ColosSEAum - a devlog for an unfinished game

We made a Hades-like game for the Bigmode Game Jam 2026, but we tried to bite off more than we could chew.

You can try it out the game on itch.io.

The theme of the jam was slick:

This theme takes inspiration from last year's Bigmode game jam judging livestreams, where slick was given as high praise for clever or interesting moments. So one option is to simply design something that, when encountered, would make the player say 'you were slick for that.'

The team consisted of Jan Kuželík (dev), Adrien Ramanana Rahary (3D modeling), Alžběta Volhejnová (visual design), Vojtěch Volhejn (dev), and yours truly (dev).

The process started with a brainstorming session in Figma, which I wasn't part of. My favorite artifact from that phase is this analysis of what "slick" can mean (zoom to see the games):

We were aiming for the fast/schmooving quadrant. The idea that we settled on was a game similar to Hades in its combat system, but cranking up the slickness element. Weapons would only last for a short time before breaking, like in Superhot, encouraging the player to switch weapons often, throw them at enemies, and pick up weapons from defeated enemies.

The setting/lore is that you're in debt to an oil baron and you must fight in an arena to repay your debt. If you're slick enough to impress the spectators, they'll throw more weapons and other pickups your way.

Also, the whole game ended up having an underwater theme somehow. Not complaining!

Here are a few screenshots from the Discord channel to illustrate the dev process.

And a little video, showing features that Honza (Jan) implemented: schmooving with bullet time, teleportation and particle effects.

To me, the lesson is: 3D is still hard! We already had one unsuccessful attempt at making a 3D game for a game jam.

The Bigmode Game Jam ran for 10 days, making it the longest game jam we've done yet. Personally, I still prefer the weekend format of Ludum Dare and similar jams: dedicating one intense weekend to the jam and then being done with it. With this long time frame, I can't spend all of my waking time on the jam – for example, I have to go to work. Towards the end, it starts to feel like homework. I didn't have as much time to contribute to the jam as I hoped.

One thing I noticed is that I'm using AI coding tools a lot more heavily for this than I was for other game jams (I think the last one I did was in October 2024). Since my Godot is extremely rusty, my contribution ends up being mostly vibe coding different features using Claude, trying only to lightly steer it when I notice it's going for a bad design.

Although I find AI coding assistants incredibly helpful, here I did feel a bit sad that I'm not learning as much about Godot as I might otherwise if I was forced into actually learning things using trial-and-error and reading documentation.

Overall, it was a fun experience, I learned a bit more about Godot and 3D, and I also saw a person proficient at 3D modeling (Adrien) at work for the first time. I was impressed at what's possible so quickly!