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

One more triple ko Draw
http://www.lifein19x19.com/viewtopic.php?f=45&t=17751
Page 1 of 1

Author:  jaeup [ Sat Aug 29, 2020 5:04 am ]
Post subject:  One more triple ko Draw

Attachment:
OhJeongWoo_vs_LeeChangSeok.png
OhJeongWoo_vs_LeeChangSeok.png [ 170.27 KiB | Viewed 9642 times ]


I just learned that one Samsung Cup preliminary game ended with a draw today. (Oh Jeong Woo_vs_Lee Chang Seok) I couldn't find the full sequence. The above figure shows only the last few moves.

First, it is not such an "exciting" triple ko. It is difficult to see that 1~11 are the best moves and there are no other choices. Second, even after 11, it is not so clear that following the triple ko sequence is inevitable. Anyway, they decided to do so and the game ended. A rematch is scheduled tomorrow. (A rematch is the most common resolution for a triple ko draw, but it is sometimes counted as 0.5 win in a team match.)

I've always wondered "exactly when does the draw occur?". When the whole board repeats once or twice?

As usual, the Korean rule text does not provide a clear answer to this question. (Japanese or Chinese rule is not any better for that matter.) The interpretation I heard is quite interesting: the draw occurs when 'the whole board repeats and both players are not willing to play otherwise'. Well, "intention" is a subjective term and I doubt if automatic rule application is possible.

Anyway, the lessen is that they want to wait until any hope to avoid infinite repetition disappears, then they call a draw. (In practice, after taking a few ko's, they realize that their opponents do not want to yield. Then, they look for referees to report the situation.)

Author:  jann [ Sat Aug 29, 2020 6:30 am ]
Post subject:  Re: One more triple ko Draw

This is why triple ko draw differs from superko (triple ko ban): it doesn't really matter exactly when you consider the draw to be finalized, so it is actually applicable in practice. I think the moves can in theory considered to continue forever, just no need to actually play them.

Author:  Harleqin [ Mon Aug 31, 2020 4:11 pm ]
Post subject:  Re: One more triple ko Draw

Technically, I'd not call it a draw if a re-match is scheduled. I'd say that it is a game without result. A draw would imply something like half a win for both.

Author:  Matti [ Mon Aug 31, 2020 11:40 pm ]
Post subject:  Re: One more triple ko Draw

Harleqin wrote:
Technically, I'd not call it a draw if a re-match is scheduled. I'd say that it is a game without result. A draw would imply something like half a win for both.

Technically, it is a draw. However, in this case the draw is not treated as a tie.

Author:  jann [ Tue Sep 01, 2020 6:16 am ]
Post subject:  Re: One more triple ko Draw

Scheduling re-match would also possible for score draws (with integer komi), but would this make them less draws?

Author:  Harleqin [ Tue Sep 01, 2020 3:56 pm ]
Post subject:  Re: One more triple ko Draw

Matti wrote:
Harleqin wrote:
Technically, I'd not call it a draw if a re-match is scheduled. I'd say that it is a game without result. A draw would imply something like half a win for both.

Technically, it is a draw. However, in this case the draw is not treated as a tie.


OK, thanks for the hint, I lookd up that word difference now. (What I understand now is that tie implies equal score (a jigo), while draw implies “ending without either side winning”.)

However, I understand long cycles to be seen as not ending the game in e. g. japanese rules. It's just that it's accepted that the game will never end, and never have a result, and so the players just do something else. Of course, the difference to a “real” draw might be purely philosophical.

Author:  John Fairbairn [ Wed Sep 02, 2020 1:19 am ]
Post subject:  Re: One more triple ko Draw

Quote:
OK, thanks for the hint, I lookd up that word difference now. (What I understand now is that tie implies equal score (a jigo), while draw implies “ending without either side winning”.)

However, I understand long cycles to be seen as not ending the game in e. g. japanese rules. It's just that it's accepted that the game will never end, and never have a result, and so the players just do something else. Of course, the difference to a “real” draw might be purely philosophical.


