We Need to Go Deeper

We Need to Go Deeper is a co-op roguelite built around the chaos of running an undersea vessel with friends. Think FTL's room-to-room crisis management crossed with a Jules Verne aesthetic, except instead of reading event text you are physically running across the submarine to fix whatever is currently on fire. Each player takes a station - weapons, engine, helm, medical - and the loop revolves around keeping all of those plates spinning simultaneously while biomes get progressively more hostile. It is not a grand-strategy game by any stretch, but it has more systemic depth than its pixelart presentation suggests. The progression layer is where the actual build decisions live. Between runs you unlock new submarine hulls, weapons loadouts, and character classes, each of which changes how the crew should divide labor. A smaller hull with upgraded cannons plays very differently from a sprawling vessel where someone has to sprint three screens to reach the breach in the hull. The unlockable roster means early runs feel intentionally limited - not broken, just constrained - and that is the right way to structure a roguelite. The question of which upgrades to prioritize is genuinely interesting, even if the answer becomes clearer after a dozen runs. Biome variety is solid. The game cycles through distinct undersea environments with different enemy archetypes and environmental hazards, so the mid-game does not collapse into repetition the way a thinner roguelite would. Enemy design occasionally leans on simple stat inflation in later biomes rather than introducing new mechanical challenges, which is the game's clearest shortcoming. The AI for solo play exists but is functionally a placeholder - this is a multiplayer game wearing a singleplayer option as a courtesy hat. If you do not have at least one other person to play with, lower your expectations significantly. For groups of two to four, the communication overhead is the actual gameplay. Calling out incoming torpedoes, delegating repair duties, deciding whether to surface for healing or push deeper for better loot - these micro-decisions create genuine tension without requiring any mechanical skill ceiling beyond basic coordination. The tutorial is brief and functional rather than thorough, which means a new player dropped into someone else's lobby will spend the first run confused about the reactor. That is a mild onboarding gap worth flagging, but the systems are not so complex that an hour of play does not resolve most questions. At 6,000-plus Steam reviews sitting at 86% positive, the community verdict is consistent: this is a reliable co-op night with friends, not a solo contemplation machine. The mod ecosystem is present but modest compared to what strategy players might expect. Replayability is real, though a dedicated group will see most of the content within thirty to forty hours and the question of long-term staying power depends entirely on whether your group enjoys optimizing the same loop or needs constant new content drip. Diego, Scout Team