Ups

This results in the death of Black's group at the top (edited my posting accordingly in the meantime).

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

I am happy with your answer Cassandra because it is really how I would like to read the rule. But the point is that by reading carefully this rule I do not find clearly who will start the confirmation phase:

Article 7. Life and death
1. Stones are said to be "alive" if they cannot be captured by the opponent, or if capturing them would enable a new stone to be played that the opponent could not capture. Stones which are not alive are said to be "dead."

I agree that "if they cannot be captured by the opponent" may imply that the attaker will start the confirmation phase but it is not stated clearly. In addition the rule defines "alive" stones and then the rule define "dead" stones being stones which are not "alive". That may also imply that the defender may start the confirmation phase to prove her stones are alive (they cannot be capture by the opponent).

Let's consider a provocative example

Click Here To Show Diagram Code

[go]$$B Confirmation phase
$$ -------------
$$ | . a . X O |
$$ | X X X X O |
$$ | O O O O O |
$$ | . . . . . |
$$ | . . . . . |
$$ -------------[/go]

Assume a game ends in this position.

Black : you fail to play at "a". By starting the confirmation phase, I can prove my black stones cannot be captured => black stones are alive
White : you fail to play at "a". By starting the confirmation phase I can prove I can capture black stones => black stones are dead

Who is right according to the strict application of the rule? I prefer your view Cassandra (ie. white is right) but the rule does not say explicitly who (the defender or the attacker) will start the confirmation. A lot of examples are done in the rule but for each of them the defender has never a good move at her disposal. That means that the examples given does not really helps.
My example is only here to show that the rule does not explicitly explain how the confirmation phase should start.

Dear Gérard,

I would like to assume that this apparent ambiguity is due to the "known" English translation(s) of the Nihon Kiin 1989 Rules, which do not include each and every implicit aspect (or have taken such for self-evident). But it does NOT exist in Japanese!

The relevant fraction of the original text is
"相手方_の_着手_により_取られない_石, ..."

An English translation of the (by me) seperated Japanese "words" (word by word) might be
"Opponent_'s_move_by_not capturable_stone, ..."
and should be read from the back, resulting in
"A stone that cannot be captured by a move of the opponent, ..."

In Japanese, there is no explicit distinction between "stone" and "group".
An analogous transfer for "group" would be (according to my rudimentary understanding of Japanese)
"A group that cannot be captured by the opponent, even if he moves first, ..."

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

So both players passed while being aware that the position was unsettled and then called a referee to settle the matter?
I suppose this might be a case where a referee could declare the game a loss for both players.

This would be the case, only if both players refused to declare that the game has ended.
But this in not the case in Gérard's example.

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

Thank you Cassandra for this interesting translation.
BTW, do you know if a better translation has been made since this (quite old ) version pointed by the link http://www.cs.cmu.edu/~wjh/go/rules/Japanese.html ?

I don't understand. Both players passed, which implies they both declare the game ended. How can players ever pass and at the same time deny that the game has ended? That makes no sense (and it violates article 9).

To me, Gerard's example looks like a case where the players colluded to troll the referee, so declaring a loss for both seems appropriate.

Oops It looks my wording was incorrect. I wrote:


Obviously there is a contradiction saying "game ends in this position" and at the same time saying we "start the confirmation phase".
Probably I should have written "game stops in this position". Surely I am not an expert in wording and so I am not able to help you a lot.

No, I don't.

But I am afraid that there is none.

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

Article 9. End of the game
1. When a player passes his move and his opponent passes in succession, the game stops.
2. After stopping, the game ends through confirmation and agreement by the two players about the life and death of stones and territory. This is called "the end of the game."
3. If a player requests resumption of a stopped game, his opponent must oblige and has the right to play first.

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

Two passes only stop the game. Resumption is possible, but if requested J89 has the opponent start in it.

So with missed key move it is possible a player does not agree to end the game but (either player) cannot afford to request resumption either, hence the both lose hack (there may also be a logical problem here).

About translation, IIRC the current one is not bad as a translation (Davies), but was probably aimed at players not theorists. So instead of expecting rigid definitions of all details you need to use both common go sense and common English sense (similarly to the original Japanese likely).

