This has been bothering me for a while.

Consider the basic 4-4 joseki below:

Click Here To Show Diagram Code

[go]$$Wc White escapes
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O X a . . . . . .
$$ | . . O X X O . X . . . .
$$ | . . O X . . . . c , . .
$$ | . . . X . 1 . . . . . .
$$ | . . O b . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

It's often said that here is the standard way for White to move out if he wants to. Yet I haven't found a standard way for Black to respond to this move, though a is often suggested. This keeps Black connected and prevents a White descent at the same point. The problem here is that White is then free to play something like b or c. This immediately undermines the only advantage Black got out of this joseki, namely influence.

Is there a way for Black to get a good result after ?

I was taught to peep and then g18.

Two things to consider:
- with , it means B already has a move elsewhere.
- how B replies to depends on the rest of the board.

Depending on the surrounding position, it may be possible to lean on White's corner and make a fight out of it.

Click Here To Show Diagram Code

[go]$$Wc
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O X . . . . . . .
$$ | . . O X X O . X . . . .
$$ | . . O X . . . . . , . .
$$ | . . . X . 1 . . . . . .
$$ | . . O 2 . . . . . . . .
$$ | . . 3 4 . . . . . . . .
$$ | . . 9 5 6 0 . . . . . .
$$ | . . . 8 7 . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

This may be a mistake, as white doesn't have to connect against the peep, but instead takes the big push on the outside. Now the aji of black's cut at c is significantly reduced as white b is sente, whereas it is not had black directly connected at a.

Click Here To Show Diagram Code

[go]$$Wc Peep then block
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O X a . . . . . .
$$ | . . O X X O . X . . . .
$$ | . . O X . . 2 . . , . .
$$ | . . . X b 1 . . . . . .
$$ | . . O 3 c . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

Black can consider pushing once and then playing the peep and connect. The idea of the push is to prevent white playing there to build influence on the side, but still the turn after the push is a nice point and it is somewhat distasteful to push and tenuki as black loses the option of other moves in this area such as the approach at a. The idea of the peep is to make white heavy (and reducing eyeshape) before connecting and make those 3 stones a liability to save. However, white might not connect right now and just be content that she has made black play that slightly aji keshi push.

Click Here To Show Diagram Code

[go]$$Wc Push then peep
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O X . 6 . . . . .
$$ | . . O X X O . X . . . .
$$ | . . O X . 5 4 . . , . .
$$ | . . . X . 1 . . . . . .
$$ | . . O 2 . . . . . . . .
$$ | . . 3 . . . . . . . . .
$$ | . . a . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

Something I noticed in your diagram is there is no marked black extension stone on the top side. If this (or other black support in this area) is already in place then the large scale fight from continuing to push on the corner Sverre mentioned becomes more appealing. But without this stone I would even raise questions about white starting at . The normal aim for white in that shape is to play the checking extension of a, aiming at attacking black's group with b next (black can of course answer a but may feel that is giving white a good exchange). So if white plays black does not necessarily need to feel so sad about simply connecting at b as white is losing some other options in this position.

Click Here To Show Diagram Code

[go]$$Wc Extension?
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O X b . . . . . .
$$ | . . O X X O . X . a B .
$$ | . . O X . . . . . , . .
$$ | . . . X . 1 . . . . . .
$$ | . . O . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

I'm well aware that the marked extension makes almost ridiculous. And in this position I would feel much better connecting at b.

This joseki, however, has been played without the marked extension for ages with both sides satisfied. And I've encountered a few times in actual games and never a. I assume most players in the lower ranks (I'm trying to get enough won games to go up) aren't aware that a is proper, or don't know how to follow it.

Far from criticizing, I have to say your post has been very insightful. Thank you.



I do love starting large, one-sided fights. I'll keep this diagram in mind when I have support on the right.

It's not ridiculous with the extension in place, in fact it's a proper reduction. What would be ridiculous is white playing at b. My rule of thumb for follow-ups with that one-space pincer joseki is if black doesn't add the extension then look forward to a and then b, with the extension that jump of 1 or the push out from the corner (e) are simple continuations. More complex is the attachment at c, which is made more effective if white has the checking extension at d. There is some coverage of these ideas in Kim Sung-Rae's book "After Joseki".



Yes, you don't need to add the extension and a common reason for playing that pincer joseki is to get sente, However, the influence group is not that strong and makes no points yet so adding a move to stop white hassling it is big; Lee Chang Ho was fond of it as it fitted his calm style.

I'll defer to Kim Sung-rae of course.

This is one of my favorite examples of a joseki that might be a joseki for pros but is not an even result for kyus. Black's influence is kind of vague.

