Go Strategy
Go Strategy(苑田勇一流基本戦略 ISBN4-14-016103-4) by Sonoda Yuichi, NHK (2001), 222 pages.
This book is a refreshing and somewhat unconventional treatment of go strategy. Its organizing skeleton is a collection of strategic precepts and concepts to which the lecture-sized sections are devoted. As a teaching vehicle, each section focuses on some whole-board position for which the principal issue is identified, and for which a proposed line of play and variations are analyzed in light of the featured precept as well as those introduced earlier. The author is always careful to show how to handle tactical complications which could arise from following his advice. (He also shows what can happen if you don't follow his advice.)
That much is conventional. What is unconventional is that Sonoda has distilled his own formulation of basic precepts, some with unfamiliar points of emphasis and some sounding downright paradoxical. Of course the goal is to give the reader a fresh perspective on the game. Sonoda remarks that occasionally his recommended lines of play are rather different from conventional wisdom; but nonetheless they should feel natural; and at any rate, "there are many paths to the top of the mountain."
His first principle is
- playing near live stones is small, playing near not (yet) live stones is large.
He regards this as the simplest concrete formulation of efficient play, and the single most important thing that the reader will learn from the book. But wait! How near? And what is live -- is it only rock solid with two eyes, or does it include a single isolated 4-4 stone not yet under attack? Or what else? Well, that depends on the position. This precept is invoked in just about every section of the book. It is the author's Prime Directive and it applies, in his hands, to situations where you might not suspect it is a relevant issue.
Here are some more of Sonoda's precepts:
- Don't make three eyes. (So cause your opponent to do so by forcing capture of a sacrifice stone.)
- Solidifying live stones is OK. (Refers to opponent's stones, as in a leaning attack.)
- Attack means allowing escape.
- Enclosing territory makes it smaller.
- Forcing your opponent to enclose territory increases your profit.
- Don't attack, don't defend.
This last one sounds especially strange, but Bob McGuigan suggests (in the forum) that, in light of diagrams displayed below, it could be understood as "don't just attack, don't just defend." Along with this precept, Sonoda addresses the point that sometimes attacks are too severe or too prolonged or from the wrong direction or aimed at the wrong target or without any profitable objective; and that too often, threatened territory is defended early in the game when it should be traded for something else. This is all elaborated in a seventy page chapter, with sections devoted to specific sub-precepts.
In another chapter he devotes sections to sabaki, seriai (pushing and contact battles between groups of stones), and the relative strengths of various third and fourth line extensions.
One chapter is a "Crash course on 3-3 point and the corner" -- he feels that amateurs are surprisingly unfamiliar with variations from standard invasions under the 4-4 point, and associated life and death issues.
The final chapter analyzes a small collection of 4-stone handicap positions.
This is an exciting book -- it has taught me altogether new ways to think about positions, and his recommendations all feel natural, as he promises. After only one reading it has already helped my game, and I even have fun reinterpreting analyses in other books via Sonoda's framework. But learning how and when to apply the precepts is a long-term process, and I suspect that I'll be rereading it when I'm well past shodan (if I ever get that far).
Sonoda says that his book is aimed at a wide range of players, whether or not at dan level. It is certainly not a first introduction to strategic issues. Speaking as a 3-kyu (AGA) I would guess that while Go Strategy can surely be read with appreciation by single-digit kyu players, even post-dan players would be happy to learn to play as effectively as Sonoda illustrates in his examples.
Here are some sample positions from the book, with summaries of the analysis.
Don't attack, don't defend (four diagrams)
In this common 4-stone handicap scenario, white has just played the marked stone. Sonoda dismisses black responses at either a (played out of "fear") or b because white easily gets excellent results, as he shows. But he doesn't like the frequently recommended pincer response at c either!
For instance, the displayed sequence might occur, followed by Ba and Wb. But black has been forced into an uncomfortably low position and so, according to Sonoda, the severe pincer of
has backfired and he cannot recommend it.
The looser two-space pincer shown here ultimately imperils the pincer stone and so cannot be recommended either.
Sonoda's preferred response is the calm, three space pincer shown here, right in the center between white and black. He devotes many pages to variations and continuations showing why. But basically,
is far enough from white so that black can fight hard on the upper right without having to watch his back. White does have room to extend to a ("attack means allowing escape"), but the extension is cramped and not so profitable. If white does turn around and invade at b, say, then
is close enough to the bottom to stay connected, and the lower side territory obtained by white is too small to compare with black's resulting thickness.
Don't enclose (two diagrams)
If black starts to enclose her lower right territory with , the displayed sequence could follow. But white has also considerably benefited from the total exchange. Black must not play this way.
Instead, black should lean into white's moyo with . If white is so timid as to respond with
at
then black immediately plays
at a, threatening white's lower left; and assuming white defends, black retains sente to play at
, ultimately getting much more profit than in the previous diagram.
But white must not play this way either! Instead, white should respond to with a counter reduction leading, perhaps, to the displayed sequence. Both sides have been reduced, but "absolutely" the second diagram is better for black than the first. In addition, black has serious invasion aji at b, as detailed by Sonoda in twelve diagrams.
The sample diagrams "Entering a large moyo" in the review of Ishikura's How to Break Out of Beginning Kyu Levels (Immediate Results) also illustrate the "don't enclose" principle.
Playing near live stones is small (two diagrams)
This is the first whole-board position considered in the book. It has just been established that white's upper right corner group is unconditionally alive. Black to move. and a are both considered, but
is judged better because it is farther from white's live corner group.
Following the joseki in the lower left, (and the extension to b following
) are adjudged as in the right direction because black's upper right group is thin, owing to the cutting points, and could use some support: playing near non-live stones is big.
These moves could be justified without Sonoda's particular formulation, of course; but he wants to get the reader to start thinking in these terms.
Following , black plays the crosscut of
and
("skillful play"), and then plays to sacrifice
and
. He intends to force white to play near the live group.
The battle runs for another 25 moves, with additional sacrifices by black, and is far too long to reproduce here. But in the end black has a nice wall stretching along the line containing and
, taking sente to play at a, while above the wall white has a huge, highly compressed group filling much of the corner and stretching to within one space from the right corner group. Sonoda says he likes to describe white's over-concentration as "making three eyes."
(Sonoda does not discuss what white mistake contributed to such a dismal outcome; but one might presume that was too near the live white stones and should have been on the side instead of the top.)
Lynx Indeed, the was bad and the attachment on top and cut is the standard tesuji to force W into overconcentrated shape. There are some very scary variants, however, where W resists, and if W is strong enough on the outside, the cut may be overplay.
Don't just enclose (three diagrams)
Sonoda feels that in this common reduction sequence, by responding to the cap of with
, black has been led badly astray by the knight's response proverb.
He recommends not responding directly to , but closing the corner instead, because the corner is bigger, and starting at the bottom instead of the top because white's lower position is thin ("non-live stones") and a move there is more of a threat. If next
, then black can also enclose the top, and with this sequence, black has played much more efficiently than white.
If black prefers speedier play then in place of the previous the attachment
, "solidifying live stones", is also fine. Taking sente to invade the bottom, black can later anticipate Ba, Wb, Bc. The two marked white stones are low priority.
-- FredK
See also