Odd Cases 🤔 in the Japanese Rules

[cdavis.7m]

@jannn

There is no such thing as other stones “already deemed to be alive”

There IS such a thing because the players are human and they understand what “confirmation” means. Kakunin 確認, “confirmation”, is in the rules. There is obviously no such thing as “string-by-string, checking capturability without enable for each string” because nothing even close to this statement is written in the rules. If you’re going to talk about the rules, there’s no need to make stuff up. Just read and follow.

@jannn

It is possible to invent different interpretations for enable, and as someone wrote a Japanese referee might even ignore the written rule and apply his own informal judgement, unconstrained by the need of logical consistency (like you).

So you’re saying that it can be hard for people to understand their own flawed logic? Is that right, huh? Interesting.

@jannn

Just like in example 4, where it is the approach move that enables the save (and new stones) already, the actual capture itself is not even needed for that. But this is irrelevant - W is alive there because the capture cannot be performed without enabling the save.

Read the rules again. It cannot be the “approach move”, which is a Black stone, that “enables” White to play a new uncapturable stone. It must be a capturable White stone that so-called enables White to play a new uncapturable White stone.

By the way, Example 4 is seki with dame. My position is that Variation A and B have dame. Example 4 does not support your position (hint: none of the examples support your position).

@jannn

This would only matter if you would be interested in L/D for the position AFTER the ko was already taken. But for scoring the stopped position you need L/D for that position, not for a different position a few moves later. And from the scoring position capturing the corner string would necessarily enable a new stone.

I am not looking at the L&D after. I have no idea how you could come to that conclusion. Actually, I do have an idea of how you could do it, but that’s besides the point. The point is that your use of the term “necessary” is incorrect. The 5 stones do not necessarily have to be captured to “enable” the new uncapturable stone. Because in this position the 5 stones are not captured yet White has placed new uncapturable stones.

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.

Because the 5 stones do not even have to be captured for new uncapturable stones to be played, then the 5 stones cannot be said to “enable” the new uncapturable stones. It is the 1 stone that “enables” the the uncapturable stones. Therefore, the capture of the 5 stones adds nothing to the player’s confirmation of L&D status. The 5 stones are dead and so they have dame.

[jannn]

@cdavis.7m

It cannot be the “approach move”, which is a Black stone, that “enables” White to play a new uncapturable stone. It must be a capturable White stone that so-called enables White to play a new uncapturable White stone.

Nonsense. W (or a W stone) cannot enable W (himself) to do anything. When B tries to capture W, it is always a B move that enables W to play a new uncapturable stone. In a simple snapback, a B move (capturing W stones) enables W to play under them.

The phrase “capturing them would enable” is impossible to misunderstand to this extent. Capturing is a B action.

@cdavis.7m

The 5 stones do not necessarily have to be captured to “enable” the new uncapturable stone.

Backwards. It doesn’t matter whether the capture is necessary (ie. the only way) for enabling the new stone, or is there also another way to enable it.

1. “capturing them would enable X” = capturing necessarily makes X possible

2. capturing is necessary for enabling X = there are no other way to make X possible

This is elementary logic. The rule is the former - what matters is whether the play of the new stone (necessarily and inavoidably) becomes possible if the capture is attempted.

[cdavis.7m]

@jannn

When B tries to capture W, it is always a B move that enables W to play a new uncapturable stone. In a simple snapback, a B move (capturing W stones) enables W to play under them.

The phrase “capturing them would enable” is impossible to misunderstand to this extent. Capturing is a B action.

I would not word it this way but I do not disagree with this different statement, though it does not match the Japanese wording which rely on the stones being assessed, not the capturing move. I still disagree with the statement " it is the approach move that enables the save" regarding Example 4. In that example, the approach move is not capturing any capturable White stones, so it is not “enabling.” I would say that it is the capturable (living) stones that “enable” new uncapturable stones.

@jannn

1. “capturing them would enable X” = capturing necessarily makes X possible
2. capturing is necessary for enabling X = there are no other way to make X possible

This is elementary logic. The rule is the former - what matters is whether the play of the new stone (necessarily and inavoidably) becomes possible if the capture is attempted.

