Talk:Go (board game)

From Citizendium, the Citizens' Compendium
Jump to: navigation, search
This article is developing and not approved.
Main Article
Related Articles  [?]
Bibliography  [?]
External Links  [?]
Citable Version  [?]
To learn how to fill out this checklist, please see CZ:The Article Checklist. To update this checklist edit the metadata template.
 Definition The world's oldest game that is played in its original form, with a documented history of over 2,500 years. [d] [e]

NOTE: The {{editintro}} template atop the article page should be removed prior approval.

Expand the character section?

This section is too short. I've never played Go, and would still have no idea how to do so after reading this section. I can infer that pieces don't move once placed, or are perhaps flipped depending on adjacent pieces, in order to control territory, but I don't know... Anton Sweeney 05:49, 31 December 2007 (CST)

Perhaps the best way to address wei-qi beginners would be by providing a diagram by a diagram a whole game on a 5x5 board, to the very end. First couple of diagrams would be of a good size, say 2x2 inches, and the remaining could be 1x1 inch, so that they would not take unreasonable space.
I could provide an example but it'd be better if it was done by a truly strong wei-qi player. I would even impose on the first player (on blacks) to start with a non-central first move, to make it more interesting.
How strong is a "truly strong wei-qi player" in your opinion. I'm 1 kyu. In general it is probably difficult to find a professional 5*5 game because nobody plays that format. Perhaps we use a 9*9 board with a game played by professionals as an example?Christian Kleineidam 13:22, 1 January 2008 (CST)
I meant that a player like you would make up a game. It's an easy and pleasant task, but a strong player will do it with a better feel, will provide a better (fictitious) game. The game should be aesthetic, nice, and should illustrate the main, elementary aspects of go (like scoring, but that's granted).
Also, I don't know wiki graphics. However, if somebody would provide one diagram, wtih one white and one black stone, then perhaps anybody could edit such a diagram to create new diagrams(???). Wlodzimierz Holsztynski 05:25, 1 January 2008 (CST)

Could we just take diagrams from the Go wiki, Sensei's Library? They have hundreds. Their license is different but it looks to me like we could provided we give credit. Sandy Harris 05:10, 7 June 2009 (UTC)


I took the liberty of cleaning up the terrible grammar and spelling errors in the Rules section to some extent.. i don't think i changed any intended meanings. Bram De Clerck 18:44, 31 December 2007 (CST)

