It is currently Mon Mar 18, 2024 11:19 pm

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 97 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next
Author Message
Offline
 Post subject: Re: Moonlight life and go rules
Post #21 Posted: Tue Jul 13, 2021 9:45 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
With pure theoritical point of view, AGA rule allows to build a very strange seki with only one black stone!
Answer in the hidden part
You begin the game by
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . 3 . . . . . , . . . |
$$ | . . . . . . . . . 1 . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]
:w2: tenuki

and you finish the game in the following position, after a very long serie of black passes:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------------------------
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O X O O O O O O O O O |
$$ | O O O O O O O O O . O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ | O O O O O O O O O O O O O O O O O O O |
$$ ---------------------------------------[/go]


In japonese rule black will have a very large win due to all these dead white stones.
But we are in AGA rule and black is not allowed to take the white stones. Then it is a seki with a large win for white!
BTW it is a seki with no eyes and only one neutral point!

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #22 Posted: Tue Jul 13, 2021 12:30 pm 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
Oddly enough I actually had a game yesterday that showed why things should work like above (pass is valid ko threat, several but not infinite resumptions are possible).

The game had a mannenko (not converted to direct ko) and were nearly over. Opponent had more threats but only I could safely connect the ko (making seki). Opponent (probably fearing that I would convert to direct ko if allowed the last ko capture there) tried to fight the ko - and leave it open, which would also deny me a point for the last capture. There were few dame so I would have needed to pass soon (then he would pass as well).

After some discussion he let me take the ko and connect it, but if he would insist on trying I would have needed 5-10 resumptions (the number of his extra threats) to finally win the ko and the last point/prisoner.

A similar rules crisis is mentioned on sensei's page, but apparenty this can happen in everyday playing as well.

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #23 Posted: Thu Jul 15, 2021 1:39 pm 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Click Here To Show Diagram Code
[go]$$W White to play
$$ --------------
$$ | O X 1 X X |
$$ | . O X X X |
$$ | O O X X X |
$$ | X X X . X |
$$ | X X X X . |
$$ -------------[/go]
:b2: pass
:w3: pass

jann wrote:
Japanese rules may be relatively problem-free here (as most rules except strict superko). IIRC resumption is defined and should be possible.

Whether the ko can be retaken after pass (in normal play) is not completely clear but may be guessed as passing for a ko (then retake it) is also the basis in hypothetical play.


Oops I am just remebering the following post made by Bill: https://lifein19x19.com/viewtopic.php?p=259748#p259748
I was rather satisfied to say that black can retake the ko when resumption occurs but Bill seems to think that the ko ban is still in effect.
Does that mean that the question remains?

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #24 Posted: Thu Jul 15, 2021 4:02 pm 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
IIRC passes lifting bans under Japanese seemed to be the majority opinion even that time. And as I wrote above, the other way seem to break even very common situations like a mannenko with few dame.

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #25 Posted: Fri Jul 16, 2021 12:51 am 
Honinbo

Posts: 10905
Liked others: 3651
Was liked: 3374
Gérard TAILLE wrote:
Click Here To Show Diagram Code
[go]$$W White to play
$$ --------------
$$ | O X 1 X X |
$$ | . O X X X |
$$ | O O X X X |
$$ | X X X . X |
$$ | X X X X . |
$$ -------------[/go]
:b2: pass
:w3: pass

jann wrote:
Japanese rules may be relatively problem-free here (as most rules except strict superko). IIRC resumption is defined and should be possible.

Whether the ko can be retaken after pass (in normal play) is not completely clear but may be guessed as passing for a ko (then retake it) is also the basis in hypothetical play.


Oops I am just remebering the following post made by Bill: https://lifein19x19.com/viewtopic.php?p=259748#p259748
I was rather satisfied to say that black can retake the ko when resumption occurs but Bill seems to think that the ko ban is still in effect.
Does that mean that the question remains?


Not in Japanese resumption. There the ko ban is lifted. But my impression, which could easily be wrong, is that the ko ban remains in effect under Korean rules.

