Go (game)
| Go | |
|---|---|
 | |
| Chinese Name | |
| Pinyin | Wéiqí (wei2qi2) |
| Wade-Giles | Wei-ch'i |
| Chinese | 圍棋 |
| Japanese Name | |
| Romaji | Igo |
| Kanji | 囲碁 |
| Korean Name | |
| Revised Romanization | Baduk |
| McCune-Reischauer | Paduk |
| Hangul | &#바둑 |
Go is a strategic, two-player board game originating in ancient China between 2000 BC and 200 BC.
Go is highly popular in East Asia, and play on the Internet has served to greatly increase its popularity throughout the rest of the world in recent years. In China it is called weiqi, in Korea its name is baduk, and in Japan igo (ee-go), which gave rise to the English name Go from the Japanese character 碁. The Chinese name translates to "Chess of Surrounding (One's Opponent)".
Philosophy
According to the history of Go, the game was used as a teaching tool after an ancient Chinese Emperor designed it for his son, who he thought needed to learn discipline, concentration, and balance. The son went on to become the first great player, a good emperor, and a balanced human being. Another suggested genesis for the game states that in ancient times, Chinese warlords and generals would use pieces of stone to map out attacking positions.
Go is deep, as playing against any stronger player will show (Depth of the game established by ELO ranking in Go) With each new level (rank) comes a deeper appreciation for the subtlety involved, and for the insight of stronger players. Beginners often start by randomly placing stones on the board, as if it were a game of chance — and they inevitably lose to experienced players. But soon an understanding of how stones connect to form strength develops, and shortly afterward a few basic joseki (corner sequences) are understood. Learning the ways of life and death helps to develop one's situational judgement.
Further experience yields an understanding of the board, the importance of the edges, then the efficiency of developing (in the corners first, then walls, then center). Soon, the advanced beginner understands that territory and influence are somewhat interchangeable — but there needs to be a balance. Best is to develop more or less at the same pace as the opponent, in both territory and influence. This intricate struggle of power and control makes the game highly dynamic.
Rules
Basic rules
Optional rules
Optional Go rules might set the following among other aspects:
- compensation points ("komi") for the right to start alternate play
- compensation stones ("handicap") placed on the board before alternate play (this allows players of different strengths to play interesting games)
- thinking times
For details see Go rules. Also see Go rules of play and strategy.
Strategy
Basic strategic aspects include the following:
- Connection: Keeping one's own stones connected means that fewer groups need defense.
- Cut: Keeping opposing stones disconnected means that the opponent needs to defend more groups.
- Life: This is the ability of stones to avoid their removal. Usually life requires at least two "eyes" for a group of stones.
- Death: The absence of life, resulting in the removal of a group.
Also see details of Go strategy and tactics, Go rules of play, and optional rules.
Nature of the game
Although rules of Go can be written so that they are very simple, the game strategy is extremely complex. Go is a complete-knowledge, deterministic, strategy game like chess, checkers (draughts), and reversi, although its depth exceeds even those games. Its large board and lack of restrictions allows great scope in strategy, as decisions in one part of the board are influenced by a seemingly unrelated situation in distant parts of the board, and moves made early in the game can shape the nature of conflict hundreds of moves later.
The game emphasises the importance and tensions of balance on multiple levels. To secure an area of the board, it is good to play moves close together; but to cover the largest area one needs to spread out. To ensure one does not fall behind, aggressive play is required; but playing too aggressively leaves weaknesses undefended that can be exploited. Playing too low (close to the edge) secures insufficient territory; yet playing too high (far from the edge) allows the opponent to invade. Many people find the game attractive for its reflection of the polarities found in life.
See Go strategy and tactics for an introductory explanation of how to play well, and the Go demonstration game.
It is commonly said that no game has ever been played twice. This may be true: On a 19×19 board, there are about 3361×0.012 = 2.1×10170 possible positions, most of which are the end result of about (120!)2 = 4.5×10397 different (no-capture) games, for a total of about 9.3×10567 games. Allowing captures gives as many as 107.49×1048 possible games, most of which last for over 1.6×1049 moves! (For two comparisons: the number of legal positions in chess is estimated to be between 1043 and 1050; and physicists estimate that there are not more than 1090 protons in the entire visible universe.)
Computers and Go
Although attempts have been made to program computers to play Go, success in that area has been moderate at best - development in this area has not reached the level of Chess programs. Even the strongest programs are no better than an average club player, and would easily be beaten by a strong player even getting a nine-stone handicap. This is attributed to many qualities of the game, including the "optimising" nature of the victory condition, the virtually unlimited placement of each stone, the large board size, the nonlocal nature of the Ko rule, and the high degree of pattern recognition involved. For this reason, many in the field of artificial intelligence consider Go to be a better measure of a computer's capacity for thought than chess. See the article on Computer Go for details.
