J89's pass-for-ko: Misinterpreted in the Western Go World?

For discussing go rule sets and rule theory
kvasir
Lives in sente
Posts: 1040
Joined: Sat Jul 28, 2012 12:29 am
Rank: panda 5 dan
GD Posts: 0
IGS: kvasir
Has thanked: 25 times
Been thanked: 187 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by kvasir »

Cassandra wrote:
jann wrote: That particular example needs a bit of caution though, with dubious sentences like this:
Note that if White 7 was a pass for the ko 6/11 on the right, Black could have played 8 at 12 and captured the white group on the right.
This note -- that contains an obious mistake -- has been deleted in the meantime from the Japanese original.
Should it ever have been in the legal text at the time of the translation into English.
My point was only that if you wish to find ambiguity in the translation vis-à-vis the original it is not enough to stumble on a single word somewhere, you need to take into account if the correct interpretation can be determined by for example the commentary. In this case the translation has multiple examples of passing for a specific ko before recapturing, and these same examples are in the original.

The translation appears to be more precise in specifying which ko is being passed for than the original. I checked with google translate and surprisingly this example refers to passes for the "top corner" and "bottom corner" so there is something different going on there that might be interesting to find an explanation for.

Some of the commentary is also puzzling in that "confirmation" has the players fighting over a double-ko elsewhere on the board without addressing why there is not an infinite cycle (except the incorrect "note" that Jann pointed to that doesn't appear to be found in the original).
jann
Lives in gote
Posts: 445
Joined: Tue May 14, 2019 8:00 pm
GD Posts: 0
Been thanked: 37 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by jann »

The other linguistic question may still remain though:
jann wrote:"passing once for that particular ko capture" might mean the ko can be freely (according the normal ko rule) taken by the player afterwards, without passing for it next time. (edit: for clarity)
In Japanese the relevant sentence reads: "Tadashi kō o torareta kata ga torikaesu kō no sorezore ni tsuki chakushu hōki o okonatta nochi wa, arata ni sono kō o toru koto ga dekiru."

Davies: "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."

My Japanese is not good enough, but "ONCE" does not seem to be emphasized in Japanese. Still it seems possible to interpret both ways (nochi ... arata ni), and I'm not sure if this is clarified anywhere in the commentary.

OC, if passing is really only required once for each ko and not for later recaptures there, most of the double ko problems could disappear.

EDIT much later - the commentary on 7.2 actually gives a hint:
If a player whose stone has been captured in a ko has passed for that particular ko ...
the situation for that ko is the same as if the game had been resumed: the player may now capture in that ko again
So a ko pass is related to a resumption there. This might also hint that pass for a ko is supposed to be a one-time action, after which it reverts to a normal ko for the player (behaves like in resumption). Still not decisive evidence oc - the traditional interpretation also remains possible (best would be an official clarification).
Last edited by jann on Tue Dec 14, 2021 10:49 am, edited 5 times in total.
Gérard TAILLE
Gosei
Posts: 1346
Joined: Sun Aug 23, 2020 2:47 am
Rank: 1d
GD Posts: 0
Has thanked: 21 times
Been thanked: 57 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by Gérard TAILLE »

jann wrote:The other linguistic question may still remain though:
jann wrote:"passing once for that particular ko capture" might mean the ko can be freely (normally) taken by the player afterwards, without passing for it next time.
In Japanese the relevant sentence reads: "Tadashi kō o torareta kata ga torikaesu kō no sorezore ni tsuki chakushu hōki o okonatta nochi wa, arata ni sono kō o toru koto ga dekiru."

Davies: "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."

My Japanese is not good enough, but "ONCE" does not seem to be emphasized in Japanese. Still it seems possible to interpret both ways (nochi ... arata ni), and I'm not sure if this is clarified anywhere in the commentary.

OC if passing is really only required once for each ko and not for later recaptures there, most of the double ko problems could disappear.
Surely I cannot read japanese and I am not able to guess the intention of the rule author, but take the following position:
Click Here To Show Diagram Code
[go]$$B
$$ +---------------------------------------+
$$ | X . X O . . O X . X O . O X . . . . . |
$$ | X X O O O O O X X O O O O X . . . . . |
$$ | X O O X X X O X O . O X X X . . . . . |
$$ | . X X X X O X X X O O X . . . , . . . |
$$ | . . . . X O X . X O X X . . . . . . . |
$$ | . . . . X O X X O X X . . . . . . . . |
$$ | . . . . X O O O O . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]
With the curent James Davies translation white stones at the top are dead and with Thomas interpretation they are alive.
If you are able to say which is the expected result for this position then you will automatically know if, on this point, James Davies translation is correct or not.
Note that, basically, this position corresponds simply to moonshine life position.
jann
Lives in gote
Posts: 445
Joined: Tue May 14, 2019 8:00 pm
GD Posts: 0
Been thanked: 37 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by jann »

