Martian problem

[Bill Spight]

Hm. This would mean the big B chunk is uncapturable in actual play, resumption or not, which seem too big of a defect to be true (not to mention the paradox of dead stones that are uncapturable in actual play, if B would be smaller). Literal translations aside, are you sure Japanese pros would agree with this?

My guess is that Japanese pros don't care.

But if you go back to pre-20th century rules questions, when rulings were given by the best players around, Moonshine Life, of which this is an example, although an unusual one, has sometimes been considered alive, sometimes dead.

Most pros never run into situations where the niceties of the rules matter, so the payoff for studying those niceties is miniscule.

The idea of relinquishing the right to play goes back at least to Yasunaga's "Draft Constitution" (1932), which said that play ends with relinquishing the right to play three times in succession. I.e., three successive passes.

[jann]

I'm aware of debates over this in the past, but was under the impression that it was more or less resolved in a reasonable way (passes are moves that lift ko ban). I would also take the official method of hypothetical play (passing for a ko) as an evidence that this is the intended meaning (and W did pass for the ko in the game already). Maybe the Korean rules / practice could also offer some hints here.

Besides, if resumption would only be possible with the earlier ko ban still in effect (despite two intervening passes), why would hypothetical play start without it?

[Gérard TAILLE]

When I first read the japonese rule I understood that no points of territory are counted in case of seki.

With a second reading of the rule I now have some doubt because my "old" definition of seki (in 1970s) seems to have change with the new defintion (1989?)

The point is the following

$$B
$$ -----------
$$ | O d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------

Click Here To Show Diagram Code

[go]$$B
$$ -----------
$$ | O d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------[/go]

All stones are alive. Point "a" is a dame and as a consequence the black group and the white group marqued with a circle are in sekis. I conclude point "b" and "c" are not territory but because the five remaining white stones are not in seki the two points "d" are territories for white. Then white wins by two points. Is that true?
If yes isn'it a difference between the new rule (1989) and the old one?

[Bill Spight]

When I first read the japonese rule I understood that no points of territory are counted in case of seki.

With a second reading of the rule I now have some doubt because my "old" definition of seki (in 1970s) seems to have change with the new defintion (1989?)

The point is the following

$$B
$$ -----------
$$ | O d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------

Click Here To Show Diagram Code

[go]$$B
$$ -----------
$$ | O d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------[/go]

All stones are alive. Point "a" is a dame and as a consequence the black group and the white group marqued with a circle are in sekis. I conclude point "b" and "c" are not territory but because the five remaining white stones are not in seki the two points "d" are territories for white. Then white wins by two points. Is that true?
If yes isn'it a difference between the new rule (1989) and the old one?

It is plain by observing Japanese practice that, although the new in seki rule produces some strange results, all of the White stones on this board are considered to "have" dame.

When the '89 rules came out, I expected that the pros would start filling in all the dame before ending play, to make sure that all of their territory was counted. That didn't happen, as the rules were interpreted to allow the informal filling of dame. Years later some problems arose because of that ambiguity.

The 1949 rules was criticized for having a number of special rulings. AFAICT, the 1989 rules were intended to eliminate special rulings and rationalize the 1949 rules as much as possible, while affecting professional play as little as possible. It would not do to tell the current title holders that they weren't playing by the rules.

[Bill Spight]

I'm aware of debates over this in the past, but was under the impression that it was more or less resolved in a reasonable way (passes are moves that lift ko ban).

I am not aware of any debates over this. Yasunaga's proposed rules did not consider a pass to be a move, but had a superko rule that did not allow repetition of the whole board, except in case of a pass. That's ambiguous, but the next rule only makes sense if the pass lifts ko bans. The next rule says that 3 consecutive passes end play. The 3d pass is unnecessary unless the 1st pass lifted the ko ban.

The Ing rules, Tromp-Taylor rules, AGA rules, and, I believe, New Zealand rules consider a pass to be a play. Of these rules, only under Ing rules does a pass lift a ko ban (for fighting kos). That is one reason that the Ing rules require 4 passes to end play.

In the 1990s I proposed a rule under which a pass lifts ko bans. To prevent infinite repetition under such a rule, I had repetition of a whole board position by the same player end play. Normally play ends by 3 passes, just as Yasunaga proposed. (I was unaware of Yasunaga's rules until later.)

I would also take the official method of hypothetical play (passing for a ko) as an evidence that this is the intended meaning (and W did pass for the ko in the game already).

Under Japanese rules passing for a ko is a property of the hypothetical play to determine life and death. You can't pass for a ko in actual play. Hypothetical play eliminates the need for the special ruling of the 1949 rules that required kos to be filled before ending play.

Maybe the Korean rules / practice could also offer some hints here.

Jaeup is the expert on Korean rules.

Besides, if resumption would only be possible with the earlier ko ban still in effect (despite two intervening passes), why would hypothetical play start without it?

