Supremacy: The Empire Game With One Bad Decision in It

Probe Software’s Supremacy: Your Will Be Done shipped for Amiga and Atari ST in January 1990, with C64 and MS-DOS conversions following over the next two years, published by Virgin under a title that became simply Overlord in North America — a rename that tells you nothing about what the game actually asks of you. It looks like a space war. You have a solar system, a rival empire, warships to buy and worlds to take. The marketing sold conquest, designed by David Perry and Nick Bruty at Probe. The design underneath sold something narrower and more interesting: a lesson in delayed consequence, dressed as a galaxy.

The empire is a supply chain

Every colony you hold has to be kept fed and powered — farming stations for food, solar satellites for energy, mining stations for fuel — while a colony’s growing population generates the tax revenue that actually pays for your fleet. New worlds don’t start ready to inhabit either: an atmospheric processor has to terraform a planet before a colony can take hold there at all, which means every expansion is a standing investment before it produces anything back. None of this is optional bookkeeping sitting behind the war. It is the war, and the rival empire fighting you for the same system is playing exactly the same supply game against you at the same time.

Neglect the farming and energy side of a colony to fund a war chest instead, and nothing happens immediately. The colony keeps running. Battle cruisers keep launching, built and armed with a mix of homing missiles, ballistic missiles and hover tanks. Turns later, population growth stalls, tax income falls with it, and the industrial output you were counting on to keep building ships starts falling too, because a shrinking, underfed colony can’t sustain the output a healthy one could. This is the one bad decision the title promises, and it is rarely a single dramatic choice. It’s a supply line left unattended while your attention was on the war map instead of the colony screen, made in the opening turns of a campaign that runs for hours. By the time the population figures confirm the mistake, the fleet you built on the back of it is already committed — in transit, in a fight, or holding a border that no longer has the tax base to keep replacing losses.

Why the delay is the design

Most strategy games of the era gave you instant feedback: build a unit, see the unit, spend the resource, watch the number drop. Supremacy’s economy runs on a lag between cause and visible effect that most players never plan for on a first session, because nothing in the interface warns you it exists. The colony screen shows current output, not the trend that output is on. You have to infer the trend yourself, from memory of what you set production toward and how long ago, which is a much harder task than reading a single number off a dashboard.

That lag is the actual challenge of the game, more than the AI opponent contesting the same system. The opposing empire plays a competent-enough game of grabbing neutral worlds and building fleets, but it doesn’t out-think you. Your own economy does that, if you let a production choice run unmonitored for too long. The interesting comparison is M.U.L.E., the economics game that still teaches — Ozark Softscape’s 1983 four-player trading game built its whole tension around visible, immediate auctions and production numbers everyone could see and react to in real time. Supremacy takes the opposite approach: it hides the same kind of resource logic behind a delay long enough that by the time you notice the shape of the mistake, the shape is already load-bearing.

The knock-on effect reaches the fleet screen too. Battle cruisers cost fuel to build and fuel to move, so a fuel shortfall arriving on the heels of a food shortfall doesn’t read as one problem with two symptoms — it reads as two separate crises arriving at once, because the interface never draws the line connecting them for you. Working that connection out yourself, once, is what separates a first campaign that collapses in the middle third from a second campaign that doesn’t.

The ships are the reward, not the game

It’s worth being honest about where Supremacy is weaker: home-computer versions give you no direct control once a battle cruiser engages an enemy world’s defences, and the fight resolves on the numbers you arrived with rather than on anything you do in the moment — the later NES conversion would add direct control over missile launches and hover-tank deployment, a layer the earlier home-computer editions simply don’t offer. You commit a fleet armed with whatever mix of missiles and tanks your colonies could afford, the game resolves the exchange, and you find out afterwards whether the economy behind that fleet was sound. This can read as a letdown if you came to Supremacy expecting a war game with the drama up front. It plays better if you read the combat as a verdict — a receipt for choices made turns earlier on colony screens you may already have stopped watching closely. The war is the scorecard. The economy is the game.

That framing also explains why exploration and colonisation matter more than they first appear to. Terraforming and settling an extra world isn’t really about the territory; it’s about diversifying which colony is carrying which part of your production load, so a shortage on your home world doesn’t starve the whole empire at once. Spreading risk across worlds is the actual defensive move Supremacy wants from you, and it’s rarely explained, only discovered — usually after the first campaign that didn’t spread it has already gone wrong, ground down all the way to a final defence of your own starbase.

