Is there a free or relatively cheap computer game with an instructor to show me some of the finer points?
You are in luck, go is very easy to learn in the sense that it only takes a few minutes to pick up the rules. Of course, being a good player is a different matter entirely, but probably you suspect this.
There are some ok youtube videos on learning the rules of go, but I think finding some software to practice with is a good idea. I don’t know if there is some really great software for learning that is free, but if you use some software in conjunction with some of the tutorials out there, you should be up to speed really quickly.
Most beginners are pushed to play 9x9 go in order to get the feel for the game. The group that makes the AI “Many Faces of Go” has a free version which plays 9x9 with you:
If it is anything like it was in 2004, it will get you up to speed quite fast.
There are many go servers, but a popular non-asian server is KGS. It has a lot of rooms for beginners, like “Beginners room” and “KGS Teaching Ladder.”
http://www.gokgs.com (java)
Good luck and have fun!
Might want to mention that searching the internet for “go” is, well, kind of impossible. You can try these instead: go game, igo, weiqi, baduk.
I think the best all-around English language resource is the wiki site Sensei’s Libary:
In any particular aspect, there is always some other better resource, but I’ve never seen a better general resource.
I’d strongly advise against spending any money on the game as a beginner. If you gain some strength and find you love the game, the money you could drop on it is seemingly without limit, so believe me, don’t rush on that particular path.
I’ve been trying to learn go as well (i.e. learn to not be terrible at it.) I learned the basics from online tutorials, but I agree that playing is the only real way to “get it”. I’ve been practicing with “goDroid” on my android phone. If you’ve got an android phone, I highly recommend it. After lots of practice, I managed to beat it on 9x9… unless you count Komi and Captures. 
Still haven’t had a real win, but I moved up to 13x13 anyway, to try to understand how and when to shift focus to different parts of the board.
Good luck! I am a huge gamer, and this “simple to learn” game has given me the most trouble to learn to play well. It’s amazing…
Most material I’ve ever seen online is for players beyond what a new player would call “a beginner.” See below for more on that topic, but I’d guess almost all material is aimed at people 15k to 5k. Most go material everywhere is like that: self-study, rather than directly-taught play, is considered the only way to progress.
It is hard to overstate the depth of this game. It seems impossible it could be so: very simple rules, pieces in play aren’t moved, all pieces are indistinguishable… but the depth is just astounding.
For those not in the know, just as a general reference point, here is the commonly-accepted scale in Japanese terms:
[ul][li]Absolute novices are considered, nominally, 30 kyu. This is almost totally meaningless, but the scale starts somewhere. 30k is “I have probably played the game once or twice; I know the rules.”[/li][li]Ratings increase one level at a time, from 30k (kyu) to 29k, …, 2k, 1k. Then the dan levels start at 1d, 2d, …, up to 7 or 9 depending on the organization. I think all online servers go to 9. In theory, anything more than this is a professional dan, but professional ranks don’t work like amateur ratings.[/li][li]Statistically, ratings below 25k are very, very unstable because people usually either decide they don’t like the game, or play enough to improve beyond this level fairly quickly.[/li][li]In an even game, where both players are the same strength, black moves first and white gets 6.5 points (called komi) as compensation for going second. One difference in strength, often called a “stone” (e.g. “he’s three stones stronger than me”) means white gets only 0.5 points (tiebreaking purposes) and the weaker player takes black. Two difference in strength means black gets to play two moves before white plays, and so on.[/li][li]This handicap system is, generally, considered to yield even play at least to 6 stones difference. Generally, no one plays 19x19 games with more than 9 stones handicap, regardless of strength difference. Apart from instructors, playing with someone that much stronger than you has a good chance of not being pleasant or particularly enlightening.[/li][li]This handicap system, online, is what defines the relative strengths. If you win, on average, about half of your games when giving 2k two stones, you’re 1d. [/li][li]The last point is telling in another way: go will always be hard. If it isn’t hard, your rating increases until it is hard again! Expect to lose half your games for the rest of your life. [/li][li]In a year of fairly dedicated but not “second job” level of play, one can expect to make it from 30k at the outset to anywhere between 9 kyu and 1 or 2 dan. A not small number of people will never be dan level. Your “typical strong club player” is probably about 2k to 3d, depending on the club, the rating organization, etc.[/li][li]It is not uncommon to consider all kyus “beginners” and all dans “amateurs” but I think this use is becoming non-standard.[/ul][/li]
I’m trying not to flood this thread, honest, but I love the game and am happy whenever I can share what I know. Hope it is interesting for you all.
I read of a study a few years ago meant to determine which board games were most complicated, in the following sense: take the best player in the world, determine which opponent he would score 75% against over the long haul, then look for the guy against whom player B would score 75%, and so on until you get down to the rankest of rank amateurs. In chess, for example, they found this took about 14 steps, while in go it took 40 to get there.
Have a look in your area for a Go club. For example, there’s the Canadian Go Association, American Go Association, etc.
