Thoughts on the Ages of Go

The way I see it, we can divide the last 500 years of Go history into four roughly equal periods (and a fifth that we’re currently in.)

The Heroic Age
(1612 – 1702), 90 years

Begins with the founding of the four major Japanese Go Houses and ends with the death of the first Kisei, Dosaku.

The Interregnum
(1703 – 1827), 124 years

Begins with the death of Dosaku and ends with Jowa, the second Kisei, becoming head of the Honinbo house.

The Classical Period
(1828 – 1907), 79 years

Begins with Jowa becoming head of the Honinbo house and ends with the death of Honinbo Shuei, excluding the entirety of the Shin Fuseki.

The Athletic Period
(1908 – 2017), 119 years

So-called because of the transition of Go from an “art” to a “sport”. Begins with the death of Honinbo Shuei and ends with the victory of AlphaGo over Ke Jie at the Future of Go Summit.

The Mechanical Period
(2018 --)

Begins with AlphaGo’s victory over Ke Jie.


What do you think?

6 Likes

Very interesting. Clearly you know a lot more about the history than I do, and I can’t really offer meaningful comment on the past.

However, it’s interesting that you chose the victory of AlphaGo over Ke Jie as the turning point for the age of AI. Although Ke Jie may have been considered the strongest human challenger of that time, I think the match with Lee Sedol could be argued to have had much more direct impact and cultural significance, given that it happened earlier and people generally regarded that as enough proof that the age of strong Go AI had dawned.

6 Likes

I actually thought that Lee Sedol vs AlphaGo had been in 2017 and Ke Jie vs AlphaGo in 2018. Then I read that Ke Jie vs Alpha was in 2017, and so I figured the two matches had been at opposite ends of the same year. Which is weird since I watched both of them live. I think it’s my mind rejecting how much time has passed :stuck_out_tongue:

4 Likes

Why the last 500 years though?

1 Like

Because I’m not familiar enough with Go before 1500 to make distinctions between periods. There are no undisputed kifu from before then, limited archeology, and even people talking about Go in historical records might really be talking about another game, like Chinese chess, because the term is ambiguous. The history of Go in ancient times is very murky.

5 Likes

Disclaimer: I have absolutely no idea about Go history (and hardly any about Go :woman_shrugging:).

Is there any historical accuracy/ distinct era/ highlights in Go as it expanded from China to Japan and Korea?

1 Like

I have no idea. Certainly Japan and Korea will have what they believe to be the oldest boards discovered in their countries, so they have an idea, but I don’t really know about the timepoints. I think we are talking about somewhere between 900 and 1300CE for Go entering Japan, and earlier for Korea.

2 Likes

Hey bugcat - you seem pretty knowledgeable about this topic.

Besides the names of the Go Giants that bookend each part of each era - can you throw a few more words regarding the ways in which the game itself underwent evolution? I know very little about this and I’d like to know more.

For instance, I’ve read that the legendary Chinese Go players were focused on making big moyos and winning by as large a margin as possible - winning by a narrow margin was somehow “in poor taste” and indicated that one did not have the daring and imagination of a true master, etc.

Can you share some more tales of how the game itself changed, or what elements were prioritized / de-prioritized over time? Are there some classic games that you can post SGFs of so that we can get a better idea of how the game evolved over these periods?

Please and thank you! Fascinating topic :slight_smile:

6 Likes

@bugcat Basically, please write anything you would like on the subject, you peaked my interest and I’m intrigued. :blush:

1 Like

Sure, I would love to.

As I said, the history of Go in ancient times if very unclear. We have Go boards that are 19x19 but we also have Go boards that are 17x17. Maybe 19x19 replaced 17x17; maybe they were played contemporaneously in different regions or by different social classes. We don’t know.

There are historical references to Go, but they are ambiguous. For instance, there is a famous saying by Confucius that goes something like “Don’t even gamblers and Go players have something to do with their time?” and this used (controversially) to hang in the Nihon Kiin’s headquarters. But we don’t really know when this said, if it was Confucius who said it, or if he was even talking about Go. The term he uses could well have referred to Chinese chess, or perhaps both Chinese chess and Go, or maybe even a game that no longer exists.

For a number of reasons, it seems to me that the claim that Go is 4000 years old is an exaggeration. I think a game of Go that we would recognise as having the same basic principles is probably no older than around 2000 years. But earlier than that, the same equipment may have been used to play different games like gomoku, or it’s been suggested that there was a use in divination.

Books have been written that are said to be from this or that date, but when I looked into it they often turned out to only exist now in copies that are much younger – like is the case with the Vedas. So you can’t really determine their age without having a very good understanding of the linguistic evolution of Chinese languages. At some point, Go spread from China to three other countries: to Tibet, to Korea, and to Japan, probably in that order. I envisage this happening somewhere around 1000CE but I don’t have any proof to hand.

