Some weird tsumego

Ferran · Post by **Ferran** » Fri Sep 13, 2019 12:06 pm

I've been having a bit of a weird week, schedule-wise, so I've tried to use the time to restart my DDK tsumego. On one of the books I've found several quirky solutions, for my understanding. Now, I *know* I see some tsumego differently, that my answer "works" but is not always the preferred option. Sometimes I understand why, sometimes I don't. In this case, I *think* some of the answers are wrong.

If I may,

Click Here To Show Diagram Code: [go]$$ $$ ---------------------- $$ | . . . O X . X O . | $$ | . . . O X X . . O | $$ | . . , O O O X X X | $$ | . . . . . . O O O | $$ | . . . . . O . . . |[/go]

Click Here To Show Diagram Code: [go]$$c What is white's level? $$ ---------------------- $$ | . . . . O . . . . | $$ | . . O O X O . . . | $$ | O . X O X . O . . | $$ | O O O X X . . O . | $$ | X X X . . . . . . |[/go]

This one shows something that always bugs me of these things. White's level is alternatively about par with Black's or significantly lower.

I don't understand this kind of solution. 4 is one move away. If Black can see seven moves away, surely White can see a single one? Then, 4 at a, 5 captures at the marked stone. Should Black atari at b, then 6 at, well, 6 has already connected and, can play at 7 for a connection outside. Capturing a becomes gote. The specific example aside, there are many problems where Black can't be greedy, but White is nothing but. There's not "damage control", of sorts. And I'm not sure what part is me being DDK, what part is just the way things are done...

Another example of the same kind...

Click Here To Show Diagram Code: [go]$$c Another one $$ ---------------------- $$ | . . . . . X . . . | $$ | . O O O . O X . . | $$ | . . . X O . . X . | $$ | . O . X . O O X . | $$ | . . O . X X X . . |[/go]

Why would white ever play 2? It's self-atari, it makes no sense. It's one move away. 2 at 3, 3 at 2, 4 at a. Or even 2 at a, then pick whichever's left after black plays 3 at either 2 or 3. I don't know, but the point is that we wouldn't allow Black, in an exercise, to self-atari at the very first move, would we? As I understood tsumego, both sides are supposed to try the better move available. I understand a book for DDK can't ask for the moon, but this is looking a single move ahead... So, what am I missing? This is not the only book I've had the feeling with, it's just that the first example really puzzles me.

Thanks. Take care.

PS: Last question.. for now. Why does BBCode insist on numbering the lines of those diagrams with comments?

Cassandra · Post by **Cassandra** » Fri Sep 13, 2019 12:31 pm

Ferran wrote:
$$ Tsumego solution
$$ ----------------------
$$ | . . . . . X . . . |
$$ | . O O O a O X . . |
$$ | . . . X O 1 2 X . |
$$ | . O . X 3 O O X . |
$$ | . . O . X X X . . |

Click Here To Show Diagram Code
[go]$$ Tsumego solution $$ ---------------------- $$ | . . . . . X . . . | $$ | . O O O a O X . . | $$ | . . . X O 1 2 X . | $$ | . O . X 3 O O X . | $$ | . . O . X X X . . |[/go]

Why would white ever play 2? It's self-atari, it makes no sense. It's one move away. 2 at 3, 3 at 2, 4 at a. Or even 2 at a, then pick whichever's left after black plays 3 at either 2 or 3. I don't know, but the point is that we wouldn't allow Black, in an exercise, to self-atari at the very first move, would we? As I understood tsumego, both sides are supposed to try the better move available. I understand a book for DDK can't ask for the moon, but this is looking a single move ahead... So, what am I missing? This is not the only book I've had the feeling with, it's just that the first example really puzzles me.

Tsumego books show the strongest / hardest resistance in the solutions, not necessarily the "best" moves being played in a real game. Also, side effects on the endgame are not considered.

