Scoring
With Japanese rules, you score the territory you have surrounded plus the prisoners you have captured: this is known as "territory scoring". With most other rules, you score the territory you have surrounded plus the points that you occupy. The SST rules implement area scoring by using sets with known numbers of stones. The result of the two methods is usually the same to within one point: the proof of this is left as an exercise for the reader. North American rules allow scoring by either method.Some reasons why the result may not be the same:
- One player may have passed while the other filled in her own territory.
- The two players may not have moved the same number of times because Black moved both first and last.
- There may be handicap stones.
Some rules (including North American) address point 2 by requiring White to play the last move of the game.
Some area-scoring rules (including North American) address point 3 by subtracting a point per handicap stone from Black's score.
Does it cost me a point to make a defensive move inside my
territory?
Making an unnecessary defensive move inside your own
territory is obviously a waste of a move. Near the end of the game, when
there are no moves worth making, this does not matter. But under Japanese
and IGS-PandaNet rules, an unnecessary defensive move also costs a point.
Under other rules, it does not.
Note that under rules other than Japanese, filling a dame gains a point. So under Japanese rules, you should never make unnecessary defensive moves inside your own territory. Under other rules, you should only do so after all the dame have been filled, and it may be helpful to do so as it may make clear what is alive.
Seki
The Japanese rules say that this position is seki, and no territory is counted in a seki. Other rules say that White counts the two points she has surrounded here. (The SST rules used to count the 2-1 points pro rata, making this position worth one-and-a-third net to White, but this has been changed.)
Figure 1. A seki.
Suicide
In a position like this, White may have a ko threat. She can play between the two marked stones, suiciding three stones and threatening to kill the black group. But under Japanese, Chinese, North American and IGS-PandaNet rules, suicide is forbidden.Here are some much less likely positions in which suicide is a good move, if allowed.
Figure 2. A suicidal ko threat.
Repeated position
All rules agree in preventing two-cycles of
repetition, in which one player moves, the other player moves, and the
position is restored to what it was before. All those listed here do so by
the familiar ko rule, which forbids a move that would restore the position
to what it was just before one's opponent's last move.
But there are other, much less likely, ways in which a position can be repeated. These necessarily have a cycle length greater than two moves. One is triple ko. Others are cho-sei, and the molasses ko. Here are some more examples, and discussion, by Robert Jasiek and by Harry Fearnley.
Under Japanese rules, if such a repetition occurs, the game is a non-event, and the players will normally play again.
Under Chinese rules, I am told by a Chinese Professional player that a
repeated position is treated as a drawn game, with the victory being
shared.
According to the Chinese rules as presented in
"The Go Player's Almanac" (Ishi Press 1991), such repetition is
"forbidden", and when it happens, "the referee may declare a draw
or a replay". This does not make sense to me. Unfortunately I cannot read
Chinese and check the original. The intention may be to forbid
repetition whose only intention is to prevent the game from ending (such
repetition can only work under area-scoring), while handling repetitions
such as triple-ko as a draw or as a void game.
Under North American, and New Zealand, rules, such repetition is explicitly forbidden. A player may not make a move which brings about a position that has ocurred earlier with the same player to move. There are three ways of forbidding such repetition: see below.
The SST rules attempt to restrict such repetition, and do so in a very complicated way, which involves a distinction between "fighting kos" and "disturbing kos", and many other concepts which I do not understand.
The IGS-PandaNet rules allow such repetition. However, there is a maximal move number for each game after which it must be adjourned. (This limit is currently 1000 moves.)
"Bent four in the corner"
In this position, if Black plays on
the vacant 2-1 point, the result will be a ko fight with White to find the
first ko threat. Normally Black will leave this until the game is almost
over and she has eliminated all White's ko threats. She will then be able
to kill the marked group.
But some ko threats cannot be eliminated.
The Japanese rules state that the marked white group is dead regardless of the position on the rest of the board, and may be removed at the end of the game.
Other rules say that the position is to be played out. Normally the result will be the same as under the Japanese rules. But if Black has more uneliminable ko threats than White, the result will be different.
There is a more complicated position which works the same way as "bent four".
If the marked white "bent four" group is allowed to live when White has an uneliminable ko threat, you might think that the marked white "moonlight life" group should be allowed to live if White has an infinite number of uneliminable ko threats (i.e. a double ko). However no ruleset allows the moonlight life group to live.
Figure 3. "Bent four in the corner".
Figure 4.
Moonlight life.
Komi
The SST rules state that komi is 8, but that Black wins
jigos.
With integral komi, jigo is possible. This may be thought undesirable (we want most games to give a positive result, we don't want pre-arranged jigos) or desirable (more scope to tournament organisers in assigning prizes, the correct result between two perfect players must be jigo).
Placement of handicap stones
Japanese and IGS-PandaNet rules
require the handicap stones to be placed on the star points.
North American rules give placement on the star points as the default, but permit tournament directors to specify free placement if they wish.
Other rulesets allow Black to place them where she chooses.
Compensation for handicap stones
Under the Chinese and North
American rules, Black gives White compensation for handicap stones, so
that the area which they occupy is not counted. Under Chinese rules, where
n stones are given, n/2 is added to White's score and
n/2 is subtracted from Black's score. Under North American rules,
where n stones are given, n-1 is added to White's score.
Under SST and New Zealand rules, no such compensation is given.
Under Japanese rules, this is not an issue, as area is not counted.
The effect of this is that an n-stone handicap is n points larger under SST, NZ and IGS-PandaNet rules than under J or C rules.