Surviving Mars

Every colony sim has to answer one question: what is trying to kill the colony, and when does it stop? Surviving Mars, developed by Bulgaria’s Haemimont Games (with technical work from Abstraction) and published by Paradox Interactive, has an unusually sharp answer to the first half and a badly muddled answer to the second.

Released on March 15, 2018 into a genre already crowded with Banished clones and Factorio disciples, Surviving Mars found its niche by refusing to start with people. No colonists land first. Machinery lands first — unmanned rockets carrying RC rovers, drone hubs, and enough prefab to bootstrap a power grid on a procedurally selected patch of Martian regolith. Humans are a late-game reward, not a starting condition, and that inversion is the single smartest structural decision the game makes. The first act is a pure infrastructure puzzle: wind turbines and solar arrays feeding moisture vaporators and MOXIE oxygen plants, cables and pipes threaded across broken terrain, with a drone economy hauling metal, concrete and polymers between them. Nothing is alive yet. Everything can still fail.

Paradox positioned it as the mid-budget, high-concept counterweight to their grand strategy portfolio — an approachable sim with genuine mechanical teeth — and the market broadly agreed. The original earned a solid critical reception (a Metacritic of 76) and a Steam audience that reads as “loved for what it does, forgiven for what it doesn’t.”

That original is no longer for sale. It was delisted following the 2025 remaster, Surviving Mars: Relaunched, which has been received coolly — dragged down by launch bugs and, more pointedly, by asking longtime owners to repurchase content they already had. The game got a fresh coat of paint and lost some goodwill in the trade.

• Drone-First Logistics: Drone hubs, RC rovers and shuttle networks form an autonomous hauling layer to configure rather than command. Building placement is the gameplay — a hub sited fifty meters wrong starves a production chain hours later.
• Mission Sponsors & Commander Profiles: NASA, ESA, SpaceY, China, or a rogue billionaire. Each rewrites the funding model, research rate, rocket capacity and starting cargo. This is the closest the game gets to a difficulty slider, and it’s a genuinely good one.
• Domes and Individual Colonists: Every colonist is a discrete entity with traits, specializations, comfort needs and morale. Dome composition is a real balancing act — and a mismanaged one can spiral into a dome-wide suicide crisis.
• Randomized “Mysteries”: A mid-run narrative event injects a story crisis — sentient machines, alien artifacts, the spheres — that reframes a stable colony as a horror scenario. The best content in the game, and the most underused.
• Terraforming & Expansion Content: Later releases push toward a breathable atmosphere, underground caverns and asteroid mining, converting the endgame from maintenance into planetary engineering.
• Deep Mod Support: Extensive, well-adopted, and quietly load-bearing for the game’s long-term value.

The Gameplay Loop

The opening act works, and it works for a specific reason: scarcity with visible causality. When a colony fails early, the game makes the cause legible. A cold wave froze the water pipes; the vaporators stopped; the oxygen tank drained; the spare polymers for redundancy went to a second concrete extractor instead. That chain reads end to end. There’s no black-box happiness stat quietly poisoning the city. A resource decision gets made, the planet punishes it, and the line can be traced.

Dust storms, meteor showers and cold waves aren’t set dressing — they’re targeted attacks on the colony’s topology. A meteor strike severs a cable and suddenly the drone hub that would repair the cable has no power to run its drones. That deadlock, the infrastructure loop eating its own tail, is the best thing Surviving Mars does. Few builders model failure cascades this cleanly.

Then the colonists arrive, and the game changes genre without telling anyone.

The Difficulty Cliff

This is the central flaw, and it isn’t subtle. Once a colony achieves self-sufficiency — reliable power surplus, water margin, a functioning food chain, enough drones to self-repair — the game stops having an opinion about what anyone does. Disasters become nuisances the automated repair network absorbs without player input. The research tree, which early on gates survival, becomes a queue of percentage improvements to systems that no longer strain. Money stops mattering. The tension of the early hours is simply gone by mid-game.