Click Here To Show Diagram Code: [go]$$ Real game sequence $$ ---------------------- $$ | . . . . . X . . . | $$ | . O O O 2 O X . . | $$ | . . . X O 1 . X . | $$ | . O . X 3 P P X . | $$ | . . O . X X X . . |[/go]

In a real game, as a matter of course, White would connect at 2 immediately. Thereafter, Black cuts at 3, and White's two marked stones are caught in a snapback.

Cassandra · Post by **Cassandra** » Fri Sep 13, 2019 12:36 pm

Ferran wrote:
$$c Seki!!!
$$ ----------------------
$$ | . . . O X . X W . |
$$ | . . . O X X 1 a W |
$$ | . . , O O O X X X |
$$ | . . . . . . O O O |
$$ | . . . . . O . . . |

Click Here To Show Diagram Code
[go]$$c Seki!!! $$ ---------------------- $$ | . . . O X . X W . | $$ | . . . O X X 1 a W | $$ | . . , O O O X X X | $$ | . . . . . . O O O | $$ | . . . . . O . . . |[/go]

The book claims this is a seki, 'a' being the unplayable point (there must be a name for that). I see it unconditionally alive. What am I missing?

Black cannot capture White's two marked stones in the corner, so these are alive, so the four-point area in the top right corner is NOT Black territory, so Black has only one eye. White cannot capture Black's group as well, so the position is seki.

lightvector · Post by **lightvector** » Fri Sep 13, 2019 6:17 pm

@Ferran - just to additionally chime in about some of the reason for why the problems are like this. When you present a problem, you sometimes have two (or more!) variations after the first correct move that are often (but not always) distinct, of the following nature:

1. Opponent backs down immediately if the tesuji truly works and says "okay you win, I will just limit the damage".

2. Opponent "resists" and forces you to prove that you know how to finish the fight and/or find all the remaining moves in the sequence accurately. Force you to prove that you know *why* the tesuji truly works. The opponent of course gets a worse result because they resisted and still died.

In a real game opponent should usually choose 1. But for testing that you can read accurately, most problem answers show 2, because knowing why the whole sequence works lets you test the accuracy of your deeper reading. Sometimes, you will find you get the first move right but get the later moves wrong, because you didn't truly understand why it worked (and without understanding, there is nothing stopping you from also trying the that first move even in situations where the first move in fact doesn't work!).

I do wish more problems also showed examples of 1 though. Sometimes it's challenging in its own right to figure out how to play on the receiving end of a tesuji. And it isn't always even a mistake to allow a tesuji against yourself if you needed to get sente elsewhere, so it does come up.

Bill Spight · Post by **Bill Spight** » Fri Sep 13, 2019 8:04 pm

Ferran wrote:I've been having a bit of a weird week, schedule-wise, so I've tried to use the time to restart my DDK tsumego. On one of the books I've found several quirky solutions, for my understanding. Now, I *know* I see some tsumego differently, that my answer "works" but is not always the preferred option. Sometimes I understand why, sometimes I don't. In this case, I *think* some of the answers are wrong.

My 2¢.

$$c Seki!?
$$ -------------------
$$ . . . O X . X O . |
$$ . . . O X X 1 a O |
$$ . . , O O O X X X |
$$ . . . . . . O O O |
$$ . . . . . O . . . |

Click Here To Show Diagram Code
[go]$$c Seki!? $$ ------------------- $$ . . . O X . X O . | $$ . . . O X X 1 a O | $$ . . , O O O X X X | $$ . . . . . . O O O | $$ . . . . . O . . . |[/go]

The book claims this is a seki, 'a' being the unplayable point (there must be a name for that). I see it unconditionally alive. What am I missing?

The "unplayable point" is called a shared dame or an internal dame.

Seki is unconditionally alive.

(Note: That is the traditional way of using the term unconditionally alive. In some contexts people use the term to mean that a group is alive even if its owner always passes. But in problems it means that a group is alive with correct play against any attack, except possibly in a ko fight.)

$$c What is white's level?
$$ --------------------
$$ | . . . . O . . . .
$$ | . . O O X O . . .
$$ | O . X O X . O . .
$$ | O O O X X . . O .
$$ | X X X . . . . . .