Reminds me of:
http://gooften.net/2012/01/26/excerpt-f ... ic-joseki/

Isn't influence always vauge? It seems to be part of the definition of the term.
If it were specific, it would be territory.

_________________
Help make L19 more organized. Make an index: https://lifein19x19.com/viewtopic.php?f=14&t=5207

I think it's 'vague' in the sense that it's difficult for us kyu players to find the appropriate way to make use of it.

Edit: I think another problem is that it's not only a small wall, but compound that with the fact that like most walls it is not 'thick' makes it a potential liability.

No longer is influence ambiguous. Apply my definitions to get precise values. Apply part of them to get precise values for specific aspects of influence.

I do kind-of agree with you. Black's influence can become a weak group if you aren't careful. My opinion of this joseki is that locally speaking I prefer white (solid shape with territory) so it's only okay for black if the direction of the pincer makes sense globally. I've not seen it said so explicitly by professionals, but looking at their games I do get this feeling as you often see the one space pincer tenuki-d, the thinking behind which I imagine to be "Okay, so because you were scared of me making a mini-chinese / Kobayashi / whatever nice formation you decided to play the one-space low pincer to prevent that, but to do so you had to choose a territorially inferior result so I'll go do something else like make a shimari and likely come back to 3-3 later".



No, by vague I understand snorri to mean not particularly good influence. An example of non-vague influence would be the standard 3-3 invasion under 4-4 joseki which produces strong influence with good shape. I expect if you used Robert's exciting definitions you would find this has a bigger number.

Click Here To Show Diagram Code

[go]$$Wc Clear influence
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O X . . . . . . .
$$ | . . O X . X . . . . . .
$$ | . . O X . . . . . , . .
$$ | . O X . . . . . . . . .
$$ | . O X . . . . . . . . .
$$ | . . X . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

No, under Robert influence is a compound object consisting of several numbers ( 5 if I remember correctly ). One position compared to another might be superior in some numbers but inferior in other. His influence does not provide a total ordening. This significantly complicates the application of this concept. IMO.

There's nothing vague about this influence!

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Influence stone difference = 7.

It is the reality that influence is not basic and simple. The definition identifies the minimally necessary information if one wants to get a reasonably complete information. E.g., knowing that a moyo is surrounded by only stones of its player is insufficient information. One must also know whether and how well the opponent can invade and live or connect (by running out).

Although reality and so the definition is complicated, application can be made as easy as possible for a purpose of application! In particular, one can consider only 1 or 2 of the 5 (or 6) parameters, if already they provide sufficient information for the purpose of application.

Remember that, although 5 numbers can be involved, their interesting values tend to be restricted to -2 (or smaller), -1, 0, 1, 2 (or greater), *. (Usually, the territory-parameter needs only the values 0 or 1.)

5 values sounds like a lot, but what do they describe?
- How well does Black live?
- How well can Black connect?
- How well does White live?
- How well can White connect?
- Does the player, who obviously is the one with territory potential on an intersection, already have territory there or does he need to make (only) another play to get territory there?

These are questions one should be asking oneself and answering anyway. So why not answer the questions accurately by determining the numbers, when this accuracy is useful for a particular purpose of application?

If one neglects to reflect these questions, then one makes too rough judgements such as "Black has a moyo here. I do not care if a reinforcement is neeeded. If White invades, I pray and try to kill." - Good strategy comes from (more) accurate judgement.

It is sometimes sufficient to apply only the other definition, that of influence stone difference. It relies on only 1 number.

And same for this? This influence is not so good as less eyeshape and doesn't reach as far to the right. Would it differ in another of your metrics?

Click Here To Show Diagram Code

[go]$$Wc Not quite so good influence
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O X . . . . . . .
$$ | . . O X X . . . . . . .
$$ | . . O X . . . . . , . .
$$ | . O X . . . . . . . . .
$$ | . O X . . . . . . . . .
$$ | . . X . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

If you want to measure how far influence reaches (rightwards) and how alive the thickness is etc., then the 3-parameter thickness definition and the 5-parameter influence definition (applied to empty intersections right of the wall) are possible tools while the influence stone difference does not assess it at all.

For influence stone difference, your position has the value 6 because one black stone does not contribute significant influence (removing the stone has an only marginal effect). Those that do are marked here:

Click Here To Show Diagram Code

[go]$$Wc Stones creating significant influence
$$ -------------------------
$$ | . . . . . . . . . . . .
$$ | . . O O B . . . . . . .
$$ | . . O X B . . . . . . .
$$ | . . O B . . . . . , . .
$$ | . O B . . . . . . . . .
$$ | . O B . . . . . . . . .
$$ | . . B . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . . . . . . . . . .
$$ | . . . , . . . . . , . .
$$ | . . . . . . . . . . . .
$[/go]

In all of them, although part of the parameters might still be the same.