(Referenced by 222br)

: ( 222fb)
Black gives atari at the upper edge.

:
But White extends into Black's corner, sacrificing the two stones at the top.

:
Black must capture now.

:
White further attacks Black's hidden weaknesses in the upper left corner.

:
Black peeps at the second line, threatening to cut at .

:
As a matter of course, White has to connect.

:
Black strengthens his position at the upper edge.

:
White gives atari at the left edge in return.

:
Black is forced to connect.

He cannot afford to give counter-atari with , because he would be unable to win the ko fight after White captured with .

:
Now, the time is ripe to cut Black's forces apart.

After having finished all the corrections in some subvariations below, we noticed that White could optimise her starting position by capturing the nakade with beforehand. Black would be forced to fill it with some stones that would become additional prisoners in one of the subvariations below.

:
Black pushes at the second line.

Jumping to , instead, would end even worse for him.

This later part of the variation tree - with the exchange of in the upper right for in the lower left - was analysed and commented first by us.
Please understand it as a status of our research on the problem from a time when we did not have the new KataGo (2020) network available, only enriched with a few annotations that reflect our current state of knowledge. We hope for your kind understanding that we did not entirely update this section here.

You will find an up-to-date analysis of this position at an earlier subvariation ( 222bv).
That part of the variation tree is two moves earlier, but the absence of the marked exchange does not really matter. It will be easy for you to transpose the sequences from there to here.

: (A 222fd ; B 222fi ; C 222fj)
White's hane results in difficulties for Black to approach her stones from above, using the special properties of the corner.

Further attacking Black's corner with pushing at from the outside - without the preparatory hane in the corner - would be overdoing things. This mistake would lose White two points overall, compared to the alternative subvariations here.
As mentioned above, if White wanted to choose this line of play, she should have captured the nakade before cutting Black's forces in the upper left corner apart. This procedure would compensate for the loss of two points that she suffered here.

Continuing with the Main Semeai, e.g. with a move at instead, would give Black the opportunity to cut the height of White's win in half.

Peeping at the upper edge, with , gives Black the opportunity to start a semeai with White's centre group, but which he will lose.

: ( 120 222fk)
It is tempting for Black to push at the top now.

Continuing with the Main Semeai, however, would be better for him, cutting the height of White's win in half.

: (123 222fm)
White reduces the liberties of Black's cut-off group.

As we know now, blocking at would be playable, provided White chooses the correct continuation (which we did not know at the time when we initially commented this subvariation).
Even without that knowledge, White's win would still be higher than in the solution, while being worse compared to the sequence following here.

Alternatively, White could take Black's large group at the right a decisive liberty by connecting at , punishing Black for having just lost one tempo in the Main Semeai.
However, the fight in the centre might become quite difficult, and White has simpler options for achieving a large win.

:
White connects, winning the fight for life and death with Black's stones below.
Please note White's approach-move liberty at .

:
Black forces White to connect her three stones below, ...

:
... before capturing White's stones at the top.

:
White continues in the lower left corner.

Please note that the hanezeki will remain untouched, as both large groups live independently.

According to our estimations, ...

White will win the game by about 15 points.