Life In 19x19

Hello everybody! In the book "Opening theorie made easy" by Hideo Otake 9p I read the following:
Enclosing a corner takes priority over a side extension. Hideo gives 20 opening principles in his book (the above is in the 1st chapter "Corner Enclosures aim at side extensions") which I find very helpful as I am so used to it from 30 years of chess.
There are e.g. the classical openings, which should be learned especially by beginners. Something like "the Spanish Game", or the "Queen's Gambit". So far I have not found anything like that in Go. It is also difficult to find games that illustrate these 20 principles. At least I can't find anything.
For me, master games I find everything that can be played like that.
Does anyone have a tip for me where I can find games for beginners that are really instructive?

Go doesn't have full-board named openings like chess, instead you have the standard corner joseki, some of which have names, plus various named openings that take up a side or two (typically for black) like Chinese opening, Kobayashi opening, or san-ren-sei. San-ren-sei is still a san-ren-sei if white changes her 2 first corner moves from say 4-4s to 3-4s, though how it continues will depend on them. Otake's book will be a bit dated on some principles vs the latest understanding of high level go from AI (e.g. side extensions you mention are now seen as less valuable than back then), but most of it will still be good ideas and I still recommend it to beginners. If you want to find games following these kinds of principles, I recommend to look for older Japanese pro games from mid to late 20th century, Otake himself is a prime candidate as he was known for his clean and elegant style with an emphasis on good shape. Kobayashi Koichi from the 80s - 90s is another.

Uberdude wrote:Go doesn't have full-board named openings like chess, instead you have the standard corner joseki, some of which have names, plus various named openings that take up a side or two (typically for black) like Chinese opening, Kobayashi opening, or san-ren-sei. San-ren-sei is still a san-ren-sei if white changes her 2 first corner moves from say 4-4s to 3-4s, though how it continues will depend on them. Otake's book will be a bit dated on some principles vs the latest understanding of high level go from AI (e.g. side extensions you mention are now seen as less valuable than back then), but most of it will still be good ideas and I still recommend it to beginners. If you want to find games following these kinds of principles, I recommend to look for older Japanese pro games from mid to late 20th century, Otake himself is a prime candidate as he was known for his clean and elegant style with an emphasis on good shape. Kobayashi Koichi from the 80s - 90s is another.

Many thanks!!!

https://senseis.xmp.net/?GeneralOpeningPrinciples

is a good set of principles, with the most helpful in my opinion being:

* Corners, then sides, then center
* Do not get hemmed in
* Play urgent moves before big moves; (aka stabilize existing groups before moving into undisputed area)
* Take the last big point
* Play away from thickness and Push the enemy towards your thickness
* Play dual-purpose moves
* Your opponent's good move is your good move

You will find separate pages for each of those in above mentioned index.

These "general" opening principles are not general but, if seen as advice for beginners, later require unlearning.

In the opening, play what you want as long as you also consider the global picture and avoid (big) mistakes, such as choosing the smaller or dying big.

Let me use the states of matter as an analogy:

In the opening, a go position is similar to a gas. There are only a few pieces on the board and they have a maximum amount of space and freedom.
Then it becomes more like a fluid: the middle game is more crowded, but very dynamic with lots of interactions between pieces.
Towards the end of the game the board fills up more and more and everything gradually becomes fixed like a solid.

A game of chess progresses more or less in the reverse order,
from a fixed position in the opening with a maximum number of pieces on the board,
towards a very dynamic middle game with lots of interactions between pieces,
then towards the endgame with less and less pieces on the board, that have a maximum amount of space and freedom.

So perhaps the opening of go is more like the endgame in chess.
I think the endgame of chess also has lots of general principles and theoretical considerations, but chess endgames don't have names like chess openings. I think that is because there is just too much variation in chess endgames to name every position. So that would be my explanation for go openings in general not having a name.

But there are names for common patterns in go (joseki). I suppose chess also has named patterns (besides names for openings).

Uberdude wrote:Go doesn't have full-board named openings like chess, instead you have the standard corner joseki, some of which have names, plus various named openings that take up a side or two (typically for black) like Chinese opening, Kobayashi opening, or san-ren-sei. San-ren-sei is still a san-ren-sei if white changes her 2 first corner moves from say 4-4s to 3-4s, though how it continues will depend on them. Otake's book will be a bit dated on some principles vs the latest understanding of high level go from AI (e.g. side extensions you mention are now seen as less valuable than back then), but most of it will still be good ideas and I still recommend it to beginners.

Otake's ideas are fairly standard. IMHO, they are more than dated. I would strongly suggest looking at pro games from 2018 onward. The pros have learned to play better openings.

If you want to find games following these kinds of principles, I recommend to look for older Japanese pro games from mid to late 20th century, Otake himself is a prime candidate as he was known for his clean and elegant style with an emphasis on good shape. Kobayashi Koichi from the 80s - 90s is another.

Here I would strongly suggest looking at games from the mid to late 19th century. Why? Two reasons: 1) That's when these ideas were almost fully developed. 2) They are almost all validated by AI, in context. The reason, I think, is that the initial 4-4 in the corner was not popular back then. In that case, the corners were mostly enclosed or approached before extensions on the side were made. It is the early extensions and pincers that are questionable or even mistakes, as a rule.

