modern joseki question

Calvin Clark · Post by **Calvin Clark** » Mon Feb 08, 2016 11:10 am

I'd like some thoughts on this. I'll be the first to admit I'm over my head, and this is a warning to players who copy pro moves without understanding them.

In the following game, I played white. We both chose a pretty modern pattern, up until the cut at

which I've never seen. I am not sure if this is a mistake, and if it is, how to punish it:

Click Here To Show Diagram Code: [go]$$W is black's cut a mistake? $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 4 3 . . | $$ | . . . O . . . . . , . . . 1 . X . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 2 O . . | $$ | . . . . . . . . . . . . . . . 5 6 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . O . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . , X . . | $$ | . . . . . X . . . . . X . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

In the game I continued as follows...

Click Here To Show Diagram Code: [go]$$B game $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . 9 8 0 . . | $$ | . . . . . . . . . . . . . . . X O . . | $$ | . . . O . . . . . . . . . O 7 X 4 . . | $$ | . . . . . . . . . . . . . . 5 2 6 . . | $$ | . . . . . . . . . . . . . . 3 X O . . | $$ | . . . . . . . . . . . . . . . O 1 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . O . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . , X . . | $$ | . . . . . X . . . . . X . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

...but I did not like the result for white. (However, I am open to opinions that white might still be okay here. Even though white's corner is slightly bigger, black got sente and the thickness seems to work a bit too well with the micro-chinese framework.) I invaded twice, died the second time, and resigned, so I am wondering if the real problem is this early.

Click Here To Show Diagram Code: [go]$$B game continued $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . X O O . . | $$ | . . . . . . . . . . . . . . . X O . . | $$ | . . . O . . . . . . . . . O X X O . . | $$ | . . . . . . . . . . . . . . X O O . . | $$ | . . . . . . . . . . . . . . X X O 2 . | $$ | . . . . . . . . . . . . . . . O X 1 . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . O . . . . . . . . . . . . . . . . | $$ | . . . . . 3 . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . , X . . | $$ | . . . . . X . . . . . X . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

For reference, the following pattern (I am not sure if they call it joseki yet) is sometimes played. (Another option is

at

)

Click Here To Show Diagram Code: [go]$$B modern pattern $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . X O . . | $$ | . . . O . . . . . 8 . . . O . X 2 . . | $$ | . . . . . . . . . . . . . . . 3 4 . . | $$ | . . . . . . . . . . . . . . 1 X O . . | $$ | . . . . . . . . . . . . . . . O 5 6 . | $$ | . . . . . . . . . . . . . . . . 7 . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . O . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . , X . . | $$ | . . . . . X . . . . . X . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

I feel that somehow I am misunderstanding the point of the hane at

in the orginal diagram. In many games, it seems that white can try to get some kind of group on the right side that disrupts black's formation.

In that spirit, maybe something along the following lines is a better idea? Of course, I'm playing myself here. I am not sure black would really play this way.

Click Here To Show Diagram Code: [go]$$B break up framework? $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . X O . . | $$ | . . . O . . . . . . . . . O . X . . . | $$ | . . . . . . . . . . . . . . . . 5 . . | $$ | . . . . . . . . . . . . . 8 . X O 7 . | $$ | . . . . . . . . . . . . . . . O 1 3 . | $$ | . . . . . . . . . . . . . . . 4 2 6 . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . O . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . , X . . | $$ | . . . . . X . . . . . X . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Uberdude · Post by **Uberdude** » Mon Feb 08, 2016 11:34 am

Did you consider 4 at 5 in the second diagram to take the corner? Also your knowledge of the normal joseki should allow you to deduce that you don't need to play the gote hane connect of 8-9-10 but the 2nd line atari and then tenuki. It is very painful to give black that move in sente (but is it really sente? )

Uberdude · Post by **Uberdude** » Mon Feb 08, 2016 11:58 am

Calvin Clark wrote:I feel that somehow I am misunderstanding the point of the hane at in the orginal diagram.

I expect so. It is quite subtle and complicated, but an excellent case study of shape, move order, and engineering beneficial exchanges. I will explain later when I have more time.