The genre it’s arguing with

Supremacy arrived at the tail end of a run of British space-empire games that treated the galaxy as a spreadsheet with a starfield bolted on, and it’s fair to place it as a bridge between that home-computer strategy tradition and the American 4X games that would formalise the genre over the following decade. Master of Orion II: the 4X peak arrived five years later with turn-based clarity, a research tree you could actually see coming, and combat you could plan against rather than merely react to. Supremacy’s real-time economy, running whether or not you’re watching the screen, is a rougher and less forgiving idea than anything in that later, tidier lineage — closer to a simulation you’re trying to steer than a system you’re meant to master outright.

That roughness is also the reason the game rewards a second playthrough more than most strategy titles of its generation. Knowing the lag exists changes how you read the colony screen entirely. You start checking trend rather than total, watching for the direction a number is moving in rather than trusting the number itself, and the whole interface — which felt opaque and unhelpful on a first attempt — starts reading like a set of warning lights you’d simply never learned to watch. Modern strategy design has mostly solved this exact problem with trend arrows and projected-shortage warnings built straight into the UI. Supremacy predates that convention by years, and the absence is close to the entire lesson: it forces you to build the mental model the modern interface would otherwise build for you, and that model sticks with you longer for the effort.

The studio behind the numbers

Probe Software built much of its reputation through the eighties on licensed arcade conversions, the kind of contract work that pays a studio’s bills without necessarily building its identity. Supremacy reads like a different kind of project under Perry and Bruty — an original design the team clearly had room to think through properly, rather than a conversion brief with a deadline stapled to it. Reviewers at the time agreed: Amiga Format and CU Amiga both scored the Amiga version in the low nineties, with Jeroen Tel’s soundtrack singled out as some of the best SID and Amiga music of the period. The confidence shows in exactly the places a rushed licence job would have cut corners: the colony-management screens are deep enough to reward the extra hour spent studying them, and the supply economy is consistent enough that once you understand the shape of the trap, you can see it coming on every subsequent world you settle.

Virgin’s publishing muscle got the game onto four platforms across three years at a time when cross-format ports frequently meant a diminished experience on anything that wasn’t the lead platform. The C64 conversion is honest about its limits, and the MS-DOS version arrived over a year behind the 16-bit originals — but the underlying economic model survives both downgrades intact, which says something about how load-bearing the supply logic was to the design from the start. A game that leans on spectacle rarely survives a conversion to a machine with a fraction of the colour palette and none of the processing headroom. A game that leans on a numbers problem travels remarkably well, because the numbers don’t change shape depending on the hardware rendering them.

The lesson outlives the game

What Supremacy leaves behind, more than any specific memory of a fleet battle or a colonised world, is a small piece of design literacy: the habit of asking what a number is doing over time rather than trusting what it says right now. That’s a transferable skill in strategy games generally, from city-builders where a happiness stat quietly bleeds out over several in-game years before the riots start, to grand strategy titles where a treasury can look healthy for a dozen turns before a war debt catches up with it. Supremacy isn’t remembered as fondly today as some of the games that came after it and solved the same interface problem more gracefully, but it got there first, and it got there without softening the lesson with a warning light. You were meant to learn to watch the trend because the game wouldn’t watch it for you. Most modern strategy design would consider that a usability failure. Played today, knowing what to look for, it reads as close to the entire point.

What actually plays now

Emulated Amiga and ST versions run cleanly through the usual routes, and the interface — dense, text-and-readout heavy, unapologetically a management screen before it’s a war screen — holds up better than the combat, which is the part of the game that has aged into a fairly simple resolve-and-wait affair on the home-computer versions. Go in expecting a resource-management puzzle with a war attached rather than the reverse, and the pacing makes sense. Go in expecting Dune II-style unit command, and the game will feel slow and undirected, because it was never trying to be that. The C64 version trims the ambition to fit the machine and is worth trying only after the 16-bit editions, where the extra colour and the faster redraw make the colony screens legible enough to actually watch closely — which, as it turns out, was always the point.

Spoilers below

The genuine trap in a full campaign is pouring a colony’s output into battle cruisers to rush an early fleet, which is the single most tempting move the game offers and the one the design punishes hardest — food and energy supply on your home world falls behind just as the fleet you built with it starts taking losses against the rival empire’s border defences, and there’s no faster way to rebuild a starving colony than there is to rebuild a destroyed fleet. The campaign’s endgame, a push all the way to the enemy’s own starbase in the system the game calls Yottsu, is unreachable for an empire that let this happen early and never recovered its supply base. Recovering from it is possible but slow, and most losing campaigns trace back to exactly this choice, made once, forgotten, and left to run for turns longer than anyone would choose to leave it if the game had simply told them what it was doing.