Quite strange result in area counting rules

Gérard TAILLE · #1

Click Here To Show Diagram Code: [go]$$W White to play $$ -------------------- $$| . O . X O . O . . | $$| X . X X O O O O O | $$| X X X O O M M M M | $$| O O O O M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$ -------------------- $$[/go]

Under japonese rules no doubt the black stones are dead.
But what about area scoring rules?
Can you build in the marked area, an environment such that black can save her stomes in the upper left corner? I believe it is really possible! An interesting challenge for all those interesting by curious results.

Gérard TAILLE · #2

Gérard TAILLE wrote:

Click Here To Show Diagram Code: [go]$$W White to play $$ -------------------- $$| . O . X O . O . . | $$| X . X X O O O O O | $$| X X X O O M M M M | $$| O O O O M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$ -------------------- $$[/go]

Under japonese rules no doubt the black stones are dead.
But what about area scoring rules?
Can you build in the marked area, an environment such that black can save her stomes in the upper left corner? I believe it is really possible! An interesting challenge for all those interesting by curious results.

Click Here To Show Diagram Code: [go]$$W White to play $$ -------------------- $$| . O . X O . O W . | $$| X . X X O O O O O | $$| X X X O O M M M M | $$| O O O O M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$ -------------------- $$[/go]

I am now sure to be able to save the black stones if I add the white marked stone to my initial diagram.
Without this stone I still believe it is possible but I am not sure. I will continue to search.

luigi · #3

White can capture the Black stones at any time regardless of the scoring system.

Pio2001 · #4

Hi Gérard,
In pure area rules, there should be no such things as dead or alive stones. As clearly stated in some versions of area rules, each player gets "one point per stone of his colour still present on the board" (plus one point per territory i.e. empty intersection surrounded by stones of his colour only, plus komi, plus handicap compensation).
The only thing that matters is if the stone in "on the board" or not. The game of go in its simplest form.

But such positions are nonetheless interesting. In this perspective, they are tsumego or strategic problems, rather than rule beasts. In area scoring, rule beasts are rather found in endless cycles, or in shapes like four bent in the corner, or as Chihyung Nam say, in positions "when only one player can start a ko and the other can do nothing other than kill himself" (showing a position equal to the four bent, but not in the corner !).

In your diagram, if White would have one eye, I would have considered invading the large space, forcing White to invade the 3-3, while Black prevents the White stones to form a second eye.
But since White already has two eyes, I see no way Black can live.

Gérard TAILLE · #5

Click Here To Show Diagram Code: [go]$$W White to play $$ -------------------- $$| . O . X O . O . . | $$| X . X X O O O O O | $$| X X X O O M M M M | $$| O O O O M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$| M M M M M M M M M | $$ -------------------- $$[/go]

As Luigi or Pio2001 said it seems very improbable that black can save her stones. It was not very easy but now I have found an environment such that black can save her stones in area scoring.
That proves that you cannot easily say that a group of stones is dead without any conditions. In addition the environment I found does not look an unrealistic one!
Just a small hint: in japonese rule white can wait till confirmation phase to prove the black stones are dead which is obvious. In area scoring white has to take effectively the black stones BEFORE the final pass and there difficulties may occur.

Pio2001 · #6

With the anomalous french superko rule (Natural Situational Super Ko rule, or NSSK), we can turn anything alive by creating positions where no stone should be played anywhere on the board under the penalty of loosing another group, thus White would loose something else if she plays to actually capture the black stones, but this is an anomaly of NSSK, not of area scoring.

le_4TC · #7

Would this work?

Click Here To Show Diagram Code: [go]$$c $$ -------------------- $$| . O . X O . O . . | $$| X . X X O O O O O | $$| X X X O O X X X X | $$| O O O O X X . X X | $$| O O X X X . X O O | $$| O O X X X X O O . | $$| O O X X X X O X X | $$| O O X X X X O O X | $$| O O X X X X O . . | $$ -------------------- $$[/go]

If top left and bottom right is seki, white is losing by 5 points (4 points if she plays B8), so I'm working under the assumption that white can't afford to pass and end the game. I'm not sure how different rules handle disputes and continuation of play - but even in a continuation, surely white has to play first in the upper left if she wants to prove the black stones dead?

