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

Gérard TAILLE · Post by **Gérard TAILLE** » Thu Sep 16, 2021 12:50 pm

jann wrote:Compare your example to lightvector's. In both cases one side has a decisive play that makes miai of two local kos and creates life on double ko, thus safe in normal play.

safe in normal play ????
lightvector's position looks like

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

which is a death double ko isn't it?

jann · Post by **jann** » Thu Sep 16, 2021 1:01 pm

Click Here To Show Diagram Code: [go]$$B :w4: above :b1: $$ ---------------------- $$ | O . O . X 2 3 O O . O $$ | O O O X O X O O O O O $$ | X X X O 1 O O . O . . $$ | X . X O O O O O . . . $$ | . X X . . . . . . . . $$ | X X . . . . . . . . . $$ | . . . . . . . . . . . $$ | . . . . . . . . . . .[/go]

W corner safe on double ko (just like B in your example, after starting both edge kos simultaneously).

OC if you add a remote double ko seki, life on double ko may not be favorable since it allows the opponent to go for triple ko. But J89 calls both W in lightvector's and B in your original example dead even if no double ko seki exists - which is already wrong (and differs from Korean).

Correcting myself from above posts: the miai doesn't work (at least for B uncapturability) after 2 W reinforcement moves, only earlier, since B doesn't get to go first to create it.

Gérard TAILLE · Post by **Gérard TAILLE** » Thu Sep 16, 2021 2:00 pm

[quote="jann"]OC if you add a remote double ko seki, life on double ko may not be favorable since it allows the opponent to go for triple ko. But J89 calls both W in lightvector's and B in your original example dead even if no double ko seki exists - which is already clearly wrong (and differs from Korean).

Let's try to analyse the position WITHOUT double ko seki:

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

In normal play black cannot be killed, even if it is white to play.
What about confirmation phase? It seems black is always dead even in J89-pass-once-per-ko! Do you agree?
BTW, OC, in GT territory rule there are no territory because GT territory rule is based on normal play.

jann · Post by **jann** » Thu Sep 16, 2021 2:09 pm

Gérard TAILLE wrote:
$$B
$$ -----------------------------------------
$$ | . . . . . O O O . . O O X . X O . . . |
$$ | . . . . . O O X O O X O O X X O . O . |
$$ | . . . . . . O X X X X . X . X O . . . |
$$ | . . . , . . O O O O X X X X O O O O O |
$$ | . . . . . . . . . O O O X X X O X X X |
$$ | . . . . . . . . . . . O X . X X . X . |
$$ | . . . . . . . . . . . O O X X X X O X |
$$ | . . . . . . . . . . . . O O O X . O O |
$$ | . . . . . . . . . . . . . . O X X X O |
$$ | . . . , . . . . . , . . . . O O X O . |
$$ | . . . . . . . . . . . . . . . O X O . |
$$ | . . . . . . . . . . . . . . . O X X O |
$$ | . . . . . . . . . . . . . . . O O O O |
$$ | . . . . . . . . . . . . . . . . . O O |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------
Click Here To Show Diagram Code
[go]$$B $$ ----------------------------------------- $$ | . . . . . O O O . . O O X . X O . . . | $$ | . . . . . O O X O O X O O X X O . O . | $$ | . . . . . . O X X X X . X . X O . . . | $$ | . . . , . . O O O O X X X X O O O O O | $$ | . . . . . . . . . O O O X X X O X X X | $$ | . . . . . . . . . . . O X . X X . X . | $$ | . . . . . . . . . . . O O X X X X O X | $$ | . . . . . . . . . . . . O O O X . O O | $$ | . . . . . . . . . . . . . . O X X X O | $$ | . . . , . . . . . , . . . . O O X O . | $$ | . . . . . . . . . . . . . . . O X O . | $$ | . . . . . . . . . . . . . . . O X X O | $$ | . . . . . . . . . . . . . . . O O O O | $$ | . . . . . . . . . . . . . . . . . O O | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]
In normal play black cannot be killed, even if it is white to play.

