A Short History of Number Puzzles: From Latin Squares to KenKen
A turtle shell, a Swiss mathematician, a Dell magazine, and a Yokohama classroom walk into a grid.
Before you pencil a single digit into today's grid, consider that you are joining a lineage thousands of years long. The history of number puzzles is not a tidy straight line from one clever inventor to the next. It is a family tree, with roots in ancient China, a branch in eighteenth-century Switzerland, a twig in a Manhattan puzzle office, and fresh growth in a small classroom in Yokohama.
The shared DNA is simple: take a grid, fill it with numbers, and obey a rule that admits exactly one answer. From that single idea grew the latin square, the puzzle we now call sudoku, and eventually KenKen. Here is how the pieces connect, with the real dates attached.
The oldest grid: magic squares
The first ancestor is the magic square, and the most famous is the Lo Shu, a three by three arrangement from ancient China. Legend hands it a lovely origin: a turtle crawled out of the Luo River with a pattern of dots on its shell, and those dots spelled the numbers 1 through 9. Arrange them the Lo Shu way and every row, every column, and both diagonals add up to 15.
The tidy details of the myth came later, but the documented square is genuinely old, appearing in Chinese texts more than two thousand years ago. What matters for our story is the idea it planted: numbers are not only for counting. Slotted into a grid under a rule, they become a puzzle, a small locked box that only one arrangement can open.
Euler and the latin square
Why we call it Latin
Fast forward to the 1700s and the Swiss mathematician Leonhard Euler (1707 to 1783). In a paper published in 1782 he studied grids filled so that each symbol appears exactly once in every row and every column. He used Latin letters for the symbols, so he named the objects carres latins, latin squares. That constraint, one of each symbol per line, is the exact skeleton every modern number grid still hangs on.
Euler also left behind a teaser that outlived him: the 36 officers problem, which asks whether six regiments of six ranks each can stand in a six by six grid with no repeats down any row or column. He suspected it was impossible, and much later mathematicians proved him right. The lesson is that serious mathematics and Saturday-morning puzzling grew from the same seed.
A latin square is a grid of n symbols in which each symbol occurs exactly once in every row and exactly once in every column.
Number Place: the sudoku origin story
The latin square sat mostly in math journals until an American puzzle-maker gave it a jolt. In May 1979, Dell Magazines printed a grid called Number Place, most likely built by Howard Garns, a retired architect from Indiana. Garns took the latin-square rule, carved a nine by nine grid into nine boxes, and pre-filled a handful of clues. He died in 1989, years before the world ever learned his name.
The puzzle crossed the Pacific, and in 1984 the Japanese publisher Nikoli renamed it Sudoku, short for a phrase meaning roughly the digits must stay single. Then a retired judge named Wayne Gould spotted it in a Tokyo bookshop, spent six years writing software to generate puzzles by the thousand, and talked The Times of London into running one in November 2004. Within months the craze had circled the planet.
- 11979: Number Place appears in a Dell puzzle magazine in the United States.
- 21984: the publisher Nikoli reprints it in Japan under the name sudoku.
- 32004: Wayne Gould's computer-generated grids reach The Times of London, and the global boom begins.
KenKen adds the arithmetic
Sudoku proved that a latin square plus a few clues could hypnotize millions. KenKen, invented in 2004 by the Japanese math teacher Tetsuya Miyamoto, added the missing verb: arithmetic. Instead of pre-printed clues, the grid is divided into outlined cages, each stamped with a target and an operation. Your job is to make the digits inside a cage add, subtract, multiply, or divide to that target, all while keeping the latin-square rule intact.
Miyamoto had built the puzzle for his own students, calling it Kashikoku Naru Puzzle, a puzzle that makes you smarter. The export name doubled a word for cleverness: KenKen, cleverness squared. It reached The Times of London in 2008 and The New York Times the year after, where the crossword editor Will Shortz called it the most addictive puzzle since sudoku. If the rules are new to you, our how to play guide is the fastest way in, and the ArKKade has grids at every size.
One family, one very good habit
Line the branches up and the through-line is obvious. Every member of this family asks you to reason, not guess, toward a single answer that was waiting all along. That may be why the habit sticks, and why researchers keep studying it.
Adults who did number and word puzzles more often scored as though their brains were up to a decade younger, on measures of short-term memory and problem-solving speed.
That study is a correlation, not a cure. It cannot prove the puzzles caused the sharper scores, and no grid prevents dementia. A separate Duke University trial did find crosswords beat some computer brain-games at slowing memory decline in at-risk adults. So enjoy the puzzle for the click of insight it hands you. Ready to add your own branch to the tree? Go play a grid, or keep a fresh one coming with Premium.
From a turtle shell in a Chinese river to Euler's Latin letters, from a Dell magazine to a Yokohama classroom, the history of number puzzles reads like one long argument that thinking is its own reward. The next chapter is the grid in front of you. Fill it in.