Talk:Go (game)/Archive 5

This is an archive of past discussions about Go (game). Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 1

←

Archive 3

→

Level of complexity

"It is extremely unlikely that any Go game played by humans has ever been duplicated by others."

Um, oh really? This is a pretty lofty sounding claim that seems to be easily disprovable. What if, at some point in history, somebody recorded their moves and then deliberately played the same game twice? What about for training purposes? IMHO, this claim is way too vague to be worthwhile without at least sourcing an article that explains the concept in more depth. I'd suggest rewording or striking it. Kazim27 04:21, 2 January 2007 (UTC)

intentional replaying of professional games occurs all the time as a study method. What the quote (which has been previously discussed per inclusion if you bothered to look) refers to is the exact replication through real time game play of another game. The idea that the exact replication of every single move through chance on a board of 360 points has never happened is not outraegous in the least. But honestly, there is no second-party source verifying nor denying, and it is kept for inclusion because it is a popular arguement for the level of complexity in the culture of Go. Feel free to add something about the claim being unsourced/unstudied. VanTucky 22:18, 2 January 2007 (UTC)

I have already stated that I have played the same game twice. Unfortunately this led to me being told I was not a Go player; quite a revelation. I imagine playing against a computer program is another easy way to replicate a game. Harry Fearnley developed a method of beating one program by starting a relatively obscure type of ko.--ZincBelief 16:54, 8 January 2007 (UTC)

This claim, like most of the article, absolutely needs to be sourced. —ptk✰fgs 18:38, 8 January 2007 (UTC)

Here we have four people, including me, speaking gainst this silly claim, some disproving it. "No game played by humans has been duplicated by others" -- meaning what, duplicated by non-humans? Aside from the inept phrasing, this comment damns Go with faint praise evenm if true, and it is NOT true. Let's say that "there are many more possible Go games than there are subatomic particles in the known universe." Easily verified and much more impressive in my experience when teaching beginners. Or just drop this stupid statement. kibi 16:06, 12 January 2007 (UTC)

I agree Kibiusa. the subatomic particle statement is actually provable considering scientists actually know the estimated number of protons in the visible universe (just fyi, they dont mean that there is an invisible universe as well). lets get a source, I've seen one before but we'd have to search. VanTucky 18:39, 12 January 2007 (UTC)

One thing that needs to be corrected is the mathematics, another is a way of describing in meaningul terms the magnitude of the number of games (about 10⁷⁶⁶) that can be played. I have no idea where the concept of 10^10^48 came from but to me it seems bogus. For example and I know I do not play by the same Go rules as everyone, but by the rules I play by black gets 181 stones and white 180 and once they are gone they get no more, no matter how many have been captured, so that leaves a maximum number of moves of 361, not 10 to the 48. However, using 120 factorial and squaring it to represent a non-capture game of what were they thinking, 120 moves? 241 moves? seems highly questionable. The correct mathematics for 120 moves is that you have 361 places to put the first stone, 360 for the second, a total for the first 120 moves of 361 factorial divided by (361-120) or 241 factorial or 1.4 x 10^768 / 9.8 x 10^470 = 1.4 x 10^297, and for 200 moves 1.4 x 10^768 / 7.6 x 10^286 = 1.9 x 10^481, for 300 moves 361! / 61! = 2.8 x 10^684, and for 361 moves 361! = 1.438 x 10^768. I accept that out of those perhaps only 0.012 are legal, leading to 1.7 x 10^766 for an estimate of the possible games. Capturing does increase this number because the number of possible moves after a capture may increase rather than decrease. No reference is given for how the number of capture games was increased from 10^397 to 10^567. If someone can give a quantification of the increase over non-capture that would be helpful, although not necessary. Another article on the web quotes the number of Go games as about 10^750, and you can see that when you are dealing with such a large number the difference between 750 and 770 is not very significant. An upper limit would be to assume that for each piece played the previous piece was captured leading (with 361 moves total) to 361 x 360^360. Absurd as this is it still is only 6.7 x 10^922. Please note that the above has been revised and corrected.

This leads me to my second point, how to make meaningful such a large number? For one thing while there are a few orders of magnitude more protons than atoms atoms are far easier to understand. The number of atoms is on the order of 10^80, which is a valid comparison for Chess, but definitely not Go. To say that it far exceeds it is like saying that the size of the galaxy far exceeds the size of my shoe, or the size of the galaxy far exceeds the size of the earth. While both are true, neither gives any meaning to the size of the earth vs the size of a shoe. What I am proposing is the words "far exceeds the number of atoms in the known universe squared to the fourth power". When you square 10^80 you get 10^160, raise that to the fourth and you get 10^640. Surely most people can identify with that being a very big number, and leave in far exceeds, because you have to add 126 digits to get to the actual number. The whole point is that it is a whole lot bigger than Chess or any other game and not bigger by just 3 or 4 times more complex, not just a few orders of magnitude, it is bigger by many hundreds of orders of magnitude.—The preceding unsigned comment was added by 199.125.109.78 (talk • contribs) 10:14, 1 February 2007 (UTC1)

Have you read the referenced research paper explaining where the number comes from? I would be interested in your opinion as to why it is bogus.--ZincBelief 12:34, 1 February 2007 (UTC)

As stated, I am assuming a total of 181 black stones and 180 white ones, so the maximum number of moves is 361.

The maximum number of 361 moves is too low if the Go Almanac is correct in it's record of the longest recorded professisonal game - I don't recall the exact length, but it was over 400 moves. Replenishing of stones would be accomplished by a prisoner swap, which wouldn't affect the score. As a lousy player, I fall victim to snapbacks & playing under the stones, resulting in a lot of moves on previously played points.Twslandlord 01:44, 3 March 2007 (UTC)twslandlord

This nonsense nitpicky addition about "squared to the fourth power" will not be included for several reasons. Not being a math person, I trust the opinon voiced by other wikipedians that steward this page to the point that the math is fuzzy. Second, even if it is true, the grammatical incorrectness of saying "the number of possible games far exceeds the number of atoms in the known universe squared to the fourth power" is astounding. It is completely redundant. To say that it far exceeds the number of atoms in the visible universe includes the number of atoms squared to the fourth power by definition. It seems to me that the general consensus is againgst inclusion. VanTucky 08:05, 2 February 2007 (UTC)

No, "the number of atoms squared to the fourth power" is in fact a different number, but that's beside the point. There is an editor who seems practically obsessed with adding ever more towering mathematical hyperbole to the sentence. The fact is that we don't need that hyperbole, for a number of reasons:

"The number of atoms in the universe squared to the fourth power" does not offer the reader any better idea of the magnitude than "the number of atoms in the universe". Our best estimates of the number of atoms are just that -- estimates. This comparison is used to give the reader a rough idea of how huge the game tree is, nothing more.
This is an encyclopedia, not a solutions manual. For the purposes of this article, "the number of atoms in the universe" and "the number of atoms in the universe cubed to the fifth power" are the same number, that is, "very huge".
It makes the sentence obnoxious and cumbersome to read. Tacking on piles of exponents and junk to the sentence does not help the reader understand the scale of the number any better. It just makes the sentence harder to finish and the concept less clear by the time you're done.

The current revision still suffers from more of the same "mathematics WOW!" cancer; I'll be reverting it shortly. If you can provide a reference, please simply list the numbers from that reference in plain, uninflected scientific notation without the use of multiple exponents, factorials, or other sensationalistic/unreadable junk.

Remember -- this is an encyclopedia, not the Discovery Channel. We're just trying to tell the reader how many legal positions there are, not scintillate him with irrelevant arithmetical expletives. Thanks. —ptk✰fgs 16:02, 2 February 2007 (UTC)

Please consider the following:

Level of complexity (take two)

Numerical estimates tell us that the number of possible games of Go ^[1] far exceeds the number of atoms in the known universe squared to the fourth power. Because of Go's simple mechanics of play, all legal positions^[2] are also positions resulting from possible games.

Ah, I see I have been reverted. Allow me, then, to briefly note that I dispute this connection. If anything, this piece of information would be relevant to searches by computers. There are many games with absurdly large state space which aren't that deep. Whether a state space is ludicrously large, ridiculous large, or merely absurdly large, is just not a very important factor. Therefore it seems to me this not connected to elementary strategy, and probably not to advanced strategy either, and I would like to see more explanation. Thank you for your time. --216.183.151.123 03:23, 23 March 2007 (UTC)
- I agree with the anon here. The size of the game tree is much more important in the computer go section, and the number of possible games is next to meaningless in terms of strategy. —ptk✰fgs 04:22, 23 March 2007 (UTC)
  - I have altered the wording slightly,. The number of possible games does illustrate something about how complex the game can be. Humans have to use a game tree as well as computers. What do you think of the current wording?--ZincBelief 10:49, 23 March 2007 (UTC)
    - I agree with Zinc here. One little mention in the nature of the game section works with the current wording. It's not exactly vanity or factual inaccuracy to have one sentence. VanTucky 17:20, 23 March 2007 (UTC)
      - I also agree with Zinc, the only thing is, humans don't have to rely ONLY on the game tree. Obviously, they have their own level of perception and understanding that a computer cannot match or replicate for that matter. the problem with the computer is that it all it has to work with is past games that have been recorded and put into its programming. The ability to recognize new formations and apply them in ways computers cannot. that is where computers are limited. Other than that, the current wording is fine. A little elaboration would be nice, but not required.Devildope 17:27, 23 March 2007 (UTC)

World Best Player

It should be noted about the world's best player. Oyo321 22:54, 2 January 2007 (UTC)

There isnt really A best player. there are a group of half-dozen or so top international title holders that are basically the current top winners, but to call any single player the best is stretching it. VanTucky 04:46, 3 January 2007 (UTC)

