This Game Took Me 2.5 Years To Make

In 2023, I downloaded the Godot game engine, and started tinkering with an idea inspired by some of my favorite movement games. The melee combat of Ghostrunner, the speedrunning of Neon White, the deep and satisfying movement kit of Ultrakill, and the vibes of Quake. For my game, I worked on a movement and combat system where each mechanic was carefully woven together, where the blood of your foes powers your deep and satisfying movement kit. A game where blood

equals speed. I never could have predicted the series of events that happened next. over

equals speed. I never could have predicted the series of events that happened next. over

100,000 demo downloads, some of the biggest streamers playing and talking about the demo, and best of all, a passionate group of players forming that actually loved the game and were wanting to see it released. 2 years later, I finished it.

[Music] Bloodthief is the game I've always wanted to make. I started basically only knowing how to code and learned everything else along the way. From the inner workings of the Godot game engine to making 3D art, the story, the lore, and most importantly, the game design required to actually make this project fun to play. I've quit my job, done all my own marketing,

and basically put my whole heart and soul on the line to make this project happen. And today, it's finally done, and it's for sale on Steam. This video is about the incredible terrifying journey I've gone on to bring my game to the world. Starting from the very beginning. [Music] So,

the original idea for Blood Thief was pretty simple. Hotline Miami plus Quake. Taking the

super fast one hit and you're dead gameplay loop of Hotline Miami and putting it in a dark fantasy 3D world. And from the very beginning, I wanted it to be a melee game with a big focus on crazy movement. I downloaded the Godot game engine mainly because it was free, open source, and honestly just looked like a ton of fun. I took the your first 3D game tutorial and by the end

of it, I felt ready to start. So, I popped open Blender to start modeling my player character and yeah, so one huge issue early on was that I just sucked at Blender. I'm really trying to figure out where my head was at when I gave the main character of my dark fantasy game like a t-shirt and jeans. But I guess it really just goes to show how far I've come in this whole process. So, it

and jeans. But I guess it really just goes to show how far I've come in this whole process. So, it

took a few tries and a lot more YouTube tutorials, but eventually I ended up with this. The first 3D model ever done for Bloodthief. Is it good? No. But you know what's even better than good? Good

enough. But now I still needed a player character. I had absolutely no interest in modeling another character, so I just chopped off the knight's head and gave him a new texture. And would you look at that? We got a player character. For animations, I use Mixamo for the night, which is pretty easy

that? We got a player character. For animations, I use Mixamo for the night, which is pretty easy to set up. But for the player character, I needed to do the animations by hand. And well,

so now I had a character and an enemy, so it was time to start making some combat. From the very beginning, I knew I wanted Blood Thief's combat to be ultra juicy. So I spent some time on making enemy deaths super dramatic. I made this nice blood splatter effect. I made their material flash when you hit them. And I made this absurdly complicated Godot animation tree to make it so no

matter what direction you hit them from, they always fly away in the correct direction. Okay,

nice. This is looking pretty good. So, back then, I had a theory I wanted to test. If you've ever played Hotline Miami, you know that the mechanics of that game are insanely simple. There's not that many ways to attack, and the main challenge of the game is really about how you move through the level. So, the way I saw it, I already had it so you could kill enemies. That was all the

the level. So, the way I saw it, I already had it so you could kill enemies. That was all the tech I needed to test that this concept was going to be fun in 3D. So, I threw together this little two room level, and yeah, it was boring as hell. Okay, successful experiment. I verified there's still work to do. The game just needed more stuff. The first thing I added next was just

some more movement options for the player. I added a slide ability, which provided the muchneeded Bloodthief protagonist feet reveal. And I also added some basic wall jumping. And I also added these archer enemies. Archer, I barely know her. Also, from the very beginning, sliding in Blood Thief has kind of worked in a funky way. First of all, sliding increases your speed, which doesn't

physically make any sense. And also, you actually can still fully rotate while you're sliding, which looks kind of hilarious in the game, but it feels amazing. And since that's all I really care about is the gameplay, we're keeping it. As soon as I added the sliding, a new problem arose, which is that enemies were basically brick walls, and if you slid into them, they just stopped you, which

felt terrible. So, I made it so enemies could be knocked down, which stunned them for a bit,

felt terrible. So, I made it so enemies could be knocked down, which stunned them for a bit, and they'd eventually lift themselves back up. At this point, the first enemy's animation tree was officially starting to look like a Gordian knot. But it will be fine. Surely this won't become a problem later. No, of course not. And don't call me Shirley. So, around this time, I posted the

problem later. No, of course not. And don't call me Shirley. So, around this time, I posted the first ever devlog for Blood Thief, and it got way better traction than I was expecting. So,

