BQM 535

    Keywords: Joseki, Question


A (not-so-well) known joseki (B7 just a possibility, may be at a as well.)  
A joseki  

Tapir: I have trouble in understanding the sequence after from W12 onwards, any takers? Why B13 and not the cut, why W14?

Comparison to old joseki

The old joseki  

Uberdude: For reference here is the old version of this joseki. (In fact there are some other variations, for example B5 can be extension at c to stop white taking the corner). Black can then continue by playing at a or b. The downside of a is that white's marked stones still has lots of aji so black isn't thick here. Sometimes white will save them immediately.

The old joseki: continuation - somewhat overconcentrated for black  

If black cuts at b he is more solid, but white can force him into a rather overconcentrated shape as shown. Note that the white stones on the outside can be treated as kikashi and sacrificed if needed.

So, now lets compare these results to the (newer) joseki in the original question. In the new joseki white's corner is a bit bigger, but black is thicker on the outside due to the extension at B19. This extra thickness is generally judged to be of greater benefit to black than the extra points white got in the corner (afterall white is alive in both cases).

The new joseki, again  

The reason B13 is not the cut is then it would revert to the old joseki, which is not as good for black. W14 takes a liberty off the black stone preparing the trade. With B15 he abandons the corner in order to make the cut and get sente again to extend for thickness at B19, but before doing that black plays B17 in sente (otherwise he extends at W18 and white's eyes in the corner are gone).

Not so good 14  

W14 here is not good: it ends up like the old joseki but instead of giving black a ponnuki from capturing 1 stone, he captures 2 and is much stronger. Also the marked white stone has lost a liberty.

Cut in old joseki?

B5 in the old joseki cuts  

Dieter: Uberdude's analysis prompted me to ask why Black would not cut here with B15 while B13 at a allowed him to do so.

Uberdude: Interesting question, and not something I've thought about before (the previous point about the new joseki being thicker and thus better is something I heard about from others, the following is purely my own thoughts so more chance of error).

13 becomes dame  

If we assume the similar-to-joseki sequence as below, in which white captures black's original 3-4 stone and black gets thickness on the outside, it is obvious that B13 is poorly placed as it was played as a cutting stone, but then the other cutting stone was sacrificed leaving it as dame. Black could possibly take sente now, but that leaves white with the option to save her stone at a which is not only large in terms of territory, but also means white+square is a cutting stone again and could be saved in the future. White can also push through at b for some more profit.

Worse than joseki  

If black defends at B19 to prevent white saving white+square then it is like the joseki, but black has played on dame at B13 instead of a move with yose value at a. So this is unconditionally worse than the joseki.

Black pulls out corner stone  

Black may pull out his corner stone, and if white plays W18 then driving through with B21 and B23 gives black a thick shape in sente. If he extends at a it ends up like the joseki but white has even more points in the corner so that's worse than joseki. Instead black can tenuki now which is a big plus, but white gets the atari at a which is a nice kikashi. This seems a fair result to me.


Instead of taking the corner, white may choose to fight by extending at W16. The sequence shown seems a reasonable continuation.

Fighting part 2  

Black can then exchange B26 for W27. Black will then probably continue to attack the white group on the outside, perhaps at a. However, note that white can capture a stone at b for some profit (and a 2nd eye should it be needed) and if she does so there is some aji of a cut at c. Black can fix this cut in sente by playing atari there straight after W26, but would rather not as it wastes a ko threat and gives the wall on the outside more liberties (it's unlikely that wall will face damezumari, but it's nice to keep that possibility).

Fighting with B13 hane below not cut  

If black had played the hane instead of the cut for black 13, white's extension to W16 is also possible and the same sequence follows.

Fighting B13 hane part 2  

However, after the B26 W27 exchange, the marked hane stone is now perfectly placed to stop white capturing at a. This is a considerable gain for black. The downside is white can save a stone at b, but this is just 2 points and is heavy as it makes no eyes and loses liberties. So this diagram is better for black than the one 2 above.

Black refuses to give up corner

Black stubborn in new joseki  

Tapir: What I need to understand most is what happens if Black tries to resist the exchange like this or in any other way? (How exchanges happen and become inevitabe and are prepared to be so is a topic I generally do not understand very well.) If I am Black I feel like I want to avoid this exchange, because settling a medium corner in sente, when Black played first there, looks favourable to White for me.

Andy: This B15 is not good because white doesn't need to defend after forcing at a due to the presence of white+square.

Black stubborn in new joseki  

Uberdude: Holding on to the corner lets white split black's groups. This leads to a fight such as shown which look good for white. Black is not even 100% alive in the corner once white plays a.

Black stubborn in new joseki  

However, white is not alive yet either, so jumping out and letting black live in gote in the corner and then playing another move on the outside to attack black's heavy marked group seems a way to keep the game simple and good for white.

White tries harder  

In fact the kosumi might be slack; perhaps white can jump further to W28 and make miai of killing the corner and killing the outside group with W30. There are some cuts in the knight's move, but white could perhaps force with the a - d exchanges (which do help black on that side) to patch the weakness.

Tapir: Thank you for your comments. I agree - though I doubt all amateurs read to this point before rejecting the resistance variation. Btw. is Black alive at all after White a? I see ko at best or a bent four pseudoseki = one that allows attacking the outside first, but dies in the end of the game? (I believe some analysis presented here should find its way back to 34Point54ApproachOneSpaceLowPincerSeparate - is CamelCase? disabled?)

BQM 535 last edited by AndyPierce on March 9, 2011 - 04:44
RecentChanges · StartingPoints · About
Edit page ·Search · Related · Page info · Latest diff
[Welcome to Sensei's Library!]
Search position
Page history
Latest page diff
Partner sites:
Go Teaching Ladder
Login / Prefs
Sensei's Library