Proper sabaki example

[John Fairbairn]

For those of you who think sabaki means making "light and flexible" shape (other threads and SL refer), I invite you to tell us what is "light and flexible" about the following example. The game, broadcast last week, was headed "Splendid sabaki by Shida" and the commentary was by Rin Kanketsu.

The sabaki sequence was described as beginning with moves 75 and 77, Black "worming into the corner to achieve sabaki". The sequence ended on move 119 with a trade which was adjudged favourable to Black, giving him a territorial advantage. On the way, Black 89 was praised as a good move.



Any move can be described as flexible, of course. But where is the lightness, the gracefulness of the ballerina, the funwari-ness of the fairy cake?

Black went on to win, having "coped" in a difficult situation.

[CDavis7M]

Thanks for sharing. This situation feels special and surely brings new meaning to sabaki. But I don't see it as a counterpoint to "sensei's" understanding. It seems light and flexible to me. Flexible in that Black had options they were willing to accept and light in that Black did not play heavily but setup kos which were able to be sacrificed.

[John Fairbairn]

light in that Black did not play heavily but setup kos which were able to be sacrificed.

That's a reasonable point, but what I see as the problem is that the usual definition, the one I object to, is "light, flexible shape", which is why I referenced the other threads.

I don't see how you would end up with an outcome such as in this game if you set out to make sabaki by making light shapes.

The other problem with the usual English version is that it adds that whatever is done is to "save a group". That is mostly taken by readers to mean shinogi, which means giving a group definite two-eyed life. Sabaki does not go that far - if it is local it refers to temporarily settling an endangered group (or, often, grouplet), which may still lack two eyes. But as the example here shows, even that sense of locale does not have to apply. You can "cope" in other ways, by e.g. making a trade.

Using the verb "cope" leads you in the right direction. Using the noun "shape" does not. Lightness is a red herring.

[CDavis7M]

Using the verb "cope" leads you in the right direction. Using the noun "shape" does not. Lightness is a red herring.

There definitely seems to be misconception (over-generalization?) on shape and use of sabaki for situations that are more like shinogi. But thinking more about this position -- is this really sabaki (coping)?

Looking at move 75 and 77 and the result in the corner it sure seems like an attempt at sabaki resulting in a trade. But going back to move 59 it seems like something different. The entire sequence on the left seems less like sabaki and more like Black realizing that there is enough room in the corner to cause trouble. Maybe Shida hoped that White would defend and he could take over the left side? But when White played strongly Shida started laying out the picnic blankets amidst White's cherry blossoms. Less like sabaki, more like Hanami shiito o motsu. Could even be uglier with blue tarps that I see pictures of.
Did I butcher Japanese even worse than Sensei's?

Hanami.png (138.79 KiB) Viewed 19680 times

[jeromie]

The book Strategic Concepts of Go, by Nagahara, has a definition of sabaki that fits right in with what you're saying. It stresses that sabaki means "development," and in go the term means "the development of stones in a dangerous situation in a kind of quick, light, and flexible way." (There are those light and flexible terms!) But it also stresses that "...sabaki does not refer to shape; rather, it is a way of playing. However, the idea of shape is often related to the method of sabaki. Sabaki is often used to avoid omoi katachi (heavy shape)."

I can't speak to the accuracy of the Japanese translations, but I think this captures both the active nature of sabaki and its association with shape.

The crux of your argument here seems to be that the Japanese loan words that have made it into English go parlance no longer carry their original subtlety. I don't think that's surprising, though it is certainly important to recognize if we are continuing to import material from Japanese professionals and authors.

[John Fairbairn]

While toying with the idea of putting out a small book on Shuwa (to celebrate the fact that the GoGoD database now (though not yet issued) has the complete games of Shuwa - 600), I came across another example of "proper" sabaki. It occurred to me that this example might also explain why the western misconception arises.