Go Seigen is considered by many to be the greatest player of the 20th century and one of the greatest of all time. Lee Chang-ho is the best player overall for the past 10 years for the number of international titles he has won(19). --Jiuduang 06:52, 15 February 2007 (UTC)

There's too many to really narrow it down to one. Each era has it's shining player (Dosaku, Seigen, Chang-ho etc) but they can't be compared. CanbekEsen 07:42, 15 February 2007 (UTC)

Last Remaining Citation-Needed

Korean players have had an edge in the major international titles, winning 23 tournaments in a row between 2000 and 2002. Can anyone provide a source for this. Is there something in gobase, msoworld, or a report in a magazine.--ZincBelief 16:21, 16 January 2007 (UTC)

yeah, needs a source. but isnt it pretty common knowledge that Korea produces more top winners than China or Japan recently? or is there an ongoing reversal of this? VanTucky 17:41, 16 January 2007 (UTC)

Regardless, it needs to be cited. —ptk✰fgs 17:52, 16 January 2007 (UTC)

for sure, but who wrote it? probably someone in the distant past of the article. maybe we can replace it with something similar if we can find a citation for more recent tournament statistics. VanTucky 18:48, 16 January 2007 (UTC)

This citation came straight from GoBase. There was a numerical citation requested which was explained adequately in the Senseis Libary page provided, containing a link of a published paper. --ZincBelief 17:05, 18 January 2007 (UTC)

nice work Zinc. VanTucky 19:21, 18 January 2007 (UTC)

Depends on the definition of "recently". Chinese players are current title holders of the LG Cup, the Samsung Cup, the Ing Cup. and the Asian TV Cup. The next winner of the Chunlan Cup will be either Gu Li or Chang Hao. --Jiuduang 06:33, 15 February 2007 (UTC)

Speaking of citations... What book of chapter 17 is go mentioned? I have a copy of The Analects in front of me now and cannot find any reference. Perhaps citations should be from the actual source they claim to be from. In other words, it is stated that go is mentioned in The Analects but actually the footnote leads to a magazine that claimed this. The game may in fact have been mentioned but I think a citation to the actual source would make sense. Either that or the phrase could be re-written as "it is mentioned by such and such magazine to be mentioned in..." 66.91.115.104 13:20, 27 March 2007 (UTC)

I must say I didn't notice that the Analects was actually mentioned in the article. I will take a look at this shortly. We should note that one citation is still requested to explain the contraversy surrounding the introduction of clocks into the game.--ZincBelief 13:26, 27 March 2007 (UTC) <edit> Ah those Analects , don't have any Confucius to hand though so I can't check myself.--ZincBelief 13:29, 27 March 2007 (UTC)

Standard technical jargon

The claim that "some terms have no equivalent in Asian languages" and the Go therefore lacks a "standard technical jargon" is incorrect. See the book Contemporary Go Terms for details. There are no important terms in any Asian language that do not have equivalents in other languages. Generally speaking, concepts which are difficult to express in English have become known by their Japanese equivalents -- e.g. sente, gote, aji, sabaki. Some Chinese and Korean authors, seemingly more interested in asserting the superiority of their own language than in developing a simple lexicon, insist on using alternate terms for the same concept, causing confusion among readers who are familiar with other names for those concepts. kibi 16:36, 23 January 2007 (UTC)

I don't really agree with that point. The use of Japanese terms is not widespread enough to claim that they form a standard technical jargon in my opinion. Paticularly now Korean terminnology is becoming more popular, as published literature in the korean tradition for the english language market appears.--ZincBelief 17:07, 23 January 2007 (UTC)

I disagree. though I may not move in Korean and Chinese Go circles, here in the states even the chinese and korean players use terms like aji and sente. Its just that Japanese is the language that happened to filter into western go playing circles, so these particular japanese words have become in fact (like Kibiusa said) a kind of "standard technical jargon" despite adolescentinternational rivalries. its a fact, let it go. and even though janice kim might publish books using korean terms or something does not mean there that it is becoming actually more popular in usage. The japanese may have forbidden the use of the native language when they occupied korea, but that doesnt mean that overnight koreans stopped speaking korean. VanTucky 19:06, 23 January 2007 (UTC)

The book by Nam is interesting: I wasn't aware of it (I talked on go terms at the first ICOB in 2001). But the point is that say hane in Japanese doesn't translate into a unique Chinese term (but three); and anyway what English speakers call hane might be hane or osae.

That was a misunderstanding and misquotation by User:Kibiusa: the article read

There is also no exact equivalence of concepts in different Asian languages, meaning that Go is still without a standard technical jargon.

Please note that equivalence of concepts in different Asian languages is something quite different from some terms have no equivalent in Asian languages.

Charles Matthews 14:13, 2 February 2007 (UTC)

Traditional go equipment

Reading through it, it looks like it needs a big rewrite. It's just POV everywhere. CanbekEsen 12:49, 12 February 2007 (UTC)

For sure the first paragraph is poor, but can we have a little reason on citing sources? Any material that is challenged or likely to be challenged needs a reliable source, which should be cited in the article. Specifically, which lines are likely to be challenged, none have been chalenged so far. --ZincBelief 14:54, 12 February 2007 (UTC)

It is preferably made from the rare golden-tinged Kaya tree (Torreya nucifera), with the very best made from Kaya trees up to 700 years old.

This sounds like a matter of taste. If it's universal, cite it.

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.

This sentence makes significant claims about the ecological and economic infrastructure behind go equipment production. Cite these claims.

Most players find that the lower price does not justify this distraction.

Bordering on POV-pushing, but acceptable if rephrased and cited.

Traditionally, the board's grid is 1.5 shaku long by 1.4 shaku wide (455 mm by 424 mm) with space beyond to allow stones to be played on the edges and corners of the grid. This often surprises newcomers: it is not a perfect square, but is longer than it is wide, in the proportion 15:14. 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 suggested reason is that the Japanese aesthetic finds structures with geometric symmetry to be in bad taste.

Beyond the traditional dimensions, which should be easy to cite, this paragraph offers an extended subjective commentary on the equipment. If this is to remain, it must be cited.

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 colours that would make equal-sized 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.

Some uncited speculation and subjective commentary.

It is best to take only one stone at a time as you decide where best to play. It is permissible to strike the board firmly to produce a sharp click. Many consider the acoustic properties 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.

Some folks say Wikipedia isn't an instruction manual. Well, most folks actually. Maybe we should just ditch this block altogether.

Did I miss anything? —ptk✰fgs 17:31, 12 February 2007 (UTC)

Have you read the existing citations before requesting seperate citations for these? This reminds me of a previous instance...go, chess and backgammon --ZincBelief 19:25, 12 February 2007 (UTC)

GA Review

Please. NO. Naked. U. R. Ls. Aside from the lack of title and author and whatnot, they also lack retrieval dates, which means that anyone trying to use a dead link will have abs--Rmky87 19:13, 3 March 2007 (UTC)olutely positively NO idea what to tell the Wayback Machine.--Rmky87 22:48, 22 February 2007 (UTC)

I reformated 3 of the reference's urls, but then got completely bored of doing it and stopped.--ZincBelief 15:58, 23 February 2007 (UTC)

Piotrus also seems to have helped out formatting the links, and that no longer appears to be a problem. --Yono 23:53, 27 February 2007 (UTC)

Okay, then. I shall get to reading it shortly.--Rmky87 22:14, 2 March 2007 (UTC)

Failed GA

I will give you this: the article is not the subject of edit wars, the images are mostly in the public domain and the one image that is not is used with the permission of the creator. That said, there are entire sections that don't have citations. And this sentence here, "Despite the fact that Go originated in Ancient China, Westerners first learned about this game from Japan, so it is commonly known in the West by its Japanese name." makes me, say, "Which ones were those?" A look at the history section did not tell which Westerner or Westerner introduced it to other Westerners or when that happened.

I'm sorry.--Rmky87 19:13, 3 March 2007 (UTC)

Whole sections uncited, I wonder if you read the existing citations before making that remark. The sentence you quote is badly written though yes. It should indicate that the modern pioneers of the game learnt about the game from Japan, which led to it being called Go. --ZincBelief 19:43, 3 March 2007 (UTC)

How many people are going to have to say this before you start to consider that it's true? How long is it going to take? The article is undercited! Instead of arguing about it — and since you seem to have access to reference sources we don't — why don't you just add <ref> tags to the statements that the sources support? —ptk✰fgs 20:32, 3 March 2007 (UTC)

You have twice asked for citations in a section without reading the citations there, which provided all the explanations required. Fundamentally, I don't understand this requirement for citations in every sentence. It seems appaling bad style. As somebody who has research papers published I find it odious. It is not true. Plain and simple. If it really the policy of wikipedia, then wikipedia is flawed in my opinion. —The preceding unsigned comment was added by ZincBelief (talk • contribs) 21:28, 3 March 2007 (UTC).

ptkfgs, I am unable to comprehend how an article with 42 referenceds is undercited. Rmky87, I'm not sure what criteria this article fails. It is factually accurate and verifiable, it includes inline citations, it includes sources for information likely to be challenged. The Origin of the Name and the Basics Rules sections that you cite as unreferenced are so because they simply state facts that are unlikely to be challenged. --Yono 07:03, 4 March 2007 (UTC)

Looks like we have to cite every sentence, sad. CanbekEsen 17:13, 4 March 2007 (UTC)

How about just every paragraph? And really, would it kill you to mention that which Westerners first got into go and introduced to the rest of the Western world?--Rmky87 04:45, 5 March 2007 (UTC)

Oskar Korschelt is accepted as the westerner to introduce Go to the western world. There are no others that are known of. CanbekEsen 05:24, 5 March 2007 (UTC)

I think this part was poorly written, but it's been changed already, just a minor edit. It does seem a pretty trite point though. If the game is known by its Japanese name in the West, it's pretty obvious as to why that would be. Maybe we should pile in with references as to where Lasker and others learnt the game from? All the early books refer to the game as Go. —The preceding unsigned comment was added by ZincBelief (talk • contribs) 10:40, 5 March 2007 (UTC).

