2012 Go Symposium




As of June 28

List of Abstracts

Siddhartha Avila. The Art of Go in Education. View Abstract
Paul Barchilon. Developing Youth Programs. View Abstract
Jacques Basaldúa. Advances in M-eval: A Multivariate Evaluation Function for Opening Positions in Computer Go. View Abstract
Terry Benson. Spreading Go - Rules and Attitudes Which Make That Difficult. View Abstract
Kyle Blocher. The Value of Moves: Sente, Gote, and Combinatorial Game Theory. View Abstract
Dalsoo, Kim. A Study of Star and Flower Points. View Abstract
Roxanna Duntly-Matos. "Playing Under and Pushing Through the Stones": Privileging Transnational Go Network Formation and Minority Leadership from Lower to Higher Education. View Abstract
Peter Freedman.Teaching Children to Become Go Players. View Abstract
Roy Laird, PhD (USA). Play go and grow: Why every school and library should have a go program. View Abstract
Marc L. Moskowitz. Go Play Go: The Challenges of Categorization for an Increasingly Globalized Game. View Abstract
Francois van Niekerk. New Work on MCTS Parallelisation. View Abstract
Israel Rodriguez Nava. Cultural Considerations in the Spread of Go in Western Countries. View Abstract
Peter Schumer. Design of a College Go Course. View Abstract
Peter Shotwell. The Origins of Go Strategies in Classical Chinese Grammar. View Abstract
Bill Spight. How to Estimate Territory: Research into the Influence of Immortal, Partially Immortal, and Dead Stones, and the Estimation of Territory on the 3x3 Board. View Abstract
Ichiro Tanioka. The Origins of 19x19 Go-Boards. View Abstract
Thomas Wolf. Basic Seki in Go. View Abstract
Chen Zu-yan. The Art of Black and White: Weiqi in Chinese Poetry View Abstract
Chen Zuyuan  陈祖源. The Evolution of Japanese Territory and Chinese Area Scoring Since the Tang Dynasty. View Abstract





Roy Laird, PhD (USA). Play go and grow: Why every school and library should have a go program.

We live in an "evidence-based" world. Teachers, librarians and others who work with youth need to know the benefits of incorporating Go into their programs. Organizers need hard facts to convince them. I will summarize widely accepted research on the positive influence of chess programs on youth development in many ways, noting specific factors that make Go a comparable or superior arena for growth and development, and citing research that suggests that playing go may produce actual differences in the brain.




Marc L. Moskowitz. Go Play Go: The Challenges of Categorization for an Increasingly Globalized Game.

This year The American Go Association will hold the first professional rank competition in the United States. To commemorate this important event it therefore seems fitting to pose a few questions for further thought concerning the direction that America's Go should take. First, I would like to question whether "Go" is really the best term to use in English. Given that it is a homonym for the verb "to go" this can at times create syntax bordering on a Laurel and Hardy skit ("go play go", for instance). In the internet age, search engines read the word "Go" as the verb with predictably frustrating results.

A second issue, also inherited from Japan, is the dan-ranking system. This system has been adopted throughout the world but in interviewing Weiqi players in China I found that there is in fact a great deal of dissatisfaction with the current system. Many people I interviewed in China posited that it is an unjust system. They referred to nationally imposed age limits, so that if a male does not achieve a dan rank by the age of sixteen, or win a national competition by the age of 24, he will not be allowed to have a professional dan rank regardless of how well he plays after that age. The fact that one can never lose one's rank is also problematic, for some people with professional ranks do not compete for a living and some people who have made their careers by playing at Weiqi competitions have not been granted a professional rank. In my interviews, several people suggested that it might be better to use a system that is more akin to baseball or international chess in which one’s cumulative score, based on ongoing competitions, is kept.

It is part of America's history to be a new frontier that addresses problems in new and innovative ways. Because of Weiqi's relatively small numbers in the US it seems unlikely that we will redefine the theory of the game in the next few decades. Yet we may be able to influence the Weiqi sphere by reforming our categories, both in terminology and ranks.




Peter Shotwell. The Origins of Go Strategies in Classical Chinese Grammar.