In the position below, where White (Shuwa), in obvious difficulties at the top, has just played the triangled stone, the commentary is headed "An idea for coping" (sabaki no suji). This move could be described as "light and flexible", though the meaning would also be rather frothy, too! In loose English we might describe this as a "sabaki move". But that would be highly ambiguous. Maybe it's that ambiguity that has led to problems. The important point is that sabaki/coping has NOT happened yet. The triangled move is simply seeking sabaki/coping. It is only the seeking move that is light and flexible (after all, when in trouble you don't throw good money after bad). The final SHAPE may be light but is more usually stodgy. But even then it's not a fairy-cake lightness. It's more a Yorkshire pudding compared to a suet dumpling.



The coping process begins with this triangled move and is complete only with the position below.



Again, you would be hard pushed to describe this suet dumpling as "light and flexible". But White has COPED. The heading for the commentary on this portion of the game is "Coped plus...! (sabaita ue). The plus refers to the fact that, in addition to having coped, White has come away with sente because Black is now forced to play A.

The full GoGoD game is 1842-12-18e (vs Hayashi Hakuei) if you want to look it up. At this stage, it's not likely to appear in the book.

[dany]

White has come away with sente because Black is now forced to play A.

imho more like a gote L12

[dhu163]

irresistable.

I think 89 is surprising to play what looks like a heavy stone when your opponent can cut in the corner, but it is using the bonus power you have in the corner if you control the 1st line for a ko. In this case, blocking off W's center of the 3 is more of a vital point.

I agree with pretty much everything here. Sabaki starts off light, often allowing perhaps 3 different options for continuation, but when your opponent takes one, if you don't tenuki then you need to choose another rather than fighting for the same space, leading to "heavy committal shape" that leans on opponent, trying to claim local territory. I think the upper left corner here is relatively light for B as even if W kills everything, it isn't too much bigger than what W already had.

In the given example, W tried to kill everything rather than giving any way locally, so B had to fight. It was more a matter of committing to local play and choosing the right shapes.

I think Go (especially in the west) lacks in good definitions, perhaps because it is more important to be able to apply them. But for the purpose of conversation, I try:

heavy shape: moves that solidify local territory at the cost of leaving forcing moves by the opponent.
heavy strategy: moves defending or reinforcing a weak group rather than sacrificing it.

Why is "heavy" used? Because sente/gote around valuable weak points means the position will attract moves by both sides, but especially the opponent. It is risky and very costly if it doesn't work, and falling heavy objects cause damage.

This is similar to but slightly opposed to shape which is just gote and doesn't imply more moves will be played locally.

edit: I can't say I've ever been sure what sabaki means beyond a label.

[CDavis7M]

While toying with the idea of putting out a small book on Shuwa (to celebrate the fact that the GoGoD database now (though not yet issued) has the complete games of Shuwa - 600).

I would buy it!

The full GoGoD game is 1842-12-18e (vs Hayashi Hakuei) if you want to look it up. At this stage, it's not likely to appear in the book.

No worries. This game is already in my book on Shuwa I've been typesetting an A4 (folded to A5) booklet of Shuwa's Castle games just for fun, and just for my own studies.

I don't have any commentary on this game to tell me that it was sabaki and I definitely did not recognize this position as initiating sabaki. Perhaps because I have misconceptions, or I'm just not familiar enough with the technique. But I think it's because White was the instigator rather then merely defending (which may not actually matter for "sabaki" as it's used in Japanese).

Shuwa1.PNG (92.22 KiB) Viewed 19361 times

Putting aside the notion of light and flexible, I think that this position wasn't feeling like sabaki to me by the time the marked stone was played because White started the fight with 56 and because White had already dropped so many stones in the area by 70. Playing 56 when Black already had the extension/pincer seemed more like starting a fight than "coping." But I guess the end result and techniques are the same regardless of how the players got there (invading the opponent's position vs neglecting your own group).