You’re misunderstanding what “enable” means because you are trying to think about “logic” and not about “reading comprehension.” In order for the 5 White stones to be captured, the 1 separate White stone must be captured. But if Black is attempting show the L&D status of 5 stones, why does Black capture a different stone? Because the capture of the 5 stones depends on the capture of the 1 stone. Once the 1 stone is captured, White is already enabled to play the new uncapturable stones. By the definition of the word “enable,” capturing the 5 White stones cannot be said to “enable” something that is already possible.

[elsantodel90]

Jannn, what about the position in the first go seigen dispute in Rule disputes involving Go Seigen at Sensei's Library?

It seems that by your logic, unless I am misunderstanding it, the first example would be deemed alive by Japanese 1989 rules “enable” clause, and black would not need to fill the ko. This is because even if white takes the ko first, black immediately plays at “a”, and even though white can connect the ko, there is no way for white to avoid black getting a new permanent stone at “a”.

However, this is a well known precedent, and the current rule (meaning, the way the Japanese pros and referees actually interpret the 1989 rules, because we have these particular positions as well known examples) says that black must fill and defend that ko, and cannot claim that extra point, regardless of ko threats.

In effect, the “enabled stone at a” does not count for the purpose of “trying to keep a direct ko open”. In the Jasiek 2003 model, this is in fact a very specific special case, for which two extra definitions are introduced (capturable-3 and local-3), because the natural “enabled” definition in Jasiek 2003 (that implied by capturable-2) happens to count stones like “a” (and also stones like in the example currently being discussed here: all of them are in the local-2 of the relevant stones), and thus would deem it alive. Thus the extra rule “the game cannot end with a capturable-3 ko stone” (precisely like the one in the go seigen example), which seems to try to “fit” the Jasiek 2003 definition of “enable” into this Japanese tradition where “these stones in the grey-area in a direct ko do not count, you cannot leave the direct ko open” (but this special rule fails to detect the examples that are under discussion, because in those it is not the ko stone itself that is a capturable-2 ko stone, but a close by related group also involved in the direct ko).

In particular, I fail to understand how you would make this following part apply to the positions at hand, but however not apply neither to the go seigen case nor to a one-sided dame, unless you invoke locality arguments like the Jasiek 2003 models tries to do (the player can force his one-sided dame whenever he wants, and the opponent cannot prevent it, regardless of whether the opponent goes for a capture of any group or not)

@jannn

Backwards. It doesn’t matter whether the capture is necessary (ie. the only way) for enabling the new stone, or is there also another way to enable it.

1. “capturing them would enable X” = capturing necessarily makes X possible
2. capturing is necessary for enabling X = there are no other way to make X possible

This is elementary logic. The rule is the former - what matters is whether the play of the new stone (necessarily and inavoidably) becomes possible if the capture is attempted.

[jannn]

@elsantodel90

Jannn, what about the position in the first go seigen dispute in Rule disputes involving Go Seigen at Sensei’s Library?

It seems that by your logic, unless I am misunderstanding it, the first example would be deemed alive by Japanese 1989 rules “enable” clause, and black would not need to fill the ko. This is because even if white takes the ko first, black immediately plays at “a”, and even though white can connect the ko, there is no way for white to avoid black getting a new permanent stone at “a”.

In that position black can play a permanent stone at “a” regardless of what W does. It was always possible, was never preventable, thus not enabled by (made possible by) the capture.

There can be various ways to enable the play of the new stone, but there must also have been a way to prevent it originally (and OC originally means starting from the stopped/scoring position - that we want L/D of - not from a later one).

@elsantodel90

In particular, I fail to understand how you would make this following part apply to the positions at hand, but however not apply neither to the go seigen case nor to a one-sided dame

B must show either a sequence that captures without W playing a certain permanent stone, or another sequence that shows how W could have played that stone anyway, even if B tried to prevent it (again, starting from the stopped/scoring position). One sided dame is useless.