I was feeling great at this point. Time to add more features. So, like I said, I also added wall jumping. And naturally, this led to me trying to do these like jumps off the wall and perform these

jumping. And naturally, this led to me trying to do these like jumps off the wall and perform these high-flying attacks. But it was actually really difficult to actually pull off hitting an enemy

high-flying attacks. But it was actually really difficult to actually pull off hitting an enemy out of the air. And so that's when I added this air dash to the game. Feels good in my opinion.

I also added a bow and arrow at this point, which bow and arrows are cool. Players will love this.

So at this point, there was one glaring issue with the game. My level design looked like this.

basically just giant cubes. Thus, we now enter the part of this video titled Guy who's never designed a level in his life tries to make a game where good level design is fundamental to the success of the project. Okay, so because I'm a programmer by trade, my first attempt at making levels was to try to use geometry nodes to try to make it easy to procedurally generate levels. And I actually

got pretty far with this approach. Too far, one might say. But then something happened around this time. I downloaded the legendary Arcane Dimensions mod for Quake and played maybe the greatest video

time. I downloaded the legendary Arcane Dimensions mod for Quake and played maybe the greatest video game level ever conceived by anyone, let alone a video game mod. Ad Sepiler. It completely swept me off my feet and gave me probably my favorite 45 minutes of gaming I'd had in years. And it

was in that moment that I knew that procedural generation was just not going to work for Blood Thief. What makes games like Quake work so well is the handcraftedness, the attention to even the

Thief. What makes games like Quake work so well is the handcraftedness, the attention to even the finest details. So, I trashed those freaking geometry nodes and went back to the drawing

finest details. So, I trashed those freaking geometry nodes and went back to the drawing board. I couldn't help but wonder, well, if Arcane Dimensions was a mod, how did they do it? Turns

board. I couldn't help but wonder, well, if Arcane Dimensions was a mod, how did they do it? Turns

out, pretty much all Quake Mappers use a free and open- source software called Trench Broom. And lo

and behold, there's a Godot plugin for converting Trench Broom maps into Godot maps. I set my project up and the difference in quality of my maps immediately skyrocketed. It was just easier to block stuff out, create interesting layouts, and just add dynamic interactions. I also started watching a bunch of tutorials on making retro textures. And this is when you could start to

see whispers of Blood Thief's aesthetic starting to form. And for the first time, I could actually see a path towards making something actually pretty good. And so I continued on for 5 months.

So, at this time period, I'd gotten myself in some kind of a rut. After

working on the game for a better part of a year, no one had played it besides me, and it was just really hard to tell if I was on the right track or not. So,

I texted my friends and asked them if they wanted to play. They all came over and I'm just going to slide in. Was that like one hit? I got them all. That was sweet. use shift

as like an ultimate shield, but it's like, oh, this is so, oh my god, I have obliterated. It

got to a point where we're all kind of just making fun of the game together and enjoying how goofy and jank it was. And even though the game itself sucked, it really was just fun as hell seeing my friends interact with it. Plus, most importantly, for the first time in months, I had tons of new ideas for things to work on to make the game better and more fun.

One of the biggest takeaways I got from the day was that slide and air dash were too OP and could just be spammed at will, which created this super trivial gameplay that wasn't really particularly interesting. This is when I came up with the idea for the blood bar, which basically made

interesting. This is when I came up with the idea for the blood bar, which basically made it so all of your abilities in the game were powered by blood, which you got from killing enemies. Just this simple change did so much for making the gameplay more interesting. It added

enemies. Just this simple change did so much for making the gameplay more interesting. It added

a sort of economy to battles where now instead of just spamming everyone with your abilities, you had to be a little bit more wise about who you chose to dash and when you chose to slide.

