Area or territory – A brief history of go rules

Published 7 Aug 2020 by antti (last edited 14 Aug 2020)
tags: tldr, rules

(The historical information in this article is based on The History of Go Rules by Chen Zuyuan.)

For a game that has mostly remained unchanged through its history, go has a lot of different rulesets. You may have heard of Japanese rules and Chinese rules, but how about Korean rules, aga rules, New Zealand rules, Tromp–Taylor rules, or Ing rules? Why do all these different rulesets exist, and what is different between them?

Once upon a time in China, go was (most likely) a stone-capturing game. At some point, the Chinese started to count the living stones on the board rather than the captured stones to determine the winner; we might dub this ‘stone scoring’. Subsequently, the notion of ‘territory’ (i.e., an intersection in which the other player cannot have live stones) was born, and the game-end process sped up. Since both players play the same number of moves, you get the difference of living stones from the captured stones; so, instead of filling the board with stones and counting them, you essentially get the same result by comparing the players’ captured stones and surrounded territories. This is called ‘territory scoring’.

Later, in Ming dynasty China, around the 16th century, Chinese go players came up with another way of counting the score. This was at least partly motivated by the fact that territory scoring penalises players from playing inside their own territories when all neutral points was filled, even though there might be unfinished business inside (such as the bent four in the corner). The Chinese started comparing the sum of a player’s stones and territories against half the sum of the intersections on the board; if the sum is larger, then the player has logically won the game, and vice versa. This is nowadays called ‘area scoring’.

Chinese players traditionally started the game from this setting, with White playing first.

Around year 700, when the Chinese were still using territory scoring, go spread to Japan, and so the Japanese also came to play go with territory scoring. The Japanese then gradually altered the rules of the game, for example abolishing the practice of starting the game from the preset position shown above, called zuozi, and the group tax. The latter was a byproduct of stone scoring: originally, only living scores counted towards a player’s score, so in the end the board was filled with stones; but every group has to leave two empty intersections for its eyes. Therefore, when the ‘group tax’ applies, every group costs a player two points. This rule made the territory scoring process more convoluted, which is probably the main reason why it was abolished.

A 9×9 game has ended. Black’s area (stones + surrounded territory) sums up to 46 intersections and White’s to 35 intersections. By area scoring, Black wins by 11 points.
Here, the board is filled for stone scoring. Black has 44 stones to White’s 31 stones; Black wins by 13 points.
In territory scoring, we instead count the empty intersections surrounded by the players. Now we need to add one captured stone at a to Black’s score. Now Black has 24 points while White has 14 points, so Black wins by 10 points. If the group tax is taken into account, two points are reduced from Black’s score and four points from White’s for a black win of 12 points.

During the centuries that followed, the Japanese discovered interesting features of the territory scoring, which led to special rules such as ‘the bent four in the corner is dead’ and ‘three points without capturing’. By the 20th century, the Japanese were proud to say that they ‘had taken the old Chinese game of go and perfected it’.

Elegant? Under Japanese rules, if a game ends with this shape on the board, Black is awarded exactly three points for the a stones! This rule is called torazu sanmoku (‘three points without capturing’) in Japanese.

In the mid-20th century, the Korean Baduk Association was created. Korea also had a long history of go played with their own rules, called sunjang baduk, which involves a special preset position and scoring rules. However, as the Japanese players of the time outclassed that of Korea, the Korean Baduk Association promptly decided to implement their own, slightly modified version of the Japanese rules which they dubbed ‘Korean rules’.

Korean players traditionally started the game from this setting, with Black playing first.
In the scoring system of sunjang baduk, all superfluous stones are removed from inside the players’ territories, as long as no stones are left in atari. Then, the number of empty intersections is compared: Black has 28 points to White’s 16 points, so Black wins by 12 points.

A bit later, for similar reasons, China followed suit. The Chinese Weiqi Association was formed and traditional Chinese rules were adjusted; the stone-placement practice and the group tax were abolished. The Weiqi Association however chose to implement area scoring, perhaps to emphasise that go is, and always has been, a Chinese game.

Closer to the end of the 20th century, more variations of the go rules started popping up. By this point, a dichotomy between Japanese rules, based on territory scoring, and Chinese rules, based on area scoring, had been formed, and subsequent rulesets were mostly either combinations or modifications of the two. Possibly following the two countries’ (and Korea’s) lead, some of the new rulesets also carried country names, such as the New Zealand and aga (American Go Association) rules.

Now, in the present day, the names ‘Japanese rules’ and ‘Chinese rules’ still form a dichotomy for ‘territory or area scoring’, as both rulesets have a firm culture-historical basis. Some other rulesets, such as the Korean, aga, or New Zealand rules, on the other hand, mainly describe where they are used, with a much lighter go-political weight. But how appropriate are these ruleset names in the globalised world, and especially online?

Continued here.


This post has no comments so far. Please log in to comment!