Hypothetical play is not actual play. It only exists as an alternative to an encore of actual play. One reason for hypothetical play is to force the player who took a ko to win it, by removing the ko ban on her opponent. That's also why the player who originally took the ko cannot play a ko threat and take it back. Hence, the pass for ko rule in hypothetical play. Japanese hypothetical play is a strange beast.

[Gérard TAILLE]

It is plain by observing Japanese practice that, although the new in seki rule produces some strange results, all of the White stones on this board are considered to "have" dame.

Bill, can you explain why on this board the white stones have dame?
Maybe it is again a translation issue? The text I am referring to is the following:

Article 8. Territory
Empty points surrounded by the live stones of just one player are called "eye points." Other empty points are called "dame." Stones which are alive but possess dame are said to be in "seki." Eye points surrounded by stones that are alive but not in seki are called "territory," each eye point counting as one point of territory.

$$B
$$ -----------
$$ | Q d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------

Click Here To Show Diagram Code

[go]$$B
$$ -----------
$$ | Q d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------[/go]

All white stones are alive.
point "a" is clearly a "dame point".
point "b": it is an empty points surrounded by the live stones of just one player; that means it is an "eye point".
points "d" are also "eye points" for the same reason.

The triangle white stone has two eyes points ("b" and "d") and no dame => triangle white stone appears alive without being in seki.
And it is the same for the four non marked white stones.
Where am I wrong?

[jann]

I would also take the official method of hypothetical play (passing for a ko) as an evidence that this is the intended meaning (and W did pass for the ko in the game already).

Under Japanese rules passing for a ko is a property of the hypothetical play to determine life and death. You can't pass for a ko in actual play.

Sure, but this is a clear indication that in Japanese rules a pass CAN (possibly) allow subsequent ko recapture without a board play changing the position.

Gérard: look for deeper meaning of "possessing" dame (note this is not defined, probably intentionally).

[Gérard TAILLE]

I would also take the official method of hypothetical play (passing for a ko) as an evidence that this is the intended meaning (and W did pass for the ko in the game already).

Under Japanese rules passing for a ko is a property of the hypothetical play to determine life and death. You can't pass for a ko in actual play.

Sure, but this is a clear indication that in Japanese rules a pass CAN (possibly) allow subsequent ko recapture without a board play changing the position.

Gérard: look for deeper meaning of "possessing" dame (note this is not defined, probably intentionally).

Could somebody, knowing japonese language, tell us what was really meant by the text translated by
Stones which are alive but possess dame are said to be in "seki."

It seems almost unbelievable to see a very simple and precise definition of a group of stones, of a liberty, of a dame, of an eye point and here a definition of a seki which is very difficult to grasp though it is of great importance for understanding what a territory is.

[Bill Spight]

I would also take the official method of hypothetical play (passing for a ko) as an evidence that this is the intended meaning (and W did pass for the ko in the game already).

Under Japanese rules passing for a ko is a property of the hypothetical play to determine life and death. You can't pass for a ko in actual play.

Sure, but this is a clear indication that in Japanese rules a pass CAN (possibly) allow subsequent ko recapture without a board play changing the position.

Can possibly = could. You are right. The Japanese could have written the rules so that you could pass as a ko threat for a particular ko in actual play, but they didn't.

[Cassandra]

Could somebody, knowing japonese language, tell us what was really meant by the text translated by
Stones which are alive but possess dame are said to be in "seki."

As far as I understand it, the construction is a bit more recursive that you probably are aware of.

Alive STONE(s) that are neighboured to DAME, are in SEKI.

O n l y   points that are surrounded by alive STONE(s) that are   n o t   in SEKI are TERRITORY.

This implies that ALL points that are neighboured to (also surrounded by, even eye-points) STONE(s) that are in SEKI are DAME.

This implies that you have to occupy   a l l   DAME neighboured to one of your (two-eyed) alive groups to become the surrounded points TERRITORY.

+ + + + + + + + + + +

All White stones in your example (as well as the Black ones) are in SEKI.

[Bill Spight]

OK, let's use these definitions to make a single definition of territory.

Article 8. Territory
Empty points surrounded by the live stones of just one player are called "eye points." Other empty points are called "dame." Stones which are alive but possess dame are said to be in "seki." Eye points surrounded by stones that are alive but not in seki are called "territory," each eye point counting as one point of territory.[/quote]

Here is the Japanese text, and my translation.

一方のみの活き石で囲んだ空点を「目」といい、

An empty point surrounded by living stones of one side is called a moku,

目以外の空点を「駄目」という。

An empty point other than a moku is called a dame, pronounced DAH-MEH.

駄目を有する活き石を「セキ石」といい、

Living stones which possess a dame are called seki stones,

セキ石以外の活き石の目を「地」という。

Moku of living stones other than seki stones are called territory.

地の一点を「一目」という。

One (empty) point of territory is called one moku.