Playing well against the computer is one thing; playing a human who knows what they are doing is a very different matter. Also, the human can teach you much more about the game. I was playing very well against ‘The Many Faces of Go’ - then a very patient, average player from the Toronto Go Club handed me my ass on a platter one Sunday afternoon. Over coffee, we went over the game we had just played; I learned more that afternoon than in several months of computer play.
The AGA has “The Way To Go” which is a good introduction
http://usgo.org/usa/waytogo/index.html
Brian
That seemed high to me, so I dug in a little bit. The model used by the KGS go server would indeed make 1 rating difference equivalent to something like 70-75% chance of the stronger player winning, which means in fact that 40 is not only plausible, it is probably too small, if we go from the strongest professional players to the noobiest noob in Noobtown. Here is a pro’s rating graph on KGS, mostly some exhibition games and non-serious stuff (Chinese 2 dan professional):
http://www.gokgs.com/graphPage.jsp?user=milanmilan
That’s 10d, though the system caps nominally at 9d. MilanMilan is very strong but there are definitely stronger pros.
Definitely. Though recent AI advances have created some amazingly strong bots. There was a man-machine challenge issued by a strong amateur way back in 1997 (I think). He thought no bot could beat him in a 10-game match under tournament-ish time settings by 2010. That held up, and the one challenger paid the $1000 for taking the bet—I think it was Many Faces of Go, actually. Well, it is 2012 now, and there was a 5-game exhibition followup. It was set for best of five, and the score was 3-1 in favor of the bot. “Zen19N” is the name of the bot, here are the records:
http://www.gokgs.com/gameArchives.jsp?user=tromp&year=2012&month=1
So, they’re starting to get there. Both Zen and Crazystone are 4/5 dan on KGS, though they still play “bottish” moves and still aren’t excellent teachers.
I have a collection of those books from Ishi Press - the most useful one I’ve found is ‘Life and Death’, by James Davies. One of the big concepts he gets across is developing the ability to look at a section of the board and tell right away - is this group alive, is it dead or is it in flux? What is the one move that will kill it/give it life?
One of the ways in which my friend clobbered me was to bluff his way out of my sente. He would place a stone somewhere else on the board, somewhere I thought was settled. I fell for it every time - ‘What’s he doing that for? I’ve got more eyes than a potato over there; I must have missed something…’ and instead of continuing my pursuit, I’d shore up something unnecessarily and hand sente over to him. I’m not sure if computer programs have picked up on that ‘never give a sucker an even break’ trick that ruthless humans innately possess.
If the stronger bots overplay, they do so only accidentally. There are some common sequences that are known not to work that do get played a bit, because to find out they don’t work you’d have to read ahead like 15 moves or more, which is basically impossible for most everyone. But, they’re studied positions, where the possibilities have been explored and responses generally known. I have this whole theory on why bots play them but it is really tangential.
The strongest bots also tend to play tenuki too much; at least, according to most sensibilities. I also have some deep opinions on this. In the end, though, their fighting ability is very good and their ability to hold onto a lead in the endgame is extremely strong. Most games where professionally-researched positions aren’t screwed up, the human has to win early, strategically. If the bot smells a respectable lead it can hold onto it like nobody’s business. They’re probably like 1k-2d in fuseki, 4d in fighting, and 7d in yose. (These are not my estimates, but estimates of 5-7d players who are familiar with the bots. I wouldn’t know 7d yose from a hole in the wall.)
The thing that helped me the most, besides discussing some games with a couple of strong players I know*, was the book ‘Lessons in the Fundamental of Go’ by Toshiro Kageyama.
The book covers the fundamentals (duh) very well. The thing about Go, for me at least, was changing my thinking. I was used to games where the attacks were more direct, sacrifice was never ever done and the fight was for the whole board.
The last one may need some explanation. When I first started playing I would get to focused on a particular fight and end up winning the fight but losing the game. I wasn’t looking at the whole board.
If you can, play against humans. The Go programs I have played usually focused too much on fights and not on the whole board. The only Go program I have played recently is one on my Android phone that I can kill fairly easily on the highest setting giving it 9 stones. Once I figured out what the program tried to do every game, kill groups, I knew how to beat it.* *
You won’t win fights with any decent Go program. They can tree out the fights pretty well. But Go isn’t a game about fights most of the time.
Playing a stronger player who can go over the game after is a huge plus. The thinking is so different than other games that I have played that getting an understanding of why the moves were made helped a huge amount.
Last, play a lot of games and have fun. Try something different every game.
Slee
- The guys I learned from were ~4 or 5 Dan at the time. I played them for a while and didn’t do much. I didn’t play them for a while but studied the Fundamentals book. When I played them again I had moved up to about 1 or 2 k. Huge difference. I haven’t really played in a couple years so I am sure I suck again…
** I haven’t looked at the newer programs lately. I don’t have anyone local to play and I don’t like playing on the computer as much as in person. Plus life is busy and my fiancee prefers I hang out with her instead of my computer. Apparently the programs have gotten much stronger. Many Faces was pretty weak when I last played it.
I have a computer Go game called “Enter the Rising Sun, Stone Master”, which I recommend. However, I don’t see it available on Amazon. I also have Konami’s “Hikaru No Go 2” for Gameboy Advance. However, it doesn’t seem to work on my GBA.