Tsumego problem from one of my recent games

xela · #1

Click Here To Show Diagram Code: [go]$$Bc White to play $$ | . . . . . . . . $$ | . . O . . . . . $$ | . X O 1 . . . . $$ | . X X O . . . . $$ | . O O X X . . . $$ | . . O . . . . . $$ | . O X X X O . . $$ | . . O . . . . . $$ | . . . . . . . . $$ | . . . . . . . . $$ +----------------[/go]

White's best reply to :b1:

? (Yes, mistakes have been made by both sides up to this point. And I didn't get the right move in the game: many more mistakes were made by both sides.)

Edited to add: all ladders are in white's favour.

Bill Spight · #2

I would call this a "joseki" problem, not a tsumego problem, I suppose it arose like this.

Click Here To Show Diagram Code: [go]$$Bc Old joseki $$ | . . . . . . . . $$ | . . . , . . . . $$ | . . . . . . . . $$ | . . . . . . . . $$ | . . . 6 . . . . $$ | . . 2 5 7 . . . $$ | . . 8 . . . . . $$ | . . . 1 9 4 . . $$ | . a 0 . . . . . $$ | . . . . . . . . $$ | . . . . . . . . $$ +----------------[/go]

elsewhere

This is a double approach joseki. :w10:

is old fashioned. The modern (but pre-AI) play is the keima at a.

Click Here To Show Diagram Code: [go]$$Bcm11 Old joseki, continued $$ | . . . . . . . . $$ | . . . , . . . . $$ | . . 8 . . . . . $$ | . 7 4 9 . . . . $$ | . 5 3 W . . . . $$ | . 6 O X X . . . $$ | . . O . . . . . $$ | . 2 1 X X O . . $$ | . . O . . . . . $$ | . . . . . . . . $$ | . . . . . . . . $$ +----------------[/go]

I suppose the ladder works.

But I doubt if White can afford to abandon the :wc:

stone.

Click Here To Show Diagram Code: [go]$$Wcm20 This ladder $$ | . . . . . . . . $$ | . . b , . . . . $$ | . . O 2 . . . . $$ | . X O X . . . . $$ | . X X O 1 a . . $$ | . O O X X . . . $$ | . . O . . . . . $$ | . O X X X O . . $$ | . . O . . . . . $$ | . . . . . . . . $$ | . . . . . . . . $$ +----------------[/go]

is a well known tesuji. Actually, there are two ladders, aren't there?

EdLee · #3

Quote:

I suppose the ladder works.

Thanks, Bill.

The local study is interesting; also curious about the whole board evaluations.

Uberdude · #4

xela, I could imagine you made the wrong reply to black 3, but to black 1?!

xela · #5

Uberdude wrote:

xela, I could imagine you made the wrong reply to black 3, but to black 1?!

The curse of looking ahead: I could visualise the position after black 3, saw that the ladders were against me, but couldn't see the next move for white. So I didn't go down that path. Instead I went for a trade (there was interesting stuff happening in the top left corner too, but I can't remember enough of the game to be worth posting).

Bill Spight · #6

xela wrote:

Uberdude wrote:

xela, I could imagine you made the wrong reply to black 3, but to black 1?!

The curse of looking ahead: I could visualise the position after black 3, saw that the ladders were against me, but couldn't see the next move for white. So I didn't go down that path. Instead I went for a trade (there was interesting stuff happening in the top left corner too, but I can't remember enough of the game to be worth posting).

A couple of things. First, just because Black has the option of capturing some White stones is not a reason not to sacrifice them.

As you said, you made a trade. If you give up the stones on the left side, the five Black stones to the right are still cut off and attackable. As it happens, the damezumari is very bad for those stones. But if not, White could still attack them. Giving up the single :wc:

stone, however, unites all the Black stones. That stone is hard to sacrifice.

Second, something that I noted without reading is the vulnerability of the bamboo joint. If it is threatened, the threat and connection take away two dame. So in the position you show, the stones connected by a bamboo joint appear to have 6 liberties. But you have to subtract two for the connection, leaving 4 liberties (or allowing the cut). Wilcox says that stones with five dame are alive in a fight. That's not always so, but stones with only 4 dame are vulnerable. Extending from the :wc:

stone leaves them with 3 dame (or allows cut). Then the hane leaves them with 2 dame. Bingo!

