The "no ban in double ko" rule to solve pass-for-ko issue ?

[kvasir]

Btw - NOT as serious suggestion

I think we often don't use the best representations and it gets in the way of our understanding.

For me ko is the following:

1. a point (with a stone) where capture is banned if nothing else is captured at the same time, or no point.
2. a rule to update #1

So stating that the ko is not the same is meaningless.

One could easily extend to a list of points (i.e. j89 pass-ko) and double points (i.e. double hot stones ala Ing), if one can come up with a rule for #2.

Also superko is hardly representable as a ko, by this representation.

[CDavis7M]

$$B :b5: pass :b7: pass
$$ -----------------
$$ | 4 8 6 X . O . |
$$ | X O X X X O O |
$$ | O O O O X X O |
$$ | O O O 3 O X O |
$$ | X X O O X X O |
$$ | . X X O O O . |
$$ | X . X X X O O |
$$ -----------------

Click Here To Show Diagram Code

[go]$$B :b5: pass :b7: pass
$$ -----------------
$$ | 4 8 6 X . O . |
$$ | X O X X X O O |
$$ | O O O O X X O |
$$ | O O O 3 O X O |
$$ | X X O O X X O |
$$ | . X X O O O . |
$$ | X . X X X O O |
$$ -----------------[/go]

is allowed due to the "no ban in double ko" rule

The Japanese rules do not allow for White to play 6 after Black has passed at 5. If White does not pass (thereby stopping the game and beginning the confirmation stage) then the next action in the game must be a play by Black. A stone is only allowed to be played after the play of another stone (except when confirming life and death status, then a ko may be retaken by a player after they already passed for that ko).

If these example plays are being made after the game has been resumed (after it was stopped and life and death status was not agreed to), then this fact will indeed confirm that Black is dead.

[jann]

So stating that the ko is not the same is meaningless.

Why? J89 pass for (a particular) ko relies on that meaning, and in the quoted examples some recaptures may or may not be forbidden (without passing for them first) depending on this.

Extreme example: opponent captured me in a ko, then whole ko shape is destroyed, neighbouring stones also captured, then several moves later an identical ko shape reappears at the same intersection. Can I capture there without passing for it first? Is it still the "same ko" the opponent captured in earlier?

[jann]

A stone is only allowed to be played after the play of another stone

This idea is from the time when pass was not a move, and was ruled out decades ago, even before J89. In modern rules a pass is a play, and it is perfectly possible to play on the board even after the opponent passed (that's why only two passes stop).

Also, since J89 confirmation introduced pass for a ko, it cannot at the same time be claimed that a pass is not a play, or ko recapture would not be generally allowed after a pass (a pass is valid ko threat). Even in resumption, it is mentioned that a player may start the resumed game with a pass (potentially followed by a board play of the opponent).

BTW this old approach led to problems with mannenko (and was the root of some historical disputes).

[CDavis7M]

In modern rules a pass is a play, and it is perfectly possible to play on the board even after the opponent passed (that's why only two passes stop).

That's different from what I read on the Nihon Kiin website. Here are the versions I am looking at:
https://www.nihonkiin.or.jp/match/kiyaku/
https://www.nihonkiin.or.jp/match/kiyaku/zenbun.html

Thanks for pointing out the explanation to Article 9. Indeed the explanation (presumably given the same weight as rules) does allow for playing after a pass upon resumption of a stopped game. And of course, after the game has been stopped Article 7 allows recapturing a ko after a pass when confirming life and death. But I do not see any allowance for playing after a pass besides these exceptions. And I see nothing that allows a player to continue playing move after move while their opponent continues to pass again and again.

Article 2 only allows for playing one stone one after the other until both players abandon playing (pass). It does not allow playing a move after a pass. And the explanation of Article 2 specifically states abandoning play (pass) is a declaration to "stop the game" ("対局の停止"), which is covered in Article 9. How can it be that opponent can continue to play moves after the other player has requested proceeding with stopping the game under Article 9?

By the way, I also don't see anything suggesting that "a pass is valid ko threat." I only see the exception to the ko rule that allows for recapture of a ko after a pass when confirming life and death (in Article 7). Maybe a similar concept but "a pass is valid ko threat" seems like it's broader than the rule allow.