And with those changes, for the first time since starting the project, I was starting to feel confident in this idea. So, I posted on my Discord that there was going to be an open play test. And

all right, surprise surprise. People I found on the internet were a little bit harsher with their feedback than my best friends. But I'm so glad I did this play test because it basically kicked off a period of development where Bloodthief turned from my little hobby project into something that was actually pretty good because after showing my work to no one for so long, finally having

a captive audience that actually cared about my updates turned me into a highly motivated feature machine. And every feature I added was basically addressing some issue players were specifically

machine. And every feature I added was basically addressing some issue players were specifically running into. For example, players didn't like that there was no way to save yourself if you fell

running into. For example, players didn't like that there was no way to save yourself if you fell short of a ledge. So I added a ledge climb. On the other end, players kept overshooting obstacles, so I added a ground slam so you could ground yourself quickly in a pinch. It also had the side effect of allowing you to blow guys up in large quantities. Everyone hated the bow and arrow, so I turned it

into a wrist crossbow that shot explosive arrows. Lesson learned. People like blowing stuff up.

Also, because Blood Thief is a movement game, I gave each of these abilities a movement twist.

Jump while climbing a ledge, and you'll vault over it. Jump while ground pounding and you'll bounce.

And the explosive crossbow could be used to rocket jump. I also added a parry ability because players kept rage quitting after getting sniped out of the air and having no defensive option to change their fate. In this period, I also significantly tightened up the existing mechanics. I actually

their fate. In this period, I also significantly tightened up the existing mechanics. I actually

read a book on game design called Game Fuel by Steve Swink, and this gave me a ton of great insights as to why mechanics like the wall run and the air dash felt terrible to play. I reworked the wall Run's controls to be much more streamlined and easy to learn. And I also added a bunch of subtle animations to tell your subconscious what state the player is in at any given moment.

Awesome. Way better. And remember when I said that I liked how the air dash felt? Well, turns

out it had so many issues that I don't even know where to start. But I talk about it in depth in this devlog. In summary, the player just was not getting enough information as to how it actually

this devlog. In summary, the player just was not getting enough information as to how it actually worked and what they were allowed to do at any given moment. It also was just insanely buggy and there was just a billion edge cases that my simple 20line prototype script wasn't handling. So,

I proceeded to completely rework it and turn it from a little 20-line prototype into this 600line behemoth in the codebase that I'm scared to think about when I'm home alone at night. But now, it feels great. So, hey, I regret nothing. The end. So, now I want to take a little bit of a step back

feels great. So, hey, I regret nothing. The end. So, now I want to take a little bit of a step back and talk about what was going on in my personal life around this time. So, up until this point, I was doing all this work at night and on weekends while working a full-time job. I

didn't hate my job, but I was completely obsessed with Blood Thief. It was all I wanted to work on.

This was also around the time that the YouTube channel was starting to get a lot of traction and it was eating away at my soul that there were players eagerly awaiting new features to come out and I was stuck at work. I actually talked about this with my boss and ended up taking a pay cut so I could go part-time for a while and work on the game. This was an amazing arrangement, one

that I probably could have done forever. But then the return to office movement came, remote work started to go away and people working part-time like me were asked to come back to full-time.

the smart safe thing to do would have been to go back to work. But in that period going part-time, I had made so much progress on Bloodthief. I'd grown the channel so much, and I really felt like the dream was starting to become a reality. It felt like I'd been dealt a once-in-a-lifetime hand, and this was my chance to go all in and bet on myself. I was lucky enough to have savings that

I'd accumulated over the years, a supportive wife who also works, and I was also just in a period of my life where I didn't have a lot of costs. So, I did kind of a crazy thing and quit my job. [Music]

Okay, so now that I'm unemployed, I finally had time to address this massive elephant in the room, which was that Blood Thief was a low poly retro game that had the resource consumption of Battlefield 6. So, with Bloodthief, optimization issues can basically be broken down into three