I was talking about the third possible interpretation (pass for ko only needed once for each ko, for the first recapture - see above), which makes W dead here (and will likely agree with Davies / traditional interpretation in most cases, except a double ko cannot be abused in confirmation).
kvasir
Lives in sente
Posts: 1040
Joined: Sat Jul 28, 2012 12:29 am
Rank: panda 5 dan
GD Posts: 0
IGS: kvasir
Has thanked: 25 times
Been thanked: 187 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by kvasir »

Gérard TAILLE wrote:
Surely I cannot read japanese and I am not able to guess the intention of the rule author, but take the following position:
Click Here To Show Diagram Code
[go]$$B
$$ +---------------------------------------+
$$ | X . X O . . O X . X O . O X . . . . . |
$$ | X X O O O O O X X O O O O X . . . . . |
$$ | X O O X X X O X O . O X X X . . . . . |
$$ | . X X X X O X X X O O X . . . , . . . |
$$ | . . . . X O X . X O X X . . . . . . . |
$$ | . . . . X O X X O X X . . . . . . . . |
$$ | . . . . X O O O O . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]
This results in repeated positions if white insists on always taking in the double-ko. The j89 rules are flawed in that they do not compel an end to the game, the players can agree on a "no result" during the game per article 12 but now we are in "confirmation" and there appears to be nothing in the rules about the possibility of repeated positions during "confirmation" but the players can resume the game and try to agree on "no result". I don't see how passing for a ko has anything to do with this if white just insists to always capture in the double-ko.

You just can't talk about things like "the expected result" when there is nothing to compel a unique result.
jann
Lives in gote
Posts: 445
Joined: Tue May 14, 2019 8:00 pm
GD Posts: 0
Been thanked: 37 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by jann »

kvasir wrote:I don't see how passing for a ko has anything to do with this if white just insists to always capture in the double-ko.
W cannot capture in double ko continuously, he will need to pass soon (allowing B to progress with the other ko, using his blocked ko advantage given by the rules). Pass for ko matters here because - with the usual interpretation - it can also be abused to make a continuous cycle (W can now take in double ko and pass for the other repeatedly, and this is forcing B do the same). This is not a problem if pass for ko only needed once for each ko, though (my third possible interpretation above).
kvasir
Lives in sente
Posts: 1040
Joined: Sat Jul 28, 2012 12:29 am
Rank: panda 5 dan
GD Posts: 0
IGS: kvasir
Has thanked: 25 times
Been thanked: 187 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by kvasir »

jann wrote:
kvasir wrote:I don't see how passing for a ko has anything to do with this if white just insists to always capture in the double-ko.
W cannot capture in double ko continuously, he will need to pass soon (allowing B to progress). Pass for ko matters here because - with the usual interpretation - it can be abused to make a continuous cycle (W can now take ko and pass for the other repeatedly, and this is forcing for B to do the same). This is not a problem if pass for ko only needed once for each ko, though (my third possible interpretation above).
I should have made a diagram as follows:
Click Here To Show Diagram Code
[go]$$B
$$ +---------------------------------------+
$$ | X 2 X O 1 . O X 4 X O . O X . . . . . |
$$ | X X O O O O O X X O O O O X . . . . . |
$$ | X O O X X X O X O 5 O X X X . . . . . |
$$ | 3 X X X X O X X X O O X . . . , . . . |
$$ | . . . . X O X . X O X X . . . . . . . |
$$ | . . . . X O X X O X X . . . . . . . . |
$$ | . . . . X O O O O . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]
It seems to me that black can't make any progress now if white passes for :b5: because this forces black to pass for :w4:.
Click Here To Show Diagram Code
[go]$$W :w1: pass :b2: pass
$$ +---------------------------------------+
$$ | X W . O X . O X O 4 O . O X . . . . . |
$$ | X X O O O O O X X O O O O X . . . . . |
$$ | X O O X X X O X 3 X O X X X . . . . . |
$$ | X X X X X O X X X O O X . . . , . . . |
$$ | . . . . X O X . X O X X . . . . . . . |
$$ | . . . . X O X X O X X . . . . . . . . |
$$ | . . . . X O O O O . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]
Black still hasn't passed for the marked stone and won't have time to do so.
jann
Lives in gote
Posts: 445
Joined: Tue May 14, 2019 8:00 pm
GD Posts: 0
Been thanked: 37 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by jann »

