined within that framework, if you can recognize groups who want things but simultaneously recognize that those things are incoherent within the framework, the hope is that seeing that contrast might make it clearer why some groups or designers could be better served by rejecting (at least that sense of) balance as a design goal.

So as a meta-goal for the conversation, my success conditions are: A lesser success would be that the claims which previously seemed incoherent to you now seem consistent with at least some set of assumptions, even if they're different than the ones you'd bring to a design problem. A greater success would be that you'd find a way to extend how you're thinking about games and balance such that it becomes possible to consider this sort of experientially-focused design, such that you could e.g. in the future suggest subsystems and alternatives to people who have those things as their goal without the suggestions falling flat or being rejected.

As an example point, you've described the 4ed skill challenges system as - to your design goals - a good example of balanced design. However, to this sort of goal of providing transformative experiences that need to be lived to really have their full effect, the 4ed skill challenge system has a problem - it takes a set of disparate contexts or situations and places them under the same mechanics, effectively making subsequent engagements with the skill system more similar to each-other than they would be otherwise. I think if I said that at the start of the conversation, it'd be hard to even get across why one would want such a thing. After this spinoff conversation, I would hope that it would at least be raised to the level of 'I wouldn't personally want that, but if you want that I am able to understand the conditions that would have to be satisfied to provide it'. And, again, at a greater level of success, for you to be able to conclude 'I see why X approach to balance conflicts with that goal, but here's Y approach to balance that would still be compatible' for example.

(And, as a meta-meta thing, I think this line of conversation has made more forward progress to some kind of mutual understanding and even potentially actionable outcomes than the previous debate about 'is D&D 3.5ed a bad game?' which has been done to death on these forums and won't actually change what anyone runs or plays, so I'm personally having more fun discussing this).



Well, I mean, then it's not really a goal for a tabletop RPG, which was sort of my point. I'm interested in talking about games, since this is the Roleplaying Games forum, and our focus topic is balance (with the more-refined "what is balance, and how much is worthwhile?" question).


I think my players would probably be fairly put off if they were told that they didn't spend the last 12 months playing a tabletop RPG and could not discuss it here because there were aspects of the activity that put stress on the formal game theoretic sense of what a 'game' is... If one were to sit in on a session, it would be hard to say that its not some kind of tabletop gaming activity, even with internal strategies and consistent outcomes and stuff like that; its just that if you were to take the rules and run it yourself, it'd almost certainly end up being a different experience in the end, and copying over those internal strategies wouldn't work very well because they'd be highly tuned to the particular context of those players.



Then...yeah I guess I just don't consider that a game. If you play it twice and necessarily generate effectively identical outcomes....it's not a game. More like a really entertaining personality test or self-introspection questionnaire? Regardless, though, while it is a worthy subject of discussion, I just...don't really have anything to say about it in the context of "Roleplaying Games," balance, and TTRPGs in general.

I'd also argue that that's really not very life-like--a second go-round at life should at least in theory look pretty different, right? That's why belief systems with reincarnation don't expect you to be a carbon-copy of the person you were before, just similar. (Frex, the Dalai Lama.)


I think that's in agreement with what I'm saying? That basically, if you live in a world where everyone plays 'life' once, then life works a certain way. Take that same world, but let everyone play life N times (and remember the past experiences), and the world as a whole would look wildly different to the extent that at least some of the strategies or insights which would apply to the non-reincarnation world wouldn't apply very well to the reincarnation world.

Play Limit Break once and you get a game (or whatever we want to call it) about progressing past your own psychological boundaries. Play it a second time and your (and all other players') psychological boundaries have already been progressed past. I could go into details, but maybe a simple approximation would be easier to talk about concretely. Imagine the game: 'two players each pick a finite real number of their choice - higher wins'. A good strategy for that game might be to dive for named numbers in math journals, but once you've done it once, that strategy is burned and can't be used with the same opponent again. A decent meta-strategy might be that you want to escalate as slowly as possible while still maintaining a lead, but even that will only take you so far.



Edit:
To give a different example of something I consider extremely worthwhile, that is similar to all this, but that I am not interested in discussing here and now: Writing exercises. Have you ever used the "Exquisite Corpse (https://en.wikipedia.org/wiki/Exquisite_corpse)" writing prompt method? Some might call it a "game," but it would be rather pointless to try to come up with "strategy" for it--the whole point is to break anticipated patterns and force unconventional, unexpected thinking through random association or "telephone"-like lack of symmetry. Either the word-sequence is in a fixed pattern (like "The <adjective> <noun> <adverb> <present-tense verb> a/n <adjective> <noun>," so that a guaranteed-grammatical but possibly very surreal sentence results), OR you can only see the previous word (thus connotations/expectations can change sharply from one word to the next). Either way, the whole intent is to generate surprising things, and the only "victory" condition is being inspired to write more formally as a result.