Click Here To Show Diagram Code
[go]$$c What is white's level? $$ -------------------- $$ | . . . . O . . . . $$ | . . O O X O . . . $$ | O . X O X . O . . $$ | O O O X X . . O . $$ | X X X . . . . . .[/go]

This one shows something that always bugs me of these things. White's level is alternatively about par with Black's or significantly lower.

I agree that that is a ridiculous question.

$$c 4 at the marked stone
$$ --------------------
$$ | b a 5 6 O 7 . . .
$$ | 1 3 O O X O . . .
$$ | O 2 B O X . O . .
$$ | O O O X X . . O .
$$ | X X X . . . . . .

Click Here To Show Diagram Code
[go]$$c 4 at the marked stone $$ -------------------- $$ | b a 5 6 O 7 . . . $$ | 1 3 O O X O . . . $$ | O 2 B O X . O . . $$ | O O O X X . . O . $$ | X X X . . . . . .[/go]

I don't understand this kind of solution. 4 is one move away. If Black can see seven moves away, surely White can see a single one? Then, 4 at a, 5 captures at the marked stone. Should Black atari at b, then 6 at, well, 6 has already connected and, can play at 7 for a connection outside. Capturing a becomes gote. The specific example aside, there are many problems where Black can't be greedy, but White is nothing but. There's not "damage control", of sorts. And I'm not sure what part is me being DDK, what part is just the way things are done...

You can't say that you have solved a problem unless you have an answer for every reply of the opponent. The sequence shown in the diagram is necessary to show that :b1:

solves the problem. So is the sequence where White plays :w2:

at 3, but that obviously gives up.

Unfortunately, the book asks not for Black to capture four stones, but how strong White is. Apparently it does not say that Black is to play and capture some White stones. So what is a solution? :scratch:

Another example of the same kind...

$$c Another one
$$ ----------------------
$$ | . . . . . X . . . |
$$ | . O O O . O X . . |
$$ | . . . X O . . X . |
$$ | . O . X . O O X . |
$$ | . . O . X X X . . |

Click Here To Show Diagram Code
[go]$$c Another one $$ ---------------------- $$ | . . . . . X . . . | $$ | . O O O . O X . . | $$ | . . . X O . . X . | $$ | . O . X . O O X . | $$ | . . O . X X X . . |[/go]

$$c
$$ ----------------------
$$ | . . . . . X . . . |
$$ | . O O O a O X . . |
$$ | . . . X O 1 2 X . |
$$ | . O . X 3 O O X . |
$$ | . . O . X X X . . |

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

Why would white ever play 2? It's self-atari, it makes no sense. It's one move away. 2 at 3, 3 at 2, 4 at a. Or even 2 at a, then pick whichever's left after black plays 3 at either 2 or 3. I don't know, but the point is that we wouldn't allow Black, in an exercise, to self-atari at the very first move, would we? As I understood tsumego, both sides are supposed to try the better move available. I understand a book for DDK can't ask for the moon, but this is looking a single move ahead... So, what am I missing? This is not the only book I've had the feeling with, it's just that the first example really puzzles me.

is not self atari. :b3:

is atari.

at 1 would be self-atari, AKA connect-and-die.

Again, the sequence given is necessary to show that :b1:

solves the problem. Also, :w2:

gives up no points locally. Here is the result of that sequence of play.

Click Here To Show Diagram Code: [go]$$c 3 White prisoners, 1 Black prisoner $$ ---------------------- $$ | . . . . . X . . . | $$ | . O O O W O X . . | $$ | . . . X O B . X . | $$ | . O . X B . . X . | $$ | . . O . X X X . . |[/go]

And here is the result when :w2:

simply connects.

