Rules poser in the Kisei

A mid-game pass was played for the first-time ever in a seven-game title match in Japan in Game 5 of the 32nd Kisei. It was part of a conundrum for both players in a position that has long been highlighted by western rules theoreticians as a major flaw in the Japanese 1989 rules.

The problem arose on Day 2, that is 28 February 2008, when Cho Chikun (Black) had established a small lead, reversing the advantage enjoyed by Yamashita Keigo at the end of Day 1. As is his wont, Cho was in time trouble. He had entered the 10-minute byoyomi period as early as move 119, and by the time of the incident he was down to his last five minutes (and 1-3 down overall at that). He had also been grumbling and yelping non-stop until the endgame, when he quietened down. The analysts in the press room had also ended their labours, thinking the game was effectively over.

The only people on duty now, apart from the players, were referee Sakaguchi Ryuzo and commentator Imamura Toshiya, who had been chosen simply because they were both from Kyoto where the game was being played. Normally, the referee's job is a sinecure. On this occasion Sakaguchi was to have a little excitement.

Cho suddenly exclaimed, "What can I do? I can't count this!" He struck his knee and made the press room look up in astonishment. The position was as below, with White about to play 292. Yamashita was not quite in byoyomi.

Cho had already made a mistake in byoyomi by allowing White to play at the square-marked point (move 244 - it was Black's privilege to play there). To add to his time troubles, that made Cho flare up in rage, according to observer Goto Shungo, and they had been playing the knotty ko at A for some time. Cho was trying to make Yamashita give up a point by filling in the ko and so ending for ever the possibility of Black getting a play at B (having taken the ko, of course).

Having seen Black play 291 at the triangled point (a worthless dame), Yamashita decided to pass. Apparently he actually said, "Pasu." He was also implying that he would say pass every time in future that Black played a dame. In other words, Yamashita believed he did not have to make a repair (teire) move in the lower right.

This was at one minute past seven in the evening. Cho then said to Yamashita, "So we need the referee, eh?" Yamashita said, "Yes," quietly. Cho then turned to the scorekeeper - Sakamoto Yasuo 2-dan - and said, "Please call the referee." Sakamoto rushed off to the press room. There was apparently the sound of an explosive "What the ..!" from there. Sakaguchi then came into the playing room, and Yamashita, after waiting for him to take his seat, again announced, "Pasu." Since Cho declined to say pass in his turn, the game had to continue. It went on up to a quarter past seven. Sakaguchi meanwhile was reading the rule book.

It is not clear from the account in Gekkan Go World, to which we are indebted, exactly what made Cho desist beyond White 310, beyond a realisation that there is no way he can turn this into a real ko (incidentally, the internet version of the game we have seen differs from the printed version). Nor is it clear whether Sakamoto actually made any ruling. By the sound of it, there was no need to.

When the game was counted up, the one point they had been fighting over was irrelevant. Black had won anyway by 3.5 points. Yamashita, who was not in time trouble, knew that but as Kisei he had an obligation not to sully the title by playing an unnecessary fill-in move. He had to play the best move, and for once "pass" was the best move. Whether he was also playing mind games you will have to decide for yourself, but we would suggest that was highly unlikely.

The rules today are in part the result of two similar prior incidents involving a ko and a fill-in move. One was in 1948, when Iwamoto Kaoru believed he did not have to fill in a final ko because he had a surfeit of ko threats. It was more of a theoretical dispute, because his opponent, Go Seigen, won by either 1 point of 2 points depending on whether the fill-in took place. The referee on that occasion was Segoe Kensaku and he came up with the ideal decision: "White, Go Seigen 8-dan, wins by 1 or 2 points."

Japanese referees have to bear in mind what happens to umpires in Japanese baseball games. They get beaten up. In one famous case, the poor umpire changed his decision about a dozen times, depending on who landed the last good punch, in the space of being pummelled for an hour.

Segoe also received flak for his decision, and the Yomiuri Shinbun (the sponsor), quoting an internal memorandum attributed to Honinbo Shusai that "Where there is a surfeit of ko threats, it is not necessary to fill in," later announced a correction: "White won by 1 point" (on 10 July, two days after the game was completed). Although this was seen as enforcing the "traditional" view, Shusai had apparently disagreed with his predecessor, Honinbo Shuei, on this. Fans also had their say, and it was this incident that led to the first complete set of written rules in Japan, published on 2 October 1949. This was really thanks to Baron Okura Kishichiro who was famously able to bang heads together. He is sorely missed.

But these rules reversed the 1948 thinking and declared that a true ko needed to be filled in at the end of the game, regardless of the ko-threat situation. The rules committee apparently felt that Shusai's byelaw lacked clarity.