I think it's wonderful that we have this method, and that it's always fruitful to discuss this and other koan-like "break your chains" methods...just not in a discussion about balance in RPG design.

I haven't used that, but it sounds interesting. I guess I'd say that a large part of my reason to engage in tabletop RPGs at all is to find those kinds of 'break your chains' moments in things that can only be understood through experience (because having other elements that are not-yourself involved means that you can get shoved in directions you weren't able to see just by pure thought).

[ezekielraiden]

It's relevant to understanding the motivations of people who engage with tabletop RPGs, D&D included. I think a lot of the circular conflict in this thread between e.g. you and Quertus is that his motivations for engaging in this activity have little to do with the question of whether it's formally a game.
But...okay. Let me be as direct as possible.

How can we discuss things that, as you yourself said, cannot be discussed?

That's my whole point here. I'm not constantly circling back around this because the play experience has to be mathematical. I've given examples where the test is mathematical, but the game is not--your very own Nomic example, where there is a clearly mathematically-testable aspect to it, based on survey data about what people want to do when exposed to the rules. The designers of Nomic very clearly wanted a particular, and communicatable, discussable, experience when playing the Nomic game, and designed it in such a way to naturally trigger that experience. Thus, the most fully "un-game game" example you could give besides Limit Break (which...I honestly can't even find anywhere, so do you have a link?) was still perfectly well described by what I'm talking about.

That's why I'm so deeply confused about this idea that there is this enormous swathe of gaming/design-goals for which ANY form of discussion, ANY form of analysis, ANY form of strategy, is COMPLETELY alien.


The point of this spinoff is that I'm trying to make the boundaries of that blind spot clearer by way of examples that more clearly belong to it, but otherwise have aspects in common with the kinds of design problems that one might go about solving in order to ensure that one's group has a good time (whether they're technically playing a game or not).
I don't actually SEE it as a blind spot though! That's been my whole point. I don't see how design considerations don't apply to Nomic, and I feel like if I could actually look at the rules for Limit Break, I could point out exactly the kind of thing I described above, where there are still statistical things you can do to test if what you're doing is achieving the goals you set out for. Some mechanics are themselves mathematical, and thus admit direct analysis. Other mechanics are not mathematical in themselves, but we can collect data about user response and overall experience in such a way as to learn whether the players at least believe they are having the experience you want them to have.

So as a meta-goal for the conversation, my success conditions are: A lesser success would be that the claims which previously seemed incoherent to you now seem consistent with at least some set of assumptions, even if they're different than the ones you'd bring to a design problem. A greater success would be that you'd find a way to extend how you're thinking about games and balance such that it becomes possible to consider this sort of experientially-focused design, such that you could e.g. in the future suggest subsystems and alternatives to people who have those things as their goal without the suggestions falling flat or being rejected.


As an example point, you've described the 4ed skill challenges system as - to your design goals - a good example of balanced design.
Well, I'd call them an okay example. They required some tweaking after launch, and you kinda needed to "play beyond the rules" (not contradicting any of them, but setting up further rules and ideas than what the strict limits of RAW included). So: balanced, but only okay balance. They could admit a significant amount of improvement by asking carefully-designed survey questions about the experience itself (mostly because, as I said, I have had multiple DMs who "got" the rules, even ones theoretically "brand new" to 4e, and thus saw the Better Thing beyond the system-as-it-was.)


However, to this sort of goal of providing transformative experiences that need to be lived to really have their full effect, the 4ed skill challenge system has a problem - it takes a set of disparate contexts or situations and places them under the same mechanics, effectively making subsequent engagements with the skill system more similar to each-other than they would be otherwise. I think if I said that at the start of the conversation, it'd be hard to even get across why one would want such a thing.
I don't even understand what the phrase means. What is a "transformative experience"? How on earth can mere procedure create such "transformation"? That would be like having a procedure that, of necessity, simply by following its motions, one could achieve Buddhist enlightenment. The whole point of something like that is that it is in defiance of procedure. But the only constituent OF rules is procedure.


