Let's Talk Game Mechanics!

[Saki_Sliz]

In terms of mechanics there are styles that are proven to work and be lots of fun, like Trainers and life-sims, I guess thats why we see them so often.

agreed, and I think the other important aspect is that, due to the nature of these games and mechanics, a greater percentage of the artwork can be for lewd purposes. IE a diablo style game (dungeon crawling, skill tree and loot) would need to have a lot of work done for the various weapons or loot, which does not contribute to the lewd aspect of the game. meanwhile a Visual Novel which can reuse common art assets (ie standing pose) means you can get better bang for your buck out of your art assets, and from what I have heard from other people, art tends to be one of the greatest limitation factors in making games. either one can't get art, or because they can't get art they have to rework their design around this limitation (ie text based game).

I know when I first wanted to get into nsfw game design, I was trying to work around figuring out game mechanics I both liked and could get the best bang per buck out of art assets. But this was years ago at this point and I don't remember much of my notes from back then. Not to mention, there is a range of art to consider, UI, character, background/world, and depending on how well you can access each of these, you may considered the bang per buck issue again depending on game design choices. for example the rpg would need a lot of world art, but can be simple if you just use as asset pack, but a nv that needs custom locations could be more challenging.

while having a world to explore is cool and all, I think it is one of the more challenging game to support just because of the art need, but game systems that focus on characters (vn's, trainers, life sims (aka advance nv's) ) can focus on the art that matters, and that is why I think we see them a lot. Also text based games, those are common as they are the easiest to make (at least effort to results is pretty direct) and have the lowest need for art.

 

[DuniX]

at any one time while playing a game, you can either be enjoying the game mechanics, reading the story, or enjoying lewd content. What I want to make, and what it seems other people often complain about is, we want everything, at the same time. great game mechanic, great story, great lewd content. The issue I find is that most of the time, you can only do one thing at a time, at best two. But when you try to mix them, it does not make something better, it just makes it two lesser things, half game, half lewd.

I am not so sure that that is the case.

If you look at what makes stories interesting Conflict, Drama, Struggle they already map to games pretty well, Plot is basically equivalent to Progression, the accumulation of power for you as well as other characters/enemies. As for Porn is a Powerful Motivator by itself as an Reward. The fact that we can go through all the pointless grinding just to get to then next sex scene should be obvious enough.
All can be mixed pretty well while having good pacing.

I think the problem is more the premise/setting that can be faulty that does not fit the game mechanics and causes friction.

this may be a case where the fighting requires the player's full attention (time sensitive, reactive), and reading acts as a distraction,

It's fine in a turn based game. I think dialog in combat could be great to make the battle more meaningful and the characters more humanized then just some expendable units.
I want them to struggle, fear, rage, cry, despair while I slowly crush them to death.
Adding a surrender mechanic which you can convince the enemy to give up I think is also interesting.
Surrender of course means more porn.

 

[Saki_Sliz]

I am not so sure that that is the case.

If you look at what makes stories interesting Conflict, Drama, Struggle they already map to games pretty well, Plot is basically equivalent to Progression, the accumulation of power for you as well as other characters. As for Porn is a Powerful Motivator by itself as an Reward. The fact that we can go through all the pointless grinding just to get to then next sex scene should be obvious enough.
All can be mixed pretty well while having good pacing.

I think the problem is more the premise/setting that can be faulty that does not fit the game mechanics and causes friction.

