(;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Chinese]SZ[19]KM[7.50]
C[Duan Rong 7p Using joseki in real games 5/130
Tianyuanweiqi
围棋 定式在实战中的应用
段嵘
29:41 mm:ss

Translated by Daniel Hu UK 5d
胡怿伦
20190805

Hello everyone. Welcome to the lesson. Today we will continue discussing using joseki in real games. We will continue discussing the Sword of Muramasa joseki.]
;B[qd]
;W[od]
;B[ld]
;W[pg]C[In this position, there is a move that white has to watch out for. Although this black play isn't a very good move, but white must also know how to counter it.

...

So regardless of bump or attach then bump, they are all moves that are very rare now. What other moves are there now?
]
(;B[oe]
;W[ne]
(;B[of]
(;W[pd]C[In the last few sessions, we discussed the variations after the hane and bump.]
;B[qf]
;W[qc]
;B[qe]
;W[nf]
;B[og]
;W[rc]C[When white has the ladder, we discussed the variations regarding the descent.

Today we'll do a simple review.

Even when black doesn't have the ladder, black can choose to make a ladder breaker. White has to defend both against the possibility of a ladder and also the possibility of black sacrificing.]
;B[nd]
;W[md]
;B[nc]
;W[mc]
;B[ob]
;W[oc]
;B[nb]
;W[mb]
;B[pb]
;W[pc]
;B[me]
;W[kb]C[The judgement of this depends on the whole board]
;AW[cq]
;B[dp]C[When black has a severe ladder breaker. It doesn't have to be capturing a group. Even this against a 3-3 playing two moves in  row is a severe ladder breaker.]
;W[]
;B[cp]
;AE[cp][cq]AW[dp][pp][fq]AB[dc][ci][cn]
;W[]
(;B[ep]C[Alternatively as we've looked at before, when white has this sort of corner, ]
;W[]
;B[eq]C[After the ladder breaker, breaking through here is a pretty severe ladder breaker.

When black doesn't have the ladder and black is captured in the upper right, black can still fight with a ladder breaker.])
(;B[qb]C[apart from a ladder breaker, white also has to watch out for a sacrifice strategy]
;W[rb]
;B[ra]
;W[na]
;B[kc]
;W[lb]
;B[ma]
;W[la]
;B[jb]
;W[oa]
;B[rd]
(;W[qa]
;B[mg]
;W[mf]
;B[lf]
;W[le]
;B[ng]
;W[me]
;B[qn]C[Even if black plays simply and sacrifices, this isn't necessarily profitable for white.]
;W[nq]
;B[on]C[black's potential is very impressive.])
(;W[mg]C[but it is worth noting that white has a counter-sacrifice strategy]
(;B[le]C[If black connects]
;W[qa]C[now black can't seal white in])
(;B[qa]C[If black goes for the kill, then white shorts the liberties.]
;W[ma]
;B[lc]C[forcing black to damage his shape in an empty triangle]
;W[le]
;B[pe]
;W[jc]C[black captures white in the corner, but white counters by capturing the outside.

This position isn't necessarily profitable for black. Even though black escaped, the loss on the outside was too big. White can be satisfied.]))))
(;W[qc]C[Other than the descent, there are some other variations. In recent years this attachment has become quite fashionable.]))
(;B[pd]C[There was a time where another move was considered very good to play. Bump now instead of extending.]
(;W[oc]
;B[of]C[If W descends, after B gets to extend here, this is better for B than all the variations we have been analysing.])
(;W[of]C[So W must atari.]
(;B[pe]C[Of course B can't connect, a classic crude shape.])
(;B[oc]
(;W[pe]C[Of course the normal judgement is that capture is good enough and completely acceptable for W.]
;B[nc]C[B ataris and connects underneath.

Although it was met with a ponnuki, B took a lot of territory below.

Normally these days we judge W as very thick.

...

So W has a lot of potential to the right. W has taken a ponnuki so W is not bad.