(And, as a meta-meta thing, I think this line of conversation has made more forward progress to some kind of mutual understanding and even potentially actionable outcomes than the previous debate about 'is D&D 3.5ed a bad game?' which has been done to death on these forums and won't actually change what anyone runs or plays, so I'm personally having more fun discussing this).
Personally I think the answer is quite simple. "Bad game" means different things to different people, hence one must provide one's definition first. If the standard is "did people have fun with it? if not, then it's a bad game," then 3.5e is just about the antithesis of a bad game, because it's historical fact that an enormous number of people had a huge amount of fun with it. But the standard is rather loose, as one can argue literally any game enjoyed by more than one person can't be a "bad game." My personal preference is: "does the game fulfill the goals its designers communicated to its players? if not, then it's a bad game." And by that standard, there's a lot of room to argue that 3.5e did badly--after all, not only its own creators, but the people who took up the mantle to keep it alive, have both admitted that in order to fulfill their goals better, they had to leave that system behind.


I think my players would probably be fairly put off if they were told that they didn't spend the last 12 months playing a tabletop RPG and could not discuss it here because there were aspects of the activity that put stress on the formal game theoretic sense of what a 'game' is...
Except I'm not the one claiming that. You did:

Strong sense of alienation when a player transfers from one table running the game to another. Inability for players to accurately describe their experiences with the game, or to agree with eachother about the nature of their experiences.
You have, repeatedly, referenced a game that cannot even be discussed. And that's why I have been pushing against this thing being, in any meaningful sense, "a game." If you cannot talk about it, how did someone tell you how to play it? You're describing a ruleset that somehow means two players cannot accurately describe to each other how it felt to play.

You have described--at least in high-concept terms--how it feels to play Limit Break. We are already discussing it now. You are not at a loss for words. Ergo, Limit Break is not one of the things I'm saying cannot be a game. If it can be discussed, if you can ask people questions about their experiences so that you can have even a hope of shaping that experience (even if the shaping you choose to seek is "not shaping it at all"), then it admits design, and can be a game.


If one were to sit in on a session, it would be hard to say that its not some kind of tabletop gaming activity, even with internal strategies and consistent outcomes and stuff like that; its just that if you were to take the rules and run it yourself, it'd almost certainly end up being a different experience in the end, and copying over those internal strategies wouldn't work very well because they'd be highly tuned to the particular context of those players.
Okay so...what does that mean? I really, truly do not understand what this part means. I honestly think you're taking an overly-restrictive definition of "strategy," here, since (as I said earlier) I am allowing that even Nomic admits strategy, it's just going to look extremely different from other games' strategy, because it relies on a different set of skills than most games (mostly creativity and spontaneity, as opposed to within-the-rules analysis and calculation). "Strategy" is not limited to within-the-rules analysis of numerical values; it embraces a much wider variety of things. "Design" is not limited to setting probabilities for an internal random number generator; it too embraces a much wider variety of things--in short, anything you can gather data about, which is a very, very large set of things. (I am not a hyper-empiricist, I genuinely believe there are questions for which the answer cannot be found solely through measurement and counting, but I am skeptical of the notion that such questions are that relevant to game design.)


I think that's in agreement with what I'm saying? That basically, if you live in a world where everyone plays 'life' once, then life works a certain way. Take that same world, but let everyone play life N times (and remember the past experiences), and the world as a whole would look wildly different to the extent that at least some of the strategies or insights which would apply to the non-reincarnation world wouldn't apply very well to the reincarnation world.
Okay, but are you therefore asserting that there is no strategy at all? That there is no way, even in principle, for two players to sit down and compare relevant parts of their experiences, and to plan for future concerns? Because that's the kind of "game" you described above. One where trying to live your 7th "life" with a different group of people is so alien, you're not sure you're still playing "life."