Is there some way to appeal the decision? Yono 23:57, 4 March 2007 (UTC)

I have listed the article for Review of GA Nomination already. --ZincBelief 00:31, 5 March 2007 (UTC)

By the way, thanks for adding the name of one of the Westerners to the article. That really helps. And I see you've added more references.--Rmky87 21:08, 5 March 2007 (UTC)

rating question

The discussion of ratings seems a little unclear to me in one spot. "If an amateur does not know his/her rank then he/she is probably less than 10 kyu" implies to me that a numerical rating under 10 indicates someone who doesn't play enough to know/care about ratings. The rest of the context implies that someone with a lower rating is better. —The preceding unsigned comment was added by 144.15.77.156 (talk) 20:57, 26 February 2007 (UTC).

The rating scale is 30kyu 29kyu .... 2k 1kyu 1dan 2dan etc . I think this information is displayed in the ratings table beside the article.--ZincBelief 10:56, 27 February 2007 (UTC) I changed it to 'no better than 10 kyu'.... I'm not sure if that put much of a dent in the confusion. This section tries to explain something in a few sentences that is very hard to describe in a few sentences. --Haikon 17:53, 3 March 2007 (UTC)

I tend to think of "kyu" as a "minus sign" in math. I say that when explaining the rating scale to others. SlowJog 21:07, 12 August 2007 (UTC)

The "minus sign" analogy can mislead people into thinking that there is a 0 rating between 1-kyu and 1-dan, I've noticed. An analogy to B.C. and A.D. years avoids that risk (well, as long as the person knows that there was no year 0. :). Goulo 16:26, 16 August 2007 (UTC)

History of game

Does anyone know how the 19x19 size came to be standard? I've heard that 17x17 & 21x21 boards have been found. It makes sense that there should've been some experimenting. I would find this interesting.Twslandlord 02:01, 3 March 2007 (UTC)

19x19 = 361. This is considered a mystic number. Also some professionals - sorry I forget which - played a few games on 23X23 boards but they felt that neither of the players really knew what was going on. Centuries of play have determined that 19x19 is the perfect size for a game of igo.SmokeyTheCat 12:18, 6 March 2007 (UTC)

yes, i have also seen what Twslandlord sees. I have seen references to smaller boards such as 5×5 boards and 7×7 boards. Maybe we should state those as well. You may have already added this to the article but i have yet to check it out.--Devildope 12:55, 16 April 2007 (UTC)

In Tibet Go used to be played on a 17x17 board. As far as I know every other place saw 19x19. If you can find a reference supporting why 19x19 came to be used feel free to add it in. However if there is nothing, probably best to leave it aside. You can chose to play Go on any size of board you like, so not really much need to mention other individual sizes unless there is something interesting associated with that board size. For example, 5x5 has been solved by computers.--ZincBelief 13:08, 16 April 2007 (UTC)

Terminology

The second paragraph seems to need a citation to me. Following on from this GA review, I've been trawling through the article seeking unreasonable material. The claims made here look like speculation. Also, if somebody could fix my botched reference on the first paragraph that'd be grand.--ZincBelief 11:12, 5 March 2007 (UTC)

It is definitely true that some Chinese and Korean players get their undies in a bundle about the use of Japanese terms, but good luck finding something n print about it to cite. Some Chinese and Korean people just don't like Japanese stuff period, just as the reverse is also true, and they carry that bias with them into this discussion. Yesterday I was talking with someone who was having a heck of a time reading the new Korean book, | Train Like A Pro: Haengma because it's loaded with Korean terms. I pointed out the glossary, which the reader found helpful, but what would have been even more helpful would have been the use of terms that he knew from other books. It's hard enough learning concepts like sente, gote and sabaki, without having to learn three different names for each one. Too bad some people think petty chauvinistic concerns are more important than simple, effective teaching. kibi 17:45, 5 March 2007 (UTC)

I agree with you. However if we can't source something about this then we probably ought to rewrite this, otherwise it looks like POV.--ZincBelief 15:02, 6 March 2007 (UTC)

Actually thinking about it, the new OroMedia series Train like a Pro (loaded as they are with korean terms) is probably good enough as a reference. What do you think?--ZincBelief 13:10, 16 April 2007 (UTC)

Train Like A Pro does not "advocate for the primacy" of Korean terms; no arguments are advanced to show that Korean terms are preferable to Japanese ones. It simply ignores the terms that Western players have become familiar with, adding to the confusion by substituting terms of its own choice. So it is not really a reference that supports the statement. How is it helpful to refer people to books that will confuse them? kibi 14:23, 16 April 2007 (UTC)

Well why else would a book entitled Train like a Pro: Pae appear on the international market, if not to thrust Korean terminology upon the world? OroMedia would have been well aware that 95% of western Go players were unfamiliar with the term Pae before. Besides, I can't think of another reference to support the claim made in this section, and nobody else has offered one up. :) —The preceding unsigned comment was added by ZincBelief (talk • contribs) 14:37, 16 April 2007 (UTC).

Exercising the brain

I put this small section in. And, no, I don't have a citation sorry. I just read it somewhere. Feel free to delete/amend it as you see fit. SmokeyTheCat 12:14, 6 March 2007 (UTC)

I remember reading that is uses a different part of the brain to chess. However I would contest the Alzheimers assertion. Learning new things is the key to combatting stagnation of the brain, not playing a mind game. Does this little tidbit of information really deserve its own section though? --ZincBelief 12:26, 6 March 2007 (UTC)

Claiming that the game prevents Alzheimers is truly egregious without a citation. I've removed the section, as it made two scientific claims without evidence. —ptk✰fgs 15:07, 6 March 2007 (UTC)

I think a {fact} might have been more appropriate than deleting the material. However I await it's reappearance with some citations. Could it be wedged into another subsection like the Nature of the Game? —The preceding unsigned comment was added by ZincBelief (talk • contribs) 15:30, 6 March 2007 (UTC).

I will try and find the reference. Until then I bow to majority.(Even if it is only two!)SmokeyTheCat 15:12, 6 March 2007 (UTC)

The article mentioned in http://www.usgo.org/EJournal/archive/20060227.htm is probably a good starting point.--ZincBelief 16:18, 6 March 2007 (UTC)

The link inside the link should also be useful. CanbekEsen 00:07, 7 March 2007 (UTC)

I have re-added this section with a reference.SmokeyTheCat 10:08, 13 June 2007 (UTC)

I am not sure whether this information is relevant in an article specifically about go. The relevant research is: Leisure Activities and the Risk of Dementia in the Elderly by Joe Verghese et al. (abstract, free registration required to see full article). Which shows that several leisure activities reduce risk of dementia (reading, board games, dancing, playing a musical instrument). So this effect is not unique to Go, but applies to all boardgames (and is mentioned in relation to chess as well). Therefore, I do not think this merits its own section, but might merit a footnote somewhere. HermanHiddema 12:26, 13 June 2007 (UTC)

GA again?

Most of the issues raised in the GA review from March 3 seem to have been fixed. Are we ready to renominate this? --Selket ^Talk 17:41, 9 March 2007 (UTC)

Is there anything anyone finds remotely challengable in the text at the moment? One reviewer did say it was a bit too listy. --ZincBelief 12:03, 12 March 2007 (UTC)

At one time the article had FA status, and the objections that cost it that status have also been addressed. Why settle for mere GA status? kibi 13:06, 20 March 2007 (UTC)

If you're feeling brave nominate it for FA status then, You get much constructive criticism from reviewers that way. :) --ZincBelief 13:47, 20 March 2007 (UTC)

Well the game image given under strategy should probably be considered as a derivative work based on the game in question. As such the copyright would be held by the LG group. 85.166.105.165 11:56, 20 March 2007 (UTC)

Strange assertion to make, only game commentary and rights to live broadcast are asserted with regard to games of Go. --ZincBelief 12:11, 20 March 2007 (UTC)

You're right, Zinc. A game record, especially from a major tournament, is a public document. It cannot be copyrighted. kibi 13:06, 20 March 2007 (UTC)

It seems to me mildly crazy that we devote as much space to time control arcana as to great players. Charles Matthews 12:19, 20 March 2007 (UTC)

This is very true Charles, length of elaborations on sidelines is one thing that should be addressed perhaps. There is still a lot of dull material only of likely interest to the hardcore player in the article, although it is very well written dull material. I wonder if a peer review might be useful for this article. --ZincBelief 13:47, 20 March 2007 (UTC)

Wikipedia:Good_article_review#Go lists concerns about the article still--ZincBelief 16:52, 22 March 2007 (UTC)

?? The GA criticisms aren't there anymore. kibi 12:52, 18 April 2007 (UTC)

Yes, that's because they are now on the archive section of that page. Outstanding problems... There is still one citation request to be provided. The references in general probably need a bit of reformatting. I don't think any other serious concerns have been raised about the article. GA status is acheivable with a little work. --ZincBelief 13:06, 18 April 2007 (UTC)

Last call for a reference for Formal time controls were introduced into the professional game during the 1920s, and were controversial. Adjournments and sealed moves began to be regulated in the 1930s. If nobody can provide one I would suggest we delete these and try to get GA status. The article is not formatted correctly to get FA status, but I think most of the requirements for GA are there right now.--ZincBelief 12:28, 10 May 2007 (UTC)

The Go Player's Almanac 2001 pp. 92-93. Are we ready for GA status now? kibi 16:20, 13 May 2007 (UTC)

Thanks, I have added that reference into the article now. If somebody wants to nominate the article for GA status go ahead. As I've already nominated it once, I don't want to renominate.--ZincBelief 09:57, 14 May 2007 (UTC)