Pls see my above comment: this is the traditional interpretation. But if passing for a ko is only required ONCE, B will (after one cycle) be freed from the duty of passing for the right (since both he and W already passed for both kos there, that part reverts to a normal double ko seki). Thus B will have time to progress on the left, passing for the ko there and then recapturing it (blocking for W).
kvasir
Lives in sente
Posts: 1040
Joined: Sat Jul 28, 2012 12:29 am
Rank: panda 5 dan
GD Posts: 0
IGS: kvasir
Has thanked: 25 times
Been thanked: 187 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by kvasir »

jann wrote:Pls see my above comment: this is the traditional interpretation. But if passing for a ko is only required ONCE, B will (after one cycle) be freed from the duty of passing for the right (since both he and W already passed for both kos there, that part reverts to a normal double ko seki) and thus B will have time to cash in his blocked ko advantage on the left, after passing for the ko there and then recapturing it (blocking for W).
I may be misunderstanding. I thought what you said implied that white also could take in the double ko without passing when he has passed before. Black would be forced to take in the double ko, else this happen...
Click Here To Show Diagram Code
[go]$$W :b2: pass for the marked stone ko
$$ +---------------------------------------+
$$ | X W . O X . O X 1 X O . O X . . . . . |
$$ | X X O O O O O X X O O O O X . . . . . |
$$ | X O O X X X O X O . O X X X . . . . . |
$$ | X X X X X O X X X O O X . . . , . . . |
$$ | . . . . X O X 3 X O X X . . . . . . . |
$$ | . . . . X O X X O X X . . . . . . . . |
$$ | . . . . X O O O O . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]
jann
Lives in gote
Posts: 445
Joined: Tue May 14, 2019 8:00 pm
GD Posts: 0
Been thanked: 37 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by jann »

I meant both sides become free to play in the double ko (which reverts to a normal double ko seki, after both have already passed for both kos there). But note this does not help W - B can still progress (even answering each W move on the right).

In a normal double ko seki W has infinite threats - but ko threats are useless for the left, where (after B recaptured it) W would also need to pass for the ko once before recapturing.

(EDIT: diagram) So after one cycle:
Click Here To Show Diagram Code
[go]$$W
$$ +---------------------------------------+
$$ | X O 4 O X . O X 1 X O . O X . . . . . |
$$ | X X O O O O O X X O O O O X . . . . . |
$$ | X O O X X X O X O 2 O X X X . . . . . |
$$ | X X X X X O X X X O O X . . . , . . . |
$$ | . . . . X O X . X O X X . . . . . . . |
$$ | . . . . X O X X O X X . . . . . . . . |
$$ | . . . . X O O O O . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]
And :w3: cannot do anything but pass (no further pass-for-ko on the right since both sides have already passed for both kos there, and OC cannot immediately recapture :b2: because of the normal ko rule). Then after :b4:, :w5: can flip the double ko one last time, B answers, then :w7: pass (for left or just normally, doesn't matter, :b8: captures).
Last edited by jann on Fri Sep 03, 2021 10:12 am, edited 5 times in total.
User avatar
Cassandra
Lives in sente
Posts: 1326
Joined: Wed Apr 28, 2010 11:33 am
Rank: German 1 Kyu
GD Posts: 0
Has thanked: 14 times
Been thanked: 153 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by Cassandra »

RobertJasiek wrote:I repeat my question for the particular position at the current moment.
Before that, let me digress a little again. That will make it unnecessary to fight your way through GT territory rules:: Honte's loops and cycles.

Let...
:w1: :b1: be ko captures / re-captures into ko-shape #1; :w2: :b2: be ko captures / re-captures into ko-shape #2; and so on...
:white: :black: be GENUINE "pass".
:wx: :bx: be "pass-for-ko" of the "traditional" J89's kind; i.e. ONE "pass" for lifting the ko-ban of ONE particular ko-shape, into which re-capturing is allowed immmediately thereafter.
:wc: :bc: be "pass-for-all-ko" of your J2003's kind; i.e. ONE "pass" for lifting the ko-bans of EVERY ko-shape, into which re-capturing is allowed immediately thereafter.
:wt: :bt: be "pass-for-ko" of the "current" J89's kind (provided my understanding is correct); i.e. ONE "pass" for lifting the ko-ban of ONE particular ko-shape, but into which re-capturing is allowed ONLY after ALL active ko-bans have been lifted before.
:b91: :w92: :b93: :w94: etc. genuine moves that do NOT capture/ re-capture into a ko-shape.

WHITE's task in the following formations is to prevent her group(s) from becoming captured. whatsoever the costs.
BLACK's task is the opposite, he has to do everything he can to capture White's group, whatsoever the costs.