As another aside, my impression is that given definition of which stones are alive and which are not, confirmation of life and death does not depend on multiple passes being used.

[jann]

I also don't see anything suggesting that "a pass is valid ko threat." I only see the exception to the ko rule that allows for recapture of a ko after a pass when confirming life and death (in Article 7). Maybe a similar concept but "a pass is valid ko threat" seems like it's broader than the rule allow.

This is not spelled out explicitly, but would be incredible inconsistency if at one point a pass could be used as a ko threat (required even), while a few moves away it would not even allow ko recapture. Also note the comment that after a pass, "IF the opponent also passes in succession" - so this is only one of the possibilities.

You may be confusing Japanese and Korean rules - the latter may still be unclear on the role of passes, even today.

[CDavis7M]

This is not spelled out explicitly, but would be incredible inconsistency if at one point a pass could be used as a ko threat (required even), while a few moves away it would not even allow ko recapture.

I must be missing the background for why using a pass as a ko threat must be required. But I don't see how "a pass can be a ko threat" under the Japanese rules because the Japanese rules do not allow for a move to be played after a pass, except as discussed above, when first resuming the game or during confirmation of life and death. Article 2 does not allow for a move after a pass. It only allows for black and white to play one after the other.

Also note the comment that after a pass, "IF the opponent also passes in succession" - so this is only one of the possibilities.

Right. When Black passes, there is also the possibility that White does not pass. In this case, the game does not proceed to Article 9 of the Japanese Rules. "If the opponent also passes" is the condition for proceeding to Article 9.

When Black passes White may pass. But there is the situation where "the opponent does NOT also pass." When White does not pass in response to Black's pass, the only action in the game allowed in the Japanese rules (after a play by White) is a move by Black (Article 2), which can be abandoned (passed). If Black passes after White has not passed in response to Blacks pass, then Black must play or pass. There is no rule to compel the players to proceed with Article 9.

You may be confusing Japanese and Korean rules - the latter may still be unclear on the role of passes, even today.

Nope. I have not read the Korean rules.

[jann]

When White does not pass in response to Black's pass, the only action in the game allowed in the Japanese rules (after a play by White) is a move by Black (Article 2), which can be abandoned (passed). If Black passes after White has not passed in response to Blacks pass, then Black must play or pass.

B passes (as he did so before, since he doesn't want to play). W would like to play but is not allowed to. So the game enters a loop of "B pass, B pass, B pass"? This seems very wrong on several layers, hard to see how could you read all this into J89 text.

I won't repeat what I already wrote, but one more thing: there are cases/shapes where only one side has things to play (several moves even) while the other has to pass. So "B pass, W play, B pass, W play, B pass, W pass" can sometimes be the ONLY meaningful way to finish a game.

[kvasir]

So stating that the ko is not the same is meaningless.

Why? J89 pass for (a particular) ko relies on that meaning, and in the quoted examples some recaptures may or may not be forbidden (without passing for them first) depending on this.

Extreme example: opponent captured me in a ko, then whole ko shape is destroyed, neighbouring stones also captured, then several moves later an identical ko shape reappears at the same intersection. Can I capture there without passing for it first? Is it still the "same ko" the opponent captured in earlier?

I am not saying that you can't think of ko as the shape surrounding the forbidden move or the forbidden capture, just that all you need to track are the stones that are forbidden to capture. The surrounding shape and even the forbidden move are not needed. If you take that then as your representation of regular ko, then it doesn't make sense to talk about changing the shape. I can admit that I didn't state anything that ruled out checking the shape when updating the ko, so technically I allowed for anything there (which does not exactly follow my thinking). What I meant by it being "meaningless" is that nothing changes in the ko representation except what stones are allowed to be captured (which is not the change in the shape that you mean).

What I am talking about is that when you want to check if a move is legal by the regular ko rule you only need to compare the set of captured stones with the ko, if it is identical (and the ko is not empty) the move is forbidden.