So black can happily pass (since he doesn't mind ending the game as is), while white is forced to play for instance H9, B8, C9. Then black starts the ko in the lower right, and plays A9 capture as a ko threat, which white should ignore. The final position is this, with black ahead by 1 point:

Click Here To Show Diagram Code: [go]$$c B+1 $$ -------------------- $$| X X . X O . O O . | $$| X . X X O O O O O | $$| X X X O O X X X X | $$| O O O O X X . X X | $$| O O X X X . X O O | $$| O O X X X X O O O | $$| O O X X X X O . . | $$| O O X X X X O O . | $$| O O X X X X O . O | $$ -------------------- $$[/go]

Note that the filled-in territories are not necessary, white will eventually have to try to take the top left group off the board, even if there are more stalling moves like H9 available.

Gérard TAILLE · #8

le_4TC wrote:

Would this work?

Click Here To Show Diagram Code: [go]$$c $$ -------------------- $$| . O . X O . O . . | $$| X . X X O O O O O | $$| X X X O O X X X X | $$| O O O O X X . X X | $$| O O X X X . X O O | $$| O O X X X X O O . | $$| O O X X X X O X X | $$| O O X X X X O O X | $$| O O X X X X O . . | $$ -------------------- $$[/go]

If top left and bottom right is seki, white is losing by 5 points (4 points if she plays B8), so I'm working under the assumption that white can't afford to pass and end the game. I'm not sure how different rules handle disputes and continuation of play - but even in a continuation, surely white has to play first in the upper left if she wants to prove the black stones dead?

So black can happily pass (since he doesn't mind ending the game as is), while white is forced to play for instance H9, B8, C9. Then black starts the ko in the lower right, and plays A9 capture as a ko threat, which white should ignore. The final position is this, with black ahead by 1 point:

Click Here To Show Diagram Code: [go]$$c B+1 $$ -------------------- $$| X X . X O . O O . | $$| X . X X O O O O O | $$| X X X O O X X X X | $$| O O O O X X . X X | $$| O O X X X . X O O | $$| O O X X X X O O O | $$| O O X X X X O . . | $$| O O X X X X O O . | $$| O O X X X X O . O | $$ -------------------- $$[/go]

Note that the filled-in territories are not necessary, white will eventually have to try to take the top left group off the board, even if there are more stalling moves like H9 available.

Yes le_4TC, that is exactly my point : when white plays to kill black stones white must give black a ko threat which will change the game in area scoring. Under japonese rule this cannot arrive because white has not to capture effectively the black stones.
As you see, assuming you use area scoring when solving a tsumego problem, then it's become far less obvious to get as result a killing seqeunce without conditions!

RobertJasiek · #9

It is area scoring, not area counting (the counting procedure is immaterial here).

Is it? Should we not rather speak of a consequence of (ko and suicide) restriction, pass and scoring rules?

Go is a global game - area scoring is a global scoring method.

Go is a global game but territory scoring (such as in Japanese style rulesets) violates the spirit by scoring locally (unless the position has exactly one global locale).

Pio2001, can you please explain why NSK has, but PSK and SSK do not have, an anomaly and in which sense is it an anomaly? There are positions with different behaviour under PSK, NSK and SSK, but such alone does not justify calling something an anomaly. There must be a more profound characterisation of an anomaly. "Can turn anything alive by creating positions where no stone should be played anywhere on the board under the penalty of loosing another group" can occur under every superko rule, can't it?

Cassandra · **#10**

RobertJasiek wrote:

Go is a global game but ...

Go is NOT a "global" game per se.

It is decided by the rules to what extent the game has a global or local character.
In particular, those parts of the set of rules are decisive with which the possibilities of infinite games are to be curtailed.

Bill Spight · **#11**

RobertJasiek wrote:

Go is a global game but territory scoring (such as in Japanese style rulesets) violates the spirit by scoring locally (unless the position has exactly one global locale).

Cassandra wrote:

Go is NOT a "global" game per se.

It is decided by the rules to what extent the game has a global or local character.
In particular, those parts of the set of rules are decisive with which the possibilities of infinite games are to be curtailed.