Calvin Clark · Post by **Calvin Clark** » Mon Feb 08, 2016 12:05 pm

Uberdude wrote:Did you consider 4 at 5 in the second diagram to take the corner? Also your knowledge of the normal joseki should allow you to deduce that you don't need to play the gote hane connect of 8-9-10 but the 2nd line atari and then tenuki. It is very painful to give black that move in sente (but is it really sente? )

My opponent pointed out 4 at 5 as you mention, but we couldn't decide whether that was any good either. I do not even consider that during the game, which is a blind spot. Perhaps I was mostly concerned about white's safety.

Regarding the hane/connect, there is something that felt unnecessary about it before I played it, but I couldn't quite put my finger on it. I thought it was okay because I would have a forcing move later, but it would much later (like endgame). I wasn't picturing the joseki result clearly enough to draw your conclusion. So the result may be chicken-style.

Uberdude · Post by **Uberdude** » Mon Feb 08, 2016 2:07 pm

So, lets study the reasons for playing that hane of white 5.

First of all, let us start with this position, in which black has made the 2 space low pincer, white high double approaches, black attaches and then white plays the 3-3 and black blocks. I will go back and look at some alternatives in this sequence later, but we have to start somewhere.

Click Here To Show Diagram Code: [go]$$B Dia 1 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . 7 6 . . | $$ , . . . 4 . 1 . . . | $$ . . . . . . . . . . | $$ . . . . . . 5 2 . . | $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . . 3 . . | $$ , . . . . . , . . . |[/go]

As black chose to cut off the 3-3 stone from the left side, the simplest and most logical way for white to continue is to connect to his right side stone. The following sequence shows a sensible continuation for both.

Click Here To Show Diagram Code: [go]$$W Dia 2 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . b a c . . | $$ . . . . e d X O . . | $$ , . . . O . X 1 . . | $$ . . . . . . 4 2 3 . | $$ . . . . . . X O 7 . | $$ . . . . . . . 6 5 . | $$ . . . . . . . . . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

Black gets a nice wall, whilst white gets some territory along the left side. Black has some weak points around

, but for now he will usually pincer the white stone on the top side. In fact, you won't find this shape in many strong player's games because it is good for black (just one such pro game in ps.waltheri.net). The shape you will find like this is with the white counter pincer low instead of high. Without going off-topic too much, that's all to do with the white hane-connect at a-b-c: if it is low black has to make the bad shape defence at d (so white increased corner eyespace in sente) and the white stone still has lots of annoying aji, but with the high move black can choose to tenuki (the white cut doesn't clearly capture anymore, though there is the clamp) or make the good shape defence at e which renders the high counter approach stone all but useless.
So white's dissatisfaction with the above result causes him to look for a way to improve upon it, leading to the hane of Calvin's question. This threatens to play a and break through black's shape, and extend is a natural way for black to defend against that. (Note that the high counter approach here makes that breakthrough even more effective that with the low counter pincer, so to answer the question yes I think result is fine for white.)

Click Here To Show Diagram Code: [go]$$W Dia 3 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . X O . . | $$ , . . . O . X . . . | $$ . . . . . . a . . . | $$ . . . . . 2 X O . . | $$ . . . . . . 1 . . . | $$ . . . . . . . . . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

If after this hane extend white continues with connecting his 3-3 to approach stone and black wedges in as before then we get the following result:

Click Here To Show Diagram Code: [go]$$W Dia 4 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . X O . . | $$ , . . . O . X 1 . . | $$ . . . . . . 4 2 3 . | $$ . . . . . X X O . . | $$ . . . . . a O . 5 . | $$ . . . . . . b . . . | $$ . . . . . . M X . . | $$ , . . . . . , . . . |[/go]

This is vastly better for white than Dia 2: black has not connected to his pincer stone and white is out into the centre with good shape. If the pincer were a 2 space high one at the marked point then at least black could get some small consolation with sealing in white with a, but in this position he simply answers at b to split black's elephant eye. Another way to see how awful this is for black is to go back to Dia 2 and play the hane later, it is as if white played this 5 and black answered with 6:

Click Here To Show Diagram Code: [go]$$W Dia 5 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . X O . . | $$ , . . . O . X 1 . . | $$ . . . . . . 4 2 3 . | $$ . . . . . 6 X O . . | $$ . . . . . . 5 . . . | $$ . . . . . . . . . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

Obviously white is being stupid here and black should play like this:

