The goals of joseki

[Joaz Banbeck]

I've seen it again and again: a 20-30K beginner posts a game, and asks for advice. The first 4-5 moves look ok: they are plays in the corner, and maybe mid-side. They could even be mistaken for professional play.

Then the battle is joined, usually in a corner, and the game goes to hell so fast that it is laughable. Players make pointlessly bad moves. Sometimes even random moves would be better. They clearly have no clue what they are doing. Occasionally they follow known josekis for a few moves, but as soon as they are out of their book, the cascade of horrible moves commences.

As a reviewer, I often find myself at a loss in such situations. The ignorance is so great that I am not sure where to begin. Sometimes, the best I can say is to read a joseki book. But that does not solve the fundamental problem. It merely delays its presentation.

What I want is a good, simple statement of goals in joseki.

My tentative try is like this:
1) Try to stay connected
2) Separate him
3) Make eye space
4) Deny him eye space
5) Ensure that you have room to run to the center
6) Block his access to the center
7) Get influence

Those are some of the things that I find myself striving for in joseki. But I am not happy with that list. It is almost certainly incomplete, it has no prioritization, and some goals are in direct contradiction with others. It is probably confusing to beginners.

Can someone else do better?

EDIT: Maybe asking for a list is a bad idea. Perhaps an algorithm for decison making? Once you are out of your book, how do you decide what to do in a corner? What guides you?

[Tryphon]

I know relatively few about joseki considering my rank, but I read some books about shapes and found myself to actually play joseki moves, just trying to make good shapes or reading some moves. Which doesn't mean I don't sometimes fall into a trap, or misread something and get a huge los at beginning.

I remarked too that joseki often show optimal moves, even if they are difficult, whereas suboptimal moves can be lot easier and sufficient at a weak level.

So, I often recommand to beginners to read about basic shapes and basic techniques of life and death, rather than joseki books.

[jdl]

Yilun Yang's lectures cover this topic in a way that I love. At my level I can understand what he's saying, but it's still difficult to grasp. In other words I can study it, but it's hard. He has two major themes.

Which joseki?
The "why" or the "strategy" part of playing josekis. This is covered heavily in his books "Whole Board Thinking in Joseki (1-2)" and a chapter in "Fundamental Principles of Go."

How to play a joseki
This is the "how," the "tactical," or the "don't get slaughtered and lose everything" part of the joseki. He teaches the tactics as a close-quarter cross-cut fight, rather than memorizing specific moves. "The Workshop Lectures: Volume 1" has a chapter on this that was eye opening to me in the same way as "Attack & Defense" was the first time I read it. He takes the taisha joseki and breaks it down move-by-move into a corner fight, explaining each move in terms of the local fight rather than "it's the book move."

This is also a running theme on his monthly KGS+ lectures. Almost every one of those will eventually have him asking about the proper thought process during a cross-cut fight, which has applications in nearly every joseki.

So, while I can't answer your question directly, I figured I'd point you to a pro who is also tackling the problem of amateurs improperly wielding these weapons.

[Mnemonic]

I guess that kind of depends on who your target audience is. There is a certain evolution that we go through when we first start playing on the 19x9. The first on is to just play like on a 9x9. This game sadly has handicap stones s it doesn't capture my point quite as well, but I think it works quite well (notice that the 4 handicap stones are the closest anyone ever got to a corner )

http://www.lifein19x19.com/forum/download/file.php?id=922

To this I think the only thing stronger players can reply is: TENUKI!!!
Which brings us to the tenuki fuseki stage and a jump to 18k

http://www.lifein19x19.com/forum/download/file.php?id=923

Finally someone sits the beginners down and explains some basic go strategy and we get to the 'invent your own joseki' stage depicted in this game.

http://www.lifein19x19.com/forum/download/file.php?id=921

[John Fairbairn]

Can someone else do better?

EDIT: Maybe asking for a list is a bad idea. Perhaps an algorithm for decison making? Once you are out of your book, how do you decide what to do in a corner? What guides you?