Edit: The current Korean rules are complicated, and there is no official English translation.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #26 Posted: Fri Jul 16, 2021 2:10 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Bill Spight wrote:
Gérard TAILLE wrote:
Click Here To Show Diagram Code
[go]$$W White to play
$$ --------------
$$ | O X 1 X X |
$$ | . O X X X |
$$ | O O X X X |
$$ | X X X . X |
$$ | X X X X . |
$$ -------------[/go]
:b2: pass
:w3: pass

jann wrote:
Japanese rules may be relatively problem-free here (as most rules except strict superko). IIRC resumption is defined and should be possible.

Whether the ko can be retaken after pass (in normal play) is not completely clear but may be guessed as passing for a ko (then retake it) is also the basis in hypothetical play.


Oops I am just remebering the following post made by Bill: https://lifein19x19.com/viewtopic.php?p=259748#p259748
I was rather satisfied to say that black can retake the ko when resumption occurs but Bill seems to think that the ko ban is still in effect.
Does that mean that the question remains?


Not in Japanese resumption. There the ko ban is lifted. But my impression, which could easily be wrong, is that the ko ban remains in effect under Korean rules.

Edit: The current Korean rules are complicated, and there is no official English translation.


That is good news, now all comments goes in the same direction: when japonese resumption is used then the ko ban is lift. It remains only the problem of the number of resumption : how avoiding an infinite number of resumption? My view was to avoid a resumption if a previous one has been occured in exactly the same situation (a loop has been detected) but I see it is a kind a superko and this is not in the spirit of japonese rule. BTW the japonese rule is really able to detect a long loop because this rule is able to conclude to a no result game (triple ko, chosei...).

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #27 Posted: Fri Jul 16, 2021 7:12 am 
Honinbo

Posts: 10905
Liked others: 3651
Was liked: 3374
Gérard TAILLE wrote:
That is good news, now all comments goes in the same direction: when japonese resumption is used then the ko ban is lift. It remains only the problem of the number of resumption : how avoiding an infinite number of resumption? My view was to avoid a resumption if a previous one has been occured in exactly the same situation (a loop has been detected) but I see it is a kind a superko and this is not in the spirit of japonese rule. BTW the japonese rule is really able to detect a long loop because this rule is able to conclude to a no result game (triple ko, chosei...).


In practice, at the professional level the number of resumptions is zero. That the question of the number of resumptions arises underscores my sense that the Japanese 1989 rules are not appropriate for amateurs. Let's look at the relevant rules.

Nihon Kiin wrote:
Article 7. Life and death
1. Stones are said to be "alive" if they cannot be captured by the opponent, or if capturing them would enable a new stone to be played that the opponent could not capture. Stones which are not alive are said to be "dead."
2. In the confirmation of life and death after the game stops in Article 9, recapturing in the same ko is prohibited. A player whose stone has been captured in a ko may, however, capture in that ko again after passing once for that particular ko capture.


In the case of the 10,000 year ko where the open ko cannot be filled, that stone in the ko mouth is dead. Capturing it does not allow a new stone to be played that the opponent could not capture. End of story. A professional who insisted that that stone is actually alive would lose face badly and be guilty of unsportsmanlike conduct.

Edit: By which I do not mean to imply that a game between two professionals would end in the unfillable ko position. The game would end with the ko filled for seki.

Conceivably, a game between two amateurs could end in such a position, which would be problematic.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.


Last edited by Bill Spight on Fri Jul 16, 2021 9:01 am, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #28 Posted: Fri Jul 16, 2021 8:57 am 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
Bill Spight wrote:
In the case of the 10,000 year ko where the open ko cannot be filled, that stone in the ko mouth is dead. Capturing it does not allow a new stone to be played that the opponent could not capture. End of story.

I'm not sure if this refers to my example, but in that case the question was not about life. If the side who cannot fill could force the ko to be left open (with no dame and less threats available for the other side), this would mean an 1 point score difference (the last prisoner) IF resumptions would not lift the ban.

Bill Spight wrote:
Edit: By which I do not mean to imply that a game between two professionals would end in the unfillable ko position. The game would end with the ko filled for seki.

Again, only if passes/resumptions lift bans. Otherwise the other side could not actually force and win the ko, and I doubt any pro would give up a point as gift.

About number of resumptions: this is not the only case where Japanese rules don't give strict definitions (possess dame etc). This has its problems but at least also has the advantage of not being wrong if common sense is used (numerous but not infinite resumptions).

