4-4 high approach joseki incomplete?

Question

I have a problem with the 4-4 high approach joseki - the left variant:

In the left variant, black ends the joseki and white can play anywhere else, i.e. has sente. But the position is pretty weak for white! Black can invade at 6-1, leaving white in a very bad position - it takes a lot of points from white and it seems like the two white stones at 4-3 and 5-3 will not form any kind of territory and are pretty much useless. So it seems that the Joseki shouldn't end here but fix the position by playing 4-2 for white, leaving the black in sente.

Or am I wrong? Why is this joseki ending in this weak position?

Answer 1

First of all, simply playing joseki does not mean the result will be even (or even good for you) on a given board.

A joseki is a sequence that is considered as often optimal given a particular situation. If the situation changes, a joseki which was fine before might suddenly become a terrible choice.

The situation depends on the whole board. In particular, ladder breakers have a huge impact, as does thickness (or the opposite, existing weak groups). Lots of other stuff can be considered in addition.

The high approach and the following moves are, to my knowledge, pretty much exclusive to only two situations:

1. As a ko threat
2. To enforce central and top side influence

If played as a ko threat, the situation is completely different, and white will certainly not continue locally in case black answers. Thus I'll talk about point 2 only.

This is played for instance when white already has a lot of central influence and feels compelled to expand it at any cost.

Comparison with the low approach

Let us suppose this situation.

$$W
$$ --------------
$$ |.............
$$ |.............
$$ |.............
$$ |...X.1...,...
$$ |.............
$$ |...2.........
$$ |.............

Approaching high instead of low mostly removes the possibility to invade at 3-3, because there is no good connection between 6-4 and 3-3 possible. If it were low, a connection would be easier to achieve in good shape: Contrast white 1 with white a.

$$W
$$ --------------
$$ |.............
$$ |.............
$$ |..3..a.......
$$ |...X.1...,...
$$ |.............
$$ |...2.........
$$ |.............

Typically, black can block at 2 and draw back to 4, which leaves white in an uncomfortable position that is actually not easy to live (depends on a ladder). Either way black will get great influence, which is contradictory to white's initial move.

$$W
$$ --------------
$$ |.............
$$ |.............
$$ |..12.........
$$ |.3.X.O...,...
$$ |.............
$$ |.4.X.........
$$ |.............

I'm not certain, but black might possibly consider simply drawing back, to continue with a or b to harass the white shape, which would not be possible if white had approached low.

$$W
$$ --------------
$$ |.............
$$ |.............
$$ |..13ab.......
$$ |..2X.O...,...
$$ |.............
$$ |...X.........
$$ |.............

So we see that the high approach has less nice continuations locally. Why is it ever played then? Because sometimes we don't really care about the corner! We are rather interested in the center and top side.

$$W
$$ --------------
$$ |.............
$$ |.............
$$ |.....1.2.....
$$ |...X....a,...
$$ |.............
$$ |.............
$$ |.............

Had we played low, black could easily pincer us with a move like 2 or somewhere around a. It would then be reasonable to enter the corner:

$$W
$$ --------------
$$ |.............
$$ |.....7.......
$$ |..35.1.2.....
$$ |..4X6....,...
$$ |......8......
$$ |.............
$$ |.............

Now black controls the outside and possibly the top side, while we "only" got some comfortable corner territory. We will see why this is not always good for black below.

$$W
$$ --------------
$$ |.............
$$ |.............
$$ |.......a.....
$$ |...X.1.b.,...
$$ |.............
$$ |.............
$$ |.............

It is possible to pincer around a or b instead of responding, but this pincer has much weaker continuations. White has a bunch of options how to continue, and it will be very difficult to refuse white center access.

$$W
$$ --------------
$$ |.............
$$ |.............
$$ |.....1.2.....
$$ |...X.4...,...
$$ |.............
$$ |.............
$$ |.............

With the low approach, it would be questionable to tenuki. Black can connect his stones and gain a quite strong outside shape once he covers white with 4.

There are a lot of additional variations of importance here, but I'll leave them to you do discover. I'll now try to answer the actual question :)

After joseki

$$W
$$ --------------
$$ |.............
$$ |..4..........
$$ |..213....5...
$$ |...X.O...,...
$$ |.............
$$ |...X.........
$$ |.............

In the above situation, white established a reasonable base for herself. As we found out earlier, black will usually follow a path similar to this, so white can expect this result with rather high probability.

$$B
$$ --------------
$$ |.....a.......
$$ |..X..........
$$ |..XOO..b.O...
$$ |...X.O...,...
$$ |.............
$$ |...X.........
$$ |.............