It is self-evident that language affects culture and thought and I will argue that this is because of the way abstract nouns were used or not used in Classical Chinese and Western languages that their users are predisposed to play wei qi or chess.

a) The presence and development of abstract nouns that could be used to codify the features of the physical and mental world by the early Greeks, influenced by their contacts with other languages, caused the creation of a "two-tiered" system of "reality." As Plato constructed his Ideas, Aristotle his Categories and Essences, and the Greco-Arabic-Christian worlds their ideas about God, the senses were demoted and considered to be untrustworthy and inferior.

b) The isolated Chinese embraced wei qi once their early period of 550 years of warfare ended in 221 BC because Classical Chinese can be construed as "non-abstract" in the sense that it postulates "one-world"—the one that is seen, smelled, felt and heard. There is no "to be"—i.e. there is no "it" in "It is raining," There is only "rain."

c) The fact that there was no outside reality forced the creation of the correlative classifying system of yin-yang in which there are no outside referents. Everything is defined in terms of everything else and nothing can exist by itself.

d) c) Thus, when using Classical Chinese, one cannot say what something is without taking into account what it isn't. If there is "Yang" there must be "Yin"—the Strong must have a Weakness— and once one knows that, if you are a strategist or war general, you are tempted to use it. As the armies in that dark period grew in size to many hundreds of thousands of men, an entire philosophy was built on expediency. As demonstrated in books like Sun Zi's Art of War, and unlike the Western concepts of aiming for conquest through direct force, getting the most for the least amount of effort was the goal. Yin could be used to manipulate yang. Not fighting was always better than fighting—the banquet table is preferred to the battlefield; don't approach strength directly; lure enemies into traps; sacrifice the small to gain the big; always think ahead to look at the large picture.

d) If this sounds like "Go-think," it is. "Nothing" obtains existence because there is "Something." The empty "eyes" of Go are what makes groups useful and existent. There are no invisible abstract qualities in go stones which are played one by one to build groups which are all intertwined and cannot exist alone. The nature of the game is yin-creative and players are advised to temper their aggression.

e) On the other hand, in Western chess, all the pieces have abstract qualities—a meaning beyond their physical reality—while the latent possibilities of their movements lurk in the background and are united in an abstract way in their effort to kill the other king by direct force.

f) The Greeks had Poelis that was played with stones, but it was a direct-force, custodial capture battle game. The Chinese have chess, but, as the earliest writing on the yin-yang of Go playing shows, only it encapsulated the Daoist, Confucian and Sun Zi-like underpinnings of their culture.




Peter Shotwell. A Form of Tibetan Mig-Mang From the West?

By Peter Shotwell

Weiqi, called "Mig-Mang" (and often misspelled "Ming-Mang"), came to Tibet from China, but there is a second game played with go sets that bears the same name since the name refers to the "many eyes" of the board. This second game features "custodial capture," a singular opening pattern and the replacement of captured stones with one's own. However, there are conflicting accounts of whether or not the stones move like the rooks of chess or one intersection at a time.


and this

but not this

Opening Position


China has many games that are played with go-like boards and stones but none that use custodial capture. However, the history of the West is littered with those that do. The Greeks had Polis or Poleis and Plato says that it "came from Egypt" which now has Seega and may have had it in ancient times, the Roman Empire had Ludus Latrunculorum and the Vikings had Hnefatafl. A problem, however, is that while the rules of Seega are clear since it is played today, the descriptions of the Greek and Roman, outside of the capture rules, are sketchy to say the least. Like so many other descriptions of early games, including the early Confucian descriptions of go, no one wrote about them in any detail because "everyone knew them." For example, as in Mig Mang, there is an open question as to whether pieces moved like rooks or one space or intersection at a time.

In any case, my thesis is that Mig Mang may have come to Tibet from the West through Grecian Bactria following the path of Silk Road-type trade, Persian Dualism and Buddhism from India.




Bill Spight. How to Estimate Territory: Research into the Influence of Immortal, Partially Immortal, and Dead Stones, and the Estimation of Territory on the 3x3 Board.

In go a local non-terminal position may be evaluated by solving its game tree. The first computer go program, by Albert Zobrist in 1968, included an influence function which was used to evaluate positions without lookahead. The decades since then have produced no consensus about influence. Modern Monte Carlo go programs play games out until they may be scored. The question of the influence of go stones upon other stones and empty points remains unsolved.

A theory of influence is being developed by studying influence in positions in which it may be precisely calculated. The influence of immortal and partially immortal stones has been studied, as well as influence on the 3x3 board. The results of this study are reported.




Chen Zu-yan. The Art of Black and White: Weiqi in Chinese Poetry

No game surpasses weiqi in the interest it has evoked among major Chinese poets. Their fascination is explored in this presentation through close analysis of a small but representative sampling of weiqi poems. Decoding these seemingly frivolous poems reveals the richness of weiqi as a source of artistic inspiration. China's great poets drew from weiqi's patterns of opposition three broad metaphors: weiqi approximates war, offers paradigms for social order, and teaches lessons about humankind's moral stake in the cosmic game.




