Talk:Go (game)
Please note that the simple rules do *not* make suicide illegal. That's an extra rule, omitting it does not change the character of the game because suicide is usually a blunder. The purpose of the simple rules is to have the minimum rules that people need in order to play what we call "Go". I've based these on the work by Robert Jasiek at [1]. -- Bignose
Why did some delete EVERYTHING I wrote about Pente and Go-Moku? Wikipedia does not function by an anoymous person's unilateral censorship. (Worse, the person who did the censoring is igonorant. Pente certainly is a Go variation.) I never would have imagined that a person could have something against a Go variation, and would violate Wikipedia protocol to delete all mention of it! In wikipedia, we discuss such things in the talk section before deleting entire sections! The deleted material WILL definately be restored. But now others are going to brought into this as well. What happens next? Deletion of mention of mancala and chess variations? The mind boggles at who would do such a thing, and why. RK
- Slow down a bit. It wasn't an anonymous edit; the revision history says it was AxelBoldt who made the change. He moved your material to Pente, with some additions. --Stephen Gilbert
The entire section titled "Other board games commonly compared with Go" seems highly irrelevent. I don't know anyone who confuses _any_ of these games with Go at all; they have nothing to do with each other at all. Are there really people who think that ALL board games are basically the same? What is the purpose of this section? And how can a rational person claim that a Go-variation like Pente or Go-Moku is just as distantly related to Go as Chess (which obviously has zero relationship to Go). It seems as if someone has a personal grudge against the games of Pente and Go-Moku! We have a case of snobbery on hand. RK
- The section is not about games which are "confused with Go", but about games which are "compared with Go". As such it is relevant. Go-Moku is not a variant of Go; the only similarity is that it is played with the same equipment. It does not deserve mentioning in the second paragraph of an article about Go. --AxelBoldt
- As the article states, people first experiencing Go will naturally try to compare it with other games they may be familiar with. These comparisons are not about which game is "better", or about confusing the game with others, but about drawing on common experience to learn something new. People often find it difficult to approach Go unless they can see where it fits in the game pantheon; though such effort may be futile, it is still information people frequently want to know. -- Bignose
Though the word spelled the same, it is a Japanese word which pronounced differently in Japanese. The 'o' sound in Japanese sounds like the O in 'Octopus'. Since the English verb 'go' is too common, people tends to pronounce the name of the games as an English word. Same problem with other words like Toyota where the O is pronounced as Oh in the western world; Karaoke is also pronounced differently from the Japanese original.
- This is true. Are you presenting this as a problem with the article, a suggestion for inclusion in the article, an interesting point for discussion, or something else? I'm not sure what you wanted to provoke with this comment. If you feel the article needs modification in some way, please feel free to do so. -- Bignose
I have a question regarding the Ko rule: is it not allowed to recreate the previous board position, or is it not allowed to recreate any previous board position? -- AxelBoldt
- The ko rule is expressed differently in different rulesets. Japanese rules effectively says "can't recapture the stone immediately", and then has extra rules tacked on to cover things like triple kos. Other rulesets use "superko", which disallows recreating any previous board position. I don't know whether we should be getting into obscure rules differences here. -- Bignose
Also, regarding the 9x9 boards: what happens if strong players play 9x9? Is it always a draw, or does black always win? How do computers fare against human players on a 9x9 board? -- AxelBoldt
- 9x9 boards are quite challenging even to professionals; however, the game is much shorter and tends to be almost entirely tactical instead of strategic.
- Computers tend to fare poorly once brute-force searching of the game tree is ineffective, which is the case well below the 9x9 board size. They also need a good position-evaluation algorithm to decide which moves to explore, which is proving very difficult to discover. -- Bignose
IIRC, I once saw two 9x9 games played by a strong professional against himself. Both had black winning by 4 points. I doubt that there is enough subtlety in the game for white to win more than 1% of the time between human experts. Computers are a completely different story. They are still so weak even at 9x9 boards that a superior computer (say, go4++) can win from either side. Top humans can still annihilate top computers even on small boards. -- Karl Juhnke
Professionals play 9x9 go with komi, just as they do for the full board. So white wins about as often as black. There's relatively little pulished about 9x9 theory, but the occasional professional commentary on a professional game shows that there's a good deal of shared knowledge about it.' Matthew Woodcraft
Right, it is typical to play both 9x9 and 19x19 with a 5.5 komi. But it isn't clear to me that 5.5 is fair, or that a fair number is the same for different size boards. It is quite plausible that a fair komi for 9x9 is 4.5 while a fair komi for 19x19 is 6.5. -- Karl Juhnke
- The distinction between each rank is, by definition, one handicap stone. In other words, the difference in rank between two players is theoretically equal to the number of handicap stones required for a balanced game between the players
I am not sure that the second sentence follows from the first; in fact I think it it quite surprising that there be a transitivity as follows: A and B are balanced with 1 handicap stone, B and C are balanced with 1 handicap stone, therefore A and C are balanced with two handicap stones. Is it actually true? -- AxelBoldt
- The terms "by definition" and "theoretically" are sprinkled liberally throughout the discussion on ranks; it is a model to allow players to quickly gauge relative strength and to allow a structure to be given to a pool of players. In fact, once players reach a certain level (somewhere in the mid-kyus) their rank is a very good indicator of who they can expect to play an even game against, and how many stones are required against players about their own level.