Over the next several centuries isolation in four countries produced four different games. On mainland Asia, games always began with a single mandatory opening. In China, this opening was fairly simple: you had to play diagonally crossed 4-4s. In Tibet’s Tibetan Go and Korea’s sunjang baduk, more elaborate placements were used. The Tibetan set stones were, in fact, much larger than the stones that could be freely placed. In all these three variants, the set stones limited building potential and encouraged fighting. In Japan, however, there was no mandatory opening. However, in both Japan and China there were a common set of required handicap openings which we still use today, which are centred on the 4-4 points. This is especially interesting since, until about 1920, the Japanese largely believed the 4-4 to be inferior to the 3-4.

Another difference was in scoring the game. The Chinese rules are what is called area scoring, which says at the end of the game “What is the maximum amount of stones that you could theoretically fit on the board? That is your score.” It was the same in ancient China, with one difference: in the theoretical stone-placing period, you were not permitted to fill in the eyes of your own groups. Thus, each group on the board actually cost a two-point “eye tax” – this again encouraged fighting, since every time you split your opponent’s stones you automatically took two points away from his score. It’s not clear when the Japanese abandoned area scoring. I think it would be relatively easy for a strong player to tell you whether or not a dame-filling phase had been left off a scored kifu, but I am not a strong player. It’s my impression that the Japanese were using territory scoring (in which only the points “within the walls” are counted) before the first reliable kifu we have from that country.


Go history becomes much richer, abruptly, in the 17th century. In 1612 the young Tokugawa Shogunate, which had been in power for less than a decade, founded the first professional Go system. It was modelled on what was called iemoto, a martial arts tradition in which four houses would compete for victory. The four Go houses were: the Honinbo house, founded by Sansa; the Hayashi house, founded by Monnyusai; the Inoue house, founded by Nakamura Doseki; and finally the Yasui house, founded by Santetsu. The houses operated similarly to sumo stables today: the head of the house, the master of his craft, would educate a large number of boarding pupils in a single residence. This system wasn’t perfect; in fact as a result of this close-proximity living many students died of cholera and tuberculosis.

From the beginning the Honinbo house was dominant and for the next three centuries it largely remained so. There were also three minor houses: Sakaguchi, allied with Yasui; Hattori, with Inoue; and Mizutani, with Honinbo. It’s not clear when they were founded or how they exactly slotted into the system. The players of the houses competed in the Castle Games, which were played (or, perhaps, only re-enacted) in the castle of the local shogun (warlord). These games were untimed and could take several days to complete. My suspicion is that it may have been considered impolite to resign, since sometimes I see games won by over ten points, which is a very large margin for a professional player.

The heads of the houses competed for the title of Meijin, which meant “virtuoso”. The Meijin was in theory the strongest player in Japan, and the only holder of a nine-dan rank (although I have before questioned whether the concept of a ninth dan rank existed.) The Meijin was sometimes, but not always, the Godokoro – this was an official post which bestowed a number of powers. Throughout the 17th and 18th centuries there was generally a Meijin.

The games of early 17th century Japan look quite alien to the modern eye. The placement of the stones tends to be very “calm” or “passive” compared to today. The players are cautious engaging the opponent at close range; they are happy to accept an “even result” and to take gote in order to solidify their groups. Third line moves took up a more greater proportion of the opening and early midgame, with the centre being thought of less favourably. However, every now and again a strikingly “modern” joseki is played, like the taisha.

The exception to this rule is our first giant, Honinbo Dosaku. Dosaku was extremely strong – for centuries he has been considered as the first Kisei, or “Go Saint”. Dosaku was perhaps the first ever player to really master the ideas of light play, efficiency, and the value of influence. He the inventor of the tewari technique, which uses move-switching to test whether the plays in a sequence are working their hardest. Into the mid-20th century, professionals put great value on studying his games.

My knowledge of Chinese games from this time is quite poor, but I’m aware of one Guo Bailing who authored three books on handicap strategy (Guanzi Pu, Sanzi Pu, and Sizi Pu), the latter two of which have been translated into English. The Chinese system at this time was much more amorphous than that of the Japanese. The primary unit was the guosho, who was acknowledged as the best player within their town or region. Exceptional players like Guo Bailing were referred to as daguosho; very strong players would travel around defeating other guosho to improve their reputation. But, as far as I’m aware, they didn’t receive any stipend from the government.


The 18th century was not a very interesting time in Japanese Go. Everyone considered themselves to be weaker than Dosaku. Strategy remained largely the same, focused on third-line moves to secure territory. Attack and defence were not seen as very important. In China, however, the opposite was true. There were two master players at this time: Fan Xiping and Shi Dingan. Fan was a quick-playing, adventurous type; Shi Dingan was more ponderous and calm at the board. All Chinese games still, however, used the ancient cross-hoshi opening and were centred on fighting skill.