For all practical purposes, you have enough grasp of the terms, but if you are seeking precision, you need to take account of cultural as well as semantic differences and associations.

Most English speakers are coming to the term 'draw' from chess and soccer. In chess a draw was originally just a game that never ended (e.g. from lack of mating material) and, to get a decisive result, the game was just replayed. But as tournament go developed, it was the British Chess Association that introduced the concept of giving a half point each. That idea proved popular. The association of a draw with a half point each thus became strong and was reinforced by using the same idea in soccer and other field sports in Britain and Europe. Originally, a win was scored 2 points for the winner and 0 for the loser, but a draw was counted as 1 point each. In recent years that has been amended so that the winner gets 3 points, but a draw still gives equal points. This idea of draw = equality is still very dominant, but is not entirely fixed in chess. There have been experiments with giving both sides different point scores depending on whether they were lack or White. And of course the idea of avoiding draws by using penalty shoot-outs has become ubiquitous in soccer. But, still, in chess and cricket, draw = equal points is still an idea that comes readily to British and European people.

It's quite different in cricket, but that excludes Europe, and so usage there is limited to Britain and the worldwide Commonwealth. There, a draw is almost entirely seen as a game that hasn't finished properly (there are players still to bat - the game is ended by bad weather or lack of time). Until the advent of one-day cricket in recent years, it was vanishingly rare for a cricket game to end with equal scores. It is still not common, but now when it occurs there is a strong tendency to call that a 'tie' as opposed to a 'draw'.

In the USA it's different. As I understand it, in field sports there, they always call a game that finishes in normal play with equal points a 'tie'. As I mentioned, that term is easily understood in British/worldwide English. But is not often used. and if it is (e.g. in cricket), we still think in terms of acceptable equality. But a tie in the USA is seen as unacceptable equality. In baseball, for example, extra innings are played until a decisive result emerges. In American NFL football (and basketball?), I believe, that overtime (the British term is extra time) is used to a decisive result, though a tie is still possible when that time expires. Even then, though, the pressure from fans or the media is such that a team is expected to make a risky play to avoid a tie. The concept of honourable draws (as in e.g. boxing) doesn't seem to fit well with the American psyche.

In Japanese baseball, the concept of extra innings also applies but with the major difference that there is a limit of three each (games are time limited for neighbourhood noise control, etc), and if the result is 'equal runs' after that, that is what it is called (同点). Or other terms such as 均衡 (equilibrium) are used. 'Tie' タイ as borrowed from English is sometimes used (and is used for e.g. a tie in music) but tends to be avoided. I imagine that's because another word pronounced 'tai' 対 meaning 'versus' is so common in baseball that it would cause confusion.

Anyway, the upshot is that Japanese people are not conditioned to use the English term 'tie' in go, and in any case go long precedes baseball. So they use a native solution.

Two terms exist, both with very different connotations.

A 'jigo' is when a game is finished and the players have equal points (with or without komi). But that does not automatically mean it is counted as a half point each in a tournament reckoning. In old go, a contest was a series a games between two players, and if the overall score difference became big enough, the handicap was adjusted. Jigos in that case were essentially just ignored. Later, when tournaments evolved, the dominant type was the win-and-continue. When a jigo occurred in that case, the game was replayed. However, sponsors got wise to the fact that some players were contriving jigos just to get an extra game and thus an extra game fee. That led to a change in tournament structure - either pyramidical knockouts or leagues. In a knockout, a jigo would still have to be replayed, but in a league it could be counted as a haf-point each. However, that was not a given. In the original Meijin league, a jigo was counted as a win for White, but as inferior to a non-jigo win. This rule famously cost Go Seigen his chance to become the first modern Meijin. In short, unlike draw in English, jigo does not have strong connotations of a 'half-point each'.