Play Limit Break once and you get a game (or whatever we want to call it) about progressing past your own psychological boundaries. Play it a second time and your (and all other players') psychological boundaries have already been progressed past. I could go into details, but maybe a simple approximation would be easier to talk about concretely. Imagine the game: 'two players each pick a finite real number of their choice - higher wins'. A good strategy for that game might be to dive for named numbers in math journals, but once you've done it once, that strategy is burned and can't be used with the same opponent again. A decent meta-strategy might be that you want to escalate as slowly as possible while still maintaining a lead, but even that will only take you so far.
But such a game still admits discussion. Playing it with a different partner doesn't result in "alienation" from one's former experiences, nor would such new experiences with a new player be alien to an old hand. It admits a fair amount of analysis (that is, uh, kind of what led to Cantor's diagonal argument, and I'd argue that "number theory before Cantor" is a pretty substantial body of analysis). Even if you ignore the purely mathematical parts, leaving just the psychology of it, there's still a lot to learn from things like "what would the average joe think of as the largest number?" or "what's the first '-illion' term that most people don't know?" (e.g. most people know "trillion," a lot of people probably even know "quintillion," but how many people know nonillion?) All of this is under the umbrella of strategy and discussion. It's amenable to design (though which such an intentionally-reduced example, I admit there aren't many avenues of design available without changing the nature of the game somehow.)


I haven't used that, but it sounds interesting. I guess I'd say that a large part of my reason to engage in tabletop RPGs at all is to find those kinds of 'break your chains' moments in things that can only be understood through experience (because having other elements that are not-yourself involved means that you can get shoved in directions you weren't able to see just by pure thought).
Okay but...doesn't that mean you see potential design even in that? You have just identified a discussion-capable, can-be-tested element: by including one or more sources of non-self-generated notions, do people get an experience of...well, you'd probably want to think about exactly how you phrase it. "Change in perspective," "broken (mental) chains," "epiphany," etc. That's still a statistical test, based on gathering user input after play. Even if the actual experience itself is always inherently unique, the concept of "epiphany" is not so anti-transferrable that we can't talk about it. That would be like saying that we can't talk about the flavor of bananas or apples just because I can't beam into your head what apples taste like to me--and, further, that apple pie recipes are impossible to design because any two people will always taste the pie a bit differently. (Of course, in food science, you're best served by catering to data clusters rather than finding a single data center, but that's going further into the statistics and psychophysics of it than I think befits the thread thus far.)

[NichG]

I'm going to try to consolidate a bit since some points are raised in multiple places in the post.



Thus, the most fully "un-game game" example you could give besides Limit Break (which...I honestly can't even find anywhere, so do you have a link?) was still perfectly well described by what I'm talking about.

...



I thought I had mentioned it earlier in the thread, but my general MO is to write a custom system for any campaign I intend to run, aiming at some particular experience or question. Limit Break is one of those. If you want to check the rules, they're here: https://drive.google.com/open?id=1pDnZ5NnSUeRwQEOuEjxMFGmT4isb9iGI

I am claiming a different design methodology, so I should explicitly state it. The design methodology basically amounts to trying to find things that I myself have difficulty understanding or anticipating, then writing the system in such a way that gameplay will center around those things and force them to be resolved. Alternately, I'll try to target some kind of standard belief or assumption which would be easy to just take as true, but then try to find a system that falsifies that belief.

With Limit Break, the thing I didn't understand which was used as a seed was the underlying motivational structure of the superhero genre - not just 'what happens' in superhero media, but ultimately why it happens and what it means. In particular, a common occurrence in campaigns I've run is that the PCs cross a threshold where things that once mattered to them become irrelevant (be it their home country, other NPCs, recurring villains, etc) unless those things also escalate. In the superhero genre, generally characters do big cosmic deeds motivated primarily by personal connections, and those connections are maintained in the spotlight. If there was something real to that, I could try to design a system around it to query that conceit, and I'd be likely to learn something in the process.

The assumption I targeted was that character power comes from mechanical options, and that sub-parts of a system with more powerful options lead to more powerful characters and vice versa. So the game was constructed to basically make any character capable of things far above the power level of pretty much any other tabletop RPG other than Nobilis - however, I wouldn't tell them too explicitly that this was the case. The rule in question was: 'as long as you can justify it to yourself in your character's theme, and as long as no one who would be directly affected by it opposes it, it automatically works' combined with 'you can at any point create a Lv0 power for free'. So a starting character with the power 'leadership' for example could create a universe in which they are worshipped by the populace and lead armies of minions out from the portal, or other such over the top things. This was the gamble on my part, the intentionally designed-in uncertainty - would the game fall apart, or would it actually work despite that?

And it ended up actually working, with a number of moments during play that were quite informative to me. So, I can say at least within the framework of my design methodology and goals, it was a successful campaign despite not being designed specifically with a mind towards balance (other than trying to make sure that there were enough wildcards in the system that it would be hard to obtain any guarantees).



But...okay. Let me be as direct as possible.

How can we discuss things that, as you yourself said, cannot be discussed?

