So... I have an idea for a game... but I believe it would be too complex to actually be made. Opinions?

[Cynicaladm]

on a more writer oriented note, I am now wondering if this game should be, from a creative point of view, be thought of as an exploration of the evolution of NPCs from the viewing point of the MC, or as an evolution of the MCs primarily... it does change the narrative angle quite a bit. so much effort and attention would go in the evolution of the NPCs and the consequences this brings to the narrative, that I almost did not consider the balancing act that this requires to not make the MCs disappear in the background. From a gaming perspective, the MC of choice is the player.. and it should be about him... so.. that's food for thought too.

I would hope that the interconnectivity and dependencies between the stories would be a little more complex than modifying one or two stats.
In the example with the wife, maybe the husband has a NTR fetish and does that with her. That means in the other timeline, maybe she can get with the lawyer but the husband will be there to watch or something. And hopefully even more stuff I didn't think about yet.
That's why I stated in one of my posts that everything should be thought about from the start.
Imagine developing a good chunk of this game with a few stats and then realizing you need something more complex to be able to handle what you want to implement next.
That would mean you have to rewrite a lot of the code, retest everything and so on. I've seen a lot of games going into redesign of code because they couldn't handle the new ideas the developer had. And so one or two updates were mostly about the rewrite of the code than anything else. Or maybe some even died.
Of course, it is possible to also make it simple and use just a few stats and check them, I think it depends on the story and what those branches contain.



I think that could also be a way to implement your complex game in a simpler and more realistic way.
If you want 4 stories, actually make 4 separate games, each with only 1 story. Start with the most interesting one and gain a fan base for the world. Games could be made to depend on each other either with the player selecting what choices he made in the other game or by reading save games or some ending output files.

that "augmented consequence" would be the whole reason for this game in the first place, so yes... I agree.
and I also agree that this should be planned in advance pretty much to completion.

That said, you pose another interesting issue by postulating that these are alternate timelines. I had not thought of that in these terms, but it is true.. the playthrough with the first character is the baseline, on which, potentially, a second character inserts his actions and experiences.. now.. the two can travel parallel to one another, but the first story influences the second, and the second, in turn (at least until the nexus event is reached) does not influence the first one.. so the relationshisp between characters, NPCs and such, are sort of lopsided. that's another thing to take into account in defining how to carefully insert additional narratives on the previous ones and still avoid for them to go into conflict with one another.
It's basically Return to the future... which is a terrible example of time travel and not how traveling in time would actually affect reality at all, but it's the closest approximation.
From a writing standpoint, that might actually be the biggest challenge... because it can be done, but it adds yet another layer of complexity and requires a bunch of further dialogue to explain certain dynamics..dialogue that changes with each permutation and order of MC playthrough....

 

[woody554]

it's a 10 year project. not sure it's a good enough idea to justify using that much time. but sure it's possible, stubborn people have done much more against all odds. the question is are you one of those 1-in-a-milllion guys?

don't take this the wrong way, but not having learned how to program kinda already points at you being one of us dreamers rather than those rare doers. but then again those kind of people wouldn't listen to anyone telling them it can't be done.

realistically though, it'll never happen. the odds for you surviving the workload of first 3 updates of a single character one-route plot (without any complex programming) are something like 1 in a 100 or less. you'll be swamped with problems you now don't know even exist, hit the wall, burn out, and that'll be it.

nothing concrete stops you from doing it though. nothing we think matters. the only obstacles are in your own head.

 

[Cynicaladm]

it's a 10 year project. not sure it's a good enough idea to justify using that much time. but sure it's possible, stubborn people have done much more against all odds. the question is are you one of those 1-in-a-milllion guys?

don't take this the wrong way, but not having learned how to program kinda already points at you being one of us dreamers rather than those rare doers. but then again those kind of people wouldn't listen to anyone telling them it can't be done.

realistically though, it'll never happen. the odds for you surviving the workload of first 3 updates of a single character one-route plot (without any complex programming) are something like 1 in a 100 or less. you'll be swamped with problems you now don't know even exist, hit the wall, burn out, and that'll be it.

nothing concrete stops you from doing it though. nothing we think matters. the only obstacles are in your own head.

as I said, this thread started as an exploration as to whether such a game could be possible, and whether it would be something anyone would care to play at all.
Also based on my curiosity about how these games are put together behind the scenes...
It is mostly a thought experiment, and not something I ever intend to actually make. I might write it out as a creative exercise, but I know that it is not something that will ever see the light of day on this forum in any videogame format.

 

[woody554]

maybe you should just make it a book instead of a game. four intertwining stories. kinda reminds me of the structure of sin city.