The other relevant term in Japanese is mushobu 無勝負. This literally means 'without win or loss' and is reserved for those cases (including triple kos) where a game does not finish in a normal count-uppable manner and so does not have a decisive result. There are no special connotations here - it is just taken literally.

Korea and China have largely followed Japan as regards pro go, and so their terms overlap to a fair degree, but there are some practical differences. For example, the Chinese did formally take the step of defining superkos and incorporating them in their rule book. But in practice no-one has taken a blind bit of notice of that rule and so triple kos and the like are treated like mushobu. The anti-superko reasoning seems to be a combination of various factors: they are relatively rare; most people don't like mathematicians interfering in the game; it's seen as too awkward to implement; there's a desire to stay in step with international partners and avoid disputes (some Oriental sponsors have operations in all or some CJK countries).

In English, the term 'void game' has become common for mushobu (and that is the one used in the GoGoD database). It is a very explicit and therefore useful term, free of too many (or the wrong) connotations. Jigo appears to be the favoured term over 'draw' or 'tie', and as far as I can make out has similar, if diluted, connotations to the Japanese usage, which also seems useful.

PS One possible idea to avoid replaying drawn games in go might be to follow the penalty shoot-out idea and to make the players do a series of tsumego problems. Or one, and the first to solve it wins.

Author:  MikeKyle [ Fri Sep 04, 2020 4:16 am ]
Post subject:  Re: One more triple ko Draw

jaeup wrote:
First, it is not such an "exciting" triple ko.


I kind of know what you mean, but I still think this is a pretty exciting situation. Thanks for sharing.

I was wondering - does anyone have any experience of tripple ko in the west?

My understanding is that western tournaments typically use superko (of some sort) and so technically once a triple ko starts you should have a sequence of (at most) 5 (?) ko captures before a player can no longer continue the sequence and needs to make a threat (an "external" threat - not capturing one of the ko's concerned.) If both players have threats then this would probably feel a bit like a regular single ko but potentially with a chain of captures instead of a single ko capture each time (although I think the players could omit this?)

I think in practive this would require some heavy refereeing - Seems like a reasonable likelihood of an accidental illegal move. I wonder if some tournament referees would actually uphold the superko rule or if they would try to find some comprimise?

Author:  jaeup [ Fri Sep 04, 2020 5:39 am ]
Post subject:  Re: One more triple ko Draw

MikeKyle wrote:
My understanding is that western tournaments typically use superko (of some sort) and so technically once a triple ko starts you should have a sequence of (at most) 5 (?) ko captures before a player can no longer continue the sequence and needs to make a threat (an "external" threat - not capturing one of the ko's concerned.) If both players have threats then this would probably feel a bit like a regular single ko but potentially with a chain of captures instead of a single ko capture each time (although I think the players could omit this?)

I think in practive this would require some heavy refereeing - Seems like a reasonable likelihood of an accidental illegal move. I wonder if some tournament referees would actually uphold the superko rule or if they would try to find some comprimise?
I have never observed a real life game with a superko fight, but here are a few lessons I can provide.

1) Make a ko threat just after your opponent captured the ko. In most cases, there is no reason to wait until the last moment, risking an unintentional illegal move. (It will also lift the burden of the referee to identify if you are repeating the whole board positions.)
2) Despite rule 1, there are situations that you must delay the ko threat for a few moves. (I know it sounds weird, but it just depends on the situation, and you can only rely on your Go skill to find the best move for each case.)

Author:  jaeup [ Fri Sep 04, 2020 6:26 am ]
Post subject:  Re: One more triple ko Draw

John Fairbairn wrote:
A 'jigo' is when a game is finished and the players have equal points (with or without komi). But that does not automatically mean it is counted as a half point each in a tournament reckoning. In old go, a contest was a series a games between two players, and if the overall score difference became big enough, the handicap was adjusted. Jigos in that case were essentially just ignored. Later, when tournaments evolved, the dominant type was the win-and-continue. When a jigo occurred in that case, the game was replayed. However, sponsors got wise to the fact that some players were contriving jigos just to get an extra game and thus an extra game fee. That led to a change in tournament structure - either pyramidical knockouts or leagues. In a knockout, a jigo would still have to be replayed, but in a league it could be counted as a haf-point each. However, that was not a given. In the original Meijin league, a jigo was counted as a win for White, but as inferior to a non-jigo win. This rule famously cost Go Seigen his chance to become the first modern Meijin. In short, unlike draw in English, jigo does not have strong connotations of a 'half-point each'.