Uberdude · #7

The shortage of liberties in this bamboo is very similar to this non-joseki variation punishment http://josekipedia.com/#path:pdqfqindpf ... ohphoioepe (though weirdly josekipedia doesn't continue with the ladder at q18 going for an unncessary compromise!), which explains why joseki is q15 pole connection even though it makes an empty triangle, and bamboo at p16 is worse.

xela · #8

Bill Spight wrote:

...something that I noted without reading is the vulnerability of the bamboo joint. If it is threatened, the threat and connection take away two dame. So in the position you show, the stones connected by a bamboo joint appear to have 6 liberties. But you have to subtract two for the connection, leaving 4 liberties...

This has been a blind spot for me (using the past tense because, now we've had this conversation, I'll never get it wrong again). I simply didn't see the black stones as attackable. I saw them as having plenty of liberties, with white F4 being blighted.

And here we have a nice case study: the AI told me I'd missed something interesting in the position, but the conversation with humans showed up the underlying gap in my thought process :-)

Uberdude · #9

Even if black didn't push to reduce his liberties and missed the hane for white 4, I'd still imagine trades like this:

Click Here To Show Diagram Code: [go]$$Bc $$ | . . 5 . . . . . $$ | . . O 3 . . . . $$ | . X O 1 . . . . $$ | . X X O 2 4 . . $$ | . O O X X 6 . . $$ | . . O . . . . . $$ | . . . X X O . . $$ | . . O . . . . . $$ | . . . . . . . . $$ | . . . . . . . . $$ +----------------[/go]

xela · **#10**

Here's another one. Again I didn't find the move in the game, but KataGo showed me an interesting way to connect.

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . . . . . . . | $$ | . X X X O O O . . O . . . 2 . X . . . | $$ | . . . . X X O . . . . 1 a . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

I wimped out and played :b1:

at a not realising I could jump a bit further. After :w2:

, how does black connect :b1:

to safety?

jlt · **#11**

My first thought:

xela · **#12**

Yes, that seems to work too, but KataGo showed me something much more stylish!

Uberdude · **#13**

It's easy to connect with the standard shape of kosumi or attach against the White knight move, though there may be better moves given White's thinness and Black's support above. But I would be more concerned with reading counters (or evaluating the natural trade) to this White resistance.

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . . . . . . . | $$ | . X X X O O O . . O . . . . 2 X . . . | $$ | . . . . X X O . . . . 1 . . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Uberdude · **#14**

Bill Spight · **#15**

xela wrote:

Here's another one. Again I didn't find the move in the game, but KataGo showed me an interesting way to connect.

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . . . . . . . | $$ | . X X X O O O . . O . . . 2 . X . . . | $$ | . . . . X X O . . . . 1 a . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

I wimped out and played :b1:

at a not realising I could jump a bit further.

Here is where Sonoda's theory of diagonal lines might have helped.

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . . . . . . . | $$ | . X X X O O O . . O . . 1 . . X . . . | $$ | . . . . X X O . . . . a . . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

It is obvious that :b1:

is OK, so why not a? Worth checking out.

Quote:

After

, how does black connect :b1:

to safety?

xela · **#16**

Uberdude got it. The first diagram behind the cut I thought was pretty, it's not a familiar shape for me. The second diagram is best play for both sides according to KataGo.

Bill Spight wrote:

Here is where Sonoda's theory of diagonal lines might have helped. :)

I can make a fair guess as to what the theory is based on your diagram, but do you want to tell us more?

xela · **#17**

Uberdude wrote:

It's easy to connect with the standard shape of kosumi or attach against the White knight move, though there may be better moves given White's thinness and Black's support above. But I would be more concerned with reading counters (or evaluating the natural trade) to this White resistance.

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . . . . . . . | $$ | . X X X O O O . . O . . . . 2 X . . . | $$ | . . . . X X O . . . . 1 . . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Yes, that really does get interesting!