It's fine in a turn based game. I think dialog in combat could be great to make the battle more meaningful and the characters more humanized then just some expendable units.
I want them to struggle, fear, rage, cry, despair while I slowly crush them to death.
Adding a surrender mechanic which you can convince the enemy to give up I think is also interesting.
Surrender of course means more porn.

all true stuff, I'm just saying you could come into to a problem (ie making a game) at a wrong angle, or as you say, a problem with premise/setting, but coming at it from another direction works great, ie turn based or something I played with, a slow mo mechanic so you can see nice lewd animation effect

 

[lancelotdulak]

We arent talking about the elephant in the room that kicks all our asses. Rendering time. And to be clear i dont consider vn's games at all. Theyre litereally visual novels and ive damned near gotten carpal tunnel clicking through text and scenes i didnt care about and knew i wouldnt care about. But there are a genrewe call vn's like DMD that ARE games.
I think this is what makes games like Big Brother extremely successful to the point that type of game is a subgenre of its own.
DMD has real choices , is well written .. but thats the reason it takes so much work and time to make even though the author obviously has top end hardware etc and does it full time (i believe has a team now?) and still has to be very smart/clever about keeping the number of scenes down.
I think there are a lot of talented creators here who get hit in the head hard with this limitation. I think what we all need to figure out are game mechanics systems that dont exponentially increase the renders with every choice you make. In something like UE4 it doesnt matter.. just some different text or voice. In games with rendered scenes it means everything

 

[Saki_Sliz]

We arent talking about the elephant in the room that kicks all our asses. Rendering time. And to be clear i dont consider vn's games at all. Theyre litereally visual novels and ive damned near gotten carpal tunnel clicking through text and scenes i didnt care about and knew i wouldnt care about. But there are a genrewe call vn's like DMD that ARE games.
I think this is what makes games like Big Brother extremely successful to the point that type of game is a subgenre of its own.
DMD has real choices , is well written .. but thats the reason it takes so much work and time to make even though the author obviously has top end hardware etc and does it full time (i believe has a team now?) and still has to be very smart/clever about keeping the number of scenes down.
I think there are a lot of talented creators here who get hit in the head hard with this limitation. I think what we all need to figure out are game mechanics systems that dont exponentially increase the renders with every choice you make. In something like UE4 it doesnt matter.. just some different text or voice. In games with rendered scenes it means everything

It's not just rendering time, but the whole process of, ok we have a script, now we need 24 new images, we have to find assets if it is any new location/clothes/characters and anything else that needs to be prepaired before rendering even begins. In my case a few years ago when I focused mostly on 2D art, to go from character description, to fully drawn final quality nude in 12 hours and then clothed an animateable rig with some animations in another 12 hours to be pretty good. but then if I only have 2 hours a weekday, and 4 hours weekends, it takes a full week to make a character, and if someone wants a lot of character for a story, that is at least half a year's worth of work planned out. 3D from scratch is even more time, but far more reusable, but doesn't include things like multiple hair/clothe options, or environment art. 3D with daz is a lot faster, but then you have to shift through available art resources (stores, secrete Russian servers, etc.) to try to find something that matches your vision.

With that DMD having a team, I imagine it is lot like getting a bigger pay check. your spending increases with your income, so you never save money. in the case of a team, with more on board, you can spend more time trying to do what you couldn't, custom assets, prototypes, etc. so that is more time being wasted.

Since we are in a discussion about game mechanics. Are all yall familiar with the concept of developing the Minimum Viable Product? Where you make a prototype that focuses on just the core, core, core aspect of the game. IE the Mario games would have just been a cube that moves and jumps, maybe lands on other cubes to kill them. Everyone is so focused on making an Alpha 0v0.0.0.0.0.1.a preview intro scene, that very few actually make a prototype of the game first. But then, since most of us base our games off of inspiration from other games, from mechanics that we know already work and are popular, everyone moves on to the next step of, does this system work with the context of my story and my art. But most alphas are just straight up stories, meanwhile I mentioned this to Cul, I was just thinking of making a simple comic with simple words "bitch bitch bitch, me me me" in the text bubbles to communicate the core idea across, to see if that alone is enough context to make a demo or game fun but understandable.

 

[DuniX]

Are all yall familiar with the concept of developing the Minimum Viable Product?

It kinda doesn't work for porn games.
People come here for the porn, otherwise if there was no porn it would be pointless to release here.