I can see you like writing enough, and you like planning it out. so writing it into one giant story would remove all the extraneous obstacles you're facing now, and let you do what you're already doing from day 1.

 

[Cynicaladm]

maybe you should just make it a book instead of a game. four intertwining stories. kinda reminds me of the structure of sin city.

I can see you like writing enough, and you like planning it out. so writing it into one giant story would remove all the extraneous obstacles you're facing now, and let you do what you're already doing from day 1.

I'm actually in the process of writing lore and content for the setting I am DMing a Dungeons and Dragons campaign in.. there's plenty of lore entries, locations, events and NPCs the players/characters have no guarantee of ever encountering or finding out about.

This would be an undertaking of similar scope... if I have time for it beyond thinking about it and getting feedback about it on here, I might just do that... but I actually do have a life and some daily commitments too, so it may well be that this thread is all that there will ever be.
which doesn't mean I find it any less intersting a thought experiment.

 

[DaClown]

This is really just a problem of scale.

If you pick out say 3 main characters that are playable but somehow confined from interaction with each others then you can make something buildable and playable by confining the number of characters they interact with.

Having a million playable characters with dozens or more NPCs they interact with and dozens or more of those NPCs interact with each other would produce something that even AAA studios would not likely develop well.

3 main characters with 3 supporting characters would likely not be complex enough for the effects you're hoping for. In fact, we can just exhaust the first few sets of possibilities.

2 main characters with 2 supporting characters each. If PC.A exclusively interacts with NPC.A.1 then NPC.A.2 would need to interact with PC.B, so your total number of unique characters would be 5. NPC.A.2 would be NPC.B.2. So NPC.A.1 and NPC.B.1 would be distinctly different characters. That's the boring case which forms the basis for the more interesting permutations and combinations.

If the requirement for the design isn't strict direct interaction but strict friend-of-a-friend interactions through NPC-NPC couplings then you could have 6 characters in the 2MC/2SC case.

I think the minimum interesting set for strongly coupled MCs is 3MC/3SC. Consider contrarily the 3MC/2SC case: PC.A exclusively interacts with NPC.A.1; PC.B exclusively interacts with NPC.B.1; PC.C exclusively interacts with NPC.C.1; PC.A and PC.B interact with NPC.A.2/B.2; PC.C would have to have exclusive interaction with both NPC.C.1 and NPC.C.2. 3MC/2SC would work if no NPCs are exclusive to any specific MC.

So 3MC/3SC would result in: PC.A exclusively interacts with NPC.A.1; PC.B exclusively interacts with NPC.B.1; PC.C exclusively interacts with NPC.C.1; PC.A and PC.B interact with NPC.A.2/B.2; PC.B and PC.C interact with NPC.B.3/NPC.C.2; PC.A and PC.C interact with NPC.A.3/NPC.C.3.

Core mechanic for your game would be the presence or absence of Exclusive/Special, Particular, and Universal NPCs. A universal NPC would appear in the story/gameplay-space of all the MCs. Particular NPCs would only appear in some subset of MCs' stories/gameplay-space. Exclusive/Special NPCs would appear pairwise with a specific MC.

If you go for weak coupling rather than strong coupling or in addition to strong coupling then NPCs would further classify as interacting with other NPCs or not. A non-interacting-NPC would only have interactions with the MC and not with any NPCs. A Universally-Interacting-NPC has interactions with all characters. Particular-Interacting-NPCs would interact with some but not all characters and possibly one or more MCs.

A potentially interesting game model from a purely technical standpoint--not necessarily as something fun to play--would be the existence of NPCs that do not interact in any way with the MCs but interact with NPCs that interact with the MCs.

What you're fighting against in terms of feasible development is

You must be registered to see the links

and

You must be registered to see the links

; your design has to constrain the problem practically to sub-exponential complexity in

You must be registered to see the links

,

You must be registered to see the links

, and

You must be registered to see the links

at least at runtime if not at compile/generative-grammar time.

You must be registered to see the links

realizes some of these concepts though they are secondary not primary to the game mechanics and game play in general.

What is generally being described is a kind of neural network.

 

[Cynicaladm]

And this kids, is why your dad did not do well in math at school...

