Quoridor Strategy

Quoridor hands you one decision and then asks for it again every turn until somebody crosses the board. You cannot pass, and you cannot do two things in a turn, so each turn resolves the same question: do I move my pawn, or do I place a wall? Everything a strong player knows is an answer to that question in a particular position. This page is the framework the rest of the strategy section hangs from — what the two options cost, why the answer changes as the game goes on, and which article to read next depending on what you are trying to fix.

The vocabulary worth borrowing is economic. Call the first option tempo — turns converted directly into progress — and the second terrain — walls converted into distance your opponent has to cover. Both are bought with the same thing, your turn. What separates them is what you are left holding afterwards.

The two purchases are not the same shape

A step is a sure thing: one turn, one square closer, no conditions. Nothing leaves your hands but the turn itself, which you had to spend on something anyway, and there is no reply that undoes it. What a step is not is permanent — a wall placed later can send a pawn back over ground it has already covered, which is the entire business of Traps and Cages — but that is a future expense, not a risk you are taking at the moment you move.

A wall is speculative. It costs the same turn, plus one unit from a supply of ten that never refills, and it does not buy squares directly. It buys a change to the map, which converts into squares only to the extent that your opponent's route actually runs through the ground you just made expensive. If they have a cheap way around, you spent a turn and a wall to accomplish nothing while they spent their turn and got a square — a wall gone and a square of ground lost, from a move that felt like pressure.

So the whole game is one trade at one price: squares of detour bought with a wall, against the single square a step delivers for certain. Wall Economy works out what that rate has to reach before a wall is worth placing at all, and it is a considerably higher bar than most players place walls against.

A stock and a flow

Underneath the trade sits an asymmetry in how the two resources arrive. Turns come one at a time and never run out. Walls are a stock of ten that only ever shrinks. You cannot save a turn — there is no passing, so the turn gets spent whether or not you have anything good to do with it — but you can save a wall indefinitely, which makes holding a genuine third option rather than an absence of a decision. Almost every position where a strong player does something a weaker player finds passive is a position where holding was the option being exercised.

Holding has a deadline, though, and the deadline is not the end of the game. A wall's purchasing power climbs while the board narrows and then collapses once its target is nearly home, so the skill is selling somewhere near the peak rather than at either extreme; Wall Economy traces that curve and names what pushes it up and down. Whatever is still in your hand when somebody crosses the board is worth exactly nothing, and there is no credit for economy.

Two properties of terrain

Tempo is private; terrain is public. Nobody else can use a square you have walked, but a wall stops being yours the moment you place it — it belongs to the board, and for the rest of the game it obstructs whoever needs to cross it, including you if the game turns and your own route shifts. Charging yourself for that is a step in the procedure at Wall or Advance.

Terrain also has one thing it can never do: no wall may leave a player with no route to their goal at all (The Complete Rules), so walls lengthen a journey and never end one. That is not a ceiling on what terrain is worth — walls placed in the right relationship to each other convert at rates that decide games outright, which is what Traps and Cages is for. It only means terrain is always financing: somebody still has to run the race on foot.

Which article, and in what order

The six articles are ordered so each uses what the previous one established. Read them straight through for the full argument, or jump in by symptom.

Tempo and the Race prices the first currency — how to count both shortest paths quickly, and how to read the difference between them. Start here if you lose races you were sure you were ahead in.

Wall Economy prices the second, end to end: the cost of the wall itself, the future placements each one permanently rules out, and why walls spent in planned pairs buy a far better rate than reactive singles. Start here if your walls look threatening and change nothing.

Wall or Advance is the decision procedure itself — a repeatable sequence of checks to run on your turn, instead of a feeling about whether the position calls for aggression. It is the most immediately practical article in the set, and the one to reread if your results swing.

Quoridor Openings covers the first several moves, where the exchange rate is at its worst and a lone wall on an open board usually buys nothing. Start here if you are comfortable at move ten and losing by move twenty.

Traps and Cages covers the midgame structures where a few walls placed in relation to one another force detours far larger than the sum of their parts. Start here if you keep discovering detours you never saw being built.

Quoridor Endgames covers what is left once the walls are gone and the position comes down to counting. Start here if you keep losing finishes by a single move.

What the framework does not do

Pricing a move is not the same as finding one. Everything in this section tells you how to evaluate a candidate; none of it generates candidates for you. An empty nine-by-nine board admits 128 legal wall placements — sixty-four intersections, two orientations each — and the one that returns five squares against a committed pawn is somewhere in that list whether or not you notice it. The arithmetic only starts once you have. That noticing is pattern recognition, and it comes from having seen the shape before, which is why the later articles are built around named structures rather than around more formulas.

One boundary worth stating: all of this assumes the standard two-player game, nine by nine, ten walls each. The four-player variant keeps the arithmetic and breaks the accounting — with five walls apiece and three opponents, terrain you buy against the leader is handed free to the two players you are also racing; Wall Economy prices what that does to the trade.

If any of this assumed too much, How to Play is the first-game walkthrough that comes before it, and the Glossary defines the terms these pages use in passing.