This guide explains the rules and basic tactics of go in a simplified fashion.
The rules and scoring process used in this guide differ in some details to those used in go tournaments, but the core game is the same. We use this simplified interpretation because it makes it easier to understand when a game should end and how it is scored.
There are mouse-hover tooltips scattered throughout the guide. These are meant for more experienced readers who are considering to use this guide as reference for teaching new players. The tooltips explain the differences and reasonings to the simplifications made in the guide.
In the future, this guide will hopefully be improved by including interactive exercises.
Go is a two-player perfect-information board game not unlike chess. One of the players holds black game pieces, or ‘stones’, and the other player holds white.
The game board is a grid of variable size. The 19×19 grid is most common in tournament play, but 13×13 and 9×9 are often used as well. The nature of the game is the same on any board size; in this guide, we use a 6×6 board because it leads to shorter games.
As you can also see in the above photo, the game pieces (hence, ‘stones’) are placed on the intersections on the grid. A 19×19 board may consist of 18×18 squares, but don’t let them fool you!
The number of stones a player can play during a game is not limited. In practice, a game set should have in total as many stones as there are intersections on the board. (19×19 game sets usually have 181 black and 180 white stones.)
The players take turns placing one stone on an empty intersection on the board. Black starts.
The winner of a game is determined by comparing the players’ scores after the game. A player’s score is the number of stones they have on the board. After a game ends, the players’ scores are counted, and the player with the higher score wins.
In other words: when a game ends, the player with more stones on the board wins.
The game ends when neither player wants to continue playing. Usually (especially in tournaments) this is indicated by passing: on their turn, a player may either place a stone on an empty intersection or pass. The game ends when two consecutive passes occur. In non-tournament settings, the players may skip the passing process and verbally agree to count the score.
Since the players take turns placing a stone, you may wonder if this means every game will finally end in a tie. This is however not the case because of…
This is the main rule that defines what go is all about.
When a stone gets surrounded from all four directions by the opponent’s stones, it is captured and removed from the board.
A single stone gets captured when it no longer has empty intersections next to it. Therefore, a stone on one of the sides of the board can be captured by three opposing stones, and a stone in a corner gets captured by two opposing stones. This means that stones in the corners and on the sides are more easily captured, i.e., weaker than stones closer to the centre.
The capturing examples until now were one-sided: we did not consider the defender’s point of view.
If stones of the same colour are neighbours, they share their empty intersections, or liberties. This means that neighbouring stones either live or get captured together. For example, two stones in the centre of the board require six opposing stones to capture.
This means that when your stone is one move away from getting captured – we call this state atari – you can save it by placing a stone next to it.
Congratulations, you actually now know most of the basic rules of go! Most of the content of this guide from now on is either emergent from the rules already explained, or otherwise commonsensical.
Although the rules explained up to now may seem simple, they join to create an extraordinarily deep and interesting game.
Below is a simple stone-capturing exercise. Assuming that White cannot play any stones, how many stones does it take for Black to capture all the white stones?
This exercise is a lot trickier than the previous one, because we need to predict both players’ moves. Still, give it a try before seeing the answer!
White has just played 1, and the two black stones are in atari – they only have one shared liberty remaining. Can Black save her stones?
The above exercise is one of a myriad examples showcasing that simple rules can lead to complex emergent patterns. Learning these patterns is one of the most fun parts of learning go!
Consider the position below. What happens if Black plays on the intersection marked ‘a’?
After Black plays her stone, we notice that it has no empty intersections next to it – therefore, the stone will immediately get captured and removed from the board.
It is possible for Black to play at ‘a’ in the first diagram, but it is not a good move. A player’s score is the number of stones they have on the board – a stone placed at ‘a’ does not stay on the board, and therefore does not contribute to Black’s score.
While Black’s stone placed at ‘a’ does not remain on the board, White’s does, because it connects to the white stones surrounding it and shares their liberties.
However, strategically speaking White would do better to play somewhere else than 1 above. Think about it – Black could never have had a stone at 1 anyway! Therefore, White has time to claim other intersections first, and then, when the board is otherwise filled, then he can return to play 1 without fear of Black getting a stone there first.
Consider the situation below. This is where the capturing rule gets truly interesting.
This situation is similar to above, but the white stones no longer have any external liberties. This changes the situation considerably.
For example, if White were to play at 1 like above, then all the eight white stones would no longer have empty neighbouring intersections – this means that they all get captured and removed from the board.
What if Black were to play instead? Then, after the black move, neither the white stones nor the black stone are connected to empty intersections.
In this case, we assume that the player who just played a move has privilege. Therefore, her opponent’s stones get captured first – after the opponent’s stones are removed, if any black stones still have no access to liberties, then they get captured.
In this way, ‘surrounding’ is a key determining feature in go. So determining, in fact, that the game is even named after it – the Chinese name, weiqi, or 圍棋, literally means ‘surrounding game’. ‘Go’ comes from the Japanese name of the game, igo or 囲碁, which is almost a direct Japanese translation of the Chinese name. In Japanese, the name means ‘surrounding game played with stones’.
This is the last rule we need to know, and luckily it is also fairly logical to understand.
Consider the position below. It looks like the white stone at ‘a’ is just about to get captured.
However, when Black captures the white stone, it creates a new position where it looks like Black is about to get captured.
After the white stone is captured, black 1 has exactly one liberty remaining – it, too, is in atari. If White could now capture Black again, that would lead to the original position, as below.
Because the indefinite recapturing would prevent the game from ever finishing, go has the so-called kō rule (kō is a Japanese word roughly translating to ‘eternity’). The kō rule states that ‘a move that recreates a previous whole-board state is forbidden’.
In practice, this means White cannot play the recapture of 2 above, but any other move is fine.
With this, you know everything necessary in order to start playing! If (and when) you still feel confused by the rules, the best way to get into the game is to find another beginner and try it out! If you don’t have a board and stones at hand, for example the Online Go Server is a good place to find an opponent or set up a game with your friend.
This guide will later be appended with more examples and exercises relating to the end of the game.
All times are in Helsinki time (eet with summer time).
Check the current review credits balance
Public lectures on Twitch every 2nd and 4th Saturday of the month at 1 pm.
Jeff and Mikko stream on Twitch on Fridays at 6 pm.