I understand all of the words you say, some even with a degree of competence, when you combine them my eyes glaze over and I find myself putting my trust in the fact that you seem to know what you're talking about, lol.

on a more serious note, I believe these interactions can be curtailed in ways to diminish the workload and the potential combinations, whilst still making it an enjoyable experience that does not betray the core concept of the game.
Second hand NPCs would easily feature in the game as well on account of it containing a fair number of VN elements.. narrations and flashbacks that I would be using to expand the tale/enrich the character or NPC and so on, in a manner not unlike the female character in The Adventurous couple does when she asks if her husband wants to know what she got up to today, which narratively speaking (in her context) plays right into the kinks of NTR, which could be one gameplay path for a MC.
Being as I approach this from a narrative standpoint anyway, I would use VN style narration and transitions to set the stage, expand on plot points and maybe explain away contradictions that might arise in specific gameplay sequences...
This of course poses yet another balancing act challenge, that of mixing VN and active game with choices.. with the risk of displeasing both the VN enthusiasts and the more gaming/choices oriented players.. this is a balancing act the solution of which would rest entirely on pacing, I believe.

 

[Cynicaladm]

If anything, the above clarifies that this concept could only work in a carefully crafted story, and not in a free roaming sandbox environment, where recursions and all manner of narrative inaccuracies would crop up.

 

[DaClown]

Being as I approach this from a narrative standpoint anyway, I would use VN style narration and transitions to set the stage, expand on plot points and maybe explain away contradictions that might arise in specific gameplay sequences...

What I sketched out in schematic form is relevant to both programmatic approaches and to strict narrative approaches to the problem.

The implicit system of characters would be useful for procedural programming in a sandbox/open-world type game or for a procedural generative grammar for a narration engine. It is however still relevant for just estimating the minimum and maximum amount of writing that you would need to do or the number of assets or scenes that would be required to be constructed for the narrative.

Done badly, you end up writing in such a way that each discrete scene entails exponentially more scenes. This is the basic problem with tree-based stories where each decision creates mutually exclusive forks. You have 1 initial scene then 2 child scenes which have 4 scenes. The nth sequence of scenes will be 2^n in total number assuming you are choosing between mutually exclusive non-intersecting choices each time. Total number of scenes possible would be Sum(2^n) = 1 + 2 + 4 + ... +2^n = 2^(n+1)-1. Manageable for a story with say 5 choices (63 total scenes) but somewhat unmanageable for a story with 10 successive choices (2047 total scenes). Completely unmanageable even for AAA studios at around 20 successive choices (more than 2 million total scenes).

Combining a tree-base story with the graph of possible character interactions results in an implicit system of narration that can get very large very quick especially if choices are not binary but n-ary (more than 2 mutually exclusive choices at each decision branch).

This is why most AAA studios end up doing a story that has like one to three possible endings and all the choices ultimately don't really do much if anything.

 

[Cynicaladm]

What I sketched out in schematic form is relevant to both programmatic approaches and to strict narrative approaches to the problem.

The implicit system of characters would be useful for procedural programming in a sandbox/open-world type game or for a procedural generative grammar for a narration engine. It is however still relevant for just estimating the minimum and maximum amount of writing that you would need to do or the number of assets or scenes that would be required to be constructed for the narrative.

Done badly, you end up writing in such a way that each discrete scene entails exponentially more scenes. This is the basic problem with tree-based stories where each decision creates mutually exclusive forks. You have 1 initial scene then 2 child scenes which have 4 scenes. The nth sequence of scenes will be 2^n in total number assuming you are choosing between mutually exclusive non-intersecting choices each time. Total number of scenes possible would be Sum(2^n) = 1 + 2 + 4 + ... +2^n = 2^(n+1)-1. Manageable for a story with say 5 choices (63 total scenes) but somewhat unmanageable for a story with 10 successive choices (2047 total scenes). Completely unmanageable even for AAA studios at around 20 successive choices (more than 2 million total scenes).

Combining a tree-base story with the graph of possible character interactions results in an implicit system of narration that can get very large very quick especially if choices are not binary but n-ary (more than 2 mutually exclusive choices at each decision branch).

This is why most AAA studios end up doing a story that has like one to three possible endings and all the choices ultimately don't really do much if anything.

thanks for the insight.. it does clarify somewhat the scope of this concept.. and why I am quite happy to treat it as a thought experiment and nothing more.

 

[DaClown]

If anything, the above clarifies that this concept could only work in a carefully crafted story, and not in a free roaming sandbox environment, where recursions and all manner of narrative inaccuracies would crop up.

It could work on an indie budget with a small team of developers in a very tightly constrained scenario with a very small number of total characters and a very small number of possible interactions/character-relationships. Most of the characters would need to be either non-universal or non-universally-interacting to keep the number of possible interactions, relationships, and scenes/scenarios low enough to be manageable for development.

Forget building this for a contemporary city. Small island culture or a collection of small island cultures would be more appropriate. Or like a cabin in the woods. Or small teams developing habitats on the Moon/Mars. A Vault in Fallout. A remote monastery next to a couple of rural thorpes. A couple of remote farms. Crashlanded on a water planet with little to no hope of rescue.

