Yesterday's rule dispute in Korea

[Bill Spight]

Now, I watched the broadcast of the game and a few other comments on this game.

They are all saying "White does not need to reinforce because it is a one move approach ko". Well, the conclusion is correct, but do they really know why? I feel like seeing a student who solves a math problem using a given formula, but does not know why the formula is valid.

Well, players knew instinctively about double ko seki for centuries if not millenia, but it was only in 1998 that I discovered a theory that could prove it. As for this being a final position in an approach ko, Professor Berlekamp proved that with komaster theory around 30 years ago. It is too bad that his theory has not been adopted (yet) by baduk professionals.

[Bill Spight]

Is it just me or do almost all of these rules disputes boil down to: "this would not have happened with area scoring"?

Oh, and the problem is not territory scoring.

Well, territory scoring has a natural stopping point when there may be ko fights in later play. How to deal with that is a real question.

The problem is the unwritten tradition that considers life and death to be determined with both white to play and black to play from any position, and with all ko bans lifted.

Well, that is one viewpoint (and one I share) but it is not the only traditional viewpoint.

If the tradition was just to consider what could be captured if the game would have went on, it would be completely different.
But that would imply that you have one point of territory inside the last ko if you avoid filling it and if your opponent can't win the ko. No one is ready to accept that.

Well, both Honinbo Shusai and Go Seigen accepted that.

[RobertJasiek]

Apart from the usual specific mistakes of the Korean 2016 Rules, they show an unprofessional attitude. From ruleset to ruleset, they substitute some beginner mistakes by other beginner mistakes. They hope to solve rules by trial and error but this does not work as long as well known mistakes recur. Korean professional players write rules like a beginner plays the game.

Hopefully, they learn from your book, after which they can no longer pretend a language barrier.

[yoyoma]

The videos are on BadukTV's youtube channel. First one shows the prelude -- Black plays a ko threat, but white ignores it. The leads to the difficult fight in the upper left corner.
https://www.youtube.com/watch?v=O0LlIyE0RuQ

Here you can see the dispute from 2:30 to 4:30. BTW the referee who shows up is Kim Hyunghwan 8p (김형환), who will be attending the US Go Congress.
https://www.youtube.com/watch?v=OMyGhjArMIY

Also at :45 white plays atari on 4 stones instead of taking the ko (the commentator was expecting white to take the ko). It takes him 10 seconds to realize and start explaining that white will still win because white is ko master and won't have to fill the ko.

Another fun point: the players have so many prisoners that the final score is black 2, white -4. After 6.5 komi, white wins by 0.5.

[jaeup]

Hopefully, they learn from your book, after which they can no longer pretend a language barrier.

Language is not the biggest problem, I agree, but it is at least a big problem. When I asked KBA if they have Korean translation of the Japanese and Chinese rule, they simply said no. Well.. I decided to make my own translation, but I do not think they really read what I gave them.

Foreign players frequently play tournament under Korean rule, and KBA never distributes them the Korean rule. (They do not even have Japanese/Chinese/English version of the Korean rule.) To be fair, Korean players visiting Japan or China does not get a piece of paper explaining the ruleset they must play with.

I assume they learn other rules by word of mouth, well... in 21st century.

[jaeup]

Capturable is not defined, however. My guess is that alternating play is assumed, and also that stones that would be captured in a fight that belong to the winner of the fight are not considered capturable.

This is all they have: "All uncapturable stones are alive and all capturable stones are dead".

What you are thinking is correct. They are asking for the complete control of the territory you are claiming. i.e. capture all the opponent's stones in the hypothetical play. Your stone being temporarily captured in that process does not matter. The final owner of the region matters. Of course, none of these are written in the rule text. You must guess everything from the above one sentence.

The Japanese pass for ko rule is, IMHO, one of the abominations of the Japanese 1989 rules.

The Japanese rule pretends that you can play anywhere on the board during the hypothetical play, which made the special pass rule inevitable. But, now we know that "playing anywhere" is not really a good idea because there are ways to abuse such a rule. Once we decide to restrict the playable area, the special pass rule is not mandatory. Its removal is probably the right direction.

And the anti-seki is an even worse abomination. The J89 rules are too clever by half. Good for the Koreans for not following suit.

Well, I think it is probably impossible to eliminate "dead stones vs dead stones" situation by slightly modifying the hypothetical play rule. It happens in both Japanese and Korean rules. The ruleset must say something when it occurs whether one likes it or not, and treating it effectively like a seki is a reasonable conclusion. (Though, technically, it is an anti-seki not seki)

Another good feature, IMO, is that a pass is not a move, and that players are not supposed to pass until necessary, filling dame during play.

