Snapback.

_________________
A good system naturally covers all corner cases without further effort.

In my opinion, "Uttegaeshi" is not possible with a one-stone chain.

Perhaps you have another kind of move in mind with "Snapback" ?

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

Your attempt to find a wording for a basic ko rule ("a player must not immediately recapture a stone, which just captured one of his stones") has a problem with its contents much more than its wording. Your text can be read as if prohibiting also snapback. If you do not want to discuss wording, then get the contents of your rules at least right so far that your wording does not imply wrong contents.

I am precisely aware of the distinction between "Play", "Evaluate", and "Count". However, keeping that distinction 100% means that during Play the players can never know whether their moves are good or bad because they know nothing about results. All their move decisions would be completely meaningless. At the very least, the resulting scores must cross the boundary between Play and Evaluate. Similarly the result "neither victory nor defeat" and its strategic comparison score value must cross that boundary. The players need to know the result of every game situation BEFORE they make their next moves. Think about why you do make a move in your game! Because you want to increase the score, which Evaluate will determine, in your favour! This is possible only if you do have access to what Evaluate will output as results. This is so regardless of whether you do not want to continue discussion of it.

I have understood your intention "If the termination conditions of case A and case B cannot be reached, case C will apply without saying.". So, during evaluation of a particular string, let us suppose the players create a cycle and continue it forever. You do not want your "neither victory nor defeat" rule to apply here. You want your evaluation conditions to apply. However, as long as the move-sequence goes on and on, analysis is not ready yet for applying these conditions. We are in a hole: the process of rules application does not terminate. Thus I maintain that you need an extra ko rule (like "During Evaluate, a cycle with more than 2 plays always invokes case C."). The rules shall be clear also for others - not only for you, right?

Since you do not clarify "more than 2 [eyes]" or "eye", let me state that your rules cannot be understood and applied WRT "2-eyed". Since this is a core concept of your rules, they are inapplicable as a whole.

The double threat connection problem is of course not a problem for your rules. It is just a counter-example to your earlier claim that considering one versus more than one string could not lead to differences.

About the no-3-eye-formation-example: It is possible that your thinking differs. But what is it! Why don't you explain it? Just because you do not want to explain "more than 2 eyes" clearly?

Examples might help a bit, but please don't spend too much time with them to do me a favour. What I am really interested in is clear definitions. Then examples might even be superfluous.

You have not given any definition of "eye". You have used the word in phrases like "more than 2 eyes" but using it as if it were already known what "eye" is is the contrary of a definition: It is not stating any.

Once more: Try to apply your rules (the rules!) to the 360 stones example! Writing prose instead is not an application of your rules but is circumventing it. Everybody can talk informally about verbal Japanese style rules, but such is not what your rules are - they are a new written ruleset and require specific, related application.

A positional example for inconsistency in your ideas? So far I have mentioned inconsistency between your ideas and Japanese style rules. To judge inconsistency within only your rules (apart from some ko rule being valid only during some phase etc.), I would need to understand your "2-eyed". Until you define it clearly, I cannot really say much about how (in)consistent that is for itself.

While you call it Japanese style if there is just territory except in sekis, I see further requirements before something can be called Japanese style. E.g., Korean style would also meet your weak criteria. Therefore I would speak of Traditional Territory Scoring style. - When one studies Japanese style rules more closely, then one notices also things mentioned earlier: http://home.snafu.de/jasiek/wagcmod.html

Actually, I think this is ok ... "a stone" is quite specific, and removes ambiguous possibilities of applying to a snapback. If one wanted to be more clearly specific, it would be easy to state "a single stone".

The problem with just "a stone" is that it could be part of a larger string. Therefore, if one does not want to refer to positions, using "a single stone" has been good practice for basic ko rule texts.

Your result of Life and Death example No. 4 is not acceptable.

Black can take either one or the other white group, but only by letting the other live. He can choose which one to take, but he cannot choose to take both. That this miai situation is suddenly broken down, assigning both areas completely to Black and producing a swing of 24 points, is like a cheap accounting trick, where one credit suddenly appears in two locations.

Even the Nihon-Kiin 1989 rules produce an acceptable result here.

_________________
A good system naturally covers all corner cases without further effort.

This result is a choice, in the same way that writing rules so that "bent four in the corner" always dies is a choice. In the case of bent four it may similarly be true that a group is captured without compensation, even though it would be impossible to do so during actual play (i.e. without special rules such as the J1989 "Pass for specific ko as the only valid threat" during evaluation).

You may not like it, but to call it "not acceptable" is too strong a phrase.