Ad hoc Review

Looking over the article I identified 3 sections I thought were a waste of space

Level of complexity , one line, therefore not worthy of its own section
Terminology , very specialist and surely better siphoned off into the preceeding section in the article
Software Assistance , contains some rather banal trivia

What do others think? I'd rather sort out how the article reads that fiddling about with formatting --ZincBelief 13:58, 20 March 2007 (UTC)

I read an article, which is misplaced, in the AGA Journal (?) (c. 1990?), that asserted that go is superior to chess in 10 ways: 1. Go is simpler than chess because, once played, the stones do not move; 2. Yet, go is more complex because of the size of the board, 361 points compared to 64 squares; 3. The Go game record is visually holistic. The game can be shown on one page; 4. One can find compensation for a loss by good play in another region of the Go board whereas a loss of a chess piece is a lasting handicap; 5. The system of incremental handicaps at the start of the go game can equalize the fighting strength between go players; 6-10. etc.) Larry R. Holmgren 07:51, 23 March 2007 (UTC)

Is the variants section in the correct place, might it not be better placed after Rules ? --ZincBelief 11:22, 23 March 2007 (UTC)

I think if a reference can be found for the Time Control being contravertial, we can move toward GA nomination. What do other people think? --ZincBelief 11:57, 5 April 2007 (UTC)

I think the article will need a review for quality of sources. For example, this appears to be from a self-published source. This may not be a concern before GA, but it will take a while, so it would be worth undertaking soon, rather than having to re-source the whole thing during FAC. —ptk✰fgs 12:49, 5 April 2007 (UTC)

I don't think that is a particularly good example to pick. You can observe the way people play stones at any Go tournament or any Go club. The need for this reference is pretty low in my opinion, since it is basically common knowledge. Having said that, I am open to looking at the quality of other references.--ZincBelief 12:57, 5 April 2007 (UTC)

The reader should not have to go to a go tournament to verify the claims in the article. They should be attributed to reliable sources. To assert that there is a generally agreed-upon way of grasping the stones is not exactly an obvious claim, particularly to most readers in English-speaking regions (read: the audience of this article). —ptk✰fgs 13:05, 5 April 2007 (UTC)

If you look pictures of a stone being played you will see the technique in action. In the context of querying the reliability of the source, what you're saying seems dubious. It is an easy claim to verify. That the source might be self published could be said to be a bad thing, but to then assert that it must be unreliable seems in itself questionable. The source is self published by the Nihon Kiin. To argue that the Nihon Kiin, the primary Japanese professional Go players association, is engaged in an act of misleading the world in how to play Go stones seems folly. Therefore I repeat, this doesn't seem a particularly good example to pick to me. —The preceding unsigned comment was added by ZincBelief (talk • contribs) 13:27, 5 April 2007 (UTC).

That's why I said it should be reviewed, not removed. The page that's directly linked does not mention the author, and neither does the footnote. —ptk✰fgs 13:29, 5 April 2007 (UTC)

It is also a professional self published source on a non contentious issue.--ZincBelief 13:32, 5 April 2007 (UTC)

So, it's simply a matter of adding the author's name to the footnote, in this case. —ptk✰fgs 13:33, 5 April 2007 (UTC)

East versus west

Often, a comparison of Go and chess is used as a parallel to explain western versus eastern strategic thinking (despite the original forms of chess having Asian origin)... [Reference: Science, Culture, and the Game of Go (word document) Retrieved on 24 February 2007]

This reference does not support this sentence. The authors call them "facile analogies" in fact. 216.183.151.123 03:37, 23 March 2007 (UTC)

Oh dear, some work clearly needed here then. Although, we should note that the reference provides an example of somebody using this comparison, not supporting the validity of this comparison --ZincBelief 10:52, 23 March 2007 (UTC)

Worlds Best Player (2)

I wasn't getting any response on the first one so i will post it in a new area. Here is my idea for a section on the worlds best players. Why don't we just have a chart that shows the last 10 years and has the top players of those years on it. Say,

                  1997: Bla Bla Bla (Where they are from)(number of titles)
                  1998: "    "   "  (Where they are from)(number of titles)
                  1999: Bloo Bla Bloo (Where they are from)(number of titles) and so on...

I just think that it would be nice to show the top players of recent years so that people can grasp what the standings actually look like. You don't have to just put the number of titles that this person has/had won, you can put the names of the titles as well. go ahead and edit it to your liking. ORIGINAL POST(--Devildope 12:42, 16 April 2007 (UTC)) THIS POST --Devildope 12:41, 18 April 2007 (UTC)

Well, visually this would look nice. However, how do you prove who was the best player in a certain year? Go doesn't have a world professional championship. It is sadly going to be subjective to say who was the worlds best player in a certain year, and man alive, think of all those references we would have to provide! --ZincBelief 15:08, 18 April 2007 (UTC)

Well wouldn't there be a reference somewhere that would state the top title holders in the world or something to that effect. If we could simply (haha simply) find this reference and then cite it's information...Devildope 13:20, 19 April 2007 (UTC)

Possible, we'd have to show top titles in Korea, China and Japan though. If you find references let me know :)--ZincBelief 13:53, 19 April 2007 (UTC)

i'll begin my search! ^_^ Devildope 15:28, 19 April 2007 (UTC)

There is another article on Wikipedia called | List of top title holders in Go. we could just make a short list of the players that are shown here and then we could reference that page. If the readers would like to check out more, have them click on the link and be redirected to that page. How does that sound? Devildope 16:15, 19 April 2007 (UTC)

This is a dead end, for Wikipedia. There is absolutely no reliable source for a world champion, or even a national champion in countries with pros. Charles Matthews 18:59, 19 April 2007 (UTC)

well ya know what! Maybe they should pick one! Haha! Well it was worth a shot. Whatever, maybe next time. ^_^ Devildope 20:14, 19 April 2007 (UTC)

Wait! I just fixed the link for the "List of top title holders in Go" link above. It works now. anyone feel like reconsidering? Devildope 01:49, 20 April 2007 (UTC)

The other page that i mentioned above states the total titles won and the total international titles that have won. I still think that it would be a worthwhile subject to explore. Maybe not in this form but we should reconsider anyways. I know that making a decision as to the worlds best player can be considered a biased assumption and i am completely against the use of bias on Wikipedia but i think that if we fly by the facts on this one we will not be making a biased decision but a conclusion based on what the info states. I dont think that this will seemed biased in any way. We don't have to call the page the "Worlds Best Player" but we could rewrite the page as Top Title Holders in Go. Thoughts?Devildope 13:18, 20 April 2007 (UTC)

I don't want to sound like party pooper, but I have agree with Charles on this matter. I believe what we have now works well. Besides, this article was turned down from GA a few times, and one of the concerns was that the article was too "listy". CanbekEsen 19:17, 20 April 2007 (UTC)

Alright, i see your point. I concede defeat. ^_^ hey, but whatever. lets just move on to the next concern. How about those stubs. (see post below)Devildope 02:40, 21 April 2007 (UTC)

I've rewritten and expanded this section, giving some more info on historical (1600-1900) top players like Dosaku, Shusaku, Jowa, etc. I've included info on top Japan based players of the 20th century like Go Seigen, Kitani, Sakata, etc. I've added some citations to GoBase, which has lists of title results, and used these as a basis for updating the bit about international tournament winners of recente years (Lee Changho etc). I've tried to keep it reasonably brief and to avoid listy-ness. The main article referenced (Go Players) will still need a lot of work HermanHiddema 13:17, 25 May 2007 (UTC)

Any certain stubs...

...that someone would like me to research? Devildope 01:46, 20 April 2007 (UTC)

That's a better question for the appropriate Wikiproject page, where you can find a to-do list. VanTucky 21:20, 25 April 2007 (UTC)

Number of players...

In the first info box at the beginning of the article, it is stated that the game is for two players, which i know is the norm, but i have also heard of pairs Go, which has two people on a side. I don't know if it is something you want to mention. Something like "Players: 2-4".Devildope 21:07, 25 April 2007 (UTC)

Basically, other than just for simplicity, the game is only ever played by two players. Pair go is two alternating teams of players on the same board, but they are not allowed to discuss the game under tournament rules and you always play against just one opponenent literally. Also, there is already a section on altenate forms of the game. VanTucky 21:17, 25 April 2007 (UTC)

complexity sourcing

The organization claiming Go is the most complex game (the American Go Association) that I sourced says..

"2. Go is the most complex of all games. Nearly all known games have been "solved" for the computer -- that is, the strongest computer programs can defeat the best human players. Even chess now falls into this category. However, the strongest go programs, after decades of effort, are routinely trounced by Asian schoolchildren. Why is go so hard for computers? Because go is much, much more complicated than chess. There are many more possible games of go -- as much as 10 with more than 700 zeroes! -- than there are sub-atomic particles in the known universe."

So the game is very complex (as evidenced by the difficulty in programming computers to play it) because of its practically infinite variations, according to the AGA. Thus, congruent with Wikipedia sourcing standards, I changed the reasoning for the claims of complexity to match the source. If you can find a reliable published source backing the reason for being most complex that was previously there, then feel free to revert it. Until that time, the facts must match the citation. VanTucky 05:51, 18 May 2007 (UTC)

I just want to thank you for not saying anything about sub-atomic particles in the article. Since the most common element in the universe is Hydrogen, which only has about 2 sub-atomic particles, when you are dealing with a number with 700 zeroes compared to the number of sub-atomic particles which is a number with only about 80 zeroes, that extra 2 to 4 times as big is totally moot, and doesn't even get you to 90 zeroes. 199.125.109.59 06:12, 18 May 2007 (UTC)

The reference to sub-atomic particles is technically incorrect. The article does mention the estimated number of atoms in the universe however, as this is backed by the math. VanTucky 19:48, 18 May 2007 (UTC)

Where does this 10⁷⁰⁰ number come from? It's given in several places, e.g. our own game complexity article, but never with a proper citation. It's also a massive undercount: John Tromp and Gunnar Farnebäck in their paper Combinatorics of Go give the bounds 10^10⁴⁸ and 10^10¹⁷¹. I suspect someone has mistaken "game tree complexity" (the number of positions you might have to look in a full-width search that finds the game-theoretic value of the original position) for "number of possible games". Gdr 19:15, 23 May 2007 (UTC) P.S. Ah, I see it's just 361!×0.012 and then rounded down. I replaced the note with the estimates of Tromp and Farnebäck.