- Your suggestion, that one's strength will need to be measured individually against each player, was the way the Japanese professional players decided handicaps in the Edo period (top-level players would have individual handicap ratings against each other, and a record of games against each player was required to be kept to know when the handicap should change against that player).
- However, those were leisurely times, when players were sponsored by the school and would often take a week or more to play a single game, and the status of each professional player against the others was closely watched by the Imperial court, so attention to such minor detail was rewarded. These days, in a pool of thousands of players, swiftness of ranking is more valuable, and the theoretical assumption (that a difference of X ranks requires X handicap stones to compensate) is quite close to the reality of how playing pools evolve, and hence "good enough". -- Bignose
Also, is there any formalized way to proceed to the next rank, or do you rank yourself? --AxelBoldt
- Naturally there are formalised ways, but those ways are formalised differently in each country depending on the rules of the professional (and amateur) Go associations. I don't think it's appropriate to delve into detail in this entry, since it would be unbalanced without coverage of all the different associations, and terribly dull with such coverage.
- In a small pool of players without a guiding Association or other such body, one can easily fall back on the ranking system to provide a guide to relative ranks. Since two players at the correct handicap should be "even", they should each statistically win half the games. Once one player is winning consistently at that handicap, he probably needs to go up a rank; a rule of thumb commonly employed is "three wins in a row at the same handicap against the same player" is sufficient to promote the winner by one stone.
- Yes, this ignores the fact that players will tend to improve over time regardless of wins or losses. No, there is no method commonly employed for demoting players. Yes, this does lead to ranking levels that drift over time. The ranking system only attempts to measure a current pool of players against itself; it has no relevance to a different pool of players (hence the drift experienced between amateur and professional ranks, or rankings in different countries), nor to the same pool of players at some point in the distant past (today's 9 dans are likely much stronger than the Meijins of long ago). -- Bignose
This article is getting better and better. I have one request: could we have an introductory article about Go Strategy and Tactics, akin to Chess strategy and tactics, aimed at beginners who just want to know a couple of rules of thumb? -- AxelBoldt
- I've started a Go strategy and tactics page. -- Bignose
Also, the article says the rules of Go are simple, but judging from the official rule set at the Go wiki, it seems to me that when it comes to counting and judging dead and alive groups, they are actually much more complicated than e.g. the Chess rules, and in fact they don't even seem to be completely finalized. There is a big catalog of positions that have been "judged" by some organization, and these judgements are technically part of the rule set. -- AxelBoldt
- That big catalogue of positions and precedents (which is periodically added to, and still cannot cover every possible board position) is part of the Japanese ruleset only; other rulesets resolve the status of groups simply by continuing play in the case of a dispute until both players agree on the status (which neatly encompasses all possible board positions). I hope I've answered this sufficiently in the article, without compromising my desire to see the explanation be comprehensive yet simple. I really don't feel we should emphasise the differences in rulesets to a degree more than necessary. -- Bignose
Imran, you're fairly new, so you didn't know, but when you create a disambiguation page, you need to fix all of the links to the ambiguous article to point to the correct unambiguous article.
