Environmental Go Game

The world's first game of Enviromental Go was played on Tuesday April 21 1998 at the American Ing Goe Center in Menlo Park, California. The players were (White) Rui Naiwei and (Black) Jiang Zhujiu, both 9-dan professionals.

Starting time was 10:30 a.m. PST, with adjournment for lunch, and resumption at 13:30. The game was live on the Internet, on the IGS Go Server.

Another such game may be played this summer at the American Go Congress in Sante Fe, New Mexico in early August.

Game Commentary

by Professor Elwyn Berlekamp

Although we thought that we had agreed on the use of Ing rules, for reasons to be revealed forthwith, I believe the game was actually played as though Japanese or American rules were operative,

There were 40 cards numbered from 20, 19.5, 19,..., 0.5. Players were allowed to either take the top card OR play on the board. The final score was equal to score on the board, plus the total of cards you have taken, plus 9.5 komi for White. Here is the number of moves played on the board while the available card was as shown:

The last two recorded moves each filled an unfought ko of Japanese temperature 1/3; there were then many "dame" left on the board, and the filling of these was not recorded but they were instead filled alternately as part of the scoring process. Black ended up with 4 more cards than White, but Black filled the last dame, so that White had 3 more stones on the board than Black. In Ing scoring, White won by 2½ points, but in Japanese or American scoring, Black won by ½ points.

It is easily seen that in Ing scoring, the 1-point card behaves exactly like a dame, and it should be taken during that final phase of the game, not before. So from the final stages of play, it seems clear that both players were thinking in a different scoring system, such as Japanese or American.

A video tape was made of the game and the commentary session which followed.

Each player used up their full 3-hour allotment long before the game ended, and both were playing on "byo yomi". It seemed to some of the kibitzers that, during much of the endgame, each player would often take a card not because it was a well thought-out decision, but because they had used over 55 seconds of their 1 minute allotment, and grabbing the top card near the last second might have seemed to them to be a relatively "safe" thing to do.

Conclusions

In my preliminary opinion, the event was a very successful first run. We learned quite a bit. We obtained one quite interesting and very close game.

New and clearer directors' rules will need to be devised before we do the next such event. I now eagerly welcome comments and suggestions. The time limit will have to be changed to put more emphasis on the endgame. One preliminary idea, partly borrowed from chess, is the following:

2 hours each for all moves while the top card exceeds 5 (or possibly another value, to be discussed in the following paragraph). A player whose clock runs out goes onto byo yomi. But, when the 5½ card is taken, then both players receive an additional one hour on each of their clocks.

The idea is to use some specified value of the top card (e.g., "5") as the definition of when the "endgame" begins. Based on the existing sample of one game, any number between 4 and 10 would have given approximately the same "endgame". However, this particular game featured a very big (and hot!) middle game fight. A larger collection of games will be needed to get some better feeling of what is the best value of "5".

I think we were lucky that this particular game ended so close. I think many games will often be clearly decided before the endgame stage (i.e., top card <= "5") is reached. So, I'm considering ideas on how to ensure a very high level of attention to this phase of the game, even though the primary contest might already essentially be "over". One preliminary idea is as follows:

Only some fraction f (perhaps 2/3) of the total prize money is awarded to the winner of the outcome of the conventional environmental game. After the conventional game ends, before any commentary, the endgame is replayed with sides reversed, and with a new time rule, perhaps 30 or 45 minutes each plus byo yomi, and/or maybe the byo yomi is 2 minutes/move instead of only one. Then the final third of the prize money is awarded to whoever played the better endgame (a la duplicate bridge; each player gets to play each side of the same position once). In the case of identical scores on the two playouts of the endgame, the final fraction of the prize money would be equally divided.

Game Record

David Kent entered the first 90+% of the game onto the Internet in (almost) real time, but got messed up near the conclusion. Bill Spight recorded the game with paper and colored pencils. During the postgame commentary, the players replayed the entire game from memory, and it agreed with Spight's record. The game and commentary also exists on video tape.

The game record may be downloaded in Ishi or SGF format.

A second game of Environmental Go was played in July 2000.

This game appears to have been played using Japanese (territory) scoring. With Japanese scoring, a move on the board is worth a point less than with Chinese scoring.

Environmental Go

Environmental Go is Go played in the normal way, but with a stack of forty cards, numbered ½, 1, 1½, .., 19½, 20 as well as the normal board and stones. For a player's move, they may either play a stone on the board in the usual way, or take one of the cards. At the end of the game, a player's score is the sum of their score on the board and the cards that they have taken.

The spelling "goe" rather than "go" is preferred by the sponsors, the Ing foundation. It is not relevant to the interest of "environmental go(e)".

The name environmental was given by Professor Elwyn Berlekamp, for reasons connected with the evaluation of games of complete information, as developed by himself and John Conway. The "environment" of a game is the set of alternative moves, in addition to the one currently being considered. A rich choice of alternative moves affects the theoretical analysis of a move sequence, and may effect which of two moves is better (e.g. by eliminating the possibility of tedomari).

Here is more information about Combinatorial Game Theory.

Rui Naiwei

Rui Naiwei 9 dan was born on December 28, 1963 in Shanghai, China. She turned professional in China in 1985, the year she reached 7 dan. She was promoted to 8 dan in 1986 and 9 dan in 1988, the first woman ever in the history of go playing to reach that level. She moved to Japan shortly after the Tianamen Square incident, where she taught go and worked for an insurance company. She was unable to enter the Japanese professional Go association, but several newspapers sponsored matches between Rui and the top tournament players in Japan.

Rui, who is without question the greatest woman Go player of all time and one of only two women to reach the top rank of 9 dan, has recently begun representing the United States in professional Go matches. She along with her husband Jiang Zhujiu, 9 dan, has moved to the Bay Area of California.

Rui has won three of the four Women's World Go Championships (first, third and fourth), making her the current champion, about to defend her title in a few weeks time. She was unable to attend the second championship. In 1992 Rui placed third in the Ing Cup World Championship of Go, which brought together the best go players in the world from Japan, Korea, China and Taiwan. She was the only woman competing and is the only one to ever compete in a world championship such as this. This could be considered the highest achievement by a woman in any strategic board game throughout history.

Jiang Zhujiu

Jiang Zhujiu, nicknamed "Jujo", was born in Shandong province and then moved to Taiyuan City in Shanxi province. Jiang toured Japan with a Chinese team in 1980, when he scored 6-1, and again in 1982, when he scored 7-0. In 1985 he was the sensation of the Japan-China Super-Go series, a team competition based on successive individual matches. He shocked Japanese Go fans by toppling one after another the first five members of the Japanese team. His defeated opponents were all past or current members of the Honinbo and Meijin leagues, which marked them as of the professional elite in Japan; one was a former title-holder. In 1988, Jiang Zhujiu placed 5th in the Ing Cup, reckoned to be the first ever genuine world championship of Go. He was also one of the most popular TV Go-teachers in China at that time.

Rui and Jiang

Rui Naiwei and Jiang Zhujiu are now married to one another. They live and teach Go in the San Francisco area. Recently, they established the American Professional Go Association. This organization, with its eleven members, hosts the annual American Master's Go Championship, which Jiang has won three years in a row. Both are seen playing on the IGS (Internet Go Server) from time to time.

Another view of the game.