Thanks for correcting this Gdr, John Tromp and Gunnar Farneback's work is not in dispute.--ZincBelief 09:18, 24 May 2007 (UTC)

I see that our anonymous colleague does dispute it! The place to look is section 6.4 of the Tromp/Farnebäck paper, in which they show how to play a game that visits positions corresponding to the 2^154 values of the 154-bit Gray code, with 103! ways to move from one of these positions to the next. Thus contructing (103!)^(2^154) ~ 10^(10^48) different games even in this highly structured subset of legal positions. The upper bound is just the naive (maximum branching factor)^(longest possible game). Seems pretty clear to me. Gdr 12:35, 24 May 2007 (UTC)

Dear anonymous user, please help us to understand what your objections are to the Tromp/Farnebäck paper. Are you concerned about the meaning of "possible games"? Is there a problem with the rule about repeating a position? If we can understand your objections then maybe we can agree on a compromise wording. Gdr 18:30, 24 May 2007 (UTC)

I am not an anonymous user. Due to the way wikipedia allows username creation, making up a name is far more anonymous. This issue of "how many possible games" and "longest possible game" has been discussed a lot. We can all agree that the former is a really big number, however the latter quickly becomes an absurdity. If you look at the stated reference, it says that the number of moves is about N^3, which for a 19x19 board is 361^3=47,045,881 and since there are about 31 million seconds in a year that would take two people playing 16 hours a day at the rate of one second per move about 2 and a quarter years, which has already reached the realm of absurdity. Go, after all is a game between two players, not an exercise for two computers to see which can count the highest, which like go when you relax the rules is infinite. We also need to look at the issue of significant figures, because saying 47 million is only significantly different from 47,045,881 if there is a reason it would make a difference. There is not. The same is true of changing 0.012 to 0.01196, which by the way is an estimate, not an exact number anyway. Does it really matter that there are 2x10^170 or 2.082x10^170 possible legal positions? No. What is significant is the exponent, and for all practical purposes it does not matter if it was 170 or 150 or even 120, unless you started to compare chess with go for example and you needed to prove a point.

We have already covered extensively the relative difference between the number atoms in the universe (10^80, a really big number) and number of go games, a really really big number. No one wants to know about that. I repeatedly got shot down when I tried to quantify the difference. While I used the more rigorous N^L where N is the game space of 361 and L the longest game to show how huge the number gets when you move only from 361 moves to 400 moves, I used a lower estimate to show how many games a 47 million move game allows. Both are absurdly big numbers, however it is very useful to understand the relative complexity and to understand in real world terms why it is easy to never have played the same go game twice, even if a billion humans played for a billion years, or a trillion, or a googol (but not a googolplex).

It is also in an encyclopedia important to reference whare that "often quoted" 10 to the 700th number comes from. So lets keep the numbers useful, and not get into the realm of contrived games, ok? Let's also not forget that none of the 97% of the world using Windows can read the preliminary Tromp paper because it is only available as a postscript (.ps) file. 199.125.109.133 19:28, 24 May 2007 (UTC)

OK, let's see if we separate out the issues here. As far as I can tell you are happy with the results in the Tromp/Farnebäck paper: you don't dispute their correctness.

However, you object to their bounds appearing the article because, although correct, they seem uninteresting to you and have no relevance to practical play. You would prefer to present a smaller lower bound that is more understandable to the average reader.
You object to the number of significant digits in the count of positions because it's just an estimate, and because these digits are uninteresting to you, and have no relevance to practical play.

On (1), what's uninteresting to you may be interesting to others; certainly the amount of discussion of this issue suggests that it has interest to some. The best known lower bound is certainly encyclopedic. So could we have a compromise where we give both the best known lower bound, and some lower number that's more relevant to actual play?

On (2), number of positions is interesting (for some people) in its own right, not just for any practical consequences it may have. As to how good the estimate is, see page 28: "For 19 × 19, the [asymptotic formula] gives 2.08168199382 ×10¹⁷⁰, of which we can expect all digits to be correct [i.e. because the convergence is so fast]." So I think a few digits here are not unreasonable. Gdr 20:36, 24 May 2007 (UTC)

This seems to me to be a non-issue. The math is correct, and any one user's personal opinion about how interesting it is matters not at all. The issue is relevancy, and these facts were decided to be relevant long ago. If you read the past discussions, you can see that proper consensus was reached on their wording and relevancy quite adequately. The new objections seem to hold no water in terms of Wikipedia policy regarding inclusion. VanTucky 20:53, 24 May 2007 (UTC)

I agree. But finding a workable compromise with the anonymous editor would be better than edit-warring. Gdr 20:58, 24 May 2007 (UTC)

Well, first off as to the objections to using all allowable positions in the calculations instead of just those within the common bounds of human play..."Does it really matter that there are 2x10^170 or 2.082x10^170 possible legal positions?" Yes, because one of the issues that makes a discussion of the math behind Go complexity relevant is programming for computer play, which deals with numerical variations beyond the numbers of positions for normal human play. Compare it to the numbers of all possible positions in Chess, and you begin to get a good picture. As to, "We have already covered extensively the relative difference between the number atoms in the universe (10^80, a really big number) and number of go games, a really really big number. No one wants to know about that" You are simply incorrect. Prior discussion abounds on this issue, and it has been resolved that it stays as is. Not only have your concerns been addressed in previous discussion, but they don't hold water on Wikipedia. Facts are evaluated for inclusion per Wikipedia policy, and you cite no policy or guideline that prohibits or discourages inclusion. Arguing around the issue gets you nowhere (as many of us learned the hard way). VanTucky 21:11, 24 May 2007 (UTC)

PostScript on Windows: try GSview. Gdr 21:22, 24 May 2007 (UTC)

My first comment is to be rated a "good article" a lot of work needs to be done outside of this one reference. I hope that most people would recognize that and focus their attention appropriately. I began working on this reference because of it's obvious lack of relevance. My objectives were two, one to update the mathematics and second to put the results into a meaningful context. 400 moves was specifically chosen because it is meantioned in the article. I appreciate constructive criticism and look forward to making any necessary tweaks. I'm still having a hard time understanding the importance of saying 2.082 (to the bazillionth) vs 2 (also to the bazillionth). It is "the bazillionth" that is important, not the fractional exponent. You will notice however, that I figured that it was "important to you" and did not revert that part. When I said "no one cares" I was referring to attempts to quantify how much bigger 10^766 is from 10^80. I tried lots of times and was always told that "no one cares". Let's say you are given one grain of rice for the first play, two for the second and so on. While the mathematics of a precise answer after say 150 plays is trivial, the important point comes in realizing that there is not enough rice in the world long before that point. It's like if someone asks you what pi is, do you pull up pi on your calculator to ten digits or pull out your copy of pi to a billion digits and read those out to them? The point is that if it mattered you could, although that probably was not the question they had in mind. 199.125.109.48 23:04, 24 May 2007 (UTC)

So would you be happy with my suggested compromise? Gdr 23:50, 24 May 2007 (UTC)

Please be specific. What I think the article needs is a reference that is relevant to the topic. NOT a reference that says, "while the average go game lasts 150 to 200 moves, if it was continued for a googol squared moves..." I do not see anything wrong with the current version. It includes your reference and has practical numbers. If you want to include 4 digits of precision, knock yourself out. 199.125.109.92 01:10, 25 May 2007 (UTC)

We should clearly give the best known bounds on the number of games. I've rewritten the note so that it does so, but I've also given a lower bound that is more relevant to play, the 10^360 game-tree complexity estimate from Victor Allis's thesis. I couldn't find a good reference explaining the significance of the 10^766 number, so I've omitted it. Gdr 12:08, 25 May 2007 (UTC)

Hmmm, so you don't like my rephrasing, even though it addressed your concerns and was properly referenced. Do you really object to any mention of the best known lower bound? Gdr 13:20, 25 May 2007 (UTC)

Maybe I should have been more clear. Please indicate what changes you would suggest here and we can discuss them. Just because you can not find a reference does not mean that someone else can't. From simple mathematics you have 361 places to play on the first move, 360 on the second, 359 on the third and so on, other than the fact that only one percent of the positions are going to be legal, as stated in the reference. Is it too much to ask to list your proposal here on the talk page so that we can come to an agreement on it first? No I do not object to you putting in your lower bound as long as it is put into context. Tell me, what is the lower bound for someone playing by Japanese Amatuer rules (361 stones)?

The changes I suggest are the one I made here. Let me know what you object to.

Here is what I object to in the article as it stands:

The best known bounds are missing.
An upper bound of 10^2807 is given but this is known to be bad (it's smaller than the best known lower bound).
The 10^766 number is based on a couple of mistakes. Just because 1% of positions are legal, doesn't mean that 1% of 361-move sequences constitute legal games. Remember that an illegal position is one in which a string has no liberties. This means the position can never appear on the board. But that doesn't mean the moves that lead up to it are illegal, because the move sequence may lead to capture of the liberty-less group, which is fine. So the 10^766 number is clearly an underestimate even before we start to consider sequences involving playing back onto captured intersections. The best I can do for the problem of counting the number of 361-move games is to note the bounds 10^766 and 10^924.

