Author: Studio Ganzheit

I finally lost to the placeholder boss. Even though the remaining work is simple, I just couldn’t push through, so I took a break and worked on some game feature updates.

• Kay visibility: I made Kay’s scarf much longer and gave it a color that really pops. I went back and forth on what color to use, but thinking about Kay’s personality, disposition, and role, white felt right. Some users suggested jiggle bones or kawaii physics, which I interpreted as wanting physics-driven objects to stand out more — it’s a little ironic that it’s harder to understand what users mean when they use dev terminology than when they just describe how they feel. Anyway, I also increased the emissive value on the scarf material when the camera pulls back, so it reads better at a distance. Cranking emissive up close looks unnatural.

• Attune: I added a feature to read information from the environment, like reading signs. I thought it might help emphasize the game’s mood, and more importantly, I needed another way to deliver story in the threshold — the existing methods weren’t enough for various reasons. It’s a secondary feature though, completely optional for progression.

• Photo mode: while Kay looks at photos of her daughter, the player can take screenshots from a distance.

I also finished some of the gameplay for the boss’s final phase, but I think I need to swap out the placeholders for proper models before doing any more polish.

For full assembly, I’ll probably need to rewrite most of the story-related text from scratch. A lot of the dialogue was written while the story was still being built, so there are gaps and parts that no longer match. This is exhausting — I need to figure something out. I think I’ve been going at full throttle for too long. I need to start managing my energy better going forward.

Third phase of the last boss. That effect is 3 seconds long and took a week. One day went to getting a wisdom tooth pulled, about a fifth of another went to helping around the house while my wife wasn’t feeling well, and a day or two was lost to going down the wrong path. I initially exported the floor as skeletal animation from 3ds Max, advancing the animation incrementally — but with around 350 pieces, the framerate tanked. The thing is, what I actually wanted was a shockwave-driven collapse, not a physics animation. Completely wrong approach. So I just made 5 static meshes representing stages of the collapse, swapped between them, and hid the transitions with FX. For the tell effect, I went back and forth before landing on a slightly anime-inspired look. I put some care into the lens flare to keep it from looking too anime. Lens flares can’t afford to look low-res, but using a massive texture just for that felt wasteful, so I handled most of it in the material. I took a crunchy 256×256 lens flare texture, scaled it up with blur, then drew sharp radial black lines from the center in the shader to fake a high-res look. Not completely happy with the quality, but moving on for now.

I’ve been building nothing but bosses for almost three months, and staring at that placeholder tin can for over a month is making me sick. I want to finish this and get into full assembly. That’s how tedious this stretch is — I’m already thinking about the next game and the one after that. I mean, I am working hard, so why is progress this slow. Next up: the boss’s armor plates blow off and embed in the ground, and the player has to step on them to dodge shockwaves and lasers, get close to the boss, and destroy the core. I keep going back and forth — am I overscoping this because it’s the last boss, or does the last boss deserve this much effort? My head’s a mess.

Every evening I browse through the discovery queue and category pages, looking at games with surprisingly (relatively) strong numbers or ones with great visuals but unexpectedly low numbers. If there’s a demo, I play it. I’ve done this about 5,000 times by now. Here are some scattered thoughts from what I’ve noticed.

• It’s hard to find any fixed rules among games that sold in the 1k–50k range. The usual advice — your capsule should look like this, you need an early UX hook — I can’t really see any of that mattering here.
• But around the 10k mark, capsule quality does seem to hold up to a certain standard at least.
• What matters far more is fit. The game needs a point that can’t be scratched by anything else, and it needs to reach the players who care about that. The latter is especially critical. The former isn’t something you can easily manufacture — it seems to be decided when the game is born. It can develop and evolve during production, but even that feels like it was baked in from the start.
• So what’s the most important factor in achieving that fit for games in this range? Luck, I think. If you’re at least posting dev logs and entering Next Fest, luck does come around once or twice, big or small. That’s where the old saying applies: “be ready when luck comes.”
• The “readiness” here seems to be playability and build stability. This is one of the few areas that can actually be improved through effort.
• I’ve also seen games with major publishers and impressive visuals that sold shockingly poorly.
• The ratio of self-published games among those with meaningful numbers was higher than I expected. Just a feeling, though.

So what’s my conclusion?