This is what I wrote above: after 2 W reinforcements, W first seem to kill now (in normal play) by connecting once on either side. B first could create the double ko life. This position seems unsettled. (EDIT: W doesn't kill only makes unfavorable ko, see below)

Gérard TAILLE · Post by **Gérard TAILLE** » Thu Sep 16, 2021 2:18 pm

jann wrote:
Gérard TAILLE wrote:
$$B
$$ -----------------------------------------
$$ | . . . . . O O O . . O O X . X O . . . |
$$ | . . . . . O O X O O X O O X X O . O . |
$$ | . . . . . . O X X X X . X . X O . . . |
$$ | . . . , . . O O O O X X X X O O O O O |
$$ | . . . . . . . . . O O O X X X O X X X |
$$ | . . . . . . . . . . . O X . X X . X . |
$$ | . . . . . . . . . . . O O X X X X O X |
$$ | . . . . . . . . . . . . O O O X . O O |
$$ | . . . . . . . . . . . . . . O X X X O |
$$ | . . . , . . . . . , . . . . O O X O . |
$$ | . . . . . . . . . . . . . . . O X O . |
$$ | . . . . . . . . . . . . . . . O X X O |
$$ | . . . . . . . . . . . . . . . O O O O |
$$ | . . . . . . . . . . . . . . . . . O O |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------
Click Here To Show Diagram Code
[go]$$B $$ ----------------------------------------- $$ | . . . . . O O O . . O O X . X O . . . | $$ | . . . . . O O X O O X O O X X O . O . | $$ | . . . . . . O X X X X . X . X O . . . | $$ | . . . , . . O O O O X X X X O O O O O | $$ | . . . . . . . . . O O O X X X O X X X | $$ | . . . . . . . . . . . O X . X X . X . | $$ | . . . . . . . . . . . O O X X X X O X | $$ | . . . . . . . . . . . . O O O X . O O | $$ | . . . . . . . . . . . . . . O X X X O | $$ | . . . , . . . . . , . . . . O O X O . | $$ | . . . . . . . . . . . . . . . O X O . | $$ | . . . . . . . . . . . . . . . O X X O | $$ | . . . . . . . . . . . . . . . O O O O | $$ | . . . . . . . . . . . . . . . . . O O | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]
In normal play black cannot be killed, even if it is white to play.
This is what I wrote above: after 2 W reinforcements, W first seem to kill now (in normal play) by connecting once on either side. B first could create the double ko life. This position seems unsettled.

How white can kill black in normal play?

Click Here To Show Diagram Code: [go]$$W $$ ----------------------------------------- $$ | . . . . . O O O . . O O X 3 X O . . . | $$ | . . . . . O O X O O X O O X X O . O . | $$ | . . . . . . O X X X X 2 X . X O . . . | $$ | . . . , . . O O O O X X X X O O O O O | $$ | . . . . . . . . . O O O X X X O X X X | $$ | . . . . . . . . . . . O X . X X . X . | $$ | . . . . . . . . . . . O O X X X X O X | $$ | . . . . . . . . . . . . O O O X 4 O O | $$ | . . . . . . . . . . . . . . O X X X O | $$ | . . . , . . . . . , . . . . O O X O . | $$ | . . . . . . . . . . . . . . . O X O 1 | $$ | . . . . . . . . . . . . . . . O X X O | $$ | . . . . . . . . . . . . . . . O O O O | $$ | . . . . . . . . . . . . . . . . . O O | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

jann · Post by **jann** » Thu Sep 16, 2021 2:21 pm

You are right, I meant preventing the double ko life but B still has at least one local threat, so it is a winnable ko for him. Still not unconditional life though.

Gérard TAILLE · Post by **Gérard TAILLE** » Thu Sep 16, 2021 2:37 pm

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

This position with the double ko seki is really strange:
1) In normal play the double ko allows white to avoid black living without condition => black can only force NO RESULT to avoid losing her group.
2) In confirmation phase and J89-pass-once-per-ko it is the contrary : the double ko allows black to live if white add only two moves.
Interesting no?

jann · Post by **jann** » Thu Sep 16, 2021 2:42 pm

Interesting yes. I think the right way to analyze this from here would be to create a table, showing the outcome for 0-1-2-3 W reinforcement moves, with and without double ko seki, and for all potential rules (normal play, J89 pass-each-time, J89 pass-once, Korean). Would make a nice and informatible table.

Btw life on double ko is known to be vulnerable and potentially become triple ko later, even in simpler positions. Any double ko restricts further ko play, often in an asymmetrical way.

Gérard TAILLE · Post by **Gérard TAILLE** » Thu Sep 16, 2021 3:14 pm

jann wrote:Interesting yes. I think the right way to analyze this from here would be to create a table, showing the outcome for 0-1-2-3 W reinforcement moves, with and without double ko seki, and for all potential rules (normal play, J89 pass-each-time, J89 pass-once, Korean). Would make a nice and informatible table.

