Nobody Saves the World

There’s a point early in Nobody Saves the World where the game’s design intent becomes clear. You’ve been swapping between a Rat that squeezes through vents, a Ranger that snipes from across the room, and a Slug that leaves poison trails, not because you wanted variety but because the quest log made you. And then it clicks: you’re not playing a class. You’re playing a toolbox. That reframing is the whole pitch, and Drinkbox Studios built the game to sell it.

This is the Toronto studio behind Guacamelee!, a Metroidvania that understood movement as expression better than almost anything in its generation. Nobody Saves the World is a different genre — top-down action RPG, dungeon crawler, loot-adjacent — but the DNA is identical. Drinkbox builds games where the verb set is the content. Where other studios ship a skill tree and call it depth, Drinkbox ships a system and dares you to break it.

The setup is deliberately thin, and that’s the point. You are Nobody: a featureless, pallid blank slate with no face, no history, and no discernible reason to exist. You find a wand. The wand turns you into things. Meanwhile, a fungal apocalypse called the Calamity is eating the world, spreading its spores across an overworld that’s larger than the game’s cheerful presentation suggests. The narrative is a delivery mechanism for jokes and a reason to keep walking east. It’s fine. Nobody’s here for the story.

Market positioning is where this gets interesting. Released into a genre stuffed with roguelikes that demand mastery and ARPGs that demand hundreds of hours, Nobody Saves the World asks for neither. It’s a moderately sized game with roguelike-adjacent dungeon generation and zero permadeath anxiety — a paid release at $24.99 on Steam and Nintendo platforms, out since January 2022. It occupies a genuine gap: approachable enough for lapsed genre fans, deep enough for tinkerers.

• 15+ Distinct Forms: Rat, Slug, Ghost, Rogue, Robot, Dragon, and more. These aren’t skins with stat modifiers. The Rat is a stealth-and-swarm build. The Ghost phases through walls and drains. The Robot is a self-damaging turret platform. Each Form has its own attack set, movement profile, and failure state.
• 80+ Cross-Pollinated Abilities: The actual game. Completing Form-specific quests unlocks abilities that can be equipped on any Form. Bolt the Rat’s poison onto the Dragon. Give the Ranger the Rogue’s crit multipliers. The game never suggests these combinations. It just doesn’t stop you.
• Quest-Driven Progression: You don’t level by killing things. You level by completing objectives — “deal X damage with a discharge attack,” “kill Y enemies while stealthed” — which forcibly rotates your loadout and prevents the single-build ossification that kills most ARPGs.
• Procedural Dungeons with Elemental Gating: Dungeons scale and remix as you level, and many gate progress behind damage types, forcing loadout adjustments at the door rather than letting you brute-force with one build.
• Two-Player Co-op, Local and Online: Drop-in, drop-out, and designed so Forms complement each other rather than competing for the same damage ceiling.
• Jim Guthrie Score: Original soundtrack from the Sword & Sworcery and Below composer. Not background music. Actual composition.

The Gameplay Loop

Most action RPGs have a comfort problem. You find a build that works early on, and the rest of the game is a slow, pleasant slide into numbers going up. Nobody Saves the World is architected specifically to prevent this, and the mechanism is elegant: it decouples progression from combat entirely.

You don’t gain meaningful power by fighting. You gain it by completing Form quests. And Form quests are, by design, things your current build is bad at. The Ranger’s quest wants you landing charged shots at range. The Slug’s wants you poisoning enemies and letting them rot. The Magician’s wants you managing mana you don’t have. Every quest completion is a small act of self-sabotage — you deliberately equip a worse build to satisfy an objective, and in doing so, you learn what that Form actually does.

This is a genuinely clever bit of design psychology. The game weaponizes completionism against comfort. And because the abilities you unlock are portable across every Form, that forced experimentation compounds. Deep into the campaign, you’re carrying a mental index of dozens of abilities and a growing suspicion that some of them, combined, are absurd.

They are. That’s the payoff.

The Build Engineering

Here’s where Nobody Saves the World separates from its peers. The ability system isn’t a skill tree with locked branches. It’s a flat pool with slot constraints, and the constraints are what make it interesting. Each Form has a limited number of slots, some Form-specific and some universal, and abilities carry damage-type tags — light, dark, physical, elemental — that interact with enemy shields and dungeon gates.

So the question stops being “what’s strong?” and becomes “what’s strong here, given that this dungeon’s shielded enemies require dark damage and my Form’s native attack is physical?” The answer is usually a Frankenstein build the designers didn’t script: the Horse’s charge attack carrying lifesteal from the Zombie and a status effect from the Slug, tuned to a damage type the Horse has no business dealing.