Dear Robert, here come the promised diagrams to explain some of my ideas to you.

For the sake of not interfering with anything you might have in mind when reading "eye", let us call points, which must not be taken by one player, because that move would leave a chain of him with no lifeline left, as "taboo".

Click Here To Show Diagram Code

[go]$$
$$ . . . . .
$$ . O O O .
$$ . O 1 O .
$$ . O O O .
$$ . . . . .[/go]

The move of 1 is not allowed for Black, ...

Click Here To Show Diagram Code

[go]$$
$$ . . . . .
$$ . O O O .
$$ . O M O .
$$ . O O O .
$$ . . . . .[/go]

... so the marked point is a "taboo-point" of White's chain.

"Evaluate" is done for each chain of stones separately, starting with a move of the opponent (may be a Pass).


Status "2-eyed" has three sub-statuses:

(I) The chain already has at least 2 "taboo-points". Black must not play at each of them. So the chain cannot be taken off the board. Just to give a help where to look, these "taboo-points" lay "inside" the chain, "inside" not to be taken literally, as the second example shows.

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |M W M W X .
$$ |W W W W X .
$$ |X X X X X .
$$ |. . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |----------
$$ |M W W X .
$$ |O M W X .
$$ |O O X X .
$$ |X X X . .
$$ |. . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |--------------------
$$ |M W M W . X . W X .
$$ |W W W W W W W W X .
$$ |X X X X X X X X X .
$$ |. . . . . . . . . .[/go]

(II) The chain has not 2 "taboo-points" yet. Whatever Black tries, the chain will be transformed into sub-status (I).

Click Here To Show Diagram Code

[go]$$
$$ |----------------
$$ |. . W . . W X .
$$ |W W W W W W X .
$$ |X X X X X X X .
$$ |. . . . . . . .[/go]

White's chain has no "taboo-points" yet.

Click Here To Show Diagram Code

[go]$$
$$ |----------------
$$ |M 3 O M 1 O X .
$$ |O O O O O O X .
$$ |X X X X X X X .
$$ |. . . . . . . .[/go]

Now the chain has 2 "taboo-points".

Click Here To Show Diagram Code

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

White's marked chain has no "taboo-points" yet.

Click Here To Show Diagram Code

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

Now the chain has 2 "taboo-points".

Chains of sub-status (II) cannot be taken off the board, too.

(III) The chain can be captured by the opponent. Thereafter, each of the primary points of this chain will be transformed into sub-status (I).

Click Here To Show Diagram Code

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

White's marked stone can be captured by Black.

Click Here To Show Diagram Code

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

Click Here To Show Diagram Code

[go]$$
$$ |--------------
$$ |X X O . O . O
$$ |X X O O O O O
$$ |X 2 X . . . .
$$ |O X X X X . .
$$ |O O O O X . .
$$ |. O . O X . .
$$ |O O O O X . .
$$ |. . . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |--------------
$$ |. . O . O . O
$$ |. 3 O O O O O
$$ |4 W X . . . .
$$ |O X X X X . .
$$ |O O O O X . .
$$ |M O M O X . .
$$ |O O O O X . .
$$ |. . . . . . .[/go]

Now White's chain has 2 "taboo-points".

Click Here To Show Diagram Code

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

White's marked chain can be captured.

Click Here To Show Diagram Code

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

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |2 3 X X O .
$$ |X X X X O .
$$ |O O O O O .
$$ |. O . O O .
$$ |O O O O O .
$$ |. . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |4 X X X O .
$$ |X X X X O .
$$ |O O O O O .
$$ |. O . O O .
$$ |O O O O O .
$$ |. . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |O 6 . . O .
$$ |8 . . . O .
$$ |O O O O O .
$$ |M O M O O .
$$ |O O O O O .
$$ |. . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |W W . . O .
$$ |O . . . O .
$$ |O O O O O .
$$ |M O M O O .
$$ |O O O O O .
$$ |. . . . . .[/go]

Each primary point now belongs to a chain with 2 "taboo-points".


Status "stable" has two sub-statuses:

(I) The chain does not have 2 "taboo-points", but will not be captured by the opponent.

Click Here To Show Diagram Code

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

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |O O O O X .
$$ |. . O O X X
$$ |. . . . O X
$$ |O O O O O X
$$ |. O . O . X
$$ |O O O O . X
$$ |. . . . . .[/go]

Black cannot capture White's chain.

Click Here To Show Diagram Code

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

White's marked chain ...

