Life In 19x19
http://www.lifein19x19.com/

J89's pass-for-ko: Misinterpreted in the Western Go World?
http://www.lifein19x19.com/viewtopic.php?f=45&t=18350
Page 6 of 6

Author:  jann [ Sat Sep 18, 2021 10:23 am ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

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

Quote:
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).

Author:  CDavis7M [ Tue Oct 19, 2021 6:57 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

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

Quote:
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).

Author:  jann [ Tue Oct 19, 2021 7:46 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

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).

Author:  CDavis7M [ Tue Oct 19, 2021 8:40 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

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.

Author:  santo [ Mon Nov 08, 2021 8:32 am ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

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.

Author:  santo [ Mon Nov 29, 2021 4:57 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

I have not seen it mentioned in the thread (but maybe I missed it, I did not read everything completely in depth) but any "global enable" rule must include some very tricky definition of what intersection-controlling / stone placements actually count as "new stones ENABLED BY THE CAPTURE". Just "Any sort of new stone at all" cannot do because otherwise, the presence of any one-sided-dame would make all groups of the player immediately alive, which is of course not the interpretation. Even if the specific 1989 model of the Japanese tradition does not explicitly mention locality in the enabling or ko rules, it is a very firm and established principle that Japanese rules actually HAVE locality of life as a principle: any attempt of "global rule" must be very carefully written such that even though it has some global definition, the consequences of one area never "leak" to far away, independent regions (that is, regions separated by pass-alive groups). And based on precedent and universal practice we can be quite sure that if any such a consequence ever emerges, it is a defect in the rules.

https://senseis.xmp.net/?OneSidedDame

According to those I was able to consult, the Japanese stance about this would simply be that the one-sided-dame stones are "not enabled by the capture": the player could already play there to begin with, so there is no "relation" to the capture. It is tricky to find a precise definition of "enable" that correctly models this kind of "causality", and matches all precedents. I haven't been able to find a definition of "what counts as enabled" that at the same time matches all precedents for:

1) The seki in life and death example 2 in the Japanese rules

2) The 3 points without capture position

3) The three examples in this specific reply: https://forums.online-go.com/t/odd-case ... s/24109/64

4) The example in this specific reply: https://forums.online-go.com/t/odd-case ... s/24109/68

Author:  CDavis7M [ Wed Dec 01, 2021 10:58 am ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

santo wrote:
I have not seen it mentioned in the thread (but maybe I missed it, I did not read everything completely in depth) but any "global enable" rule must include some very tricky definition of what intersection-controlling / stone placements actually count as "new stones ENABLED BY THE CAPTURE". Just "Any sort of new stone at all" cannot do because otherwise, the presence of any one-sided-dame would make all groups of the player immediately alive, which is of course not the interpretation. Even if the specific 1989 model of the Japanese tradition does not explicitly mention locality in the enabling or ko rules, it is a very firm and established principle that Japanese rules actually HAVE locality of life as a principle: any attempt of "global rule" must be very carefully written such that even though it has some global definition, the consequences of one area never "leak" to far away, independent regions (that is, regions separated by pass-alive groups). And based on precedent and universal practice we can be quite sure that if any such a consequence ever emerges, it is a defect in the rules.
I just wanted to point out that this is not an issue under the Japanese Rules because such dame would be filled during/before confirming life and death status. So there is no possibility of any dame, including one-sided dame, being used to prove life. Teire are also played before confirming.

Author:  santo [ Wed Dec 01, 2021 12:46 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Is it MANDATORY to fill one sided dame? I did not know that.

Is that based on informal, intuitive understanding of what dame / one-sided-dame is?

Unless it was ruled mandatory, a player could avoid filling his one-sided dame, for it to be there on the board, and allowing it to count as "enabling" for other stones (this is the whole point basically, that such dames would never be allowed for the "enabling" rule by Japanese, no matter any other rules technicalities).

Interestingly, if filling dame were MANDATORY both for normal dame and for one-sided-dame, then such a "mandatory dame filling" rule would avoid "pseudosekis", like the one posted by Mef in https://lifein19x19.com/viewtopic.php?f ... 6&start=60

Author:  CDavis7M [ Wed Dec 01, 2021 2:00 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

santo wrote:
Is it MANDATORY to fill one sided dame? I did not know that.
Yes. Under Article 9-2 dame and teire are "hitsuyo" (required) plays. They must be played before the players confirm Life and Death. After all, dame must be played in order to differentiate between live stones without dame and seki stones that still have dame. There's no need to differentiate dame that one side can play.

And from looking at that, I learned another relevant word: "hitsuzetsu."

Author:  santo [ Wed Dec 01, 2021 2:33 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Wow, there are a few additional lines with clarifications in the official Japanese page! I have never seen a full translation for those. I mean, the only translated English text of the Japanese rules that I have seen is http://www.cs.cmu.edu/~wjh/go/rules/Japanese.html

This seems to be a direct translation of the original "main" text in https://www.nihonkiin.or.jp/match/kiyaku/zenbun.html

However, apart from that "main" text, on the side bar you can click and there is additional text! Is this additional text anywhere to be found in English, so that we can check it?

