Nuclear Option

Combat flight simulation has spent twenty years splitting into two hostile camps. On one side, DCS World and its descendants: 400-page manuals, HOTAS setups that cost more than a console, forty-minute cold-start procedures before you taxi. On the other, Ace Combat: gorgeous, weightless, a rail shooter wearing a fighter jet’s clothes. The middle ground — the space where a game respects physics without demanding you memorize a radar’s PRF modes — has been effectively abandoned. Publishers decided it didn’t exist.

Shockfront Studios apparently didn’t get the memo. Nuclear Option, currently in Early Access on Steam, is a near-future combat flight sim built around roughly a dozen fictional aircraft — nimble fighters, ground-attack jets, VTOL gunships, heavy bombers, transports — flying across 100km-wide maps of ocean, desert, mountain, and city. The fictional airframes are the first tell that this team knows what it’s doing. By inventing its aircraft rather than licensing them, Shockfront sidesteps the accuracy arms race that makes DCS modules so expensive and so slow to ship, while keeping every system plausible enough that nothing breaks the fiction.

The positioning has found an audience. The Steam reception has been strongly positive, and — the number that actually matters — sentiment has held up as the audience widened past the early adopters. That’s rare. Most Early Access darlings see their recent score sag as the honeymoon ends and the backlog of unfixed complaints accumulates.

What makes this more remarkable is the scale of the operation. This is a tiny indie team competing against Eagle Dynamics and Bandai Namco, and it’s competing on the one axis both of those competitors ignore: the feeling of flying a real machine that you can actually learn in an evening. That positioning isn’t a compromise between two genres. It’s a third thing, and the reception suggests a lot of people have been waiting for it without knowing what to ask for.

• Part-by-Part Damage Modeling: Aircraft are constructed from up to 50 individually simulated physics parts. Take cannon fire to a wing and it doesn’t decrement a health bar — the wing physically detaches, and you fly what’s left. Asymmetric thrust, lost control authority, a spin you may or may not recover from. Damage becomes a flying problem instead of a countdown.
• A Persistent Battlefield: AI fleets, convoys, SAM networks, and airfields prosecute their own war whether you show up or not. Kill a radar site and the corridor opens. Crater an airfield and enemy sorties thin out. Your strikes shift the front line, which is the promise DCS’s dynamic campaign has been making for over a decade.
• Nuclear Weapons as a Real Mechanic: Both tactical and strategic nukes, and they’re not a cutscene. Deploying one is a tactical decision with a blast radius, a delivery problem, and consequences — the ugliest, most spectacular button in the game.
• The Full Sensor Stack, Simplified: Radar and infrared missiles, guided bombs, cruise missiles, cannons, plus flares and jammers on defense. The systems are modeled deeply enough that countermeasure timing is a skill, but abstracted enough that you don’t need a checklist.
• Creation Tools: A custom mission editor and Steam Workshop support, which is the pressure valve for the thin official content.

The Flight Model Is the Whole Argument

Everything Nuclear Option is trying to prove lives in how the aircraft handle. The claim — “between arcade and study-sim” — is the most overused sentence in this genre and almost always means “arcade with a stall warning.” Here, by design, it doesn’t.

The 50-part physics construction isn’t a marketing spec; it’s the architecture that produces the game’s best moments. Because each component is simulated independently, the sim doesn’t need scripted failure states. It just computes what happens when a control surface is gone. Stalls, spins, and g-induced blackouts emerge from the model rather than being bolted onto it, and so does the far more interesting case: limping home in an airframe that’s no longer symmetric, trading altitude for control, learning by feel how much rudder it takes to keep the nose somewhere near where you’re pointing.

That’s the loop. Not the kill — the flight home after the kill.

Onboarding: Brilliant Design, Absent Instruction

Here’s the blunt contradiction at the center of this game. The controls are superbly accessible. Mouse-and-keyboard works. A gamepad works. You do not need a $500 HOTAS, and the fact that a physics-driven sim plays well on a controller is a real engineering achievement, not a concession.

The teaching is not accessible at all. There are effectively no tutorials and no campaign structure to scaffold you through the systems. So a new pilot faces a strange split: flying is intuitive within ten minutes, but understanding radar modes, missile employment envelopes, and when to punch flares is entirely self-taught. The game hands you a low skill floor and then declines to mention where the stairs are.

The community papers over this with YouTube guides and Discord. That works, but it’s an outsourced solution to a first-party problem, and it’s the single highest-leverage fix available. A well-built tutorial series would convert a meaningful fraction of bounced players into permanent ones.