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #29 Posted: Fri Jul 16, 2021 9:22 am 
Honinbo

Posts: 10905
Liked others: 3651
Was liked: 3374
jann wrote:
Bill Spight wrote:
In the case of the 10,000 year ko where the open ko cannot be filled, that stone in the ko mouth is dead. Capturing it does not allow a new stone to be played that the opponent could not capture. End of story.

I'm not sure if this refers to my example, but in that case the question was not about life. If the side who cannot fill could force the ko to be left open (with no dame and less threats available for the other side), this would mean an 1 point score difference (the last prisoner) IF resumptions would not lift the ban.


Yes. Against a stubborn amateur opponent, the game should be reopened and the ko taken and filled. if possible. If the stubborn opponent plays a regular ko threat and then takes the ko back, the game should be reopened again, in the new position. There is only one reopening for each position.

But what if the stubborn opponent plays in a double ko seki
before taking the ko back? Now there is the possibility of an infinite number of reopenings from the same position. Given the history of the Japanese rules and the official commentary, it seems plain that no such ko threat should be effective. Hypothetical play governs. And the pros plainly agree about the hypothetical play. But you can't tell this from the text of the rules. You have to rely upon the official commentary, as well.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #30 Posted: Fri Jul 16, 2021 10:24 am 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
I think mannenko + double ko seki is similar to a last endgame ko + double ko seki. If the last point decides the winner, it is triple ko (no resumptions since no stoppage either, as now both sides have infinite threats).

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #31 Posted: Fri Jul 16, 2021 3:10 pm 
Lives in gote

Posts: 418
Liked others: 9
Was liked: 83
Rank: kgs 5 kyu
KGS: Pio2001
Gérard TAILLE wrote:
Click Here To Show Diagram Code
[go]$$W White to play
$$ --------------
$$ | O X . X X |
$$ | . O X X X |
$$ | O O X X X |
$$ | X X X . X |
$$ | X X X X . |
$$ -------------[/go]


Click Here To Show Diagram Code
[go]$$W White to play
$$ --------------
$$ | O X 1 X X |
$$ | . O X X X |
$$ | O O X X X |
$$ | X X X . X |
$$ | X X X X . |
$$ -------------[/go]
:b2: pass
:w3: pass

Assume AGA rule is used. Whites stones are alive.
In this context what is the correct wording for the status of white group? It doesn't seem to be a seki with another black group (without two eyes) and, in addition, a seki implies at least one neutral point. It is neither an unconditional living group with two eyes is it? May be we can say it is a living group by miracle!


Hi Gérard,
In article 10 of AGA rules, after resumption, it is said that "If they both pass while a disagreement still exists, all stones remaining on the board are alive, and the board in counted as it stands".

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #32 Posted: Sat Jul 17, 2021 5:32 am 
Lives in sente

Posts: 1037
Liked others: 0
Was liked: 180
Sorry, but I keep looking at that and can't see how the final position cold have been reached (in ko) EXCEPT by black having passed after white had made a self atari move.

Care to present the last half dozen moves or so instead of just the final position. Say from a couple moves before the two black stones were captured.

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #33 Posted: Sat Jul 17, 2021 7:23 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Mike Novack wrote:
Sorry, but I keep looking at that and can't see how the final position cold have been reached (in ko) EXCEPT by black having passed after white had made a self atari move.

Care to present the last half dozen moves or so instead of just the final position. Say from a couple moves before the two black stones were captured.

Oops I do not see what is really your problem. Anyway here is an example of game (without any pass)


Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #34 Posted: Sat Jul 17, 2021 7:44 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
I am afraid that the style of your "game" sequence might be very infantile.

Black has had numerous chances to win that game by 25 points. Therefore -- at this childish level of understanding of the game -- it does not matter, whether White's group might be alive (but without any territory), resulting in a Black win by 13 points "only". Black will be happy to count the Smarties he got for free during the game.