When you say you want to talk about the ko changing, it means you want to talk about a set of points (possibly one set for each color) that if played clear that particular ko. Even if you can compute this set from the position at any time you have still introduced something new to the representation but I need to explain why. First, I did not state that you could compute anything from the current position (except I did implicitly state that you have the captures being made). Secondly, if I were to allow any computation when updating the ko it would allow you for example to create a ko rule that states that the "best" move in any position is forbidden -- that kind of update rule for the ko is not really what I am talking about, therefore restrictions have to be understood to apply even if not stated.


I would say that the j89 pass-ko is only different from regular ko in that now there is a set of points that can not be captured, and the update rule doesn't clear the ko when any move is made but only removes the point (the hot stone) that identifies the ko being passed for. So it is very similar to regular ko.

Extreme example: opponent captured me in a ko, then whole ko shape is destroyed, neighbouring stones also captured, then several moves later an identical ko shape reappears at the same intersection. Can I capture there without passing for it first? Is it still the "same ko" the opponent captured in earlier?

You can capture there because the ko was cleared when the hot stone was captured along with some other stones. It may look the same but there is no ko ban, basically the shape is irrelevant.

Of course I basically just tried to state a definition of what regular ko is (without stating every detail), if you adopt a different definition you may well read something into j89 that doesn't match my view

[jann]

I think even with your view you have to answer the question: in confirmation where pass-for-ko is required, does a "ko" or ko ban get cleared if the stones making the ko shape change.

This is the same as answering whether it is still the "same ko" or "that particular ko" that needs to be passed for, or a new ko that can be captured in freely (once basic ko hot stone cooled off). J89 itself is not 100% clear on this, though some interpretations more likely than others.

[kvasir]

I think even with your view you have to answer the question: in confirmation where pass-for-ko is required, does a "ko" or ko ban get cleared if the stones making the ko shape change.

This is the same as answering whether it is still the "same ko" or "that particular ko" that needs to be passed for, or a new ko that can be captured in freely (once basic ko hot stone cooled off). J89 itself is not 100% clear on this, though some interpretations more likely than others.

The ko ban is cleared when there is a pass-ko for that ko or the hot stones are captured. It is not necessary to handle the case of connecting the ko, but you may just as well state that the ko ban is cleared in this case too.

Agreed, that j89 is not 100% clear on this. If you have a different idea of what a regular ko is you may read something else into it.

But you have a point in that if one captures certain types of approach ko back and forth it may be disputable what the intention is.

For example

$$B :w4: pass for :w1:
$$ ---------------------
$$ | . . X X X X O X . .
$$ | X X O X O X O X . .
$$ | X O 1 O 3 O O X . .
$$ | O 2 O 5 O O X X . .
$$ | O . O O O X X . . .
$$ | O O O . O X X X . . . .
$$ | X X O O X . . . . . .
$$ | X X X X X

Click Here To Show Diagram Code

[go]$$B :w4: pass for :w1:
$$ ---------------------
$$ | . . X X X X O X . .
$$ | X X O X O X O X . .
$$ | X O 1 O 3 O O X . .
$$ | O 2 O 5 O O X X . .
$$ | O . O O O X X . . .
$$ | O O O . O X X X . . . .
$$ | X X O O X . . . . . .
$$ | X X X X X[/go]

I'd say that there is a pass-ko ban for and but not

[John Fairbairn]

1. The Preamble to the J89 rules tells us, and firmly stresses, that the intention was to adhere to the traditional way of play in Japan and that they are an attempt to rationalise and clarify the logic behind the Japanese way of play. There is no reference to mathematical rigour or to claim to be definitive rules (indeed they talk about being only a framework for new ideas). Therefore, any attempt to invent things like "pass is a valid ko threat" is doomed to fail on the grounds that it is not traditional. The same remark applies to several other whimsies here.

2. Several other good clues in the Preface and Preamble are ignored (though I've never seen these in English, so that may be the explanation).

3. The clarifications in the Nihon Ki-in rules booklet published in 2004 to mark its 80th anniversary likewise seem to be regularly ignored.

4. Also apparently ignored, are the implications of the 1999 Moriyama game with several passes (which may actually be an example of a confirmation phase in a pro game, though that wasn't mentioned in the commentary, and I suspect GoGoD may be the only place where you can see the full game), and the 2008 Kisei game with Cho Chikun in which a pass occurred in a dispute where player and referee disagreed, though J2003 was quoted rather than J89. [I am guessing the dates from memory.]