Japanese 1989 rules may be an abomination, but not because they use territory scoring. The thing is, kos and superkos may destroy the independence of different regions of the board, and thus, locality. Here is how correct play might go under traditional stone scoring, an "area" scoring method.

Click Here To Show Diagram Code: [go]$$Wc Stone scoring $$ -------------------- $$| 4 O 5 X O . O . 1 | $$| X 3 X X O O O O O | $$| X X X O O X X X X | $$| O O O O X X . X X | $$| O O X X X . X O O | $$| O O X X X X O O . | $$| O O X X X X O X X | $$| O O X X X X O O X | $$| O O X X X X O . 2 | $$ -------------------- $$[/go]

It is also how correct play might go under No Pass Go with Prisoner Return, which is a "territory" scoring method. It is also a global scoring method.

Bill Spight · **#12**

Change the board slightly.

Click Here To Show Diagram Code: [go]$$Wc Prisoner Return $$ -------------------- $$| 8 W 5 X O . O . 1 | $$| X 3 X X O O O O O | $$| X X X O O X X X X | $$| O O O O X X . X X | $$| O O X X X . X O O | $$| O O X 2 X X O O 9 | $$| O O X X 4 X O X X | $$| O O X X X X O O X | $$| O X X X X X O 6 7 | $$ -------------------- $$[/go]

@

After

White has 6 points, but has to play first.

RobertJasiek · **#13**

Cassandra wrote:

Go is NOT a "global" game per se.

Until successive passes, the game is global because decisions must always take into account the whole board, regardless of those rules restricting infinite play.

Cassandra · **#14**

Bill Spight wrote:

Japanese 1989 rules may be an abomination, but not because they use territory scoring. The thing is, kos and superkos may destroy the independence of different regions of the board, and thus, locality.

Yes, indeed.

The creators of the Nihon Kiin 1989 Rules should have invested a bit more time to generate a ruleset that was logically closed in itself.

What I have in mind:

With something like "When Black takes a White stone in the shape of a ko, White must not recapture the just played Black stone with the next move." only (a very "local" restriction) for prohibiting endless games, it does not seen unnatural for me to decide about the status of stones / groups "locally".

With something like "White must not play any move that would repeat an earlier board position." (a "global" restriction) to ensure the same goal, it seems natural to me to also decide about the status of stones / groups "globally".

Pio2001 · **#15**

RobertJasiek wrote:

Pio2001, can you please explain why NSK has, but PSK and SSK do not have, an anomaly and in which sense is it an anomaly? There are positions with different behaviour under PSK, NSK and SSK, but such alone does not justify calling something an anomaly. There must be a more profound characterisation of an anomaly. "Can turn anything alive by creating positions where no stone should be played anywhere on the board under the penalty of loosing another group" can occur under every superko rule, can't it?

Hi Robert,
I am talking about this situation :

Click Here To Show Diagram Code: [go]$$Bc Komi 7.5.\nFrench rules (strict NSSK).\n 3 Black passes \n4 White passes $$ --------------------- $$ | . . . O X . X X . | $$ | . O O O X X X 2 1 | $$ | O . O X X O O O O | $$ | O O X O O O . . O | $$ | O X X X X X O O . | $$ | X X . . X X O O O | $$ | . X X X . X O X X | $$ | X . X . X . X X . | $$ | . X . X X . X . X | $$ ---------------------[/go]

The only thing that the french rule says about repetitions is "a player may not, playing a stone on the board, give to the board a position that he has already given it".

After move 2 above, the game ends, but there is a dispute about the life of all black stones at the top.
According to french rules, the game resumes. Now, White has to actually capture the black stones.
Black passes.
White J9 takes the ko.
Black immediately recaptures at J8.

The move is legit : it is a board play by Black that creates a position that was never created yet by a board play of Black! This position was originally created by White 2.

That is the move I call "abnormal". The rule should never have permitted to immediately recapture in a ko. It was not the intention of the writers of the rule.
French people even told me that in this case, a referee should apply common sense and prevent Black to recapture, despite the the text of the rule allowing it.

