We start our analysis from a position, where Black has not played any of his kikashi , / , in the hanezeki at the right and , / , at the left side.

Please note that Black has eight Main Semeai Liberties in the moment, and so is one over par with regard to the Punishment Semeai Liberty Level.

:
Black's forcing move at in the upper right corner destroys White's potential ko threat of her atari at in sente. Some of the following subvariations contain a ko fight in the left centre of the board (with White having to find the first valid ko threat), so the size of each side's supply of ko threats is important.

: (A 648ab; B 648gf; C 649aa)
We decided to include Black's first kikashi in the hanezeki in the main line of this new presentation.

White cannot take advantage if Black does without this kikashi. Black might also continue by immediately capturing White's single stone in the top right, with a move at , or by largely simplifying the issue by KataGo's solid connection at . In both cases, the final scores of the game will remain the same as with this kikashi played "early".

When analysing this position with KataGo (2021i), we found that the atari at White's two stones in the lower left was the top entry in the list of favoured moves.
As it eventually turned out, the best lines of play for both sides resulted in White + 2 / Black + 4, the same scores as with the kikashi played "early".
Please note that KataGo (2021i) does no longer consider , which will benefit White in the Semeai Variation.

: ( A 648gm;  B 648go; C 648gu; D 648hm)
Black's tenuki was the main concern of our investigation; therefore we present it as part of the main line here.
In principle, Black loses two Main Semeai Liberties by this move. But becomes an approach-move liberty of his group, so the net effect of this move is a loss of one Main Semeai Liberty.

Alternatively, Black might push at in the top left, or again play KataGo's solid connection at .
Please note that Black's tenuki is not forced. This implies that Black can always use the solid connection as a fallback position, should White find a favourable enhancement in the tenuki branch of our variation tree.

Defending his lower left corner with , or blocking with in the left centre, instead, will not change the overall result of the game (i.e. the Capture Variation's score), but Black's score in the Semeai Variation will become worse.

: (D 648id; A 648io; C 648ix; B 648kh)
The wedge at is White's strongest answer to Black's tenuki after her bad-shape move.

White might start her manoeuvre at the left side of the board with the forcing move at in the lower left corner, resulting in a change in the order of moves.

Pushing at , or blocking at , instead, are White mistakes.
Starting with the forcing move in the lower left corner is even worse for her.

: (A 648ku)
Black captures White's just played stone.

Black might consider taking White's group at the left a Main Semeai Liberty by playing at , instead. However, this move will not have the expected result, but lose one point overall.

: (A 648kv)
This kikashi should be left unplayed and kept as a potential ko threat.

We only realized this fact shortly before the 2020 update of our website was completed. We ask for your kind understanding that in this late phase of our elaboration we both felt unable to reduce the move numbers by two in many diagrams, and to restructure the variation tree.

However, we thought for long that this atari could be utilised as an additional ko threat. But this is not true, as we only discovered in late 2020.

: ( A 648mk; B 648nn; C 648ns)
The order of moves is very important now.
White must play in the lower left corner first, forcing Black to answer at .
If were played after , Black would have a better option at his disposal.

However, heading for Dany's Ko by playing Dany's atekomi , instead, will increase White's score in the Semeai Variation.

Erasing Black's potential eye in the centre with the central attachment at , instead, is a White mistake.
Starting with the forcing move in the lower left corner will be even worse for her.

: (A 648nx)
Black's answer is forced.

Playing at , instead, will end in a disaster for him

: ( B 648nz; C 648oc)
For the second time in a row, White plays in the lower left corner.

Heading for Dany's ko, by playing Dany's atekomi , instead, will increase White's score in the Semeai Variation.

Erasing Black's potential eye in the centre with the central attachment at , however, will end worse for her.

: ( B 648of;  A 648or)
Black plays KataGo's solid connection "late", securing an eye in the centre.
Black loses one Main Semeai Liberty by this move.

However, this move is not correct, as Black will be pushed below the Punishment Semeai Liberty Level by White's throw-in in the lower left corner.
He simply cannot afford providing White with an additional approach-move liberty in the lower left corner as he does in this main line.

Joachim found that this is the moment - just after White played her move (=) in the lower left corner - to give atari at White's two stones with a move at , returning to what I thought to be the main line.
If White chooses Dany's Sacrifice (i.e. lets Black capture her two stones), there is no need for Black to connect at in the left centre, so he gains one move back in the Main Semeai.