I have some doubts about this section in its present shape. Perhaps it should be called: Rules--some ideas, and there should be also a solid section with rules, called Rules--precise, something like this. To help the cause, I may first write about the geometric notions involved in wei-qi. Wei-qi is such a wonderful intellectual achievement that it deserves and should be given full justice. Wlodzimierz Holsztynski 05:35, 1 January 2008 (CST)
There are no real Rules--precise in Go. There are Japanese Rules (also used in Europe), Chinese rules, Korean Rules, American Rules, New Zealand rules and Ing-Rules. At present there are Congresses that try to find International Go rules. In addition you have a bunch of other rule sets (for example the rule sets used on the different computer Go Servers). Which Go rules do you want to formulate precisely?Christian Kleineidam 09:41, 1 January 2008 (CST)
The ultimate, elegant rules were arrived at during the 1958 conference in Taiwan. I used to have their brochure. While indeed, there are all kind of unnecessary awkward formulations elsewhere, the spirit of the game is clear, and it is expressed by the mentioned Taiwanese rules. They got rid of the special cases of ko, especially of that configuration in the corner. Instead, a simple and elegant rule says that it's illegal to make a move which would repeat any previous position (not just the previous one), thus covering the special cases, like simple ko, triple ko, etc. (There is still room for two versions of this rule, which are equally profound, and which in practice are virtually the same; thus the one which is slightly simpler is preferable). Also, the Chinese scoring of counting stones + territory, s+t for short (see the configuration score in the article) is preferable to subtracting the prisoners from the territory, t-p for short, for four reasons:
  • with s+t there is no issue of gambling a large territory in order to avoid a loss of a scoring point due to playing into that territory a defensive stone; the game is more pure and profound with s+t than with with t-p;
  • s+t is simpler, as witnessed by the fact that you don't need to involve the game, as in t-s; with t+s it is enough to know the configuration only; you don't need to know about the captures made in the game;
  • with s+t one dos not need to know the number of prisoners;
  • you don't need to worry about a cat sneakily snatching a prisoner, and hiding it out of sight, when you think about the next move.
The argument for t+p counting is often that it is easier (takes less time) to count that way.
You don't have to fill the last dame points and still get the right result. That means more time for the actually game.
I also don't think that CZ is a place to make the value judgment on which counting method and rule set is the best one. Christian Kleineidam 11:39, 1 January 2008 (CST)
True, the difference for the game score is not too significant, but occasionally a difference of one score point may be extremely important to a player.
Anyway, I am almost finished with the geometric notions of the game. Now it will be very easy to write down one version of the precise rules.
BTW, there must be sufficiently precise rules or else it would make no sense to play seriously, e.g. for big money. And there are precise rules. There is just more than one (sufficiently precise) variation, i.e. more than one set of rules. Wlodzimierz Holsztynski 10:33, 1 January 2008 (CST)
You can probably get a precise description for the rules that use t+s counting. On the other hand the Japanese rules have concepts like "Bent for in the Corner Shapes" that aren't easily mathematically definable. Japanese Rules also eliminate the Ko-Threads that come from Double Ko's which means that some shapes die which wouldn't in Chinese rules with t+s counting. There are some variantions of those shapes that get complicated and where the Japanese rules don't really say what you have to do. Then it's like law, law doesn't clearly define whether someone is guilty or isn't but a judge or jury has to decide what is "right". It would be practically impossible to describe the American law (or any other law system) with mathematical precision. Japanese rules are a bit like this but fortunately for most games both players agree on the result.Christian Kleineidam 11:39, 1 January 2008 (CST)

Perhaps we should also make a separate page for the different rule sets. I don't know whether someone who just wants an overview over Go would find the mathematical description of the rules useful.Christian Kleineidam 11:39, 1 January 2008 (CST)

I think, at least for the details, we should just link to the Sensei's Library discussion of the multiple rule sets. It is a sufficiently tricky and controversial topic to be left to specialists. An overview here is fine. Sandy Harris 07:37, 21 March 2011 (UTC)

The name of the game

I never ever saw GO refferred to as "go" (lower case g). It was always GO or Go.

Also, the game was invented by Chinese. It would be perhaps nicer to refer to it as wei-qi or wei qi or Wei qi. This seems to be the tendency these day(?). Of course, it should be mentioned that for a long time (centuries?) the main place where the game was flourishing was Japan. Japanese were even bringing Chinise wei qi prodigies to Japan, and some of these Chinese became world champions. That once again shows how seriously Japanese were treating wei qi (GO :-).

In conclusion, I propose to change the name of the article to Wei qi (and at the same time to give Japan credit for its contribution to the game). Wlodzimierz Holsztynski 05:46, 1 January 2008 (CST)

According to go is the noun that describes the game in the English language. We aren't using either the original name in which chess was invented in the chess article but the English word for chess. Christian Kleineidam 09:42, 1 January 2008 (CST)
What about "go" versus "Go"? (In this discussion, elsewhere, you yourself were writing "Go", not "go").
It doesn't really matter to me. I can live with both versions. Christian Kleineidam 07:38, 2 January 2008 (CST)

Definitely Go has bigger lead. I've never heard of "weiqi" in my life. I see that the title may give less credit to the Chinese, but Japanese culture is so popular in the west that the westerners know everything that the Japanese adopted from the Chinese or the Koreans as Japanese and in Japanese names/titles. (Chunbum Park 08:49, 1 January 2008 (CST))


Note that "notes section" is for within-the-article citing, while "Bibliography" is for all the books used for this article. We need real books to write this article, not just web site links. (Chunbum Park 09:10, 1 January 2008 (CST))