Click Here To Show Diagram Code: [go]$$c Another one $$ ---------------------- $$ | . . . . . X . . . | $$ | . O O O W O X . . | $$ | . . . X O B . X . | $$ | . O . X B O O X . | $$ | . . O . X X X . . |[/go]

In the second diagram Black has captured 2 White stones in a snapback, for 5 pts. of territory. In the first diagram Black has captured 3 White stones and White has captured 1 Black stone. Adding the 3 pts. of territory on the board gives Black a net local score of 5 pts. (The difference is that White has lost a ko threat, but the local score is the same.

)

As I understood tsumego, both sides are supposed to try the better move available.

Only the solver is supposed to play correctly. The opponent can, and ideally should, try anything. It is up to the solver to counter whatever the opponent may throw at him.

Edit: OC, in a failure diagram, where the solver makes a mistake, the opponent is supposed to play correctly after the mistake, to show that the solver's play fails.

EdLee · Post by **EdLee** » Fri Sep 13, 2019 10:10 pm

Seki is unconditionally alive.

(Note: That is the traditional way of using the term unconditionally alive. In some contexts people use the term to mean that a group is alive even if its owner always passes. But in problems it means that a group is alive with correct play against any attack, except possibly in a ko fight.)

Hi Bill

Ferran · Post by **Ferran** » Fri Sep 13, 2019 11:19 pm

Cassandra wrote:Black cannot capture White's two marked stones in the corner, so these are alive, so the four-point area in the top right corner is NOT Black territory, so Black has only one eye. White cannot capture Black's group as well, so the position is seki.

I saw this last night. I was about to answer that I still didn't see it, then decided to wait if someone said something else... and it clicked. Argh. This shows I have two blindspots that spill into something else.

Thanks. Take care.

Ferran · Post by **Ferran** » Fri Sep 13, 2019 11:24 pm

Bill Spight wrote:Unfortunately, the book asks not for Black to capture four stones, but how strong White is. Apparently it does not say that Black is to play and capture some White stones. So what is a solution?

No, sorry, that question was mine. Don't blame it on the book.

Thanks. Take care.

Bill Spight · Post by **Bill Spight** » Sat Sep 14, 2019 12:51 am

Ferran wrote:
Bill Spight wrote:Unfortunately, the book asks not for Black to capture four stones, but how strong White is. Apparently it does not say that Black is to play and capture some White stones. So what is a solution?

No, sorry, that question was mine. Don't blame it on the book.

Thanks. Take care.

Oh, OK. Thanks.

jlt · Post by **jlt** » Sat Sep 14, 2019 12:54 am

I find the terminology "unconditionally alive" confusing. Black can be alive in several ways :

(a) Even if Black always passes.
(b) If both play alternatively, with White playing first.
(c) If both play alternatively, with Black playing first.
(d) If Black wins a local ko.

So which ones are "unconditionally alive"?

Click Here To Show Diagram Code: [go]$$B $$ +---------------------------+ $$ | a X a X O . O . O X c c c | $$ | X X X X O . O . O X X X X | $$ | O O O O O O O O O O O O O | $$ | X X X O . . . . . , . . . | $$ | b b O X . . . . . . . . . | $$ | O O O X . . . . . . . . . | $$ | X X X X . . , . . . . . . | $$ | a X a X . . . . . . . . . | $$ | X X X X . . . . . . . . . | $$ | . . . , . . . . . , O O O | $$ | O O O O O O O O O O O X X | $$ | X X X X X O . O O X X X O | $$ | b b b b X O O . O X d O d | $$ +---------------------------+[/go]

Bill Spight · Post by **Bill Spight** » Sat Sep 14, 2019 1:06 am

EdLee wrote:
Seki is unconditionally alive.

(Note: That is the traditional way of using the term unconditionally alive. In some contexts people use the term to mean that a group is alive even if its owner always passes. But in problems it means that a group is alive with correct play against any attack, except possibly in a ko fight.)
Hi Bill
That's an interesting historical ( "traditional" ) usage.