This situation occurs when a player creates, with a board play, the shape of a ko that is open for his opponent, and the opponent passes.
The opponent can then recapture immediately every time the first players takes the ko, as long as he permanently stops making any board play other than recapturing the ko.

NSSK is the only kind of superko that allows this.

Now, White can play D4 and F1 (Black passes every time).

Click Here To Show Diagram Code: [go]$$Bc $$ --------------------- $$ | . . . O X . X X . | $$ | . O O O X X X O X | $$ | O . O X X O O O O | $$ | O O X O O O . . O | $$ | O X X X X X O O . | $$ | X X . O X X O O O | $$ | . X X X . X O X X | $$ | X . X . X . X X . | $$ | . X . X X O X . X | $$ ---------------------[/go]

These moves look stupid, but they are actually destroying Black's territory. Now the white stones D4 and F1 are alive ! In fact, all White stones wherever on the board, with or without eyes are alive, as long as Black leading the game depends on the life of his group at the top.
The reason is that if Black makes any board play, his top group immediately dies. This black board play would create a new position that Black would not be allowed to create again, then white could kill.

The British rule, written after the French rule, avoids this problem by having two independent ko rules : one superko rule (6.2 It is illegal for a player to play so as to recreate a board position of the game, previously created by a play of the same player), and another rule that forbids to immediately recapture in a ko (6.1 It is illegal for a player to capture a single stone which itself captured a single stone of the same player on the previous move).

Cassandra · **#16**

RobertJasiek wrote:

Until successive passes, the game is global because decisions must always take into account the whole board...

This is a question of strategy, not of the rules.

Nevertheless, wouldn't it be preferable to have as less "global" restrictions by the rules as possible for a "global" game? :razz:

RobertJasiek · **#17**

Thanks for the NSK anomaly! The defensive British ko rules design avoids it nicely.

I invented NSK with the wording "A player may not use a board play to recreate a position if he has used one to create it". I did overlook your anomaly. NSK was meant to improve on the rules concept of SSK. I still think the concept of SSK is dubious. Concerning NSK, mea culpa. Every ko ruleset should at least imply the prohibition of 2-play-repetition without intervening pass(es).

Anyway, I have always thought that PSK is the best.

Gérard TAILLE · **#18**

BTW I found a second type of position, but more complex:

Click Here To Show Diagram Code: [go]$$W White to play $$ --------------------- $$ | . O . X O . O . . | $$ | X . X X O O O O O | $$ | X X X O O X X X X | $$ | O O O O X X O O O | $$ | X X X X . X O O O | $$ | O O O X X X O . . | $$ | . . O X . X O X X | $$ | X X O O X X O X O | $$ | X . X O O X O O . | $$ ---------------------[/go]

The best sequence seems the following :

Click Here To Show Diagram Code: [go]$$W White to play $$ --------------------- $$ | 3 O 6 X O . O . . | $$ | X 1 X X O O O O O | $$ | X X X O O X X X X | $$ | O O O O X X O O O | $$ | X X X X . X O O O | $$ | O O O X X X O . . | $$ | 4 . O X . X O X X | $$ | X X O O X X O X O | $$ | X 5 X O O X O O . | $$ ---------------------[/go]

pass

Click Here To Show Diagram Code: [go]$$Wm7 $$ --------------------- $$ | . 2 X X O . O . . | $$ | X . X X O O O O O | $$ | X X X O O X X X X | $$ | O O O O X X O O O | $$ | X X X X . X O O O | $$ | O O O X X X O . . | $$ | X 1 O X . X O X X | $$ | X X O O X X O X O | $$ | X O . O O X O O 4 | $$ ---------------------[/go]

pass

RobertJasiek · **#19**

Cassandra wrote:

This is a question of strategy, not of the rules.

Before even considering strategy, it is a matter of concept of how the game shall be.

luigi · **#20**

RobertJasiek wrote:

I did overlook your anomaly. NSK was meant to improve on the rules concept of SSK. I still think the concept of SSK is dubious. Concerning NSK, mea culpa. Every ko ruleset should at least imply the prohibition of 2-play-repetition without intervening pass(es).

https://www.lifein19x19.com/forum/viewt ... 45&t=11510

Quite strange result in area counting rules

Who is online