The history of Mines

Few puzzles have been played by as many people as the mine-clearing game. For a decade it came free on almost every office computer in the world, and a whole generation learned to right-click from it.

Early minefields

The idea of crossing a hidden minefield with only a count of nearby mines to go on turns up in several home-computer games of the early 1980s. In one popular 1983 game, you steered a figure from one side of the screen to the other, and a number told you how many mines lay in the squares next to it. Others set the minefield on a grid and let you mark squares you suspected. Each had a piece of the modern game; none put them together.

The grid we know

The version that settled the rules — a rectangle of covered cells, numbers counting all eight neighbours, a right-click to plant a flag, and a blank cell that opens its whole neighbourhood — was written by Robert Donner and Curt Johnson at Microsoft. It shipped in a pack of games for Windows in 1990, and from 1992 it was included with Windows itself.

That is where most people met it. It sat on hundreds of millions of computers, at home and at work, for the next twenty years. It's often said it was there to teach people how to use a mouse: precise left-clicks, and right-clicks, which were still new to many. Whether or not that was the plan, it worked.

The classic setup came in three sizes: a 9×9 beginner board with 10 mines, a 16×16 intermediate board with 40, and an expert board 30 wide and 16 high with 99. Players still race on those boards today, and the fastest clear the expert board in well under a minute.

A not-so-simple puzzle

For a game that looks this simple, it hides some deep mathematics. In 2000 the mathematician Richard Kaye showed that deciding whether a mines position is even possible — whether some arrangement of mines fits every number on the board — is NP-complete. That puts it in the same family as some of the most famous hard problems in computer science: no one knows a fast method that works on every board, and finding one would be worth a million-dollar prize.

You'll never meet that difficulty in an ordinary game, because real boards are small. But it explains why the game can be genuinely hard to reason about, even though every rule fits in one sentence.

The trouble with guessing

The classic game placed its mines at random, which means many boards can't be finished by logic. Sooner or later you reach two cells with nothing to tell them apart, and you have to pick one and hope. For a game that rewards careful thinking, ending on a coin toss always felt wrong.

Modern versions fixed this with no-guess boards: before a board is offered, a program checks that it can be solved from the opening without guessing, and throws it away if not.

Other ways to play

• Hexagonal boards, where each cell has six neighbours instead of eight.
• Three-dimensional boards, where a cell can have up to 26 neighbours.
• Wrap-around boards, where the edges join up, so there are no corners to lean on.
• Mines-only puzzles printed on paper: a grid of numbers with no clicking, where the solver marks every mine at once, much like a Sudoku.

The square-grid original is still the favourite, and that's what you'll find here — on boards shaped to fit a phone held upright.