- I know, I was correcting the links when my internet connection decided to break and I wasn't able to reconnect for a while. --Imran
Also, if it looks like almost all of the links are for a particular one of the meanings of the term, then it might be better to use a disambiguation block rather than a disambiguation page; see Set for an example of this. More information is at Wikipedia:Disambiguation. — Toby 14:47 Jul 26, 2002 (PDT)
PS: Why were the CJK characters deleted? Were they wrong? — Toby
I spot-checked a couple of them with the Unicode database, and they were good, so I'm putting them back. --LDC
Go (pronounced somewhere between English /gaw/ and /go/)
- Is this really correct? This seems to be describing the pronunciation of the Japanese word go. The English word go for the same thing is pronounced with a regular "long o" (phonemically /o/, phonetically [@U] or [oU] depending on dialect) according to the American Heritage Dictionary, and I have yet to find a counterexample. --Brion VIBBER 15:27 Jul 26, 2002 (PDT)
I would tend to agree, but it may be one of those cases where afficianados of the game might go out of their way to pronounce it in the Japanese way. I like AHD a lot, but dictionaries aren't always the last word on such things, since they are made for a general audience and often leave out domain-specific nits. --LDC
- I don't think I've noticed anyone pronouncing it the Japanese way in the UK. There are lots of Japanese terms used in go, but the pronunciations are generally anglicised. --Zundark 01:28 Jul 27, 2002 (PDT)
I removed this paragraph:
- The highly competitive qualifications for professional players have the side-effect that the two pools of players do not mix in a way that affects rankings; thus, the rankings for professional and amateur players have diverged to the point where amateurs at a given rank are much weaker than the equivalent professional rank.
This is wrong, as (in Japan, at least) the amateur ranks were always on a different scale from the professional ranks. --Zundark 10:04 Jul 29, 2002 (PDT)
David, the usage group is more common than the perhaps more technically correct string. If you wish to introduce string as the preferred usage, please preface the first use of it by saying who prefers it and why. Otherwise, it will confuse many readers. --Ed Poor
"Stones occupying adjacent points constitute a solidly connected group." British Go Association
Ed, David is right. group is more common, but it's used to mean a different thing - as David said, a loosely-connected collection of stones. The BGA page you point to is using solidly-connected-group as a set phrase. It would be best for the rule exposition not to use group on its own to describe solidly-connected stones, as this will confuse people when they see it used in the more common sense later. --Matthew Woodcraft
The following text was added to a page I-go; I've made that a redirect here, but maybe some of it is worth incorporating into the main article. --Matthew Woodcraft
Go is unique among games , it originated in China for more than 3000 years.
Introduction to the game of Go can be found here: [2]
external link :
http://leaohp1.epfl.ch/~warkent/go/golinks.html
software about Go game may be found here :
[3]
computer go,
is relative to computer AI, Computational Intelligence in Games netrul net
external links :
http://www.reiss.demon.co.uk/webgo/compgo.htm
http://www.gnu.org/software/gnugo/gnugo.html
http://www.cgl.ucsf.edu/go/ladder.html
http://www.markus-enzenberger.de/compgo_biblio.html
Above, Karl Juhnke implies that a fair komi for 19x19 may be 6.5, and for 9x9 4.5. However, I believe that this is backward. The first move has a far larger effect on a small board than on a large (some people believe, for example, that if black plays the first move on the center point on a 9x9 board, the game is won by at least half a point, assuming fair komi and that both players are 'perfect'). In fact, some computer go software uses the standard 5.5 for 19x19 and 6.5 for 9x9 (while other software gives one a choice of various komis for a single board size).
It would be interesting to see some mathematical methods for determining komi. These might result by using the mathematical ruleset on boards, inductively, from 2x2 up to whatever upper limit is practical. The resulting function of komi with board size can then perhaps be extrapolated up to 13x13 and 19x19. David 20:18 Oct 12, 2002 (UTC)
Yes, the first move has larger effect on smaller board sizes, but also 1 point of komi has a larger effect. So it is not at all clear in what way the komi should change by board size. As far as I know, the standard komi is 5.5 for 19x19, 8.5 for 13x13 and 6.5 for 9x9. For 19x19 this is generally considered to be too small, and more and more it is changed to 6.5 or 7.5, I don't think there is any such agreement about the other two sizes. Andre Engels 15:28 Oct 13, 2002 (UTC)
Simple rules * these are confusing Axel, visually and otherwise, and nobody, i mean nobody whose a native english speaker refers to groups as 'strings'. Maybe quoting some awful book. Ref: Senseis lib -Stevertigo
But those rules aren't talking about groups when they say 'string'; they're talking about collections of solidly-connected stones. I agree that it would be useful to clarify that the term 'string' is being introduced only to explain the rules, and isn't a term in everyday use otherwise. Matthew Woodcraft