However I do want to see projects that focus on the gameplay and making that work first in to a finishable state and adding placeholders for the sex scenes with image packs or using Illusion engine.
I like projects like that recent Dungeon Lord game or games like Trap Quest with multiple image variants.
Things like renders or commissions can come later.

 

[Saki_Sliz]

It kinda doesn't work for porn games.
People come here for the porn, otherwise if there was no porn it would be pointless to release here.

However I do want to see projects that focus on the gameplay and making that work first in to a finishable state and adding placeholders for the sex scenes with image packs or using Illusion engine.
I like projects like that recent Dungeon Lord game or games like Trap Quest with multiple image variants.

understandable, I wasn't too sure if it may be because everyone is focused on trying to get a patreon up and running first, or if everyone wanted to make something presentable first before sharing their work. I know I myself don't share too much. The only real MVP I've made is just a few python script to see if base game mechanics work in the console like I think they should. ie testing an ai's ability to make calculated risks or auto randomized effect loops that may be used in a future project or something. What few team project I did try to be a part of, we tried to set up place holders while we we worked but that was about it.

 

[khumak]

We arent talking about the elephant in the room that kicks all our asses. Rendering time. And to be clear i dont consider vn's games at all. Theyre litereally visual novels and ive damned near gotten carpal tunnel clicking through text and scenes i didnt care about and knew i wouldnt care about. But there are a genrewe call vn's like DMD that ARE games.
I think this is what makes games like Big Brother extremely successful to the point that type of game is a subgenre of its own.
DMD has real choices , is well written .. but thats the reason it takes so much work and time to make even though the author obviously has top end hardware etc and does it full time (i believe has a team now?) and still has to be very smart/clever about keeping the number of scenes down.
I think there are a lot of talented creators here who get hit in the head hard with this limitation. I think what we all need to figure out are game mechanics systems that dont exponentially increase the renders with every choice you make. In something like UE4 it doesnt matter.. just some different text or voice. In games with rendered scenes it means everything

I think once you have decent hardware, render times aren't such a huge issue unless your game is heavily focused on animations (like Being a DIK). You either want a second computer to do the set up/posing/lighting/etc for everything while your faster machine is rendering or you want to just lump them all together and render them as a batch while you sleep. With decent hardware your render time per image is probably less than your set up time except for things like animations that have tons of images that are nearly identical (and only require minor tweaks during setup). So rendering them while you sleep would mean you spend zero time waiting for renders.

In my case render times are an issue because my hardware sucks. A complicated scene for me will usually take longer than 2 hours to render. Sometimes more like 8 hours. That's per image (1050Ti 4GB). Someone with a 2080Ti could probably do that in more like 20 minutes per image. So it takes me a week to render a scene that someone with good hardware could do in a day.

For me the far bigger time sink is figuring out the story. I will start off writing out some dialog for a scene I'm working on but then I'll replay that scene over and over in my head over the course of several days until it feels right. I will replay the whole thing in my head from the player perspective and when I come across something that doesn't make sense or just doesn't feel right I'll scrap that scene and redo it until it does feel right.

The VNs you mention that you don't consider games because of the weak story are like that because the dev didn't spend enough time on the story. I do the same thing with a lot of games and just skip through the story without really reading it once it becomes obvious that the story/dialog are not much good. If your story sucks then you might as well just release a CG with a bunch of images and no actual dialog or gameplay at all.

For the games I actually get drawn into I consider the story more important than the renders and I read through everything. Bad renders will kill a game for me as well but for a good game the story is the main course. The renders are the icing on the cake.

 

[Saki_Sliz]

In my case render times are an issue because my hardware sucks. A complicated scene for me will usually take longer than 2 hours to render. Sometimes more like 8 hours. That's per image (1050Ti 4GB). Someone with a 2080Ti could probably do that in more like 20 minutes per image. So it takes me a week to render a scene that someone with good hardware could do in a day.