[elsantodel90]

@jannn

[…], or another sequence that shows how W could have played that stone anyway, even if B tried to prevent it (again, starting from the stopped/scoring position). One sided dame is useless.

But there is such a sequence here, I mean, white CAN FORCE both E18 and F19 unconditionally, from the starting position, even if black starts. I am assuming that is what you mean by “show a sequence of how W could have played that stone anyway, even if B tried to prevent it, starting from the scoring position”.

The sequence is like this: if black D19, white takes everything in fact. If black F19, white fills E17 and will take everything. If black G19, white will take with F19 and get everything. So black must take the ko at E17, only remaining option. Then, white passes to lift that koban. Now black has absolutely no way to avoid white getting E18 and F19: if black fills ko at E18, just white G19 and white will capture everything. If black F19, white G19 makes it a dead ko and captures everything. If black G19, white retakes ko at E18, and now black can only self atari so white will capture everything (if black passes for ko, white fills E17 and still gets everything).

The only remaining line to analyze is black takes E17, white passes for the ko, and black plays D19. Then white takes the ko at E18, and since black is in atari and cannot retake ko, capturing at B19 is the only move to not lose everything. After what white plays F19.

In absolutely every case, not matter what black does, white can get bot E18 and F19, unconditionally, from the starting position. In the analysis we never cared at all whether the corner was captured or not, our goal was focused on getting E18 and F19, and we can unconditionally secure both from the starting position. So it does not satisfy the “[…] but there must also have been a way to prevent it originally” condition.

By the way, a similar analysis to this shows that in the Torazu Sanmoku position, the new stones that the small (one-stone) group can place after capture outside the 1-stone-string itself, are also unavoidable already from the starting position (in this same sense: make it your goal from the starting position to place those permanent stones and calculate the sequence, not worrying whether there is or is not capture of the group, and you will find that you can force them).

So actually requiring them to be preventable in some alternative line of play would deem that group as dead in the Torazu Sanmoku seki (the position would however end up as a seki / antiseki by chance, because one string would be alive and the other dead, but both with dame so 0 points. But it points to the fact that matching all of these enable results with the Japanese tradition seems tricky, without saying also that “you cannot leave a ko open that would only be considered alive because of the enable rule and placing new permanent stones outside the group itself”).

[jannn]

@elsantodel90

white CAN FORCE both E18 and F19 unconditionally, from the starting position, even if black starts. I am assuming that is what you mean by “show a sequence of how W could have played that stone anyway, even if B tried to prevent it, starting from the scoring position”.

W cannot play an uncapturable stone at E18 if B tries to prevent it. You are missing B’s best resistance.

[yebellz]

@jannn

You are missing B’s best resistance.

I’m missing it as well. Could you clarify?

[jannn]

Pass.

[yebellz]

What do you mean? Sorry, can you explain the sequence and context where black should pass?

[jannn]

For W to play an uncapturable stone at E18, it must be empty. B can prevent the play of a stone there simply by not helping W.

[yebellz]

Could you please clarify the context in which you are speaking? I cannot follow your argument. It seems that you originally suggested:

@jannn

You are missing B’s best resistance.

in response to something specific that @elsantodel90 said, like in refutation of a particular line of play. However, I don’t see any flaws with what @elsantodel90 has said, nor understand how your argument aims to refute it.

[jannn]

I mean playing a stone at E18 only becomes possible when B takes the ko, but he is not obliged to do that.

@jannn

B must show either a sequence that captures without W playing a certain permanent stone, or another sequence that shows how W could have played that stone anyway, even if B tried to prevent it

B cannot show a sequence that captures the corner without W playing a new uncapturable stone at E18. He also cannot show a sequence that shows how W could have played that stone if B tried to prevent it, because W can not play an uncapturable stone at E18 if B just passes in the original position.

[yebellz]

Your argument still seems a bit vague, but let me try to dissect what you’ve written.

@jannn

B cannot show a sequence that captures the corner without W playing a new uncapturable stone at E18.