That's my whole point here. I'm not constantly circling back around this because the play experience has to be mathematical. I've given examples where the test is mathematical, but the game is not--your very own Nomic example, where there is a clearly mathematically-testable aspect to it, based on survey data about what people want to do when exposed to the rules. The designers of Nomic very clearly wanted a particular, and communicatable, discussable, experience when playing the Nomic game, and designed it in such a way to naturally trigger that experience.

...

That's why I'm so deeply confused about this idea that there is this enormous swathe of gaming/design-goals for which ANY form of discussion, ANY form of analysis, ANY form of strategy, is COMPLETELY alien.

I don't actually SEE it as a blind spot though! That's been my whole point. I don't see how design considerations don't apply to Nomic, and I feel like if I could actually look at the rules for Limit Break, I could point out exactly the kind of thing I described above, where there are still statistical things you can do to test if what you're doing is achieving the goals you set out for. Some mechanics are themselves mathematical, and thus admit direct analysis. Other mechanics are not mathematical in themselves, but we can collect data about user response and overall experience in such a way as to learn whether the players at least believe they are having the experience you want them to have.

...

Except I'm not the one claiming that. You did:

You have, repeatedly, referenced a game that cannot even be discussed. And that's why I have been pushing against this thing being, in any meaningful sense, "a game." If you cannot talk about it, how did someone tell you how to play it? You're describing a ruleset that somehow means two players cannot accurately describe to each other how it felt to play.

...

Okay but...doesn't that mean you see potential design even in that? You have just identified a discussion-capable, can-be-tested element: by including one or more sources of non-self-generated notions, do people get an experience of...well, you'd probably want to think about exactly how you phrase it. "Change in perspective," "broken (mental) chains," "epiphany," etc. That's still a statistical test, based on gathering user input after play. Even if the actual experience itself is always inherently unique, the concept of "epiphany" is not so anti-transferrable that we can't talk about it. That would be like saying that we can't talk about the flavor of bananas or apples just because I can't beam into your head what apples taste like to me--and, further, that apple pie recipes are impossible to design because any two people will always taste the pie a bit differently. (Of course, in food science, you're best served by catering to data clusters rather than finding a single data center, but that's going further into the statistics and psychophysics of it than I think befits the thread thus far.)

...

But such a game still admits discussion. Playing it with a different partner doesn't result in "alienation" from one's former experiences, nor would such new experiences with a new player be alien to an old hand. It admits a fair amount of analysis (that is, uh, kind of what led to Cantor's diagonal argument, and I'd argue that "number theory before Cantor" is a pretty substantial body of analysis). Even if you ignore the purely mathematical parts, leaving just the psychology of it, there's still a lot to learn from things like "what would the average joe think of as the largest number?" or "what's the first '-illion' term that most people don't know?" (e.g. most people know "trillion," a lot of people probably even know "quintillion," but how many people know nonillion?) All of this is under the umbrella of strategy and discussion. It's amenable to design (though which such an intentionally-reduced example, I admit there aren't many avenues of design available without changing the nature of the game somehow.)


For all of this, I'm going to go back to a point I raised earlier, and ask to be very specific about my wording here. When specifying those goals, I'm specifying a direction to move towards, not an end point which we must be at. Similarly, I'm not specifying that something be impossible to discuss, but just that discussion should not suffice to capture the experience of the thing and that players who have played at different tables end up finding that they disagree about what playing the game is actually like. This is not the same as saying that the game is impossible to discuss.

Also, whether this set of things means that design is impossible is a point of disagreement between us, so while this is incoherent from your definition of design, it isn't actually incoherent from my definition of design. That is to say, I am not at all claiming these things cannot be designed - I'm in fact claiming that designing them is not at all a problem, and that doing so is relatively easy in the design methodology I use (which is not based first and foremost on mathematical testing).

Since you claimed that these goals are incoherent, I'm taking that as you holding the position that these goals are not subject to the design methodology you are forwarding. If you've changed your mind on that and have decided that they aren't incoherent, then we can go from that point. If they feel simultaneously designable and incoherent to you, then that should seem paradoxical to you.

(As far as alienation in the real number game, if I just had a game where we started with Graham's number and went from there, and the next game my opponent said 'a trillion!', or worse handed me a terabyte harddrive filled with code that outputs a 1, executes Busy Beaver of 10^12, and outputs a zero at every step then I'd feel like it's not the same game, even though it has the same rules.)