Robert Jasiek's two Joseki books probably have all you will ever need and he has probably done it as well as you will ever see in this niche market. Admittedly Volume 2 has lists on about 67 of 266 pages, and some pages have two lists or lists that fill a whole page. The list proportion is much higher if you take out the many example pages and count just the pure text. There is also the business of adapting to terms such as n-connection and -n-connection, or q-territory, and so on. But, offputting as that will sound to some, if you treat the books as a vade mecum or reference book you will find that everything there has a structure of sorts and is consistently defined, and it is also reasonably complete. The examples in Volume 2 are also excellent, as are some of the definitions.

I'm not sure a dan player would learn much (though he'd probably get a stimulus for further study from a new way of looking at several things), and I certainly don't accept Robert's claim that a professional would benefit from it. But teachers looking for the sort of "curriculum" you want would be especially well served, and I'd say that any player up to low dan level would find it invaluable in the sense that, if you encounter a difficulty somewhere, you can turn there and get, almost always, another way of looking at things, something which seems to be important for true learning.

In fact, I'd go further and say that every serious go player up to about 1-dan should buy at least Vol. 2. Unless you like lists you probably won't enjoy reading it as continuous prose (note, though, that Robert has made a big attempt with the examples to overcome that aspect), and you may initially resent your outlay of hard-earned cash. But so long as you treat it as a reference book and actually use it (as I say, I see it literally as a vade mecum where you could, say, take this book with you to tournaments to check things between rounds) you will soon feel you have recouped the modest price, and it will also probably remain one of the books you won't want to give away. If you think about it for a moment, western go is very hard up as regards reference books.

Vol. 2 is physically as well presented as Vol. 1, and the English, though still not perfect, is much improved. In fact, it reads extremely well stylistically in many parts. It is well indexed.

One awful mistake needs correcting, though. Haengma is not pronounced (page 146) haa-aeng-maa. (This has the potential to become the buy-oh-yoh-mee de nos jours.) It's only got two syllables and hengma will do as well as anything.

If you don't actually want to buy a book, jb, I'd suggest reverting to the sadly neglected twin pillars of Japanese joseki pedagogy. Teach some basic patterns and then offer test positions on choosing the right joseki for each position. Pupils will soon sense what each one is for and will then get a feel for moves in new patterns they learn. Most westerners seem just to manacle themselves to the one pillar of memorising josekis. Admittedly there's not a huge amount of literature here to form the other pillar, but there is some. There was a series of books on this in landscape format, as I recall. And there are, of course, teachers like yourself. You truly can become a pillar of the community!

[Laman]

'Having two large groups is better than having one small group.' go proverb

if you talk about 20-30k, i am afraid your goal is unachievable, because such beginners can get even the best decision-rule wrong. (just few days ago i looked at some of my KGS games when i was about 18k and it was sad, how much my stones and groups had to suffer before i learned something )

nonetheless:

- is my group safe? else reinforce (extend, connect, enlarge eyespace, play contact move, sacrifice)
- is his group safe? else attack
- is my important stone in danger (important stone = stone separating two potentially weak groups)? help it
- is his important stone in danger? capture it

- is there an obvious big move (ie. move getting much territory/influence)? take it
- is there an obvious and safe way of preventing opponent from getting a big move? play it
- is there an obvious good looking move other not already mentioned? give it a try
- does your choice go well with your position elsewhere (doesn't it encourage your opponent to erase your influence / to build strength against your weak group)? consider other move (and same reversed)

- apply same evaluation to future moves - if you see something bad, consider other move
- tenuki

i guess my list is not very good, i teach mainly by examples and correcting actual mistakes. big trouble with writing such decision tree is that you have to rely on correct evaluation of relatively complex terms, otherwise the algorithm breaks

is this group weak or strong? is this stone important or not? is it in danger or already dead?

anyway, good luck with your effort, would be nice

EDIT before posting:
i knew something like this would be a topic for Robert Jasiek but i forgot about his books. so solving mentioned troubles is hard, as i said (according to the number of lists in the books), but certainly not impossible

[Toge]

What I want is a good, simple statement of goals in joseki.