• Increase iteration speed, ship faster, release more often — to test different game hooks.
• Without dropping quality — so I’m ready when luck comes.
• But then why is this boss taking so damn long.

Nearly done with the second phase of the last boss. That was genuinely hard. A lot of things didn’t work the way I expected, and what I could see was wrong didn’t always match what I felt was wrong, which made it confusing and easy to get lost. Now heading into the third and final phase. Truly the last of the last. Of course, once the last phase is done I still need to design the boss visually, but that should be less of a maze than the gameplay side.

I keep thinking about what I want to make for the next project, which I guess means this one really is in its final stretch. Even though there’s still 2–3 months of work left including the boss, plus testing and polish after that. Games I want to play keep coming up too, but my PC can’t keep up anymore — and computer prices have gotten so high I can’t even consider buying one. Never thought I’d be in this spot.

Finished building the weak point for the last boss, T6, and got the player able to attack it. The original design had no direct combat — the player would interact with devices on the boss’s body to destroy it while parkouring up. It’s shifted into something closer to a traditional action game boss now. Next up is the phase transition: when it gets back up, the shredders on its arms swap out for something else and it starts doing absurd attacks. I want to get the boss done and move into full assembly. The story is locked in, all the elements are in place, and I can finally replace all the placeholders — though there’s a good amount of polish ahead. Hoping to wrap up around summer and start testing.
I found out a bit late that the world is a sea full of nothing but manchineel islands. Not sure it would have made much difference even if I’d known sooner. It’d be nice to be an iguana that can happily eat manchineel fruit, but I wasn’t one. So I got on a raft and headed out to sea. I don’t know the faces or names of the people who enjoy my game, but maybe because of that, sometimes they feel like people on the same boat, and sometimes like the wind pushing it forward. That’s what my gratitude toward players means. Even if it never reaches them, it’s something I feel.

Still working on the last boss. The gameplay is gradually shifting from the original plan, but I’m liking this direction better. Unlike the previous one, this one’s design is tightly tied to the gameplay, so I’m building the gameplay first with a mock-up. Going through the core elements one by one — the boss’s basic attacks are mostly done, and next up is building its weak points. The legs are handled procedurally, so even if the quality takes a small hit, it made the animation work much easier.

The weather got hot out of nowhere. I thought my room would stay cool, but the moment the temperature went up it betrayed me. The monitors make it even worse. I have a circulator running but it’s not doing much — just adding noise. If it’s already like this, I have no idea how I’m going to survive the actual summer.

Working on the new boss. I’m building it with multi-legged locomotion so the player can climb its legs using wall jumps. The body is driven by skeletal animation while the legs are handled with procedural animation. I’ve done procedural movement plenty of times before, but I keep getting greedy with it and ending up rolling back. When you’re solving things programmatically, you unconsciously start avoiding skeletal animation, and I think that’s the number one way I overcomplicate things.

The two most common types of messages I get while making this game are marketing offers and music composition offers. Marketing offers — I can almost tell it’s a marketing DM just from how it opens. The pattern is always the same: no introduction, just a casual greeting with no mention of what they actually want. Composition offers come with a portfolio, but with zero consideration for the direction of the music already in the game, and I’ve never once found anything relevant in those portfolios. Some even push with “do you know who I am.” You don’t even know who you are — how am I supposed to?

Anyway, promoting a game is incredibly hard for me too. I try different things, introduce the game here and there, but sometimes I wonder — is my outreach any different from theirs? Am I reaching out to the wrong places? But then I’ve tried everywhere I can think of, so where else is there? A lot of thoughts like that. It actually gets harder a year after the demo launch, not easier. “Launching is the best marketing” feels more real by the day. But you still need some visibility before launch so people even know it exists. In the end, all these thoughts and attempts always loop back to the same place: just finish it. Focus on making it.

Boss work is taking longer than expected. Dental treatments keep eating into my schedule, wiping out a day here and there. The gameplay design for the boss is mostly done — I just want to build it out quickly, but that’s easier said than done. Can’t wait to get this done.

Finally done. This one took a long time too — not as long as the previous boss, but still. The video work I had to fit in along the way probably made it longer. It’s been the hardest stretch of this project so far. Not because the work itself was hard, but because juggling everything in parallel was stressful. Anyway, it’s done. I’ll handle tuning later during testing.

