How Snake Went From a 1970s Arcade Cabinet to 400 Million Phones

The snake game did not begin on a phone. It began in a plywood arcade cabinet in 1976, two decades before Taneli Armanto wrote a single line of code for the Nokia 6110. The concept traveled through arcade halls, home computers, office PCs, and programming textbooks before landing on the device that made it famous. Each stop along the way stripped something away or added something new, until the game arrived at a form that was perfectly suited to a 1997 mobile phone.

This is the story of that 20-year journey: where the snake concept came from, how it evolved across platforms, and why the technical constraints of a business phone turned out to be exactly what the game needed.

Blockade (1976): the first game with a growing trail

Blockade was released by Gremlin Industries in October 1976. Gremlin was a small arcade manufacturer based in San Diego, California, competing against Atari, Midway, and a handful of other companies for floor space in bars and arcades across the United States.

The game was simple. Two players each controlled a cursor that moved continuously across a grid, leaving a permanent trail behind it. You could change direction (up, down, left, right) but you could not stop. The goal was to force your opponent to crash into a wall, your trail, or their own trail. The last player moving won the round.

Blockade did not have food, scoring, or a single-player mode. It was a pure two-player competition, closer in spirit to Tron light cycles (which would not appear until 1982) than to the Snake game people know today. The cabinet used a monochrome CRT display and simple four-directional joysticks.

What Blockade established was the core mechanic: a moving element that leaves a trail, and the challenge of navigating without hitting anything. That mechanic would survive every adaptation that followed. The details changed, the platforms changed, the audience changed. But the trail never went away.

1977 to 1979: the arcade clones and the food mechanic

Blockade's success (it was profitable, though not a blockbuster) sparked immediate imitation, which was standard practice in the arcade industry of the 1970s. Within a year, several clones appeared.

The most important was Hustle, released by Gremlin themselves in 1977 as a follow-up. Hustle added a critical new element: dots scattered across the playing field that players could collect for points. Eating a dot made your trail grow longer. This was the first time the snake mechanic included food collection and a growing body, the two features that would define every subsequent version of Snake.

Other arcade entries included Checkmate (Midway, 1977) and various unnamed trail games that appeared in European and Japanese arcades. None of them achieved major commercial success. The arcade market in the late 1970s was dominated by Space Invaders (1978) and its descendants, and trail games could not compete for attention.

But the genre was quietly migrating. By 1978, home computer versions were appearing. The arcade was the birthplace, but the snake concept would find its real audience elsewhere.

1978 to 1985: Snake on home computers

The late 1970s and early 1980s saw an explosion of home computing. The TRS-80 (1977), Apple II (1977), Commodore PET (1977), BBC Micro (1981), ZX Spectrum (1982), and Commodore 64 (1982) all created platforms where hobbyist programmers could write and share games.

Snake variants (usually called Worm or Snake) became a staple of this ecosystem. They were ideal beginner programming projects: simple rules, grid-based movement, no complex graphics. Computer magazines like Byte, Creative Computing, and the UK's Computer & Video Games published type-in listings for Worm games that readers could enter by hand and run on their machines.

These home computer versions introduced a crucial change. Where Blockade and its arcade descendants were two-player competitive games, the home computer versions were single-player. Instead of trying to trap an opponent, you guided a worm around a field collecting food, growing longer, and trying to survive as long as possible. Your enemy was your own body and the walls. This single-player food-collection format became the definitive version of the snake concept.

The shift made sense for the platform. Home computers often had only one keyboard, and many owners were the sole computer user in their household. A single-player game that could run on minimal hardware and be coded in BASIC was exactly what the audience needed. The multiplayer arcade format evolved into a solitary puzzle, and that puzzle was more compelling than the original.

1991: NIBBLES.BAS brings Snake to every PC

When Microsoft released MS-DOS 5.0 in June 1991, it included QBasic, a free BASIC interpreter, along with two sample programs: GORILLA.BAS (a turn-based artillery game) and NIBBLES.BAS (a snake game). Both were included as programming examples to demonstrate QBasic's capabilities, not as commercial products.

NIBBLES.BAS was a text-mode game that rendered the snake and food using colored characters on an 80 by 25 grid. It supported one or two players and included nine levels with different wall layouts. The snake moved at a steady pace, the player collected numbered food items, and the game ended when the snake hit a wall or its own body.

MS-DOS 5.0 shipped on millions of PCs worldwide. NIBBLES.BAS became, by accident, one of the most widely distributed snake games before the Nokia era. Office workers found it. Students found it. Aspiring programmers found it and, more importantly, read the source code, modified it, and used it as a launching point for their own projects.

NIBBLES.BAS also established a distribution pattern that would prove prophetic: a game included free on a device that people bought for other reasons. People did not buy MS-DOS 5.0 to play NIBBLES. They bought it to run their PC. The game was a bonus. Six years later, the same pattern would play out on a much larger scale.

1996: a phone that needed a game

By the mid-1990s, Nokia was the world's largest mobile phone manufacturer. The company was headquartered in Espoo, Finland, and employed tens of thousands of engineers working on the next generation of GSM handsets.

In 1996, Nokia was developing the 6110, a premium business phone. The device had a 32-bit ARM processor, a monochrome LCD screen measuring 84 by 48 pixels, and limited onboard memory (estimates place the total available RAM for applications at roughly 8 kilobytes). The product team wanted to include a few small games as a quality-of-life feature. Not as a selling point, not as a product differentiator, but as something businesspeople could do during downtime.