I try to keep render times between 40 seconds and 4 minutes just in case I want to do animations, but I know a lot of the time in testing renders can take upwards of 12 minutes, but I can only get these times through various optimization techniques and a GTX1080TI + 8 thread CPU, and a second computer to render other frames (I use blender so I can use multiple components and computers to render). The idea is that the math works out such that, if you are doing a 24 frame per second animation, each minute it takes to render an image = days to render a minute of animation for a single computer (half that if two computers and so on). so keeping things below a minute is critical for longer animations, which is what I prefer to mess around with. so with it taking 2 hours, or 120 minutes, it would take a third of the year to render 1 minute of animations, assuming 24 frames per second, and not just using a simpler key frame only animation which can be between 3 to .3 frames per second (an 8 to 80 x speed up in time)

I will replay the whole thing in my head from the player perspective and when I come across something that doesn't make sense or just doesn't feel right I'll scrap that scene and redo it until it does feel right.

or worse you just give up on the game like i do cuz I can't perfect everything.

 

[SerokMinmas]

We arent talking about the elephant in the room that kicks all our asses. Rendering time.

Is it really that bad? I haven't messed with DAZ and 3D stuff yet (It's on my list) But my dad was a designer and used a bunch of 3DSMax back in like 2009, rendered some pretty amazing scenes in a few hours, of course the resolution and detail wasn't all that good because 2009 but I was under the impression render speed improved over 11 years. But I guess that the increase in resolution and detail countered the increase in performance.

And to be clear i dont consider vn's games at all.

Well they are on the fringe of what could be considered games, you could make a RenPy "game" with absolutely no choices and the question of how gamey is the game is valid, but once you get choices and options and branching paths I think the discussion is settled and they are indeed games.

I dont really care if the no-choice RenPy games are considered a game or not, since the community the developer will be part of will be the same as the one with hundreds of options and branching path.

But by that metric, CYOA books aren't games since the communities are different, but they are basically physical versions of VNs.

Guess the actual answer is fuck it, they are part of our community, the resources to write and publish their things are the same, so in the end it doesn't really matter.

In my case a few years ago when I focused mostly on 2D art, to go from character description, to fully drawn final quality nude in 12 hours and then clothed an animateable rig with some animations in another 12 hours to be pretty good.

Funny how you can never escape from art taking time, either in the computer with rendering times or in 2d handmade art, its always going to take a lot of time.

I try to keep render times between 40 seconds and 4 minutes just in case I want to do animations, but I know a lot of the time in testing renders can take upwards of 12 minutes

That is for 1080p full scenes, with scenery, global illumination and such?

For me the far bigger time sink is figuring out the story. I will start off writing out some dialog for a scene I'm working on but then I'll replay that scene over and over in my head over the course of several days until it feels right. I will replay the whole thing in my head from the player perspective and when I come across something that doesn't make sense or just doesn't feel right I'll scrap that scene and redo it until it does feel right.

Making things feel right is something I struggle with a lot, writing is way more difficult than most people think, lately I've gotten a lot more respect for writers.

 

[Saki_Sliz]

That is for 1080p full scenes, with scenery, global illumination and such?

Full path tracing, and I go for 1440p since that is my monitor size. I have thought about 4K but it takes too long. I can understand renders taking much longer, but I do a lot of work to optimize speed, managing light bounces, sampling techniques, denoising tools, custom shaders, background script optimization, and using command-line interfaces. Without all that work a default render would take 2 hours, and this is why I use blender not daz, just because I know how to do this with blender where as I can't have as much control with the default iray render. One could argue a lack of quality, but it is only really noticeable when compared against a ground truth other than that it is imperceivable.

