This is my third finished and polished game, and it started as a pure learning exercise. Mythic Dash is heavily inspired by an arcade classic, which meant diving into basic AI and pathfinding—building directly on what I learned while finishing Break Bricks.

I began by researching the original arcade game to understand how its AI and pathfinding worked. My first version was a near pixel-perfect recreation using integer values to keep things simple. Once the player could move around the grid, I switched to floats for smoother movement and finer speed control. That decision made everything much harder than expected! No more simple equality checks—you suddenly need epsilon values and slop to let characters navigate the maze cleanly. It took a ton of tinkering, but the end result is buttery-smooth movement and responsive controls at any frame rate. I tested it at 144, 60, 30, and even 15 FPS, and it feels great across the board.

At that point I realised that if I wanted to actually share the game, I needed to move away from the original arcade graphics. After a few days hunting for free assets, I discovered the fantastic Art assets by Shade. That pulled the whole project toward a Southeast Asian theme, which led me to matching sounds and music that feel authentic and cohesive.

I also learned how to use Tiled, which saved me from writing my own level editor. I loved working with it and plan to use it again. Loading the maps and tilesets was trickier than expected though—I’ll probably rewrite that part in the future. It works reliably now, but it was definitely hacked together.

Sound design turned out to be its own little art form. Getting music and effects to sit together nicely takes real patience. The coin pickup sound went through multiple revisions before it stopped giving me a headache! (My first game, bloc, still makes me laugh—I called it “I know a sound effect that will get on your nerves!” I’ve definitely come a long way.)

AI and pathfinding took a lot of time. I actually scrapped my first attempt because I couldn’t figure out why the creatures kept misbehaving. Even now they occasionally do something weird, but after hours of testing it never breaks anything serious. The best investment was building a debug overlay so I could literally watch what the creatures were thinking in real time. Logs only get you so far.

Once I had one playable map that felt like the arcade original, I started expanding. The tileset was built for RPGs and included traps, so I added those and designed each level around introducing a new one. This is the part where I’m really curious for player feedback—how my intentions line up with what people actually experience and perceive.

Adding the key/door mechanic created a fun new challenge: the basic AI couldn’t handle locked-off areas and would get stuck looping. My solution was to make doors passable for the player but still solid walls for the creatures. As a bonus, it gives the player a welcome breather from being chased.

I ended up with seven levels, each introducing mechanics progressively. Balancing was tough because I wanted to keep the spirit of the arcade original—simple to understand, brutally hard to master. To make it more approachable for modern desktop play, I added three difficulty levels. Surprisingly, small tweaks (like 10% faster player or slower enemies) make a huge difference. I also implemented a level unlock system so you can jump back in where you left off, though you only get full run credit if you start from level 1 and finish in one sitting.

Arcade Mode is the original brutal difficulty. Speedrunning isn’t usually my thing, but I added determinism and frame-perfect timing support, so it should be a treat for anyone who is.

Interestingly, getting something playable took a few months, but expanding it into seven full levels only took about a week. That’s because I put real effort into designing a solid base engine and architecture upfront. Even though I work in C, I use object-oriented patterns where they make sense—the player and creatures share the same base class. The player took a week to implement; basic creature movement took one evening.

The final stretch—proper game states, title screen, options, menus, high scores, etc.—was more of a slog. The fun creative part was done, but all the “boring” polish still needed doing. Development slowed, but I stayed determined to ship something I could be proud of.

What started as a learning project became a polished little hybrid arcade game that I genuinely enjoyed making—and more importantly, that’s fun to play. So I decided to put it out on itch.io and see what others think. The launch has been quiet (six downloads on day one, two of which were me and a friend), but I’m not discouraged. I’m getting active in relevant communities, looking for playtesters, and considering game jams.

I’m also pretty happy with the codebase overall. Next up I’ll rewrite parts of the engine backend and improve the interface. My next project is a top-down shooter, so I’ll need solid two-directional scrolling and much stronger AI/pathfinding. Plenty more to learn!

Thanks for playing Mythic Dash. I’d love to hear your thoughts, suggestions, or even bug reports. Feedback is pure gold at this stage. Happy dashing! 🐉