What replaces it is administration. Late-game play devolves into colonist micromanagement: shuffling specialists between domes, hunting down the one unemployed botanist, tuning filters in an interface built for small populations that gets crowded as the colony grows. This is the exact failure mode of the genre, and Surviving Mars falls into it face-first. The Mysteries are the intended answer — a scripted crisis to reintroduce stakes — but they’re one-shot per run and heavily front-loaded in impact. A rogue AI is thrilling the first time and a known quantity the second.

The expansions patch this unevenly. Terraforming toward a breathable atmosphere gives the late game an actual objective, which is more than the base game manages. But it’s a goal, not a threat. Nobody is being tested anymore; a progress bar is being filled with excellent art direction.

Onboarding Friction

The tutorial is inadequate and the UI is worse. Surviving Mars asks players to reason about power grids, water networks, resource priorities, drone assignment radii and shift schedules, then explains perhaps half of them and buries the rest in tooltips of wildly inconsistent quality. New players don’t fail because the game is hard. They fail because drone hubs have a coverage radius that’s easy to never find out about until a third dead colony.

That’s a design failure, not a difficulty setting. The systems are elegant; the game just refuses to show its work. Community guides and mods do a substantial amount of teaching that should have shipped in the box.

The retro-futurist art direction is the game’s most durable asset — clean, optimistic, mid-century, all white polymer curves and geodesic glass against rust-colored rock. It looks like a 1960s magazine’s idea of the future, executed with restraint. Domes read as fragile bubbles of green in an enormous indifferent landscape, and that contrast does more emotional work than any of the writing. The soundtrack matches it: sparse, warm, a little melancholy, and paired with in-fiction radio that undercuts the isolation just enough.

The UI is the opposite. It’s dense, flat, and organized around what the systems are rather than what the player is trying to do. Finding a specific colonist, diagnosing why a factory is idle, or auditing a resource shortfall all take more clicks than they should. Form and function aren’t fighting here — function just never showed up.

Performance is the recurring technical concern, and it’s structural: large, mature colonies are expensive by design. Every colonist is simulated as an individual pathing agent, and the cost compounds exactly when the sprawling megacolony the game spent the whole run encouraging has been built. A game of this simulation depth is heavy on hardware precisely at the moment of success. Relaunched targeted this, but its cool reception suggests the remaster traded old instability for new.

Steam is the lead platform — it is the only platform. There is no iOS, Android, or Switch version; searches of the Apple lookup API, Google Play and nintendo.com surface only unrelated Mars-themed titles. Treat any store listing claiming otherwise as a knockoff.

That exclusivity suits it. This is a mouse-and-keyboard game in its bones: precise cable routing, drag-placed pipe networks, rapid camera work over uneven terrain, and constant context-menu diving. A controller would be a compromise; a touchscreen would be a joke.

The Steam ecosystem also carries real weight here. Workshop mod support is extensive and load-bearing — UI overhauls, difficulty rebalances and quality-of-life fixes patch several of the problems above, and a modded install is a materially better game than a vanilla one. Less charming is the storefront itself: Paradox’s aggressive expansion cadence produced a long, confusing DLC ledger, and the delisting of the original in favor of Relaunched means new buyers face a premium base price plus a content catalog, while veterans were asked to pay again for what they owned. That decision, more than any bug, explains the remaster’s cool standing.

Surviving Mars is a good game with a structural defect it never fixed across a full remaster. The opening is excellent — a tense, elegant survival puzzle wrapped in some of the best art direction the colony sim genre has produced. The ending doesn’t exist. The colony gets safe, the planet stops caring, and the player is left doing HR for a thousand people through a menu system that wasn’t built for it.

That’s still worth playing for the right audience. The original’s warm reception is honest: this is a game people love intensely for a specific stretch and then drift away from, without resentment. Buy it for the first act. Install mods for the second.

The purchase itself needs care. Relaunched at a premium price with a lukewarm reception is a harder recommendation than the original ever was — wait for a sale, check whether current patches have settled the launch issues, and don’t assume the remaster is strictly the better version. Owning the 2018 build with expansions is, for now, arguably the stronger position, and the fact that a publisher managed to make that true is its own small indictment.

Haemimont built a machine that conveys the thinness of Martian air. It just never figured out what to do once the colony learned to breathe.