I'm interested by your claim that under the Japanese amateur rules, games end when players run out of stones. Can you support this? Gdr 14:01, 25 May 2007 (UTC) P.S. I see that you have contributed numbers to our article on game complexity, or at least I presume so from the matching IP address. It would have been useful if you had made these edits a Wikipedia account, then it would have been possible to ask you for references. Can I appeal to you to get an account?

I would be happy to include a bound of 10^766, which by the way I do not think is a lower bound, and 10^924. I think reference should be made to what happens if you go to 400 moves should also be included, as it is meantioned in the article as the longest go game, and as to "all practical considerations" a more appropriate statement might be, "if play were continued for an impossible number of moves (exceeding a trillion years)". As to the amatuer rules, those are the rules that I have always been taught, it wasn't until I started researching different rule sets that I figured out which ones they were.

The following is not an endorsement of your proposal, it is simply a statement of your proposal so that it can be discussed.

Victor Allis (1994). Searching for Solutions in Games and Artificial Intelligence (PDF). Ph.D. Thesis, University of Limburg, Maastricht, The Netherlands. ISBN 9090074880. Allis notes that typical games between experts last about 150 moves, with an average of about 250 choices per move, suggesting a game-tree complexity of 10³⁶⁰. For the number of possible games, with all considerations of pratical play laid aside, see John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go". This paper gives lower and upper bounds of 10^10⁴⁸ and 10^10¹⁷¹ respectively. It also shows that on a 19×19 board, about 1.196% of board setups are legal positions, which makes 3³⁶¹×0.01196 = 2.082x10¹⁷⁰ positions in total.

The following is not an endorsement of the current reference. It is presented so that it also can be discussed.

John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go". This paper shows that on a 19×19 board, about 1.196% of board setups are legal positions, which makes 3³⁶¹×0.01196 = 2.082x10¹⁷⁰ positions. 361 moves give a total of approximately 361!x0.012 = 1.7×10⁷⁶⁶ games. Allowing captures, and unlimited stones could give as many as 2.1x10²⁸⁰⁷ games, each having up to 47 million moves whereas 400 moves yield on the order of 1.2x10¹⁰²¹ possible games.

Fortunately not very many people from this ISP contribute to Wikipedia, or should I say not many or any vandalize from this ISP, so I have no interest in purchasing an account. Asking for references here should be sufficient.199.125.109.91 16:07, 25 May 2007 (UTC)

Can I encourage you to specify your objections to my proposed wording, so that we can progress? If I understand you correctly, your objections are now:

That the wording "all considerations of practical play laid aside" is not strong enough to indicate the size of the numbers here.
That the 10^766 number must be included.

I'd be happy to concede your (1) although I think we could come up with better wording. But I'm not happy with (2) for the reasons given above (number only applies to a particular sub-case, i.e. counting the games with 361 alternating moves, and is not well supported even in that case). I also don't think "the rules that I have always been taught" is good enough evidence for the existence of a rule of finite stone supply in the Japanese amateur rules. I'd like to see something written down, like this ruleset. Gdr 16:24, 25 May 2007 (UTC)

The number 10^924 is just the naive upper bound 361^361. But really you shouldn't include it for the same reasons as you shouldn't include the 10^766 number: applies only to a sub-case, and not properly referenced. Gdr 16:35, 25 May 2007 (UTC)

Let me restate my objective (not objectives). How is the statement that there are more games than atoms in the universe referenced? One by stating the possible number of legal positions. That is a very important reference. Two by relating mathematically how many possible games can be played (by either two humans or two computers). Since we can anticipate that the age of the universe is less than a trillion years and we can anticipate that no computer will ever exceed say a Trillion Teraflops (physics tells us that the number is much lower), it is not practical to spend a lot of space debating a number of moves (all contrived I might add) that exceeds 10^36. What is useful, however is to mathematically show how many games are possible for a game of 150, 200, 250, 361, 400, or even 361^3 moves (47 million)(which is supported by the reference).

http://home.snafu.de/jasiek/wagc.html (iii) There should be 181 black and 180 white stones. Since all will not usually be needed in play, however, the game may be played with a smaller number of stones. [not a larger number]199.125.109.91 18:03, 25 May 2007 (UTC)

I think that two sets of numbers should be presented: first, the best known bounds; second, a set of bounds that's relevant to practical play, difficulty of programming a computer, etc. Do we agree on this at least?

That rules page doesn't say what you think it does, as you can see from the section on the end of the game, which mentions only three possibilities: two passes in succession, resignation, and running out of time. No mention of termination by running out of stones. Gdr 18:25, 25 May 2007 (UTC)

Also, that link is outdated (Those are the 1979 WAGC rules). And the mention of equipment is not so much a reference to the rules of the game, as it is an tournament organisational issue (These are tournament rules, not just game rules)

See eg the 1997 rules under Right To Play or 2003 Rules under Alternation for a more up to date view on this.

Most rulesets assume an unlimited number of stones for both players (I'm not sure on Ing rules). HermanHiddema 18:37, 25 May 2007 (UTC)

Well pardon me for not following the rules for 20 years. Actually I do think that it means exactly that. Why else would it state 181 for black and 180 for white? Ask someone who is Japanese. I am well aware of other rule sets that state that a sufficient number of stones are supplied. I was trying to research what rules I had been taught, and found that they conformed to Amatuer Japanese rules. The above reference is just from a quick google search, and not necessarily the ones that I had found. 199.125.109.91 18:45, 25 May 2007 (UTC)

The reason why we doubt it means what you think it means is that if the game ended when the players are out of stones, then the players might not agree about who is the winner. This would need to be handled by resumption of the game, hypothetical play, or by appeal to the referee. However, the rules you cite specify none of these procedures. My guess is that the author simply didn't consider this issue as being important. In your 20 years, has this situation ever arisen (game ended by lack of stones, players unable to agree on the winner)? If so, what did you do? Gdr 18:59, 25 May 2007 (UTC)

No game I played has lasted that long. Long before then territories are so clearly defined that there is no difficulty in determining the winner. Besides, if you can't agree on the winner, just laugh and play another game. 199.125.109.91 19:54, 25 May 2007 (UTC)

Complexity reference

I don't think adding the 0.012 factor adds anything significant, the more accurate number is mentioned in the Tromp/Farnebäck paper, so I have removed it. I have also removed the reference to Senseis Library. The referenced page contains several different estimates and can only confuse an uninformed reader. I'm looking for other references for this number. HermanHiddema 21:14, 25 May 2007 (UTC)

Gosh two orders of magnitude isn't significant? Someone wanted 0.01196. 199.125.109.91 21:52, 25 May 2007 (UTC)

Be nice :-) HermanHiddema 22:11, 25 May 2007 (UTC)

Perhaps we can reference section 5.12 of the rec.games.go Rules FAQ. It is not very clear, but does seem to contain relevant material HermanHiddema 21:31, 25 May 2007 (UTC)

I certainly do not think that 761 was a typo for 361, however it has only been recently that ordinary people had access to 361! if they even knew what it was. Hey my calculator only goes up to 10^499 199.125.109.91 21:52, 25 May 2007 (UTC)

10^761 is very close to 10^768, so that might well be the source of that number. Or perhaps it's someone confusing 768 and 361 and producing 761.

I also found out that the rule faq on the homepage of Robert Jasiek is out of date. The newest version posted to rec.games.go is http://groups.google.com/group/rec.games.go/browse_thread/thread/cd4aea68741a41ce/72203c4da6f24649?lnk=st&q=rec.games.go+rules+faq&rnum=2&hl=en#72203c4da6f24649 which has a reworded section 5.12 HermanHiddema 22:11, 25 May 2007 (UTC)

John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go". This paper shows that on a 19×19 board, about 1.196% of arrangements of pieces are legal positions, which makes 3³⁶¹×0.01196 = 2.082x10¹⁷⁰ positions. Considering only games up to 361 moves long with no "play under the stones" (i.e. no play on points where a stone has been captured) we get about 361!×0.01196 ~ 10⁷⁶⁶ games. For the number of theoretically possible games, including ones far too long to play in practice, Tromp and Farnebäck give lower and upper bounds of 10^10⁴⁸ and 10^10¹⁷¹ respectively.

User:Gdr's preferred wording:

Victor Allis (1994). Searching for Solutions in Games and Artificial Intelligence (PDF). Ph.D. Thesis, University of Limburg, Maastricht, The Netherlands. ISBN 9090074880. Allis notes that typical games between experts last about 150 moves, with an average of about 250 choices per move, suggesting a game-tree complexity of 10³⁶⁰. For the number of theoretically possible games, including ones impossible to play in practice, see John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go". This paper gives lower and upper bounds of 10^10⁴⁸ and 10^10¹⁷¹ respectively. It also shows that on a 19×19 board, about 1.196% of board setups are legal positions, which makes 3³⁶¹×0.01196 = 2.082x10¹⁷⁰ positions in total.

Nah, I like the following better, and please do not expect to reach consensus on this soon!

John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go". This paper shows that on a 19×19 board, about 1.196% of placements are legal positions, which makes 3³⁶¹×0.01196 = 2.082x10¹⁷⁰ positions. 361 moves gives a lower bound of approximately 361!x0.012 = 1.7×10⁷⁶⁶ games and an upper bound of (how did you get that number?). Allowing captures, and unlimited stones for 400 moves yield on the order of 1.2x10¹⁰²¹ possible games, and more than 2.1x10²⁸⁰⁷ games, for 47 million moves. It is also shown by Tromp that more than 10^10⁴⁸ specially contrived games could last an impossible time of 10⁴⁸ moves.

