Seven Day Roguelike Jam 2025

Recently, I took part in the Seven Day Roguelike jam, or the “7DRL.” I’ve been a watcher of this jam after seeing Terry Cavanagh talk about his entry a few years back, but I’ve never had the time to really give it a go. Lucky for me, I have no job right now, so I have more time than I normally would to work on something like this.

Before the jam started, I did a little brainstorming here and there about ideas, but a few old concepts I’ve had in mind for years kept insisting I give them some attention. The one I settled on wasn’t originally a roguelike in my mind, but after thinking about it a bit, I realized it was probably perfect for some classic rogue elements.

The Concept

At some point, while staring at some dice years ago and trying to come up with a simple game idea, I inadvertently came up with one that was much more complicated. The basic concept was that you play as a six-sided die, and the layout of the pips determines which spaces around your die are impacted by actions you take. Movement would involve rotating your die across to a neighboring grid square and the new set of pips on top would be what you’re stuck using for better or worse.

Early look at the die moving from one space to the next

Throw on top of this foundational idea a huge pile of elemental effects, character progression, and equipment shuffling and the prototype in my head became very complex. For a jam game, though, all of that extra cool stuff was tossed out quickly in the hopes of having a reasonable scope.

Back in a Jam

Turns out the basic scope was still quite a lot to get done for this. First of all, even though I am unemployed, I still have a family, and the sense that I should be working through some home project backlogs as well as, you know, finding a job. So, even with all of those things, I suspect I was able to get in more than forty hours of pretty solid work on this project, which is pretty good for something I was mainly doing “for fun.”

In addition to real-life time constraints, I was also a bit held back by my lack of existing frameworks, tools, and even code snippets to pull from. When I’ve previously done game jams, I had worked on similar projects in close proximity, so I at least felt comfortable pulling things like state machines and audio controllers in from those. It’s been so long since I’ve jammed, though, that it felt like digging through any of my old stuff would take a long time, and I also know anything I created more than three years ago is going to seem just terrible to me now. That means I started with a blank slate, for better or worse.

I was able to get a rough architecture idea together with some basic game and turn management without much trouble. It all had to be adaptable though, because I wasn’t even sure of some important design elements from the start like turn resolution order, enemy movement, etc. I was moderately successful at this, and by the end felt pretty good about how I would do it all so much better if I could start all over—game jam enlightenment.

Since this was a roguelike, I also had to create some level generation logic, trying to take into consideration the way your die would have to move around, especially if trying to “line up” attacks against surrounding enemies. This logic is definitely in need of a revision. It seemed to work out just fine in all of my tests, but the level layouts were generally less interesting than they could be. I thought I could come back to this at the end after some playtesting and refine the level output, but playtesting was barely even possible in the time limit so I was just glad this level generation worked reliably at the end.

Gif showing debug tests of level generation — not quite the “final” version

Other than that, my main challenges were setting up different enemy pathing routines and making sure your attacks worked correctly based on your current pip pattern. 

Results

So, did it all work out? Well, of course not! It’s a game jam! I think it was somewhat successful, though, and I came away pleased with what I got done. Getting a working game posted within the time limit and having it use a new level generation system and a pip-based battle mechanic were pretty huge. The art is passable, and the free assets and color palette I used definitely give it a look I am pleased with—except for the player’s shadow disappearing sometime during the last fifteen minutes of the jam for unknown reasons.

Screenshot of the gameplay

It had been quite a few years since I’d really committed to a game jam, and a lot of lessons learned came back to me. I definitely need to have some generic frameworks ready for use next time, and ideally would work with at least a couple other people to get something that feels more polished at the end. I’ll probably also need a good reason to do another one soon, because honestly, it wasn’t a lot of fun to have this kind of time pressure applied voluntarily. It feels completely different when you’re crunching on finishing a game for a job and trying to help a team than when you are just doing it for your own satisfaction. Still, I’m glad I did this one, and I’m not ruling out another one sometime soon.

Thanks to anyone who gave the game a try, and let me know if you like the core mechanic or not. I might revisit it later with more of the many ideas that got quickly chopped from this version due to the tight schedule.