I understand that your insight that the rules specifically say that all dame and teire MUST be played is to be found here in the "detailed version" of rule 9: https://www.nihonkiin.or.jp/match/kiyaku/kiyaku09.html . There at least it seems that this word you mention, "hitsuyo", appears (必要).

I can't read Japanese, so having a translation for this extra official text that I was not even aware existed would be wonderful, but I am afraid maybe no one has done so yet?

Author:  santo [ Wed Dec 01, 2021 2:46 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

This is huge for interpretation of the rules, because then if the final position that players leave on the board is not really that "final", as it is not if it can be argued that Teire, dames and all other "necessary protection" moves must be filled before life and death determination starts (even if done by the referee and not the players themselves, at least thinking in a western amateur tournament with a dispute, where "someone" has to apply the "Japanese rule" and the referee is the only one that makes sense once the players are in dispute), then some situations where those "necessary moves" cannot be played during the game stop being a problem.

For example, pseudosekis like the one I linked disappear: The referee fills whites one-sided dames before life and death determination, thus the position is not seki, even though white would prefer to avoid filling the dames to make seki.

Similarly, a case for the mannen ko + infinite ko threats discussion in https://www.lifein19x19.com/viewtopic.php?f=45&t=18396 can be made: depending on the exact definition of "teire" or "missing protective moves", the arbiter might say "ok, they could not fill this ko during the game because one player 'abused' their infinite ko threats to disallow the other player from filling and completing this moonshine-life-like-ko, but since this is still a missing move to make the clean seki and complete the one-sided position, I am going to play it out before then applying life and death".

I mean, the fact that strict life and death analysis as per the rules would not start *exactly* from the positions the players actually left on the board was not an option to me. But once that is not the case and extra moves that the players did not play can be added, because "dame filling and defenses are mandatory before applying life and death", a whole new world of possibilities open. Of course it is tricky because there is no exact definition of what is a "protective play"... but given moonshine life and this dame-precedents, I would assume a referee would complete the mannen ko even in the presence of infinite ko threats, as part of this "mandatory filling" and according to most precedents that locally judge the mannen ko as if the seki is played out.

Author:  santo [ Wed Dec 01, 2021 2:57 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Sorry, I just found that in the rules I linked, there is extra text commentary between the "main text" and the "examples and precedents". So the full text is available it seems, and there it says about dame and teire. I must have read that once many years ago, and then completely forgot and only checked the "main" text again.

Author:  CDavis7M [ Wed Dec 01, 2021 5:00 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

santo wrote:
...http://www.cs.cmu.edu/~wjh/go/rules/Japanese.html... the rules specifically say that all dame and teire MUST be played is to be found here in the "detailed version" of rule 9: https://www.nihonkiin.or.jp/match/kiyaku/kiyaku09.html. There at least it seems that this word you mention, "hitsuyo", appears (必要).
Yes, that is the page I was looking at. I haven't actually read the English translation. I see you found the additional text in the English translation but you might be interested to know that at least one example in the online Japanese Rules was changed compared to the older English translation (it was Example 18, also mentioned below). This was discussed somewhere here recently. I haven't looked myself. Other things might be different as well.

Two parts of the Japanese Rules, which are not in the English translation you linked, that I found enlightening are the Preample by the Committee and the Summary of the Revision. I think failure to read these portions has led to some misconceptions.

santo wrote:
I can't read Japanese, so having a translation for this extra official text that I was not even aware existed would be wonderful, but I am afraid maybe no one has done so yet?
You can search here and see that we almost had a well-known and respected translator work to translate the rules as the Japanese would read them. But we are too ornery to have nice things. As for myself, I am learning Japanese and so I use computer translations and dictionaries (jisho.org and Kenskusha's Jp-En dictionary).

santo wrote:
For example, pseudosekis like the one I linked disappear: The referee fills whites one-sided dames before life and death determination, thus the position is not seki, even though white would prefer to avoid filling the dames to make seki.
I don't see any issue with Mef's pseudoseki from the other post. My understanding of the Article 9-2 explanation is that the play of dame and teire are not actually moves in the game. So it's not like one player could demand to resume from the not-the-actual-game position after dame and teire are played. The real issue in the pseudoseki seems to be that White wants Black to make a mistake. Maybe a referee would be needed if the pseudoseki White player were obstinate.