Most books that you could use as good sources are written in Chinese, Japanese or Korean. There are "How to play Go"-books but those books focus on specific parts of the game and I don't think they make good sources for the history of the game or it's character. On the other hand I think that a PDF that is published by or by the Strategic Studies Institute of the US Army War College can be used as a scholarly source.Christian Kleineidam 10:34, 1 January 2008 (CST)
There is a notable comparison between Go and chess in Deleuze and Guattari's A Thousand Plateaus [Mille Plateaux], in the chapter "1227: Treatise on Nomadology - The War Machine", p. 389-90 [London: Continuum, 2004].Jeffrey Scott Bernstein 13:47, 1 January 2008 (CST)

Geometric concepts of the game

"Strong go players are always aware of the number of liberties of each group of stones on the board." Strong players won't be aware of the specific count of liberties if it's >10. In addition a strong player counts "effective liberties" where he counts the number of moves that the opponent has to make more than him to catch the stones if he defends the stones. A group with a single 5 point eye has 3+2+2=7 effective liberties. I would suggest to drop that sentence. Christian Kleineidam 18:28, 1 January 2008 (CST)

Yes, this statement was a (sloppy :-) simplification and an overstatement. I'll formulate it more carefully (I felt uneasy when I was writing it). Wlodzimierz Holsztynski 02:55, 3 January 2008 (CST)
I have simply removed the offending statement (it was just a side remark). Wlodzimierz Holsztynski 05:29, 3 January 2008 (CST)

A version of precise, complete rules of go

You should add which go rules (Taiwanese from 19xx?) you are describing here and note that this rule set isn't used anywhere in the world. I think you haven't forbidden suicide at the moment in the rules.

Yes, thank you. I forgot, sorry (I'll fix it in a few hours; I have to go now :-). Wlodzimierz Holsztynski 22:17, 1 January 2008 (CST)

In addition the stones don't get removed immediately, which allows a white to transform a big black "double-eye" into a single eye:


B W 0 W B

B B B B 0




B B B B 0

Black can't respond by building two eyes and white can play there again locally to kill the group ->


B 0 W 0 B

B B B B 0

Every Go rule set that allows suicide removes the suicide stones immediately to prevent this from happening.

I didn't mean to allow any suicide :-) I was simply tired and forgot to write one more postulate. I've already done it by now. Thank you for pointing the omission. Wlodzimierz Holsztynski 02:51, 2 January 2008 (CST)

To be truly complete a rule set should also include komi. Christian Kleineidam 18:29, 1 January 2008 (CST)

I mentioned it by now. It'd be nice if someone would expand the section about komi, titled: Who is the winner?, as well as the section about handicap, titled" Games between players of unequal strength. Wlodzimierz Holsztynski 02:51, 2 January 2008 (CST)

Why so mathy?

Shouldn't a lot of the mathematical discussion in this article be moved to a new article titled "mathematical aspects of Go" or "an introduction to graph theory using Go as an example"? This does not seem like a good way to introduce Go to people who have not played it before. As much as we might dread the thought, many people are interested in Go but aren't interested in mathematics. It's ok if the rules we give aren't totally precise; people who are interested in judging Go tournaments can learn precise rules later. --Warren Schudy 19:59, 1 January 2008 (CST)

I agree. (Chunbum Park 00:05, 2 January 2008 (CST))