Use of computer networks to allow humans to meet, discuss games, and play one another, although generally considered inferior to face-to-face play, is becoming much more common. There are servers and software to facilitate this; see Additional Resources below for more information.
Other board games commonly compared with Go
Go appears to stand apart among games in its rules and gameplay; it is difficult to find another board game which could be considered of the same "family" as Go. However, on learning about the game, people will attempt to compare it with other games they may already have experienced. This is a list of some games that are played with similar equipment or come from the same area.
- Variations of chess
- Western chess: This game dominates Western culture as the pinnacle of strategic game play; its history in the culture stretches back many centuries. By comparison, Go has only been known to Western culture as a challenging strategic game since the 1950s.
- Shogi: Early Western literature often made the error of referring to Go as "Japanese Chess". The Japanese do have their own variant of Chess, called Shogi; it is much more similar to the other Chess variants than to Go. Shogi schools were founded in Japan about the same time as Go schools, but the game never received as much favour as Go.
- Xiangqi: This is the Chinese variant of Chess, usually called "Chinese Chess" by Westerners. Like most Chess variants, it has great depth of strategy, but bears few similarities to Go in game play.
- Janggi: This is the Korean variant of Chess, usually called "Korean Chess". It is also very different from Go in game play. Go and Janggi are the two main board games played in Korea, but Go is arguably more popular as there are professional Go players and Cable Television channels devoted to Go.
- Hex is a connection game which is often compared to go; though it can have some depth, its object is entirely different.
- Reversi: Marketed by one game company as "Othello", Reversi bears superficial similarity to Go, with black and white circular pieces, an undifferentiated grid for a board, simple rules, and a goal of covering more of the board than the opponent. The game play is quite unlike Go, however, and the depth of strategy in Reversi is not comparable to Go.
- Gomoku, Renju and Pente: Played with the same equipment as Go (a 19x19 grid, black and white stones), in these games the goal is to create five stones in a row. The rules are thus completely unrelated, and the game style is much shorter and less strategic than Go.
- Alak is a Go-like game restricted to a single spatial dimension.
Traditional Go game equipment
Although one could play Go with a piece of card for a board and a bag of plastic tokens, Go players pride themselves on their game sets.
The traditional Go board (called a goban in Japanese) is solid wood, about 15-20cm thick, and stands on its own attached legs. They are preferably made from the rare golden-coloured Kaya tree (torreya nucifera), with the very best made from Kaya trees up to 700 years old.
Players sit on reed mats (tatami) on the floor to play. The stones (go-ishi) come in matching solid wood pots (go-ke) and are made out of clamshell (white) and slate (black) and are extremely smooth. The natural resources of Japan have been unable to keep up with the enormous demand for the native clams and slow-growing Kaya trees; both must be of sufficient age to grow to the desired size, and they are now extremely rare at the age and quality required, raising the price of such equipment tremendously.
In clubs and at tournaments, where large numbers of sets must be maintained (and usually purchased) by one organization, the expensive traditional sets are not usually used. For these situations, table boards (of the same design as floor boards, but only about 2-5cm thick and without legs) are used, and the stones are made of glass rather than slate and shell. Bowls will often be plastic if cheap wooden bowls are not available. Plastic stones could be used, but are considered inferior to glass as they are generally much lighter, and most players find that not even the cheaper price justifies their unpleasantness.
The dimensions of the board (traditionally the grid is 45.45cm long and 42.42cm wide, with space beyond to allow stones to be played on the edges and corners of the grid) often surprise newcomers: it is not a perfect square, but is longer than it is wide, roughly in the proportion 12:11. Two reasons are frequently given for this. One is that when the players sit at the board, the angle at which they view the board gives a foreshortening of the grid; the board is slightly longer between the players to compensate for this. Another reason is that the Japanese aesthetic finds any structure which is perfectly symmetrical to be in bad taste, and the board is not made a perfect square for this reason.
Traditional stones are made so that black stones are slightly larger in diameter than white; this is probably to compensate for the optical illusion created by contrasting colors that makes the white stones appear larger on the board than black stones. The difference is slight, and since its effect is to make the stones appear the same size on the board, it can be surprising to discover they are not.
The bowls for the stones are of a simple shape, like a flattened sphere with a level underside. The lid is loose-fitting and is upturned before play to place opponent's stones captured during the game. The bowls are usually made of turned wood, although small lidded baskets of woven bamboo or reeds make an attractive cheaper alternative.