Shuwa2.PNG (113.54 KiB) Viewed 19361 times

Anyway, I do think that this example hammers the point home much better than the previous example - sabaki is "coping" whether the play is heavy or light.

--------

By the way, I've been putting small biographies of Shuwa's opponents into my booklet based on SL and the GoGoD Onomasticon and I have been having trouble figuring out the players ranks in the Castle Games, like for Hakuei and Sanchi IX and so on. Sometimes GoGod has a previous game with a rank. But is it remotely possible to guess a player's rank based on their games with players having known ranks? Of course it might be hard to know whether they played on tagaisen or senaisen.

[CDavis7M]

I think Go (especially in the west) lacks in good definitions, perhaps because it is more important to be able to apply them. But for the purpose of conversation, I try:

heavy shape: moves that solidify local territory at the cost of leaving forcing moves by the opponent
heavy strategy: moves defending or reinforcing a weak group rather than sacrificing it.

Why is "heavy" used? Because sente/gote around valuable weak points means the position will attract moves by both sides, but especially the opponent. It is risky and very costly if it doesn't work, and falling heavy objects cause damage.

I always thought of light and heavy in terms of their burden on the player "carrying" the stones. But it's related to what you describe. A series of moves that solidify local territory at the cost of leaving forcing moves would be a heavy burden for the player to carry until the group has "coped." Whereas a move played with the intention to possibly sacrifice it has little burden on the player because they intended sacrifice as an option from the start. The player has light feelings over the death of their light stone.

[CDavis7M]

By the way, looking at this game again I remember what interested me the most was actually Hakuei's play in the lower-right. He plays 125 maybe threatening to connect to his dead group in the center, or threatening to make trouble on the lower side. And then he does make trouble and actually succeeds in taking away most of the lower right corner right before he decides it's still not enough, and he resigns. So was 107 (right of 137) on the upper side a mistake? White 108 and 110 seem to already make it difficult for Black. Maybe Black should have attempted a connect towards the left first. Or does the turning point in the game go back to Shuwa's sabaki in the upper right?

Shuwa3.PNG (54.08 KiB) Viewed 19365 times

----------

And if I can just keep talking about Shuwa. There's a game where Shuwa plays Black against Yasui Sanchi IX (White) that has this interesting sequence by Shuwa:

$$W Moves 80 to 89
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . 9 . . . . 5 6 . . . |
$$ | . O O . O . . 2 . . O . O . X 8 . . . |
$$ | . . X , . . . . 0 , . . . . 7 , X . . |
$$ | . . . X O O . X . 4 . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . 3 . . . . . . . . O X . . |
$$ | . . . . . . . . . . . . O . . . . . . |
$$ | . . X X X X . . . . . . . . O O X . . |
$$ | . X O O O O X . . , . . . X . , . . . |
$$ | . X X O . . 1 O . . . . . X O . O . . |
$$ | . . . . . X . . . . . . . X X O . . . |
$$ | . . O O . . O . X . . X . . . . . . . |
$$ | . O X O O O . . . . X . X . X O . . . |
$$ | . X X X X . . . X . O X O . X X X . . |
$$ | . . X , O . O . . , . . . X . , . O . |
$$ | . . . X O . . . X . . O O . O O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------

Click Here To Show Diagram Code

[go]$$W Moves 80 to 89
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . 9 . . . . 5 6 . . . |
$$ | . O O . O . . 2 . . O . O . X 8 . . . |
$$ | . . X , . . . . 0 , . . . . 7 , X . . |
$$ | . . . X O O . X . 4 . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . 3 . . . . . . . . O X . . |
$$ | . . . . . . . . . . . . O . . . . . . |
$$ | . . X X X X . . . . . . . . O O X . . |
$$ | . X O O O O X . . , . . . X . , . . . |
$$ | . X X O . . 1 O . . . . . X O . O . . |
$$ | . . . . . X . . . . . . . X X O . . . |
$$ | . . O O . . O . X . . X . . . . . . . |
$$ | . O X O O O . . . . X . X . X O . . . |
$$ | . X X X X . . . X . O X O . X X X . . |
$$ | . . X , O . O . . , . . . X . , . O . |
$$ | . . . X O . . . X . . O O . O O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