 

[khumak]

Full path tracing, and I go for 1440p since that is my monitor size. I have thought about 4K but it takes too long. I can understand renders taking much longer, but I do a lot of work to optimize speed, managing light bounces, sampling techniques, denoising tools, custom shaders, background script optimization, and using command-line interfaces. Without all that work a default render would take 2 hours, and this is why I use blender not daz, just because I know how to do this with blender where as I can't have as much control with the default iray render. One could argue a lack of quality, but it is only really noticeable when compared against a ground truth other than that it is imperceivable.

You have a relatively fast GPU as well though. If I render anything for less than 20 minutes it looks absolutely terrible. With a scene optimized to reduce complexity and memory utilization I can get some simpler scenes to look pretty good in 20-30 minutes but I have to do things like restrict myself to only a single light source, remove all reflective surfaces, delete any objects that aren't really necessary and are just there for decoration, etc. A render like this one with multiple light sources, reflections, etc takes me over 8 hours to render and even then it still has some fireflies unless I run it through a denoiser:



Edit: Was curious so I recreated a similar scene (didn't have the one above saved anymore and this is what it looks like for me after 4 mins). It's worth noting as well that the image above has several light sources to brighten up certain areas while the one on the bottom only has a single spotlight as a light source so it's actually easier to render than the one above. This is the same characters and the same room but not quite the same camera angle and different lights:

 

[Saki_Sliz]

Edit: Was curious so I recreated a similar scene (didn't have the one above saved anymore and this is what it looks like for me after 4 mins). It's worth noting as well that the image above has several light sources to brighten up certain areas while the one on the bottom only has a single spotlight as a light source so it's actually easier to render than the one above. This is the same characters and the same room but not quite the same camera angle and different lights:

yeah, one of my optimization techniques is to minimize lights, I only really use 3, with 1 major key light and 1 fill light. Infact I'll often render the character and the scene with different lighting so that each look their best but take the least amount of time. I also minimize indirect lighting, and instead fake it using a fill light, which is extremely soft (meaning the light is physically large, like 8 to 15 feet in diameter, usually less than 200 watts of energy).

I took your 4 minute render and ran it through blender's denoiser and got this

you will notice some errors on small things, like the lamp, and the wall is smooth, you can't see the texture of it, but assuming you didn't know about these details or looked at the image too long, it may be passible.

If possible, if you can render a normal map of the image, the denoiser can be even more accurate, but it can still have artifacts and actually usually the simple denoise shown above is what most people prefer.

Blender uses the intel denoiser, and there are even better denoisers out there on the internet, I think blenderguru the youtube has a nice video covering different denoisers.

denoising certainly helps with saving time. I also use branch path tracing, rather than just path tracing found in iray and most other engines. branch path is slower, but better quality, and while lower samples look much much worse than the same sample count for normal path tracing, the denoiser seems to do a good job with both type equally, not to mention blender recently came out with an auto sampling that will dynamically continue to sample areas that need more samples, so it is smarter not just throwing more samples at the whole image. Not to say blender is great, I do think iray/daz does better glass than blender out of the box, but with a bit of work and customization glass can look pretty good in blender.

edit, you said it takes 2 to 8 hours without a denoiser, not sure what type you use but I thought I would run your first image through it as well.

again I think it smoothed the wall texture, and again that can be something corrected if a normal map was also available. I don't know if daz is able to output a normal map of the scene.

 

[DuniX]

Why are we talking about rendering in a thread about mechanics?

 

[Saki_Sliz]

Why are we talking about rendering in a thread about mechanics?

I think it was because it was talking about how rendering times can impact game design choices, such as if it takes so long to make art, then to get the best bang per buck out of your time, you want to design a game and game mechanics that limit the amount of art needed, thus certain game types are more common than other types, ie vn's and life sims vs a diablo like game or simple vs the amount of time between patreon updates and making sure you have enough content each update to keep the masses happy. Art is a major bottle neck in game design, accessibility, cost and time being the issue. Trying to minimize these impact only serves to help any project.

