Learning something new can be a straightforward ...
By Rob Van Zeijst
Learning something new can be a straightforward process if simple, or ordinary, knowledge is involved. But complex systems require another approach, often more than one. Learning by rote is part of the process, but it is not enough. Take languages, for example. It is virtually impossible to learn every sentence and combination of words, so one should learn basic syntax, some vocabulary and a few rules of thumb, and start experimenting with real-life partners. Go is a game that has far too many possibilities for even the most advanced computers to handle. To learn go, you need to know the basic rules, attain the ability to read out possibilities and sequences, and use some rules of thumb.
General rule for extensions
For extensions along the third line, the optimal extension from a lone stone is a regular two-space jump.
The rule of thumb for extensions is to take the length (the number of stones) of the wall and add one for the optimal extension. In the case of a lone stone, this means the ideal extension is 1+1=an extension of two spaces. For a wall that is two stones long, the best extension is 2+1=a three-space extension.
Diagram 1: This is the situation when this rule is followed. An extension to A is too inefficient while an extension to B is too far because it is loose and weak.
Diagram 2: In the case of the two connected marked stones, Black's ideal extension is a three-space jump. White will find it very difficult to invade at A.
Diagram 3: Should White invade at 2, Black can easily bottle up the invader with 3. After that, the sequence through 7 seems likely. As it is, without outside assistance, White cannot escape. Instead of 2, White can try an attachment at A, but the result will be roughly the same. With a partner, try some different combinations in this position.
Diagram 4: For a wall of three stones a four-space extension is optimal. If White invades at 2, Black 3 will usually be the right response. White 4 looks like a clever move, but black 5 through 11 will contain the white invaders. This is also a good position for some sparring with a partner.
Solution to last week's problem
Reference 1: Before we discuss the solution, let's first determine why Black did not make a second shimari (corner enclosure) by playing at 1. If he had, White would extend to 2 and pressure would be lifted on White's marked stone. Since White's stones are settled now, White can then easily invade the top, e.g., at A, followed by black B, white C, black D, white E, black F and white G. The strength of Black's marked stones now serves little purpose.
Solution: In the game, Black played 1 and White extended to 2. The reason for this three-space extension is that White has forcing moves against Black's marked stone, such as 4 through 10. So in fact, White has a wall here and is not afraid of a black invasion at A. The game proceeded as shown and the result was roughly even.
Reference 2: If White plays the ordinary two-space extension, black 2 is a good move, defending the corner and making points while putting pressure on White's stones. If White does not answer right away, Black will probably play at A, followed by white B, black C, white D and black E. This puts even more pressure on White and builds up huge center thickness. This is bad for White.
Problems
Problem 1: This black position is quite usual. At this point, White often invades on the right, but where? Is A better or B? Consider some variations.
Problem 2: What about this position with opposing shimari? Should Black next extend to A, B, C, D, or as far as E or even farther? Consider the advantages and disadvantages of all these possibilities.
Van Zeijst is a four-time European go champion and European representative at the Fujitsu World Championship.