Click Here To Show Diagram Code: [go]$$W Dia 6 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . X O . . | $$ , . . . O . X 1 8 . | $$ . . . . . . 4 2 3 . | $$ . . . . . . X O 6 . | $$ . . . . . . 5 7 . . | $$ . . . . . . . . . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

So this explains why, if black answers the hane with extend, when white connects he should not wedge. So the next idea for black is to cut. White defends by connecting (if he plays atari on the cutting stone black could either extend, or counter atari where white should connect and break through munching the corner). This threatens to push through and cut at a. (Note that white does not yet threaten to capture the black cutting stone with c-g due to the presence of black's pincer stone making f an atari). Black has 2 ways to defend against this threat, the solid connection of a or the bamboo connection of b. The joseki move is the solid connection, but often when you want to connect you prefer the bamboo joint if it is more efficient: here it is nice to take a liberty from the white counter-approach stone (it doesn't take a liberty from the white corner wall but that's not so useful). However the bamboo joint makes B turn into an empty triangle which should set alarm bells ringing. This empty triangle is not a purely theoretical drawback, but a sequence can show it to be a disaster.

Click Here To Show Diagram Code: [go]$$W Dia 7 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . B O . . | $$ , . . . O b X 1 . . | $$ . . . . . . a 3 . . | $$ . . . . . X X O . . | $$ . . . . . g O 2 d . | $$ . . . . . . f c e . | $$ . . . . . . . X h . | $$ , . . . . . , . . . |[/go]

If we recall from Dia 7 white couldn't capture the black cutting stone because f was an atari. But if black makes the bamboo joint then white can exploit black's bad shape. If when white pushes up with 1 black defends with 2 (he could also hane which is more complicated) then white crushes black with the following sequence ending in a ladder:

Click Here To Show Diagram Code: [go]$$W Dia 8 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . c . . . | $$ . . . . . . X O . . | $$ , . . . O X X O . . | $$ . . . . 7 b a O . . | $$ . . . . 3 X X O . . | $$ . . . 6 4 1 O X 2 . | $$ . . . . 5 8 9 . . . | $$ . . . . . 0 . X . . | $$ , . . . . . , . . . |[/go]

If black made the solid connection and white tries the same thing he is simply caught in a net and black's wall is safe with 4 liberties:

Click Here To Show Diagram Code: [go]$$W Dia 9 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . X O . . | $$ , . . . O . X O . . | $$ . . . . 7 . X O . . | $$ . . . . 3 X X O . . | $$ . . . 6 4 1 O X 2 . | $$ . . . . 5 8 9 . . . | $$ . . . . . 0 . X . . | $$ , . . . . . , . . . |[/go]

So to conclude the joseki, after black connects solidly, white plays the 2nd line atari now that it is clear he can't capture it with the other atari. This move is now a double sente that increases white's eyespace. White is now alive (and it's important to note that black b is not sente to kill the corner) and will tenuki, often to around a (to prevent black using his wall to attack the top side stone). If we look at the original hane and black's exension answer, we can see that the white cutting stone is not yet captured, and black would probably prefer the marked stone to be at c to securely capture white. However that extension is not a useless stone as it makes it rather heavy for white to pull out the cutting stone anytime soon, but white will often look to utilise that stone later when dealing with black's moyo, for example if white can build a group on the right side then he may persuade black to spend a move at d to secure the capture.

Click Here To Show Diagram Code: [go]$$W Dia 10 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . b . . . | $$ a . . . . . X O . . | $$ , . . . O . X 1 . . | $$ . . . . . . 4 3 . . | $$ . . . . . B X O . . | $$ . . . . . c O 2 5 . | $$ . . . . . . . 6 . . | $$ . . . . d . . X . . | $$ , . . . . . , . . . |[/go]

Now let's compare one of Calvin's possible sequences following the cut after the hane with the joseki above. They are pretty similar, so at the sdk level I would say that reaching this result was reasonable enough at resulting in something joseki-like and not being tricked too badly by that unusual cut. However, it is slightly better for black than the joseki because

has taken a liberty from the white counter pincer, without suffering the consequence that if he does that bamboo joint in the joseki then white can capture the cutting stone rather than playing the 2nd line atari.

