Muck

I have watched enough live-service games quietly rot that a small, finished, self-contained survival roguelike feels almost radical in 2024. Muck is exactly that. Solo Norwegian developer Dani built it in roughly a month to answer a YouTube comment that said he could not make a multiplayer survival game, and the result somehow crossed 10 million downloads without a single paid expansion, season pass, or battle royale battle pass in sight. That origin story matters to me because it shapes what the game is: tight, honest, and unbothered by monetisation pressure. The loop is straightforward but genuinely mean. You drop onto a procedurally generated island with nothing, spend the day chopping wood, mining Iron and Mithril ore, smelting bars at a furnace, and crafting tools up the tier ladder through Steel, Adamantite, and eventually the absurdly named Obamium. When night falls, goblins, kobolds, golems, and buffed elite variants swarm your position, and enemy strength scales with every day you survive. Killing mobs earns gold coins you spend on loot chests containing passive powerups like lifesteal, Bossbane, Flame Aspect, and Nightrend bonuses. That powerup layer is the Risk of Rain 2 DNA showing through, and it is the reason the game has real run variance beyond map seed luck. Two soft class paths exist, melee and ranged bow builds, both sharpened by defeating bosses: Big Chunk drops Chunkium for its armor set, Gronk rewards its own materials, and the dragon Bob waits at the end of your boat-repair run as a hard final check on whether your powerup stack is actually ready. Death resets everything, no saves, no persistent unlocks. Multiplayer is where the cracks become funny rather than frustrating. Up to eight players can share a lobby, but more players means more enemies and harder scaling, so inviting a chaotic friend can absolutely destroy a run you had under control. There is no dedicated server infrastructure, so sessions are peer-hosted and lag is occasionally a real problem, especially on higher player counts. The room-code system works, but it is barebones. Guild tooling, matchmaking, guild tooling, any of it, does not exist, because this is not a live game. It is a pickup co-op game that happens to support up to eight people in a tent. The criticisms are real: the low-poly visuals are functional but rough, the island variety is limited especially in later updates that never quite filled out the world, and Dani has since moved on to other projects, meaning Muck is effectively version-locked at 1.3. There will be no new biomes, no seasonal content, no second final boss on hard difficulty. For a live-service brain that craves the next patch, that can feel like a dead end. For someone who is tired of games that never end, it is a feature. A full run takes 45 to 90 minutes. The whole game fits in 100 MB. It runs on hardware from a decade ago. The speedrunning community has been active with it for years, which tells you the underlying mechanical loop holds up to real scrutiny. If you are the kind of player who measures a game by its seasonal model and content roadmap, Muck will disappoint you in about twenty minutes. If you are the kind of player who remembers games like WildStar or Landmark and how much time you poured in before the servers went dark, a completely free, finished, offline-capable co-op game with a genuine boss loop and no expiry date is genuinely refreshing. Bring a friend, pick a seed, and lose an evening to it. Yuki, Scout Team