Chen Zuyuan  陈祖源. The Evolution of Japanese Territory and Chinese Area Scoring Since the Tang Dynasty.

Japanese rules of territory scoring are notorious for having many flaws and to try to improve them is an important subject. Through the interpretation of ancient Chinese texts, the rules of the Tang dynasty (618-907 AD) have been found to be the origin of territory scoring. However, it was originally a simplified form of stones scoring—that is, it was territory counting based on equal stones being played by both players, leaving the legacies of "three points without capturing" and "empty points in seki cannot be counted". Originally, Japanese rules used that principle, however it was lost in the process of development over the centuries and all of the troubles therefore arose. The most natural way to resolve this problem is to restore the "equal stones" concept, a process which also proves that the modern amendments of the AGA, FFG and BGA are correct.




Dalsoo, Kim. A Study of Star and Flower Points.

The modern Go board has 9 star (flower) points but some Go boards from past times did not have a star point while others had up to 25. The usage of star points now is generally for handicap games but to add a stone on a star point does not exactly represent the difference of Go strength. Thus, this paper is to examine star points, flower points and other markings of various board games and to clarify the original purpose of them. The conclusion is that the original purpose of star points does not seem to be for Go but for the entrance and check-points of race games.




Francois van Niekerk. New Work on MCTS Parallelisation.

Due to physical constraints, parallel hardware, such as multi-core and cluster systems (sometimes referred to as supercomputers), has become increasingly commonplace. In order to take advantage of this hardware, we parallelise algorithms such as Monte-Carlo Tree Search (MCTS), the dominant algorithm for Computer Go.

In this talk, the general workings of MCTS and the main approaches to MCTS parallelisation are presented. Results from an implementation of tree parallelisation for multi-core systems and root parallelisation for cluster systems are discussed. Finally, new developments in MCTS parallelisation are discussed and we consider the path forward.




Israel Rodriguez Nava. Cultural Considerations in the Spread of Go in Western Countries.

Besides a brief summary of Go's development in Asia and the first accounts found in Europe, I mainly review Schädler's (2003) and Van Ees' (2005) hypothesis on the difficulties Go faced in the spread to Europe. While I partially agree with some, I propose the main reason that Go did not spread earlier to the rest of the world was Go's historical development itself—it was a game reserved for the highest and most educated classes in Asia until the first years of the XXth century.

In the second part, I review the recent history of Go in Japan, Korea and China, analyzing its quick development, opposed to the slow development in the West, mainly focusing on the Confucian tradition and the long history and high regard Go has had in Asia for hundreds of years.

Finally, on the last part, I ponder the various challenges a globalized society present to both the West and the East, and discuss some ideas and relatively "new" approaches to help spread Go in Western Countries.




Jacques Basaldúa. Advances in M-eval: A Multivariate Evaluation Function for Opening Positions in Computer Go.

Computer go has experienced great advances with the introduction of Monte-Carlo Tree Search (MCTS). This video presentation describes how MCTS works and gives a robust solution to a central problem in computer go. Also, it analyzes how knowledge can be implemented in an MCTS engine and takes a closer look at its strengths and weaknesses, explaining why the opening is one of the weak points. A second topic is about extracting knowledge from large collections of professional and strong amateur games and how ideas developed by go experts can be implemented in a program. Finally, we explain our solution called M-eval—how the idea behind it evolved as the program got stronger, what useful information it gives about positions, and how strong it make the program. Our presentation is not just for computer scientists, but is intended for a wide audience of go players with an interest in computer go.





Kyle Blocher. The Value of Moves: Sente, Gote, and Combinatorial Game Theory.

The concept of sente and gote is a natural one to Go players, yet whether or not moves are sente, gote or both (ambiguous) fluctuates throughout the game according to the board’s overall temperature. This talk will explain the basics of miai counting, a technique for determining values of moves that are easily composable, then use a straightforward, no-nonsense technique on several interesting examples to show when, and how long, the orthodox moves in a given position are sente, gote, or ambiguous. This technique is scalable in theory, but bounded by the requirement that the effects of all moves on the thermograph of a position be completely known before analysis; this means as computing power improves, computer programs' ability to further compute the exact move order without brute force will expand beyond the endgame and creep into the middle game.




Paul Barchilon. Developing Youth Programs.