And it would help a lot to think in terms of developing a generative grammar for most of the possible narratives; at least something that creates the skeletons of the narratives like a sheet breaking down the scene in terms of who or what is included in it. Handwriting the important scenes but automatically generating the bulk of it. Corruption of Champions succeeded at the basic concepts for this.

 

[Cynicaladm]

Forget building this for a contemporary city. Small island culture or a collection of small island cultures would be more appropriate. Or like a cabin in the woods. Or small teams developing habitats on the Moon/Mars. A Vault in Fallout. A remote monastery next to a couple of rural thorpes. A couple of remote farms. Crashlanded on a water planet with little to no hope of rescue.

And it would help a lot to think in terms of developing a generative grammar for most of the possible narratives; at least something that creates the skeletons of the narratives like a sheet breaking down the scene in terms of who or what is included in it. Handwriting the important scenes but automatically generating the bulk of it. Corruption of Champions succeeded at the basic concepts for this.

or.. you know... a bunch of people that don't necessarily hump everything that moves and they interact with, lol... we may live in a metropolis, but we are all social islands anyway, and have limited interactions with relatively small social circles and meaningful interactions with even fewer than those.
Much like what happens in a D&D setting, the population of the world counts millions, the relevant ones are those that interact with the PCs, and of those, most are only sketched and have a couple of infos and relevant functions but nothing else. The small group of active and relevant NPCs are what constitutes the small island culture/social circle in your example.
The fact that they phisically navigate a loaded metro carriage doesn't add extra labour... because the other commuters are scenery, not NPCs.

When you speak about generative grammar, which I had to google (I'm not native English and I didn't recognise the term),are you talking about the programming side of things, to "auto generate" code, or also the plot/textual side of it?

 

[ihl86]

What I sketched out in schematic form is relevant to both programmatic approaches and to strict narrative approaches to the problem.

The implicit system of characters would be useful for procedural programming in a sandbox/open-world type game or for a procedural generative grammar for a narration engine. It is however still relevant for just estimating the minimum and maximum amount of writing that you would need to do or the number of assets or scenes that would be required to be constructed for the narrative.

Done badly, you end up writing in such a way that each discrete scene entails exponentially more scenes. This is the basic problem with tree-based stories where each decision creates mutually exclusive forks. You have 1 initial scene then 2 child scenes which have 4 scenes. The nth sequence of scenes will be 2^n in total number assuming you are choosing between mutually exclusive non-intersecting choices each time. Total number of scenes possible would be Sum(2^n) = 1 + 2 + 4 + ... +2^n = 2^(n+1)-1. Manageable for a story with say 5 choices (63 total scenes) but somewhat unmanageable for a story with 10 successive choices (2047 total scenes). Completely unmanageable even for AAA studios at around 20 successive choices (more than 2 million total scenes).

Combining a tree-base story with the graph of possible character interactions results in an implicit system of narration that can get very large very quick especially if choices are not binary but n-ary (more than 2 mutually exclusive choices at each decision branch).

This is why most AAA studios end up doing a story that has like one to three possible endings and all the choices ultimately don't really do much if anything.

Your huge scene number is based on the presumption that each scene would generate 2 new scenes. I think you made it too abstract of a problem when calculating the possibilities and didn't think of it in a concrete way.
I am having trouble imagining this type of multiplication of consequences for game decisions. Unless you do it just to increase the complexity.
I feel like there are only a few big decisions that could indeed influence and branch the story further along.

Regarding the Sandbox implementation, imagine a Sims game where you can play with multiple characters from a city of wherever they live.
You can play with a family today and grow some relations and if you play with another family the next day, those influences exist and you can make other type of interactions based on what you played with the first family.
So, of course it can also be done in a sandbox type.
You just need to make the "rules of engagement" between characters very clear and program them that way.

I mean, in the end, how many type of scenes/interactions can you have between two characters. It is a limited amount and it certainly won't get too far. I mean, if two character have slept together 2 times or 6 times, they 7th interaction doesn't have to have a different dialogue/scene. In a perfect scenario, they would be more familiar with each other, the more interactions they have, but that would also be true for any of the single MC games out there and they are not implemented that way.

I think you guys got so caught up in the abstract that you forgot the point of this. Just try to put on paper some real characters/interactions consequences and see how many would really exist. And don't just make it complicated for the sake of making it complicated. Think of the simplest way to do it that can still do service to a good story.

 

[Felicityskye]

Your game idea feels somewhat like the game, Zero Escape: Zero Time Dilemma. Without major spoilers, this VN splits a group of people into 3 teams who are trapped in a facility. And the point of the VN/game is to find a way to escape. Each team's choices effects the other teams. You can purposely play as one team and be blinded to the default choices the game gives the other teams till the end, but to get the perfect ending you have to play as all the teams.