Life kind of feels like a long fight with things you can’t control. I’ve been the type who beats myself up trying to do something about what I can’t. I’ve also seen people who, when they can’t control somebody, find a way to grind them down or shove them into themselves, or just label them as an enemy. Looking back, I came to understand it as fear — fear of the things you can’t do anything about.

Tangent aside, the numbers are something I can’t really control either. They refuse to budge no matter how hard I push, then suddenly jump when I’m doing nothing. The good thing is that I’ve gotten better at letting go of what I can’t control, so it naturally flows into “just focus on finishing the work.” Next up is a boss called T6. Last boss of the game — wonder how rough it’ll be, or maybe now that I’m not glued to the numbers, I can wrap it up quickly.

Boss hit reactions, boss attacks, and player hit reactions are mostly done. Just the death sequence and phase transitions left. The end is finally in sight — this one was rough. Since even the movement is built as patterns, instead of a behaviour tree I just need to decide what state comes next at the end of each one. Assembly should take a day or two, tuning maybe 3–4 more days.

I also fixed two bugs that had been sitting around for almost two years. Both turned out to be Unreal bugs. First one: a scene component attached to a movable character would occasionally behave as if it were static. It only happened sometimes and I couldn’t reliably reproduce it, so I never managed to track down the cause — until I caught it while building the new boss camera. Attaching a fresh scene component fixed it. What’s strange is that if you attach a static mesh component to a static mesh actor in the background, the component gets added as movable, but if you attach a component to a movable actor like a character, it sometimes attaches as static. Second bug: in shipping builds only, Niagara beams were getting culled. I’d only seen one report and couldn’t reproduce it — until today, when it suddenly showed up in my own shipping build. I tried everything but couldn’t fix it. People online say beams aren’t reliable, so I replaced them with spline meshes. Still need to test, but it feels great to clear two bugs that old.

The first boss took 2 months. This second one has taken 5 weeks so far including modeling, and maybe ¼ to ⅓ of that time went to trailer and PR work. So roughly a month of actual boss work, with 3–4 days left. Bosses feel like tangled bundles of video clips — not the most fun thing to build — but the build time is going down, which is something. I want to finish the last boss in May and wrap the build in June. I’ve picked up a lot of know-how on this one, so for the next game I’d like to try something simpler that I can finish in 1 to 1.5 years.

Submitting to a game show taught me a lot. I’d been pretty naive about it. Turns out game shows are closer to an advertising market than a place that gives indies a leg up. Submission costs are steep and the math doesn’t really work for a game my size, so I ended up dropping it. I just released the trailer on its own. Feedback’s been mixed — it’s a tone I hadn’t tried before. I’ll do better next time.

Anyway, the end is getting closer and that feels good. Whether it ends well or not, finishing something is its own reward.

Boss patterns are mostly done. Hit reactions, tuning, and phase assembly are next. Video work every week means dev progress is slower than I’d like. I submitted something with a different feel than before to a game show. It’s exclusive so I can’t share it yet, but I’ll post it separately whether it gets in or not. Too much work went into it to just sit on it. Building bosses has always been the hardest part for me, and it still is. One more to go after this and I’m not sure when I’ll get to it. It’s not that it’s difficult — it’s more that this one part of dev feels a bit removed from what I enjoy. I love making games, but actually building out the bosses is the one piece that gets to me. Anyway, still grinding away at it.

I tried diversifying my social media a bit by posting a YouTube video in a new style, but it didn’t really work. I also tried a shorts with a different style, but the response was worse than usual. Oof. YouTube is probably going to stay an archive for me. Making the videos, uploading, and responding to comments took 2-3 days out of the week.

The final boss of the Threshold is still moving, slowly. The idea here is to make Bulletwalk a central part of the fight, both using it and figuring out how to recharge it. About two patterns left, I think. There’s still player hit detection, AI, and tuning to do.

Even so, after about five years of development counting the previous game, things are starting to accumulate — pipelines, code, whatever — that I can reuse going forward, and that makes the work feel worthwhile. UE4.27 is getting comfortable in my hands. UE5 feels too heavy for my machine and I don’t trust my optimization skills, so I’m imagining maybe switching to UE6 someday after I make enough money to upgrade my computer. Or I might just stay on 4.27.

Boss work will probably keep moving slowly next week too, since I’ll be preparing a video for a game show submission. It’s my first time trying. We’ll see if they take it.