Taneli Armanto, a software engineer at Nokia, was assigned to create three games for the 6110. He developed Snake, Memory (a card-matching game), and Logic (a code-breaking puzzle). The games had to be tiny in code size, use minimal processing power, and operate within the phone's existing memory budget without affecting core functions like call handling.

The engineering constraints that shaped the game

The Nokia 6110's hardware imposed constraints that would have killed most game concepts. But for Snake, each constraint turned out to be an advantage.

The display: 84 by 48 pixels, two states only. Each pixel was either on (dark) or off (the greenish LCD background). No grayscale, no color, no backlight. This eliminated any possibility of complex graphics, detailed sprites, or visual decoration. Everything on screen had to be functional. The snake was dark rectangles. The food was a dark rectangle. The border was dark rectangles. That extreme visual clarity meant the player could read the entire game state in a single glance, which was essential for a game where the snake never stopped moving.

The input: a numeric keypad. The 6110 had no directional pad, no joystick, and no touchscreen. Armanto mapped the four directions to the number keys: 2 (up), 4 (left), 6 (right), 8 (down). These keys formed a rough diamond shape on the keypad that worked intuitively for directional control. More importantly, the keypad naturally limited the game to four directions. There was no temptation to add diagonal movement, analog speed control, or other complexity that the input could not support.

Memory: roughly 8 kilobytes for game code. The phone's RAM was shared between the operating system, the radio stack, the phone book, text messaging, and any additional features. The games received a tiny slice of what was available. This meant no elaborate level designs, no stored assets, no animation frames. The game state was a short list of coordinates (the snake's body segments) and a single food position. That was all the memory could hold, and it was all the game needed.

Battery: the game could not drain the phone. The 6110 was a communication device first. Any game that noticeably reduced battery life would have been cut before shipping. Snake's minimal rendering (updating only the pixels that changed each frame) and simple computation (checking a few collision conditions per tick) used negligible power. A player could spend 30 minutes playing Snake and see no measurable battery impact.

Why Snake worked on a phone when other arcade games would not have

Armanto could have adapted any number of classic game concepts for the 6110. He chose Snake, and that choice was more consequential than it might appear.

Consider the alternatives. A Space Invaders clone would have required a fire button, fast-moving projectiles, and enemies with distinct visual profiles, none of which the 6110's input or display could support well. A Pac-Man clone would have needed a large enough maze to be interesting, ghost AI, and smooth animation across the entire screen, all demanding more memory and processing than the phone could spare. A puzzle game like Tetris could have worked (and later did appear on phones), but it required a wider screen and more responsive input than the keypad could offer for piece rotation.

Snake fit the phone's constraints naturally, not despite them. The game needed only four-directional input (which the keypad provided). It rendered only a few changing pixels per frame (which the processor could handle). It stored its state as a simple list of coordinates (which fit in the available memory). And it was playable in one to three minute sessions (which matched how people used their phones). No other classic game concept mapped to the 6110's limitations as cleanly as Snake did.

The design principles that made the game satisfying (instant understanding, visible progress, fair difficulty) were amplified by the phone's constraints, not diminished by them. The tiny screen made the growing snake more dramatic. The limited input made the controls more intuitive. The lack of save states made each session feel self-contained and low-commitment.

December 1997: launch, and the accidental phenomenon

The Nokia 6110 launched in December 1997 with Snake, Memory, and Logic pre-installed. Nokia's marketing focused on the phone's business features: infrared data transfer, customizable profiles, and its slim form factor. The games were mentioned in passing, if at all.

Users found Snake anyway. The game was accessible from the phone's main menu in two taps. There was no login, no tutorial, no terms of service to accept. You pressed a key and the snake started moving. The zero-friction access was the result of the phone's simplicity, not a deliberate growth strategy. But the effect was the same: once people found the game, they played it. And they talked about it. And the people they talked to already had the game on their own phones.

Nokia noticed. The 6110 sold well, and Snake was generating word of mouth that no marketing budget could have purchased. Nokia began including Snake (and later Snake II) on every phone they manufactured. As Nokia's market share grew, so did Snake's installed base. By 2002, Nokia held over 35 percent of the global mobile phone market. Every single one of those phones had Snake.

Between 1997 and 2007, Snake appeared on over 400 million Nokia devices. At least 350 million unique people played it. For a decade, it was the most-played video game on the planet, though few people thought of it as a video game. It was just the thing you did with your phone when you were waiting for something.

From Blockade to 400 million phones: what the journey reveals

The 20-year path from Blockade to the Nokia 6110 is a story about a game concept finding the platform it was built for. Blockade was too limited for long-term engagement (two players, no scoring, no progression). The home computer versions were too niche (hobbyist audience, no distribution). NIBBLES.BAS reached a larger audience but still required someone to actively seek out a game on their PC.

The Nokia 6110 solved the distribution problem in a way no previous platform could. The game was on the device when you bought it. You did not have to find it, buy it, download it, or install it. It was just there. And the phone's constraints, which would have limited most games, shaped Snake into something more focused, more readable, and more satisfying than any of its predecessors.

That combination of frictionless distribution and constraint-driven design is the reason Snake reached 400 million phones while Blockade is a footnote in arcade history. The core mechanic was always good. It took 20 years to find the right platform. When it did, the result was the most widely played game of its era.

You can play the game that 400 million phones delivered. This browser recreation preserves the original 1997 version with pixel-perfect accuracy. Read the how to play guide for controls and difficulty modes, or check the leaderboard to see how your score stacks up against players worldwide.