There is even an art to placing a Go stone, held between the tips of the outstretched index and middle fingers and striking the board firmly to create a sharp click. Many consider the acoustic properties of the wood of the board to be quite important. The traditional goban will usually have its underside carved with a pyramid called a Heso recessed into the board. Tradition holds that this is to give a better resonance to the stone's click, but the more conventional explanation is to allow the board to expand and contract without splitting the wood. A board is seen as more attractive when it is marked with slight dents from decades – or centuries – of stones striking the surface.
The Go world
Ranks
In countries where Go is popular, ranks are employed to indicate playing strength. From about the sixteenth century, the Japanese formalised the teaching and ranking of Go. The system is comparable to that of martial arts schools; and is considered to be derived ultimately from court ranks in China.
Beginning players today start at a rank of between 25 and 30 kyu. The kyu ranking then decreases as the player becomes stronger, dropping down to 1 kyu. Since beginners will commonly progress through elementary concepts quickly, it is difficult to set a solid kyu ranking for new players. Players who have progressed through the kyu ranks and passed the 1 kyu mark are then ranked at 1 dan, sometimes called shodan. The player then could advance through the amateur dan ranks up to amateur 6 dan, which only few players achieve. At that that playing level the ranks for professionals start with professional 1 dan going up to 9 dan (also called ping or p' as in 9 p to avoid confusion between a 1p professional 1 dan and a weaker amateur 6 or 7 dan.. The distinction between each amateur rank is, by definition, one handicap stone.
In other words, the difference in rank between two players is theoretically equal to the number of handicap stones required for each player to have an even chance of winning. Beating this handicap consistently is the indicator that a player's strength has improved, and his rank should be adjusted upward by one stone, thus changing the number of handicap stones required.
In China, Japan, and Korea, there are two distinct ranking sets, one for professional players (who receive a fee for each game they play, bonuses for winning, and fees for other related activities) and one for amateur players. In the Japanese professional ranking system, distinction between ranks was traditionally considered to be 1/3 of a handicap stone (making the difference 3 pro dan equal to one amateur dan). The strength of new professionals (1 dan) was usually comparable to that of the highest ranked amateurs. Currently the professional ranks are assumed to be more bunched together, covering not much more than two amateur dans; so that pro 1 dans win some games against 9 dans. There are also a number of amateur players acknowledged as having pro 6 dan understanding of the game.
In Japan, amateur ranks are only recognised up to 8 dan (before the year 2001, only amateur ranks up to 7 dan were recognised). Within European Go Federation, ranks are recognised up to 7 dan. The American Go Association currently recognizes rankings up to 6 dan amateur.
Player pools that do not regularly mix (such as different countries, or online versus real-life player groups) often result in divergent playing strengths for the same rank level. Players asked to give their rank will therefore often qualify it with "in my country" or "on this Internet server".
In Germany and The Netherlands a "classes"-system (German: "Klassen") was established. It comprised a further subdivision into Ky?Dan halfgrades with classes 18 and 17 = amateur 1 dan with the 17 being on the stronger side. It is still in use for club ladders etc. where you pro-/demote after a won/lost game.
Go Ranking with ELO
As in other sports where you cannot hold an all-against-all championship to establish who is strongest (Chess, statistical Backgammon, physical Racquetbal etc.) a statistical method is employed. The ELO rating system is nowadays quite used in Go (e.g. ELO by the EGF). An average 1 dan should have Go Rating about 2100 and the difference between grades is set to 100. Hence, a rating of 2100 would coincide with an amateur 1 dan who has a 2/3 winning chance against just anyone with a recent rating of 2000 (an average 1 kyu, a recently improved 2 kyu or, e.g. a 1 dan who has lost many games and so rating points).
Similarly the same player only has a 1/3 winning chances against any player with a recent ELO rating 100 points higher than oneself. Here one can see the difference between ranks kyu/dan, being the highest personal achievements and an actual rating. A player having been 3 dan 15 years ago will know the necessary strategies and concepts for teaching weaker players, but when it comes to actual play this person will certainly lose rating points when again entering the tournaments (where the rating points are measured) against players with the same grade.
Winning Probabilities The rating indirectly represents the probability of winning against other rated players. This probability depends only on the difference between the two players' ratings as follows:
rating probability
difference of winning
---------- -----------
400 .919
300 .853
200 .758
100 .637
50 .569
0 .500
-50 .431
-100 .363
-200 .242
-300 .147
-400 .081
Converting Elo Ratings into Go Ranks The Elo system can be used establish a Go grade (kyu/dan) by a national promotion commission or implement Go ranks on Go servers (on the internet).