Click Here To Show Diagram Code

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

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |. X O O O X
$$ |X . O M O X
$$ |O O O O X X
$$ |X X X X X .
$$ |. . . . . .[/go]

... will not get 2 "taboo-points", so it remains in status "stable".

Click Here To Show Diagram Code

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

White's marked chain can be captured by Black.

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |X 4 . X . .
$$ |X 2 3 X X X
$$ |O X X X . .
$$ |O O O O . .
$$ |. O . O . .
$$ |O O O O . .
$$ |. . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |. O 5 X . .
$$ |6 O X X X X
$$ |O X X X . .
$$ |O O O O . .
$$ |. O . O . .
$$ |O O O O . .
$$ |. . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |M W B X . .
$$ |O W B X X X
$$ |O X X X . .
$$ |O O O O . .
$$ |M O M O . .
$$ |O O O O . .
$$ |. . . . . .[/go]

Only 2 of the primary 4 points are part of a group with at least 2 "taboo-points". So the chain cannot get the status "2-eyed". But will be "stable".


Status "removable" contains, what cannot reach neither status "2-eyed" nor "stable":

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |W W 1 X . .
$$ |X X X X . .
$$ |. X . X . .
$$ |X X X X . .
$$ |. . . . . .[/go]

White's marked chain can be captured by Black, ...

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |C C X X . .
$$ |X X X X . .
$$ |. X . X . .
$$ |X X X X . .
$$ |. . . . . .[/go]

... and none of its primary points will become part of a "2-eyed" White chain.

Click Here To Show Diagram Code

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

White's marked chain can be captured by Black ...

Click Here To Show Diagram Code

[go]$$
$$ |------------
$$ |. B X . X O
$$ |X . X O X O
$$ |X X X X O O
$$ |O O O O O .
$$ |. . . . . .[/go]

... and its primary point will not become part of a "2-eyed" White chain.

Click Here To Show Diagram Code

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

White's marked chain can be captured by Black ...

Click Here To Show Diagram Code

[go]$$
$$ |--------------
$$ |. O 5 O X O .
$$ |O O O X X O .
$$ |X O X . X O .
$$ |4 X X X X O .
$$ |X X O O O O .
$$ |O O O . . . .
$$ |. . . . . . .[/go]

Click Here To Show Diagram Code

[go]$$
$$ |--------------
$$ |. O X 8 X O .
$$ |O O O X X O .
$$ |7 O X . X O .
$$ |O X X X X O .
$$ |X X O O O O .
$$ |O O O . . . .
$$ |. . . . . . .[/go]

... and its primary point will not become part of a "2-eyed" White chain.

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

As I stressed several times, it is important to have consistent rules.

It is not important, if one likes the result of the evaluation of a very special situation (which may never arise in actual games). The result must fit into the chorus of all the other ones, provided by the rule set.

The result for 1989 Nihon Kiin rules example 4 fits into what I mean with the "local" view. There are no effects across living groups. Undoutably Black's tree-stone chain in the corner is a living one, so the "local" principle applies. And gives a valid result.

As Bent-Four-in-the-Corner is concerned, there are even examples within the 1989 Nihon Kiin rules, which end with the Bent-Four living in Seki or living unconditionally. There is no special rule claiming "Bent-Four-in-the-Corner" is dead. The "death" of "Bent-Four" in the overwhelming majority of cases results from the Nihon Kiin evaluation procedure alone.

The Nihon Kiin's specific rule of explicitely passing a particular Ko before retaking it is - in my opinion - a somewhat miscarried written trial to secure the evaluation being local.

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

Herman, bent-4 does not need a pass for ko rule. An ordinary basic ko rule during the analysis suffices!

I have done it already.

Let's have a second try.

It goes without saying the the 360 stone chain cannot get the status "2-eyed". There is only one point left that could be awarded.

There is no other chain on the board to evaluate.

Independently of our result of the evaluation of the 360 stones chain (which can only be "stable" or "removable"), there is no territory on the board (which can only be found within "2-eyed" chains).

Should the result of our evaluation be "removable" (what I do not think), there is nothing to be taken off the board, because the "removable" chain is not situation inside a "2-eyed" one.

During "Evaluate" you cannot repair what you might have missed during "Play".

If you forgot during "Play", for example, to capture a 50 stones string (that otherwise could have been connected to a living group), you will not get these 50 stones as prisoners during or after "Evaluate".

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

Locally, yes, but with unremovable ko threats a player can claim that it cannot be killed without compensation elsewhere. If you want bent-4 to be dead always, regardless of the global position, and without compensation elsewhere, then your rules most make provisions for it. It is a choice you can make when writing rules. Some rules makers prefer not to have bent-4 always dead, which is also a valid choice.