In Japan, the first signs of a new era approaching were shown by Yasui Senchi Senkaku, now referred to as the Grandfather of Modern Go. Playing at the turn of the 19th century, he was the first player to display a powerful attacking style. I’ve reviewed a few of his games, one in detail, and the way he went after his opponents’ group using central influence was very impressive. I can’t stress highly enough the difference between Yasui Senchi’s new technique of hard attack and the older, calmer style that it unleashed itself on.


In the early 19th century, our second Kisei takes his place on the stage. His name was Honinbo Jowa. He was an unusual figure: unlike most Go players, he was low-born, perhaps the son of a fish merchant. He was also rather old, becoming a professional at about 30 years old. He was, however, the heir to Senchi’s style: Jowa was a brilliant reader, a master of life and death positions, and committed to attack. He was very comfortable in fighting positions; he found it very easy to settle his groups under pressure; and he always found the time to harass his opponent’s groups. I like to say that Senchi invented the attacking style, but Jowa brought it to its rightful position. In 1827 he became the head of the Honinbo house. Jowa’s games are brilliant; they are full of incredible fights and astonishing struggles for survival. To this day he holds the record for the most dead stones in a professional game, which I believe was around 200.

Here began a convoluted and disreputable tale in which Jowa attempted to manipulate various senior figures of the Go community in order to take the title of Meijin-Godokoro, which (although he was acknowledged as the stronger player of his time) was denied him for political reasons. You can read all about this in Invisible: The Games of Shusaku. Jowa became the Meijin in 1831, but a mere eight years later his engineering came to light and he was forced to retire.

His heir was Josaku (who was unexceptional), but his heir was Shuwa, who was the opposite of Jowa. Shuwa was excellent at calm and quiet territorial play. He was very good at simply taking points beneath strong groups, repeatedly throughout the game, generally avoiding large fights. To say that he is one of Cho Chikun’s favourite players conveys his style well enough. Shuwa was in turn succeeded by Shusaku, the third Kisei. Shusaku is regarded highly for two reasons. Firstly, because of his great respect for “honour” in regard to his parents and employers. Secondly, for his unbeaten streak playing Black in the Castle Games (which were, of course, played without komi.) Personally I can’t really identify a style in Shusaku’s play, and I find his games a little dull. Certainly Shusaku was a strong player, but I’m not convinced that he was stronger than Shuwa: for reasons of etiquette he refused to play against Shuwa with the white stones. Shusaku died very young from cholera or tuberculosis, aggravated by grief-fasting.

Shuwa’s heir was Shuho. Shuho, like Jowa, had an attacking style: however, I get the impression that Shuwa’s technique was more controlled and restrained than Jowa’s. There isn’t the impression that Shuwa himself has risked so much on the fight that he has induced; rather, he seemed to bully the opponent’s groups in a way that minimised damage to himself. Shuho had wanted to participate in the Castle Games, however they were shut down in 1861 as a result of the Meiji Revolution. So, Shuho toured Japan with his half- (or perhaps adoptive) brother Shuei, playing Go. They were probably also giving many teaching games for money. Shuho wrote an opening book called Hoen Shinpo. I’ve been told that this book contained new ideas for an “expansive” fuseki, but I’ve never read it. Shuho also tutored the first ever serious Western player, Oscar Korschelt, after whom our A1 coordinates are still named today.

At this time, the modern structure of professional Go began to show itself. Shuho collaborated with Nakagawa Kamesaburo (one of Jowa’s sons) to create, in 1879, an organisation called the Hoensha, which was the first professional organisation outside of the iemoto system. The Hoensha published Igo Shinpo, the very first Go magazine. This annoyed Shuei, who was a traditionalist; the two fell out and later reconciled. Shuho then died, leaving Shuei as his heir.

Shuei was the last of the great classical players. He was, in my opinion, one of the strongest players in history. Interestingly, in contrast to today’s professionals who peak young and burn out quickly, Shuei has a mediocre player until he was around fifty years old. In the 1890s, though – perhaps as a result of having powerful students – he became incredibly strong. His style was unique. He was neither a hyper-aggressive attacker like Jowa, nor a calm bean-counter like Shuwa.

To quote Takagawa Kaku: “In particular his games as White after he reached 7-dan hide a fathomless strength amidst a serene and well-balanced flow, which we can perhaps describe as flexible on the outside, unbendable on the inside. In brief, the characteristic of Shuei’s skill at go is not just about local things such as his skill at evaluating positions or his skill in seeing moves, but his mastery of go itself.”


I’d really love to go on, but I don’t have the energy ~o~

19 Likes

That post deserves a round of applause! Great stuff!