Battlefield 6. So, with Bloodthief, optimization issues can basically be broken down into three major buckets: rendering, navigation, and physics. These are the three big FPS hogs, and all the code I write in the game. 90% of the time pales in comparison to the impact that these three things have. So, the first thing I did was address the lowhanging fruits in these three buckets. For

have. So, the first thing I did was address the lowhanging fruits in these three buckets. For

rendering, I set up occlusion calling, which makes it so the renderer doesn't waste time drawing objects that are already blocked from view. I also started using baked lighting at this point, which basically pre-calculates all the lighting on all the surfaces in the game and saves it to a giant texture. That way, it doesn't have to be computed in real time. The big downside to

baked lighting is that every single time you make an edit to your level, you also have to re-ake the lighting. Godot's baked lighting system, for whatever reason, is actually like pretty fast. So,

the lighting. Godot's baked lighting system, for whatever reason, is actually like pretty fast. So,

this actually wasn't that bad. Like, it only takes like 10 15 seconds for me to bake lighting.

Weirdly though, I get different results when I bake lighting on different machines for a few specific levels in Blood Thief. So, I have no idea what's going on there, but if you're like a baked lighting Godot expert and you're watching this, hey, maybe leave a comment and let me know what the hell's going on. Okay, so now after adding occlusion calling and baked lighting, I

already improved my FPS by like 100 frames in most scenarios. Low hanging fruits harvested. Okay,

now for navigation where players FPS could tank to like 10 frames per second in scenes like this where there's a whole bunch of enemies running around. A little background, most games use what are called pathfinding algorithms that allow enemies to basically navigate towards the player even when there are walls and stuff that they have to navigate around. So, for example, if I'm

here and the enemy is here, my enemy code will ask Godot's navigation server where he should go next, which will probably be somewhere like here because this is the next logical point along the path to the player. The problem is the algorithms that are used to calculate this path are expensive.

the player. The problem is the algorithms that are used to calculate this path are expensive.

The key insight here though is that enemies don't need to recalculate their path every single frame.

So this was an easy change. I basically just made it so instead of querying the navigation server every frame, I query the navigation server every 20 frames or so. So basically three expensive navigation calls per enemy per second. The problem with this though is that every single enemy was using the same code. So while 19 out of 20 frames were smooth on that 20th frame all the

enemies would do the expensive thing all at once causing a huge stutter. I posted this to YouTube and Tavish McKewin in the YouTube comments posted an even better solution. If I maintain a count of all the enemies in the level and give each enemy a unique ID starting from zero, I can use the magical modulus operator to perfectly schedule out all the expensive path updates in a perfectly even

fashion. Basically, with this formula with this, say we have four enemies. Enemy zero updates on

fashion. Basically, with this formula with this, say we have four enemies. Enemy zero updates on tick 0, 4, and 8. Enemy 1 updates on ticks 1, 5, and 9. Enemy 2 updates on 2, six, and 10, and so on. Ah, now everything is perfectly spaced out. Beautiful. Thanks, Tavish Mchuwin. Okay,

what about physics? So, in the early days, anytime something exploded in Bloodthief, all the little giblets that formed were all rigid bodies. And I learned the hard way that rigid body physics is crazy expensive. Take this scene for example, where there's things breaking all around us. Everyone's FPS used to tank at this part. To optimize this, I basically went through the whole

us. Everyone's FPS used to tank at this part. To optimize this, I basically went through the whole thing and made it so only the breakables that were front and center were using rigid bodies and all the breakables that were far away or in the periphery used a much worse but also much cheaper home brood option written by yours truly. As a result, this scene looks a little bit

worse technically if you're looking really really closely, but your FPS is like three times better.

And now the optimization that is probably the most important one that I did, the restart code. So,

in Blood Thief, you die a lot, like a lot a lot. So, I knew that restart had to be fast. At first,

when you hit restart, literally the entire scene was completely reloaded, which took a few seconds.