I can't get rid of the feeling that this kind of discussion is comparable to that about an AI who is supposed to solve a task for which it has not been trained.
Neither must the AI be blamed for her poor performance, nor a rule set for not giving a 100%-answer to a question that is has never be designed for.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #35 Posted: Sat Jul 17, 2021 7:50 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Click Here To Show Diagram Code
[go]$$W White to play
$$ --------------
$$ | O X . X X |
$$ | . O X X X |
$$ | O O X X X |
$$ | X X X . X |
$$ | X X X X . |
$$ -------------[/go]

AGA rule : we saw that in the position above white can live by taking the ko. That's true but it does not look to be in the spirit of the go game. Comparing to japonese rule this result is due to the fact that, when in AGA rule the play is resumed after two passes, all the ko bans (superko) are not lifted. Do you know if it was a real choice of the rule (in order to avoid adding some complexity to the rule)?

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #36 Posted: Sat Jul 17, 2021 8:03 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Cassandra wrote:
I am afraid that the style of your "game" sequence might be very infantile.

Black has had numerous chances to win that game by 25 points. Therefore -- at this childish level of understanding of the game -- it does not matter, whether White's group might be alive (but without any territory), resulting in a Black win by 13 points "only". Black will be happy to count the Smarties he got for free during the game.

I can't get rid of the feeling that this kind of discussion is comparable to that about an AI who is supposed to solve a task for which it has not been trained.
Neither must the AI be blamed for her poor performance, nor a rule set for not giving a 100%-answer to a question that is has never be designed for.


I agree with you but the purpose here was not to play good or bad moves but only to consider pure theoritical point concerning the rule. In itself the game I showed is really completly stupid in practice but the point was only to show that the final position can be reach with legal play.
Obviously I did not understand correctly Mike Novack issue. Sorry for that.

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #37 Posted: Sat Jul 17, 2021 8:29 am 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
Gérard TAILLE wrote:
Comparing to japonese rule this result is due to the fact that, when in AGA rule the play is resumed after two passes, all the ko bans (superko) are not lifted. Do you know if it was a real choice of the rule (in order to avoid adding some complexity to the rule)?

It is very hard for pass to lift bans under superko as this breaks the essence of superko and reinstates most problems it tried to solve (like sending2-returning1 abuse for area scoring).

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #38 Posted: Sat Jul 17, 2021 1:34 pm 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
Gérard TAILLE wrote:
but the point was only to show that the final position can be reach with legal play.

"Legal" play ALONE is not sufficient for a discussion of rule sets at the very high level that you apparently intend to do.

Utilising positions that can be reached only after several "infantile" mistakes (of both sides) cannot be the right method for qualified discussions at a "professional" level. Intending to do so disqualifies yourself.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #39 Posted: Sat Jul 17, 2021 2:49 pm 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Cassandra wrote:
Gérard TAILLE wrote:
but the point was only to show that the final position can be reach with legal play.

"Legal" play ALONE is not sufficient for a discussion of rule sets at the very high level that you apparently intend to do.

Utilising positions that can be reached only after several "infantile" mistakes (of both sides) cannot be the right method for qualified discussions at a "professional" level. Intending to do so disqualifies yourself.


Does that mean you do not like unrealistic problems for which the initial position can be reached only by several "infantile" mistakes ?

Top
 Profile  
 
Offline
 Post subject: Re: Moonlight life and go rules
Post #40 Posted: Sat Jul 17, 2021 7:52 pm 
Lives in sente

Posts: 757
Liked others: 114
Was liked: 916
Rank: maybe 2d
jann wrote:
Gérard TAILLE wrote:
Comparing to japonese rule this result is due to the fact that, when in AGA rule the play is resumed after two passes, all the ko bans (superko) are not lifted. Do you know if it was a real choice of the rule (in order to avoid adding some complexity to the rule)?

It is very hard for pass to lift bans under superko as this breaks the essence of superko and reinstates most problems it tried to solve (like sending2-returning1 abuse for area scoring).


This was solved years and years ago by our very own Bill Spight, no? Consider the condition "the game ends if the same player passes in a position in which that player passed before". It is easy to see how even if passes lift ko bans, that this rule still mathematically guarantees that the game is finite. It handles sending-two-returning one fine. And it even can be justified intuitively: "even when passes lift ko bans, if when I pass you have nothing better to do than cycle back the same position where I can pass yet again, clearly you have nothing you can accomplish, so the game should be over".

Depending on what you are trying to accomplish (e.g. if you want to have agreement phases or other details of game end procedure), maybe you can also consider variants. But the key essence is the condition "same player passes in a position where that player passed before".


This post by lightvector was liked by: hzamir
Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 97 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group