BillSpightsElfPositionalJudgmentExercises/Game 12 ELF's Solution

Black is ahead.  

Elf's mainline game continuation for .

Mainline game continuation (32.3k playouts)  

is big. But then there are several reasonable plays for Black, so the number of playouts per move drops precipitously. That is why Elf's variation stops so soon. There are just too few playouts to have confidence in further plays. :-| attacks the stone, and then attaches for sabaki.

Earlier in the game.

White feels his oats  

is and was unusual. Yang was 19 or 20, and was feeling his oats. He has since made 9 dan, and Sensei's Library says that "Yang uses experimental corner plays." Elf thinks that loses 12% from par, and also that loses 8% by comparison with the enclosure at a.

Black returns the favor  

Elf thinks that loses 12% by comparison with the usual joseki play, the hane at 8. Then Elf thinks that loses 9½% by comparison with the kosumi at a, which seems kind of obvious when you see it. But was joseki at the time and is still the majority choice on Waltheri.

When I was learning go was the textbook play. The reasoning was that, although it was only a 3 space extension from the 3 stone Black wall, that was OK because a White keima at b would require a response. Better just to take a slight hit. Also, Takagawa taught that the Black wall needed an extension because it had no eye potential (or very little). OC, today's bots will have none of that. Elf regards as losing 13% by comparison with c, approaching the top left corner.

Elf's mainline variation for  

After White chooses one of Elf's favorite joseki, the kick and extend one space on the third line. Then Black plays the double keima at E-07. Don't let the double keima get away, as the saying goes. To me this is the answer to the conundrum of how to support the weakish Black wall. Don't extend. Play the double keima instead. :D plays a wedge on the bottom side, "killing Black's influence," as the amateur saying goes. ;) approaches and makes some territory. makes a base.

Elf's mainline variation for , continued  

and attack the White group on the bottom side. Then and defend and force a bit of overconcentration on Black. Then invades the top right corner.

Splitting play is a mistake.  

White played the splitting play, , instead of enclosing the top left corner. Note that it is at R-09, not R-10. The reason, as Sakata explained in his writings, is that it is closer to the Black moyo on the bottom side, which means that can extend only as far as R-07 instead of R-08 (b), making a bigger moyo. Be that as it may, Elf thinks that loses a whopping 18% by comparison with a. Today's bots really devalue the sides early in the game.

Elf's mainline variation for  

After encloses the top left corner, does not play the "big" point on the right side, but makes a tight enclosure of the bottom right corner. Elf has a slight preference for this enclosure over the one at a, perhaps because it is closer to making territory. invades the top right corner. blocks on the right side and follows up with to create a large sphere of influence.

Elf's mainline variation for  

Elf regards as a minor error, losing 5½% by comparison with this approach to the top left corner. White plays one of Elf's favorite joseki, and then encloses the bottom right corner. It does not appear that at R-09 (marked) has hampered Black in any way.

Back to the game.

Questionable extension  

The extension, , was questionable at the time, without an enclosure of the top left corner. Elf thinks that it loses 9% to par. (Usually in these discussions I do not mention losses under 10%, but I think that these are of some interest, and they do add up to White's disadvantage.) Elf also considers as a minor error, losing 6½% by comparison with the enclosure at a. Takagawa wrote that it was OK to tenuki instead of making a base for after the approach, , because if Black now pincered , White could be satisfied with having prevented Black from making the ideal extension on the right side. That reasoning no longer holds up, as Black is in no hurry to play on the right side. But Elf still prefers the enclosure over the base.

Elf's variation for  

After encloses the top left corner, encloses the top right. Note that it leaves White room to make a base with the two space extension. Note also that White does not do so, preferring the Go Seigen/AI shoulder blow. Whether that is because Elf regards the two space base as weak, or because at this point it prefers central influence to territory, Elf does not say. ;)

Elf's mainline variation for  

After encloses the top left corner, extends only to L-17, inviting the White extension to J-17. After each side has made a moyo on the top side, attacks the White group on the right side while enlarging the Black sphere of influence in the bottom half of the board. Then invades Black's top side moyo. The stratagem of is interesting. :)

Back to the game.

Dirk bad?  

Elf thinks that loses 9½% by comparison with the underneath attachment at a.

Elf's mainline variation for  

=

and secure the corner, then turns to the top left corner. After White lets the top left corner fend for itself and attacks the top right corner. Note . Many players would play :wt: ( ) instead.

Back to the game.

Little Bad Peep  

Elf thinks that the peep, , loses 11½% by comparison with a.

Elf's mainline variation for  

is a natural connection for the White stones on the top side. Black has other reasonable options for , which gets 24.4k playouts. Elf's second choice is the pincer at a, which gets 13.5k playouts. White's connection is secure, so shifts to the submarine attack against the top left corner. is a calm, solid response. Then is a shape play to take away a possible eye shape for Black. It is striking to me how often Elf chooses plays that prevent the opponent from making good shape. :)

Not exactly a conclusion.

To me the main evidence of Black being so far ahead is the disarray of the White stones. On each side White has stones that don't seem to be doing much. The stones in the top left are disconnected and high. The stone on the top side is not connected to White's corner stones. The stone on the bottom side is a singleton, and White's base on the right side is weak. Elf's review of the game to date indicates that the main culprits are Elf's failure, time and again, to enclose the top left corner, the splitting play on the right side, the extension on the top side, and the peep in the top left instead of connecting to the White stone on the top side.