I don't even understand what the phrase means. What is a "transformative experience"? How on earth can mere procedure create such "transformation"? That would be like having a procedure that, of necessity, simply by following its motions, one could achieve Buddhist enlightenment. The whole point of something like that is that it is in defiance of procedure. But the only constituent OF rules is procedure.


At least within how I'm thinking about these things, there isn't a contradiction here. Seeing a movie for the first time can be a transformative experience. Hearing music for the first time can be a transformative experience. Learning to actually do the math of quantum mechanics can be a transformative experience. Even playing tabletop RPGs for the first time (and the corresponding Stockholm syndrome people have with D&D) can be a transformative experience.
It's not supernatural or mystical, just trying to be the kind of thing that people remember for a very long time because it anchors some new way of thinking that they can engage in. Things like 'eureka!' moments.

At a meta level, generally things which force your brain to operate in new ways have at least the potential for qualifying. They're going to generally be somewhat personal though, so it's harder to hit the target if you don't know the players ahead of time. Recognize assumptions or modes of operation favored by the player, create a scenario in which those things are under tension, allow the player to work out that tension themselves via the vehicle of the game. I'm also not going to say I have a 100% hit rate on this kind of thing, but even if it's say 10% per campaign per player then I think that's still worth pursuing.



Okay so...what does that mean? I really, truly do not understand what this part means. I honestly think you're taking an overly-restrictive definition of "strategy," here, since (as I said earlier) I am allowing that even Nomic admits strategy, it's just going to look extremely different from other games' strategy, because it relies on a different set of skills than most games (mostly creativity and spontaneity, as opposed to within-the-rules analysis and calculation). "Strategy" is not limited to within-the-rules analysis of numerical values; it embraces a much wider variety of things. "Design" is not limited to setting probabilities for an internal random number generator; it too embraces a much wider variety of things--in short, anything you can gather data about, which is a very, very large set of things. (I am not a hyper-empiricist, I genuinely believe there are questions for which the answer cannot be found solely through measurement and counting, but I am skeptical of the notion that such questions are that relevant to game design.)

...

Okay, but are you therefore asserting that there is no strategy at all? That there is no way, even in principle, for two players to sit down and compare relevant parts of their experiences, and to plan for future concerns? Because that's the kind of "game" you described above. One where trying to live your 7th "life" with a different group of people is so alien, you're not sure you're still playing "life."


By 'strategy' I mean long-term plans or policies of interacting with the game in order to achieve a goal or reward. Responses you map out in advance of the events which will demand those responses. In math terms we could talk about this in terms of, for example, the way that you can have strong guarantees of asymptotic zero regret methods for bandit problems, but once you add the right kind of feedbacks or non-stationarities to the problem then you can force those methods to break. But I'm not sure that's a useful analogy for us to discuss?

Keeping in mind the previous thing about the difference between sitting at an extremal point and moving in the direction of that point, I'm not requiring strategies to be completely invalidated. Rather, I'm looking to increase the degree to which strategies are invalidated. Not everything from Life 1 will be transferrable to Life 2. Less will be transferable from life to life than from, say, chess game to chess game. It's possible to devise methods by which strategies will be destabilized, largely centering around non-stationarity and incomplete information. This element in particular I think you could probably target in your design methodology, though it would as a consequence make the kind of testing you want to do more difficult or expensive (because now you really do need to test multiple campaigns-in-sequence to get one data point, rather than getting multiple data points even within a given campaign).




Personally I think the answer is quite simple. "Bad game" means different things to different people, hence one must provide one's definition first. If the standard is "did people have fun with it? if not, then it's a bad game," then 3.5e is just about the antithesis of a bad game, because it's historical fact that an enormous number of people had a huge amount of fun with it. But the standard is rather loose, as one can argue literally any game enjoyed by more than one person can't be a "bad game." My personal preference is: "does the game fulfill the goals its designers communicated to its players? if not, then it's a bad game." And by that standard, there's a lot of room to argue that 3.5e did badly--after all, not only its own creators, but the people who took up the mantle to keep it alive, have both admitted that in order to fulfill their goals better, they had to leave that system behind.


I just find it sort of boring conversation because, if you conclude with respect to some formal abstraction that 'the designers did a bad job' then so what? If you've been enjoying it, that conclusion would serve you poorly if it led you to stop. If you haven't been enjoying it, you don't need to prove that it's objectively bad in order to make the decision not to play it. So there's a high likelihood that someone discussing this honestly will be led to a point where they have to say 'okay, you can say all of that, but it seems that I don't care'. That tends to make it a frustrating discussion topic, because rather than just say that people try to find some logic to defend their tastes (or try to argue 'no, you can't possibly like that!'), and stuff goes in circles.