Btw life on double ko is known to be vulnerable and potentially become triple ko later, even in simpler positions. Any double ko restricts further ko play, often in an asymmetrical way.

Oops, a lot of work

If you try this work I promise to help you Jann (except for korean rule because I have not enough information concerning Life and Death analysis).
I am happy you found this position interesting because I had to take a lot of time to find it!

jann · Post by **jann** » Thu Sep 16, 2021 3:42 pm

No sry, I cannot even imagine doing that much, for a still error-prone table.

But I'd expect it to show some differences between pass-once and pass-each-time here and there. But also that the latter itself would give anomalous result (similarly to lightvector's example) in some cases (like in the original position with 0 W moves) - so even the "correct" ruling could be doubtful.

jann · Post by **jann** » Sat Sep 18, 2021 10:23 am

I just noticed an interesting comment in J89. In commentary on 7.2:

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

This liberal phrasing and the "as if the game had been resumed" seems present in the current Japanese text as well (with slighly more verbosity). I'm not sure how literally this can be taken, but this also supports the interpretation that passing for a certain ko in confirmation is only required once - after which it reverts to a normal ko (behaves like in resumption) for the player (thus no double ko flaw / closed loop).

CDavis7M · Post by **CDavis7M** » Tue Oct 19, 2021 6:57 pm

jann wrote:I just noticed an interesting comment in J89. In commentary on 7.2:

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
This liberal phrasing and the "as if the game had been resumed" seems present in the current Japanese text as well (with slighly more verbosity). I'm not sure how literally this can be taken, but this also supports the interpretation that passing for a certain ko in confirmation is only required once - after which it reverts to a normal ko (behaves like in resumption) for the player (thus no double ko flaw / closed loop).

This is jumping to conclusions. If you look at the actual example, the question is whether black needs to play A to reinforce after "wining" the ko. The answer is YES because otherwise black is dead.

Back to your statement, the reason that black is dead is because even though the most recent move in the game was black taking the ko, white can immediately retake the ko because this is treated as if the game were resumed with white to play (black has passed).

There is no suggestion anywhere in the Examples that passing for the ko once is sufficient such that passes no longer need to be made to retake (e.g., others moves being played is sufficient as in normal gameplay).

jann · Post by **jann** » Tue Oct 19, 2021 7:46 pm

There was no definite conclusion so no jumping either. The notice was not about the actual example but about a side comment in the commentary which may shed some light on how pass for ko is supposed to work and be interpreted ("the situation for that ko is the same as if the game had been resumed").

Also you really should use the English version, your comments doesn't really make sense and sound like you refer to some misunderstood Japanese (you mix up reinforcement in game vs reinforcement in hypothetical play (which is free), also mix up pass for ko vs recapturing after the opponent's pass).

CDavis7M · Post by **CDavis7M** » Tue Oct 19, 2021 8:40 pm

jann wrote:Also you really should use the English version, your comments doesn't really make sense and sound like you refer to some misunderstood Japanese (you mix up reinforcement in game vs reinforcement in hypothetical play (which is free), also mix up pass for ko vs recapturing after the opponent's pass).

The conclusion that was jumped to was the conclusion that this statement in the rule comments supports the linked interpretation.

Also, I don't think I was mixing things up. I understand that a reinforcement can be played after the game is stopped. But if an additional stone needs to be played to prove life and death status because playing it could begin a double ko that can't be stopped, then it might seem as if that stone is required. But it's not required to be played. The reason that it's not required is because life and death status is defined by the examples (as precedent) without any need to "play it out". I'm just starting from the board position in the examples as they are given. The way that they work is by definition.

I don't even want to get started on the original post.

santo · Post by **santo** » Mon Nov 08, 2021 8:32 am

I might have missed something in this thread, but as far as I have seen it proposes a new version of the hypothetical-ko rule and then goes on to apply it to many examples. Is there a known example where this Cassandra proposal of a new hypothetical-play ko rule gives a different life/death status than applying the Jasiek 2003 rules? That, plus the actual informal Japanese professional opinion for that position, would be the most interesting in deciding whether the interpretation can be taken as correct / so far better / so far worse than existing formalizations like Jasiek 2003. My understanding from reading the thread is that so far, the "Cassandra proposal" and the "Jasiek 2003" proposal agree on the shown examples.

Life In 19x19

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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