I think this is the essence of what @elsantodel90 is discussing. Black demonstrating the capture of the five white stones in the corner would lead to White playing new stones at E18 and F19. Thus, depending on how one interprets the “enables” clause, one could argue that those five captured White stones are alive.

@jannn

He also cannot show a sequence that shows how W could have played that stone if B tried to prevent it, because W can not play an uncapturable stone at E18 if B just passes in the original position.

If Black just passes, then Black does not do anything to demonstrate the death of the five stones in the corner, so why does it matter for White to play another uncapturable stone at E18?

What exactly is the aim of your argument? Are you saying that White needs or does not need to play another move to defend?

[elsantodel90]

I see, your point is that “E18 counts” basically because although it was a white stone to begin with, and a NEW white stone is now played there, not just leaving the original. So basically “getting to play a NEW stone below any intersection where a stone of the player was before, even if in a different group” would always count (because by the same infinite passing argument, a NEW stone can never be forced there in rulesets without suicide).

Still, the Japanese rules seem to say that in the Torazu Sanmoku, both groups are considered alive because of the enabled stones. But then even if the opponent passes, the new stones are forced, so using this same logic in the torazu sanmoku, the single-stone-group would be considered dead. So this interpretation is still problematic.

[jannn]

When B tries to show that W could have played the stone in question anyway, he doesn’t need to think about the capture. This is exactly what he tries to prove: that playing the stone was not made possible by the capture, but it was already possible in the original position.

He just needs to show a sequence where W can play the stone even against the best resistance, if B is ONLY concerned in preventing it. This is possible for things that are truly unrelated to the capture, like one sided dame, or new stones into own territory. But not possible if there actually is a causal relationship with the capture: if it was NOT possible originally, but inavoidably BECAME possible in the course of the capture.

@elsantodel90 seem to understood what I meant. Torazu3 is a well known issue for what does capturing the single stone “enables” there, discussed a few times in the past, but I doubt there is any stretch of interpretation of “enable” that would make it alive. Particularly as the commentary itself shows that W can play the same spots in all lines.

[cdavis.7m]

If anyone is interested, John Fairbairn posted his own translation of the Japanese Rule on Life and Death that “that follows the Japanese as closely as possible in every respect - vocabulary, grammar, syntax, order or whatever”.
https://lifein19x19.com/viewtopic.php?p=269950#p269950

No surprise that the word “enable” is not part of the translation. And of course there is nothing in the rule about “in the course of the capture.”

---

@elsantodel90

Still, the Japanese rules seem to say that in the Torazu Sanmoku, both groups are considered alive because of the enabled stones. But then even if the opponent passes, the new stones are forced, so using this same logic in the torazu sanmoku, the single-stone-group would be considered dead. So this interpretation is still problematic.

The interpretation is problematic. But going back to jann’s suggestion, it’s silly to talk about “Black” passing or having pass as their strongest resistance. There is no requirement that the player of the Black stones be the one to confirm the L&D status of the White stones by placing the first black stone. This should be obvious as disputes are contemplated. And trying to pass as the first “move” is not an attempt to capture, so it is not even L&D confirmation.

Second, the Japanese Rules (Ⅲ 死活確認例 | 棋戦 | 囲碁の日本棋院) definitely say that that the 1 White stone (白一子) is alive (活き石) and that the 4 black stones are alive in the Torazu Sanmoku 取らず三目 position.

---

I think understanding of the position being discussed relies on an understanding of Life and Death (Article 7-1) as described in Example 24 of the rules and an understanding of what Life and Death “confirmation” is from the definition of “confirmation” (determination of something unknown). If can be captured and newly form uncapturable stones without being captured. Then showing that and can be captured while only newly forming uncapturable stones again does not confirm anything beyond being independently alive. is extraneous to being alive. So is dead.

---

It actually should just be really simple if we think about the possibilities during L&D confirmation: Is there any way for capturable stones to be alive with territory besides snapback and nakade? If there was such a way then the Japanese Rules would have an example of it.

The 5 White stones in the corner are capturable in L&D confirmation but there is no snapback or nakade. So then, shouldn’t they be seki stones?

