Who Created Sudoku Puzzles?
Sudoku was created by Howard Garns, a retired American architect from Indianapolis, who designed it in 1979 for Dell Magazines under the name Number Place. Maki Kaji of the Japanese publisher Nikoli renamed it Sudoku in 1984 and set the conventions we know today. Wayne Gould, a New Zealand-born judge, wrote the software that took it worldwide in 2004. Three men, three countries, three decades — and the inventor died without ever knowing what he had made.
Below: what each of them actually contributed, how Garns was identified long after his death, and why Euler and the 1890s French newspapers are not the answer people think they are.
Howard Garns: the inventor
Howard Garns (1905–1989) was an architect who worked for the firm Daggett in Indianapolis, Indiana. In retirement he became a prolific puzzle constructor, contributing regularly to Dell Magazines.
In 1979 Dell published his creation as Number Place. The format was already final: a 9×9 grid, digits 1 to 9, no repetition in any row, column or 3×3 box, with a set of givens producing exactly one logically reachable solution.
Nothing important has changed since. Every Sudoku published anywhere in the world today follows the rules Garns set out in 1979.
He died in Indianapolis in 1989, aged 84. Sudoku became a global phenomenon fifteen years later, under a Japanese name he had never heard.
How his authorship was discovered
Dell Magazines did not give puzzle constructors bylines. Puzzles appeared unattributed, with contributors listed collectively at the front of each issue.
Garns's authorship was established by researchers who cross-referenced those contributor lists against the issues containing Number Place. His name appeared in every issue that carried the puzzle and was absent from every issue that did not — a pattern far too consistent to be coincidence.
The identification is now widely accepted, though it rests on that circumstantial evidence rather than on any signed claim. Garns himself, as far as anyone knows, never spoke publicly about the puzzle.
Maki Kaji: the man who named it
Maki Kaji (1951–2021) founded the Japanese puzzle publisher Nikoli in 1980 and became known worldwide as the Godfather of Sudoku.
In April 1984 Nikoli published Number Place in its magazine Monthly Nikolist. Kaji gave it the Japanese name sūji wa dokushin ni kagiru — "the digits must remain single" — which he shortened to Sudoku.
Kaji habitually named puzzles after whatever caught his attention; several Nikoli titles came from horse-racing terms he liked the sound of. Sudoku was a rare case where the name states the rule itself.
He trademarked Sudoku in Japan but not internationally — a decision he later described with characteristic good humour as having cost him a fortune. He died in Tokyo in 2021, aged 69, having lived to see the puzzle played worldwide.
What Nikoli actually changed
Nikoli did not alter the rules, but it introduced two conventions that define the puzzle's modern character:
- A maximum of 32 givens. Dell's Number Place puzzles often carried more clues. Capping them forced more logic and less filling-in.
- 180-degree rotational symmetry in the clue positions. Purely aesthetic, and now near-universal.
Nikoli also insisted on hand-made puzzles, and still does. Human constructors design a deliberate solving path through the grid, which produces a different experience from computer-generated puzzles where clues are removed at random.
Wayne Gould: the man who spread it
Wayne Gould (born 1945) is a New Zealander who worked as a judge in Hong Kong. In 1997, browsing a Tokyo bookshop, he saw a partly completed Sudoku and was immediately taken with it.
Over the following six years he wrote a program capable of generating puzzles at consistent, controlled difficulty levels — a harder problem than it sounds, since difficulty depends on which techniques a grid requires rather than on clue count.
His distribution model is what actually caused the global craze. He gave the puzzles to newspapers free of charge, asking only that they print his website address.
The Times of London ran its first Sudoku on 12 November 2004. Within three months most British national papers carried one. By mid-2005 the puzzle was in newspapers across the United States, Australia, Canada, India and continental Europe.
And what about Euler?
The Swiss mathematician Leonhard Euler is regularly credited with inventing Sudoku. He did not.
Euler studied Latin squares in the 1780s — n×n grids where each symbol appears exactly once per row and once per column. That is two of Sudoku's three constraints, and it is a mathematical object rather than a puzzle.
Missing are the 3×3 box rule, the idea of removing entries to create a puzzle, and the requirement of a unique solution. Euler supplied an ancestor, not the invention.
The forgotten French puzzles
Between 1892 and 1895, French newspapers published number puzzles remarkably close to Sudoku.
Le Siècle ran a 9×9 magic-square puzzle with entries removed in November 1892. In July 1895, La France went further: a 9×9 grid using only the digits 1 to 9, with each row, column and 3×3 sub-square containing all nine.
That is Sudoku in all but name, eighty-four years early. The puzzles disappeared from French papers around the First World War, and there is no evidence Garns knew of them — making them an independent precursor rather than a source.
The 1979 Dell puzzle, described exactly
The first Number Place ever printed ran in Dell Pencil Puzzles & Word Games #16, May 1979. Anyone expecting a modern Sudoku would find it familiar but not identical. The grid was 9×9 with nine 3×3 boxes and the rule was already the rule we use today: every row, every column and every box holds 1 to 9 exactly once. What differed was the presentation. Dell printed some of the givens in boldface and told solvers those bold digits formed a hidden word or phrase once decoded — a habit borrowed from the word-puzzle magazines Garns worked in. The instruction line was long and apologetic, as if the editors doubted readers would understand a puzzle with no words in it at all.
The clue count also varied wildly. Early Dell Number Place puzzles could carry 35 or more givens, far above the 22–30 range typical of a modern newspaper Sudoku, because nobody had yet worked out how few clues a puzzle could survive on. The symmetry was inconsistent too. Garns produced the puzzles by hand, and hand construction favours whatever arrangement resolves cleanly rather than whatever looks elegant on the page.