 

[DuniX]

I think it was because it was talking about how rendering times can impact game design choices, such as if it takes so long to make art,

It's a moot point when you consider engines like Honey Select or Unreal/Unity or even Blender has real-time shaders.
To me waiting for render time is going backwards, and it will still look like shit since you don't have the control to customize things perfectly.

 

[SerokMinmas]

Why are we talking about rendering in a thread about mechanics?

Yeah... We might have gotten a bit off-topic here, sorry about that, we should probably start a rendering thread, or move this discussion to some already existing one.

Because I do find this interesting but I wont reply to them here to try to stay on topic.

 

[Saki_Sliz]

It's a moot point when you consider engines like Honey Select or Unreal/Unity or even Blender has real-time shaders.
To me waiting for render time is going backwards, and it will still look like shit since you don't have the control to customize things perfectly.

Honey select is used for art, but then there is the concern that all HS art based games look the same (kinda like how RPGmaker games or hentai games look the same, heck even many daz games start to look the same if the artist is still new). There are other similar tools/games that generate 2D art instantly (screen shot) but I can never remember their names.

Unreal/unity takes set up time (ie to make a 3D model or make it compatible) which compared to daz can take much longer, at least from my experience trying to make custom 3D assets for unity projects. And depending on the artist, one may not be able to get the look they want. I myself like a comic book look, but don't like how it looks in real time, but slower compositing in blender matches exactly what I want only taking a few seconds (but now my game has a bunch of images and takes up way more memory than a true real-time solution would).

Blender has EEVEE, which is great, but it sucks that in 2.8 they removed the blender game engine (I don't know if you could export the game, only trick blender using the Comand line intefraces to fake making blender a game window). I know that the engine is in it's own development thread and has gotten some major updates and may be readded after 2.9, but again I don't know if that is an exportable format.

So I agree, real time art may be better and does save a lot of time in larger projects (ie AAA titles) but the amount of set up time per asset eclipses the amount of time it would take to make a few renders, so it doesn't work for smaller or quicker projects. The exception of course being to export daz assets to game engines, but I have no experience with this to know the set up time. But the limit to this is a loss of control (aka quality limitations) and the art's skill's which as far as I can tell most dev's are not artist and most artists are not devs and it is hard to get the two to commit to a project together.

 

[khumak]

Dialog is something I've been thinking about doing something a bit different with for scenes that are repeatable. I can't think of any cases where someone else did something like this but maybe there are some examples out there. My idea for a repeatable scene with dialog starts off normally with dialog playing out normally the first time where you're actually reading what the characters are saying to each other.

The next time you go to that scene though, instead of playing through dialog you've already read I would do something more like what you see in a comic book with some sort of little dialog bubble type of graphic for a border and then some flashback type images showing what that character is talking about. So instead of a long dialog rehashing a bunch of stuff you've already seen that most people are just going to hold down the control key to skip you get a few flashback style images for a scene that character has been involved in. These scenes could even be randomly generated without needing to even write any dialog for it. This would basically simulate small talk for repeat scenes without having to either replay identical dialog or write a bunch of alternate dialog for the same scene.

If you get to a point where a repeatable scene has additional content unlocked then it would play out with normal dialog again like any other scene that you encounter for the first time.

I haven't tried actually implementing a scene like that so it could very well turn out to sound better in my head than it will feel once implemented.

 

[DuniX]

The next time you go to that scene though, instead of playing through dialog you've already read I would do something more like what you see in a comic book with some sort of little dialog bubble type of graphic for a border and then some flashback type images showing what that character is talking about.

You mean something like The Sims conversions? That could work.
There are some games with brief skip scenes where they just describe what happened when there isn't anything interesting happening.
You see this in Raising Sims style games sometimes.

One Idea I had is to make an actual Procedural cutscene flashback. Since it's a visual medium and there can be parity between actions you can do and actions in the past we can show them based on a log of events.
You can even add character interpretations on what happened(telephone game) and emotional reactions for the events.