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

SIMPLE-ko & eye

Click Here To Show Diagram Code
[go]$$B
$$ +------------------
$$ | X Q . Q . O X . .
$$ | X X Q O O O X . .
$$ | . X X X X X X . .
$$ | . . . . . . . . .
$$ | . . . . . . . . .[/go]
Black starts with :b1: .

White dreams of a cycle :w1: :b1: :w1: :b1: ..., but also knows that her dream will never become true, as the sequence :b1: :w1: (that started with Black's initial move) is forbidden by the rules.

Thus, she can only hope for a cycle :white: :black: :w1: :black: :white: :b1: ..., which would effectively result in her desired cycle above, but without breaking the rules.

Black, however, likes the sequence :white: :b91: more, with :b91: filling the ko-shape.

Please note that using :black: :white: was sufficient above.
There was NO NEED AT ALL to resort to any of :wx: :bx: :wc: :bc: :wt: :bt:.



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

DOUBLE-ko

Click Here To Show Diagram Code
[go]$$B
$$ +----------------
$$ | . Y . Y O X . .
$$ | X X Y O O X . .
$$ | Q X O . O X . .
$$ | . Q O O O X . .
$$ | Q O X X X X . .
$$ | X X X . . . . .
$$ | . . . . . . . .
$$ | . . . . . . . .[/go]
Black starts with :b1: .

White already knows very well that the sequence :b1: :w1: is forbidden by the rules.
She now dreams of a cycle starting with :w2: :b2: , but has to admit that Black would never follow her, as the sequence :w2: :b2: is also forbidden by the rules.

A valid alternative for keeping her group alive is :w2: :black: :w1: :b2: :white: :b1: ...
Half the length of this cycle (= 3) is an ODD number, which means that -- in principle -- the RIGHT to initiate one of these halves alternates between both players.
Here, Black is FORCED to initiate the second half of the cycle, as :w1: was an atari (as was his initial :b1: as well). However, he might want to refrain from his second :b1:, as White played "only" a :white: before.
In order to keep the cycle going, both players must cooperate. However, none of these is forced to do so.

The other valid alternative for keeping White's group alive is :w2: :black: :white: :b2: :w1: :black: :white: :b1: ..., which effectively results in her desired sequence above, but without breaking the rules.
However, despite half the length of this cycle (= 4) being an EVEN number, cooperation between both players is also necessary to keep this cycle going, as Black will soon realise that :b2: :w1: :black: :white: is about the same as :b1: :w2: :black: :white: that did not lead to success just before, due to the symmetry of the formation. Being very well aware of this, he might not even want to start with his intial :b1: .

Please note that using :black: :white: was sufficient above.
There was NO NEED AT ALL to resort to any of :wx: :bx: :wc: :bc: :wt: :bt:.


Let's now consider what might happen when using the "path-for-ko" options mentioned above.

----------------------------------

##### :wx: :bx: #####

Black knows from above that starting with :b1: will not lead to success. Thus, he will refrain from doing so.

In this case White -- not being aware that the formation is symmetrical -- has the option of creating the following cycle on her own:
:w1: :b2: :wx: :bx: :w2: :b1: :wx: :bx: ...

:b2: is ENFORCED, as :w1: was an atari.
:wx: enables :w2: that would be an atari.
Black is FORCED to play :bx:, in order to make :b1: possible. Otherwise, his group would become captured.

White is able to keep Black busy for a while.
But sooner or later she will realise that her ENFORCED cycle has the same effect as doing nothing. And thus will stop continuing, in order to enable a status confirmation elsewhere on the board.

Please note that half the length of this cycle (= 4) is an EVEN number. This results in the SAME player being able to start every single ENFORCED half of this cycle FOREVER.
This is the very hidden unfortunate side-effect of J89's initial version!


----------------------------------

##### :wc: :bc: #####

Black knows from above that starting with :b1: will not lead to success. Thus, he will refrain from doing so.

In this case White -- not being aware that the formation is symmetrical -- has the option of creating the following cycle on her own:
:w1: :b2: :wc: :b1: :w2: :bc: ...

:b2: is ENFORCED, as :w1: was an atari.
:wc: enables :w2: that would be an atari.
However, :wc: also enabled :b1: . Therefore, Black is free to continue with playing a GENUINE move, as the danger of his group becoming captured after :w2: has vanished.

Half the length of this cycle (= 3) is an ODD number, which means that the RIGHT to initiate one of these halves alternates between both players.
The transition from the first half of this cycle to the second one is NOT forced. This means that the cycle is an OPTIONAL one.


Thus, Black would like the sequence :w1: :b2: :wc: :b91: more, with :b91: being tenuki.
White may be able to continue with :w2: :b1: :wc: :b93: , with :b93: being tenuki again. You you will easily recognise that White -- in principle -- simply continues with a meaningless local sequence, while Black is able to play several moves in a row elsewhere.