Cassandra, 2-eyed according to your examples does not need to be classified into subtypes and reference to no suicide or taboo intersections is also superfluous. It can be defined as followed:

A defender's string is "2-eyed" if he can force pass-alive stones on each of its intersections.

Note though that 2-eyed is a terribly inappropriate word and must be changed as soon as possible.

Using J2003 terms, your stable (I) could be generalized as follows, if that is your intention:

A defender's string is "stable-1" if it is not 2-eyed but uncapturable.

Using J2003 terms, your stable (II) could be generalized as follows, if that is your intention:

A defender's string is "stable-2" if it is not 2-eyed but capturable-1.

Assuming that these fit your intention, we can summarize:

A defender's string is "stable" if it is not 2-eyed but either uncapturable or capturable-1.

I am also not happy with the word stable though; it is used for too many other Go meanings.

A string is "removable" if it is neither 2-eyed nor stable.

I would like to see a proof whether the word is always justified... Until then another word is less risky.

One thing is very clear to me: Your rules text suggested something very different to me than what your recent examples suggest!

In your double ko seki sequence, why does it not end at move 4?! We already have two successive passes then! You should study sequences with single passes before the final passes though.

How do your rules apply to http://home.snafu.de/jasiek/j1989c.html Example 0000?

Concerning the 360 stones example and in view of your new implied definitions by the examples, your new failing attempt to apply your rules to the example does not need yet another attempt for the time being since the new implied definitions make the example much less interesting.

Herman, unremovable ko threats elsewhere do not change anything for J1989/J2003. To let them be effective, the unremovable ko threats must be "adjacent" to the bent-4 strings. The effect is only about points though - not about status clarification for the bent-4 strings. Before the game stop, it can make a difference. During analysis, unremovable ko threats of whichever kind do not affect the life and death status assessment of the bent-4 strings. So, no, a pass for ko rule(set) does not make bent-4 special. (If there is an adjacent n-tuple-ko, then it is not bent-4 any longer, except by fake-local shape. But J1989 do not use such fake-local-shape rules any longer. WAGC do, but they do not have pass-for-ko rules. So where is the sake-of-argument necessity for a pass-for-ko rule in case of bent-4?)

Concerning "acceptable": I find it unacceptable that the status of a group can change upon transition from "play" to "evaluation". If that could be the case, the "evaluation" would be a different game than the "play". Since the "play" is Go, "evaluation" would not be.

Yes, it is a choice, but it is a wrong choice. I am sorry for the explicit wording.

_________________
A good system naturally covers all corner cases without further effort.

Additional considerations on Cassandra's life: We have:

A defender's string is "2-eyed" if he can force pass-alive stones on each of its intersections.

A defender's string is "stable" if it is not 2-eyed but either uncapturable or capturable-1.

***

Proposition: Each 2-eyed string is uncapturable.

Proof: By definition, each 2-eyed string can be transformed into a pass-alive string. A pass-alive string implies being uncapturable. QED.

Corollary: That a string is either 2-eyed or stable implies that the string is either uncapturable or capturable-1.

Remark: Cassandra's life is almost the same as J2003-alive, except that Cassandra has yet to rediscover capturable-2 ;)

That's not really a surprise, isn't it ?


It had just been a matter of taste.


Intersting work of art.

Click Here To Show Diagram Code

[go]$$
$$ |-------------------
$$ |Q Q Q Q Q Q Q Q . |
$$ |Q Y Y Y Y Y Y Y Q |
$$ |Q Y . Y . Y Y Y Q |
$$ |Q Y Y Y P Y Y Y Q |
$$ |Q Y . P Z P . Y Q |
$$ |Q Y Y Y . Y Y Y Q |
$$ |Q Y Y Y . Y Y Y Q |
$$ |Q Y Y Y Y Y Y Y Q |
$$ |. Q Q Q Q Q Q Q Q |
$$ |-------------------[/go]

Triangled is "2-eyed", crossed is "removable".
Black has 1 point of territory, White 2 points.

Black should not have captured the two White stones.
Or played further on.
Either way, he will win by prisoners.

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

If you could hand me an example ?

It's too difficult to get through your text.

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

LOL.

Look at Example 0000 again: Tengen is capturable-2 under the Japanese 2003 Rules, i.e. alive.

And now look at Example 0000 again until you realize that your rules are a failure in their claim of being Japanese style! All really Japanese style rules must assess the (on-board-) score B+10. Else they are not Japanese style rules.

Look at Example 0000 yet again and realize why: Black is independently alive!