Thinking About Go

British Go Journal No. 10. December 1969. Page 8.

Part 1

Andrew Daly

Introduction and General Thoughts

Go is a game of thought, and thinking about it is almost invariably productive as well as interesting. Part of the charm of the game, for me at any rate, is in the theorising that is possible, and the variety of possible approaches, each with its particular attractions. In this series of articles I intend to present some of my own pet theories, the aim being not so much to expound authoritatively, but more to suggest provocative ideas which may be thought about, talked about, tried out in games, modified, rejected, or even accepted as being sound. The articles will be fairly elementary - players of about 15 kyu should not get lost - but because my aim is more to be original than to be correct (though naturally I aim at that as well) players much stronger than that should at least find something to disagree with.

Most thinking about Go is about particular aspects of the game, or about particular situations, but there are certain ideas which are absolutely fundamental to every stage of every game: of these ideas the most important is that of sente. A player has sente when he is not forced to answer his opponents last move. This means he can direct the play to whichever part of the board he judges most advantageous. Should he contrive to make forcing (sente) moves in succession, he can keep this initiative, to great advantage, as he has control of the direction the game will take, as well as the considerable local advantages derived from playing first in each situation. Forcing moves can often be achieved by threatening the opponents weaknesses - but beware! - there may be several ways to attack each weakness, or it may be better to leave the weakness to wreak its toll later in the game. It is unlikely that such a weakness would be directly defended, as this would cost sente, so leaving them for later is often a good idea.

One can regard every move as having some locally forcing effect - that is, a failure to answer it directly will cause some local loss. Even for gote (the opposite of sente) moves - i.e. those not answered directly because the local loss is considered to be smaller than the gain possible elsewhere on the board - this potential local gain will remain for sometime in the future. One should therefore try to make this locally forcing effect as great as possible, even when a sente move is impossible, to leave gains for you to make at the appropriate time later on.

The opponents weaknesses should therefore be exploited at the right time to make maximum profit from them. A fairly easy way of making such profit is by attacking a weak group in the early stages of the game. A weak group is one that cannot be captured immediately, but has not yet either two eyes or connection to a safe group. Such a group must make eyes or run away if it is attacked, and the attacker can nearly always make profit in either territory or influence from doing so. Weak groups are fundamental to opening and middle game strategy: Dia 1 shows how much emphasis professionals place on weak groups, for the recommended move is one that, though small in itself, makes an opposing group very weak.

Diagram 1

Black 1 is incorrect, although it is the largest move in terms of territory, as white 2 seizes the vital point. Black should have played at A, even permitting white B, a very large move.

For maximum effect such attacks should be made with tact, and sometimes with restraint. The group that is attacked may well survive to become strong under a direct attack, but If attacked indirectly remain weak long into the middle game as a heavy burden on all your opponents schemes.

Two weak groups simultaneously under attack make a golden opportunity, never to be missed. One of them can frequently be captured and, for moderately good players, the game is over, as the usual compensating gains made by the loser of a group have evaporated in the saving of the other group.

Another fundamental concept is the balance between present and future gains. A moderate player will usually prefer solid territory to the outside strength, or thickness, that brings gains in the future. With increasing tactical skill, however, stones further away from the immediate skirmish begin to have a greater effect on play, and the player with nearby thickness can turn the fight to his advantage. This must not be overdone, of course; a professional has said that Go is the art of balance. There Is a strong connection between this paragraph and the preceding ones, for one of the greatest gains from thickness is freedom from weak groups - all the groups can connect easily to the outside.

Any successful strategy must, however, depend on good tactics. Unless stones are used economically but effectively the most superficially subtle strategy must fail from sheer want of stones. Handicaps in effect compensate for the weaker players extravagant use of stones by giving him more to use - the stronger player can achieve the same effect with fewer stones. The art of thrifty stone placement is called katachi, meaning form or shape. The secrets of form are not to be learnt easily, for it is said that one can achieve 3-dan with this knowledge alone. The only certain way is by long and bitter experience, for, although master games are a fruitful source of ideas, these ideas are understood only after trying them out in games for oneself.

One of the commonest sources of error among moderate players is unnecessarily over-defining positions, i.e. playing on in one part of the board until nearly every point of territory has been decided. There are considerable profits to be made, and deep subtleties to be explored, by judicious tenuki (playing elsewhere) while the result of a local battle is still not quite decided. This makes matters difficult for your opponent to calculate, for even if he plays another stone it may not finish the situation completely, and meanwhile you have made definite and certain gains elsewhere on the board. Playing in this way has an unusual advantage, in that the skill of both players will improve, as they are both forced to consider more difficult positions. This is obviously another procedure that should not be overdone, and before playing tenuki you must have a very clear idea of what your opponent could gain in the situation you are about to leave.

This last paragraph is another example of an idea that has appeared twice before in this article, and indeed is fundamental to all good Go. This is the method of leaving possibilities of gains for future exploitation. These possibilities are called aji. Leaving and exploiting aji effectively is the mark of the mature Go player - Dias 2 and 3 are two elementary examples, but this is a quite difficult matter, and is best learnt from experience. It will not, however, be learnt by a player who is not on the look-out for opportunities, and who plays each position to its death before starting on the next. So be alive to your possibilities!

Diagram 2

Black 1 in Dia 2 is a typical short-sighted play, made thinking it strengthens his territory. In fact, the strengthening is marginal, and could have been done at any time, whereas White has significantly reinforced his lower group. Black should have played immediately at 3.

Diagram 3a

Diagram 3b

Black needs a play in the corner to be safe, but black 1 in Dia 3a loses the possibility of playing at A in diagram 3b, which may be very useful later. So Black should play as in that diagram, leaving the one point gain for later.


This article is from the British Go Journal Issue 10
which is one of a series of back issues now available on the web.

Last updated Thu May 04 2017. If you have any comments, please email the webmaster on web-master AT britgo DOT org.