In my opinion, it needs to be instant. I implemented a system where every entity in the game that needs to be restarted, handles its own restart individually, and everything that is static, like this wall here, just ignores restarts and does nothing. And now, because the game only has to restart a few targeted things instead of literally everything, hitting restart is instant

now. All right. So, at this point, the game runs great on my machine. I guess my work here is Wait,

now. All right. So, at this point, the game runs great on my machine. I guess my work here is Wait, what's this? A stutter. So, as it turns out, at the time, I was using Godot 4.3 and it had this

what's this? A stutter. So, as it turns out, at the time, I was using Godot 4.3 and it had this issue where every once in a while it had to do something called shader compilation, which caused this massive, seemingly unavoidable stutter. I scoured the internet looking for solutions to make it go away. And I tried probably a dozen different solutions. But the problem was

I couldn't come up with an exhaustive list of things that would trigger the stutter. What

I ultimately did to fix the issue was easily the jankiest solution to any problem I've had in Blood Thief. I put up a fake loading screen and behind that loading screen, I play a replay of myself

Thief. I put up a fake loading screen and behind that loading screen, I play a replay of myself going through the level. And in that replay, I kill all the enemies, blow up all the stuff, open all the doors, do everything I can to trigger shader compilation. That way, everything gets pre-ompiled while the loading screen is up, and when the player starts the level, everything is

already compiled and there's no stutter. This solution actually worked great. Hilariously,

Godot came out with a fix for this issue in the next version of the Godot , but I actually have found that sometimes the stutter still happens a bit. So, I actually still use this technique in a few places in the game. So, if you ever see compiling shaders when Bloodthief is loading, now you know what's actually happening behind the scenes. Okay, so let's take another step back. At

this point, it's been about 6 7 months since I opened up the play test. About 2 3 months since I quit my job, and now the game is really starting to pick up some steam. Or should I say Steam is really starting to pick up the game. The last thing I want is for this video to be like job quitting propaganda, but I just think this is an interesting little data point here. You see

this elbow right here in Blood Thief's wish list chart. That's when I quit my job. With all this extra time on my hands, I was able to level up my art skills a bit. I fixed up the player model. I

started working on adding a weapon system to the game with all sorts of cool weapons and swords that I was modeling all myself. I also was steadily getting better at level design and making environments and just steadily started giving everything in Bloodthief a visual upgrade. Also,

a dedicated speedrunning community was starting to form and people started discovering all sorts of crazy tech, some of which I created on purpose, some completely on accident. Either way,

it was incredibly exciting and really just motivating to see people dedicating hundreds of hours to mastering the game. I was still pushing updates to the play test, but at this point, I was deep in the production phase for the rest of the game and had started blocking out the levels, the enemies, and the mechanics that would make up the whole game. And all the while, there was one

big thing that I had yet to do that was weighing on my mind, the demo launch. So, Bloodthief

already had a public beta, but demos are more official and get more visibility on Steam. So, the

plan was to close the beta and have a demo be the official way for people to try the game out before the launch. From a marketing perspective, the demo launch was an opportunity for me to try to make a

the launch. From a marketing perspective, the demo launch was an opportunity for me to try to make a big splash and spread the word about the game. So, all my energy over the next few months was directed towards having the best demo I possibly could. In this time period, I added more features to the game, medals, leaderboards, ghosts. I did a weapon system where you could find artifacts in

the world, and those artifacts combined to create weapons with unique perks. The time came for me to set the date for the demo launch, February 4th, 2025. The weeks went by faster and faster until ultimately the day was upon me. I released the demo, plus a brand new devlog at the same time, and then something truly incredible happened. The demo took off, literally like 10 times bigger than

I was ever expecting or ever could have hoped for. To put it in perspective, the play test peaked out at 38 peak concurrent players. I figured I could probably get up to 50 peak concurrent for the demo launch. But my secret hope was that I could get up to 100 concurrent players because I had heard that's when your demo starts showing up on the trending free tab in Steam, which would

be a really big deal for getting visibility to the game. It got 624 concurrent players. And all

of a sudden, all sorts of amazing things started happening. Some amazing content creators started playing and talking about the game. Blood Thief started climbing the ranks on speedrun.com. And

just a massive influx of new folks started getting exposed to the game, which just lit a fire under me like I've never experienced in my entire life. And I have to admit, while exciting, all this was a little scary. Suddenly, people cared about the game. And I wanted to not only deliver, but I wanted to blow people away. I wanted to give people something that they would remember and that

they would think was special. And so I kind of dropped off the face of the earth and locked in.