RobertJasiek wrote:These "general" opening principles are not general but, if seen as advice for beginners, later require unlearning.

Well worth repeating.

Here is an opening principle that I think I am the first to formulate.

Occupy the last open corner, as a rule.

OC, today that usually happens an move 4, but it's still a good rule of thumb.

Also, in general avoid the initial 5-4 and 5-3 in a corner.

Hint: If your opponent plays an initial 5-4 or 5-3, invading on the 3-3 is usually right.

Bill Spight wrote: I would strongly suggest looking at pro games from 2018 onward. The pros have learned to play better openings.

Thanks for your insights!

In chess it better to learn the classics first.

Caveats about 19th century and early 20th century games.

They were nearly all no komi games, which affected the play in a number of ways. One way is that Black tended to play keima (small knight's move) enclosures, which are not preferred by AI today. But keima enclosures are more solid than large knight's move enclosures, which were preferred by White, and by AI today. AI also like high 2 space enclosures from the 3-4.

Another way is that White generally preferred to approach a corner instead of occupying the last open corner. Occupying an empty corner was considered to be theoretically superior to approaching a corner, but as a practical matter it was felt that the approach made more difficulties for Black.

jumapari wrote: In chess it better to learn the classics first.

These days I suppose that aspriring young players study a lot with AI. But previous generations studied the classics a lot. The openings of those classic games are considered old fashioned, but other than that, these games are of high quality.

The most famous examples are Honinbo Shusaku (who lived about 200 years ago) and Honinbo Dosaku (who lived about 350 years ago).

Here is a video review of one of Dosaku's games: https://www.youtube.com/watch?v=STr1o1JrC20

gennan wrote:

jumapari wrote: In chess it better to learn the classics first.

These days I suppose that aspriring young players study a lot with AI. But previous generations studied the classics a lot. The openings of those classic games are considered old fashioned, but other than that, these games are of high quality.

The most famous examples are Honinbo Shusaku (who lived about 200 years ago) and Honinbo Dosaku (who lived about 350 years ago).

Here is a video review of one of Dosaku's games: https://www.youtube.com/watch?v=STr1o1JrC20

Thanks!

In chess, computers have been stronger than any human for more than 20 years right now. A development which, however, was not as surprising as in Go, but was already taking place for about 20 years since the end of the 70s of the 20th century.
But up to a playing strength of at least Elo 2500 (I'm close to 2300) it is irrelevant if you try to learn something from the AI. Here the study of human games, beginning with classical music, is sufficient. It is also of little use to analyze your games with a computer. It just shows you tactical errors, but it doesn't explain concepts
Surprisingly, the opening theory was hardly changed by the AI in chess.
Magnus Carlsen is an exception. He claimed to have studied the games of Alpha Zero before the last World Cup match. He is not by chance the strongest player in chess.
There are, however, a number of strong grandmasters who are completely indifferent to computer and AI chess, and who still maintain their playing strength.

jumapari wrote: In chess, computers have been stronger than any human for more than 20 years right now. A development which, however, was not as surprising as in Go, but was already taking place for about 20 years since the end of the 70s of the 20th century.
But up to a playing strength of at least Elo 2500 (I'm close to 2300) it is irrelevant if you try to learn something from the AI. Here the study of human games, beginning with classical music, is sufficient. It is also of little use to analyze your games with a computer. It just shows you tactical errors, but it doesn't explain concepts

I think it's somewhat similar in go. Studying with AI is probably not very useful until you're getting close to dan level. Below that level you need human teachers to explain concepts and heuristics that guide human thinking about the game.

jumapari wrote: Surprisingly, the opening theory was hardly changed by the AI in chess.
Magnus Carlsen is an exception. He claimed to have studied the games of Alpha Zero before the last World Cup match. He is not by chance the strongest player in chess.
There are, however, a number of strong grandmasters who are completely indifferent to computer and AI chess, and who still maintain their playing strength.

I think there's a big difference between AI (large neural networks trained by many millions of self-play games) and engines (alfa-beta search with highly tuned human coded algorithms and evaluation functions).

In chess, engines beat humans mainly by superior tactics for decades already, while the engines' openings are still based on a database of human knowledge. So for top human players there is not much to learn from the engines in regard to the opening and strategy.

But there is currently a revolution going on in computer chess, where the superior tactics of the mighty engines is being trumped by the superior strategy of highly creative AI like A0 and Lc0 (which also learned about the opening all by themselves). I think that young aspiring chess players may learn novel strategic concepts from these AI (even in the opening), that the older generation couldn't learn from the engines.

BTW: Nobody ever managed to make a strong go engine, so go players never had to deal with that. So it's been quite a sudden jump for us from pretty crappy engines to demi-god level AI (tactics and strategy) in just a few years.

Even if some pre-AI opening are now out of fashion (sanrensei, Chinese opening, Kobayashi,...) I think it's still interesting to study them at least at kyu level. First because the ideas you learn may be useful in other situations. And second, because these openings are still played in amateur games, and an opponent who knows these openings well may be able to trick you.