In my opinion, an encyclopedia article about weiqi should play several roles, including: an introduction, a minimal tutorial, a reference!, history, a short overview of variants. Thus while there should be a non-rigorous, easy introduction to the game, there also should be a complete rules reference for at least one version of weiqi. It is important that when two novices have a different idea of a situation in the game due to their incomplete knowledge of the rules then they can patiently check the formal and complete description of the rules.
When two novices have a rules dispute your mathematical rules won't help if the two haven't decide beforehand to use this rule set. Especially since the default rules on Go events in that country probably use other rules ( provides links to the official rules for multiple countries). It would probably better to have an extra page where we have the mathematical rules and a list of rules that apply in different Go events. Christian Kleineidam 07:18, 2 January 2008 (CST)
On the other hand I would have the non-pedantic description expanded, so that the rigorous section about the geometric concepts can be moved down just before the rigorous rules description. That was my original idea. I consider the present placement as a temporary, until the notions like killing a group of stones will be written in a more leisure language.
I replaced killing for capturing but both are expressions that appear in Go books (for example Richard Bozulich). Christian Kleineidam 07:18, 2 January 2008 (CST)
BTW, I detect a certain anti-math attitude. But mathematics provides readers with the easiest language for stating weiqi rules precisely. Avoiding mathematical language would make the text rather muddy. In the somewhat advanced situation of weiqi, it's virtually a tautological equality: precision = mathematics Wlodzimierz Holsztynski 02:43, 2 January 2008 (CST)
The problem with math is that a lot of people don't understand it. Christian Kleineidam 07:18, 2 January 2008 (CST)
As a theoretical computer scientist I have a great respect for math; I use it every day! I agree that math is approximately synonymous with precision. However, IMHO the precise rules of Go are far less important to the article than the informal explanation of the rules. Math is great for precision, but precision is not the primary goal of an introduction. --Warren Schudy 11:34, 2 January 2008 (CST)


Page turns a text like below into nice graphics. Can we have something similar?

$$B triple ko
$$ -----------
$$ | . . X . X |
$$ | . . . X O |
$$ | . . . . O |
$$ | O X . O . |
$$ | X . X X O |
$$ -----------

Wlodzimierz Holsztynski 03:30, 3 January 2008 (CST)

The Go articles at wikipedia use this Go graphics template, which seems to be the best of several alternatives. I think it would be easy enough to import it here, and by easy, I mean easy for someone who knows what they're doing, which certainly isn't me. :) I suggest asking one of the constables to import it if you find it suits your needs. Richard Pettitt 21:26, 3 January 2008 (CST)
At the moment I'm trying to import it, but there is a problem. The template uses images (for every field where a stone could be placed). Those are unfortunatly uploaded by an wikipedia users with a pseudonym. At the moment seems to forbid copying those images. I made a thread in the forum about the issue,1542.0.html . Christian Kleineidam 15:55, 5 February 2008 (CST)

Core rules v. paraphernalia rules

There are the core rules which truly define the game (and for a friendly game you don't need any additional rules), and there are other rules which I will call paraphernalia rules. The paraphernalia rules should be mentioned too, and one or two can be stated, to give the gist of them. For instance, a rule states that the stones have to be lens like disks. For the tournament play this rule is important. Otherwise a guy would show up with square stones, and it would look weird to his opponent. Another rule may say that you make a pass move not by saying "pass" but by handing your opponent one of your stones from your container. All these rules may be important to a smooth run of a tournament or a serious match (where conditions should be standard and fair), but they are way less fundamental to the game than the core rules.

What I have presented, using a formal language, are the Chinese Weiqi Association core rules. It seems to me that several other national associations use the same rules, including AGA (American Go Association). They disallow: suicide and any repetition of a configuration (except for pass). They count stones + surrounded vacant points, including the surrounded seki points.

(Let me mention, that even the issue of komi, which I would consider a core rule, is not as fundamental as other core rules. Indeed, komi is somewhat arbitrary, and unnecessary when two players play two games, one game with the stones of each color. This can work for the double round robin tournaments too.) Wlodzimierz Holsztynski 04:03, 3 January 2008 (CST)

The Chinese have a rule that declares the moonshine life as dead which would sometimes live in your example. Christian Kleineidam 08:05, 3 January 2008 (CST)

A functional MRI study of high-level cognition.