I will speak from my experience about the various resources available for starting Go programs for youth in schools and libraries, and what we can do to create more programs.

Background: I was the AGF Teacher of the Year in 2006, and have been active in both school and library programs in Boulder, Co, for the past 8 years. As head of the AGF Mentor Committee, I have been in regular communication with hundreds of Go programs from around the nation. I also run Tigersmouth.org, the AGF's website for youth, and am the youth editor for the AGA E-Journal.




Peter Freedman.Teaching Children to Become Go Players.

Background. This paper presents lessons learned at a program in Portland, Oregon that began years ago as an after-school chess club in an elementary school.

Starting. It is not that difficult to get into a school to demonstrate the game of Go, or to start an afterschool club. But without significant systemic support success is likely to be limited. Three key points are 1. Go should be an integral part of the school program, not an add-on or an afterthought; 2. Parent support is an extremely critical element, and, 3. The goal should be to create Go players.

The Teacher. I have 2nd to 4th graders and not everyone can teach children at this age. A successful teacher must have, or develop, classroom management skills, love children, be able to connect with them, and gain their respect.

The Classroom:

Teaching methodology.

Once they learn the rules they are ready to go. However, the teacher should limit air time – even with bright, engaged children, you will quickly find you are talking to yourself if you teach too much.

Children love to win. (Don't we all?) Winning is positive reinforcement. When I play them individually I often ask, 'Do you want to win?' If so, I arrange it, i.e., talk them through to a win.

Other incentives include ladders, team practices, matches, tournaments and prizes.

Classroom Management.

If only one person is in the session, doing an opening lesson and then managing the room is a full-time job. If there are two, one can manage and the other can teach one or two children at a time by reviewing a game, teaching a joseki, etc.

Results. Some of these children now play independently on KGS and in the library, where there are boards, stones, Go problems, and the Hikaru No Go manga series. Go is now an integral part of the school program and parents, whether they play or not, understand its importance and its educational values, so they support their childrens' interest in the game.




Peter Schumer. Design of a College Go Course.

At Middlebury College, I teach two versions of an introductory go course – a twelve week writing focused first year seminar and a separate four week intensive winter term course. The main goal of each course is to introduce the students to an important and fascinating game, its rich history and cultural setting, and to foster the students' appreciation of both the creative art and deductive reasoning aspects that are so central to the game.

In this talk, the format and grading of the course, its syllabus, and the specific reading and writing assignments will be discussed. We will also briefly show how the game itself is introduced and developed. Class projects dealing with counting, ladders, tesuji and break out problems, endgame play, as well as shape and efficiency of stones will be described.

Finally, I will present an example of a course pack that I have created which includes a variety of go-related essays, history, go vocabulary, poetry, annotated example games, and many short biographies of famous go players along with these pictures to emphasize the role of go in Chinese and Japanese art.




Roxanna Duntly-Matos. "Playing Under and Pushing Through the Stones": Privileging Transnational Go Network Formation and Minority Leadership from Lower to Higher Education.

This work addresses the psychosocial relevance of the game of Go in the creation of transnational and enduring community-based networks that disrupt taken-for granted hierarchies of knowledge. (i.e. lower and higher education). As a result, it promotes leadership, commaraderie and success among minority and often-marginalized populations throughout their academic trajectories. The connection to the American Go Foundation nonprofit community efforts and the work of Yasutoshi Yasuda are used as a vehicle to foster reflexivity and humility. In turn, their work strengthens and makes informal, semi-formal and formal social relationships possible across educational hierarchies and national boundaries. Yasuda’s international efforts in conjunction with Freirean theory of emancipatory education and praxis become powerful venues for decreasing the achievement gap in a context of multicultural and multilingual validation. This is accomplished through the promotion of youth participation in the Go Cultural Ambassador Program and its resulting mentorship networks.




Siddhartha Avila. The Art of Go in Education.

A case study of the Educational and Artistic Research Center located in Mexico City. There the core work relies on artistic expressions such as painting, music, drama and dance and other practices like yoga and meditation. This develops the whole and individual aspects of children and gives deepness and meaning to their learning so they can succeed in academics. A Go program has been added to the Kindergarten and Elementary schools since learning the game is recognized as a discipline that contributes to the development of key elements of human life such as cognitive, emotional, artistic and social skills. In other words, it is highly effective in achieving the goals and solutions that a 21st century education demands of us—one of them and probably the most important being the harmonic relationship between human beings and our planet Earth.




Terry Benson. Spreading Go - Rules and Attitudes Which Make That Difficult.

