1992
Dune II invented the RTS, and then everyone forgot the interesting part
Almost every convention you think of as “how strategy games work” was decided by one Westwood game in 1992. It also had one idea the genre quietly dropped, and it was the best one.
Dune II: The Building of a Dynasty came out in 1992, and it is one of the most quietly influential games ever made — in the specific sense that if you have played any base-building strategy game in the last thirty years, you have been playing a modified version of it without necessarily knowing.
It was not the first real-time game with strategy in it. Herzog Zwei got there in 1989 and deserves its credit. But Dune II is the one that fixed the template, and the template held.
What it decided
Every one of these was in that game, and every one of them is still the default:
- Select units with a mouse, order them with a mouse. Obvious now. Not obvious then.
- Build your base from a sidebar menu, rather than by selecting a building and drilling into it. This is why you can watch the battle while building — a subtler decision than it looks.
- One harvested resource, collected by a dedicated unit and turned into a single currency.
- A tech tree, where each structure unlocks the next.
- Fog of war over unexplored ground.
- Three asymmetric factions sharing a common core, so a campaign could be replayed from another angle.
- Mission briefings between levels, so the campaign was a story rather than a difficulty ramp.
Westwood then made Command & Conquer out of essentially the same design three years later, better in every respect, and the genre was set.
The part that did not survive
Here is the thing I actually want to argue about. Dune II’s economy was a vehicle, and the genre has mostly moved away from that, and it was a mistake.
The spice harvester was slow, expensive, unarmed, and it had to drive out into the open, sit still in a dangerous place, and drive back. In the Dune setting it also had a worm eat it occasionally, which is the single best expression of "resource extraction carries risk" any game has managed.
Compare that to how most modern RTS economies work. Worker units shuffle between a mineral patch and a depot fifteen metres from your town hall, inside your own defended perimeter. Or a number simply goes up. Both are economically identical to the harvester, and both are dramatically inert.
The vehicle harvester turns the map into a place. There is somewhere your money is, somewhere your base is, and a road between them that both players can see and neither fully controls. Nearly every interesting aggressive decision in the genre falls out of that one fact.
Think about what it gives you for free:
- Raiding has a target. You do not have to invent a reason to attack early — there is a 700-credit unarmed vehicle out in the open, and both players know it.
- Expanding is a real decision with a real cost. A closer field is more money per minute; a further one is safer or richer. That is a genuine trade with no correct answer.
- Defence has a shape. You are not defending "your base", you are defending a route. That is a much more interesting spatial problem.
- Fields run dry, which puts a clock on the map and stops the game becoming a fortification contest.
Worker-based economies got a lot of things in exchange — finer-grained economic control, a smoother difficulty curve, better competitive depth. I am not claiming StarCraft is wrong. I am claiming something specific was traded away and hardly anyone noticed, because the thing traded away was drama rather than depth.
What I kept
When I built TANKS, this is the mechanic I was least willing to modernise. It works exactly as it did in 1992, with 2026 conveniences:
- Ore fields are scattered around the map and each holds a finite amount.
- A harvester routes to the nearest one on its own, mines for about six seconds, drives home and unloads 500 credits.
- A round trip to a mid-map field is roughly 22 seconds, which you can time, which means you can plan around it.
- It is unarmed and it has 400 hit points, so it survives being shot at for just long enough that pulling it out is a real decision rather than a formality.
The result is that almost every match has the same shape as the ones I remember: the fighting is not really about the fighting, it is about whose harvesters are still running. Which is what Dune II worked out in 1992, and which the genre has spent thirty years slowly abstracting away.
If you want to play the originals rather than read about them, the games like Dune II page has the list — Dune Legacy is a free open-source reimplementation of the actual thing, and OpenRA runs the 1998 remake.
I made the game this is all about. It is free, there are no advertisements and there is nothing to buy — but if you got something out of reading this, there is a tip jar, and it pays for the next batch of levels.
Reading about it is slower than playing it.
TANKS runs in this browser tab. No download, no account, no launcher.
Questions people actually ask
What was the first real-time strategy game?
It depends on the definition. Herzog Zwei (1989) is usually credited as the first real-time game with strategic base and unit management. But Dune II (1992) is the game that established the template everything since has followed: mouse-driven selection, sidebar base building, a harvested resource, a tech tree and fog of war.
Why is Dune II important?
Because Command & Conquer, Warcraft, StarCraft, Age of Empires and effectively every base-building RTS descend from its structure. It was made by Westwood Studios, who then made Command & Conquer out of essentially the same design.
Read next
Games like Dune II
The 1992 game that invented the genre, and everything that inherited its harvester-and-sidebar loop.
Nothing ever replaced Command & Conquer, and I have been looking for twenty years
The genre split into esports and 4X, and the thirty-minute base-building game in the middle quietly stopped being made.
The RTS economy guide
How many harvesters, when to expand, how to protect income, and why the richer player usually wins.