Click Here To Show Diagram Code: [go]$$B Dia 11 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . . X O . . | $$ , . . . O 7 X 4 . . | $$ . . . . . 5 2 6 . . | $$ . . . . . 3 X O . . | $$ . . . . . . O 1 8 . | $$ . . . . . . . 9 . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

Also if white does the 2nd line hane-connect, maybe black should just answer as below to keep his group thick, as I think these exchanges help black increase his thickness more than they help white increase his territory (seeing as white was already alive in the corner and the black descent there was gote, if it were sente then white here would have more value as a reverse sente).

Click Here To Show Diagram Code: [go]$$Wm10 Dia 12 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . 2 1 3 . . | $$ . . . . 4 . X O . . | $$ , . . . O X X O . . | $$ . . . . . X O O . . | $$ . . . . . X X O . . | $$ . . . . . . O X O . | $$ . . . . . . . X . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

The reason being if black tenukis (maybe reasonable in the opening) then white can do this (maybe black should tenuki instead of 15/17):

Click Here To Show Diagram Code: [go]$$Wm10 Dia 13 $$ --------------------+ $$ . . . . . 8 . . . . | $$ . . . . 7 2 1 3 . . | $$ 9 . . . 5 6 X O . . | $$ , . . . O X X O . . | $$ . . . . . X O O . . | $$ . . . . . X X O . . | $$ . . . . . . O X O . | $$ . . . . . . . X . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

This analysis suggests the exchange of the hane for extension is beneficial for white, so why doesn't he play it before the 3-3? The answer is that black would answer the 3-3 differently to turn it into a bad exchange.

for

has become bad here, making black stronger in the centre and making the weakness at a more severe (it's possible for white to tenuki but defence at b/c would be honte).

Click Here To Show Diagram Code: [go]$$W Dia 14 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . c . . . a 5 3 . . | $$ , . b . O . X . . . | $$ . . . . . . . 4 . . | $$ . . . . . 2 X O 6 . | $$ . . . . . . 1 . . . | $$ . . . . . . . . . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

This hane for extension being a bad exchange also occurs in similar double approach josekis (with various pincer placements), with the low double approach black's extension makes the checking extension of a on the top side a powerful follow-up with the aim of b (in case it's not obvious white's cant cut it off with c or else his corner dies due to

being in the perfect place to make f capture). This is why the more modern double approach josekis see white taking the corner before playing the hane, he wants to avoid making the bad (for white) hane-extend exchange, but equally black may then answer the hane differently in order to try to turn the 3-3 and black's block into a bad exchange for white.

Click Here To Show Diagram Code: [go]$$W Dia 15 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . c b . . . . | $$ . . a . O d 5 3 . . | $$ , . . . f e X . . . | $$ . . . . . . . 4 . . | $$ . . . . . 2 X O 6 . | $$ . . . . . . 1 . . . | $$ . . . . . . . . . . | $$ . . . . . . X . . . | $$ , . . . . . , . . . |[/go]

Compare this to if black played the tiger mouth above after white played the 3-3 first without the hane, a rather soft move by black. White is now happy to connect up on the top. There is still the bad aji at a, but black can't really exploit this yet as he also has bad aji in whites approach stone. If black fixes that with b white can fix with c and is obviously happy not to have made the marked exchange (note that when white did make the hane black fixed with the 2nd line atari rather than b here to avoid giving white the atari at d).

Click Here To Show Diagram Code: [go]$$W Dia 16 $$ --------------------+ $$ . . . . . . . . . . | $$ . . . . . . . . . . | $$ . . . . . a 3 1 . . | $$ , . c . O . X . . . | $$ . . . . . . . 2 . . | $$ . . . . . M X O . . | $$ . . . . . . M b . . | $$ . . . . . . . d . . | $$ . . . . . . . X . . | $$ , . . . . . , . . . |[/go]

If white really wants to connect the 3-3 to the top side stone for global strategic reasons, then the hane for extend exchange can be justified in making a locally bad exchange in order to reduce black's choice and encourage him to let you connect that way, but it's a bad exchange that pros tend to avoid these days.

Uberdude · Post by **Uberdude** » Wed Feb 10, 2016 12:41 am

The end.

Sennahoj · Post by **Sennahoj** » Thu Feb 11, 2016 8:42 am

great explanations, thanks Uberdude!

Life In 19x19

modern joseki question

modern joseki question

Re: modern joseki question

Re: modern joseki question

Re: modern joseki question

Re: modern joseki question

Re: modern joseki question

Re: modern joseki question