The other relevant term in Japanese is mushobu 無勝負. This literally means 'without win or loss' and is reserved for those cases (including triple kos) where a game does not finish in a normal count-uppable manner and so does not have a decisive result. There are no special connotations here - it is just taken literally.
I am not sure if "持碁 jigo" has a proper correspondence in Korean, so I will save my words about it, but I have some comments on "mushobu 無勝負 무승부" because Koreans also use this term and I believe its meaning is practically identical in both countries. (Probably it is the same also in China, but I do not know Chinese well, so I may be wrong.)

In direct translation, 無 = No and 勝負 = Result and it is not surprising that the word "No Result" is used in the English translation of the Japanese rule you have read. However, I do not recommend to regard a 無勝負 as if the game is nullified and/or must be replayed.

When a soccer game score is 2-2 after 90 minutes, or a baseball game score is 7-7 after 12 innings, Japanese and Koreans call such a game 無勝負. In most cases, the meaning of the word is closer to "no winner" rather than "no result". (Though in rare occasions, it may be used when a game is indeed nullified due to some technical reasons.) Whether the game is replayed or treated as 0.5 win and 0.5 loss is a matter of the tournament (or league) organizer's technical call.

Author:  John Fairbairn [ Fri Sep 04, 2020 9:37 am ]
Post subject:  Re: One more triple ko Draw

Quote:
In direct translation, 無 = No and 勝負 = Result and it is not surprising that the word "No Result" is used in the English translation of the Japanese rule you have read. However, I do not recommend to regard a 無勝負 as if the game is nullified and/or must be replayed.


I may be missing something, but this sounds to me like a distinction without a difference. It's rather like analysing the legal phrase "null and void". You can make a distinction if you really have to: a contract with no signature is null (no value) but a contract where you have corrected the signature may be void. But in practice people, even lawyers, treat the phrase as a tautology. If I heard two people talking about a triple ko game and one called it "no result" and the other a "void game" I'd assume they were talking about exactly the same thing.

Furthermore, neither phrase would indicate to me how the game would or should be handled by a tournament director. Practicalities take over. In go, there are various possibilities but none involve punishment.

But that's in go. I'd expect subtle differences in other games/sports, as I indicated above. And as I also indicated, there are cultural differences. In soccer, a 2-2 score is handled simply by giving 1 point to each side. There is no element of "punishment". In America (except perhaps for ice hockey? Not sure it's not a game I watch.) there is an element of punishment or disapproval in that players (e.g. baseball) have to play on and on until a decisive result emerges or (e.g. NFL) they have to play on for a bit and are frowned on if they don't get a decisive result. That's alien to British (and European?) people. Japanese baseball sits somewhere in the middle. There's a small element of punishment in having to play extra innings up to 12 (or 15 in playoffs?) but if the score is 7-7 at the end of those, it's not really a draw, but a no result or void game, depending on personal preference. Being British, I'd prefer to avoid "no result" because 7-7 would be a valid result in a soccer game and I have soccer in my blood. I'd call it a void game (or maybe use the, for me, slightly foreign word 'tie') and I'd be bolstered in that by the fact that such games are indeed void (i.e. have no effect) in Japanese baseball because they are not counted in the WIN% stats and so (I believe) can detrimentally affect standings at the end of a season - the punishment element.

To go back to the semantic element, shoubu of course, like many binomes, has the standard CJK problem that it can mean either 'victory and defeat' or 'victory or defeat' and that can lead to subtle differences in perception.