Your idea is definitely doable, but the logistics of such a game or VN could be a nightmare for an amateur or even a small indie dev team, depending how big the overall story and interactions you're aiming for.

 

[DaClown]

or.. you know... a bunch of people that don't necessarily hump everything that moves and they interact with, lol... we may live in a metropolis, but we are all social islands anyway, and have limited interactions with relatively small social circles and meaningful interactions with even fewer than those.
Much like what happens in a D&D setting, the population of the world counts millions, the relevant ones are those that interact with the PCs, and of those, most are only sketched and have a couple of infos and relevant functions but nothing else. The small group of active and relevant NPCs are what constitutes the small island culture/social circle in your example.
The fact that they phisically navigate a loaded metro carriage doesn't add extra labour... because the other commuters are scenery, not NPCs.

When you speak about generative grammar, which I had to google (I'm not native English and I didn't recognise the term),are you talking about the programming side of things, to "auto generate" code, or also the plot/textual side of it?

Latin is a generative grammar.

A common form of a generative grammar used in game designs and narratives is the

You must be registered to see the links

. Often in text-based games, the writers and the programmers will work together to develop a kind of grammar that will reliably produce appropriate skeletons for the construction of scenes and characters. Again, Corruption of Champions demonstrated this and developed it rather extensively with its paperdoll system, so you can click on a character's description.

Even if you don't rely on a computer to generate the mad lib, you are as a writer free styling such sheets on the regular.

What I am describing exists explicitly on the side of programming and is useful for the auto-generation of computer codes, but it is applicable and implicit in many forms of linguistics and is relevant for narration generally.

Basically, the game would follow multiple parallel paths.

At the beginning, you would have to decide to play one of several characters, let's say 4 of them.
...
These stories are independent from one another, effectively being 4 separate games, with characters that each explore different kinks (or occasionally overlapping/similar ones, depending on the choices made by the player).
The twist, such as it is, would be that they live in the same area/city/setting, and that some, or several, of the NPCs they interact with, are tied to at least one of the other characters.. so maybe the wife of character nr 1 works for character nr 3, or is a teacher at the school of character nr 2... maybe the ex wife or the daughter of character nr 3 is friends or in a relationship with character nr 2... or maybe the fourth character knows several of the NPCs the other MCs interact with for professional reasons... or all of the above and more... and then each of them have access to several NPCs that are not in the lives of the other MCs, with authonomous plotlines.

What you are describing here is a fragment of the generative grammar of the game narrative that you're interested in developing or at least exploring.

You lay out the rules for how to create valid and meaningful sentences and scenes (represented by combinations of sentences and sequences of images or sounds) that are relevant to the game design you are specifying.

My initial post in this thread is part of my attempt to translate what you describe in your opening post into a formal grammar that can be mapped to numerical models. Given that we're talking about a text-based game in say a visual novel then this corresponds to theories about the length of

You must be registered to see the links

representing the game design as it necessarily must to be implemented as a video game in a computer.

 

[Cynicaladm]

Gotcha.. or at least the gist of it where it pertains to the programming side. The rest is fairly clear.

The DnD jokes are funny and absolutely true... that would be entirely the case if this wasn't a programmed thing, with only a certain number of options each time leading to predetermined outcomes (however many those may be).. there will be no case of
"you're in a tavern, at a table in the corner, Lord Armpit of Summerisle is taking his tea.. his weapons are, for once, nowhere to be seen. After a year long research, you have the killer of your family at your mercy"
"I want to find out where the maid sleeps and if she is wearing underpants"
"but... the lord... your family... your life's mission..."
"underpants!"

That said, the generative grammar approach to the programming, is well and truly beyond my grasp, in terms of being able to do something with it..nor would I really be interested in that side of things, because it would require me to learn how to do a lot of things that I would be doing primarily for this one project... and it would be akin to learning how to speak again, remapping my brain almost... a bit more than I would want to chew on.
Even though it is an interesting aspect of this thought experiment, I feel unqualified to even expound on that side of things with anything but guesswork at best.
On the more narrative, literary and character creation side of things, I can see how it relates, but I have a hard time establishing if I would need to use the concept and the logic of generative language consciously, if it is something I already sort of include in my creative and thought process, or if it would rather impede my instinctive/trained by repetition way of approaching the act of creation and of putting structure and logic in characters and their development.
I guess I will have to read up on it to familiarise myself with it and the logic underpinnings and implications... which I might do just because you've peaked my interest.
Usually, when I do write, I am rather free form.
right now, on my actual running project (the DnD setting I mentioned in previous posts), I started with a mapmaking tool, made a world map, cut out a chunk of it and gave it a name, then zoomed in to where the action would take place.. now I am describing a village, with the various locations and NPCs populating it (again, only main ones, and only in generic terms), a few generic descriptions of the neighbouring locations, main concepts for the generic background (empire, gods, my personal fluff on the various main races populating said empire)...
Mind you, the players might still pull an underpants on me, take a boat and go somewhere else entirely, and in that case I will have to improv, but that's freeform and dynamic, and DnD..
This is much more self contained and limited to the degree I would want or need to limit it to make it work. once the skeleton of the beast is laid down, I can always stick to the muscles, bones and skin, or I can get creative, add a tentacle or change the hair colour, or expand on the story of how the tentacle came to be.. and it would depend on the player if he wants to watch and read the VN scene of the sexy tentacles, or if he doesn't care and want to progress with the main plot. but ultimately, it's still going to have to walk on the number of legs I decide it has... and that's all the players can make it do.