[ezekielraiden]

@NichG: If it pleases you, I can do a full reply to your post, but there's a thing I was reminded of tonight that, I think, may help show why I don't see the gap you seem to. That is, I see it as a sliding scale where some of the stuff you're looking for is difficult to design for, but still something that works within the framework I'm talking about. It's not a difference of kind, but of degree--and, as I think D&D and PF have shown, it's not like even the obviously-mathematical parts are really that easy to design either.

Specifically, I'm thinking of the Thanksgiving Day, 1983 episode of Sesame Street, an episode that actually aired before I was born but which still manages to be moving today as an adult. That is, this is the episode where the cast explains to Big Bird that Mr. Hooper died, and won't be coming back, and that there is no "good reason" for it, it's just...because.

This was a serious issue, and the crew knew they had to deal with it correctly--a stumble on a topic of this magnitude would be bad for the audience and the show. So they did their homework. They consulted with psychologists, writers, religious experts, and more, and they made sure to make use of test audience children. In other words, they did do a significant amount of real statistical testing, backed up by careful, respectful understanding of the topic and how hard it could be to communicate it effectively to children. Test-audience children were carefully interviewed before and after, and were in other ways tested to determine what effect the episode might have, even though dealing with something as complex (and I suspect you'd agree) obviously transformative as "accepting death" is just about the definition of individual and non-transferable.

This, to me, is a perfect example of a carefully-designed experience, that admitted very important rigorous (and statistical) testing, while still ultimately being about exposure to a difficult, painful concept for young children who might struggle to understand it. Admittedly, a TV show is not a game, so there's no rules-construction and no equivalent to "multiple trials," etc. But as far as a comparison goes between an episode of a TV show and the design of a game? This seems good enough.

It's also worth noting that the scene was only filmed with one take--the cast couldn't bring themselves to go through it a second time. But in doing it that way, they made their reactions as thoroughly honest as possible, and on review, they realized that that was very important. You see some of the characters tear up, struggle to speak, look away, etc. Big Bird presents the perfect image of the insistent child, frightened and upset while the adults struggle with a subject profoundly difficult for them and the child. Showing children that there are things that can even make adults sad, and for which even adults have no good answers and no power to change the situation, was extremely important. Based on their pre-airing interviews and post-airing studies, the overwhelming majority of children understood the message, paid attention through the whole scene, and had a fruitful and positive conversation with their parents about the subject of death--and this was millions of children, because it was aired on Thanksgiving Day, when most parents would be able to have such a conversation with their children.

So...yeah. That's a transformative experience that was very carefully designed, with extreme care and consideration put to the specific questions asked during the design process. It included elements that cannot be "designed" in any meaningful way, like the emotional reactions of the cast members, but which could with review be understood for how they would influence the intended result (communicating to children about death).

Anything that cannot be grasped by this kind of design--anything sufficiently difficult to share that no meaningful data can be collected about the experience--would need to be pretty dang alien. If even teaching children about death (and the emotions and problems that result from it) is quite meaningfully "designable," I'm not really sure that there's all that much within the "transcendental experience" umbrella that can't be "designed" but is still worth pursuing. It requires genuine creativity and effort to ask the right questions for this kind of design. You have to work hard and maybe be unusually smart to do it, but you can still do it.

[AdAstra]

Gotta agree with Ezekielraiden. Just because something is impossible to quantify in objective terms, doesn’t make it inimicable to design, discussion, and testing. Something as hard to quantify as say, the taste of a really good and complex meal, can still be tested, discussed, refined, rebuilt, and tested again. Language is ALWAYS an imperfect description of the world around us. You can never describe something and be sure the other person is thinking and feeling exactly as you do. Chances are they are not. Testing and discussion are still possible, though

[NichG]

@NichG: If it pleases you, I can do a full reply to your post, but there's a thing I was reminded of tonight that, I think, may help show why I don't see the gap you seem to. That is, I see it as a sliding scale where some of the stuff you're looking for is difficult to design for, but still something that works within the framework I'm talking about. It's not a difference of kind, but of degree--and, as I think D&D and PF have shown, it's not like even the obviously-mathematical parts are really that easy to design either.