5. Japanese explanations don't use weird expression like pass ko bans or cycles. The one phrase that seems to crop up over an over again in justifying rulings is "stones other than those that alive are dead", which I think comes somewhere in the actual rules. I don't see that concept being applied here.

6. The word for 'same' is douitsu, i.e. 'one and the same' or 'identical', not 'similar'. IOW rely on the Japanese, not the English.

Following all these pointers (the Preamble above all), I'd say more focus on what Japanese practice was and what Japanese words mean would be beneficial, rather than trying to say what they SHOULD be and mean.

Otherwise, what comes to my mind are those cases on Youtube where people take in a stray puppy and rear it like a dog, only to end up surprised when it turns into a wolf, fox or bear.

[kvasir]

The only English translation says this in the preamble:

The Nihon Kiin and Kansai Kiin hereby revise the Nihon Kiin's Rules of Go formulated in October 1949 and establish the Japanese Rules of Go. These rules must be applied in a spirit of good sense and mutual trust between the players.

The part about "good sense" and also "mutual trust" is very important.

But we can also distinguish between what actual happens when we finish a game and what the Nihon Kiin rules actually say -- without pretending that the later is the right way because it is not.

I think the reason why we don't discuss much what actually happens are
1. There aren't many examples available in English.
2. Usually we do apply the "good sense" clause.
3. Most probably actual disputes aren't "rule lawyered" so some interesting / wild ideas are never heard -- provocative ideas on the other hand are often more interesting and generate much response.

Otherwise, what comes to my mind are those cases on Youtube where people take in a stray puppy and rear it like a dog, only to end up surprised when it turns into a wolf, fox or bear.

I haven't seen those videos but I once spent a great amount of time searching for footage of wolfs on youtube, almost all footage (excluding professional nature documentaries) was some lesser canine. The most beautiful golden jackals or cayotes (in America) somehow mistaken for a much bigger animal. Maybe it is the same with rule discussions, it is not really a big bad wolf, we are just exchanging ideas here.

[John Fairbairn]

The only English translation says this in the preamble:

The Japanese version is two pages long and (from memory) doesn't mention good sense and trust. My recollection is that they were put in a version for westerners because they knew what westerners get up to (as this forum proves).

But we can also distinguish between what actual happens when we finish a game and what the Nihon Kiin rules actually say -- without pretending that the later is the right way because it is not.

I don't actually understand this, but taking a stab at it, I repeat two points: (1) The NK rules are "right" in the sense of the (Japanese) Preamble, that they adhere to the Japanese traditional way of play, i.e. right for them and not what westerners think they should do; (2) they are indeed "wrong" in the sense that their own rules were not followed precisely in at least the two games I referenced, i.e. they can be confusing even for pros - though let it be noted the sky did not fall in when these incidents occurred.

Since almost all western players in practice follow what they think of as Japanese rules, it seems to be just as important to establish what the Japanese thinking on these is as it is to have new ideas or variant interpretations tossed about by western rules mavens.

[kvasir]

I suppose you refer to what I would call the "introduction" which is not part of the translation for some reason and I was referring to what I like to call the "preamble" which is the text immediately before the first numbered paragraph in the rules (which is just two sentences and is also there in the Japanese). Introduction and preamble are I think mostly synonymous, but I don't know what else to call the unnumbered paragraph that is right there before everything else.

I don't actually understand this, but taking a stab at it, I repeat two points: (1) The NK rules are "right" in the sense of the (Japanese) Preamble, that they adhere to the Japanese traditional way of play, i.e. right for them and not what westerners think they should do; (2) they are indeed "wrong" in the sense that their own rules were not followed precisely in at least the two games I referenced, i.e. they can be confusing even for pros - though let it be noted the sky did not fall in when these incidents occurred.

I think that is about what I meant. There is an understanding that defines the game we play, then the actual text sometimes fails to convey the same understanding. This is no surprise really and it is not necessarily a problem. I don't object to that it is more fruitful to actually understand the real use of these rules.

I'd certainly want to understand how people come to ignore somethings that follows from these rules while embracing the same formal semantics in other cases.