HTML Need feedback on game with Lilith's Throne-inspired interactive sex system

Eroniche

Formerly 'TrueSandbox'
Dec 23, 2019
5
8
Hey there, first-time poster here. I've been working on a small project, a text-based game where your actions should affect the world (creative, I know), for about half a year now. I figured I'd ask you all for some feedback.

Lilith's Throne is one of my all-time favorites, but a big problem with the game is that the sex system gets so repetitive. Once you've done it a dozen times, you know how practically every sex session will play out with every character, except for a few that have unique interactions like Pix and Ralph. This means that all the characters without unique interactions (this includes many named quest characters) are reduced to the same generic horny personality, which is both really boring and also very immersion-breaking to have Aunt Lilaya behaving the same as a random cheetah morph from the club.

I want to make a system that is similar mechanically but much more personalized. My goal is to have this scene be driven by the female character's personality.

Under this system, each character will react to your actions differently; each has their own likes and dislikes, loves and hates (lust/anger); each has their own unique style of dialogue that isn't just "fuck me" x10. In general, really, sex with each character will have a different "vibe".

Also, the aftersex will be slightly different depending on what you do during it. And if you're an asshole to them, they will end sex early, but that's just a little thing for flavor/immersion.

I only have a shallow implementation of this system for one NPC right now, but you can see how I won't be able to just reuse the same dialogue and descriptions for a different character like Lilith's Throne does. As is the goal, I would have to write all-new content based on each character's personality.

This system still took a surprisingly long time to make, though, which is why I'm asking for your feedback to see if it's worth continuing or not. I'm no stranger to deleting months of work if I can't produce good enough results.

I've attached the game below; it's a small download, just an HTML file and some images. You can get to the sex scene in like 2 minutes max, so please give it a try and let me know what you think. Is it interesting or does it feel the same/worse than Lilith's system? What should I change, add, or remove?

To be clear, I'm only planning to use this system for special scenes in the game. I did still write some good ol' fashioned sex scenes that aren't interactive. I'm attempting to get the best of both worlds here, though feel free to only play to the interactive scene since everything else is still highly unfinished.

You don't have permission to view the spoiler content. Log in or register now.
 

osanaiko

Engaged Member
Modder
Jul 4, 2017
2,663
4,942
I was interested to learn about the "Penn and Teller" magic concept - one of the primary ways that they (and other magicians) achieve the illusion of "magic" is by the investment of a unbelievable amount of time/energy in their tricks. Normal people would not even consider personally spending so much time in preparing, so they mentally (subconsciously perhaps) dismiss the possibility that the performer actually did e.g. practice some motion 20,000 times over two years in preparation and can now do something "impossible" consistently.

A specific example of this is the following story: Penn had a trick where he would get an audience member to pluck a card, memorize it, and return it to the deck. Then he would do some simple card manipulation and reveal the wrong card. When the audience member said it was the wrong card, it should be e.g. three of hearts, Penn would then lead them outside to the park, dig a hole through pristine grass, and pull the correct card out of a buried metal case. The secret was that he had actually buried 52 separate cases in various locations in the park 6 months previously (to allow the grass to regrow), and just had to dig at the correct location.

Anyway, the relation of this idea to your post is: in my opinion the way to make an *amazing* version of this type of randomized / auto-generated sex system would be to write an unreasonable number of text variations and/or flexible routes - so many that only a heavy user would have a significant chance of encountering the same phrases more than once.

Update: downloaded your demo, and had a look. My thoughts:
- nice writing
- played one run through the mary sex scene
- wow, that's a lot of options for hands in the sex scene. nice!
- sad there was is no repositioning, that's one of my favorite parts, to explore the possibilities. (and I see in your caveats section that you did that intentionally. sadface)
- it's hard to tell from a player point of view if there are any specific results from your efforts to "build the scene based on character personality". i guess that would only come from the contrast to lilith's throne where (as I understand from your explanation, it;s been a very long time since I played it) every scene ends up feeling similar because the same text / progression / options are used for various NPCs. As far as that goes, I guess you must have been successful to some extent if you have customized this mary scene to her character, but we'd have to see several more scenes with different chars to experience the results.
 
  • Like
Reactions: Eroniche

Eroniche

Formerly 'TrueSandbox'
Dec 23, 2019
5
8
Hey, thanks for the reply! Interesting analogy with Penn and Teller. I've been a fan of their work, but I haven't heard about how much time they invested into their tricks. The example with the buried cases is just absurd; no wonder nobody would expect that to be trick, though I suppose investing half a year into waiting for plants to grow is still easier than practicing one trick 20,000 times.