also.. scaroused.. lol

 

[Junisdottir]

It's possible but sounds pretty convoluted. I think you'd run into more problems than its worth making it

 

[DaClown]

On the more narrative, literary and character creation side of things, I can see how it relates, but I have a hard time establishing if I would need to use the concept and the logic of generative language consciously, if it is something I already sort of include in my creative and thought process, or if it would rather impede my instinctive/trained by repetition way of approaching the act of creation and of putting structure and logic in characters and their development.

Writers have the luxury of not having to be explicit about this kind of thing. Programmers generally don't.

Usually, when I do write, I am rather free form.

that would be entirely the case if this wasn't a programmed thing, with only a certain number of options each time leading to predetermined outcomes (however many those may be).. there will be no case of

You're sorta right and sorta wrong. In a strict formal sense, you're absolutely right; the players have no choice but to execute the actions explicitly written into the game model.

In an informal sense, you're wrong. Players disagree with developers all the time, and the players may be disappointed with the implementation or the constraints, so it is a very common pattern that the players begin to wonder about creating their own game. A better game.


This whole thread being an example.

What ends up happening is that you either fail to attract interest or fail to retain interested players or you attract and retain interest while ending up in situations where there are demands to expand content or mechanics you had not planned on in the first place.

I have a hard time establishing if I would need to use the concept and the logic of generative language consciously, if it is something I already sort of include in my creative and thought process, or if it would rather impede my instinctive/trained by repetition way of approaching the act of creation and of putting structure and logic in characters and their development.

As a writer in a project, it would be better in general for you to write without explicitly using the concept in the same way that you don't necessarily spend most of your time wondering explicitly about things like noun vs verb vs adjective. Better to generate too much and have to edit it down into the constraints of the programming languages or game engine than to over constrain your writing in the first place.

It is however something that needs to be either implicitly or explicitly laid-out when working together with a programmer to get it implemented into a format that can be automatically (re)generated and formatted for the player to view and interact with through a UI.

Your huge scene number is based on the presumption that each scene would generate 2 new scenes. I think you made it too abstract of a problem when calculating the possibilities and didn't think of it in a concrete way.
I am having trouble imagining this type of multiplication of consequences for game decisions. Unless you do it just to increase the complexity.
I feel like there are only a few big decisions that could indeed influence and branch the story further along.

The scene numbers that I state in that post are based on the assumption of the mutual exclusivity of the choices; they represent significant, consequential choices which can not have the same outcome without contradiction to the original premises. The model I am discussing there is representative of a maximum of complexity that can result from a typical boolean structure and represents substantially a system of proofs (thus all possible logical consequences of a first order predicate logic or equivalently the material consequences of a restricted set of premises).

Almost all actual narrative structures are far more constrained than that and most "choices" in a game or visual novel are not really mutually exclusive or consequential. My quantification on that specific example is "Done badly".

Quite a few games I've played here and elsewhere have a problem where they start with a few characters then escalate the situation. In a relatively open environment the escalation of the situation results in logical involvement of more and more characters; this results in plot holes large enough to drive starships through and can result in development delays on small person teams that become intractable. Accidental Woman is in this class; the dev wrote themselves into a situation where they have two different games represented by the female xor male protagonist respectively which take place in a whole-ass town over the course of months.

A lot of games have the problem that the situation they pick escalates rapidly towards either save the world or destroy it.

The whole point of that particular abstraction being an example of what not to do.

I mean, in the end, how many type of scenes/interactions can you have between two characters. It is a limited amount and it certainly won't get too far. I mean, if two character have slept together 2 times or 6 times, they 7th interaction doesn't have to have a different dialogue/scene. In a perfect scenario, they would be more familiar with each other, the more interactions they have, but that would also be true for any of the single MC games out there and they are not implemented that way.