Specifically, I'm thinking of the Thanksgiving Day, 1983 episode of Sesame Street, an episode that actually aired before I was born but which still manages to be moving today as an adult. That is, this is the episode where the cast explains to Big Bird that Mr. Hooper died, and won't be coming back, and that there is no "good reason" for it, it's just...because.

This was a serious issue, and the crew knew they had to deal with it correctly--a stumble on a topic of this magnitude would be bad for the audience and the show. So they did their homework. They consulted with psychologists, writers, religious experts, and more, and they made sure to make use of test audience children. In other words, they did do a significant amount of real statistical testing, backed up by careful, respectful understanding of the topic and how hard it could be to communicate it effectively to children. Test-audience children were carefully interviewed before and after, and were in other ways tested to determine what effect the episode might have, even though dealing with something as complex (and I suspect you'd agree) obviously transformative as "accepting death" is just about the definition of individual and non-transferable.

This, to me, is a perfect example of a carefully-designed experience, that admitted very important rigorous (and statistical) testing, while still ultimately being about exposure to a difficult, painful concept for young children who might struggle to understand it. Admittedly, a TV show is not a game, so there's no rules-construction and no equivalent to "multiple trials," etc. But as far as a comparison goes between an episode of a TV show and the design of a game? This seems good enough.

It's also worth noting that the scene was only filmed with one take--the cast couldn't bring themselves to go through it a second time. But in doing it that way, they made their reactions as thoroughly honest as possible, and on review, they realized that that was very important. You see some of the characters tear up, struggle to speak, look away, etc. Big Bird presents the perfect image of the insistent child, frightened and upset while the adults struggle with a subject profoundly difficult for them and the child. Showing children that there are things that can even make adults sad, and for which even adults have no good answers and no power to change the situation, was extremely important. Based on their pre-airing interviews and post-airing studies, the overwhelming majority of children understood the message, paid attention through the whole scene, and had a fruitful and positive conversation with their parents about the subject of death--and this was millions of children, because it was aired on Thanksgiving Day, when most parents would be able to have such a conversation with their children.

So...yeah. That's a transformative experience that was very carefully designed, with extreme care and consideration put to the specific questions asked during the design process. It included elements that cannot be "designed" in any meaningful way, like the emotional reactions of the cast members, but which could with review be understood for how they would influence the intended result (communicating to children about death).

Anything that cannot be grasped by this kind of design--anything sufficiently difficult to share that no meaningful data can be collected about the experience--would need to be pretty dang alien. If even teaching children about death (and the emotions and problems that result from it) is quite meaningfully "designable," I'm not really sure that there's all that much within the "transcendental experience" umbrella that can't be "designed" but is still worth pursuing. It requires genuine creativity and effort to ask the right questions for this kind of design. You have to work hard and maybe be unusually smart to do it, but you can still do it.

I generally agree with this post. I think the only point of deviation is perhaps what we're each considering to be the 'hard part'. In the sesame street episode example, the transformative element was essentially already in hand, and the people involved understood that it would be so already (because e.g. they could remember when they first learned about death as kids, or telling their own kids about death). So the design question was then about how to modify the reception of that - sort of polishing the technical details around an already solid core.

In my eyes, it's an interesting design challenge to, e.g., 'find something as impactful as learning about death for the first time, that we as adults do not yet know'. Since we probably can't guarantee finding such things, how do we maximize our chance of happening upon them? One has to be very careful in using historical data to guide such a search, since it's easy to overfit: 'Minecraft was a transformative experience, so if I make a Minecraft clone it will also be transformative' doesn't work - therein lies the path to another fantasy heartbreaker. I'm not going to say it's impossible to use historical data for it, but it has to be used very indirectly and with open eyes to the changing context in order to have any hope of generalizing.


Gotta agree with Ezekielraiden. Just because something is impossible to quantify in objective terms, doesn’t make it inimicable to design, discussion, and testing. Something as hard to quantify as say, the taste of a really good and complex meal, can still be tested, discussed, refined, rebuilt, and tested again. Language is ALWAYS an imperfect description of the world around us. You can never describe something and be sure the other person is thinking and feeling exactly as you do. Chances are they are not. Testing and discussion are still possible, though

Non-designability wasn't my claim, though you'll probably have to wade through 2-3 pages again to trace how we got onto this. Basically it goes back to something Ezekielraiden originally said about defining balance as matching quantifiable targets, and the discussion which followed had to do with establishing whether or not that position implied that balance (in that sense) equaled design.