The Rules of Go, Explained Simply

Go is one of the oldest board games still played today, and it has a reputation for being complicated. It isn’t. The full rule set fits on an index card, far shorter than chess. What takes time is learning to see the board the way experienced players do, and that only comes from playing. This page covers the rules; how to play Go covers the thinking that goes with them, move by move. Prefer to learn by doing rather than reading? The interactive tutorial covers this same material on a real board, one move at a time.

The board and the stones

Go is played on a grid of lines. Stones sit on the points where lines cross, not inside the squares. That trips up a lot of beginners coming from chess or checkers. One player has black stones, the other has white. Black moves first.

The board comes in three common sizes: 9x9 for a quick game, 13x13 for a medium one, and 19x19 for the full traditional game. The rules are identical on all three. See which board size to start with if you’re not sure.

An empty 9x9 Go board with one black stone near the bottom-left and one white stone near the top-right, showing the grid of intersections where stones are placed.
Stones go on the intersections, not the squares.

The goal is simple to state: surround more empty territory than your opponent, counting both the empty points you’ve enclosed and the stones on the board (this site uses area scoring, and how scoring works explains it in full).

Turns alternate, one stone per turn, on any empty point you like. There’s no forced move and no piece that has to advance. A stone never moves once placed; it stays until it’s captured or the game ends. If you don’t see a useful move, you can pass instead of playing, and the game keeps going until something ends it, which the last section on this page covers.

Liberties: the air a stone needs

Every stone needs empty space to survive. The empty points directly next to a stone (up, down, left, right, but not diagonal) are called its liberties. A stone in the open middle of the board has four liberties. A stone against the edge has three. A stone in a corner has two.

A single black stone in the middle of a small board with its four surrounding empty points marked, showing the four liberties a stone in open space has.
The marked points are this stone's liberties.

Stones of the same color that touch each other (again, only up/down/left/right, not diagonally) form a group, and the group shares its liberties. Two separate stones out in the open have four liberties each, eight between them. Push them together into one connected group and they share six liberties total: fewer than eight, because the point between them is no longer empty, but the whole group now has to be surrounded on all six remaining sides before either stone is captured. That trade-off is why connected stones are harder to kill than scattered ones, and why so much of Go strategy is about deciding when to connect and when to stay spread out.

Capture: how stones come off the board

When a stone or group has no liberties left (every neighboring point is filled by the opponent), it’s captured and removed from the board immediately.

A white stone at the center with black stones on three of its four sides and only one liberty remaining, about to be captured.
One liberty left. One more black stone and it's captured.
The same position after black plays the fourth surrounding point: the white stone has been removed from the board, leaving an empty point surrounded by black.
Black fills the last liberty. The white stone is removed.

Capture applies to whole groups at once, not just single stones. A three-stone wall with one shared liberty left goes down in one move, same as a lone stone. The size of the group doesn’t protect it, only the number of liberties does.

You can’t place a stone on a point that would leave it with zero liberties. That’s called suicide, and it’s illegal, with one exception: if placing the stone captures an opponent group first, freeing up liberties, the move is legal. That exception is what makes the next rule necessary.

Ko: why you can’t just recapture forever

Sometimes a single stone captures a single stone right back, over and over, in a loop that never ends. Go blocks this with the ko rule: you can’t make a move that would recreate the exact board position that existed before your opponent’s last move. If you’re caught in a ko, you have to play somewhere else first, then you’re free to retake it.

A white stone boxed in by black on three sides with one liberty left at the marked point, the classic ko shape.
White has one liberty left, at the marked point.
Black has played the marked liberty, capturing the white stone and leaving an empty point where white is temporarily forbidden from recapturing.
Black captures. White can't retake at the marked point yet: that would repeat the position from before black's move.

Ko fights show up constantly in real games and can decide who wins, but you don’t need to master them to enjoy playing. Just remember the rule: no immediate repeats. Since you’re blocked from retaking right away, players often play a ko threat elsewhere first (a move so dangerous the opponent has to answer it) before coming back to fight for the ko. If the opponent ignores the threat and retakes the ko instead, you carry out the threat and come out ahead either way.

Eyes: the shape that can’t be captured

An eye is an empty point completely surrounded by one player’s stones. A group with two separate eyes can never be captured, because the opponent can’t fill both eyes at once without playing a suicide move on the second one, and suicide is illegal. This is the whole foundation of staying alive in Go.

A ring of black stones with one fully enclosed empty point in the middle, forming a true eye that white cannot play into.
A true eye: completely surrounded, safe from a single capture attempt.

Not every enclosed-looking point is safe, though. If the diagonal points around an eye aren’t controlled too, the opponent can sometimes break in. That weaker shape is called a false eye, and it can’t defend a group on its own.

A black shape around an empty point where black holds the straight sides but not enough of the diagonal corners, showing a false eye that looks safe but is not.
A false eye: the corners aren't secure, so this point alone won't keep the group alive.

A group needs two true eyes (or a bigger safe space that works out to the same thing) to be unconditionally alive. This is the idea behind most fighting in Go: attack a group by threatening its eye space, defend a group by securing it.

There’s one more case worth knowing: sometimes two groups from opposite colors are tangled together, neither able to make two eyes, but neither able to attack the other without dying first either. Both groups just sit there, alive, for the rest of the game. That standoff is called seki (mutual life), and it’s rare enough that you don’t need to hunt for it. Just don’t panic if you ever see a group survive without eyes.

When the game ends

Play continues until both players pass in a row. Passing means you skip your turn because you don’t see a move worth making. Two passes end the game, both sides agree which stones are dead (stones with no way to survive), and the empty territory plus remaining stones are counted. How scoring works walks through a full count on a real board.

You can also resign at any point if the position is hopeless. Plenty of games end this way once one side is clearly ahead, rather than playing out every last point.

Because Black moves first, Black gets a real advantage before White plays a single stone. To balance that out, White gets a fixed head start added to their score at the end, called komi. This site uses the standard modern value, 7.5 points. The half-point rules out a tie, so every game has a winner.

You now know every rule in the game. What separates a beginner from a strong player isn’t more rules. It’s reading the board better, which is what strategy for beginners starts building.

Try it on the board →