----------------------------------

##### :wt: :bt: #####

Black knows from above that starting with :b1: will not lead to success. Thus, he will refrain from doing so.

In this case White -- not being aware that the formation is symmetrical -- has the dream of creating the following cycle on her own:
:w1: :b2: :wt: :bt: :w2: :b1: :wt: :bt: ...

:b2: is ENFORCED, as :w1: was an atari.
:wt: is mandatory for becoming able to play :w2: that would be an atari. However, :wt: is NOT the ONLY precondition that must be fulfilled to really being able to do so!
:bt: is mandatory for becoming able to play :b1: , if Black ever wanted to do so (but which is very hazardous to assume).
All former ko bans have been lifted, so White is able to play :w2:, starting another have of White's desired cycle, with both ko-shapes interchanged.

However, is was Black's move :bt: that FINALLY made :w2: a valid move. Black knows for sure that White is hindered to play :w2: without.
Therefore, Black likes the sequence :w1: :b2: :wt: :b91: more, as there is no danger of his group becoming captured.

By capturing once into a DOUBLE-ko, White stops herself from being able to capture into that DOUBLE-ko ever again -- in principle.
There will will be NO cycle, unless Black co-operates. Thus, such cycle will be an OPTIONAL one.



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

DOUBLE-ko & LARGE eye

Click Here To Show Diagram Code
[go]$$B
$$ +--------------------------
$$ | . O . O X O . . O O X . .
$$ | O O O X X O X X X O X . .
$$ | X O X . X O O . . O X . .
$$ | . X X X X O O O O O X . .
$$ | X X O O O O X X X X X . .
$$ | O O O X X X X . . . . . .
$$ | X X X X . . . . . . . . .
$$ | . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . .[/go]
Black knows from above that starting with :b1: will not lead to success. Thus, he will refrain from doing so, and play :b91: into White's LARGE eye at the right, instead.

In this case White -- not being aware that the DOUBLE-ko formation at left is symmetrical -- has the dream of creating the following cycle on her own:
:w2: :b1: :white: :b2: :w1: :black: ...
Half the length of this cycle (= 3) is an ODD number, which means that the RIGHT to initiate one of these halves alternates between both players.
Here, Black is NOT forced to initiate the second half of the cycle, as White played "only" a :white: before.
In order to keep the cycle going, both players must cooperate. However, none of these is forced to do so.

Black would like the sequence :w2: :b1: :white: :b93: :w1: :b2: :white: :b95: etc. more, in order to make sure that White understands that her one-eyed group at right could be taken off the board. Sooner or later, White will admit that continuesly re-capturing into the DOUBLE-ko only delays a result that cannot be prevented anyway.

Please note that using :black: :white: was sufficient above.
There was NO NEED AT ALL to resort to any of :wx: :bx: :wc: :bc: :wt: :bt:.


Let's now consider what might happen when using the "path-for-ko" options mentioned above.

----------------------------------

##### :wx: :bx: #####

Black knows from above that starting with :b1: will not lead to success. Thus, he will refrain from doing so, and play :b91: into White's LARGE eye at the right, instead.

Even in this case (please note that :b91: was NO ATARI) White has the option of creating the following cycle on her own:
:w1: :b2: :wx: :bx: :w2: :b1: :wx: :bx: ...

:b2: is ENFORCED, as :w1: was an atari.
:wx: enables :w2: that would be an atari.
Black is FORCED to play :bx:, in order to make :b1: possible. Otherwise, his DOUBLE-ko group would become captured.

White is able to keep Black busy FOREVER!
Sooner or later Black will realise that the ENFORCED cycle that White initiated will make it IMPOSSIBLE to EVER return to White's LARGE eye at the right. He finally admits that White's groups are "uncapturable".

This undesirable result is a direct consequence of the very hidden unfortunate side-effect of J89's initial version!
Please note that Black would have to reduce the liberties of White's group at right to at least TWO BEFORE THE GAME STOPS, if he wanted the status confirmation declaring White's formation to be "dead".


----------------------------------

##### :wc: :bc: #####

Black knows from above that starting with :b1: will not lead to success. Thus, he will refrain from doing so, and play :b91: into White's LARGE eye at the right, instead.

In this case White dreams of creating the following cycle on her own:
:w1: :b2: :wc: :b1: :w2: :bc: ...

:b2: is ENFORCED, as :w1: was an atari.
:wc: enables :w2: that would be an atari.
However, :wc: also enabled :b1: . Therefore, Black is free to continue with playing a GENUINE move, as the danger of his group becoming captured after :w2: has vanished.