We discussed that white will only chose the high approach if she is fine with yielding local profit. We can assume white to have more influence in the other quarters of the board, possibly a large framework of stones on the fourth row.

You correctly recognized that black 4 is a very big move - if white were to play there, the corner territory white is very important in endgame would be white's. At the same time, 4 opens up continuations against white, which we will look at now.

$$B
$$ --------------
$$ |...5314......
$$ |..X..26......
$$ |..XOO....O...
$$ |...X.O...,...
$$ |.............
$$ |...X.........
$$ |.............

Let's start with the monkey jump. I would expect something similar to the above line. White is confined to a narrow space, and locally, it seems like white is too tight. However, white was also made quite strong - the white group will find life without problems (possibly playing a tiger mouth shape with 6 instead).

White being strong means black won't have a target to attack, which further implies black won't be able to start a running battle with a white group. A running battle might be a great way to destroy white's framework (which we implied earlier). Black might be a little sad to lose this opportunity.

In case everything else on the board is settled and black is winning, the monkey jump is likely a good idea. Just don't play it too early.

$$B
$$ --------------
$$ |.............
$$ |..X..........
$$ |..XOO..1.O.a.
$$ |...X.O...,...
$$ |.............
$$ |...X.........
$$ |.............

We could also invade at 1. This invasion may be difficult to handle for white. This means that black should consider this option instead of blindly monkey-jumping and forcing white to become very solid.

My knowledge of this particular shape is too thin to explain much about it. You may be interested in looking it up or exploring it in your games. Obviously, a stone around a will be greatly helpful for b. I would consider a checking extension to a as probably sente for black.

Note that there are other invasions to consider in addition.

$$B
$$ --------------
$$ |.............
$$ |..X.....e....
$$ |..XOO.c.dO.a.
$$ |...X.Ob..,...
$$ |.............
$$ |...X.........
$$ |.............

Black b or c are conceivable. With support at a, black can also think about moves like d or e.

If white is afraid of an invasion, she might simply extend less far:

$$W
$$ --------------
$$ |.............
$$ |..4..........
$$ |..213...5....
$$ |...X.O...,...
$$ |.............
$$ |...X.........
$$ |.............

With 5 being one space closer, it will be almost impossible to invade. However, white had to slow down her speed considerably, and obviously the monkey jump will leave white even more concentrated.

Result

We found that white will only play the initial joseki if she strives for certain influence. Locally, white is worse off. If white is not fine with that (i.e. the influence is not worth it), then white should simply not pick the high approach.

Your proposed move at 4-2 is large yose and also somewhat strengthens the white group. However, it would be contrary to white's original goal and thus should only be played later.

Variation

Note that it is also possible to play a high extension (this applies to quite a lot of josekis, actually).

$$W
$$ --------------
$$ |.............
$$ |..4..........
$$ |..213........
$$ |...X.O..5,...
$$ |.............
$$ |...X.........
$$ |.............

Once space closer and higher is fine as well. Obviously, this aims even more for the center. White is unlikely to gain much territory at all in this situation, but she won't care, because her whole attention is on the center.

Answer 2

The high approach is not for territory, but always a special approach due to a particular scenario of the whole board.

One purpose of high approach is to attack. Refer to this, the 3rd game of the Kamakura jubango between Go Seigen and Minoru Kitani. White 12 is the high approach. The idea is to cut the black 1 (4-4) from black 11. A low approach is not appreciated as black 11 is quite near to black 1, white has no chance to settle down, so, white would rather take the high approach, which is easier to run to the center and more preferable for the potential fight.

The game did not go as the Joseki, but the strategies of the two is too detailed to be explained here. Interested, please refer to Go's book on mid-game fighting.

Another possible purpose of high approach is to build the moyo. Refer to to http://eidogo.com/#xYvNft8F , a Fujitsu Cup game between Norimoto Yoda and Nie Weiping. Black 41 is the high approach. Nie trapped Norimoto Yoda on the right side and build a black "wall", so he took the high approach to build up the moyo. If black take a normal (low) approach, white would happily counteract by 5-6 to limit the black moyo, or even step further to a pincer.

This purpose is possible as, comparing to the low approach, black (4-4) is not easy to counteract by a pincer or kenuki (play at other places). For example, if black kenuki, white could attach on the other side of the (4-4), possibly another high approach, this double-high-approach is much more serious than a double-low-approach.