It isn't about how many times you do an interaction, but how many types of interactions there are. Or really how many distinctly different states there are to the individual elements or members of a game/narrative model and how many distinctly different states there are of the entire game/narrative model.

If each main character has one and only one outcome such that the OP's originally stipulated 4 playable characters represent 4 separate storylines then the game is going to generally be boring because the rest of the OP's notions about varying NPC stats across the 4 separate storylines isn't going to matter. The OP's original statements indicate an interest in the logical or material consequences of allowing more variations to the outcomes of each of the playable character's stories such that their stories form substantially some kind of graph/network rather than some kind of line or curve.

For a more limited example, let's say that the outcome of each playable character's story depends strictly on the specific order in which you play them in.

Right there you have 24 distinctly different endings. Practically speaking, some of those are going to be indistinct enough that you can merge them, but this serves to illustrate the point. Either you're going to have extraneous playable characters or you're going to have quite a few outcomes that grows proportionate to the number of characters involved in the system of play.

If each character can only have 2 outcomes then we still get 8 outcomes total for 4 characters, and the interplay of each of the 4 characters is substantially less interesting because there isn't room for a distinctly different outcome for each of the possible permutations of the character interactions.

Even if we're not talking about the number of final scenes, we are fundamentally interested in the number of possible and probable intersections between the network of the 4 playable characters and the NPCs they interact with.

All of this allows us to put constraints on the problem to find a tractable and definite system that we can develop into a narrative structure or game model.

 

[Cynicaladm]

Writers have the luxury of not having to be explicit about this kind of thing. Programmers generally don't.


View attachment 930483

You're sorta right and sorta wrong. In a strict formal sense, you're absolutely right; the players have no choice but to execute the actions explicitly written into the game model.

In an informal sense, you're wrong. Players disagree with developers all the time, and the players may be disappointed with the implementation or the constraints, so it is a very common pattern that the players begin to wonder about creating their own game. A better game.

This whole thread being an example.

What ends up happening is that you either fail to attract interest or fail to retain interested players or you attract and retain interest while ending up in situations where there are demands to expand content or mechanics you had not planned on in the first place.

From a commercial standpoint, which is actually my professional field, in a very simplicistic way, and without entering in the field of promotion, marketing, trending, "artificially creating a need for the product" and such, once a product is finished and distributed, it's win or bust.. it either is well received or it fails spectacularly... or at most it becomes a sleeper hit/cult phenomenon.
But then, this exercise was never about appealing to a large audience but rather trying to make the best possible theoretical/abstract project around the basic concept of the game... or rather, that's what it has become on page two of this thread that I started mostly from a spur of the moment "let's test if this theory is feasible" act of curiosity.
If this were a case like other games that are modified according to the reception of the fanbase who increase or diminish their patronage accordingly, you would be right in having the concern, which is purely financial in nature, or rather, one of resource management and the balance between resources invested and return on the investment.
I am perfectly aware that this would have, out of necessity, to be the end goal, but this "potential game project" doesn't have to worry itself with real world financial implications... and, anyway, we would be talking about a completed product, not one susceptible to variation following the whims of Patreons.
It follows, that this would merely be the execution of my vanity project, or not even that.. the exploration of the concept of the game in a mindspace where the approval of the audience is immaterial to the execution, as are financial terms, and the role of the player is that of... satisfied end user and nothing more.. barely more than a placeholder or an achievement to unlock.

In other words, the objective of the exercise is more a debate of the nature of game theory (limited of course to this specific context of gaming, totally not the game theory in the academical/mathematical sense of the definition).. so..

how does this interconnecting storyline model of a game finds it's best balance and execution possible according, admittedly, to math and programming issues which I grasp only partially (and in terms of upscaling resources and tasks), and according to what, in my very personal judgement, makes for a good story and storytelling and good gaming experience.. (hence my thoughts on balancing the various parts, not so much to appease/appeal to more players, but rather to present a well balanced product that satisfies my requirements for playability and story worth telling)..
The player is there to represent someone who would be ideally satisfied with the end result.. but whether he thinks the game is crap and he can and wants to try to do better, is a real world concern, much like the financial worries.. and doesn't factor in the exercise, if you get what I mean, because should the player disagree, then, in a real and hypothetical sense I might have "failed" at the task.. but on the other end, I might still be satisfied with the end result, or satisfied with the fact that I have come up with the best product I could come up with.. whether any of this ever gets committed to page, to an attempt at production or even to a player experience or not. (spoiler: it won't)
And since this whole thing begins and will probably never go any further than my brain, I will still feel pretty satisfied with myself (and thank you for having come along on the ride with me )