I do not agree that it is a good feature, especially for a territory scoring rule. You may want to make a pass just because you ran out of ko threats and no dame is left to be filled. The rulemakers of Korea hinted that it is the situation that "a pass is necessary", and they are likely to accept a pass in that situation. Not surprisingly, they never formally defined the term "necessary".

My preference is to introduce two types of passes, one for the ko capture and one for the game end. Some clever thoughts must be added to prevent abusing such a rule. The last chapter of my book introduces such a ruleset. (I will introduce it again when the book translation is finished. For now, only the rule text is available in English without the commentary.)

[RobertJasiek]

Instead of complicating the rules by introducing special rules (such as different types of passes), the aim should be to have the same rules for regular play and playout. It is, however, impossible for territory scoring without pass-fights: they closest I have come up with is the Simplified Japanese Rules: http://home.snafu.de/jasiek/sj.html Nevertheless, further complications should be avoided.

[Bill Spight]

Capturable is not defined, however. My guess is that alternating play is assumed, and also that stones that would be captured in a fight that belong to the winner of the fight are not considered capturable.

This is all they have: "All uncapturable stones are alive and all capturable stones are dead".

What you are thinking is correct. They are asking for the complete control of the territory you are claiming. i.e. capture all the opponent's stones in the hypothetical play. Your stone being temporarily captured in that process does not matter. The final owner of the region matters. Of course, none of these are written in the rule text. You must guess everything from the above one sentence.

You don't have to just guess. You can use your experience with life and death in play and problems. One reason, I think, that Western rules tend to be clear is that they are written for people with no experience.

The Japanese pass for ko rule is, IMHO, one of the abominations of the Japanese 1989 rules.

The Japanese rule pretends that you can play anywhere on the board during the hypothetical play, which made the special pass rule inevitable. But, now we know that "playing anywhere" is not really a good idea because there are ways to abuse such a rule. Once we decide to restrict the playable area, the special pass rule is not mandatory. Its removal is probably the right direction.

It isn't just that localization is not easy to define, it is that sometimes it is nearly impossible. It also changes the nature of the game. That is, kos, by their nature break localization. Why should kos in hypothetical play adhere to localization. (I know why that decision has been made, but it is not necessary. You can have an encore in which ko threats may be removed. Hypothetical play is especially problematic for inexperienced players.)

And the anti-seki is an even worse abomination. The J89 rules are too clever by half. Good for the Koreans for not following suit.

Well, I think it is probably impossible to eliminate "dead stones vs dead stones" situation by slightly modifying the hypothetical play rule. It happens in both Japanese and Korean rules. The ruleset must say something when it occurs whether one likes it or not, and treating it effectively like a seki is a reasonable conclusion. (Though, technically, it is an anti-seki not seki)

Again, it is something that alters the nature of the game, and drastically so. Perhaps it arose as a clever way to enforce not passing unnecessarily without saying so. But it has had unintended consequences. (At least I hope they were unintended.)

Another good feature, IMO, is that a pass is not a move, and that players are not supposed to pass until necessary, filling dame during play.

I do not agree that it is a good feature, especially for a territory scoring rule. You may want to make a pass just because you ran out of ko threats and no dame is left to be filled. The rulemakers of Korea hinted that it is the situation that "a pass is necessary", and they are likely to accept a pass in that situation. Not surprisingly, they never formally defined the term "necessary".

Actually, there is a deep connection between no pass baduk and territory, in that some concept of territory emerges from not passing, although these concepts may differ depending upon the form of no pass baduk. (See viewtopic.php?p=197451#p197451 ) If only necessary passes are allowed, a pass is necessary if a play would reduce one's territory, or if there is no legal play.

IMX, some rules beasts arise because someone made an unnecessary pass. Eliminate unnecessary passes and you eliminate those beasts.

My preference is to introduce two types of passes, one for the ko capture and one for the game end. Some clever thoughts must be added to prevent abusing such a rule.

IMHO, rules cleverness is not a plus. And one example is using passes to end play. Instead of saying, "Please keep playing," say something like "I don't have a play," or "Shall we stop?" That signals that you are ready to end play. OTOH, you might say, "I can't take the ko," in which case you are not proposing to end play.

OC, if passes do not lift ko or superko bans, then you can get moonshine life positions. There are non-clever ways to handle that problem, such as Yasunaga's three pass rule or Ing's four pass rule.