My tentative try is like this:
1) Try to stay connected
2) Separate him

- Even these two statements are sometimes contradictory. When you cut a group apart, the very move that separates those groups is now in itself separated from rest of the stones. Only experience can tell what kind of cut is appropriate. Watching 20k game can be frustrating, because the mistakes happen in so many levels that giving appropriate advice is difficult.

I think beginners would benefit most from simple exercises like those in GoChild's entrance level. The understanding must begin from bottom. It's only after you thoroughly understand that you could capture some stone or extend somewhere else to live, that the player understands that neither move is necessary right now. In this example giving advice like "don't play atari" doesn't work without recognizing where the player's own understanding goes. At end of the road the understanding is so deep that the famous ear-reddening move is not 'strange' or 'quite nice' but rather 'essential' to player's mind. I'm in the position where I can recognize most good moves as good, but sometimes wonder why exactly they are played at the time they are or how some other move is refuted.

[Bill Spight]

I've seen it again and again: a 20-30K beginner posts a game, and asks for advice. The first 4-5 moves look ok: they are plays in the corner, and maybe mid-side. They could even be mistaken for professional play.

Then the battle is joined, usually in a corner, and the game goes to hell so fast that it is laughable. Players make pointlessly bad moves. Sometimes even random moves would be better. They clearly have no clue what they are doing. Occasionally they follow known josekis for a few moves, but as soon as they are out of their book, the cascade of horrible moves commences.

As a reviewer, I often find myself at a loss in such situations. The ignorance is so great that I am not sure where to begin. Sometimes, the best I can say is to read a joseki book. But that does not solve the fundamental problem. It merely delays its presentation.

What I want is a good, simple statement of goals in joseki.

My tentative try is like this:
1) Try to stay connected
2) Separate him
3) Make eye space
4) Deny him eye space
5) Ensure that you have room to run to the center
6) Block his access to the center
7) Get influence

Those are some of the things that I find myself striving for in joseki. But I am not happy with that list. It is almost certainly incomplete, it has no prioritization, and some goals are in direct contradiction with others. It is probably confusing to beginners.

Can someone else do better?

EDIT: Maybe asking for a list is a bad idea. Perhaps an algorithm for decison making? Once you are out of your book, how do you decide what to do in a corner? What guides you?

It sounds like you have a good research project here, one that will be worth a couple of stones to you.

I began to study joseki as a 2 kyu. As a 2 dan I still had a pretty poor understanding. The point, as you say, is how to play unfamiliar lines. That's go, folks!

If dan players have trouble explaining joseki mistakes, it is because we still have a lot to learn.

I would not recommend joseki books to DDKs.

[CnP]

There was a series of books on this in landscape format

http://senseis.xmp.net/?WholeBoardThinkingInJoseki

Thanks also John F for providing another book review - it's very useful to have an impartial review of these from such a respectable source (they're next on my shopping list).

[perceval]

i ll second tryphon here and i believe our opinion does matter because we are the closest to those fabled 20k-30k beginners to have posted on this thread:
in doubt play good shape meaning really basic shape (the notion of shape is rathe vague): avoid empty triangle, avoid putting yourself on head of 2 stones, try to have tiger mouth, ponuki, use the table shape, nobi, keima. This turns out to be a list also but i think this one is more visual and easy to integrate than an actual dry list of item with an actual priority order. i always tought that algo-like way of thinking is tedious and thus hard for humans.
A corollary which indeed is on your list is stay connected if possible
That should be enough most of the time (your were referring to 20k-30k in your post ).

[ethanb]

There was a series of books on this in landscape format

http://senseis.xmp.net/?WholeBoardThinkingInJoseki

Thanks also John F for providing another book review - it's very useful to have an impartial review of these from such a respectable source (they're next on my shopping list).

Those two books are GREAT. It breaks down move by move and presents the same local situation multiple times as different problems in different fuseki. And when applicable, multiple answers are marked correct, as it may be a stylistic difference rather than an absolute "this is wrong here" case.

