This article contains affiliate links. We may earn a small commission at no extra cost to you.
The Nintendo Entertainment System had 2 kilobytes of RAM.
Two kilobytes. That is roughly the size of this article's first three paragraphs. Inside that absurdly small box of memory, Shigeru Miyamoto built Mushroom Kingdom. He built Hyrule. He built the running man in red overalls who would become the most recognised fictional character on earth.
He didn't do it in spite of the constraints. He did it because of them.
The golden age of video game design — roughly 1984 to 1999 — produced some of the most ingenious, enduring, and beloved interactive experiences in human history. And it did so almost entirely by accident. The designers of that era were not trying to build philosophical systems of reward and failure, patience and persistence, spatial memory and problem solving. They were trying to get a game to run on hardware that could barely remember its own name. The philosophy emerged from the limitation. The greatness emerged from the poverty of the tools.
This is the story of how the most primitive machines imaginable produced the most durable design masterpieces — and what exactly made them so hard to forget.
The Man Who Built Worlds in Kilobytes
Shigeru Miyamoto was born on November 16, 1952, in Sonobe — a rural town near Kyoto, Japan — where he spent his childhood doing something that would later define an entire medium. He explored.
He roamed forests, waded through rice paddies, and — on one occasion that lodged permanently in his imagination — pushed into a dark cave in the hillside beyond his house. He stood at the entrance for a long time before stepping in. The darkness, the unknown geography, the slow reveal of what lay ahead: that feeling of cautious discovery became the emotional core of The Legend of Zelda, released in 1986.
Miyamoto had no coding experience when he joined Nintendo in 1977 as the company's first staff artist. He studied industrial design — not computer science. He drew manga. He built toys. When Nintendo's president Hiroshi Yamauchi handed him his first game design assignment — salvage a failing arcade cabinet called Radar Scope by converting it into something new — Miyamoto responded with Donkey Kong. The year was 1981. The game had one button and could not scroll. Every constraint in it was the direct result of what the hardware could physically do.
But the constraints weren't frustrations to Miyamoto. They were parameters. A game, he believed, should teach itself. It should introduce every rule through play rather than instruction. It should use the smallest number of elements — repeated, combined, and varied — to create the richest possible experience. On early consoles, he had only kilobytes of memory to deliver entire worlds. Every sprite, every movement, every sound had to be optimised until it earned its place in the code.
The first level of Super Mario Bros., released in 1985, uses just nine gameplay elements. Nine. And yet the first 30 seconds of World 1-1 — the mushroom, the Goomba, the pit, the block, the jump — taught an entire grammar of play to millions of people who had never held a controller before. Miyamoto called it "teach the game by playing the game." He never used a manual if the world itself could show you what to do.
Technology is not the point. The point is what you do with it. A new chip doesn't make a game fun — a good idea does.
— Shigeru Miyamoto, Nintendo game designer and creator of Super Mario Bros. and The Legend of Zelda
Three Lives and a Hard Lesson
The three-lives system — one of the most iconic features of 1990s gaming — was not a design philosophy. It was a practical inheritance from the arcade.
In an arcade cabinet, lives were the mechanism that extracted money from players. Die three times, insert another coin. When home consoles arrived, the lives system came with them — not because designers had theorised its value, but because it was simply the convention that game designers knew. The NES cartridges that carried these games had no viable save system. The battery-backed save memory that appeared in The Legend of Zelda in 1986 was a genuine technical breakthrough, expensive and relatively rare. For most games, there was no mechanism to store progress. When you died, you went back. When you ran out of lives, you went back further. When you turned off the console, everything was gone.
Miyamoto, in a 1998 interview with the Japanese games criticism magazine Game Hihyou, described his design philosophy around difficulty with characteristic clarity: a game is better if you have to start the level again, because it increases intensity and makes the experience more urgent. The risk of death before the final moment is what makes arrival at that moment matter. Remove the stakes and you remove the feeling.
This was not cruelty. It was architecture. A Super Mario Bros. level that the player cannot fail to complete teaches nothing about its own geography. A level that kills you — at the same pit, the same Goomba, the same falling platform — three times before you clear it encodes that geography permanently. The punishment was the instruction. The restart was the lesson.
The technical limitation of no saves produced, accidentally, a design principle of rare power: earned progress. You did not skip past difficulty. You could not. Every inch of the map existed in your memory because your failure had written it there.