Half the length of this cycle (= 3) is an ODD number, which means that the RIGHT to initiate one of these halves alternates between both players.
The transition from the first half of this cycle to the second one is NOT forced. This means that the cycle is an OPTIONAL one.


Thus, Black would like the sequence :w1: :b2: :wc: :b93: more, with :b93: being played into White's eye again.
White may be able to continue with :w2: :b1: :wc: :b95: , with :b95: being played into White's eye again. Sooner or later, White will understand that her one-eyed group at right could be taken off the board, and that continuesly re-capturing into the DOUBLE-ko only delays a result that cannot be prevented anyway.

----------------------------------

##### :wt: :bt: #####

Black knows from above that starting with :b1: will not lead to success. Thus, he will refrain from doing so, and play :b91: into White's LARGE eye at the right, instead.

In this case White has the dream of creating the following cycle on her own:
:w1: :b2: :wt: :bt: :w2: :b1: :wt: :bt: ...

:b2: is ENFORCED, as :w1: was an atari.
:wt: is mandatory for becoming able to play :w2: that would be an atari. However, :wt: is NOT the ONLY precondition that must be fulfilled to really being able to do so!
:bt: is mandatory for becoming able to play :b1: , if Black ever wanted to do so (but which is very hazardous to assume).
All former ko bans have been lifted, so White is able to play :w2:, starting another have of White's desired cycle, with both ko-shapes interchanged.

However, is was Black's move :bt: that FINALLY made :w2: a valid move. Black knows for sure that White is hindered to play :w2: without.
Therefore, Black likes the sequence :w1: :b2: :wt: :b93: :white: :b95: :white: etc. (with :b93: :b95: also played into White's LARGE eye at the right) more, which will easily prove that White's stones at the right can be taken off the board. White has NOT been given any opportunity to lenghten this process!

By capturing once into a DOUBLE-ko, White stops herself from being able to capture into that DOUBLE-ko ever again -- in principle.
There will will be NO cycle, unless Black co-operates. Thus, such cycle will be an OPTIONAL one.



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

DOUBLE-ko & CHAIN of FALSE eyes
Click Here To Show Diagram Code
[go]$$B
$$ +--------------------------------
$$ | . O O . O . O X . X O . O X . .
$$ | O X X O O O O X X O O O O X . .
$$ | X X . X X X O X O . O X X X . .
$$ | . . . , X O X X X O O X . . . ,
$$ | . . . . X O X . X O X X . . . .
$$ | . . . . X O X X X O X . . . . .
$$ | . . . . X O O O O X X . . . . .
$$ | . . . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . . . .[/go]
Black knows from above that he cannot kill White's group in the DOUBLE-ko directly.
Thus, he starts with :b1: capturing into the ko-shape in the upper left corner.

White knows already that the following TRIPLE-ko cycle will bail her out:
:w2: :b3: :w1: :b2: :w3: :b1: ...
( :w2: :b3: and :b2: :w3: are played in the DOUBLE-ko at right)
Half the length of this cycle (= 3) is an ODD number, which means that the RIGHT to initiate one of these halves alternates between both players.
Here, Black is NOT forced to initiate the second half of the cycle, as :w1: before was NO ATARI.

However, as he is unable to attack White's CHAIN of FALSE eyes from the inside, there is nothing else but a "pass" that he can play.
In this case, the cycle will become
:w2: :b3: :w1: :black: :white: :b1: :w3: :b2: :w1: :black: :white: :b1: ...

Please note that half the length of this cycle (= 6) is an EVEN number. This results in WHITE being able to start every single half of this cycle FOREVER.
This is the ONE AND ONLY application case that makes a SPECIAL REGULATION necessary!!!




However, instead of installing a general overall SINGLE-ko ban, the following might have been easier, more user-friendly and less prone to errors:
By the rules of status confirmation, "MOONSHINE LIFE" is declared "dead"!!!



Another proposal with the same characteristics :
Capturing into a DOUBLE-ko must never be used as a ko-threat for a SINGLE-ko during status confirmation!!!



Let's now consider what might happen when using the "path-for-ko" options mentioned above.

----------------------------------

##### :wx: :bx: #####

White's central group that is attached to Black's DOUBLE-ko group has MORE than TWO liberties!
Black knows from above that he will be unable to approach this group, in order to give atari. Thus, he declares defeat outright.

This undesirable result is a direct consequence of the very hidden unfortunate side-effect of J89's initial version!

----------------------------------

##### :wc: :bc: #####

White knows from above that he will be unable to prevent Black from approaching her central group at the upper edge, which is attached to Black's DOUBLE-ko group, so giving atari. Thus, she declares defeat outright.