And yeah, exactly as you said, the end goal would be to make it interesting by brute force: just write so many variations/routes that most people won't ever explore it all (while at the same time being wary of overwhelming the player with too many options...)

I appreciate that you tried the demo!

Eventually there'd be multiple locations where you can initiate scenes for a character, and they'd all come with a specific position, so I'm hoping that can be a substitute for repositioning (and would admittedly take less work).

Your last point is what I was concerned about, so I'm glad you voiced it. It would be easier to tell the difference if I had demos for more than one character, but it does seem like that individual scene still needs more flair around it. Although I do contrast it to Lilith's Throne, it has to stand on its own merits to be worth anything, so I'll try to figure out what else is missing from the scene.
 
  • Like
Reactions: osanaiko

Quintillian

Member
Apr 15, 2019
134
251
Always nice to see someone else working on something like this. Good luck with your game!

As for feedback, I'm not an experienced writer by far, so I'll limit to my opinion as user. I did a couple of runs, there was some repetition here and there but overall great experience. Can't tell if the cozy feeling I got with the scene with the Angel was thanks to her personality or if all the writing for all npcs will give out that vibe though.

For the options, I'm found myself actually ok with there being no repositioning. Each action having its own position made the interaction flow more naturally, so I could just focus on what I felt like doing without the meta-gaming thought of setting up the scene for more actions to unlock. I still think, some actions should be dependant on others so you already seem to be at a good compromise since the human mind is amazing at filling it the blanks from imagination.

Finally, on the more technical side, I'm so curious how you are implementing this for my own selfish reasons. I'm thinking of adding a system like this on a game I'm making. I made a prototype using a Finite State Machine for storing and transitioning between states, and a Fuzzy Pattern Matching System for rules based decision making ( and also avoid nasty, nested, and convoluted if-statements).

And on a final, final, note. At the end of the day, I don't think there are shortcuts to the writing part. At most you could tag each piece of writing with a certain theme so searching for variation is easier, but you are still going to need to have that extensive corpus of authored content to begin with.
 
Last edited:

Eroniche

Formerly 'TrueSandbox'
Dec 23, 2019
5
8
Hey, thanks for the detailed reply! Honestly I wasn't sure if I'd get any useful feedback from this thread (first-time poster and all) so I'm grateful for you and osanaiko's input.

"cozy" is definitely the vibe I was going here, so I'm glad that worked. The other characters aren't necessarily cozy, but still fit within the tone of this game (leaning light-hearted, not dark or serious).

I agree, it's good to have actions depend on each other. My goal is to create "realistic" but interesting interactions, or at least to achieve two goals: one, that the available actions make sense in the situation (based on pose and what everybody's doing) and two, that the player gets to do what they want (so writing more available actions, some of which are gated behind conditions). You can't do whole-body repositioning but you can move your hands and other stuff around, basically.

Technical aspect: Although I hadn't thought of it in those terms, the system I have is indeed similar to using a Finite State Machine with Fuzzy Pattern Matching.

Here, the states are the locations of either character's hands, mouth, or penis, and you can transition between states if it physically possible (so regardless of mood, lust, or other subjective factors). Your available actions are just gated behind if-statements checking for locations, so like you can't start kissing her if you're already in the kissing state. Pretty straightforward.

The fuzziness comes into play with how the other character (Mary) reacts to your actions, so for example if she is angry (her internal anger value is above a threshold), then she will act resentful in her dialogue/behavior. So there are still if-statements checking for anger/lust values but I don't think it's too bad, though I'm sure there are better implementations I haven't thought of.

Overall it's just a giant switch-case statement to handle all of your available actions, some logic to semi-randomly choose a response for Mary from an available pool, then some other misc. logic that I won't bore you with, like increasing the player's lust gain over time or randomly failing Mary's action if her lust is too high.

Unfortunately yes, like you said, there's not really a shortcut to creating more content. It's an interesting challenge though, and after all if it was easy, everyone would be doing it.

P.S. It's cool that you're working on a project too! If you ever want feedback on it or whatnot, I'm always open.
 

Quintillian

Member
Apr 15, 2019
134
251
Hey, thanks for the detailed reply! Honestly I wasn't sure if I'd get any useful feedback from this thread (first-time poster and all) so I'm grateful for you and osanaiko's input.

"cozy" is definitely the vibe I was going here, so I'm glad that worked. The other characters aren't necessarily cozy, but still fit within the tone of this game (leaning light-hearted, not dark or serious).