This deals with my first objection (no mention of the best known bounds), although it has an incorrect figure for the number of moves in the longest constructed game. However, it still doesn't deal with my other objections, namely that the 10^766 number is a lower bound for a sub-case (exactly 361 alternating moves with no play under the stones); and the 10^2807 is pointless for inclusion here because it's a lower bound that we know is bad. You're also laying it on a bit thick with the "unlimited", "specially contrived" and "impossible". No-one disagrees with this, so one mention is adequate. Gdr 17:11, 25 May 2007 (UTC)

That subcase of 361 moves is pretty important, because it is the most often quoted estimate of games. I took out a few of the extra words. One thing that I would like to emphasize is how do we shorten up the reference to make it more readable and still cover all of the important points? 199.125.109.91 19:00, 25 May 2007 (UTC)

Well maybe if I take out the Victor Allis number and you take out the 10^2807 number, we get the following:

John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go". This paper shows that on a 19×19 board, about 1.196% of arrangements of pieces are legal positions, which makes 3³⁶¹×0.01196 = 2.082x10¹⁷⁰ positions. Considering only games exactly 361 moves long with no "play under the stones" (i.e. no play on points where a stone has been captured) we get about 361!×0.01196 ~ 10⁷⁶⁶ games. For the number of theoretically possible games, including ones far too long to play in practice, Tromp and Farnebäck give lower and upper bounds of 10^10⁴⁸ and 10^10¹⁷¹ respectively.

I can't say I like this much because of the need to give so many caveats about the 361-move games. But if it's acceptable to you then we should go with it because at least it's an improvement over what we have now. Gdr 19:10, 25 May 2007 (UTC)

I see no need for any caveats. There are 361 places to play. While AGA etc. allows games longer than that it is a useful number for calculation purposes, easily understood, and rarely exeeded. However if you take out the Victor Allis paper it should probably be added to the article as an external reference as follows:

Victor Allis on searching for solutions.

You don't see the need for showing how many games there really are (in normal play)? Or how many a 400 move game would give? I guess my point about there being no need to include Tromp's upper limit is that it is in the paper and since the lower limit is totally impossible, there is no need for any upper limit. I also want to caution that there are no readily looked up references in this reference [scratch that I see that the senseis reference is still included]. Can we hope that a pdf version will become available after it is published? 199.125.109.91 19:32, 25 May 2007 (UTC)

The caveat is essential because play under the stones is normal; avoiding it is an artificial restriction imposed to make the computation easy.
I do see that it would be nice to show "how many games there really are (in normal play)". But since we don't know what that is, we can't. We can give an upper bound for games of some arbitrary length: thus there are no more than 361^361 ~ 10^924 games with 361 moves and no more than 361^400 ~ 10^1024 games with 400 moves (assuming no passes). If you'd prefer either of these to the 10^766 number we could switch.
If we're going to give the Tromp/Farnebäck lower bound, we should give the upper bound too to indicate how much is still unknown about the problem.
Have you tried GSView yet? Or online PS-to-PDF converters, e.g. here?

Gdr 19:57, 25 May 2007 (UTC)

It might be a good idea to add something like that as a link as it does allow anyone for example in a public library who can not install a .ps viewer to access the article. 199.125.109.91 22:20, 25 May 2007 (UTC)

I am not making any assumption about where the stones are played when I say 361 moves. For a mathematician an upper limit is always important. For a player it is moot if they can never achieve the lower limit, i.e. the lower limit is also an upper limit. Can we get any other opinions on whether a caveat is important? Vantucky? Are you still there? Ptkfgs? It seems horribly pedantic for me. 199.125.109.91 20:24, 25 May 2007 (UTC)

It's implicit in the use of the factorial function that play under the stones is not being considered. Think about it: 361! counts the number of permutations of the board positions, with no position being repeated in any one permutation. Hence no position is played on twice in any game in this count. Gdr 20:30, 25 May 2007 (UTC)

I think all ofthese suggestion ar far too complex, this is the kind of information I would expect in an article specifically on Go complexity, not on the general Go article for wikipedia.

The van Allen reference only mentions go in passing and does no serious research into the number of possible games
The Senseis Library reference is a mess, mostly discussion between different players with very little actual useful content.
The Tromp/Farnebäck reference is serious research, but is mostly far too theoretical for this page.

How about:

The number of legal positions on the Go Board is in the order of 3^361(10¹⁷²), the number of possible games is in the order of 361! (10⁷⁶⁸). These theoretical bounds are more rigorously explored in: John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go".

HermanHiddema 20:31, 25 May 2007 (UTC)

I don't like your proposed wording because we know better numbers than the ones you give. Why not give the best known results? In particular, the number of possible games is not "in the order of 361!" I appreciate that this material is esoteric, but it is interesting and encyclopedic. And it's only a three- or four-line note. I think we can probably get something satisfactory to nearly everyone with a bit of effort.

I agree that the Sensei's Library page is a mess, but it's the only reference we have for the relevance of the 10^766 number that the anonymous editor wants to keep in the article. Not sure who you mean by "van Allen" -- is that a typo for Victor Allis? Gdr 20:41, 25 May 2007 (UTC)

Oh sorry, yes, typo, I meant Victor Allis :-) HermanHiddema 20:53, 25 May 2007 (UTC)

Regarding my suggestion: We may have better (more accurate) numbers, but for the average Wikipedia reader there is not a whole lot of difference between 10^360 (Allis), 10^766 (361! * 0.012), 10^768 (361!), 10^924 (361^361), 10^1024 (361^400), etc (I am aware that mathematically, 10^768 is insignificant compared to 10^924, but that's not my point). Such issues are best left to a separate article. I therefore propose we use two simple numbers, 3^361 and 361!, because the average reader can quickly understand how these numers are formed, and they are good enough as estimations. HermanHiddema 20:53, 25 May 2007 (UTC)

I have to say that I like the reasoning that HermanHiddema has given and the simplicity of the wording. 199.125.109.91 20:58, 25 May 2007 (UTC)

Go complexity has been submitted as a new article. See http://en.wikipedia.org/wiki/Wikipedia:Articles_for_creation/Today#Go_complexity

The proposed wording may be simple, but it's incorrect. And I think you underestimate the proportion of go players who can follow a bit of maths. So maybe we can compromise by putting simpler numbers first and the best known numbers second, like this:

The number of board positions is at most 3³⁶¹ (about 10¹⁷²) since each position can be white, black, or vacant. There are at least 361! games (about 10⁷⁶⁸) since every permutation of the board positions corresponds to a game. Tighter bounds are given in John Tromp and Gunnar Farnebäck (2007). "Combinatorics of Go". This paper shows that there are 2.082×10¹⁷⁰ positions, and finds lower and upper bounds on the number of possible games of 10^10⁴⁸ and 10^10¹⁷¹ respectively. See Go complexity for details.

(Note that it is reasonably straightfoward to show that there are at least 361! games. Choose a permutation. Play moves in the order given by the permutation. If this results in a legal game, we're done. Otherwise, eventually there's an illegal move because of the suicide rule (N.B. not the repetition rule since we never play on the same intersection twice). When this happens, the player to move passes and play continues with the sides swapped. This works because a move that was suicide for white must be legal for the black (if it were suicide for black, then it would have been a capture for white and so not illegal). Hence every permutation corresponds to a game. This avoids the heuristic argument that led to the number 10^766.) Gdr 23:16, 25 May 2007 (UTC)

Can you guys keep this section in order, please? It's hard to follow if you keep editing stuff at the top. Gdr 00:19, 26 May 2007 (UTC)

I fail to see the relevancy of including the number of games that have more than 400 moves. I understand the mathematics, which is why a separate page has been created, but those mathematics assume that there are more than 400 moves. 10^48 moves, and outside of the realm of the main go article.199.125.109.41 00:38, 26 May 2007 (UTC)

I like the current wording below. Clear, concise, to the point. HermanHiddema 17:07, 26 May 2007 (UTC)

I would also like to note that though I agree that a considerable proportion op go players can follow a bit of math, I do not think that that is very relevant. This article should not be aimed at go players, it should be aimed at a more general public interested in reading about Go in an online encyclopedia. This article can be considered the front page for everything about go, and as such should not contain too much detail on any topic (eg, no discussion of japanese/chinese/tromp-taylor/ing rules, etc). I think it is an excellent idea to have a separate topic on Go Complexity for those that are interested in the math behind it, but I would like to keep such things off the front page. HermanHiddema 17:07, 26 May 2007 (UTC)

Final suggestion for complexity reference

The number of board positions is at most 3³⁶¹ (about 10¹⁷²) since each position can be white, black, or vacant. There are at least 361! games (about 10⁷⁶⁸) since every permutation of the board positions corresponds to a game. See Go complexity for more details, which includes much larger estimates.

Any objections? Please add them to the Complexity reference section above.