In short, in go, I personally will stick with 'void game' but will accept that 'no result' has equal standing. But I would be a bit unhappy with triple ko in pro go being called a 'draw'.

Author:  Javaness2 [ Fri Sep 04, 2020 12:09 pm ]
Post subject:  Re: One more triple ko Draw

John Fairbairn wrote:
Most English speakers are coming to the term 'draw' from chess and soccer. In chess a draw was originally just a game that never ended (e.g. from lack of mating material) and, to get a decisive result, the game was just replayed. But as tournament go developed, it was the British Chess Association that introduced the concept of giving a half point each. That idea proved popular. The association of a draw with a half point each thus became strong and was reinforced by using the same idea in soccer and other field sports in Britain and Europe. Originally, a win was scored 2 points for the winner and 0 for the loser, but a draw was counted as 1 point each. In recent years that has been amended so that the winner gets 3 points, but a draw still gives equal points. This idea of draw = equality is still very dominant, but is not entirely fixed in chess. There have been experiments with giving both sides different point scores depending on whether they were lack or White. And of course the idea of avoiding draws by using penalty shoot-outs has become ubiquitous in soccer. But, still, in chess and cricket, draw = equal points is still an idea that comes readily to British and European people.


I got a bit confused on reading this paragraph. Why did the British Chess Association introduce the concept of a draw to tournament go I thought, and wasn't it the British Chess Federation anyway. I see though, that a chess tournament in Dundee is seen as the first to introduce a half point instead of demanding a replay. Indeed, back in 1867 there was a British Chess Association. https://www.chessscotland.com/documents ... e_1867.htm

The practice of awarding 3 points for a win and only 1 for a draw surely has little merit in go. It couldn't encourage attacking (exciting) play. If anything, a draw has significantly more to recommend it in go than it would in chess, particularly with the foul practice of stalemate still retching over the board.

Author:  jaeup [ Fri Sep 04, 2020 11:13 pm ]
Post subject:  Re: One more triple ko Draw

John Fairbairn wrote:
I may be missing something, but this sounds to me like a distinction without a difference. It's rather like analysing the legal phrase "null and void". You can make a distinction if you really have to: a contract with no signature is null (no value) but a contract where you have corrected the signature may be void. But in practice people, even lawyers, treat the phrase as a tautology. If I heard two people talking about a triple ko game and one called it "no result" and the other a "void game" I'd assume they were talking about exactly the same thing.

In short, in go, I personally will stick with 'void game' but will accept that 'no result' has equal standing. But I would be a bit unhappy with triple ko in pro go being called a 'draw'.
Actually, my claim is exactly the opposite. Koreans at least think that, if a professional team plays a soccer game, they have three possible results: a win, a draw and a loss. A draw is one common result, and nothing is canceled, nullified or void.

The official website of K-League(our Premier League) shows three letters for the results, 승(勝, win), 무(無, draw), 패(敗, loss). (I know one may still say 無 may be translated as no result rather than draw, but believe me. If you pick one average Korean and ask him if a 2-2 soccer game is a draw, they will simply say yes without hesitation provided that he knows the English word 'draw'.)

In the official website of Korean Baduk Association, the result between Se Dol Lee and Gu Li is shown as 23 승(勝, win), 1 무(無, draw), 25 패(敗, loss). I know they made one quadruple ko, and the KBA treats it like a draw in a soccer game. I do not like to associate words like 'void' or 'nullified' for the quadruple ko game. My Japanese is not so excellent, but the way I understand it, they also won't have a trouble associating 無 to a draw because they use the same letter 無 for a 2-2 soccer game or 7-7 baseball game.

I just realized that there is a good example of a real 'nullified game'. In a recent game between Jeonghwan Park and Tingyu Fan, there was a dispute about mouse malfunctioning and time loss, and the game is indeed nullified(무효, 無效, 'not valid' is the expression I saw on the newspaper) and replayed on the next day. KBA website shows only 8 wins and 6 losses between the two players. No 무(無, draw) is associated for that game.