Want to Relive It?
Handheld Retro Game Console — 220 Classic Games Preloaded
Rechargeable battery, 3.0" LCD screen, 220 preloaded retro games. The closest thing to that Saturday morning feeling — no cartridges, no TV required. A great gift for anyone who grew up with the originals, or wants to show a kid what gaming used to feel like.
Check Price on AmazonThe Soviet Puzzle Built on a Computer With Less RAM Than a Calculator
In June 1984, in a dark room at the Dorodnitsyn Computing Centre of the Soviet Academy of Sciences in Moscow, a speech-recognition researcher named Alexey Pajitnov was doing something he was absolutely not supposed to be doing.
He was building a game.
Pajitnov was working on an Elektronika 60 — a Soviet clone of an American computer architecture, rack-mounted, with no built-in display and a total RAM of 4 to 8 kilobytes. The machine had no graphical capabilities whatsoever. It couldn't draw shapes. Pajitnov's first version of Tetris was rendered entirely in text-based characters: brackets and spaces arranged on a terminal screen to suggest falling blocks.
He had been inspired by a childhood puzzle game called pentominoes — twelve geometric shapes made of five squares each, arranged into a rectangle. The digital version reduced the shapes to four-block combinations — tetrominoes — because five was too complex for the hardware to handle. The constraint produced the design. The impossible machine produced the perfect game.
Tetris was completed on June 6, 1985. Within two weeks, Pajitnov later recalled, "every single computer in Moscow" had a copy. It spread without the internet, without marketing, without advertising — through floppy disks passed hand to hand through a country with no commercial software market. People could not stop playing it. Pajitnov himself admitted he could not stop playing his own prototype. He knew, from the first hour, that he had found something that would not let go.
What made Tetris different from almost every game before or since was that it had no end state anyone could reach. There was no princess, no final boss, no credits screen waiting as a reward. It simply became faster, and harder, and more demanding, until the player finally ran out of space. The loop was its own point. The satisfaction was pure pattern — the quiet snap of a line completing and vanishing, the brief reprieve before the next piece fell.
Zelda's Open World Was a Memory Accident
The Legend of Zelda, released in 1986 for the Famicom Disk System in Japan, was the first major console game to include a battery-backed save system — allowing players to preserve progress between sessions. It was also, in its design, something that had never existed before: an open world.
Miyamoto's childhood of exploration in the Kyoto countryside had given him a specific feeling he wanted players to experience: the sensation of not knowing what lay beyond the next hill, the courage required to enter the dark cave, the reward of discovery earned through persistence. He told interviewers that Zelda, for all its epic mythology, was fundamentally about hiking.
But the open world of the original Zelda was also shaped, profoundly, by what the hardware could not do. When Miyamoto and his team worked on the Super Famicom sequel — A Link to the Past, released in 1991 — he had envisioned a truly open-ended game with branching paths and multiple simultaneous solutions. The memory constraints of the cartridge made this impossible. Features were cut. The ambition was pressed into a shape the hardware could carry. The bomb-able walls that revealed hidden rooms — a fragment of the original open-world vision — survived into the final game and became one of its most beloved design features.
What emerged from those forced compressions was a kind of design purity that later, more powerful hardware made harder to achieve. When you have infinite storage, you add more. When you have 2 kilobytes, every element in the game has to justify its existence. The economy of constraint produced masterworks that later generations of far more capable machines have struggled to equal.
Go Deeper
The Legend of Zelda: Hyrule Historia
Make sure to check out the other installments in this unparalleled collection of historical information on The Legend of Zelda franchise with the New York Times best selling The Legend of Zelda: Art & Artifacts and The Legend of Zelda: Encyclopedia. Also look for The Legend of Zelda: Breath of the Wild — Creating a Champion for an indepth look at the art, lore, and making of the best selling video game!.
Check Price on AmazonMiyamoto argued in a 1998 interview that a game is better if the player must restart a level after dying, because it increases intensity and makes the experience more urgent. The inability to save progress — a hardware limitation — had accidentally become his most powerful design tool.
One Screen, One Couch, One Shared World
There was something else the 1990s gaming environment produced that was not a design decision at all.
It was social.