Black could also survive the Punishment Semeai, if he secured his second eye for his left-hand group by connecting at at the lower edge. But his score in the Semeai Variation would become worse, compared to Joachim's suggestion above.

:
White throws-in in the lower left corner.

This move pushes Black below the Punishment Semeai Liberty Level.

:
Black plays his forcing move in the lower left too late here.

Please take notice that I was a bit lazy with the presentation of these two exchanges in many subvariations, where I did not display these four moves explicitly. Simply because I took them for granted, as an integral part of the Main Semeai.
Additionally, I took things much too easy in that I did not consider, whether a White tenuki at the left (i.e. "Dany's Sacrifice") would push Black's large group in the upper right below the Punishment Semeai Liberty Level, as it is the case here.

: ( A 648ot; B 648ou)
We chose White's connection here, but - after Black lost a decisive liberty in the Main Semeai by connecting at - this is a grave mistake.
White loses two Main Semeai Liberties by this move.

White will not connect, but choose Dany's Sacrifice, instead. However, she must be careful with the continuation.
It is correct for her to continue with the Main Semeai, e.g. with in the upper right.
Blocking at the upper edge, with , would be a grave mistake on White's side.

We ask for your kind understanding that in the late phase of our elaboration, in which we discovered this error, we felt no longer able to reorganize the variation tree accordingly.

: (137 648ov)
Black might change the order of and .

: (A 648ow)
Black captures White's single stone at the upper edge.

Simply continuing with the Main Semeai, by occupying one of White's liberties at the left, will worsen Black's score in the Semeai Variation only.

: (B 648ox)
White blocks.

Here, she might also continue with the Main Semeai, instead, e.g. with .
Please be aware that this might not be true in other positions on the board, where Black might be able to spend a surplus Main Semeai Liberty.

: (A 648oy)
It is best for Black to connect.
Black loses one Main Semeai Liberty by this move.

If Black does without connecting, White will not capture at "early", but wait until after her decision point, so improving her score in the Semeai Variation by two points.

:
White continues with the mutual occupation of opponent's liberties in the Main Semeai.

:
Usually, we will not explicitly display the rest of the Main Semeai's liberty occupying sequence.

:
Instead, in the final diagram of each subvariation, we will show Black's (in the Capture Variation only) territory-destroying moves in the vicinity of his large group in the upper right, usually starting with White's atari at Black's single stone in the lower right.

You will recognise that Black here neither answers White's atari locally, nor pushes at any of the points and . This implies that he is not able to do so; otherwise he would fall prey to the Punishment Semeai.

We will illustrate the number of these moves as
"(number of reductions played)/(number of possible reductions) reductions"; in this main line:

Black must not play any reduction (which is 0/0 reduction).

Please note that connecting at counts twice, as this move costs Black two liberties, not only one (as and ).

Black will often not use his potential to full capacity. This will be wise in those cases, where his score in the Capture Variation is better than in the Semeai Variation, and might result in a reduction of the overall result of the game. Any of these moves left unplayed provides White with an advantage of one point in the Capture Variation (i.e. one point of territory), but will cost her one point in the Semeai Variation (i.e. one prisoner).

At the end of the previous diagram, it is Black to move. Therefore, we can calculate the relationship of liberties in the Main Semeai.

Black needs seven moves at the left to reach White's decision point (she will have to choose between the Capture and the Semeai Variation):

• White has four external liberties left ().
• White will not answer three of Black's moves inside her nakade (splus, lplus).

Black's group at the right has 14 liberties left:

• There are 11 external liberties ().
• Additionally (here due to his potential eye) we have to count three approach-move liberties ().

This means that Black is ahead by seven liberties, barely enough to not suffer from the Punishment Semeai.

We use our 2019 Solution for comparison.

Capture Variation:

• Black captured one White stone at the top edge.
• Black destroyed one point of White territory at the top.
• Black gained one point of territory in the left centre.
• Black's cumulated loss in the lower left corner is two points.

On the left side of the board, Black gained one point in total.

• White gained two points by capturing Black's single stone in the lower right.

On the right side of the board, White gained two points.

In total, White gained one point.

White + 1 => White 2

Semeai Variation:

• Black captured one White stone at the top edge.
  
• Black destroyed one point of White territory at the top.
• Black gained one point of territory in the left centre.
• Black's cumulated loss in the lower left corner is two points.
• Black captured one additional White stone for a gain of two points.

On the left side of the board, Black gained three points in total.

• Black gained two points by not connecting his single stone.

On the right side of the board, Black gained two points.

In total, Black gained five points.

White + 1 => Black + 4