----------------------------------

##### :wt: :bt: #####

White knows from above that he will be unable to prevent Black from approaching her central group at the upper edge, which is attached to Black's DOUBLE-ko group, so giving atari. Thus, she declares defeat outright.


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

Your request
Click Here To Show Diagram Code
[go]$$B :w8: pass
$$ +---------------------------------------+
$$ | X X Q . Q . O X Q M O . O X . . . . . |
$$ | . X X Q O O O X Y Q O O O X . . . . . |
$$ | X X . X X X O X T Y O X X X . . . . . |
$$ | . . . , X O X X Y O O X . . . , . . . |
$$ | . . . . X O X . X O X X . . . . . . . |
$$ | . . . . X O X X O X X . . . . . . . . |
$$ | . . . . X O O O O . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]
:ex: resulted from White capturing into the upper DOUBLE-ko shape with :w2: .
Black has NOT yet played a "pass" for that ko-shape.

:et: resulted from White playing a "pass" for the lower DOUBLE-ko shape.
Before, this point was occupied by :ex: , too, which resulted from Black capturing into that ko-shape with :b3: .

NOT for EACH of the ko to re-capture has been played a "pass" yet. Thus, White must not re-capture into that lower of the DOUBLE-ko shapes, where one of her stones has been captured in a ko long before.
Last edited by Cassandra on Fri Sep 03, 2021 10:59 pm, edited 1 time in total.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
User avatar
Cassandra
Lives in sente
Posts: 1326
Joined: Wed Apr 28, 2010 11:33 am
Rank: German 1 Kyu
GD Posts: 0
Has thanked: 14 times
Been thanked: 153 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by Cassandra »

Cassandra wrote:... another suggestion for you:
Click Here To Show Diagram Code
[go]$$B
$$ +-------------------+
$$ | X O . X . X O . . |
$$ | X O X X X X O . . |
$$ | . O X O O O O . . |
$$ | O O X O O O O O O |
$$ | X X O O O O O O O |
$$ | X X X X O O O X O |
$$ | . . , X X X X X X |
$$ | . X . X . . . . X |
$$ | . . . X . . . . X |
$$ +-------------------+[/go]
A snapshot from a game on the 9x9 board; komi = 7.5; no prisoners.

:b1: What do you think, has White already passed for stopping the game?

To make the suggestion clear:
:w1: As a rules expert, will you ever be asked to do a status analysis for this position?
:w2: If so, what could have been the questioner's motive?
Dear Gérard,

Apperently, the THREE questions above were too difficult? Well, here comes ONE simpler replacement:

:b2: What do you think, which of the FOUR snapshots below has been taken AFTER the respective game stopped?
(Rest of the board "pass-alive", including the visible border stones, to be clear.)
Click Here To Show Diagram Code
[go]$$B Snapshot A
$$ +---------------------------------------+
$$ | X . X O . . O X . X O . O X X ? X O ? |
$$ | X X O O O O O X X O O O O X ? X X O O |
$$ | X O O X X X O X O . O X X X X X X O ? |
$$ | . X X O X O X X X O O X O O O O O O O |
$$ | X X O O X O X . X O X X O X X X X X O |
$$ | . X O O X O X X O X X O O X ? X ? X O |
$$ | X O O O X O O O O O O O X X X X X X O |
$$ | X X X O X X X X X X X X X O O O O O O |
$$ | X ? X O O O O O O O O O O O ? ? ? ? ? |
$$ | ? X X ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | X X X ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ||[/go]
Click Here To Show Diagram Code
[go]$$B Snapshot B
$$ +---------------------------------------+
$$ | X O . O X . O X . X O . O X ? ? ? ? ? |
$$ | X X O O O O O X X O O O O X ? ? ? ? ? |
$$ | X O O X X X O X O . O X X X ? ? ? ? ? |
$$ | X X X X X O X X X O O X ? ? ? ? ? ? ? |
$$ | ? ? ? ? X O X . X O X X ? ? ? ? ? ? ? |
$$ | ? ? ? ? X O X X O X X ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? X O O O O O O ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ||[/go]
Click Here To Show Diagram Code
[go]$$B Snapshot C
$$ +---------------------------------------+
$$ | X . X O . . O X . X O . O X ? ? ? ? ? |
$$ | X X O O O O O X X O O O O X ? ? ? ? ? |
$$ | X O O X X X O X O . O X X X ? ? ? ? ? |
$$ | . X X X X O X X X O O X ? ? ? ? ? ? ? |
$$ | . X ? ? X O X . X O X X ? ? ? ? ? ? ? |
$$ | . X ? ? X O X X O X X ? ? ? ? ? ? ? ? |
$$ | X X ? ? X O O O O O O ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |[/go]
Click Here To Show Diagram Code
[go]$$B Snapshot D
$$ +---------------------------------------+
$$ | X . X O . . O X . X O . O X X ? X O ? |
$$ | X X O O O O O X X O O O O X ? X X O O |
$$ | X O O X X X O X O . O X X X X X X O ? |
$$ | . X X O X O X X X O O X O O O O O O O |
$$ | X X O O X O X . X O X X O X X X X X O |
$$ | . X O O X O X X O X X O O X ? X ? X O |
$$ | X O O O X O O O O O O O X X X X X X O |
$$ | . X X O X X X X X X X X X O O O O O O |
$$ | X ? X O O O O O O O O O O O ? ? ? ? ? |
$$ | ? X X ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | X X X ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? |
$$ | ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ||[/go]
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Gérard TAILLE
Gosei
Posts: 1346
Joined: Sun Aug 23, 2020 2:47 am
Rank: 1d
GD Posts: 0
Has thanked: 21 times
Been thanked: 57 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by Gérard TAILLE »