Playing time

The infobox gives playing time as 10 minutes to 16 hours, which I think is confusing. 10 minutes is too short (this is 5 minutes each), 16 hours is exceptionally long. Perhaps it should say something like:
informal games 20-90 minutes
amateur tournaments 2-5 hours
pro tournaments up to 16 hours
Note that the Chess page uses a similar contruct. HermanHiddema 15:49, 25 May 2007 (UTC)

Whatever you do please keep the caption short. Making it smaller would also help (250px perhaps)199.125.109.91 16:14, 25 May 2007 (UTC)

10 minutes is for ten apiece, and for regular casual matches is not uncommon. Ten mins is however the minimum before the time officially becomes a "blitz" game. That is also why it is the min. 16 hours are commonly for pro tournies, but it is technically incorrect to say it is always or only for pros. VanTucky 16:26, 25 May 2007 (UTC)

I think that the box should mention total playing time, not playing time per player, as this is the common convention on board game related pages (Chess uses this convention too). And although amateurs can play with longer time limits too, this is so uncommon that it does not merit mention in the infobox in my opinion (too much detail for a box that is supposed to briefly list the main game characteristics). HermanHiddema 18:05, 25 May 2007 (UTC)

Agreed that only total time should be stated. I think that too much information is included in the info box that should be relegated to the main article. I remember how the 16 hours got added, but it always seemed to be a bit wordy. 199.125.109.91 18:33, 25 May 2007 (UTC)

I agree that total time should be what is included, but this article isn't just about amateur players. Limiting the time scope to only the average time of amateur enthusiasts is ludicrous. It's not comprehensive of the scope of playing times. Contemporary pro matches often take hours and hours; amateurs regularly watch the games via TV or internet broadcast. I don't think any Chess article editor would encourage not mentioning professional game times (especially if there was such a significant difference).VanTucky 19:00, 25 May 2007 (UTC)

So you are saying it should say 20 minutes to 16 hours? By the way I know that it starts with a blank board but it might be better to say "None" instead of "No setup needed" for Setup time - what no board needed??? 199.125.109.91 19:16, 25 May 2007 (UTC)

Yes, that would be my idea. And you're right about the "none" for set-up. VanTucky 19:24, 25 May 2007 (UTC)

I've updated the playing info, stating 20-90 minutes for casual play, 2-6 hours for tournament. I think mentioning realistic casual playing times is important, as this is relevant to the largest number of people (non players possibly interested in starting playing). 2-6 hours covers most tournaments (including many professional events using 3 hours per player). I've kept the 16 hour figure as a footnote, as this is an extreme (though interesting) case. HermanHiddema 19:47, 25 May 2007 (UTC)

Looks good to me. Clear, comprehensive, and concise. VanTucky 19:53, 25 May 2007 (UTC)

Any objection to reducing to 250 px, which allows a lot more room for the lead in paragraph? Also "tournament" in the footnote is redundant. 199.125.109.91 20:07, 25 May 2007 (UTC)

No objection, the page looks fine to me, but I have a large resolution screen. Also I agree tournament is probably redundant in the footnote. HermanHiddema 20:16, 25 May 2007 (UTC)

I've taken the liberty of reducing the image width to 280px, any smaller has no real effect, as the infobox title doesn't get narrower than that. HermanHiddema 17:15, 26 May 2007 (UTC)

infobox width

Whatever you do please keep the caption short. Making it smaller would also help (250px perhaps)199.125.109.91 16:14, 25 May 2007 (UTC)

Any objection to reducing to 250 px, which allows a lot more room for the lead in paragraph? Also "tournament" in the footnote is redundant. 199.125.109.91 20:07, 25 May 2007 (UTC)

No objection, the page looks fine to me, but I have a large resolution screen. Also I agree tournament is probably redundant in the footnote. HermanHiddema 20:16, 25 May 2007 (UTC)

I've taken the liberty of reducing the image width to 280px, any smaller has no real effect, as the infobox title doesn't get narrower than that. HermanHiddema 17:15, 26 May 2007 (UTC)

I see you've changed the image width of the go picture in the infobox to 250px with reason that it 'works better for opera'. However, under my Firefox the image is now needlessly narrow, the box title bar is wider. I've looked in to why this happens, and it is a browser issue. The infobox has a outside width of 26em, and an inside width of 24em (which is therefore also the width of the titlebar). Widths in em depend on font size (em is calculated from this), so the width of the infobox depends on browser font settings and browser font size calculations. I suggest we change the width so that it looks best on a default install of Internet Explorer under Windows XP (largest audience). Opinions? HermanHiddema 09:11, 30 May 2007 (UTC)

I've changed the width of the image back to 300px, this is the default infobox image width for wikipedia (see Template:Infobox), so I think we should follow it for consistency. HermanHiddema 14:15, 30 May 2007 (UTC)

Please revert (to 250px). Five hours was not long enough to wait for opinions. HTML normally needs to be checked in at least four browsers to avoid making things painfully difficult for some users. In the case of Opera I get these little skinny columns of text when you go even to 280px that make reading the article very difficult. I suggest that instead of targetting the largest audience we target at least 3 sigma of the audience (99.7%). Otherwise we get things like, don't worry about blind people, after all why would anyone blind be using the internet? Just click on the image if you want to see it better. You can do that, I can't make it smaller. Do you see any text to the left of that infobox at Template:Infobox? 199.125.109.91 21:46, 30 May 2007 (UTC)

Please note that my above two statements are independent. I first noted that the image is needlessly narrow on firefox (and on IE, as I've since tested). The second comment was after I found that 300px was used on the Template:Infobox page, which caused me to revert to that as the default value pending discussion. I did not intend to close the discussion with that. HermanHiddema 08:00, 1 June 2007 (UTC)

I have no real objection to narrowing the image, so feel free! I am however curious where those 'skinny columns of text' appear. It doesn't seem logical they should (from looking at the html + css), so I'm interested in what Opera is doing HermanHiddema 08:00, 1 June 2007 (UTC)

Thanks, and I apologize for being presumptive. I have no idea what Opera is doing. However it I did note that Netscape does not get better below 280px. I would suggest picking a photo that shows the go board better, I like ones that show the side of the board, or one that shows the stones closer up. Also I like ones of historical game positions. 199.125.109.75 14:40, 1 June 2007 (UTC)

Pictograms

I do not think that it can be a coincidence that the three strongest go-playing nations by far, (Korea, China and Japan) are the only three nations that still use pictograms. Go requires a subtle sense of visual space and it seems that using pictograms might stimulate this sense. I would like to put in a reference to this but fear it might be called OR. Is there anywhere that such a reference might be apt. I would welcome other opinions about this please. SmokeyTheCat 11:27, 4 June 2007 (UTC)

I think this is a coincidence, but if you have a reference to a serious scientific study on the matter, I would be very interested. There are far more likely reasons that these nations are the leading nations:

Long availability: All three of these nations have been playing Go for over a thousand years (China since at least 500BC, Japan and Korea since at least 700AD), whereas the rest of the world has only played the game for perhaps 100 years.
Number of players: China has 10 million active players, Korea 9 million and Japan 3 million. Compare Europe, which has only 20,000 active players total.
Starting age: Players in these countries start much younger, there are schools full of 7-10 year olds in China who are already of amateur dan level. The 2007 World Amateur Championship was won by a 13 year old Chinese 8 dan. Most western players start playing in college.

Also note that Japan does not solely use pictograms. It uses many Kanji, borrowed from the Chinese, but the Japanese language is alphabet based (Hiragana and Katakana, which are different writing styles for a 47 symbol alphabet) HermanHiddema 13:17, 4 June 2007 (UTC)

Korea also uses Hangul, a phonemically based written system, almost exclusively. So much for that theory, sorry . . . kibi 13:52, 4 June 2007 (UTC)

Just a note Chinese characters are not pictograms necessarily. Go to Chinese character article on Wikipedia for more info. People seem to have that misconception too often...

Promoting a healthy brain

There is some evidence that that the playing of go prevents to the onset of Alzheimer's disease and dementia. [1] This seems to be because go uses both sides of the brain while chess only uses one. To play go one needs spatial awareness and aesthetics and as well a pure calculating ability. -- preceding was moved from article to talk page for discussion. 199.125.109.23 15:28, 13 June 2007 (UTC)

I almost completely rewrote the section to comply with a NPOV, and I also fixed the citation. VanTucky 16:38, 13 June 2007 (UTC)

See my earlier comment under 'Exercising the brain' above, and the reference given there. The citation to Yomiuru Shimbun is not a very reliable one, being just a newspaper column by a go player with (AFAIK) no specific medical knowledge on the subject. HermanHiddema 17:38, 13 June 2007 (UTC)

I was not aware that my original contribution was POV but I am happy with the section as it stands now. Thanks VanTucky.SmokeyTheCat 11:29, 14 June 2007 (UTC)

While it might be completely true that some board games do good to your brain, I think this is simply off-topic. Also, medical fact like this needs considerably more scientific proof than a go professional who says go professionals do not get dementia. So I'd just say "It has been said go prevents dementia / Alzheimer's disease" and move the discussion about preventing dementia and Alzheimer's disease onto pages discussing them. Only if go has remarkably good effect on your brain, it should be mentioned on this page, but even then, very shortly. 193.110.108.67 12:51, 14 June 2007 (UTC)

Solid scientific support can be found in Leisure Activities and the Risk of Dementia in the Elderly by Joe Verghese et al. (abstract, free registration required to see full article). I agree however, this does not merit its own section. I will try to find a way to incorporate a short note on this into the article later tonight. HermanHiddema 14:42, 14 June 2007 (UTC)

However the abstract also says that it is not clear whether leisure activity reduces dementia or that those with dementia just stop playing games. Therefore the fact that no one who plays go has dementia is possibly non-consequential. 4.233.134.161 14:42, 22 June 2007 (UTC)

infobox image

To me the current infobox image is rather stark and jarring, if we're going for a more arty, angled photo instead of something straightforward why not do something more along the lines of the following?

File:Go-game-ear-reddening.jpeg

The Ear-reddening game.

Maybe not this one, but I would like to explore choosing a different image. Here's a link to the Commons page for more ideas. VanTucky 00:36, 20 June 2007 (UTC)

I quite like the current image and would probably have selected it from the commons set as a replacement. It gives an idea of the size of the board but retains an artistic angle. I guess I don't see anything better currently on commons, do you have one of your own? ∴ here…♠ 02:44, 20 June 2007 (UTC)

Image would better improve if camera f-stop was higher so all image was in focus at foreground and background. 4.233.143.116 13:15, 3 July 2007 (UTC)

will email article

Found this as browsing: The Art of Black and White: Wei-ch'i in Chinese Poetry Zu-Yan Chen.Journal of the American Oriental Society.

If it's useful, email me (link on my user or talk page). Will send to you. Ling.Nut 22:19, 27 June 2007 (UTC)

> It can also be found on the AGA site as pdf. HermanHiddema 12:03, 2 July 2007 (UTC)

Cool. Ling.Nut 12:27, 2 July 2007 (UTC)