I think the link should stay in the article. Maybe we can word that pargraph in a weaker way: "There is a study that suggests ... but had only 6 participants"? I think if someone wants to learn about Go the little cognitive research that's out there should be referenced by our article.Christian Kleineidam 21:23, 4 January 2008 (CST)

This seems reasonable to me. The reason I removed the study was because it was placed as a reference for a statement that directly compared Go to chess even though the study wasn't comparing Go to chess. I'm going to be off-wiki for a few days, but I'd be happy to help create a paragraph on the neuroscience aspect of Go when I get back. Richard Pettitt 23:50, 4 January 2008 (CST)
I added a new section (which probably should be merged somewhere else in the article.) Let me know what you think. Richard Pettitt 01:53, 10 January 2008 (CST)

Weiqi and Occam's razor

Weiqi is an exceptionally instructive illustration of Occam's razor principle. I hope that at least one of the important weiqi organizations applies the simple but profound core rules. Even if not, I would still just mention in the encyclopedia main entry article about weiqi that there are in practice different sets of rules used by different organizations, while the core rules of a standard game can be expressed simply as follows:

  • players of black and white stones start with an empty board; they put alternatively single stones of their color on vacant points, starting with the player of the black stones; instead of dropping a stone a player may say pass; two consecutive passes (i.e. a pass by one player, followed by an immediate pass by the other player) end the game; thus the player of the black stones makes the first move, then the player of the white stones makes the second move, etc.;
  • after a stone is put on a vacant point, all (if any) dead groups of stones of the other color are removed from the board; it is not legal to put a stone on a vacant point such that a dead group of stones of its own color is created without creating a dead group (one or more) of stones of the other color; the configuration of stones after the k-th move, and after removing the dead groups (if any) is called the k-th configuration of the game;
  • except for three consecutive configurations at the very end of the game, and otherwise except for any two consecutive game configurations, no two game configurations can be identical;
  • the score of the game is the score of the game's last configuration, which is the number of black points (black stones and black vacancies) minus the number of white points; the player of the black stones wins when score exceeds komi; otherwise the player of white stones is the winner.

Remark  A player who cannot put a stone legally on the board is forced to say pass.

That's all. All other special rules about multiple ko and special configurations are only blemishes on the beautiful game of weiqi (Go). The Occam's razor rules. Wlodzimierz Holsztynski 21:09, 5 January 2008 (CST)

Question about strength

I just rewrote the stuff on scoring, added some on handicaps, and added a section titled "Ranking". Comment solicited.

In particular, I ended the ranking section with:

The overall range is enormous. A top pro wins easily against a 1 dan amateur at 9 stones; the 1-dan wins easily against a 10 kyu at 9 stones, and the 10 kyu wins easily against a beginner at 13.

Being about 10 kyu myself, I'm quite certain the last two clauses are accurate. However, I'm not absolutely certain of the first one. Anyone care to comment on that? Sandy Harris 05:25, 7 June 2009 (UTC)

Someone else added:

"The difference in playing stength between the average professional 1 dan and the average 9 dan is also a lot less than one handicap stone."

Is that accurate? I thought that pro ranks were about 1/3 stone part. Sensei's Library says it is a lot more complex than that. One post suggests a 2-stone difference between 1 dan and 9. Sandy Harris 02:57, 13 June 2009 (UTC)

I think that should be one of:
  • "The difference in playing strength between professional ranks is a lot less than one handicap stone."
  • "The difference in playing strength between a professional 1 dan and a 9 dan is nowhere near eight stones."
Anyone got comments before I change it?

"Eastern" games

Shogi (Japanese chess) is played on the squares, isn't it? Anyway, chess is Eastern (Indian) in origin. Peter Jackson 12:08, 21 March 2011 (UTC)


Standard usage is that this means a number of stones near each other, whether connected or not. Peter Jackson 12:13, 21 March 2011 (UTC)

I see from Wikipedia talk pages that the situation's more complex. There's no standard term for a connected group of stones. They are variously called group, chain, link, string (...). But "group" also has the broader meaning, for which there's no other standard term. Peter Jackson 11:04, 23 March 2011 (UTC)

Certainly Peter is right about the usual usage of "group".
Personally, I prefer Bradley's terms [1], group for related stones and unit for connected stones. Sensei's Library uses chain for connected stones [2], though, and arguably we should follow that usage. Sandy Harris 11:56, 23 March 2011 (UTC)
In all commentaries on games, "group" is used for a configuration of stones related to each other, not necessarily connected. For instance, even a few loosely distributed stones after a joseki form a group. --Peter Schmitt 18:34, 23 March 2011 (UTC)