Points Go rank ------ ------- 2300 3 dan 2200 2 dan 2100 1 dan
2000 1 kyu 1900 2 kyu 1800 3 kyu
1500 6 kyu
1000 11 kyu
500 16 kyu
0 21 kyu
-100 22 kyu -200 23 kyu -300 24 kyu etc.
Because of the low confidence in (lower) amateur ratings (high fluctuation in the outcome) but high confidence in pro ratings (stable, consistent play), the K-value in the formula
where Rn = new rating, Ro = old rating, S = score, Se = expected score and the constant C is 30 and 20 for players below ELO= 2.400 (e.g. 4 dan) and set C = 10 for players above. For Go, the variance of the Normal Probability Function is set at 100 ELO points and the constant K depends on the rating of the players involved.
In the ELO system the total number of ELO points of the player pool is only changed by changes of the pool and the players itself, e.g. newcomers and improvers. Because Go is a zero-sum, full information game the ELO points won by the winner are the ones lost by the loser and its maximum is the constant K above.
Because of the few played ranked even games between amateurs and professionals, it is difficult to establish a single ELO table where both pools are represented. Using the rough equivalence between handicap stones and rating difference (1 stone = 100 ELO points) a professional 9 dan (9p) would have +/- 2900 ELO points. For the winning chances of amateurs against professionals in even games the following has to be considered.
In Go the winning chances are not fully cummulative: On average the rating differences will be congruent with the ranking differences in the same system (same club, country or server). Go is usually played on even because players prefer it and resp. tournaments are organized that way. However, besides that point, the table with the winning chances set over the rating differences will only work in practice up to a small ranking difference correctly (ELO difference of up to 300-400 points. If players with a higher inequality, , e.g. ELO difference = 900, normally requiring a handicap of 9 stones would meet in an even game, then the chances of the weaker player are not even a mere 0.0006 (0.06%) to win this game, rather they are zero! This can be explained by the fact that the probabilities of risks and gains can be moderated by the stronger player.
While in chess you had to take the full risk and must capture the opponent's king and, more difficult, avoid the possibility of draw, which in Go is exluded by the komi. An average game of Go lasts for 240 moves (= 120 double moves; compared to 50 in chess). There are more opportunities for weaker player to make sub-optimal moves for confirming the result (a loss). The stronger player would win on average with + 140 points (empirically established) in an even game with a person s/he has even chances on 9 stones.
The stronger player gains already with 1/2 points more (cf. komi) a win. By moderating the risk exposure and steering for a lower win in points (not a 50% chance for 140 points), this player increases the probability for this win. This behaviour increases the chance from 99.94% (1-0.06%) to 100%.
The ability of transforming small advantages (e.g. 5 points; eqivalent to the first move or half a grade difference) into sure wins increases much with playing strength. What is true for amateurs with an ELO difference of 900 holds as much for professionals with a smaller ELO difference. That is also the reason why the constant C in the formula above is set smaller for professionals.
Depth of the game
- The ELO rating depth also states something over the depth of the game. The total depth of a game is defined by two end points of the range of skills: the total beginner and the theorethical best play by an unfallible, allmighty creature. The total beginner, yet playing according to the simple rules can safely be set at 30 kyu. Theoretical best play could result in the strength of an imaginable 13 dan according to measurements of standard deviations among professional games.
Only taking 20 kyu and 9 dan as endpoints makes Go until now the deepest game. A rating difference of 2900 ELO points from (Gu Li) to a 20 kyu with 100 ELO points is a difference in insight into the game by 29 times the standard deviation.
Top players
Although the game was developed in China, in recent centuries the strongest players in the world have come from Japan. However, top players from China (since the 1980s) and South Korea (since the 1990s) have reached the same or an even higher level. Nowadays, top players from these three countries are of comparable strength, although top Korean players seem to have an edge, dominating the major international titles. All three countries have a number of professional Go tournaments. The top Japanese tournaments have a prize purse comparable to that of professional golf tournaments in the United States. Tournaments in China and Korea are less lavishly funded.
Players from other countries have traditionally been much weaker, except for some players who have taken up professional courses in one of the Asian countries. This is attributable to the fact that details of the game have been unknown outside of Asia for most of the game's history. A German scientist, Otto Korschelt, is credited with the first systematic description of the game in a Western language in 1880 AD; it was not until the 1950s that Western players would take up the game as more than a passing interest. In 1978, Manfred Wimmer became the first Westerner to receive a professional player's certificate from an Asian professional Go association. It was not until 2000 that a Westerner, Michael Redmond, achieved the top rank awarded by an Asian go association.
