The two replies above show we still have a long way to get the Japanese concept of thickness over properly in the west. A bit depressing after so many years. It’s also disappointing that so few people tried to help the OP. I suppose that may be in part a consequence of the increasing redditisation of go talk and the waning of interest in the game for some people in the wake of AlphaGo.
Even if you accept Robert’s characterisation (and I don’t) it’s only the equivalent of describing a picture in terms of “it’s a collection of colour blobs+ lines, makes use of perspective, has a theme, etc” when the inquirer really wants to know how to paint.
Let me quote a recent book by a professional and then do an explication de texte to illustrate some of the missing or misinterpreted points above.
I’ve briefly reviewed this bok here already: “How many stars is this thickness worth?” by Yokota Shigeaki. Here’s part of the preface.
Atsumi means “thick aji” [1]. This is the topic of this book.
There is a notion that a strong group is what is meant by atsumi, but a weak group can also have atsumi, and this can change according to circumstances [2]. In addition, if even strong atsumi is not used skilfully, the end result will be unsatisfactory.
Strong atsumi exhiibts its effectiveness the earlier it is created in the opening. Moves that make thickness stronger can even be bigger than extensions on the side. [3]
Amateurs are often perplexed about how to use string thickness skilfully, but if the thickness is yours it means you can attack the opponent’s groups severely and build an advantageous position smoothly by punishing him if he has omitted even a single move. [4]
Let us unpick this and contrast it with what has been offered so far.
[1] Yokota makes the important distinction between atsumi and atsusa. The two are mixed up in RJ’s definition. (Atsumi is written 厚味 where the character 味 is read in combinations as -mi, and meaning -ness, but the word could be expanded into two words 厚い and 味, with 味 read aji when it stands alone.)
[2] Yokota makes the case for treating atsumi thickness as worth 1, 2, 3 or 4 stars. 1 and 2 are weak atsumi (meaning it is too early to use it as thickness) and 3 and 4 are strong (and can be used as thickness as is). In this scheme, even a single stone has thickness. The important point is that if you are choosing to play in a thick way, you have to decide which of your groups are going to have thickness and they have to start with a single stone. You may change your selection “according to circumstances” but at some point you still have to start the ball rolling.
[3] What then follows from [2] (and is a major part of the book) is reinforcing your thickness, which takes time (and planning). The earlier you do it, the better. So time is a very important element in two different ways. Time – which is the basis of the “how” - is missing from the “what” definition above.
[4] The key word here – and again missing from the definition above – is “attack” (which is also a dynamic, i.e. time-based, aspect). What is missing from the Yokota book, deliberately obviously, is the word “territory.”
The method behind the book, says Yokota, has two elements: (1) learning to evaluate the strength of the atsumi, i.e. giving t 1 to 4 stars; (2) reinforcing the atsumi and using it. He adds that the goal is to make correct use of the atsumi according to its degree of strength.
Not in the preface, and not in either of the posts above, but in the body of the book is a very frequent reference to a “base “ (not territory). Yokota’s four star system is based on the following guidelines:
1 star: A group that has neither height nor a base
2 stars: A group that has some height but no base
3 stars: A group that is well endowed with height OR a group that has both height and a base
4 stars: A group that is well endowed with both height and a base.
There is actually a little more to each definition that you are expected to pick up intuitively by working through the copious examples. For example, the three topside black stones below are classed as 1 star because White can peep at a and b and so Black is deemed not to have significant height even though, when he connects, he ends up with 5 stones (the urgent move here for Black, incidentally, is c, to make a base (nb again: base, not territory).
In contrast, in the next diagram Black still has just three stones at the top but these are classed as 2 stars, because there is no peephole. Black still needs reinforcement, of course, at either a or b.
As an example where there is no base but the group can still earn 3 stars (i.e. is strong), see the white group in the upper right in the next diagram.
Nevertheless, the ideal move for White next is a which constitutes a base of sorts – you have to do a bit of neural networking to adapt your definition of “base” as you experience more an more examples. But, still, Yokota is never talking about territory. In fact he wants Black to invade at b, so that White can attack at c. I remember being derided here some time back for insisting that “thickness is not thickness unless it can be used as thickness”. The examples by Yokota are the sorts of things I see in multiple Japanese books, and that’s why I still insist on that notion.
Let us look at another of his examples that will illustrate a common misperception.
The black thickness here is counted as 3 stars (well endowed with height). I think it’s fair to say that many amateurs would think along the lines the “stand on two extend three” rule and so be thinking that a wall of five stones means you extend six spaces, to a. I can even imagine quite a few thinking, “Ooo, that’s a bit far – maybe b is better must protect my territory.” Worse, they may be thinking that they now have some territory – they haven’t, they’ve just got a base. Which is fine in a way, but just turning a strong group into a stronger group without any interaction with the opponent smacks of inefficiency.
Yokota’s move for Black is A below:
If White invades at B, Black C is essentially forcing D, which means what we have ended up with is two Black walls of 5 + 2 = 7 stones, from which an 8-space extension can be regarded as ideal, as here. White B has thus come into the sider’s web (dynamic tewari can be applied here, of course).
This does NOT mean White is dead. In fact, Takemiya (just to show this is not Yokota on his own) says in such positions Black’s aim is not to kill White but to let him live with just two eyes. The result would be White gets territory (yay!!) but just two points, whereas Black gets the equivalent of a dozen or more free handicap stones on the outside (not to mention sente).
The “stand on 2, extend 3” rule is very often misunderstood because it is taken to mean that’s how far you can extend to make a safe territory. No, it means that’s how far you can extend and still ensure a connection, and that is the purpose of a base.
I think I’ve probably given enough info so that those interested can make use of Yokota’s book even if they don’t know Japanese (knowing characters like “right answer” though will help enormously). But for the OP a few other remarks may be in order.
“Playing thickly” is not an ideal way to express the idea. It makes it sound lke a single concept. Rather there are several ways to indulge in a thick way of playing. Yasui Senchi is very different from Takemiya (who rejects “thick” in favour of “natural” anyway), who are different from Hane Naoki and Takao Shinji, wh are different from each other but have played each so often and have both written books on thickness (i think at least one is available in English), so that their games between each other are an excellent resource. They are all different from Cho Chikun who sometimes claims to have a thick style, but he is referring to atsusa and not atsumi. But whatever the individual differences, when it is atsumi that is being used, the main characteristic of a thick style of play is building a platform from which to attack early, and so to maintain the initiative (i.e. if you lose the initiative you have probably made weak atsumi). Because you maintain the initiative you end up
late in the game making territory "by accident."
Finally, it is safest to banish the word “influence” (except in the phrase “sphere of influence” in a moyo context), and if you do follow one of the several (Japanese) methods for counting thickness (atsumi), do remember they are not counting territory, except in the notional sense that a big-point move in the opening is worth 15 points.
Thank you for your detailed exegesis!
As I said, thick is a subtle term. When I became shodan I was amazed at how little I understood the game. (I am no longer amazed at that.