GitHub Game Off 2025: Build a Game in November, Theme is WAVES

TL;DR

• GitHub's annual Game Off jam runs through November with the theme "WAVES"
• Submit by December 1, 2025 at 13:37 PST — any language, any engine, solo or team
• All you need is a public GitHub repo with your source code and an itch.io submission
• Perfect for first-time game devs or veterans looking to ship something weird in a month

The Big Picture

GitHub's Game Off has been running since 2012, and it's one of the few game jams that actually cares about your code. Not just the playable build — the actual repository. This year's theme is WAVES, and the interpretation is wide open. Ocean waves. Sound waves. Wave defense. A hand gesture. A surge of enemies. Wavy hair. Whatever.

The format is simple: build a game in November, push your code to a public GitHub repo, submit to itch.io by December 1 at 13:37 PST. Then the community votes on gameplay, graphics, audio, innovation, and theme interpretation through January 8. Winners announced January 10.

What makes Game Off different from other jams is the emphasis on open development. Your source code lives on GitHub. You can use templates and existing tools, but the bulk of the work happens during the jam period. It's not just about shipping a game — it's about showing how you built it. That makes it a solid forcing function for learning version control, collaborating with a team, or just getting comfortable pushing code publicly.

The rules are intentionally loose. Use any engine. Any language. Work solo or with a team. AI-assisted development is explicitly allowed, which means you can lean on tools like GitHub Copilot to speed up boilerplate, generate placeholder assets, or debug your physics code at 2 AM.

How It Works

The mechanics are straightforward. Create a GitHub account if you don't have one. Head to the Game Off 2025 page on itch.io and join the jam. You can sign in with your GitHub account if you don't already have an itch.io profile.

Once you're in, create a public repository for your game. This is where your source code lives. Push commits as you build. The repo doesn't need to be polished — it just needs to exist and contain the code that powers your game. You can use any license you want. Open source is encouraged but not required.

When your game is ready, submit it on itch.io before the December 1 deadline. After submissions close, voting opens. Participants rate each other's games across six categories: overall, gameplay, graphics, audio, innovation, and theme interpretation. Voting runs until January 8, 2026 at 13:37 PST. Winners get announced two days later on the GitHub Blog and social channels.

The theme — WAVES — is intentionally vague. GitHub's suggestion list includes a space shooter with gravitational waves, a coastal survival game with tsunamis, a tower defense with enemy waves, a skateboard game on a sine wave, a rhythm game, a vaporwave racer, a physics puzzler with energy waves, or a nostalgic remake. But those are just prompts. The best Game Off entries usually ignore the obvious interpretations and do something unexpected.

You can work alone or assemble a team. The Game Off 2025 Community on itch.io is active, and there's a community-run Discord for finding collaborators or asking technical questions. If you're stuck on ideas, GitHub suggests asking Copilot: "What are some fun games I could create with the game jam theme, WAVES?" It's a decent starting point if you're staring at a blank repo.

What This Changes For Developers

Game jams are a forcing function. You have a hard deadline, a loose constraint, and no room for scope creep. That makes them one of the best ways to actually finish a project instead of endlessly prototyping. Game Off adds a layer of accountability by requiring a public repo. Your commit history is visible. Your code is out there. That's uncomfortable for some developers, but it's also how you get better.

The AI-assisted development rule is worth paying attention to. Most jams don't explicitly call this out, but GitHub does. That means you can use Copilot, ChatGPT, Claude, or any other LLM to generate code, write shaders, draft dialogue, or debug logic. The line is blurry, but the spirit is clear: use whatever tools help you ship.

For developers who've never touched game dev, this is a low-stakes entry point. You don't need Unity experience or a degree in computer graphics. Plenty of participants use lightweight engines like Pygame, LÖVE, or Phaser. Some build for retro hardware like the NES or Game Boy. A few write their own engines from scratch. The format doesn't care. It just wants to see your code and a playable build.

The voting system is community-driven, which means your game gets played by other developers who just spent a month in the trenches. The feedback is usually constructive. The winners aren't always the most polished games — they're the ones that nail the theme, try something weird, or just feel fun to play.

Try It Yourself

If you're new to game development, pick an engine that matches your existing skills. Godot is a solid choice for 2D or 3D games with GDScript, C#, or C++. It's open source, lightweight, and has a gentle learning curve. Unity is better if you're targeting mobile or need access to a massive asset store. Unreal Engine is overkill for a jam unless you're chasing cinematic visuals or already know Blueprints.

For web developers, Phaser is a natural fit. It's JavaScript-based, runs in the browser, and works well for 2D arcade or platformer games. If you're a Python developer, Pygame is the obvious starting point. It's not fancy, but it's fast to prototype with and has decades of tutorials.

Rust developers should look at Bevy. It's a modern, data-driven engine with an ECS architecture that feels native if you're already comfortable with Rust's ownership model. Go developers have Ebitengine, which is simple and performant for 2D games. Dart and Flutter developers can use Flame for mobile-first 2D games.

If you're coming from Java or Android, libGDX is familiar territory. For retro-style 2D games, HaxeFlixel is a strong choice. Lua developers have two solid options: LÖVE for lightweight 2D games and Defold for cross-platform projects with built-in tools and an active indie community.

The engine matters less than the constraint. Pick something you can learn in a weekend, then spend the rest of the month building. The goal is to ship, not to master a new toolchain.

The Bottom Line

Use this if you've been meaning to finish a game project but keep getting distracted. The deadline and public repo create enough pressure to actually ship. Use this if you want to learn version control in a low-stakes environment where broken commits and messy code are expected. Use this if you're curious about AI-assisted development and want to see how far Copilot can carry you in a creative project.

Skip it if you're already crunching on a commercial project or don't have 20-40 hours to spare in November. Skip it if you hate deadlines or find public repos stressful. The voting period is optional, but the submission deadline is hard.

The real opportunity here isn't winning. It's the forcing function. A month is long enough to build something interesting but short enough that you can't overthink it. The theme is loose enough to go weird. The rules are minimal. And the community is full of developers who just want to see what you made. If you've been sitting on a game idea or just want an excuse to write code that isn't work, this is your window.

Source: GitHub Blog