When it lands, the feeling is unmistakable — you didn’t find a build, you built one. The game gave you parts and got out of the way. There’s a specific pleasure in engineering something the developer never explicitly sanctioned, and Drinkbox clearly designed for exactly that. Steam reviewers keep landing on the same phrase — “frictionless fun” — and the sharper read is this: the friction is intentional and front-loaded, then it evaporates. The onboarding for each new Form is a speed bump. What comes after is a downhill run.

Where It Slips

The quest-driven engine that makes the first two-thirds sing starts grinding in the last third. Late-game objectives lean on volume rather than ingenuity — kill 100 of a thing, deal N thousand damage with a mechanic you already mastered. The design intent is obvious: keep the rotation going. But rotation only works when each stop teaches you something. Late in the campaign, the quests stop teaching and start counting, and a system built to fight tedium starts generating it.

The difficulty curve has spikes, too. Not the good kind that force reinvention — the kind where a dungeon’s gating collides with an under-leveled Form and you’re grinding for a level, not solving a problem. And a handful of Forms never fully click. The Slug is deliberately awful to control, which is a joke that works once. The Magician’s mana economy is fiddly enough that its abilities are better slotted elsewhere and the Form itself is easy to abandon. When you’re shipping fifteen-plus Forms, two or three duds is a reasonable tax. It’s still a tax.

Combat itself is competent rather than exceptional. It’s serviceable top-down action — readable, responsive, never demanding the precision Guacamelee! asked for. The depth here is in the loadout screen, not the dodge roll. If you want mechanical execution to be the challenge, this is the wrong game. The challenge is the puzzle of the build, and the combat is where you check your work.

Drinkbox’s art direction remains one of the most underrated in the business. The style is a thick-lined, flat-shaded cartoon vocabulary — closer to a Cartoon Network model sheet than to pixel-art nostalgia — and it does real mechanical work. Every Form is silhouette-distinct at a glance, which matters enormously when you’re mid-fight and your last three transformations blur together. The Calamity’s fungal corruption reads as visually wrong against the game’s candy palette, sickly purples blooming across otherwise sunny terrain. That’s art direction serving the fiction instead of decorating it.

The UI is the weak link, and it’s the predictable one: a game with 80+ abilities across 15+ Forms eventually becomes a database problem. The loadout screen is built for a catalog this size and gets crowded as it grows — functional, cleanly laid out, but requiring more menu-diving than the fluidity of the actual play deserves. Comparing abilities means remembering them, not seeing them side by side. It’s not broken. It’s just that the menu is where the least joyful minutes happen.

On the technical side, a top-down game of this scope is light on hardware by design, and Drinkbox has shipped enough cross-platform titles to know where the ceilings are. The modest demands are part of why it fits so naturally on portable platforms.

Steam is the legibility pick. On a large display, the loadout screen stops being a chore and starts being a workbench — you can actually read ability tooltips without leaning in. Controller is strongly recommended over mouse-and-keyboard; this is a twin-stick-shaped game and the KB/M mapping is a courtesy, not a design. The game’s structure — bite-sized dungeons, no permadeath, constant save states — is essentially handheld-native, which makes it a natural fit for portable PC play.

Nintendo Switch suits the game’s rhythm well: a dungeon fits a commute, a Form quest fits a coffee break. The docked/handheld duality also cleanly resolves the one real tension: manage builds while docked on a TV where the text is readable, then run dungeons in handheld. The cost is the text size in handheld mode, which is rough for a game this menu-dense, and the visual fidelity dips slightly. Neither breaks it. This is a game designed for the way you play on Switch, not the way you play on a gaming PC.

Local co-op is a Switch strength (two Joy-Cons, done). Online co-op is a Steam strength. Pick your poison based on how you plan to bring a second player.

Nobody Saves the World is what happens when a studio takes a gimmick seriously. Shapeshifting is usually a mid-game power-up or a boss mechanic — a novelty that gets three levels of screen time and then retires. Drinkbox built a game where it’s the foundation, the progression system, the difficulty curve, and the punchline all at once. That’s not a small thing to pull off, and the fact that it works for most of its runtime is a real accomplishment.

The back third is the problem. A game this smart about fighting boredom shouldn’t run out of ideas and start counting kills, and the shift from clever objective to checklist is jarring precisely because the earlier hours set such a high bar. That tension — between the joy of the first Dragon-with-Rat-poison build and the fatigue of hour twenty — is the honest shape of this game.

We rate it a solid, flawed-in-ways-that-matter recommendation: buy it for the right audience, play it until it stops surprising you, and don’t feel obligated to finish New Game+. The best hours here are worth the ones that aren’t.