See also: Go players
History
See main article history of Go
The origins of the game are unknown, but the oldest surviving references come from China in the 6th century BC. Except for changes in the board size and starting position, has essentially kept the same rules since that time, which quite likely makes it the oldest board game still played today. It had reached Japan by the 7th century, and gained popularity at the imperial court in the 8th. By the beginning of the 13th century, the game was played in the general public in Japan.
Early in the 17th century, the then best player in Japan, Honinbo Sansa, was made head of a newly founded Go academy (the Honinbo school, the first of several competing schools founded about the same time), which developed the level of playing greatly, and introduced the martial-arts style system of ranking players. The government discontinued its support for the Go academies in 1868 as a result of the fall of the Tokugawa shogunate.
In honour of the Honinbo school, whose players consistently dominated the other schools during their history, one of the most prestigious Japanese Go championships is called the "Honinbo" tournament.
Historically, Go has been unequal in terms of gender. However, the opening of new, open tournaments and the rise of strong female players, most notably Rui Naiwei, has in recent years legitimised the strength and competitiveness of emerging female players.
Around 2000, in Japan, the manga (Japanese comic) and anime series Hikaru no Go has popularized Go among the youth and started a Go boom in Japan. In January 2004, the Hikaru no Go manga also began running in the American periodical Shonen Jump, in an attempt to spark a similar following in the United States.
Mathematical theory of Go endgames
Elwyn Berlekamp and David Wolfe have developed a mathematical theory of the late endgame in Go based on the combinatorial game theory of John Horton Conway. It is outlined in their book, Mathematical Go (ISBN 1568810326). Whilst not of general utility in most play, it greatly aids the analysis of certain classes of positions.
See also
- Computer Go
- Games played with Go equipment
- Go concepts
- Go handicap
- Go players
- Go proverb
- Go rules
- Go strategy and tactics
- List of free Go programs
- List of Go topics
- Nihon Ki-in
External links
Learning Go
- The Interactive Way To Go is an excellent resource to learn the basics; includes Java applets.
- British Go Association's introduction to Go
- How To Play Go
- Samarkand is owned by Janice Kim 3p and has a very good beginner introduction.
- goproblems.com has over 1000 problems for practice in a Java applet.
- GoBase has a collection of 750 problems used in Korea to help develop players of amateur dan strength.
Resources
- Sensei's Library is a wiki devoted entirely to the game of Go - it even has special markup for displaying Go patterns.
- The Usenet newsgroup rec.games.go has its own FAQ document, the rec.games.go FAQ.
Go history
Go organisations
- Korea Go Association
- American Go Association
- Canadian Go Association
- British Go Association
- List of Go associations in Europe
- International Go Federation
Internet Go
- Go can be played on the Internet against opponents from around the world on numerous Go servers.
- The Kiseido Go Server is such a server, complete with its own easy-to-use Java client and introductory material.
- The Internet Go Server, the "original" server. Several official and 3rd party clients are availible.
- The Legend Go Server, located in Taiwan, with its own English client.
- The Dash Go Server, Korean server with an English windows client.
- The Tygem Go, located in Korea, with its own client.
- The Oro Go, located in Korea, with its own client.
- The Dragon Go Server, a turn-based server run on open source software.
- The LittleGolem, another turn-based server, this one centred around tournament play.
Recorded games
- The popular SGF file format is used to exchange Go lessons and recorded games.
- Fuseki.Info - Online professional Go games database (more than 31000 games). Contains game records, game lists, fuseki and joseki trees.
- Several free reading and authoring programs are listed at Gobase's SGF editors list
- gobase.org also hosts a database of professional Go games in SGF format (free registration required, which takes 1-2 days to process)
- A smaller collection of professional games in SGF format is available without registration at Andries E. Brouwer's Go Games.
- Amateur games are reviewed at The Go Teaching Ladder.
Go software
- GNU Go, free Go-playing engine
- gGo, SGF editor and client for IGS, in Java; and native variants, qGo and glGo (has a 3D display)
- CGoban1, Go client (Linux, etc)
- CGoban2, Go client in java
- Goban, standalone (against GNU Go) and Internet Go client for Mac OS X
- HandyGo, J2ME Go client that runs on java-enabled cellphones and PDAs.
- BiGo Assistant - BiGo Assistant is Go (Baduk, Weiqi) games (professional and amateur) database software. It allows searching by fuseki, joseki, positions and game information fields.
- WikiTeX Go supports SGF for inserting go directly into Wiki articles.