Recently B's bump way of playing has also become quite rare. Pretty much noone chooses this.]
;W[]
;B[rf]
(;W[rg]C[If B plays knight's move, W can attack block.])
(;B[rg]
;W[rf]
;B[qf]
;W[qg]
;B[re]
;W[rh]
;B[sf]
;W[qi]C[In this way, W can also seal B in.]))
(;W[nc]C[It has happened that in some situations, W has a strange way of playing, a move that opens ones mind and is also not bad. Not only not capturing but allowing B to capture.]
;B[nd]
;W[md]C[It is mostly because M16 is just perfectly cut off like this that W has such an unusual way of playing.]
(;B[od]
;W[me]
;B[mc]
;W[lc]
;B[nb]C[Of course, a necessary prerequisite for playing like this is that W can ladder M16.]
;W[mb]C[If W has the ko threat advantage, W now ataris on.]
;B[nc]
;W[kd]C[If B connects, W ladders this. W successively sacrifices two stones, and next hug captures this stone. In some situations it is a very interesting and worth considering variation.])
(;B[mc]C[If B doesn't connect and ataris, then of course this depends on which side has the advantage of ko threats.]
;W[od]
;B[nb]C[If W has a rather good advantage in ko threats, B can only connected under.]
(;W[pe]C[If W has a very large ko threat advantage, they can directly capture here.]
;B[lb]C[If B can't even win the O16 ko, then ]
;W[pc]
;B[ob]
;W[qc]C[It seems quite a loss for B, there are still problems in the corner.

The side has already been forced to retreat by W's captures.])
(;W[nd]C[If B has the ko threat advantage, W can only connect.]
;B[lb]
;W[pe]
;B[re]C[B tiger mouth protects, W can capture in sente, but B can defend over here. This is also where B can gain some territory, also very interesting.])))))))
(;B[pd]C[This bump has almost never been seen in professional play. But perhaps it can be called a trick move. If white doesn't know how to counter it, white may lose a lot.

[DH\] even though leelazero doesn't like it much, if I understand correctly, alphazero thinks it is not so bad, maybe better than P15, but R14 is still optimal [/DH\]

...

So the main reason this bump move is rarely played by top players is because it isn't hard for W to counter it. W only needs to remember to extend up P15.

This move is normally rejected.]
(;W[oc]C[If white descends

...

So this isn't a very good choice.]
;B[oe]
;W[ne]
;B[of]C[perhaps this is falling into black's plan. White is cut and it feels like white suffers a setback.]
;W[md]C[White's tiger's mouth defends the cut]
;B[le]C[If white doesn't play, black turns and white is sealed in again.]
;W[nf]C[White has no choice but to push to get out.]
;B[og]C[White has to play crude moves that damage the other stone.]
(;W[ng]C[To get out comfortably, white has to keep pressing, but this is too crude.])
(;W[lc]C[If you don't press, ]
;B[kc]
;W[lb]
;B[ng]C[In the future, when black turns, it is very uncomfortable shape for white to get out.]))
(;W[oe]
;B[oc]
;W[nc]
(;B[ob]C[If black descends, that's a bit soft]
;W[md]C[white simply tiger bumps.]
;B[lc]
;W[nb]
;B[]
;W[qb]C[If black doesn't play, then white's placement is very bad aji for black.])
(;B[nd]C[If black cuts, white has two options.

W only needs to consider atari or extend. B can't profit much.]
(;W[mc]
(;B[md]C[black can't connect]
;W[pc]C[white captures])
(;B[ne]C[black's sticking is sente, threatening the ladder.]
;W[of]
;B[ob]
;W[md]
;B[me]
;W[le]
;B[ng]
;W[kd]
;B[lf]
;W[ke]C[White captures a stone and is pretty well settled on the top side. And black doesn't have a very good way to handle the right side. There is no especially severe attacking combination.

