Tesuji puzzles

Problem #4, White to play. There are several ways to live but one is better than others.

Capture d'écran 2024-04-02 083123


Answer to Problem 4: B18.

I’ll continue with a simple problem.
Problem #5, White to play. What is the shape move?

Capture d'écran 2024-04-07 213641


Ooh, pretty problem, nice thread in general. C11 ? ^^



1 Like

Lower right in the original 4, I actually wonder if trying to connect is not best as it’s gote and still leaves a ko, whereas simple “wrong” capture is clean and sente.

1 Like

@Uberdude You mean this ko?

Capture d'écran 2024-04-08 102622

But in this variation Black captures L2 anyway, and White would need to take gote in order to live by ko which is picnic for Black?

1 Like

Yes, but it’s not such a picnic: if white wins the ko looks like n3 next kills.

1 Like

Was wondering about this also, but I think n3 doesn’t kill, unless I’m missing a way to win the ko which doesn’t leave k1 as self-atari for white ? (White needs to be able to play k1 also to kill the group)

N3 M3 lives

1 Like

N3 N2 also ^^

It seems like both responses work unless white has a way to finish the ko which allows them to play K1 to remove the L2 eye ^^

(M3 and N2 are miai for the N1 eye.)

So that said, it seems the ko is gote, or requires a further move in gote after setting it up, for white to win.

ApparentIy I did not take sufficient care with that puzzle.
I guess I should have given black’s G3 stones one more liberty, like by moving white E4 one spot to the left.

1 Like

Hi ^^

I know you’ve probably seen the tesuji now, but just responding for future reference or other readers of the thread who are curious about the ko sequences – oh; I understand now what you were reading, and E1 for B to tie up W’s liberties isn’t apparent at first glance ^^

(either as a ko threat or after W connects at H1 – G2 F1 E1 or G2 F1, ko threat, w ignores and fills H1, B throws in at E1 in sente)

I intuitively thought W may be able to threaten to kill with the ko first at first glance too, it almost seems like it from the first-line connections everywhere.

It’s an interesting tsumego in its own right !^^

One of us must be reading a wrong sequence. I can’t follow what you are talking about.

Capture d'écran 2024-04-09 081215


I believe Uberdude was reading this extension at f1 potentially allowing White to finish the ko in sente (by threatening a follow-up at n3, then filling k1 to take the second eye away after black responds to n3).

But black can throw in to keep white’s liberties tied up so that k1 would be self-atari, like this :

Or like this in sente, after the ko is won by white (move 3 elsewhere as a ko threat) :

So finishing the ko isn’t sente for white to kill the lower right, thanks to the tesuji at e1 to tie up white’s g1 group liberties, though it seems possibly so at first glance.


Even when I fix that liberty mistake, I’d say there remains another issue.

I should have made black’s G3 group bigger so that saving those in gote with J2 would be more clearly better than sacrificing those in sente with K2 when black has a surplus of big ko threats.

Black keeping sente:

I suppose that a good quality problem would not have such ambiguities.

1 Like

After G2 F1, black can just K1 and there’s no ko! :slight_smile:


Yes, that may be the case for a problem aimed at DDK of that level ^^

On the other hand, exploring and seeing ambiguities and different applications to different cases in positions can help us get stronger ^^

(It was fun to explore them in any case ^^)

Ah, right ^^ I forgot to recheck that. So it isn’t working at all for w to not tie up b’s liberties there. :grin:

Gross hallucination on the part of two of us ^^

1 Like

I meant that instead of G2, white can capture at K3 for a big ko, so black can only play K2 & H2 when he has a surplus of big ko threats:

Edit: Oh, you were talking about a different variation with G2. Sorry.

1 Like

Problem #6, Black to play. Find the next three moves.


1 Like