(Referenced by 98; 133; 136; 225; 534; 717; 819; 834)

: ( 219)
Black does not play at here, as in the previous variation, but captures White's stones in the lower left, just before the Crosscut Sequence in the centre is completed. This turned out to become surprisingly complicated. Over a long period of time we saw no possibility to follow the well known proverb "In a semeai, capture the ko last.", which worked so well with Black's move at ( 225).

:
White must try to capture some Black stones, and to live with her group in the top left, to avoid being forced to capture the hanezeki's tail.

:
As is shown for the C-variation for of our solution ( 532), Black cannot eliminate his weaknesses in this area, which will enable decisive White ko-threats later, with a move at .

: (A 366)
Black does not play at , to save his stone (= ), which would follow a suggestion of Cheng ( 534). This hanging connection of here will lead to much more complicated variations than those with the solid connection, because this move also threatens the eye-space of White's group on the right.

: (A 370)

:
White must initiate a ko here, instead, starting with an atari at .
This is the only way to chance the relationship of liberties in the semeai on the right in White's favour due to Black being able to do without Michael Redmond's kikashi in the bamboo joint, shown as a sub-variation below. We revealed this potential development lately in 2015, so we ask for your kind understanding that we did not yet adjust the order of sub-variations here.

In 2020, we learned that there was a miscalculation in one of our following sequences, resulting in the statement above. However, it will be OK to follow the sequence shown here.

: (94 379)
Playing immediately at , instead, leads to a similar result as with this kikashi inserted.

KataGo prefers doing without the exchange at the upper edge.

: (99 380)
Black can start with a move at , too, resulting in a change in the order of moves.

: (A 381)
Black could try to cut at , instead, but will not succeed either.

: (109 382)
:
Black should do without this kikashi in the bamboo joint, and play at immediately.

We have to admit that there was a miscalculation in the subvariation. In fact, changing the order of moves – as suggested above – will not make a difference.

: (B 383)
Starting an "early" ko, with a move at here, is more attractive for White than with the semeai in the previous section. Here, Black has captured White's lower left side already, so there are only a few ko-threats remaining against White's group in the upper left corner.

:
Black prepares the ground for a "quick" ko-threat against White's marked group in the top left corner, which (if unanswered) would allow him to win the semeai between this White group and Black's group in the lower right corner. He plays this sequence at the latest moment that is possible before being forced to connect the hanezeki's tail.

:
White's centre group has five liberties left, which seems to be the optimal moment to start a "late" ko. Please note that White could have already chosen an "early" ko successfully, e.g. with move .

: (137 386)
Black is forced to use one of his very valuable remaining ko-threats before connecting the hanezeki's tail.

It does not help Black to immediately play at , instead.

Ko:

:
It would not change the matter to give atari at , instead.

:
White captures the ko for the last time, and leaves Black helpless, who has no valid ko-threats left. White's group in the upper left corner has too many liberties, so White could ignore a Black move at , for example. White will still win the semeai with Black's connected group on the lower edge. White's centre group has four liberties left, so Black's atari at on the upper edge is no valid ko-threat either. In the centre, neither nor will pre-empt White connecting the ko, so winning the game.

Please note that White still has the double ko-threat at in the upper left corner.