What keeps pros awake at night ? To most amateurs, pros are almost superhuman, and they cannot imagine them worrying about anything. But there are, in fact, some things pros care about very much, things they study carefully. The most important ones are shape, style and reading ability, respectively known as katachi, suji and tsume-go in the jargon of the go world.

How does this apply to improving your own go ? It's very simple-replay the first hundred or so moves of top pro games several times. The first time, just play according to the diagram. Then play the game without looking at the diagram, if possible.

The third time you play the game, start actively thinking about reasons and alternatives for every move. The reason for doing it this way is simple. If you think about the game from the very first time you play it through, you will use your own knowledge as a reference point for understanding the game.

But by playing without actively thinking about the game, you start to see the moves as part of the flow of the game. You unconsciously change your point of reference.

Tobi, kosumi, keima

Today's study material comprises a game between rising star Jiro Akiyama, 7-dan (Black), and veteran Kimio Yamada, 8-dan.

Diagram 1 (1-18) :

White 12 is the first move that deserves attention. The normal approach would be at 17 or one point higher. However, 10 is low, and therefore the lower side is relatively unimportant. Black 15/white 16 is a normal exchange. Again, white 16 is on the third line. In other words, both the upper and lower sides are uninteresting in terms of immediate development. That leaves the right side, so White plays 12, 14, and later 18.

Diagram 2 :
Let's look at how to become stronger: be ambitious, develop quickly and keep/take the initiative. But don't overextend and don't leave behind weaknesses.

Black 19 does all of these things by playing a tobi (one space jump) to expand his moyo (potential territory) on the left side. And here is the bonus-while doing so, he also threatens to invade the upper left corner.

Seen in that light, white 20 seems to be a natural move, as it defends the corner. But it doesn't do just that-it threatens to invade at A, threatening next to connect underneath at 29, jump out to B or attach at C. Black can't defend against all of these threats at the same time, so a white invasion at A is a very real threat.

As mentioned before, the lower side is uninteresting, therefore black 21 is an excellent move. It expands his moyo on the left, while making White respond at 22. This is necessary, because a black move at D next would be intolerable.

Black 23 is another excellent move-a move at E might seem a natural extension, but it lacks ambition. It would give White the opportunity to extend to F, and that is not a gainful exchange for Black. But there is more to black 23-it has attacking potential.

White 24 expands White's local position while aiming at an invasion on the upper side. Black 25 is a kikashi-it forces White to add a stone (26) to a position that is strong already, and it strengthens Black's own position. Black 27 is the follow-up to black 23. See References 1 and 2. White 28 and 30 ward off the threat posed by black 27. Black 31 expands his moyo again, while threatening to play 33 to connect his territories on the upper and left sides. Combining moyos is a powerful strategy. White 32 is an invasion and it prepares for an attack in the upper right corner and on black 23. Black 33 is easy to understand, but white 34 is sente (Black needs to answer at 35). White then attacks the weakness between 31 and 33.

Reference 1 :
Black 1 is severe. If White follows this sequence, his whole group will be in danger. Note that the marked black stones are well positioned to keep White at bay.

Reference 2 :

White 2 is an attempt to resist. If Black follows through with the moves to 13, again, his marked stones are in good spots. Even though White's stones will stay alive, Black's thickness is superior, and White gets the short end of the deal. However, there is no telling whether Black will indeed play at 3, because at that time he can also choose to play at 7, 10 or 4, to destroy White's position from the inside.

Problem 1: White 28 and 30 fend off an invasion in Diagram 2. Is a black clamp above white 20 not possible? How does white 30 fit in the picture ?

Problem 2 :
If it works, white 32 may be good, but what would the usual invasion point be in this shape ?

Problem 3 :
Carefully consider the moyos of both sides and try to find a move that expands White's moyo while reducing Black's.

Problem 4 :
After answering Problem 3, you might come up with a better move for black 33. The keima, or knight move, between 31 and 33 does not make a very good shape. How can this be improved ?

Solutions to last week's problems

Solution 1 :
Pushing with white 1 is correct. Black needs to answer with 2, allowing White to connect his stones with 3. Black's four stones are now in dire straits. He might avoid a quick death by playing at A, followed by white B, black C, or alternatively, by starting at C. Either way, though, Black will suffer.

Reference 3 :
White 1 is wrong, as Black just escapes with 2. Note-pushing through with A won't help White. Black plays B, followed by white C, black D, white E, black F, and Black is connected.

Solution 2 :
Black 1 is correct. What-ever White does, Black keeps a strong position that splits White's groups.

Reference 4 :
Black 1 looks like the vital point, but it leaves behind a potentially deadly weakness. White will first strengthen himself on the left, in order to cut off four black stones from black 1.

Solution 3 :
Leaning on the marked white stone is a good strategy for Black. White needs to reinforce with 2, allowing Black to capture with 3.

Reference 5 :
Attempts to resist black 1 are futile. This is just one example of what can happen: after black 17, White's group is captured.

Solution 4 :
Black 1 and 3 are correct responses to the white intruder. If White plays at 4, Black next leans on the white stone with 5 and 7. White needs to escape with 8, allowing Black to cut at 9, and ravage white's position on the left.

Reference 6 :
Black 1 and 3 are terrible. In effect, this allows White to drive through Black's position, and cut off the marked stone.

Reference 7 :
The sequence black 1 through 7 is sometimes a clever combination, as it avoids getting split as in the previous examples, but allowing White to capture a stone is usually a bad idea, as it gives White too much thickness.