Thanks for bringing such complex subject matter to life!

Please feel free to revisit this topic as time and energy allow - I’d love to read more any time.

4 Likes

More.

(whenever you can :slightly_smiling_face:)

2 Likes

In particular, I would be very interested what brought you to think that go is much younger than we think. :slight_smile:

3 Likes

I found this interesting history of the rules that explains how the ancients used stone scoring and how this might have morphed into (as the common ancestor of both) the modern Japanese and Chinese rules that we use to today.

3 Likes

I thought LS-AG was 2015. I checked and it was march 2016, a year before Ke Jie-AG (if im not missing anything).

1 Like

As far as I’m aware, there is no good evidence for it being anything like 4000 years old.

What we’ve got to go on when investigating the earliest history of Go, according to Sensei’s Library and Wikipedia, is:

  1. A couple of legends placing its invention in the 3rd millenium BCE, attributing it to Emperor Yao or Shun.

  2. A mention by Confucius in Analects ca. 6th century BCE; I’ve been told before that the term he used was ambiguous and could have also referred to other games like Chinese chess

  3. In the 3rd century BCE a Go player was apparently mentioned by an author Mengzi in a book called Mencius. Again, it seems that another game could have been referred to. There is a lot of subtlety and ambiguity in terminology (eg. in the 19th century Europeans called Go chess and the Chinese called chess Go) and I’ve got very little skill in Chinese.

  4. A book called Yi Zhi by Ban Gu, purportedly from the 1st century CE. I don’t know anything about this but it seems doubtful that any surviving copy is more than a few centuries old, so dating relies on linguistic assessment of ancient Chinese.

  5. A broken piece of a pottery Go board from “206 – 24 CE”

  6. A complete Go board dated to “25 – 220 CE”

So to me, this points to a possible origin or popularisation of the game about 2000 years ago. Compare what evidence we have to the proof we have of the twenty squares game or “royal game of Ur” being played widely in the Middle East from 3000 BCE onwards. In this case we have a number of boards from across a wide area (but admittedly, a description of the rules only arrives very late in something like 100 BCE). But this might just say that Go, a perfect-information “art”, was played by the upper classes of China whereas the twenty squares game, a chance-based game well-suited to gambling, was popular with a large group of lower-class Middle Easterns.

As I said, I find the majority of history of Go to be quite opaque and sparsely-documented, and I didn’t make any concrete assertions about who was playing it where and when because I don’t think they can be made.

5 Likes

Here is an excerpt from this GuoXue article discussing the origin of Go:

据春秋时期的《左传》记载:“卫献公自夷仪使与宁喜言,宁喜许之。大叔文子闻之,曰: ‘……今宁子视君不如弈棋,其何以免乎?弈者举棋不定,不胜其隅,而况置君而弗定乎?必不免矣。……’”
这里包含了众所周知的一个成语“举棋不定”。文中的“弈者”指的是下围棋的人,在古代“弈”专指围棋,许慎《说文解字》就有解释:“弈,围棋也,从丌,亦声。”。《左传》大约成书于公元前548年,这里记载的故事发生在公元前559年。具体的故事内容,大家可以去翻看一下《左传》。可以说明的是,在据今二千五百年前,围棋已经为大家所相互熟悉的事物了。值得一提的是,《左传》中有关围棋的记载是迄今为止最早的文字记载。
在这个基础上再做一个简单的推论,围棋的形成与发展再到为大众所熟知,必然需要一个相对漫长的时期,也就是说,围棋的起源要在这个时间点上溯若干年,也许是500年,也许是1000年,这正是需要有心人来考证的事情。

I won’t bother translating the entire excerpt (Google translate might be able to get you close enough to understand some of it) but the more interesting bit is the discussion of the story in Zuo Zhuan which is the earliest reliable reference to Go.

The story itself was written in 548 BCE and is about an event that took place in 559 BCE. Note that the character used for Go in this document is 弈 a character that is used specifically for weiqi.

The quoted article points out that the way Zuo Zhuan references Yi points to Weiqi being widely known by that time.

1 Like

Both Wiktionary and Jisho define 弈 as Chinese chess, though.

https://en.wiktionary.org/wiki/弈

(And yeah, I know Jisho is for Japanese and that Wiktionary only has Japanese and Korean translations, and we’re talking about Chinese.)

1 Like

I think that must be more of an issue of poor translation into English. Even the Chinese dictionaries define it as the ancient name for weiqi and then have the English translation as Chinese chess

4 Likes

In ancient times “弈” specifically referred to Weiqi. Xu Shen’s “Shuo Wen Jie Zi” [early 2nd century Chinese dictionary, Han dynasty] has the explanation: “Yi, weiqi, from 丌, 亦 sound.”

4 Likes