santo wrote:
I mean, the fact that strict life and death analysis as per the rules would not start *exactly* from the positions the players actually left on the board was not an option to me.
My understanding (some disagree) is that not only does life and death analysis not "start from the position the players actually left on the board," but that there is no "starting" from any position at all. I don't see any basis for so-called "hypothetical play" actually happening during Life & Death confirmation under the Japanese rules. In the Preamble and Summary the Committee states that the revisions create rules having a consistent rationale to appease Westerners and that they have dismissed inconsistent rulings (Torazu sanmoku - “three points without capturing”). The examples of play to show L&D are simply explaining how the rulings have consistency from the same rationale (re-taking a ko requires a pass). There is no suggestion that this type of play would actually take place during L&D confirmation. See Example 18 where black's upper-left group is deemed dead (Bent 4) and so there is a seki-collapse. At this point in the Examples, Bent-4 has already been defined as being dead according to Example 7-1, and a seki-collapse position has already been defired, so there is no need to "hypothetically" play it out. Furthermore, if players found a position like Example 18 during L&D Confirmation and they actually tried to "hypothetically" play it out, they might endlessly take and retake the double ko between passes without getting around to creating the bent 4. But this is not how L&D confirmation works. Note that the L&D examples have a particular play-style/goal. They are not playing to "win" (e.g., endlessly retaking a ko, or recreating a position that cannot be killed), they are playing to show whether groups of stones can be captured or not (e.g., Example 13). So anyway, I think discussion of the Japanese Rules based on "hypothetical play" is misleading.
Image

Author:  jann [ Wed Dec 01, 2021 7:05 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

CDavis7M wrote:
santo wrote:
Is it MANDATORY to fill one sided dame? I did not know that.
Yes. Under Article 9-2 dame and teire are "hitsuyo" (required) plays. They must be played before the players confirm Life and Death. After all, dame must be played in order to differentiate between live stones without dame and seki stones that still have dame. There's no need to differentiate dame that one side can play.

One sided dame usually occurs in sekis to begin with, so filling them doesn't change anything.

Dame fill itself cannot be mandatory since it could be suicidal in sekis, and even with normal dame there are cases where the color of the stone filling it could affect the outcome. So dame fill can only be based on the players' decisions - and dame that neither side wants to fill signals seki.

This is complicated a bit further with the rules allowing "informal dame fill", ie. after normal game moves - but only on agreement from both players (this is emphasized in the commentary). Dame fill is always optional, but often in the interest of at least one player ("filling in the dame to confirm territory", again mentioned in the commentary).

Author:  santo [ Wed Dec 01, 2021 7:21 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

"Filling dame" in this context clearly means filling those dames that are not suicidal, those of the normal groups that can be filled without dying because of the liberty reduction.

It is obviously a highly "informal" and "based on intuition" thing, not formal. It is mandatory to fill "those dame that are not the one in sekis", in order to avoid "false" sekis. Otherwise the phrase that it is mandatory in the rules makes no sense.

"One sided dame usually occurs in sekis to begin with, so filling them doesn't change anything." , they do in the pseudoseki example linked above. My understanding is that the Japanese would never rule that as a seki (just like Katago 3-phase rules), but since only one player can fill during the actual game due to ko threats (filling by the other player enlarges the ko, making the threats not big enough), it is in the interest of the player that can fill dame to not do so, in order to "make it look like seki" and win (because he loses less points than the opponent if everything is declared seki even though the groups clearly have two eyes). This is solved again if we understand that the rule is "no, for life and dead you must fill those dame, even though they are one-sided-dame and it is not in your interest to do so because you prefer seki". At least, that seems to fit better with the way that the Japanese play.

"there are cases where the color of the stone filling it could affect the outcome", the pseudoseki is one such case, but I am calling it a "one sided dame" and not a normal dame precisely because of that: because one player could fill anytime, with the only consequence being "now there is no dame, so no chance to call all of our clearly-two-eyed-groups seki" while the other has negative consequences of filling immediately during the main game. Do you have any other example where "who fills a normal dame changes the result"? I would probably not call those "normal dame", more like an unresolved position, or a one-sided dame, depending on the position.

Author:  jann [ Wed Dec 01, 2021 7:41 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

santo wrote:
It is mandatory to fill "those dame that are not the one in sekis", in order to avoid "false" sekis. Otherwise the phrase that it is mandatory in the rules makes no sense.

See the commentary (on article 8 for example). It speaks of filling in dame to make eyes territory as an optional act. Any position can be left with dame and thus as seki (there are no specific rules for "true seki" to begin with, it is up to the players' intentions). But filling dame is often gainful (to switch from seki to territory). This can be done in the game or after it (the latter only on agreement).

Quote:
Do you have any other example where "who fills a normal dame changes the result"?

I recall a few examples where enlarging strings matter, but nothing at hand.

Author:  santo [ Wed Dec 01, 2021 7:43 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Yeah, "Enlarging strings matter" is exactly what happens in the pseudoseki example linked before that I described as an example of "one sided dame".

Interesting indeed.

Author:  jann [ Wed Dec 01, 2021 7:46 pm ]
Post subject:  Re: J89's pass-for-ko: Misinterpreted in the Western Go Worl

Iirc there also were examples where neither wanted to enlarge/fill, so not necessarily one-sided.

Page 6 of 6 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/