Scoring

At the end of a game of Go, assuming neither player has resigned first, a numeric score is determined for each player. After adjustment for komi, the higher scoring player wins, or the game may be tied (jigo). The numerical difference in score does not matter; a win is a win.

While Go is basically played the same world-wide, there are several different rulesets in use. One of the ways in which rulesets differ is in the scoring. Here we use scoring to refer to the definition of score. The point is to distinguish it from the actual procedure used to find the score, referred to as counting. This is a useful, if not universally followed, distinction. (Note, though, that counting has other usages as well, including counting during the game to see who's leading, generally known as positional judgment.)

There are two main scoring systems: Area scoring and Territory scoring.^[1] Which system is used will not normally affect who wins the game, and the difference between the margin of victory for the two methods will be zero or at most one point in the vast majority of games.

Various counting methods can be used to find the score under each scoring system.

More detailed discussion of the main scoring systems is found at Territory and Area Scoring. The information below is meant as an overview.

Area Scoring

In area scoring, your score is:

• the number of empty points which only your stones surround

• plus the number of your stones on the board

Area scoring is used in certain rulesets, notably Chinese Rules, AGA Rules and Ing Rules.

To determine the score with area scoring, Chinese counting is generally used. An alternative method is Ing counting.

Example: Assume each player has had 100 turns with no passes, this means they have played 100 stones each. At the end of the game there are 70 white stones surrounding 45 territory points, and 60 black stones surrounding 35 territory points.

White's score is 70 + 45 = 115; black's score is 60 + 35 = 95; the margin of victory is 20 points to white.

Territory Scoring

In territory scoring, your score is:

• the number of empty points which only your stones surround

• minus the number of empty points surrounded in seki (depending on the rule set)

• minus the number of your stones that have been killed (whether captured during the game, or removed from the board as dead stones at the end of the game)^[2]

Territory scoring is used in Japanese rules and Korean rules, and in several on-line servers.

To determine the score with territory scoring Japanese Counting is used.

Example: Following on from the example above, we know that white has captured 40 stones (as black played 100 stones yet only 60 remain, 100 - 60 = 40). Similarly black has captured 100 - 70 = 30 stones.

White's score is 45 - 30 = 15; black's score is 35 - 40 = -5; note that the margin of victory is the same 20 points to white.

The Gun Eight provides a nice example of the difference between area scoring and territory scoring in a commonly occurring situation.

Other Scoring Systems

There are a number of other scoring methods of historical and theoretical interest. In particular, Stone Scoring, used during the Ming Dynasty in China, on into the 20th century, is a variant of area scoring. Prisoner Scoring is a similar scoring method, but based on dead rather than alive stones. Korean Sunjang Baduk uses a variant of territory scoring.

Equivalence scoring has been devised in such a way that Chinese or Japanese counting yields the same result and returns the area score. It is used in the AGA rules. Theoretically it is suited as a counting method for all rulesets that determine the standard Area Score.

There are some house rules in Go where losing by a landslide counts as more than one loss. Mahn Bang is such a scoring system, used in Korea.

Although not a scoring issue per se, when gambling on games, the payoff may be defined as a multiple of the number of points the winner won by; this is called mego in Japanese.

See scoring system for a Korean article.

[1] The terms territory scoring and area scoring were coined by James Davies to differentiate Japanese style scoring from Chinese style scoring. Davies used those terms in translating Ikeda Toshio's book, On the Rules of Go in 1991, but the book is a compilation of a series of articles written in the late 1960s.

[2] When one counts a territory score, one's score is reduced by the number of one's own captured prisoners. Usually this is done by filling in. An older edit of this page had one's score being increased by the number of one's opponent's stones one has captured. The game results would be the same, but in practice, the former method is used.

Patrick Traill: I recall reading a rule-set in which dame in a seki are scored in proportion to the number of adjacent stones of each colour. This means that points can be worth 0 ¼ ⅓ ½ ⅔ ¾ 1 to one player, the difference in score is a multiple of ⅙ and one player’s score is a multiple of a twelfth; the result is to reduce the probability of jigo at the cost of extra difficulty in counting. Can anyone shed light on this? I felt fairly sure this was Ing, but see nothing about it under Ing Counting.

Herman: AFAIK, this was in the 1986 version of the Ing rules, since superseded by the 1991 and 1996 versions, which do not use fractional scores any longer (but do seem to use proximity scoring)

Bass: In the early Ing rules, it was actually even more bizarre than that: the fractional seki points were awarded to neighbours of empty regions. This means you could also have denominators with prime factors larger than 4, like in this real world game here (example 2): https://homepages.cwi.nl/~aeb/go/games/games/Ing/Ing01.html