The Content Problem

Nuclear Option’s systems are far ahead of the content built on them. Single-player scenarios are thin and mission variety is limited — the most consistent criticism of the game, and it’s fair. You are given a superb sandbox and comparatively little authored reason to keep entering it.

The persistent AI battlefield partially covers this. Because the war runs without you, sorties generate their own stakes, and the strategic layer means a mission has consequences beyond the score screen. But dynamic systems produce situations, not stories. There’s no campaign arc, no escalation, no authored sequence that teaches you a new capability and then tests it. The shape of the war reveals itself quickly.

The mission editor and Workshop are the intended answer, and the community has been productive. Still — Workshop content is a supplement to a content plan, not a substitute for one. Modders can build missions. They can’t build a campaign’s sense of consequence.

Multiplayer and the AI

Multiplayer is player-hosted, for co-op and PvP, and it inherits every classic weakness of that model: no guaranteed stability, and at off-peak hours a browser that’s a ghost town. When a populated server is running, large-scale PvP over a 100km map is the game at its best. When it isn’t, you’re back to the sandbox. Buy this for the multiplayer alone and you’re gambling on your timezone.

The AI is the other rough edge. It’s competent enough to make the persistent war legible, but its decisions can read as bugs rather than tactics, which undercuts the immersion the flight model works so hard to build. It’s the seam where Early Access is most visible.

The Developer

One asset doesn’t appear in any feature list: Shockfront ships. The update cadence has been frequent and substantial, and the developer is visibly responsive to the community. In Early Access, that’s not a nice-to-have — it’s the entire basis on which you decide whether to buy now or wishlist and wait. The track record here justifies buying now.

Nuclear Option is not a beautiful game in the way a modern AAA sim is beautiful, and pretending otherwise would be a disservice. Textures are functional. Terrain across the ocean, desert, mountain, and city maps reads clearly from altitude — which is the altitude that matters — but won’t survive close inspection on the deck.

What the art direction gets right is legibility and scale. On a 100km map, you need to instantly parse what’s a threat, what’s a target, and what’s terrain, and Nuclear Option’s clean, slightly austere presentation is built to deliver that. Missile smoke trails, the arc of a cruise missile going in, and the appalling bloom of a tactical nuke all carry weight. The nuke in particular is a genuine spectacle — the one moment the visuals reach past functional.

Night operations with NVGs deserve specific credit. The green-phosphor pass changes how you fly, forcing sensor reliance and turning terrain-following into real work rather than an aesthetic filter.

Performance is the quiet win on paper. Simulating 50 physics parts per aircraft across fleet-scale AI battles on a 100km map sounds like a recipe for a stuttering mess, and the engineering discipline required to keep it light on mid-range hardware is the least-discussed achievement in the package. Expect UI rough edges typical of Early Access — menus and readouts lack polish — but the simulation itself is built to hold up under load.

Steam, PC only, and the platform choices are deeply intertwined with the game’s design.

Input flexibility is the headline. This is a sim you can play on mouse and keyboard, which sounds trivial until you consider that the genre’s incumbents treat a HOTAS as a hardware requirement. A gamepad works well too. Full stick-and-throttle setups are supported and reward the investment, but they’re an upgrade path, not an entry fee — the reason Nuclear Option’s audience is many times the size of what a comparable sim would draw.

Steam Workshop integration is the other platform-native strength, and it’s load-bearing given the content situation. One-click mod subscription lowers the friction of community missions to nearly zero, which is precisely what a content-light sandbox needs.

The player-hosted multiplayer, meanwhile, is where PC-native design cuts both ways. No dedicated official servers means no guaranteed population, but it also means the community can run persistent campaigns, custom rulesets, and private squadron nights without waiting for the developer. That’s the traditional PC bargain, and if you’ve played sims before you know the terms.

Steam’s refund window is the practical advice here: it’s a low-risk way to find out whether the flight model clicks. The design suggests it will.

Nuclear Option is the rare Early Access game that has already justified its existence. The design thesis — that there’s a huge audience for a combat flight sim with real physics and no homework — was supposedly disproven by a decade of publisher indifference. The sustained positive reception suggests the thesis was right and the market was simply not being served.

The flaws are real and won’t be softened here. The absent tutorial is a genuine design failure — it actively costs the game players who would otherwise have loved it. The single-player content is too thin for the quality of the sandbox it sits in. The AI needs work. Multiplayer depends on strangers being awake.

But every one of those is a content-and-polish problem, and content-and-polish problems get solved by developers who ship. The hard part — a physics model that produces emergent, learnable, deeply satisfying flight — is already done, and done well by a studio with a fraction of its competitors’ resources. We rate it a clear recommendation with real reservations: buy it if the pitch appeals, and go in knowing the tutorial and campaign are on you.