This next part of the devlog is basically about the period of time from the demo launch until now. A period where I did do a demo update and a devlog. Mainly just to remind people that I still

now. A period where I did do a demo update and a devlog. Mainly just to remind people that I still exist and I'm still working hard on the project. But for the most part, this period of time was just me at the computer working on the game day in and day out for about 7 months straight. And

this time I worked on so much content that every fiber in my body wants to tell you about. But

I'm holding back because it would just spoil the experience. And there's so much in the game that I'm so incredibly proud of. And I just want you to experience it for yourself. But I'll give you the bird's eye view. A main story with 31 levels, over 20 weapons, each with their own unique perk, a diverse cast of enemies, over 100 secrets, and just so many cool surprises and extra content

that I just can't wait for you all to see. Because my process from the very beginning has been very play test ccentric, I met with friends on pretty much a weekly basis, showing them my prototypes, continuing to tweak, hone, and polish. As for writing the story, my cousin John has helped me out a ton, and throughout the development, we have been honing a story and a lore that I'm hoping

will be satisfying to piece together as you play. And all the while, AAM's Laser, the genius behind Blood Thief's OST, was forging away at one of the most legendary video game soundtracks I've ever heard. Fun fact, in the early days of Blood Thief, I was a fan of AAM's laser and connected to him

heard. Fun fact, in the early days of Blood Thief, I was a fan of AAM's laser and connected to him just through a cold email on a whim. And now we have this amazing relationship that I just really value. And it's another wholesome example of why gamedev is the best. And over the past few months,

value. And it's another wholesome example of why gamedev is the best. And over the past few months, I've been doing testing with a private group of players that I put together that represented the full spectrum of what I felt a Blood Thief player could be. From ultra high skill speedrunners to people who played the demo for 30 minutes all the way down to people who are just vaguely interested

in the game but haven't tried it yet. and I've just been pushing builds almost every day, polishing, iterating, polishing, balancing more secrets, more items, more lore, and more polishing. It was also around this time that I revived Blood Thief's translation effort, which I have literally never talked about on this channel, but I'd love to make a video on this eventually because localization just has a lot of really interesting challenges.

But all I'll say for now is that Blood Thief will be available in 10 different languages, and I'm just extremely grateful for everyone who pitched in to volunteer for this. though. Thank

you to everyone who volunteered to translate Blood Thief. Without you, Bloodthief would only be in English, and that would have sucked. Okay, at this point, the game is basically done. There

were just a few last minute things to make sure it got added, such as SteamCloud integration, achievements, controller support, and after that, the game was done. I uploaded one last build, sent it to Steam for review, which I proceeded to fail. Oops, forgot to change the word demo to game. Sent it for review again, passed. And that brings us to today. Right now,

to game. Sent it for review again, passed. And that brings us to today. Right now,

the game is truly done. And if you're watching this video, it's released on Steam. Hopefully,

it's going all right. But honestly, writing this devlog has caused me to do some reflecting on this journey. And yes, I'm nervous for the big day. But for the most part right now, I'm just feeling

journey. And yes, I'm nervous for the big day. But for the most part right now, I'm just feeling a lot of pride in this project. And I'm just so happy with how it turned out. It makes me so happy to know that players are going to go on an epic adventure. The game is challenging, intense, palm, sweatinducing chaos. But at times, it's funny, scary, even sad. I've put so much of myself and

sweatinducing chaos. But at times, it's funny, scary, even sad. I've put so much of myself and what I like about art and games into this. And no matter what other people feel about the game, I know that I made something that I personally love. I keep going back to this fantasy in my head that a million years from now, some hyper sensient metahuman might dig up an old PC with

Bloodthief installed on it, and if they give it a try, they might be filled with the same emotions that people playing today might have. excitement, wonder, fear, pain, and ultimately triumph. That's

what's so cool about games. This adventure will always be there for anyone who wants to embark.

This thing exists now, and for that, I'm proud. As for what's next, the only thing I'm thinking about is just supporting Blood Thief for the next little while. I have a feeling that means modding and mapping support, but we'll see what the community wants. All right, that's all for this devlog. If

you want to support me and the channel, go buy Blood Thief. All right. Thanks everyone. Bye.