Why the puzzle failed in America and succeeded in Japan
Number Place ran in Dell magazines for five years without becoming a phenomenon. The reason is mostly structural. Dell sold puzzle magazines to an audience that came for crosswords, cryptograms and word searches; a wordless logic grid was the odd item near the back. It had no name recognition, no newspaper slot, and no champion pushing it.
Japan supplied all three. Nikoli's readership was already trained on language-free logic puzzles — Nikoli built its business on them — so a number grid needed no explanation. Nikoli also ran a reader-submission model: solvers sent in puzzles they had built themselves, the magazine printed the best, and a construction culture formed around the format within a few years. That culture is what produced the aesthetic rules the world later inherited.
Was Sudoku invented or assembled?
The honest answer is that Sudoku as we know it was assembled from parts that already existed, and Garns is the person who assembled them. Latin squares had been studied for two centuries. Magic-square and number-placement puzzles had run in newspapers since the 1890s. The 3×3 box constraint — the piece that turns a Latin square into a puzzle a human can actually solve by reasoning rather than trial and error — is the contribution that makes Number Place a distinct invention rather than a reprint.
That box constraint matters more than it looks. Without it, a partially filled Latin square gives you almost no leverage: you can eliminate along a row and a column and nothing else. Add the box and every cell sits at the intersection of three overlapping constraints, which is what makes techniques like cross-hatching, hidden singles and pointing pairs possible. Garns did not write about why he added it, but the effect is the whole reason the puzzle works.
The people usually left out of the story
| Name | Contribution | Usually credited? |
|---|---|---|
| Dell editors, 1979 | Accepted and printed an unnamed wordless puzzle for five years | No |
| Nikoli's reader-constructors | Developed symmetry and clue-count conventions through competition | Rarely |
| Nobuhiko Kanamoto | Nikoli editor credited with shortening the name to Sudoku | Sometimes |
| Michael Mepham | Built puzzles and wrote the guides that taught British solvers the techniques | Rarely |
| Will Shortz | Traced the Dell contributor lists that identified Garns posthumously | Sometimes |
Will Shortz's role is worth stating plainly, because without it Garns would still be anonymous. Dell never printed bylines. Shortz compared the contributor credit lists across issues, noted that one name appeared in every issue carrying a Number Place and was absent from every issue without one, and identified Howard Garns that way. It is detective work on a magazine masthead, and it is the only reason the inventor has a name at all.
What Garns never knew
Howard Garns died in Indianapolis in 1989. Number Place was still a minor Dell feature. Nikoli had renamed it five years earlier, but nothing about that had reached the American magazines, and Garns received no royalty, no credit line and no indication that his puzzle had travelled. The global boom arrived fifteen years after his death.
He was 84 when he died, a retired architect who built puzzles in his spare time. The most-solved logic puzzle in human history came from someone who never learned he had made it.
Who deserves the credit?
| Person | Year | Contribution |
|---|---|---|
| Leonhard Euler | 1780s | Latin squares — the mathematical ancestor |
| French newspapers | 1892–1895 | Independent near-identical precursor, then forgotten |
| Howard Garns | 1979 | Invented the puzzle; rules unchanged since |
| Maki Kaji / Nikoli | 1984 | Named it, capped clues, added symmetry, built the culture |
| Wayne Gould | 1997–2004 | Generator software and free distribution to newspapers |
Garns is the inventor in the strict sense. But a puzzle nobody plays is not much of a puzzle, and without Kaji it would have stayed an obscure Dell filler, and without Gould it would have stayed a Japanese speciality.
Frequently asked questions
Who invented Sudoku?
Howard Garns, a retired architect from Indianapolis, created it in 1979. Dell Magazines published it as Number Place without crediting him by name.
Did a Japanese person invent Sudoku?
No. Maki Kaji of Nikoli named it and popularised it in Japan from 1984, but the puzzle itself was invented in the United States five years earlier.
Who is the Godfather of Sudoku?
Maki Kaji (1951–2021), founder of the Japanese publisher Nikoli. He gave the puzzle its name and established the conventions that define modern Sudoku.
Did Howard Garns know Sudoku became famous?
No. He died in 1989, fifteen years before the puzzle became a worldwide phenomenon under its Japanese name.
How was Howard Garns identified as the inventor?
Researchers noticed his name appeared in Dell Magazines' contributor lists in every issue containing Number Place and was absent from every issue without it.
Who brought Sudoku to the West?
Wayne Gould, a New Zealand-born judge working in Hong Kong. He found the puzzle in Tokyo in 1997, wrote a generator program, and gave puzzles free to newspapers in exchange for a credit link.
Did Euler invent Sudoku?
No. Euler studied Latin squares in the 1780s, which have row and column constraints but no 3×3 box rule, and are mathematical objects rather than puzzles.
Why is it called Sudoku?
It is short for the Japanese phrase suji wa dokushin ni kagiru, meaning "the digits must remain single". Su means number and doku means single.
Did Maki Kaji make money from Sudoku?
Far less than one might expect. He trademarked the name in Japan but not internationally, so publishers worldwide used it freely.
What did Nikoli change about the puzzle?
Two conventions: a maximum of 32 givens, and 180-degree rotational symmetry in the clue positions. The rules themselves were left exactly as Garns wrote them.
When did Sudoku first appear in a newspaper?
French newspapers ran very similar puzzles in the 1890s. The modern puzzle first appeared in The Times of London on 12 November 2004.
Is Sudoku a Japanese game?
The name is Japanese and Japan made it popular, but the puzzle is American. Nikoli introduced it to Japan in 1984, five years after its US publication.
Play a piece of history
Enter any grid into our free online Sudoku solver to solve it instantly, in the format one retired architect invented in 1979.
Next, read: Where Was Sudoku Invented? · When Did Sudoku Become Popular?
