The History of Sudoku

How a constrained Latin square became the world's favourite logic puzzle.

Sudoku looks ancient, but the puzzle we know today is barely fifty years old. Its story runs from 18th-century mathematics, through a 1970s American magazine, to a Japanese naming and a British newspaper craze that carried it around the world.

Here is how a humble grid of numbers travelled from Leonhard Euler to your morning paper — and your phone.

Latin squares: the mathematical ancestor

The deep ancestor of Sudoku is the Latin square — an n×n grid filled with n symbols so that each symbol appears exactly once in every row and every column. The name honours the 18th-century mathematician Leonhard Euler (1707–1783), who used Latin letters as the symbols in his work on these arrays.

A finished Sudoku is simply a 9×9 Latin square with one extra rule: each of the nine 3×3 boxes must also hold the digits 1–9 once. That added box constraint is what turns a free-form Latin square into the puzzle we recognise, and it is why every valid Sudoku is also a Latin square but not the other way round.

Number Place: the modern puzzle, 1979

The modern puzzle was most likely designed by Howard Garns, a 74-year-old retired architect from Connersville, Indiana, and first published — anonymously — by Dell Magazines in 1979 under the name "Number Place". It already carried the now-familiar 9×9 grid, nine 3×3 boxes and a single unique solution reachable by logic.

For a few years it stayed a quiet American magazine puzzle, with no hint of the global fame to come.

Japan names it Sudoku, 1984

The puzzle reached Japan through Maki Kaji and the publisher Nikoli, which ran it in the magazine Monthly Nikolist in April 1984. Kaji called it "Sūji wa dokushin ni kagiru" (数字は独身に限る) — "the digits must be single" — and the name was soon shortened to Sudoku.

Nikoli also refined the puzzle into the form solvers prize today: symmetric patterns of givens and a modest, hand-tuned clue count, so each puzzle is both elegant to look at and solvable by pure reasoning.

The 2004 global boom

The worldwide craze began with Wayne Gould, a retired judge from New Zealand who was living in Hong Kong. He spotted a part-finished Sudoku in a Tokyo bookshop in 1997 and spent about six years writing a computer program that could generate fresh, uniquely-solvable puzzles on demand.

Gould persuaded The Times of London to print his puzzles, and it launched Sudoku on 12 November 2004. Within months the puzzle leapt to newspapers across Britain, the United States and beyond, then onto websites and apps — where it remains one of the most played logic puzzles in the world.

Why it endures

Sudoku has outlasted most puzzle fads for a simple reason: it needs no language and no arithmetic, only logic. The digits are just symbols, so the same puzzle works in any country, and a single unique solution means every step can be reasoned out rather than guessed.

That blend of universal rules and pure deduction is what carried it from a Latin square to a daily habit for millions.