This is also a rather clear way for white to play.]
(;B[oi]
;W[qi]C[White just needs to come out. Black is also in danger.])
(;AB[pm][pp]C[But if the right side is black's area of influence and white is rather nervous, then white can also choose a different way to play.])))
(;W[pc]C[[DH\] leela says this variation is 20% worse than N17, with -7% this move. [/DH\]

acting from the corner]
;B[ob]
;W[pb]
;B[nb]
;W[rc]C[B mustn't let W get too many liberties in the corner.]
;B[qc]
;W[qb]
(;B[rd]
;W[rb]
;B[rg]
;W[mc]C[Even though W's corner isn't alive, this extension is difficult for B to counter.]
(;B[ne]
;W[pe]
;B[qf]C[Although B's sticking is sente, W's turn is also sente and if B answers, the sticking becomes gote.]
;W[mb]C[This doesn't work for B.])
(;B[mb]
;W[pe]
;B[qf]
;W[lc]
(;B[lb]C[If crawling repeatedly on the 2nd line all the way, then even if B can kill with T17, B is at a big disadvantage, clearly a big loss.]
;W[kc])
(;B[kc]
;W[lb]
;B[kb]
;W[md]
;B[la]
;W[me]C[Up to here, there are still problems in the corner. W can throw in N19 with Q19 sente, W is already alive.]
;B[ma]C[Then B must fix to kill]
;W[rh]C[W has the initiative to act and this side doesn't work for B. B is already squeezed only the first line at the top and it is very hard for B to deal with this attachment.])))
(;B[rb]C[If blocking doesn't work, then the only remaining is to push and cut. This is a famous tesuji, which can capture W locally, colloquially called tombstone shape.]
(;W[ra]
;B[sb]C[descend and then short liberties]
;W[sc]
;B[pa]
;W[sa]
;B[rb]C[A basic tesuji.]
;W[sb]
;B[oa]
(;W[qa]
;B[rd]C[W all dies.])
(;W[rd]C[But even to be captured into this, this is better for W, which is a bit unexpected. Crawling across is sufficient.]
(;B[qe]C[There is no point to extending.]
;W[re]
;B[qf]
;W[qa]C[W has another eye and lives, and B can't stand W N17.])
(;B[qa]
;W[qe]
;B[rb]
;W[qb]C[Although 3 cutting stones have been captured, after W recaptures, everyone can make a judgement.

This was B's corner originally but now W's points aren't necessarily less than B's. Both side's number of moves are the same.

W's shape seems to be thicker than B's. B still leaves a problem at N17 which W can make use of with sente around K17. 

B didn't get any profit from first playing in the corner, so this is a clear counter plan for W.])))
(;W[sb]
;B[ra]C[Golden chicken standing on one leg and W is already dead.])))
(;W[ob]C[[DH\] leelazero says this is 2% better than N17, just simply to compromise. [/DH\]]
;B[pc]
;W[md]
;B[ne]
;W[me]
;B[nf]
;W[mc]
;B[of]
;W[qk]
;B[jc]
;W[qf]))))
(;B[of]C[There is the elephant jump out. This seems to have been first played by a Japanese 9p a long time ago called Kajiwara Takeo.

This can clearly also cut W. This is maybe more difficult to deal with than the other moves.]
(;W[pe]C[Normally W cuts the elephant jump. This developed into a joseki like variation.]
;B[pf]C[Of course B must cut.]
;W[qe]
;B[qf]
(;W[rd]C[If W captures the corner, ]
;B[rf]
;W[qc]C[This is an exchange between territory and influence.

B cut down through. As to how much value this has, it still depends on harmony with surroundings.]
;AB[pp]C[If this corner is B, then ]
;B[pk]C[Then B's profit is very reasonable.])
(;W[rf]C[Normally considered stronger than hane in the corner is hane here.]
(;B[rd]C[B doesn't feel comfortable to die so easily in the corner, intending to leave some aji.]
;W[re]
;B[qg]C[W has many ways of capturing now. B just watches how W captures.]
(;W[qc]C[Normally the move considered to capture more cleanly is this clamp. It is relatively clear, with not much room to mess around. Hane is met by cut.]
(;B[nd]C[The point of the descend was to profit like this with some aji remaining.]
;W[nc]C[The tightest move is hane.]
;B[oc]C[At this point, cut is a tesuji.]
(;W[pc]
;B[oe]C[Now B can probe W once.]
(;W[ob]
;B[pd]
;W[oc]
;B[rg]C[B gains this turn in sente]
;W[rc]C[B gets an influence in sente. When the right is rather big, this is also a variation to consider.])
(;W[pd]C[If W doesn't want to give the S13 sente, then]
;B[mc]C[B gets this tiger sente.