...and yes, the outcomes would need to at least feel/appear distinct enough from one another to justify multiple playthroughs and to be significant and apparent, otherwise, indeed the whole concept would be easily reduced into a limited set of completely railroaded endings, which would invalidate the player agency that made them choose one path over the other... which goes against my DnD player sensibilities and against the very purpose of having interconnecting storylines that affect one another significantly during playtime.
then again, we are also not writing War and Peace here.. so endings don't necessarily have to have the gravitas and complexity of a grand piece of literature, which this wouldn't be.. I like to think I'm good with words, but I am not that conceited, lol

 

[ihl86]

The scene numbers that I state in that post are based on the assumption of the mutual exclusivity of the choices; they represent significant, consequential choices which can not have the same outcome without contradiction to the original premises. The model I am discussing there is representative of a maximum of complexity that can result from a typical boolean structure and represents substantially a system of proofs (thus all possible logical consequences of a first order predicate logic or equivalently the material consequences of a restricted set of premises).

Almost all actual narrative structures are far more constrained than that and most "choices" in a game or visual novel are not really mutually exclusive or consequential. My quantification on that specific example is "Done badly".

Quite a few games I've played here and elsewhere have a problem where they start with a few characters then escalate the situation. In a relatively open environment the escalation of the situation results in logical involvement of more and more characters; this results in plot holes large enough to drive starships through and can result in development delays on small person teams that become intractable. Accidental Woman is in this class; the dev wrote themselves into a situation where they have two different games represented by the female xor male protagonist respectively which take place in a whole-ass town over the course of months.

A lot of games have the problem that the situation they pick escalates rapidly towards either save the world or destroy it.

The whole point of that particular abstraction being an example of what not to do.


It isn't about how many times you do an interaction, but how many types of interactions there are. Or really how many distinctly different states there are to the individual elements or members of a game/narrative model and how many distinctly different states there are of the entire game/narrative model.

If each main character has one and only one outcome such that the OP's originally stipulated 4 playable characters represent 4 separate storylines then the game is going to generally be boring because the rest of the OP's notions about varying NPC stats across the 4 separate storylines isn't going to matter. The OP's original statements indicate an interest in the logical or material consequences of allowing more variations to the outcomes of each of the playable character's stories such that their stories form substantially some kind of graph/network rather than some kind of line or curve.

For a more limited example, let's say that the outcome of each playable character's story depends strictly on the specific order in which you play them in.
View attachment 930516
Right there you have 24 distinctly different endings. Practically speaking, some of those are going to be indistinct enough that you can merge them, but this serves to illustrate the point. Either you're going to have extraneous playable characters or you're going to have quite a few outcomes that grows proportionate to the number of characters involved in the system of play.

Ok, now I understood better what you were saying.
My point would be that from a practical point, as you said, you can merge some endings, results of interferences from other characters together in the first place.
I would also like that I find the premise of a characters ending depending on the order of play between the four characters very hard to even think about and put on paper. Less alone implement.

If each character can only have 2 outcomes then we still get 8 outcomes total for 4 characters, and the interplay of each of the 4 characters is substantially less interesting because there isn't room for a distinctly different outcome for each of the possible permutations of the character interactions.

I don't think I necessarily agree with this statement. For example, one of the characters could be a playboy. Maybe his goal in life is to have a harem of women at his disposal. Playing the other 3 characters correctly will allow him to have his perfect harem ending where he gets to live with the women he wants. This would mean that you need to play with 3 characters in a certain way to give this character 1 ending but I still think that would be interesting. And each of the interaction of other characters count towards this.
Another example could be that a character wants to revenge his family's death. Maybe you need to do specific things with some or all of the other's characters paths if you want for him to really get revenge and live through it. If you only do 2 of the 3, he could get his revenge and die happy, while if you do all 3 parts he could get his revenge and live through it or something. If you do none, the doesn't get his revenge but he gets to find love again or something and lives his life with a feeling of not feeling complete.
While the endings and influences are a little different, they are still kind of the same, the variation is small enough that it is doable. Of course, this is just referring to ending, as you also stated, there are more possible interactions and not just final scenes.
But I think that should not necessarily be a problem. You could make most of the characters real decision making/split path type of choices the ones dependent on other characters. The default could be the less than ideal outcome, while the best outcome happens only when the other character "helped".
Of course, this only proves that this idea is even harder to write because now almost everything has to be made with purpose and thought through.
Moreover I don't think there is such a large audience for this in the adult world community. If the game is completed and very well made, sure, probably on steam it could get a fanbase and it's just rewards, anywhere else, I don't know.
It would be a lot easier to make a stupid story, good graphics, Milfy City type of game or even multiple of those until one has success than this monster of an undertaking.