It took all of 10 years for the new rules to be tested in practice. Go Seigen was again involved, this time in a game in 1959 against Takagawa Shukaku. In this case the result depended on the winner of the dispute. Go claimed he did not have to play a fill-in move in the centre as Black had too few ko threats. The adjudication went against him and Takagawa won by half a point. That was obviously in line with the 1949 rules, but Go pointed out he was not a member of the Nihon Ki-in and had never agreed to their rules. He was compensated by a long unkept promise to review the rules, although the dispute was well aired in print.

The next stage in the controversy never actually happened, but it could have done - and if it had happened it would have been an even better example than the Kisei case as it would have affected the result. It would also have been under a different ruleset, of course, as this happened - or did not happen - in the Honinbo titlematch in 1970. There was an unfilled ko and Sakata looked hard at it, before deciding he had to fill in. From his reaction afterwards, however, it seems that he thought he had won anyway and did not have Yamashita's qualms about sullying the record with an unnecessary move. Actually he lost, but did not complain. It was the go historian and Kido editor, Hayashi Yutaka, who started a dispute later in the Mainichi Shinbun about whether the final fill-in move was actually necessary. He pointed out that the rules specifically said that a "true" ko had to be filled in but they made no mention of a yose-ko, which was what this game featured.

There was a bit of controversy in the press, and Hayashi was swatted down by the Nihon Ki-in who sent in Kano Yoshinori to bat. What the incident did highlight, however, was the fact that the Nihon Ki-in had not kept its promise to review the rules.

Since we are often at pains to counteract the western hubris that rules discoveries are made here and not in Japan, we may in passing point out that Hayashi's main article was entitled "Tradition and logic in the Japanese rules of go - discussion of the endgame, dame and local life and death."

To return to the Kisei game, since this is a rare but actual example of resolution under the 1989 rules (a revision generally regarded as being both a sop to western opinion and a pig's breakfast - and no doubt there's a connection there), we should follow it through in detail.

You will need to know the precise rules first. We will be literal here rather than use the paraphrases of the Go Almanac version.

Article 7.2: (1) Recapturing the same ko when confirming life or death after the "cessation of the game" of Article 9 is prohibited. (2) However, the player whose stone has been captured in the ko may take that ko again after having given up his right to move in respect of that particular ko.

Article 9.1: At the point in time when a player gives up his right to move and then the opposing player also gives up his right to move, the "cessation of the game" occurs.
Rules commentary: Giving up the right to move (the so-called "pass") is a declaration that play has ceased, but if the opposing player then also gives up his right to move, play does cease. The point in time when both players have expressed their intention of giving up the right to move is the cessation of play.

Article 9.2: After the cessation of play, both sides confirm the life or death of stones, and the territories, and the game is completed by agreeing on this. This is referred to as a "completed game." Gekkan Go World commentary: In this case, Cho had two options: pass or make a move.

Article 9.3: If, after cessation of the game, one player requests a resumption of the game, the opposing player has the right to play first, and he must comply with this.

Now to the actual sequence in the game, as outlined by Gekkan Go World, which calls this the "process towards the completed game."

1. Yamashita passes (Article 9.1 - gives up right to move).

2. Cho either moves (continuation of game: move 293) or, eventually, also passes (Article 9.1 - cessation of the game)

3. Confirmation of territories and life or death (Article 9.2). Both players agree and that leads to the completed game, or one side disagrees and requests resumption of play, and when play resumes the opposing player has the right to move first (Article 9-3).

This does not actually seem to address the question of whether the ko needed to be filled in. It is more a description of the fact that Cho eventually gave up disputing it. There is a ruling in the commentaries on the rules that says a fill-in move is not required in a yose-ko, so the actual course of the game is correct. It would seem that Cho was not entirely familiar with the 1989 rules, but they are actually clear enough in this particular case.

This incident will no doubt provide ammunition for both sides of the western rules debate. On the one hand it's a real rarity (three and a half cases in a century) and even arguably the world's best player apparently did not bother to learn the precise rules. On the other hand, even though the messy rules could be said to have worked on this occasion, those who have a compulsion to tidy up rules will bring out the red pen again.

In that connection, we were saddened to read recently of an example in Japan of what we think is the wrong with the obsessive approach to rules. An amateur proudly illustrated how he exploited the rules to win a lost game. You will need to know that Japanese amateurs normally play sudden death and do not use byoyomi, and that they have no place for rules such as it being permissible to stop the clock if capturing more than three stones - there, all captures have to be made while the clock is ticking. This amateur realised his opponent was close to the time limit and so played out an unwinnable ko which, however, had to be resolved by the opponent having to capture a large number of stones - during which time his clock ran down. The amateur attracted a lot of opprobrium, but at least had the grace to admit he was a pretty unpleasant individual.

Now doesn't Yamashita's concern not to sully a game record with a blemished move seem even more chivalrous by comparison? And he'd obviously learnt the rules!