My first instinct is the "crude" push and cut:

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . 1 O 4 . . . . . . | $$ | . X X X O O O . . O 2 3 . 5 O X . . . | $$ | . . . . X X O . . . . X . . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

I guess the white stones to the left are probably captured, but black gives up the lower right corner in return.

KataGo says the above diagram is not bad (third best of three moves, all within the margin of error), but its first choice move is the other push:

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . 3 2 O . . . 4 . . . | $$ | . X X X O O O . . O 1 . . . O X . . . | $$ | . . . . X X O . . . . X . . 5 6 . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Same sort of trade, but more complex (lots of variations to explore), larger scale (black loses more of the right side but gets more compensation at the bottom).

Second choice is this striking picture:

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . . . . . . . | $$ | . X X X O O O . . O . . 2 1 O X . . . | $$ | . . . . X X O . . . . X . . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Continuation:

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . 2 . 4 . . . | $$ | . X X X O O O . . O . . O X O X . . . | $$ | . . . . X X O . . . . X 1 3 5 . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

and then white tenuki (and I'm not going to show you what a horrible mess I made in the top half of the board...)

Uberdude · **#18**

First thing I'd consider is what if black takes the bait? Sometimes just taking the sente ponnuki can be good enough for losing the stone, and here it does solidify the 2-space shimari (which you normally don't want to attach against), and whilst I think black is not bad I think he could be even better (but more complicated so if leading big might choose this to simplify and win).

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . . . O . . 5 . . . . | $$ | . X X X O O O . . O . . 4 3 . X . . . | $$ | . . . . X X O . . . . X 6 2 1 . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

If white played 4 I was imagining the cleaner trade:

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . 3 2 O . . . 4 . . . | $$ | . X X X O O O . 5 O 1 . . . O X 6 . . | $$ | . . . . X X O . . . . X . . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

and there's also the question of what happens now? a? Looks like white is in more trouble than black.

Click Here To Show Diagram Code: [go]$$Bc $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . 3 2 O . . . . . . . | $$ | . X X X O O O . . O 1 . a . O X . . . | $$ | . . . . X X O . . 4 . X . . . . . . . | $$ | . . . X . O . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

maybe can even sacrifice 1 and ask if the whole thing is alive?

Click Here To Show Diagram Code: [go]$$Bcm5 $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . . O X . X . . . . X . . . X . . . | $$ | . O . . X . X . . X . . . . . . . . . | $$ | . X O O X . . . . X O O . . . . . . . | $$ | . X X X O O O . . O . 4 3 1 O X . . . | $$ | . . . . X X O . . O 6 X 2 5 . . . . . | $$ | . . . X . O . . . . . . 7 . . . . . . | $$ +---------------------------------------+[/go]

Bill Spight · **#19**

xela wrote:

Uberdude got it. The first diagram behind the cut I thought was pretty, it's not a familiar shape for me. The second diagram is best play for both sides according to KataGo.

Bill Spight wrote:

Here is where Sonoda's theory of diagonal lines might have helped.

I can make a fair guess as to what the theory is based on your diagram, but do you want to tell us more?

John Fairbairn has mentioned Sonoda's ideas a few times here. For instance,

https://lifein19x19.com/viewtopic.php?f ... da#p212148

https://lifein19x19.com/viewtopic.php?f ... da#p212148

where JF says this:

Quote:

Sonoda also has interesting things to say about tactics. One applicable here is that when you have more stones than your opponent you use them orthogonally and if you have fewer you use them diagonally (there is also separate advice to "cut the opponent's diagonals"). Hence the right move here for Black is N4.

I have a different Sonoda book where he talks about diagonally related plays for shinogi and sabaki. My guess is that is because diagonal relationships are more efficient for making eyes, although Sonoda does not say that.

xela · **#20**

Thanks Bill!

BTW, you posted the same link twice, but Googling for your quotation throws out this: https://lifein19x19.com/viewtopic.php?f=13&t=13754

Tsumego problem from one of my recent games

Who is online