Quoridor Strategy: Count the Race, Spend Walls Like Money
Quoridor gives each player exactly one number that matters, and most people never count it. The game, designed by Mirko Marchesi and published by Gigamic in 1997, hands you a pawn on a nine by nine board, ten walls, and a single decision every turn: step toward the far side, or place a wall to lengthen your opponent's journey. Everything else is presentation. Underneath, Quoridor is a race between two numbers, your shortest path to the goal and theirs, and every strong move either shrinks your number or grows the other one by more than it costs you.
The count is the position
Look at any Quoridor board and ask two questions. How many steps do I need, walking my best route? How many do they need? A position where you need seven and they need nine is winning for you by two tempi, full stop, and no amount of dramatic wall architecture changes that unless it changes the count. Beginners evaluate walls by how annoying they look. The count is the honest judge: a wall that reroutes your opponent from nine steps to ten while you spend a turn placing it has gained you nothing at all, because the race moved one turn closer to the finish for neither of you. A wall that turns their nine into thirteen has bought three clear tempi with one turn. That's the whole business model.
This is why the classic advice in our Quoridor and every other version is: when in doubt, walk. A step forward always improves your count by one. A wall only pays when it damages their count by two or more, and walls that merely inconvenience are worse than nothing, because you hold a finite supply.
Ten walls is a budget, not an arsenal
Each side gets ten walls for the entire game, and the deepest beginner leak is spending them like ammunition instead of money. Three budgeting rules cover most of it.
- Never wall a player who can still go around cheaply. Early on, the board is open and every blockade has a two-step detour. Early walls buy one tempo at full price. The same wall in a crowded midgame, dropped where two detours meet, can buy four.
- Walls spent on defense count double. A wall placed beside your own route, pre-empting the cage your opponent was about to build, protects every future step you take through that corridor. Watch what the enemy walls are sketching: three walls arranged near your path usually mean the fourth would complete a long detour, and blunting it now is cheaper than rerouting later.
- Holding walls late is a threat in itself. An opponent with two walls left walks the endgame nervously; one with zero is just a pawn, and you can route yourself with total confidence. Sometimes the best use of your last walls is never placing them.
One rule of the game polices all of this: a wall may never seal a player off entirely. Every placement must leave at least one path to the goal for both pawns, and our version checks it before the wall lands. You cannot imprison anyone. You can only make the journey long, which is why the count, not the cage, is the object of the game.
The face-off, and who blinks
Sooner or later the pawns meet in the middle, usually in some corridor the walls have carved. Quoridor resolves the standoff with a hop: a pawn facing an adjacent enemy may jump straight over it, or diagonally past it when a wall blocks the straight hop. The tactical point is that the hop is a free extra square for whoever engineers it, and engineered is the right word. Walking into a face-off one step ahead of your opponent often means they hop over you and land a net square better off. Count the hop into your race arithmetic before you close the gap, not after.
Endgames are about lanes
By the time most walls are down, the board has collapsed into a small number of lanes, and the finish is usually decided by a question you can ask two moves early: how many openings does my final stretch have, and can any remaining wall close the one I need? If your route to the goal row runs through a single gap and your opponent holds a wall, assume the gap will slam shut at the worst moment and keep a second lane alive, even at the cost of one step. A seven-step route with two exits beats a six-step route with one exit against any opponent still holding walls. Against an empty-handed opponent, take the six and sprint.
Quoridor won a Mensa Select award the year it appeared, and the reason is exactly this texture: complete information, no luck, and a single resource whose spending schedule is the entire skill. If the count-and-budget mindset appeals, it transfers whole to chess endgames and to the wall-reading in Tron, and the pathfinding instinct sharpens nicely in our maze. The rest of the board room lives with the other board games, and our computer opponent on Hard counts the race perfectly. Beat it by making it spend walls that buy less than they cost. That's the entire game, in one sentence.