Go strategy: from your first game to advanced play

The three questions behind every Go move

Strong players do not calculate more variations than you — not most of the time. They ask better questions. Three of them cover nearly every position.

1. Where is the biggest empty area? Play there before you play anywhere small.
2. Which groups are weak? Yours and theirs. Weak groups drive the whole middlegame.
3. Is this move sente? Does my opponent have to answer? If yes, I keep the initiative.

Level 1 — Basic strategy: what to do in your first fifty games

Open in the corners

Play your first moves on the 4-4 or 3-4 points of empty corners, then extend along the sides. Never start in the centre. This is not taste — it is the arithmetic in the diagram above.

Play away from strength, extend from weakness

Stones next to your own strong wall are wasted; stones next to a weak group are urgent. If you remember one slogan, make it: "urgent points before big points."

Learn two capture shapes: the ladder and the net

The ladder (shicho) chases a stone diagonally across the board — it works only if nothing of the opponent's colour is on the escape path, so always read to the edge before playing it. The net (geta) traps a stone loosely and never depends on what is far away, which is why stronger players prefer it.

Do not answer every move

Beginners answer locally out of politeness. If your opponent plays a move worth four points and there is a twenty-point point elsewhere, take the twenty.

Count liberties before you fight

Most beginner fights are lost by one liberty. Before starting a capturing race, count: mine, then theirs, then whose move it is.

Opening patterns worth knowing

The opening (fuseki) is not memorisation, but a handful of named patterns will give your first thirty moves a shape.

The corner points

• 4-4 (star point) — fast and influence-oriented. It does not secure the corner; it claims a direction. Best for beginners and the mainstay of modern play.
• 3-4 (komoku) — territory-oriented and slower. Invites a wider variety of joseki.
• 3-3 (san-san) — takes the corner outright and gives away the outside. Out of fashion for decades; rehabilitated by engines.

Two classic frameworks

The Chinese opening combines two corner stones with a low extension on the side, making an area that is hard to invade cleanly and hard to ignore. The Kobayashi opening is its territory-minded cousin. Both are popular precisely because they aim at a shape rather than at exact sequences — they survive contact with an opponent who does not know the theory.

What the engines changed

Since 2016, three ideas that were considered bad have become standard: the immediate 3-3 invasion under a 4-4 stone, early attachments against approach moves, and shoulder hits on the fourth and fifth lines. The common thread is that engines value flexibility and speed above the classical preference for a settled shape.

Level 2 — Intermediate: shape, sente and the whole board

At this level you stop losing games to tactics and start losing them to direction. The skills that matter change completely.

Sente and gote

A move is sente if your opponent must answer it, gote if they can ignore it. Keeping sente is how you play two big points in a row. A common intermediate leap forward is simply realising that a 15-point gote move is often worse than a 10-point sente one.

Good shape and bad shape

• Empty triangle — three stones in an L. Almost always bad: it has fewer liberties than the same three stones arranged well.
• Bamboo joint — two pairs of stones side by side. Uncuttable, flexible, good.
• Table shape / tiger's mouth — efficient connections that make eye shape as a by-product.

Shape is not aesthetics: good shape means more liberties and more eye potential per stone.

Joseki, and how to use them

Joseki are established corner sequences that end in a fair local result. Learn a handful, but learn why each move is played. The classic intermediate mistake is memorising a joseki and playing it in a position where the resulting wall faces the wrong way — a locally even result that loses the game.

Attack to profit, not to kill

Chasing a weak group is how you build territory somewhere else. If you try to kill and fail, the group lives and your attacking stones are worthless. Chase it in the direction that builds your own framework, and take the profit.

Endgame: learn to count

Most intermediate games are decided in the endgame by five to fifteen points — which is exactly the range that knowing your endgame values covers. Learn the value of a monkey jump, a hane at the first line, and the standard connections, and play the biggest first.

Level 3 — Advanced: direction, thickness, aji and ko

Advanced Go is mostly about resources that are not yet on the board: influence you have not cashed in, weaknesses you have not exploited, and ko threats you have not spent.

Direction of play

The single hardest thing in Go. Two locally correct moves can differ by twenty points because one works with your existing stones and the other faces an area your opponent already controls. Before choosing a joseki, decide which way you want the resulting wall to face — then pick the joseki that delivers it.

Thickness is not territory

A strong wall (thickness) is worth a great deal, and the classic advanced mistake is turning it into a small territory. Use thickness to attack: play a splitting move in the opponent's area knowing your weak stones have a safe wall to run towards.

Aji: leftover potential

Aji is the latent possibility inside dead or captured stones — a cut that might work later, a stone that might revive. Strong players avoid moves that "use up" aji, keeping options open until they know which one is worth the most. The correlate: when you are ahead, remove your opponent's aji even at a small cost.

Winning ko fights

A ko is decided before it starts, by the threat count. Practical rules:

• Count threats first. If you have fewer big threats, do not start the ko.
• Play your threats in ascending order, keeping the biggest in reserve.
• Creating threats is a move. A move that both gains something and makes a ko threat is worth much more than it looks.
• Ko for life is bigger than ko for territory. Value the ko by what the loser loses.

Positional judgement: count the board

By the middlegame you should be able to stop and estimate the score within about five points. It determines everything: ahead, simplify and take safe territory; behind, start a fight or invade while there is still space. Most advanced losses come from players who were ahead and kept complicating, or behind and kept playing safe moves.

Studying in the engine era

Engines will tell you a move loses four points, but not why. Use them backwards: play your own game, guess where you went wrong, then check. The errors an engine finds that you did not suspect are the ones worth studying.

Nine habits that cost the most points

1. Answering small moves automatically.
2. Making territory next to your own thickness.
3. Touching a weak stone you intend to attack — contact moves make it stronger.
4. Starting a ladder without reading it to the edge.
5. Saving three stones that were worth eight points, in gote, while a thirty-point point sits open.
6. Playing a joseki for its own sake, ignoring the rest of the board.
7. Invading when reducing would have been enough — invade when behind, reduce when ahead.
8. Refusing to count, then discovering in the endgame that you needed a fight twenty moves ago.
9. Adding stones to a group that is already dead.

Common questions