[daal]

I find it difficult to outline the fundamental principles of joseki for myself, let alone for a beginner, first of all, because there are different types of josekis. There are "peaceful" josekis, in which the corner is politely divvied up, and monster josekis which are nothing less than full-blown fights. In these, a good result is often determined by the nebulous calculation and far reaching ability to judge the relative value of territory vs. influence. What guidelines could help me do this?

A complicated joseki is hardly different than any other fight, and as such, applying principles is a matter of great skill. In his book "The Workshop Lectures, Vol 1" Yilun Yang demonstrates in one chapter how fundamental principles can guide you through a complicated joseki. One key distinction he uses to help identify the weakest stones is whether a stone is "inside." An inside stone's first priority is to get out. Invariably, cutting points appear, and stones get cut off. Then one must identify the important stones and make sure they are protected.

It does not take long before the basic principle becomes: read. Yang does however offer guidelines as to what you should be reading in some typical situations that arise during complicated josekis. For example, in a crosscut fight you should first read whether an opponent's stone can be captured. If not, you should read whether you can save both your stones, and if so, you should help the weaker (often inside) one first by increasing its liberties. If you can't you should save your most important stone.

In any case, this is clearly beyond the scope of a beginner, and advice such as "stay connected" or "try to split your opponent" will quickly lead to situations that are far too multi-faceted to easily get a handle on. As with many other situations in go, one false step can lead to disaster anyway. Joseki is not much different from the rest of the game - you've got your principles, you've got your reading, and you've got your intangibles. While beginners surely play ridiculous moves, I think the best approach is to ask them why they played them, and then to see why the result did or didn't turn out as they had hoped.

[Pingwin]

17 kyu here, and in one of my beginner books there was a part on basic joseki. I find them too hard to remember and it is adviced against to learn them by heart. But what I did get out of the ‘Joseki lesson’ was that I don’t need to play every situation as a cut-throat live or die fight. Depending on the stones surrounding the corner with the engagement I’ll now try to pick a ‘result’ and try to come up with a move that helps to achieve that. Something like ‘I wouldn’t mind a wall to the bottom to work with my stones already there’ or ‘I am a bit surrounded I’d like to take the corner and leave it at that’. Then I try to answer with a move that doesn’t self-atari and help me get what I want.

Often the opposing player denies this and I have to see what I can get out of the stones already played.

Especially the not beeing ‘default aggressive’ in every corner situation seems to help me avoid playing weird moves just to continue an attack and getting hammered as a result. (Now I get hammered 20 turns later, a significant improvement!)

[RobertJasiek]

John, when you suggest to pronounce <hengma> (and presumably that the Koreans pronounce it like that), is the <e> a German <ä>? I guess the three syllaby idea was caused by some of us Germans speaking haengma like a possible German pronunciation of aero, that is <ah-eh-ro>.

[John Fairbairn]

John, when you suggest to pronounce <hengma> (and presumably that the Koreans pronounce it like that), is the <e> a German <ä>? I guess the three syllaby idea was caused by some of us Germans speaking haengma like a possible German pronunciation of aero, that is <ah-eh-ro>.

Robert: The vowel ㅔ is conventionally transliterated as 'e'. Descriptions of how to say it in English range from the vowels in met, yea, hey, bay (in the latter cases minus the diphthong, of course). I'm not familiar with the Berlin accent, but I'd expect your 'ä' to sound about right for that.

That, however, is not strictly the vowel in haengma, which is ㅐ . This is transliterated 'ae' but is compared with English yam, cap, or as between bet and bat, or as a slightly more open-mouthed version of ㅔ. There are major problems here with which version of English is meant.

However, in modern Korean the distinction between the two has almost disappeared. In very careful diction some speakers will still maintain the difference in initial syllables, as in haeng-ma, but even Koreans find it hard to distinguish one from t'other without a context.

So you can safely assume that haengma can be (and is) said with the first vowel ('e') which is much easier to get a handle on, which ever dialect of English you use, and probably much easier for Germans with 'ä' available.

In my judgement, telling English people to say hengma would be more likely to lead to a more palatable sound to a Korean rather than hayngma, if for no other reason than that too many English speakers would be prone to lengthen the latter too much.