Looking Back
It’s been about a year since I wrote this thread on ideas I’ve tried that did/didn’t work: These didn't work for me but maybe they'll work for you. Here are some games I tried to make and failed.
Updates on that: I did end up revisiting the bunker survival game and it turned into Blueberry Bushwhacking. It is currently the most popular game I’ve ever made by a pretty significant margin, even compared to the non-fetish games I’ve made. ._. Inflation Roguelike turned into a short story that I posted on DeviantArt to limited fanfare but little more has come of it.
Outside of that update, though, I wanted to talk about three other games I’ve tried making and the various hiccups I’ve run into both mechanically and practically. Once again, I might revisit some of these in the new year.
Science Lab 13
You are a slime that has escaped containment. Sneak around, fill up guards and scientists, get deeper into the lab.
One of the things that went pretty well about this was the general architecture of this game. I have mixed results with state machines for NPCs but here it happened to work quite well. I did struggle a bit with inverse kinematics for tentacles and with player animations and model quality but the biggest cap I ran into was with how I’d put together the actual NPC behavior. The game started life as a Sims-like, so the scientists and guards would get hungry, tired, etc, according to their built-in needs, but this didn’t map well to the Metal-Gear-esque stealth-action game I had ultimately envisioned. I will probably borrow pieces of this for future games.
Into The Fold / Escape the Cycle
The SCP Foundation appreciates your willingness to participate. Previous subjects handling [redacted] experienced significant physical changes while playing [redacted]
This is a game-within-a-game. I wanted to take advantage of the simplicity of the text style from Blueberry Bushwhacking and do something semi-graphical. I’m fairly keen on self-inflation and growth and weight gain, but the split nature here of having to look down at one’s body didn’t really work or “feel” as good as I had hoped. It was also hard to really view how big the character was getting. On a few occasions I thought it might make sense for the computer to print ascii pictures of you from different angles, but this didn’t really work either. Important takeaway is that first-person isn’t necessarily great for inflation games.
Kaizo Taizo
This Dig-Dug knockoff is weirdly immersive. Where did everybody go?
The screenshots for this aren’t compelling, so here’s an image of the player avatar.
This game borrowed from the previous two, but took less of a slow text adventure / RPG form and was intended to be a faster paced shoot-em-up. A lot of things went pretty well: I quite like this stylized base model and the blend shapes on the face and body meant I could get a lot of character variations from a simple model. Unfortunately, the game feels pretty empty when I play it. Perhaps it’s just that it lacks ‘juice’, but I didn’t find the initial versions enjoyable and couldn’t think of any real way to add depth. I’ll still take the modular avatar approach, since that was great, but I don’t imagine I’m going to continue down this path because it’s very easy to get the gist of it from five minutes of play.
Key Takeaways and Learnings
- Modular avatars are a good way to get a lot of content. You run the risk of jank or “infinite oatmeal” (all unique but boring), but it helps tremendously with the character pipeline.
- Realistic is easier than stylized. It might seem counterintuitive, but the stylized approach is SIGNIFICANTLY HARDER than a realistic approach. If you buy two realistic assets they will look like they go together. If you buy two stylized asset packs, they won’t match nine times out of ten. If you’re making your own art and go stylized, then any assets you buy or make need to match that style or be modified to get there. You can reduce the impact by using the modular avatar approach above so you have less to make/modify.
- Roguelikes, text adventures, and RPGs work better for me than shoot-em-ups. Mechanically it’s easier to get an end-to-end playable game with shoot em ups, but the experience feels short lived and hollow, something that could just as easily have been a two minute animation. Roguelikes still are particularly interesting to me because of the possibility of emergent behavior.
- Breaking down a character model into pieces will make it much easier to simulate and work with, but people will hate it. I did this for Accident in the Alchemy lab and people hated the seams on the model. I think for my next game I might do it again for performance reasons.
Something I Could Not Do
Lastly, something I wanted very much to do but simply could not figure out how to make work is procedural inflation or procedural weight gain. All of the characters and avatars I’ve had so far have had baked shape keys which are interpolated through. Blueberry Bushwhacking had about ten different shape keys per character for inflation, liquid fill, penis size, breast size, belly inflation, and so on. I would like to use XPBD (extensible position based dynamics) to simulate the physics of a character filling up, which, coupled with a systems-based game, could make for an extremely intriguing evolving game with a lot of content that’s fun to explore and play for a long time. It is currently beyond my abilities as a programmer, or at least beyond my current abilities as a GRAPHICS programmer, but it is my white-whale for the new year.