[bugcat]

In his video https://www.youtube.com/watch?v=fbMZVXaDU-0 Michael Redmond writes that

Sabaki is the settling of a weak group inside the opponent's sphere of influence by means of light and flexible play

I can't think of a better source than a 9d professional who is a fluent speaker of both English and Japanese, and a professional commentator in both languages.

If Michael thinks that sabaki involves "light and flexible play", I'm sure it does.

[John Fairbairn]

If Michael thinks that sabaki involves "light and flexible play", I'm sure it does.

Precisely. But also a bit beside the point, and it might be added that Otake adds rather more necessary gloss to this in his book on lightning go, and Otake also happens to be a rather higher level 9-dan.

The source of the long-running problem is that the Go Player's Almanac of 1992 gave the following definition for sabaki:

Making light, flexible SHAPE in order to save a group.

My majuscule emphasis. This definition has appeared in countless places since, and been believed. To our detriment. But at least the Almanac acknowledged its mistake and the 2001 edition changed the definition to the following:

Managing a weak group of stones so that it does not become a burden, e.g. by giving it a viable or flexible shape, or sacrificing part or all of it.

This is unobjectionable (and better than Redmond's shorthand), although actual examples may be necessary for some people. The 2001 Almanac also newly includes an entry for the pure verb sabaku:

To manage, defend, develop, or unload a weak group. To parry an attack.

I'd quibble a little bit with the last sentence (confusion with shinogi can arise), though if you already know what sabaki is, it can be acceptable.

A reminder that this whole thread stems from a discussion of my claim that western go overemphasises static SHAPE as opposed to dynamic PLAY - Redmond's word.

[dhu163]

If I understand the Chinese translation of sabaki correctly (probably not very well if so), it seems to mean that you threaten to live locally, but depending on the opponent's responses, you are likely going to sacrifice part of the group. The implication is that if you do nothing then your opponent might capture even more of your group with even less aji. By threatening to live, you at least create a solid structure that makes weaknesses in your opponent's surround.

[John Fairbairn]

If I understand the Chinese translation of sabaki

This is a tricky area. It may be that modern Chinese players adopted a Japanified meaning for a word that already existed in Chinese go. The recent Chinese-English Dictionary of Weiqi Terms (a rather poor book in my view, with no apparent involvement of English native speakers - and it shows) gives 'sabaki' for 腾挪. The definition could do with more rigour, but at least it emphasises a process or action ('tactic', 'a way') and makes no mention of shape. Adjectivally, it opts for 'swift and flexible'. But the real problem is the entry ends "Originally is a Japanese term." No it isn't. It appears lots of time in e.g. Bugu Bian, Yi Cheng and Yi Wu. That is, in commentaries well before 1700. If anything, I'd guess the Japanese borrowed it from Chinese. I have shown in various books that there are good grounds for believing Japanese masters from Sansa onwards were familiar with various Chinese go classics.

However, a complication is that the word 勉强 (manage to do [with difficulty/effort], do grudgingly or reluctantly) is also used in the old Chinese classics with a meaning that appears to overlap with the idea of coping. They may be regional or personal variations, though I suspect not. I think a more likely hypothesis is that 腾挪 was borrowed into Japanese as sabaki, but also took over its nuances of dodging around nimbly and so were predisposed to think of sabaki as something done lightly and flexibly (what the Japanese call karui sabaki), but they would also be aware of 勉强 and that would cover the heavier variety. Wielding a claymore as opposed to an epee.

But on the whole I think adding Chinese (or Korean) nuances makes the cocktail too rich for English speakers.