The short answer
Dave the Diver was built by Mintrocket — a Nexon sub-brand — on Unity, not on a custom C++ engine, Godot, or Unreal. It launched on Windows and macOS on June 28, 2023 after Early Access in October 2022, then spread to Switch on October 26, 2023, PS4/PS5 on April 16, 2024, Switch 2 and Xbox in November 2025, and mobile in September 2026.
That boring choice explains the whole game. Unity let a small team ship a pixel-plus-3D hybrid to 10 platforms in a 4.4 GB Switch package, iterate the dive-shop loop through Early Access, and run five years of crossover DLC without rebuilding tech. Sales followed the porting: No.5 on Steam two days after launch, 1 million copies in 10 days, 3 million by January 2024, and over 5 million by November 2024.
My distinct take: Dave is not two games glued together. The sushi restaurant is Unity's garbage collector with chopsticks. Diving creates messy, perishable memory — heavy fish, limited oxygen. Serving converts it into permanent gold and reputation every night. That is why the loop feels relaxing instead of grindy, and why Unity — the king of boring cross-platform memory management — was the right moat.
Facts first: what you are actually buying
Mintrocket started the prototype in 2017 as Dave, where a diver just grabbed fish for rewards. The team called it too simplistic and restarted around resource management. It was briefly announced as a mobile National Geographic collaboration in 2018, then reannounced in May 2022 as the first game from Mintrocket, Nexon's division for games focusing on the essence of fun.
Inspiration is unusually honest: a real Jeju Island restaurant whose owner dives by day and serves by night, plus the team's own diving trips, plus Torneko's Great Adventure, Abzu, and Metal Gear Solid.
Core facts verified from primary pages:
- Engine: Unity in the infobox, with C++ runtime and C# scripting API, built for 2D/3D cross-platform shipping.
- Loop: Dive the Blue Hole twice per day — harpoon gun, guns, melee, nets and drones later — for 200+ sea creatures, then run Bancho's sushi bar at night: craft sushi, pour drinks, clean tables, refill wasabi, earn gold and reputation, unlock recipes, staff, a farm, and cosmetics.
- Pressure valve: Oxygen drains faster when you are hit. If you black out, you surface early and keep only one item. Upgrades raise weight, oxygen, and depth.
- Endgame biomes: Glacial Passage, Glacier Zone, and Hydrothermal Vents hold the most valuable fish and the most dangerous enemies, plus bosses, puzzles, side quests, and minigames.
- DLC as menu expansion: Free Dredge fog event on Dec 15, 2023, Godzilla missions on May 23, 2024, Balatro plus Potion Craft plus mxmtoon on Oct 24, 2024, paid Like a Dragon on April 10, 2025, In the Jungle expansion on June 18, 2026, and a Subnautica 2 crossover announced for early 2027.
- Reception: Metacritic universal acclaim from 44 critics, Famitsu 36/40 with one 10, IGN Editors' Choice, PC Gamer calling it already on his 2023 GOTY shortlist, plus best-of-2023 lists from The Guardian, The Verge, and others.
- Indie debate: Mintrocket is wholly owned by Nexon and became a wholly owned subsidiary in September 2024. Nominations for Best Independent Game at The Game Awards 2023 and Golden Joystick 2023 sparked backlash about what indie means. Organizer Geoff Keighley said independent can mean different things, while director Jaeho Hwang said he does not consider the game indie.
That last point matters for tech choice. This is corporate cozy, not garage cozy.
Why Unity and not something more fashionable?
1. Ten platforms without a porting team
A custom engine would have made Switch, PlayStation, Xbox, and mobile ports expensive. Unity already supports desktop, mobile, console, and WebGL from one C# codebase, which is why it dominates indie and mobile development.
Dave's Switch build is only 4.4 GB, 8.8 GB on Switch 2. That small footprint is a Unity specialty: 2D sprites for characters and fish layered over 3D water, rather than Nanite-scale geometry. You can read a fish's value at a glance because it is deliberately flat against a soft 3D background.
Unreal Engine 5 would have bought Lumen lighting nobody needed and a hiring bill nobody wanted. Godot would have been cheaper on license but thinner on console QA and on staff who already know Unity inside Nexon.
2. Early Access as balance lab
Mintrocket said the hardest problem was stopping any one mechanic from becoming repetitive, solved through trial and error in Early Access.
Unity's prefab and C# iteration speed fits that. Change harpoon damage in the morning, watch Steam feedback by night, push a new sushi recipe the next week. A hand-rolled engine would have turned every tuning pass into engine work.
3. Corporate hiring insurance
Here is my deliberately non-rigorous but useful idea: Unity is Nexon's internal labor market.
Nexon runs live-service giants like MapleStory and Dungeon & Fighter. Those teams already think in Unity and C#. If Dave needed an extra UI programmer for the restaurant rush or a console specialist for Switch certification, Mintrocket could borrow one without retraining them on proprietary tools.
A custom voxel engine like Teardown's or a renderer-only stack like Kenshi's OGRE would have trapped knowledge in three heads. Unity spreads it across the parent company. For a sub-brand whose mandate is to experiment, portability of people matters more than portability of pixels.
My analysis: three ideas you will not find in the wiki
I label these as interpretation, not verified fact. They are the reason I think Dave works where other cozy sims stall.
Idea 1: Oxygen is a wallet, not a health bar
In Subnautica — also Unity, also ocean — oxygen is fear. You hear breathing, you panic, you surface.
