Management sims have made me into a villain. If playing a simulator game is like playing god, then I’m certainly a wrathful one. In Factorio, I remind myself “the factory must grow” as I fight bug hordes that, understandably, attack my base as pollution saturates their settlements. In Frostpunk, I force workers to endure 18-hour shifts and sawdust-gruel meals, all while living in an impromptu shantytown.
Terra Nil is a balm for this kind of aggressive gameplay. In this “reverse city builder,” as developer Free Lives has described it on the game’s Steam page, you rewild desiccated and barren land across four major biomes in a series of four scenarios. It’s a game for this era of climate anxiety, where we’ve gone past the climate “point of no return.” Healing landscapes across Earth’s biomes is the ultimate comfort fantasy — especially amid a sea of games premised on destruction and dominion — where reversing the toll of habitat destruction comes at the click of the mouse. But the game also has an identity crisis, where these meditative tile-placement mechanics chafe against the complexity of its late-game systems.
Terra Nil’s themes of rebirth and reconstruction are translated beautifully through its delightful visuals and an ASMR-like soundscape full of clicks, rain sounds, and wind riffling gently through the grass. It’s viscerally satisfying, almost dreamlike work, slowly reviving dead, crisp-looking land with lush pine, bamboo, or mangrove forests. You revive oceans with coral reefs and thickets of kelp in which sea turtles can thrive. You rebuild ice caps, making a home for virtual penguins, even as they’re threatened in real life.
Early gameplay is purely atmospheric, in the vein of tile placement games like Dorfromantik. You start by placing windmills — and later, more advanced forms of electricity generators — followed by a building that turns desiccated land into soil, and then a building that lays a grassy field over that arable land. This phase of restoration is like a game of Tetris, wherein you try to restore as much gridded surface area as you can. The isometric-style scenario maps are procedurally generated, and fairly modest in size. Rewilding the map earns you points, depicted as leaves in the UI, which you can spend on additional structures. The only real strategy is making sure you don’t spend these points before you’re able to lay down a building that earns you more points.
In the next phase of restoration, you start to diversify these ecosystems, placing structures that can spawn forests or meadows in the surrounding land — granted it meets ecosystem requirements on that particular tile. For example, have you put grassland down on a tile, is it adjacent to an ocean or river, or have you done a controlled burn for the sake of seeding a forest? In later scenarios, this might include global requirements like humidity level or altitude.
This is where it gets convoluted: After the first scenario, the game doesn’t clearly explain the order of operations, or how these particular tools can layer atop one another, and only lets you undo the most recent building that you’ve placed. You might be waiting to place a meadow in the tundra after doing a controlled burn, for example, without realizing you should have done it in the opposite order. At this point, the scenario is lost, and you must restart. This is, in part, just how simulators are — you mess things up and you start over. But other sims tend to give earlier red flags when things are going haywire, and hint at options for digging yourself out. In Terra Nil, you discover that you’ve messed up, and that’s basically it. Furthermore, there’s also a pretty steep complexity jump between scenarios two and three, so these confusing mechanics add yet another layer of shock to the sudden failures.
The final phase of restoration — the cleanup phase — is also convoluted, though I do respect the political implications this step represents. It suggests that the work of human intervention in the environment should conclude by removing evidence of industrial presence. The animation and sound design in this section are also stellar: When a building is deconstructed, it explodes into the ether, emitting a nice crunchy sound. But the actual mechanic behind it is a huge pain, requiring you to make buildings either accessible by river or through air tram stops — which can only be built on rocks.
Suddenly, the increasingly deindustrialized landscape becomes replete with rails and artificial rivers, as you do scummy reverse engineering to attempt to remove the buildings you’ve placed, once again clogging up the landscape. You can’t beat a scenario until all the buildings are gone. I would be happy to buy into the idea that rewilding — and removing your footprint — ostensibly requires complex machinery, but the jump from atmospheric to complicated once again is grating.
It’s hard to tease out that line between satisfying complexity and convoluted complexity. Typically, weighty simulators tend to feel a little more open-ended — even if they have scenario objectives — because the buildings you place interact with one another in real time or chain to form automation. In Terra Nil, it feels stiffer, more set in stone, waiting for you to proceed with your plan. The emphasis is less on interlocking structures and more on building placement and order of operation. There are many ways you could rewild the landscape you’re given, but you’re always building on top of what you’ve already created. Where you might expect a city builder to open up and widen the amount of creative expression you’re given as the game advances, Terra Nil offers a shrinking palette of possibility. What you can place, in the late-game, depends entirely on what you’ve already done to the map. You can run out of space to create certain biomes or to lay down the tramways you need. You can back yourself into corners without even realizing it.
This isn’t the end of the world. After beating a scenario, you’re given the option to replay the biome as many times as you’d like — and thanks to procedural generation, these maps will always have some variance. Within these replays, you can be incredibly intentional, or you can put the game on “gardener” setting for a truly atmospheric experience. I’ve personally enjoyed playing it on this chiller mode, and just stuffing my biomes full of as many animals as possible.
I’m also open to the idea that the preconceived notions I have around the genre — and the way that restorative, or visually lush, games get labeled “cozy” — negatively influences my perception. To be clear, I love “cozy” games. But I also think that some people view these games pejoratively, with the assumption that joy and comfort can’t be taken as seriously as heavier topics — or that the gameplay in these games is more straightforward. I try to resist the assumptions around this label by looking for games that have something interesting to say about healing, while taking risks with the game mechanics used to execute that vision. I just think that Terra Nil is trapped somewhere between being atmospheric and complex, without a strong on-ramp for its difficulty curve.
As I played, I kept thinking, Let me love you, as I built so many pylons and tramway poles just to complete a scenario. I played the demo over and over again when it was first released last year, excited for what would come. Even if some of the gameplay feels unnecessarily rigid, I still have a lot of respect for the way this game emphasizes environmental stewardship, especially within a genre that tends to focus on the exact opposite. Despite the roadblocks, that sense of wonder is enough to bring me back into the game’s world.
Terra Nil will be released on March 28 on on Android, iOS, and Windows PC. The game was reviewed on PC using a pre-release download code provided by Devolver Digital. Vox Media has affiliate partnerships. These do not influence editorial content, though Vox Media may earn commissions for products purchased via affiliate links. You can find additional information about Polygon’s ethics policy here.