Author:  jaeup [ Sat Sep 05, 2020 1:50 am ]
Post subject:  Re: One more triple ko Draw

Earlier, I said "持碁 jigo" may not have a proper correspondence in Korean. Actually, there is one Go term, 화국 (和局) which may correspond to it. It is used when a (usually) handicapped game without komi ends with the same score. Its English translation I found is "Game of peace". Well, the Korean rule text does not include such a term, and it is more of an aesthetic expression rather than an official term used by a referee.

The reason I mention this is to describe the subtle cultural difference I am thinking now. In Korea, almost no one thinks that a Go game ending with the same score is abnormal or should be penalized. (A replay is just to save the embarrassment of the tournament organizer) 화국 (和局) is more of a word of compliment that both players played with all of their hearts and found the best result that nobody need to lose.

I think this kind of culture affects the attutude of rulemakers and rule theoreticians. Many (but not all) rule theoreticians in the West think that a game must have a winner and a ruleset accepting a draw is flawed just because of it. However, the Korean, Japanese and (practically) Chinese rules do not mind introducing a draw and think that it is a justifiable result.

Maybe it is one reason why in Korea and Japan a baseball game often ends with a draw, while in the major league they play 15, 20 or even 25 innings to decide the winner.

Author:  ez4u [ Sat Sep 05, 2020 2:00 am ]
Post subject:  Re: One more triple ko Draw

I am too poor at Japanese to have an opinion, so I will just add a little data on usage here in Japan.
In Japan, professional baseball and soccer use different terms:
Baseball
Attachment:
Pro baseball table.jpg
Pro baseball table.jpg [ 58.71 KiB | Viewed 9056 times ]

Soccer
Attachment:
J league table.jpg
J league table.jpg [ 56.61 KiB | Viewed 9056 times ]

It seems that the Nihon Ki'in keeps 無勝負 games in the official count of games played, sort of. The following picture is from the article announcing Cho Chikun reaching 1,500 wins in 2017. The yellow highlighted line lists him as having played 1,500 wins, 821 losses, 3 jigo, and 4 mushoubu. However, the circled line reports his winning percentage as 64.6%. This result appears to be obtained by excluding both the 3 jigo and the 4 mushoubu (1500 / 2321 = 64.62%). The alternative of including those seven games would have given 64.4% (1500 / 2328 = 64.43%) :) Note: just in case I did not understand what was going on, I also checked the announcement of Otake reaching 1,300 wins last year. There also the reported winning percentage matched the calculation = wins / (wins + losses), excluding his 5 jigo and 1 mushoubu.
Attachment:
Cho Chikun Record 2.jpg
Cho Chikun Record 2.jpg [ 103.29 KiB | Viewed 9056 times ]

Author:  jaeup [ Sat Sep 05, 2020 3:40 am ]
Post subject:  Re: One more triple ko Draw

ez4u wrote:
I am too poor at Japanese to have an opinion, so I will just add a little data on usage here in Japan.
In Japan, professional baseball and soccer use different terms
I see. 引き分け seems to be a more popular word in this case, and the word 無勝負 may really has a different status depending on the area of sports. It is interesting that 무승부(無勝負) in Korean definitely includes the case of 引き分け in Japanese.

Hmm.. in Japanese wikipedia, 無勝負 redirects to 引き分け, but reading the document 引き分け, the two words are (at least) not identical.

Author:  Matti [ Mon Sep 07, 2020 9:21 am ]
Post subject:  Re: One more triple ko Draw

Some people would call me a rules theoretician. I prefer to separate the rules of play (which define the game) and tournament rules. If a game end with triple ko or such, it may be treated as a draw or no result. I don't care which, once it has been decided in advance. Is there any other conceivable outcome?

John, thank you for explaining some linguistic points.

Page 1 of 1 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/