Can beginners read the rules of Go and understand territory as the object of the game - without anyone to explain it to them? Usually not. Can they understand "dame" and "dead stones" without a strong player around? No. Do the rules make it clear when you can remove stones without capturing them? No. "When the players agree," confuses as much as it helps. How many beginners give up on the game because they don't know when the game is over? The failure is not in the game. It's in the rules.

Outside Asia, Go is usually learned from a set of rules or from a teacher only slightly more advanced than the beginners they teach. The Asian rule systems use undefined concepts, are at times illogical or incomplete, and are a major reason the game has not spread more.

To successfully promote Go where there are few teachers requires an area rule system - like stone counting or the AGA rules - with a clear method for completing a game. "Stone Counting Rules" - the player who can play more stones is the winner - is a simple game which is probably very close to the ancient form of Go. AGA Rules using pass stones also allow beginners to play a game to conclusion. Everything which can be captured is captured. Every question the beginner may have is answered by play on the board by the players. No referees or teachers needed. If they like, players can count just the territory and get the same result. 

As an elite game, Go prospers in Asia. Teachers are plentiful. Many people in the culture know the game and can help beginners when they are confused. Professionals write rules they understand and use terms they like. "Dead stones" may be clear to a pro, but it is not clear to most amateurs. It's a good translation of the pro term but it’s confusing and - truthfully - inaccurate. In what way are "dead stones" really dead? They often come to life. A beginner will ask, "But didn't you say they were dead?!!" The better term is "trapped stones" which may stay trapped or may not. There are other poor choices of metaphors and rules based on arbitrary decisions which confuse beginners.

The attitude in Asia toward the game and those who play it is dictated by professionals. Pros are themselves driven to improve and their livelihood depends on players who want to take lessons to improve. This elitism permeates the culture. Ask the typical Chinese, Korean, or Japanese if they play go and – unless they are shodan or higher - they will say "no." How many players would we have, if we thought the same way? In the East the focus is on high skill and on striving for that high skill. Both attitudes are horrible for promotion of the game in the West.

We should encourage people to play go at any level. Anyone who can put stones on the board and get to the end of a game "plays go". We want our 25 kyu players to proudly say "I’m a Go player." It's a good game at every level; and, in some ways, it’s better at the lower ranks because it’s less predictable. We enthusiasts are so serious that we forget that Go is a game meant for play. It's not work. Players do not have to keep improving and our attitude as promoters should be to avoid any pressure on players to get better. If they learn, great. If they don't learn, but keep playing, even better. What matters is playing the game and enjoying the game. Some players will focus on skill and excel. But every player stops getting better at some time and we want people to keep playing even if they aren't improving.

The Asian model and Asian attitude toward Go may work where there are teachers and tournaments and a go culture, but it’s dysfunctional elsewhere. We need rules and terms and attitudes which make go attractive and fun and widen the base of the pyramid of Go players.




Thomas Wolf. Basic Seki in Go.

In the first part of the talk seki are discussed that are basic in the sense that each essential chain has two liberties, there are no ko and any move would result in the instant capture of that stone and its chain. In the second part more general seki are considered where chains have more than two liberties. This part makes frequent use of graph theory.

In more detail:

The talk starts with describing deformations of seki positions that do not change the essential structure of a seki which is summarized in a so-called Basic Seki Graph. Using such graphs it is shown that basic seki have either a linear or a circular topology.

Based on this simple structure a numerical encoding of any basic seki is possible that allows a straight forward algorithm for generating all possible topologies of basic seki and thus for creating any basic seki.

The encoding is explained with several seki positions. Further examples of seki positions describe modifications that are capable of complicating basic seki.

In the last two topics of the talk seki with chains having more than two liberties are treated. In the case that all chains have the same number of liberties several rigorous results about the existence of these seki and their construction are given.

An example of Vladimir Gurvich shows a seki where chains have different numbers of liberties. Whether this position is terminal or whether still a move is possible is very hard to judge even for professional players.




Ichiro Tanioka. The Origins of 19x19 Go-Boards.

It is known that go boards in China were 17x17 until the early 4th century AD, and that, at the latest, by the time of the Sui dynasty (581-618 AD), 19x19 had become the standard. To narrow our search within that period, this presentation will focus on two factors: 1) the changing pronunciation of Chinese words and 2) the newly adopted Lunar Calendar of the South China area ("Genka-Reki") in the 5th century AD. Utilizing many non-apparent facts, the resulting hypothesis will also include the probable year that 19x19 go came to Japan-451 AD.