The last round of this year’s Paris tournament was super dramatic, and this tsumego basically decided my game:
White to live with the bottom group.
Before you scroll down to the solution, here is a little backstory:
As described in this article, there was a lot at stake in this game. The winner would get a better placement in the tournament, (with the loser finishing around the bubble,) a better ranking for Bonus Points, of course GoR as well, but most importantly, the winner would most likely get one of the two seats for the Grand Prix Final. We almost also counted this game for the EGF Academy, but decided against it because there was already enough pressure:>
Prior to this game, our personal score favoured my opponent at 2:3. However, I won the most recent two games in a row, so it was hard to predict the result. Anyway, the mentality of the players counts a lot. People make bad decisions under pressure, so it’s about who can make the less serious mistakes.
A quick summary how we arrived to the tsumego position:
I tried to sabaki inside Black’s moyo. My opponent’s previous sequences seemed dubious to me, so I thought it would be super easy, but I was proven wrong…
Black played hane and two ataris, then spent almost his entire basic time on the push at 8.
After that, he had little time left, so he basically played the rest of the game in byo-yomi.
I have no idea what would happen if W played at A now…
…but since my opponent spent close to an hour reading it, I simply gave him credit and let him capture two stones. (He told me afterwards I would have died. Phew~)
Tsumego-wise, connecting at 13 was not necessary, but I didn’t want to let Black come back here because then his sequences would have turned out okay.
This was a misjudgement.
White’s life-and-death is not obvious at all. I have no idea if either player had better alternatives up to 30.
In any case, I spent a large portion of my basic time on my next move, and I am sure that after that Black has no way to kill White anymore. 26 was the last chance to change something in the tsumego, and 28 to let White off the hook and continue the game peacefully.
Last chance to scroll back up because here comes the solution:
Blocking at 1 is the first instinct. There is something to do against Black’s group inside.
You can still scroll up and try to find it.
The combo is descending at 5 and attaching at 7, the key move.
I forgot why I marked the stone at A, but the sequence does not work in this case. Black can destroy the eye shape at 8.
Whichever variation White picks now, White is losing by one liberty.
There is nothing White can do… Unless White had a stone at A or B.
Edit: …or unless you are cheater 1p (see below).
(Game) That is why White has to start with this hane.
This was such a decisive moment in the game… Had White chosen the other move, everything would have turned out differently.
Black (probably) cannot disconnect White in this way. Black’s outside group does not have the best shape to create liberties, so White would (probably) win this semeai.
(Game) That is why Black prevents White’s underconnection by blocking at 2 immediately. The atari at 6, giving White a ko shape, seems to be the only way to destroy this eye. (White’s descent to the left of 7 is sente against Black’s group.)
(Game) White makes Black give in at 12 using one local ko threat. The previously shown sequence comes into play.
This time White has one of the needed first-line stones for the semeai.
Thus, if Black does away with the eye shape at 1, the semeai turns into a ko. White throws in at 2, Black also eventually connects there at move 9. (Note that White does not need to capture the whole Black group, but half of it suffices to make a big eye.)
It is very fortunate that White is able throw in at 10 immediately. When Black takes at A, White can use the throw-in at B as a local ko threat.
(Game) Therefore, Black takes the stone and White gets all the sente moves by threatening Black’s own safety.
However, the story is not finished yet. 23 does not make a clean eye. Black connects at 24 and there are two spots White needs to fill to make a shape that is not a false eye.
Luckily, Black is unable to poke in at 1. White would atari at 2 and block at 4 in sente to kill Black’s corner at 6.
You can try to read out the corner. According to my calculations, it is dead unconditionally and without many liberties.
It is also notable that every move in the game sequence was necessary. If White had played at 1 without exchanging the attachment at 2, Black would have answered happily. In this case, White’s centre eye is false for Black’s corner is alive thanks to the connection at A.
Obviously, I am very proud of myself to have been able to read that far, despite all the pressure. Of course, the reading started already a few moves before the presented tsumego, in order to have the courage to connect at 13 in the initial diagrammes. Though I am still blurry about my peep at 19 and Black’s push at 26, making nakade at 28 most likely sealed the game.
After winning this game, it took some time for me to recover from the tension, but I’m sure my opponent was hit by a blow that was greater by an order of magnitude:S
Edit: Turns out White would have survived in the other variation as well, thanks the magic move at 9. After Black takes at A and White connects at B, it will be about the ko at C for which White has sufficient ko threats. Allegedly, cheater 1p saw it at first glance:p
Most likely though, I would never have found this move in the game and died.