B will probe left and right, always finding a way to profit.]
;W[ob]C[Of course, if W just submits every time, this is also a reasonable variation. Which side profits still depends on the harmony.]
;AW[gc][dd][pp]C[This is a rather unlikely comparision.

If B's outside doesn't have much use, then W can be satisfied to play simply and clearly like this.]))
(;W[ob]C[W also has another move to stubbornly resist.]
(;B[pd]C[B's strongest move now is pinch atari.]
(;W[oe]
;B[pc]
;W[pb]C[Absolute vital point or else W will all die.]
;B[rc]
;W[qb]
;B[ne]C[After extending liberties, B returns to blocking.]
(;W[rb]
;B[mc]
;W[nb]
;B[rg]
;W[sc]C[In this way, B successfully sacrificed stones. Even though W captured the stones, B got N17 and S13 in sente. W don't really feel good here.])
(;W[rg]C[W also has purposefully now capturing. This variation is now already very rare to see in games. There was a time when it could regularly be seen.

Although not played now, this is a very interesting variation in the development of this joseki that I think it is necessary to discuss with everyone to help them understand.

Able to capture, but not capture and instead crawling out. This sort of idea is worth everyone learning from and considering.]
(;B[qh]
;W[rb]
;B[mc]
;W[nb]C[If B continues extending this has a big clear difference. B turning S12 is now gote, so if B later surrounds territory on the right, the territory difference will be very big. Of course B won't be willing to let W get such clear profit.])
(;B[nb]C[B still has life potential.]
;W[mc]
;B[mb]
;W[lc]
;B[lb]
;W[kc]
;B[rb]C[B has 4 liberties on each side and W has 3, locally W is dead.

It may seem that B has captured where W could have and it is a loss, but this isn't the case. Crawling out already took this technique into account. Why choose this variation even though W knows B can capture the corner?

We will discuss this next week.]
;W[ra]C[But B must be slightly careful here.]
(;B[qa]C[Throw in is the vital point]
;W[pa]
;B[na]C[This is a clean capture.])
(;B[oa]C[B mustn't be slightly careless with one move.]
;W[na]
;B[ma]
;W[qa]
;B[sb]
;W[sd]
;B[na]
;W[kb]C[In this way B can't approach from either side, which is big trouble and rather B has been captured.]))))
(;W[pc]
;B[oe]C[This reverts]
;W[oc]
;B[rg]))
(;B[oe]C[Because if B now ataris here,]
;W[pd]
;B[mc]
;W[nb]C[W now has a better option. The corner is much larger, and B is open on the upper side. This bit of difference is quite big for professionals.

So this diagram is difficult to accept for B. ])))
(;B[rc]
;W[pc]C[W just retreats and B can gifted dead stones.]))
(;W[ob]
;B[qb]C[If just here, the corner aji is too bad. Next move to tiger Q17 lives.]
;W[pc]
;B[nc]C[The aji is very bad. W's shape feels very heavy and cramped.])
(;W[pc]
;B[nc]C[Directly diagonal, B profits from the outside with the knight's move peep. W still needs to fix the corner.]))
(;B[qg]
;W[rd]C[If B turns, then W returns to the hane. This exchange is clearly profitable for W.])
(;B[rg]C[If B blocks, W simply connects.]
;W[re]
;B[qg]
;W[ob]C[The way W captured this corner is very big. Normally modern Go players are increasing valuing territory. With this sort of diagram, definitely more players will prefer W.])))
(;W[pf]C[But W directly blocking isn't necessary failing.]
;B[oe]C[The shape is to bump and cut]
;W[pe]
;B[pd]
;W[nd]C[This is also a very complicated fight.])))