Are you familiar with Button Go ( https://senseis.xmp.net/?ButtonGo ). It is possible to implement button go using the first pass for the button, in which case it is necessary to have the first pass life ko bans but not have any effect on ending play.

[RobertJasiek]

Bill, since you know the Korean reason for local hypothetical play, tell us! (I can guess but I do not know.)

Mandatory perfect play on passing is a nightmare for clarification but the real problem is: go is competition of mistakes. Why allow mistakes on plays but prohibit mistakes on passes? Inconsistent! Oh, I get the professional intention: they want to move the first moment of perfect play clarification to the earliest possible moment, and, because perfect play is the easiest aspect of go;) , start alleged relaxation as early as possible.

[Bill Spight]

Bill, since you know the Korean reason for local hypothetical play, tell us! (I can guess but I do not know.)

I simply meant that the guesses I made were educated by experience. They were not just guesses.

Mandatory perfect play on passing is a nightmare for clarification but the real problem is: go is competition of mistakes. Why allow mistakes on plays but prohibit mistakes on passes?

You can do that if you do not overload passing by using passes to end play. As I said, using passes to end play is too clever, IMO. Letting play end prematurely causes problems that require complicated rules.

[jaeup]

IMHO, rules cleverness is not a plus. And one example is using passes to end play. Instead of saying, "Please keep playing," say something like "I don't have a play," or "Shall we stop?" That signals that you are ready to end play. OTOH, you might say, "I can't take the ko," in which case you are not proposing to end play.

OC, if passes do not lift ko or superko bans, then you can get moonshine life positions. There are non-clever ways to handle that problem, such as Yasunaga's three pass rule or Ing's four pass rule.

Are you familiar with Button Go ( https://senseis.xmp.net/?ButtonGo ). It is possible to implement button go using the first pass for the button, in which case it is necessary to have the first pass life ko bans but not have any effect on ending play.

I was mentioning cleverness of the rulemaker, not the player. I know players are quite ignorant to the actual rule, so when I devise a rule, I try hard for the players to play normally while the rule takes care of all the possible anomalies and trolling that a wicked rule theoretician may imagine and try.

I know lifting ko ban after two passes is the source of many anomalies. Someday I will explain how a rule can be designed to avoid them. Yeah, that explanation took a whole book of 360 pages (with some exaggeration).

[Bill Spight]

IMHO, rules cleverness is not a plus. And one example is using passes to end play. Instead of saying, "Please keep playing," say something like "I don't have a play," or "Shall we stop?" That signals that you are ready to end play. OTOH, you might say, "I can't take the ko," in which case you are not proposing to end play.

OC, if passes do not lift ko or superko bans, then you can get moonshine life positions. There are non-clever ways to handle that problem, such as Yasunaga's three pass rule or Ing's four pass rule.

Are you familiar with Button Go ( https://senseis.xmp.net/?ButtonGo ). It is possible to implement button go using the first pass for the button, in which case it is necessary to have the first pass life ko bans but not have any effect on ending play.

I was mentioning cleverness of the rulemaker, not the player. I know players are quite ignorant to the actual rule, so when I devise a rule, I try hard for the players to play normally while the rule takes care of all the possible anomalies and trolling that a wicked rule theoretician may imagine and try.

I appreciate that. But I agree with Ikeda that chasing anomalies is like playing Whackamole. (OC, Whackamole came after Ikeda. ) Ing's 1975 rules were simple and clear. One problem was that the superko rule could tax human ability to recognize a long cycle. But some superkos produced what Ing considered to be anomalies. He cleverly dealt with those anomalies and produced difficult to understand rules. The Japanese also produced difficult rules, J89, to deal with anomalies. Whackamole!

I know lifting ko ban after two passes is the source of many anomalies. Someday I will explain how a rule can be designed to avoid them. Yeah, that explanation took a whole book of 360 pages (with some exaggeration).

Both Yasunaga and Ing dealt with the issue, and I have also addressed it in various ways. For one simple way see https://senseis.xmp.net/?SpightRules .

The oldest known rules question is that of Moonshine Life. If you think in terms of life and death, it is anomalous in that a position with an obvious false eye can live because of a ko ban. But both Shusai and Go Seigen accepted ending play with a ko ban in effect, and even the AGA rules allow play to end in a kind of Moonshine Life position, because of the two pass rule. IMO there is a problem with overloading the pass. You seem to have a similar view.

The Japanese 1949 rules did not go along with Shusai's and Go Seigen's view about ending play. Instead they did not allow a ko banned position to remain on the board. That eliminated Moonshine Life positions. Ing rules stop play after two passes, but allow resumption without a ko ban. That also eliminates Moonshine Life positions.