A group that can be saved by a starting defensive move is not alive, the first move is what would make it alive - thus it wasn't alive (but unsettled) originally. And text says not alive => dead, NOT not dead => alive.

Oops I am a little lost Jann.
For a mathematical point of view
if "not A => B" then don't we have "not B => A" ?

The Nihon Kiin 1989 Rules contain a defition of "alive".
Everything that does not match this defition is "dead".

As a matter of course, everything that is "not dead", is "alive".

I assume that it might be possible to write a ruleset that contains a defition of "dead".
Then everything that does not match this defition is "alive".

And again, everything that is "not alive", is "dead".

Usually, the more positive first option is chosen by the authors.

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

Unsettled groups (C) could be treated alive (A) or dead (B) depending on the definition.

That is a good point Jann. We have to try to use common go sense to resolve difficult positions. Let's try. Assume a game stops in a position where playing a confirmation phase may be very difficult (look as the stupid example (sorry Cassandra) showed in https://lifein19x19.com/viewtopic.php?p=265943#p265943. That means that by playing this confirmation phase both players may very easily makes mistakes.

To simplify the discussion let's simulate such difficulties on a trivial example with trivial mistakes.

Click Here To Show Diagram Code

[go]$$W Confirmation phase
$$ -------------
$$ | . . . X O |
$$ | X X X X O |
$$ | O O O O O |
$$ | . . . . . |
$$ | . . . . . |
$$ -------------[/go]

We are in the confirmation phases (both players have passed and nobobody request a resumption of the game).

white : black stones are dead
black : I am not convinced so, please start a confirmation phase and prove I am not alive
white : OK

and it follows the sequence (with two mistakes)

Click Here To Show Diagram Code

[go]$$W Confirmation phase
$$ -------------
$$ | 1 3 2 X O |
$$ | X X X X O |
$$ | O O O O O |
$$ | . . . . . |
$$ | . . . . . |
$$ -------------[/go]

black: OK I am not alive with this sequence but may be I made a mistake. Because it is up to you to prove black is not alive I consider I can have another chance to try to live

here the second sequence

Click Here To Show Diagram Code

[go]$$W Confirmation phase
$$ -------------
$$ | 1 2 . X O |
$$ | X X X X O |
$$ | O O O O O |
$$ | . . . . . |
$$ | . . . . . |
$$ -------------[/go]

white : well you are alive but maybe I made also a mistake
black : yes OC, but you claimed black stones were dead and I ask you to prove and explain that point. If you fail that means that you are not really in a position to prove black stones are not alive => black stones are alive.

IOW : who is in charge of the proof? The attacker or the defender?
Maybe we can say both but in that case that could lead to endless discussions.
My preference is to say that the attacker is in charge of the proof. In that case, during the analysis, the attacker is not allowed to change one of her move while the defender is allowed to try several moves.

One idea to solve the problem.
The attacker starts the confirmation phase
Each player must play in say 30"
At each time the defender may change one of her move and resume the game but the number of changes is limited to say 3.
That way any player (even beginners) may give a result for their game within an acceptable time.

Some people think Japanese analysis is about theoretical perfect play. I doubt it is necessarily so, having the players play to their abilities could also work. But trying a (reasonably) few possible attacking and defending moves should fit in for both sides. Exact (time/attempt) limits are not defined - just like with (non-infinite) resumptions.

I support jann's opinion.

Go is a game of mutual agreement. If both sides are fine with their result, nothing can be wrong with it.
Remember that tango is only for TWO.

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

In tournaments, go is not a game of mutual agreement in the sense of collusion on an arbitrary result. E.g., fabrication of a jigo to share prize moneys resulted in penalty on two Russian 6d in a London Grand Prix. E.g., intentionally negotiating the winner between two Russians to let a third Russian win a French tournament on SOS resulted in penalty. J89 mutual agreement means agreeing on what perfect play analysis is also in the interest of fair tournament competition.

Dear Robert,

as I already mentioned earlier, there is NO "tournament" referenced in the Nihon Kiin 1989 Rules. You are overstretching things a bit.

"Mutual agreement" means an agreement between the two players, not with God.

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

J89 were written for pros and tournament games and contain a few tournament-like rules. I know what mutual agreement means in English but this is not the J89-intended meaning.