In an era before online multiplayer, before game-sharing, before streaming — when the experience of playing a game was fundamentally tied to a physical room and a physical screen — gaming was communal by default. You watched someone else play. You waited your turn. You argued about the map. You drew the dungeon layout on graph paper because the game could not store that information for you. You knew where the hidden room was because you and your friend had spent a Saturday afternoon trying every bombable wall in the palace until one of them gave.
The 1990s games industry understood this well enough to design around it. Street Fighter II, released by Capcom in 1991, built its entire commercial success on head-to-head competition — two players, one screen, no network required. The game sold because it was most alive when played against a human being sitting six inches away. Mario Kart, released for the Super Nintendo in 1992, created the genre of competitive couch racing that remains culturally dominant three decades later. GoldenEye 007, released in 1997, is credited with defining the template for first-person multiplayer shooters — a template built around four players sharing a single TV screen in a dormitory room.
The hardware had not provided an alternative. There was no internet. There were no servers. There was one screen. And so designers built games that were most powerful when experienced together, in person, where the social friction and cooperation and competition were the product of physical proximity rather than digital architecture.
The Decade That Accidentally Got Everything Right
By the mid-1990s, the console wars between Sega and Nintendo had pushed hardware forward faster than anyone had anticipated. The Sega Genesis launched in 1989 with its 16-bit processor. The Super Nintendo matched it in 1991 with richer colour and superior sound. By 1996, Nintendo's 64-bit machine was rendering fully three-dimensional worlds — and Miyamoto, almost immediately, faced a new problem.
Too much was possible.
In the same 1998 interview, Miyamoto spoke candidly about the challenge of abundance. His work had always been dominated, he said, by the clamoring for "more replayability" and "more content." Now that hardware could provide it, the discipline that constraint had enforced began to relax. More levels could be added. More features could be included. The economy of necessity gave way to the economy of plenty — and something of the original precision began, gradually, to dilute.
It is no coincidence that the games most consistently cited as the greatest ever made — Super Mario Bros. 3, The Legend of Zelda: Ocarina of Time, Tetris, Super Metroid, Street Fighter II — almost all date from the era of maximum constraint. Not because the designers of that era were more talented than those who followed, but because the tools they were given forced a kind of ruthlessness in design that abundance makes optional.
Pajitnov built Tetris on a machine with no display, no graphics, and less memory than a modern pocket calculator. It has now sold 520 million copies and been described as the most addictive game ever created. Miyamoto built Super Mario Bros. inside 2 kilobytes, with nine gameplay elements and a man who could only jump. That man has appeared in more than 200 games and become the most recognised fictional character on the planet.
The limits were not the obstacle to greatness. They were the condition of it.
The Point
The 1990s produced the most beloved games in history not because the designers were working toward some grand philosophy of challenge and reward, but because their hardware left them no alternative. Two kilobytes of RAM forced economy. No save system forced earned progress. One shared screen forced social play. One button forced clarity of purpose. The accident of those constraints produced the design masterpieces that every generation since has tried — and largely failed — to replicate. Shigeru Miyamoto needed to build worlds out of nothing. Alexey Pajitnov needed to make brackets and spaces feel like something worth spending your life on. The poverty of the tools was the source of the richness of what they made. That is not a lesson about video games. It is a lesson about what limitation does to imagination when given no escape route.
Sources
- Wikipedia — Tetris — en.wikipedia.org
- Wikipedia — Alexey Pajitnov — en.wikipedia.org
- Wikipedia — History of video games — en.wikipedia.org
- Wikipedia — 1990s in video games — en.wikipedia.org
- Shmuplations — Shigeru Miyamoto talks Game Design (1998), translated from Game Hihyou magazine — shmuplations.com
- CNN — Tetris: The Soviet 'mind game' that took over the world — edition.cnn.com
- Game Developer — Miyamoto: 'You can't just throw every good idea you have into a game' — gamedeveloper.com
- New York Film Academy — What Nintendo Can Teach Us About Game Design — nyfa.edu
- SJSU ScholarWorks — Luan Tran, The Impact of Shigeru Miyamoto on the Game Design Industry, ART 108: Introduction to Game Studies (2020)
- University of Toronto Mississauga — Falling Blocks: The Legacy of Tetris — collections.utm.utoronto.ca
Maya Thornton is a cultural historian and long-form writer for The Verified Post, specialising in the untold design and human stories behind the technologies that shaped modern life. She covers science history, cultural artefacts, and the accidental genius hiding inside everyday objects.



