History
OXO: The 1952 Computer Game That Played Tic Tac Toe
Meet OXO, A. S. Douglas’s 1952 noughts-and-crosses program for EDSAC and an early milestone in graphical computer games.
- Related grid games
Three-in-a-row relatives existed long before the modern paper-and-pencil rules.
- Noughts and crosses
The familiar British name and modern play appear in print.
- OXO
A. S. Douglas programmed noughts and crosses for EDSAC.
- A universal small game
The grid now travels easily across paper, classrooms, browsers and phones.
Jump to section
OXO was a 1952 noughts-and-crosses program written by A. S. Douglas for EDSAC at the University of Cambridge. It is frequently described as one of the earliest graphical computer games.
Why Douglas built it
Douglas created the program during doctoral research concerned with human-computer interaction. Noughts and crosses made a useful demonstration: the rules were familiar, a person could choose moves, the machine could reply, and the changing board could be displayed.
What the machine was like
EDSAC was a room-sized stored-program computer, not a personal console. OXO used its cathode-ray-tube display and telephone-dial input. The experience was local to the machine; it was not packaged and distributed as a commercial game.
Why OXO matters
The program joined input, game logic, computer response, and visual output in a recognizable interactive loop. That makes it historically important even though debates about “the first video game” depend on definitions and include other early experiments.
What it did not invent
Douglas did not invent noughts and crosses. He implemented an established game on a new medium. That distinction is why the broader history of Tic Tac Toe has no single inventor while OXO has a named programmer and date.
Today a browser can search the complete game tree instantly. The path from OXO to the modern computer mode is a compact example of how an old game helps each new generation learn what computers can do.
Interaction was the important demonstration
OXO did more than calculate a hidden answer. A person selected moves, the machine responded, and the changing board appeared on a display. That loop—input, state update, algorithmic decision, visible output—is recognizable in modern interactive software even though the hardware and interface were radically different.
Tic-tac-toe made the demonstration legible. An observer did not need specialist mathematics to understand whether the machine had blocked a line or chosen a square. Familiar rules turned computation into behavior.
Why accounts of “the first video game” vary
Historical superlatives depend on definitions: electronic or computer-controlled, graphical or text-based, interactive or preprogrammed, publicly displayed or built for research. OXO is securely important as an early graphical computer game, but a careful account should state the feature being claimed instead of treating every definition of first video game as settled.
It was created for a particular research machine and was not a mass-market release. Its audience and distribution were unlike a modern commercial game.
What a reconstruction can teach
An emulator or reconstruction helps modern readers understand the constraints: limited memory, specialized input, and a display designed for technical work rather than entertainment. Compare that environment with a browser version that adds touch input, responsive graphics, difficulty levels, and accessibility labels. The core state still fits in nine cells, yet the surrounding product has changed enormously.
Douglas’s achievement is clearest when described precisely: he built a documented, interactive noughts-and-crosses program on EDSAC and used a familiar game to demonstrate human-computer interaction.