From the more general standpoint of evaluation, which includes questions of life and death, a position with a score should maintain that score in subsequent play, regardless of who plays first. Since a ko ban prevents a player from playing first from the banned position, there should be no ko bans in the final position of a game. That principle eliminates Moonshine Life, as well as other, similar positions. From this viewpoint it is not the lifting of the ko ban that produces anomalies, but just the opposite: not lifting the ko ban.

The rationale for my simple rule, which I came up with in the 1990s, was this. If we allow passes to life ko and superko bans, we may get a repetition of the same position with the same player to play, with a ko ban in effect. In that case, if the player passed before, his opponent was unable, with no ko ban, to improve his score. Nor, OC, was the player who was forced to pass able to improve his score. So we may accept the current score as final.

OC, this may run afoul of the scoring rules, as distinct from the playing rules, so we may continue play to satisfy those rules in an encore. For instance, for scoring purposes it may be desirable to capture all dead stones, so play may continue for that purpose. Doing so eliminates any problems with dead stones facing dead stones, for instance. Edit: It may also be desirable, for scoring purposes, to fill all one way dame in sekis, so that they will not be counted as territory under Japanese or Korean rules.

[Pio2001]

Hi,
In my opinion, the rules of go should fulfill three conditions :

1-Novice people who read them should be able to play a game and score it without help of a more experienced player.

2-The rules of play should be translated and published by national federations worldwide, and referees should be able to use these translations and score games in any tournament without dispute.

3-The rules of go should be programmable, so that we can have software playing go.

And I would also add a wish of my own : there should be only one ruleset in the world, so that we don't have to worry under which rules we are playing.

Today, it is not possible to sell a set of stones and a board with the rules of play in game shops.
Nor is it possible for people interested to read the rules in Wikipedia and try by themselves.
At work, some colleagues told me that they have tried to play go by themselves, but gave up because they could not understand the rules.
And we can see from time to time novice people coming in this forum, asking why their software is claiming to have won a game while they think the opposite. For one people registering and asking questions here, how many just uninstall the app and try another game ?

If I look at official rulesets in the world, I can see that progressively, federations are giving up the Japanese rule and are adopting the AGA rule instead.
I think that an ideal ruleset should look like the AGA rule, but with the Chinese way of dealing with superkos : prevent the repetition of a position for cycles of two moves, and have the referee decide what to do for longer repetitions. And for softwares, prevent any positional repetition.

[Bill Spight]

Hi,
In my opinion, the rules of go should fulfill three conditions :

1-Novice people who read them should be able to play a game and score it without help of a more experienced player.

2-The rules of play should be translated and published by national federations worldwide, and referees should be able to use these translations and score games in any tournament without dispute.

3-The rules of go should be programmable, so that we can have software playing go.

And I would also add a wish of my own : there should be only one ruleset in the world, so that we don't have to worry under which rules we are playing.

Emphasis mine. In the West we need rules that raw beginners can understand, because we do not have a developed weiqi community like they have in the East.

If I look at official rulesets in the world, I can see that progressively, federations are giving up the Japanese rule and are adopting the AGA rule instead.
I think that an ideal ruleset should look like the AGA rule, but with the Chinese way of dealing with superkos : prevent the repetition of a position for cycles of two moves, and have the referee decide what to do for longer repetitions. And for softwares, prevent any positional repetition.

Quoting a recent exchange between Ed Lee and myself on This 'n' That:

Today I advocate Button Go, an intermediate form of go that reconciles territory and area scoring, and, IMHO, combines the best features of both.

Thanks, Bill. What is Button Go, and is it related in any way to AGA's passing-one-stone-per-pass ?

The simplest form of Button Go has a token called a button, which is worth ½ pt. by area scoring. At her or his turn a player may take the button instead of making a play on the board. Taking the button lifts a ko or superko ban, just as a board play does. Normally the button is taken after the last dame is filled and has the effect that it does not matter who gets the last dame.

AGA style Button Go:

You can implement Button Go with AGA pass stones by making the last pass special. If the player to make the last pass also made the first pass, she or he does not have to hand over a pass stone. White does not have to make the last pass. Because of that, we may consider that Button Go rule as a simplification of AGA rules.

See https://senseis.xmp.net/?ButtonGo

[Ferran]

Another fun point: the players have so many prisoners that the final score is black 2, white -4. After 6.5 komi, white wins by 0.5.

Sorry,

Is there a way to get the video for the full game or the SGF?

Thanks. Take care.