I agree, it's good to have actions depend on each other. My goal is to create "realistic" but interesting interactions, or at least to achieve two goals: one, that the available actions make sense in the situation (based on pose and what everybody's doing) and two, that the player gets to do what they want (so writing more available actions, some of which are gated behind conditions). You can't do whole-body repositioning but you can move your hands and other stuff around, basically.

Technical aspect: Although I hadn't thought of it in those terms, the system I have is indeed similar to using a Finite State Machine with Fuzzy Pattern Matching.

Here, the states are the locations of either character's hands, mouth, or penis, and you can transition between states if it physically possible (so regardless of mood, lust, or other subjective factors). Your available actions are just gated behind if-statements checking for locations, so like you can't start kissing her if you're already in the kissing state. Pretty straightforward.

The fuzziness comes into play with how the other character (Mary) reacts to your actions, so for example if she is angry (her internal anger value is above a threshold), then she will act resentful in her dialogue/behavior. So there are still if-statements checking for anger/lust values but I don't think it's too bad, though I'm sure there are better implementations I haven't thought of.

Overall it's just a giant switch-case statement to handle all of your available actions, some logic to semi-randomly choose a response for Mary from an available pool, then some other misc. logic that I won't bore you with, like increasing the player's lust gain over time or randomly failing Mary's action if her lust is too high.

Unfortunately yes, like you said, there's not really a shortcut to creating more content. It's an interesting challenge though, and after all if it was easy, everyone would be doing it.

P.S. It's cool that you're working on a project too! If you ever want feedback on it or whatnot, I'm always open.
I was worried I had made the terms too abstract, but you totally nailed what I meant by Fuzzy Pattern Matching. A couple of more SFW resources you may find interesting though:

This goes into detail about the Fuzzy Pattern Matching. Worth a watch if you think that switch-case statement is getting too wild. I always recommend it to anyone wanting to handle complex if-else checks. Sad truth, is that 99% of players won't even know or care how deep this feature is, specially if the overall writting is subpar, but anything that makes devs/writers lives easier is a win on its own, in my opinion.

Second, for writing, if you haven't already, you should check out Emily's Short, The Annals of the Parrigues. It is a procgen book, so not entierly your case I think, but you can download the book for free on , and at the end of the book, there is a section on how it was made. Valuable stuff for anyone working text-based systems like this.
 

Eroniche

Formerly 'TrueSandbox'
Dec 23, 2019
5
8
I was worried I had made the terms too abstract, but you totally nailed what I meant by Fuzzy Pattern Matching. A couple of more SFW resources you may find interesting though:

This goes into detail about the Fuzzy Pattern Matching. Worth a watch if you think that switch-case statement is getting too wild. I always recommend it to anyone wanting to handle complex if-else checks. Sad truth, is that 99% of players won't even know or care how deep this feature is, specially if the overall writting is subpar, but anything that makes devs/writers lives easier is a win on its own, in my opinion.

Second, for writing, if you haven't already, you should check out Emily's Short, The Annals of the Parrigues. It is a procgen book, so not entierly your case I think, but you can download the book for free on , and at the end of the book, there is a section on how it was made. Valuable stuff for anyone working text-based systems like this.
Haha, I'll admit I did have to look up fuzzy pattern matching to see what you meant.

Thanks for the resources! I did actually watch the GDC talk video just now in its entirety (aside from skipping the very last math section), and it is insightful. I already use the core system of storing facts about the world and checking criteria for dialogue, but I hadn't thought of prioritizing score/specificity over random selection, which seems like a great way of making dialogue more directly context-relevant. Going to look into that.

It seems like a simple system, and of course you still have to write a lot of variations, but it's a testament to how much immersion you can add with what boils down to rule-matching. Hell, I've played those Valve games they mentioned but never really thought about this aspect of the games.

A nice thing about interactive fiction, at least, is that the monologue you wrote can't get interrupted by a sudden zombie attack. Not unless you intentionally program random time-based events, anyway. So our IF systems are easier to make, but still good to know.

You don't have permission to view the spoiler content. Log in or register now.

The Annals of the Parrigues seems interesting, and I have great respect for the ambition involved in creating an entire world (or procedurally generated towns), and the parts I read seem surprisingly unique while still feeling naturally written.

That being said, I'll admit I'm not too interested in procgen myself, as I still think handwritten feels more personal and engaging (I mentioned Lilith's Throne originally, which uses procgen to create infinite variations of NPCs that unfortunately all feel like the same few personalities). Still, it is an interesting resource, and again I appreciate you bringing them up.
 
  • Heart
Reactions: Quintillian