----

Territory by definition means moku of living stones other than seki stones.

Substituting for moku:

Territory. Empty points surrounded by living stones of one side other than seki stones.

Substituting for seki stones:

Territory. Empty points surrounded by living stones of one side which do not possess a dame.

Substituting for dame:

Territory. Empty points surrounded by living stones of one side which do not possess an empty point other than moku.

Substituting for moku again, and simplifying.

Territory. Empty points surrounded by living stones of one side which possess only such points.

Note that this definition does not refer to dame.

[Gérard TAILLE]

Could somebody, knowing japonese language, tell us what was really meant by the text translated by
Stones which are alive but possess dame are said to be in "seki."

As far as I understand it, the construction is a bit more recursive that you probably are aware of.

Alive STONE(s) that are neighboured to DAME, are in SEKI.

O n l y   points that are surrounded by alive STONE(s) that are   n o t   in SEKI are TERRITORY.

OK with your two points above

This implies that ALL points that are neighboured to (also surrounded by, even eye-points) STONE(s) that are in SEKI are DAME.

But I do not understand this implication.
Dame is defined by the two following sentences:
Empty points surrounded by the live stones of just one player are called "eye points." Other empty points are called "dame."
You see that for defining a dame you need only to know if stones are alive or not. The status of seki for these living stones is irrelevant for defining an eye point or a dame.
As a consequence I do not see a recursivity here because the status of an empty point (eye point or dame) is determined immediately and unambigously with only one stage (with recursivity we would have several stages for determining the satus of an empy point).

$$B
$$ -----------
$$ | Q d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------

Click Here To Show Diagram Code

[go]$$B
$$ -----------
$$ | Q d d O X -
$$ | b O O O X -
$$ | W X X X X -
$$ | W W X X X -
$$ | W W a X c -
$$ -----------[/go]

BTW, if we say that white wins by two points in the above position, then we reach the same value of an area scoring!
Wouldn't it be an initeresting (unexpected?) result?

[lightvector]

Dame is defined by the two following sentences:
Empty points surrounded by the live stones of just one player are called "eye points." Other empty points are called "dame."
You see that for defining a dame you need only to know if stones are alive or not. The status of seki for these living stones is irrelevant for defining an eye point or a dame.
As a consequence ...

You are missing the point. The Japanese 1989 rules are not meant to be read and interpreted literally. They, along with the commentary examples are meant to be read holistically and to give an idea of what their *intention* is, and then to follow the spirit of that intention, even if the precise wording of the rules is inconsistent, ambiguous, or contradictory with the intention.

I earlier posted about this here:
viewtopic.php?f=45&t=17721&p=259013#p259013

[Cassandra]

This implies that ALL points that are neighboured to (also surrounded by, even eye-points) STONE(s) that are in SEKI are DAME.

But I do not understand this implication.
Dame is defined by the two following sentences:
Empty points surrounded by the live stones of just one player are called "eye points." Other empty points are called "dame."
You see that for defining a dame you need only to know if stones are alive or not. The status of seki for these living stones is irrelevant for defining an eye point or a dame.
As a consequence I do not see a recursivity here because the status of an empty point (eye point or dame) is determined immediately and unambigously with only one stage (with recursivity we would have several stages for determining the satus of an empy point).

Dear Gérard,

"eye point" is nothing that serves the definition of TERRITORY.

When determining TERRITORY, empty points of the board can have one of ONLY TWO status:
-- DAME
-- TERRITORY

STONE(s) that are in SEKI (= SEKI STONE(s)) cannot surround TERRITORY, so every empty point in their neighbourhood must be DAME.

Please note that the primary definition of DAME (= "ANYTHING BUT surrounded by alive STONE(s) of only one colour") is derived from "eye point". But the definition of "eye point" seems to be declarative only (probably for better understanding? probably a reference to the "real" world?); there is NO, ABSOLUTELY NONE, further reference on this term in the rules (with the exception of the definition of the temporary opposite). Therefore, "eye points" have nothing to do with TERRITORY.

STONE(s) that are in SEKI might be neighboured to an "eye point", but "eye point" does not exist when you determine TERRITORY. If an empty point is DAME, it cannot be TERRITORY, and vice versa.

+ + + + + + + + + + + + + + + + + + +

Yes, indeed, we have two pairs of terms:

-- eye point / DAME, and
-- DAME / TERRITORY, but

both pairs have to be seen in different context.
The first pair serves the easy detection of a first amount of "DAME", with which can be started in the second pair, for determining "SEKI STONE(s)", and probably a second amount of "DAME".

The term "DAME" exists in both pairs, but this does NOT imply that "eye point" = "TERRITORY".

[HermanHiddema]

Given how the rules have been applied, the verb "to possess" in the phrase "Stones which are alive but possess dame are said to be in seki." can be concluded to mean:

A stone possesses dame if any stone in the same group (not chain) is adjacent to a dame.