M.U.L.E.: The Economics Game That Still Teaches

There is a moment in M.U.L.E. that has never been bettered and, as far as I can tell, has never been properly copied either. The auction opens. Four cursors sit on a vertical price axis. Buyers drift down from the ceiling, sellers climb up from the floor, and the price is wherever two of you happen to touch. There is no bidding step, no going-once, no turn order. The market is a physical space and you are standing in it, moving, watching three other people move, running out of time.

Everything the game wants to teach you about scarcity happens in those thirty seconds, and it happens to your hands before it happens to your head.

M.U.L.E. was released in 1983 for the Atari 800, one of Electronic Arts’ first batch of titles — the run that also included Pinball Construction Set, Hard Hat Mack and Archon, and which arrived wrapped in the album-sleeve packaging and the “Can a Computer Make You Cry?” advertising that EA used to announce it took this seriously. It was built by Ozark Softscape, a small outfit working out of Little Rock, and designed by Dani Bunten Berry. The C64 conversion followed in 1984.

I want to be careful here: this one arrived slightly ahead of me. I came to it later, on a machine that had already been sitting in the corner for a while, which is the honest position for most people reading this. That is fine. M.U.L.E. survives the revisit better than nearly anything else from its year, and the reason is that its systems are load-bearing rather than decorative.

The colony is the point, and you are the problem

You are one of four settlers on Irata — Atari backwards, which tells you where the code started — and you take turns claiming land, buying M.U.L.E.s from the store, outfitting them to harvest one resource, and dragging them out to a plot.

The resource chain is where the design gets clever. Food buys you time: a player short of food gets less of the next turn, which is the single most elegant punishment in 8-bit design because it degrades your ability to fix the problem. Energy runs the M.U.L.E.s, so an energy shortfall stops production on plots you have already paid for. Smithore is what the store uses to build new M.U.L.E.s, which means the supply of the thing you use to produce everything is itself produced by the players. And crystite is a pure export, sold off-world at a price that fluctuates outside anyone’s control.

That last chain is the one that catches people. If nobody plants smithore, the store runs dry, M.U.L.E. prices spike, and the entire colony’s productive capacity collapses — including yours, including the player who was winning. You are competing with three other people inside a system that requires the four of you to keep it fed. The colony gets a collective score at the end and can fail as a colony; only then does the game bother to say who was richest.

So it is a competitive game with a cooperative failure state welded to it. Bunten Berry was interested in what happens to people in that space. It turns out what happens is that you start talking — offering, refusing, threatening to refuse — and M.U.L.E. extracts that talk without a single line of dialogue system, because the pressure comes from the resource graph.

The walk to the plot

Before any of that, there is the installation. You buy a M.U.L.E. from the store, outfit it for one resource, and then you have to walk it to the plot you want it on, with a timer, and it will bolt if you dawdle or misjudge the approach. A M.U.L.E. that gets away costs you the animal and the money and the turn.

This is the sort of thing a modern design meeting would delete inside a minute. It is friction with no strategic depth: the optimal play is always to install it correctly, so all the mechanic does is occasionally punish you for being slow.

It survives because of what it does to the turn’s rhythm. The land grant and the outfitting are decisions made at leisure — you are looking at a map and thinking about mountains versus river. The walk is a physical act under a clock, and it converts a plan into a commitment in about four seconds. By the time the M.U.L.E. is planted you have done something with your hands, and the plot on the board is now yours in a way that clicking a confirmation box never manages.

Every good economic game since has needed some version of this — a moment where the abstraction touches a body. Most of them cannot find one, which is why so many of them read as spreadsheets with art.

The auction is the whole thesis

Most economic games hand you a market as a menu. You open a screen, you see a number, you click buy. The transaction is arithmetic.

M.U.L.E. makes the market a room. Buyers descend, sellers ascend, the trade fires the instant two cursors meet, and the clock is running the whole time. The store sits at the edges with a fixed spread — it will always sell to you high and buy from you low — so it functions as a wall that defines the worst deal you are willing to accept.

What this produces is knowledge you cannot get from a number. You learn what a seller’s hesitation looks like. You learn that the player who dives fast is desperate and can be squeezed, and that squeezing them may cost you the colony three turns later. You learn that a shortage is not a price on a screen, it is a scramble of bodies in a narrow column with the timer bleeding out. When the store runs out of food and all four of you are buyers with nobody selling, the screen tells you the exact shape of a famine: everyone drifting down, nobody rising to meet them, nothing happening at all.

Try to build that as a menu and it becomes a lookup table. The real-time double auction is the mechanic doing the teaching, and it is why the game still earns the word economics rather than merely being about money.

The rubber band, and why calling it cheating misses it

M.U.L.E. biases its random events. Good luck — a windfall, a fortunate meteorite, an unexpected payment — lands disproportionately on whoever is losing. Bad luck goes hunting for the leader. This is stated in the design and it is deliberate.

The usual objection writes itself. Rubber-banding is a cheat; it punishes competence; it flatters the weak player with results they did not earn. Mario Kart has been taking that argument in the teeth for thirty years.

The objection is wrong here, and the reason is that M.U.L.E. is a four-player game played in a room over about an hour. A runaway leader in that format creates three bored people who have twenty minutes of nothing left and a host who has to pretend the evening is still happening. The rubber band exists to keep four humans in contention until the last auction, because the game’s actual subject is the negotiation between them, and negotiation requires that everyone still has leverage. Simulation fairness is a lower priority than the room, and Ozark chose the room.

There is also a systems argument. Because the events are biased towards the trailing player, the leader has to plan around a hostile stochastic environment while the field has a tailwind. That is a real skill test, and a harder one than optimising in a neutral world. The design taxes the lead, and the tax is legible, and you play around it.

Compare it to how Archon handles the same problem from the other direction — the board’s cycling light and dark squares mean position value keeps shifting under everyone, so no one holds a permanent advantage on the geometry alone. Free Fall and Ozark were EA labelmates working the same year on the same insight: a two-hour session needs the outcome live at the end.

What it actually teaches

Nobody comes out of M.U.L.E. knowing what a Gini coefficient is. The game teaches interdependence under competition, which is the harder and more useful lesson, and it teaches it as reflex. You come away having felt that hoarding a monopoly on energy is a strong move that can also drown you; that a shortage in one input silently prices everything downstream; that timing is a resource; that the person you most want to beat is the person you most need to still be solvent. Those are the intuitions the subject actually runs on, and M.U.L.E. installs them in an hour without once explaining itself.

It also refuses the flattery. There is no way to grind out a win. There is no build order that beats four attentive people. The game is content to let a colony fail and tell you so.

Bunten Berry spent the rest of her career arguing that computer games were a social medium and that the industry was busy building solitary ones. She died in 1998, before the online era proved her right in a way she would probably have found funny and slightly infuriating. The line she is quoted for — that nobody on their deathbed wishes they had spent more time alone with a computer — reads on the box art of her own game as a design principle rather than a slogan.

Where to play it

The C64 and Atari 800 versions are both preserved and both fine; the Atari original has a slight edge in feel. Emulation handles it, and there is a long tail of fan remakes and spiritual sequels of wildly variable quality. Ignore the ones that turn the auction into a menu, which is most of them.

The real requirement is three other people in the same room with four controllers, and that requirement is the reason the game is talked about more than it is played. Meet it once and you will understand why people who were there have never quite stopped going on about it.

For the other 1983-84 EA-shaped experiment that gambled on a systems idea over a genre, see Archon; for the mid-80s design that made a simulation out of a household instead of a planet, Little Computer People; and for what happened when one designer decided a game could be several genres at once and still cohere, Pirates!.