In Dave, oxygen is an exchange rate. Every second down there asks: is this fish worth the air it costs to carry home? Getting hit does not just hurt you, it taxes your wallet by burning air faster.
That turns each dive into a gacha pull with a pity timer. You always surface with at least one keep even if you black out, so there are no truly empty pulls. The harpoon is the pull button, weight is inventory pity, and the restaurant is where you cash out. No other Unity survival I have covered makes the meter this explicitly economic. Rust uses wipes, Valheim uses boss gates, Subnautica uses depth. Dave uses appetite.
Idea 2: The restaurant is a garbage collector
Programmers will get this instantly, designers should steal it.
Diving allocates messy objects: fish of different weights, values, and depths, plus materials and weapons. You cannot hold everything. If you die, most of it is freed.
Serving compacts it. Perishable loot becomes durable gold, reputation, recipes, and staff. Staff and the farm are basically generational GC — they handle routine serving so you can allocate bigger dives.
That is why the game feels relaxing despite having roguelike elements. Most roguelikes delete your run and call it stakes. Dave compacts your run and calls it dinner. The two-genre split is not bloat, it is memory management with wasabi on top.
Idea 3: The Blue Hole is a licensing excuse that became a feature
A single hole containing fish from all over the world is biogeographic nonsense. Glacial fish next to hydrothermal vent crabs should not share water.
As tech strategy, it is brilliant. It justifies mixing any art asset, any crossover, any new biome without worldbuilding. Dredge fog? It drifted into the Hole. Godzilla? He swam by. Balatro jokers? Bancho put them on the menu.
Unity's Asset Store logic meets fiction: if everything can live in one hole, you never need a new map, only a new depth flag and a new recipe. Compare this to open-world games that need lore reasons for new zones. Dave needs only hunger. The In the Jungle expansion finally leaves the Hole, but only after five million copies proved the trick.
A side effect: pixel-plus-3D reads as honesty. Photoreal fish would hide hitboxes and values. Flat sprites shout rarity and price. Ugliness is a price tag, just as I argued for other Unity hits, but here the tag is literally on a menu board.
What this costs — Unity's bill comes due
Unity does not give 10 platforms for free.
- No real multiplayer: Dave is single-player only. That avoids Photon or custom netcode bills that games like R.E.P.O. pay, but it also caps streamer chaos that boosted Lethal Company and PEAK. Dave grows by word of mouth and DLC, not by lobby clips.
- Physics is good enough, never great: Harpoon reeling and boss fights feel floaty compared to Unreal shooters. Mintrocket hides it with pixel charm and quick-time bypass accessibility options, but you feel the C# ceiling in late vents.
- Corporate indie tax: The Best Independent Game debate earned press, but it also painted a target. Every update is read as Nexon strategy, not solo-dev passion. Unity's democratized look feeds that — players expect a $5 Godot gem, not a polished Nexon product, and judge pricing accordingly.
None of this sank the game because the loop prints money faster than the costs compound. Small download, broad devices, constant menu reasons to return.
What indie devs should actually steal
Forget clone the sushi. Clone the economy:
- Make your meter convertible. If oxygen, stamina, or time only kills you, it is a tax. If it converts into shop progress, it is a decision. Let players cash out early for a small keep, like Dave's one-item pity.
- Ship one compaction mechanic. If your game generates messy loot, give players a nightly ritual that turns it permanent. Bakery ovens, sushi bars, and farm boxes all work. Do not make them replay the mess to feel progress.
- Use geography as a patch-note excuse. A Blue Hole, a dream, a TV station — any fiction that lets you add biomes without redrawing the world will save you a year. Add recipes, not maps, until you can afford In the Jungle.
- Prototype the second job first. Mintrocket restarted when diving alone was boring. If your core verb is thin, add the boring job that monetizes it. Management is not the garnish, it is the game.
- Borrow hiring, not just tech. If you sit inside a bigger org or a Unity-heavy city, boring tech lets you borrow testers and porters. Exotic tech makes you hire unicorns.
Try the loop in your browser first
You do not need a 4.4 GB download to feel Dave's two halves. These free browser trainers isolate each verb naturally — play one when you want to practice the matching skill:
- When you want to practice breath control and eat-or-be-eaten math, run a few rounds of Big Shark. Avoid bigger fish, eat small ones to grow, and surface with a plan — the same oxygen-versus-greed calculation as the Blue Hole, compressed into two minutes.
- When the dinner rush wipes you, reset your rhythm with Bakery Manager: Store Simulator. Scan, restock, handle cash, and keep the line moving. It teaches the exact front-of-shop flow Bancho demands: speed during rush, restock before zero, upgrade bottlenecks first.
- When recipes feel overwhelming, warm up with ABC Pizza Maker. Pick a base, layer sauce and toppings step by step, bake, and serve. It is simpler and kid-friendly, but the prepare-assemble-bake-eat sequence is the same mental checklist as testing a new sushi menu before a costly deep dive.
Play them as drills: one dive trainer, one rush trainer, one recipe trainer. Then return to Dave and notice how much calmer your air gauge feels.
Bottom line
Dave the Diver did not win because pixel sushi is cute. It won because Mintrocket paired a corporate-safe engine with an anti-corporate loop: dive alone, serve together, never waste a run.
Unity made that shippable to every screen that matters. Appetite made it memorable. If you are choosing tech for your own cozy-management hybrid, choose the boring engine that lets you add menus for five years, and spend your originality on what hunger does to oxygen.
Top comments (0)