jann wrote:I was talking about the third possible interpretation (pass for ko only needed once for each ko, for the first recapture - see above), which makes W dead here (and will likely agree with Davies / traditional interpretation in most cases, except a double ko cannot be abused in confirmation).
I believe I understand.
I tried to build an example for which this third interpretation may give a result different from J2003 but I am not completly sure.
Click Here To Show Diagram Code
[go]$$B
$$ +-------------------+
$$ | B B . O X O . O . |
$$ | B B O O X X O O O |
$$ | O O O O X . X O O |
$$ | . O O O X X O O O |
$$ | O X O O X . X X X |
$$ | X X . O X X X O O |
$$ | X O O O O O O O O |
$$ | . X O X X X X X X |
$$ | X . X . X . . X . |
$$ +-------------------+[/go]
What is the problem?
Black to play can kill all white groups => all white groups are dead.
What about black groups? All black groups are uncapturable except the four marked stones in the upper left corner.
So, the only remaining question is the following : are these four black stones alive or dead?

In J2003 they are alive due to the sequence:
Click Here To Show Diagram Code
[go]$$W :b5: ko-pass
$$ +-------------------+
$$ | B B 3 O X O . O . |
$$ | B B O O X X O O O |
$$ | O O O O X . X O O |
$$ | 2 O O O X X O O O |
$$ | 7 X O O X . X X X |
$$ | X X 1 O X X X O O |
$$ | X O O O O O O O O |
$$ | 8 X O X X X X X X |
$$ | X 6 X 4 X . . X . |
$$ +-------------------+[/go]
But what will happen with this third interpretation of the rule. It seems to me that the loop through the three ko can reappear because now white has already pass for the ko on the left => four black marked stones are dead ?
jann
Lives in gote
Posts: 445
Joined: Tue May 14, 2019 8:00 pm
GD Posts: 0
Been thanked: 37 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by jann »

It's likely there are a few cases where "pass once" and "pass each time" for a ko would differ, since they are not the same thing logically. Only J89 authors could say which was meant, but it is quite possible the traditional interpretation was the correct one (for absolute ko independence, and with the double ko pass-for-ko flaw remaining overlooked).

This example seems more tricky though: this is NOT a real moonshine life, not even a combined one. W can make this into real ko fight in actual game (for triple ko draw or life with eyes). IMO an important trait of real moonshine life is that it can always be left untouched until two passes (only L/D is in question). Never a ko fight since winning such ko is impossible for one side.
Gérard TAILLE
Gosei
Posts: 1346
Joined: Sun Aug 23, 2020 2:47 am
Rank: 1d
GD Posts: 0
Has thanked: 21 times
Been thanked: 57 times

Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Post by Gérard TAILLE »

jann wrote:It's likely there are a few cases where "pass once" and "pass each time" for a ko would differ, since they are not the same thing logically. Only J89 authors could say which was meant, but it is quite possible the traditional interpretation was the correct one (for absolute ko independence, and with the double ko pass-for-ko flaw remaining overlooked).

This example seems more tricky though: this is NOT a real moonshine life, not even a combined one. W can make this into real ko fight in actual game (for triple ko draw or life with eyes). IMO an important trait of real moonshine life is that it can always be left untouched until two passes (only L/D is in question). Never a ko fight since winning such ko is impossible for one side.
I completly agree with you Jann; "pass once" and "pass each time" for a ko would differ, and this position is not a real moonshine life.
In addition when I analyse this position with GT territory rule it is clear that the board is not black territory. The reason is quite simple; the game is not finished and white should not pass in normal play in this position.

I guess we can find a finished position with the same kind of result (a loop reappearing) but it will surely take some time to find it. ;-)
Post Reply