Table of Contents

What Is Go (Game)?

Go is a strategy board game for two players who take turns placing black and white stones on a grid, attempting to surround more territory than their opponent. The rules are extraordinarily simple — you can teach them in five minutes. The strategy is extraordinarily deep — people have been studying it for over 4,000 years and haven’t exhausted it. This combination of simplicity and depth is what makes Go unique among board games and why it’s often described as the most intellectually demanding game humans have invented.

The Rules (All of Them)

Go’s rules fit on an index card.

The board is a 19x19 grid (beginners often start on 9x9 or 13x13). Two players alternate placing stones on the intersections of the grid lines. Black goes first.

Liberties are the empty adjacent intersections around a stone or connected group of stones. A stone with no remaining liberties is captured and removed from the board.

Ko rule prevents infinite loops. You cannot make a move that recreates the exact board position that existed after your previous turn.

The game ends when both players pass consecutively, agreeing that no more useful moves remain. Territory (empty intersections surrounded by your stones) and captured stones are counted. The player with more territory wins. White receives a bonus (komi, usually 6.5 or 7.5 points) to compensate for Black’s first-move advantage.

That’s it. Everything else — the strategy that fills thousands of books and demands decades of study — emerges from these few rules.

Why It’s So Deep

The board starts empty, and each player places one stone per turn. With 361 intersections, the first move alone has 361 options. The second has 360. The number of possible game sequences is a number so large it has no meaningful comparison — more than the number of atoms in the observable universe by an absurd margin.

But complexity alone doesn’t make a great game. Go’s depth comes from the interaction between local and global considerations. Every move has both a local effect (strengthening a position, attacking a group, defending territory) and a global effect (influencing the balance of the entire board). Great Go players see both simultaneously — understanding how a move in one corner affects fights on the opposite side of the board.

The game also features a fundamental tension between influence (stones that project power across open areas) and territory (stones that secure specific areas). Playing for territory is safe but can leave you without influence. Playing for influence is bold but requires converting it into territory eventually. Balancing these competing demands is one of Go’s central strategic challenges.

History

Go (called weiqi in Chinese, baduk in Korean, and igo in Japanese) originated in China at least 2,500 years ago, with traditions claiming origins over 4,000 years ago. Confucius mentioned it. It was one of the four accomplishments expected of Chinese scholars (along with calligraphy, painting, and music).

The game reached Korea and Japan by the 5th-7th centuries CE. In Japan, Go became deeply embedded in culture. The Tokugawa shogunate (1603-1868) established four Go houses (schools) that competed for government patronage. Professional players held honored positions in society. The title of Meijin (master) could only be held by one person at a time.

Modern professional Go is concentrated in East Asia. China, South Korea, and Japan each have professional Go associations with hundreds of ranked players. Top professionals are celebrities — their games are broadcast live on television with expert commentary.

AlphaGo and AI

For decades, Go was considered the last great bastion against artificial intelligence. Chess fell to IBM’s Deep Blue in 1997, but Go’s vastly larger search space and reliance on intuitive pattern recognition seemed to resist computational approaches.

Then DeepMind’s AlphaGo defeated European champion Fan Hui in October 2015 and world champion Lee Sedol in March 2016. The match — particularly Game 4, where Lee Sedol played a brilliant move that even AlphaGo’s creators acknowledged was extraordinary — captivated millions of viewers worldwide.

The later version, AlphaGo Zero, learned to play Go entirely by playing against itself — no human games, no human knowledge. It surpassed all previous versions in 40 days. Its successor, AlphaZero, mastered Go, chess, and shogi from scratch.

The impact on human Go has been profound. Professional players now study AI moves and have adopted strategies that were previously considered unorthodox or incorrect. The AI hasn’t made human Go obsolete — it’s expanded it, revealing new possibilities in a game that people thought they’d been exploring thoroughly for millennia.

Learning Go

Go is easy to start and infinitely deep to continue. Beginners on 9x9 boards can learn the basic flow of the game in a single session. The progression to 13x13 and eventually 19x19 introduces new strategic layers — the full-size board requires balancing attention across a much larger space.

Ranking uses a system similar to martial arts. Beginners start around 30-25 kyu. Players advance through kyu ranks (decreasing numbers) until reaching 1 kyu, then enter dan ranks (increasing numbers: 1 dan, 2 dan, etc.). Amateur ranks go to 7 dan. Professional ranks go to 9 dan (the highest title).

Online platforms (OGS, KGS, Fox Go) allow playing against opponents of any level worldwide. AI tools (KataGo, Leela Zero) provide game analysis and review.

Go clubs exist in most cities and at many universities. The social aspect — playing face-to-face, reviewing games with stronger players — accelerates learning and makes the game more enjoyable.

Why Go Endures

Go has survived for over 4,000 years because it achieves something almost no other game has: a depth-to-rules ratio that seems infinite. The rules are simple enough for a child to learn. The strategy is deep enough to sustain a lifetime of study. And the aesthetic pleasure of a well-played game — the balance of stones on the board, the elegance of a strong position, the beauty of an unexpected move — gives Go a quality that transcends competition. It’s a game, an art form, and a meditation, all on a wooden board.

Frequently Asked Questions

How old is Go?

Go originated in China over 4,000 years ago, making it the oldest board game still played in its original form. Chinese legends attribute its invention to Emperor Yao (2356-2255 BCE), though the earliest reliable historical references date to around 500 BCE. The game spread to Korea and Japan over the following centuries.

Is Go harder than chess?

In terms of computational complexity, Go is vastly more complex. A standard Go board has approximately 2.1 x 10^170 legal positions — compared to roughly 10^47 for chess. This complexity is why AI took decades longer to master Go than chess. In terms of difficulty for human players, both games require years to master but exercise different cognitive skills.

Why was AlphaGo so significant?

DeepMind's AlphaGo defeated world champion Lee Sedol 4-1 in March 2016, achieving something many AI researchers thought was decades away. Go's enormous complexity and reliance on pattern recognition and intuition made it a benchmark problem for artificial intelligence. AlphaGo used deep neural networks and reinforcement learning — techniques that have since been applied across AI research.

Further Reading

• American Go Association
• International Go Federation
• Sensei's Library — Go Encyclopedia