[elsantodel90]

[EDIT] Ignore all that follows, I am blind

White can trivially show that single stone is alive, by letting all the corner die when black plays the eternal life and just capturing and connecting to the outside. I do not know why I thought that the “eternal life cycle” was forced.

==== ORIGINAL WRONG POST ====

I don’t remember anybody writing in the thread about this before, and I just noticed this (though I have a vague feeling that I had once known this before, but cannot remember): Is the “Life-and-Death Example 6: position before long life” example plainly wrong? In particular, it seems to be in direct contradiction with the explanation given in “Life-and-Death Example 25: Double-Ko Seki”.

The mistake in example 6 seems to be that they completely forgot about the existance of the single white stone, which is never mentioned nor analized. Instead, they described only the 4-stone white string (which is indeed alive). When we consider the single white stone in the 1-6 intersection, then just as the ko stones in example 25, that stone is taken and retaken an infinite number of times in the eternal sequence, and a permanent stone is never established at all. So, it must be dead by the exact same logic explained in example 25. Example 25 must be correct in fact because if those stones were not ruled dead, then it would not be seki, but Japanese rules clearly consider “double-ko-seki” a form of seki: If a double ko seki had a larger eye (with a prisoner inside), the score would vary depending on whether it is alive in seki or alive without seki, and the Japanese precedent is that it is counted as seki, no territory.

So back to example 6, the 4 white stones are alive, the black stones are all alive, but the single white stone is dead. That single dead stone means that the empty intersections are dame, not territory, and thus creates an antiseki and it is all ruled as 0 points if left as it is, since it is no longer true as the text mistakenly claims that “Black’s ten stones are therefore dead stones in territory surrounded by live white stones” (as the single white stone is needed to surround, and it is dead).

[cdavis.7m]

I noticed a misunderstanding on L19x19 which maybe you experienced at first, and I did too at first, before recognizing that the Example is simply making a particular point – the 1 stone is not important.

The examples in the Japanese are are not actually explaining how Life & Death works under the Japanese rules. Instead, the traditional Japanese rulings explain how Life & Death works. The “new” examples are merely explaining the consistency of the rationale underlying the traditional rulings. That these are simply rationales cannot be known from the English translations you linked because that translation is missing the “Introduction” section and the “Outline” appendix where the drafters of the new rules explain their goal – to theorize and clarify the rationality of the existing Japanese rules.

Similar to Example 6, there other examples in the Japanese Rules where a player has potential to play a different way and end up forcing a loop that does not show what the example is trying to show. For instance, in Example 11 (copied below), White 4 is a pass for the upper side ko, which clearly doesn’t work. But White 4 could be a pass for the lower side ko in which case it might seem that Black would need to pass for the lower side ko as well. But Example 11 is not trying to show some endless loop of passing and ko taking. It’s showing life and death status. This same issue comes up in other Examples. Maybe Example 18 if I remember correctly.

A similar issue is how the stones are analyzed. Even if the 1 white stone on its own could be proven to be dead, the 5 white stones (including the 4 living stones) are alive. I think this point comes up in a couple examples.

So yes, I agree with your Edit. White could show that the 1 stone is alive and the fact that the corner dies in that example does not actually show that the 4 stones are dead because the L/D status of the 4 stones is not being analyzed, just the 1 stone. And of course, analyzing the 5 White stones together shows that they are alive.

Example 11:

[cdavis.7m]

I meant to include links to the Intro and the outline.

日本囲碁規約 序 | 棋戦 | 囲碁の日本棋院 :
爾来四十年両先輩の先見性は見事に実証されたわけですが、国際試合の急激な増加と共に日本囲碁規約の合理化、理論的明確化の要請が関係者の間で強く呼ばれるようになりました。

Ⅳ 日本囲碁規約改定の概要（ご参考） | 棋戦 | 囲碁の日本棋院 :

1. 改定の基本方針

• （１）日本の伝統的な対局方法を遜守。
• （２）日本ルールに潜在的に含まれている合理性の理論化と明確化。