As a lay person I would've guessed:

alive = ( implied ) conditionally alive = "with correct play against any attack; i.e. it can still die collaterally in a ko"

unconditionally alive = "alive even if the owner always passes; regardless of ko (elsewhere)"

e.g. Even for this shape:
$$ shape (a)
$$ ---------------
$$ | . O . O X . .
$$ | O O O O X . .
$$ | X X X X X . .
$$ | . . . . . . .

Click Here To Show Diagram Code
[go]$$ shape (a) $$ --------------- $$ | . O . O X . . $$ | O O O O X . . $$ | X X X X X . . $$ | . . . . . . .[/go]
" is unconditionally alive" = an implicit assumption that won't kill themselves, i.e. "no incorrect plays."

As a practical matter ( not as a stickler of terminology ), I think of alive and seki as "locally, conditionally alive" and "locally, conditionally seki".

e.g. "alive; but killable in a ko"
$$ shape (b)
$$ ---------------
$$ | . . . . O X .
$$ | O O O O O X .
$$ | X X X X X X .
$$ | . . . . . . .

Click Here To Show Diagram Code
[go]$$ shape (b) $$ --------------- $$ | . . . . O X . $$ | O O O O O X . $$ | X X X X X X . $$ | . . . . . . .[/go]

If a problem says, 無条件生き, with or without the き, it means to live without ko. Outside of problems, I have only seen unconditional life used to mean alive even if the defender always passes, and only by Western writers.

Bill Spight · Post by **Bill Spight** » Sat Sep 14, 2019 1:20 am

jlt wrote:I find the terminology "unconditionally alive" confusing. Black can be alive in several ways :

(a) Even if Black always passes.
(b) If both play alternatively, with White playing first.
(c) If both play alternatively, with Black playing first.
(d) If Black wins a local ko.

So which ones are "unconditionally alive"?

$$B
$$ +---------------------------+
$$ | a X a X O . O . O X c c c |
$$ | X X X X O . O . O X X X X |
$$ | O O O O O O O O O O O O O |
$$ | X X X O . . . . . , . . . |
$$ | b b O X . . . . . . . . . |
$$ | O O O X . . . . . . . . . |
$$ | X X X X . . , . . . . . . |
$$ | a X a X . . . . . . . . . |
$$ | X X X X . . . . . . . . . |
$$ | . . . , . . . . . , O O O |
$$ | O O O O O O O O O O O X X |
$$ | X X X X X O . O O X X X O |
$$ | b b b b X O O . O X d O d |
$$ +---------------------------+

Click Here To Show Diagram Code
[go]$$B $$ +---------------------------+ $$ | a X a X O . O . O X c c c | $$ | X X X X O . O . O X X X X | $$ | O O O O O O O O O O O O O | $$ | X X X O . . . . . , . . . | $$ | b b O X . . . . . . . . . | $$ | O O O X . . . . . . . . . | $$ | X X X X . . , . . . . . . | $$ | a X a X . . . . . . . . . | $$ | X X X X . . . . . . . . . | $$ | . . . , . . . . . , O O O | $$ | O O O O O O O O O O O X X | $$ | X X X X X O . O O X X X O | $$ | b b b b X O O . O X d O d | $$ +---------------------------+[/go]

The a's are unconditionally alive. The b's could be answers to problems with the goal of making unconditional life.

hyperpape · Post by **hyperpape** » Thu Sep 19, 2019 6:28 am

Bill Spight wrote:(Note: That is the traditional way of using the term unconditionally alive. In some contexts people use the term to mean that a group is alive even if its owner always passes. But in problems it means that a group is alive with correct play against any attack, except possibly in a ko fight.)

Bill, I would have thought the lower left counts as unconditionally alive by that definition? Against any attack, black can live.

Kirby · Post by **Kirby** » Thu Sep 19, 2019 7:12 am

There are situations like a ko elsewhere on the board, where an “unconditionally alive” group is not actually “unconditionally” alive, since a player might skip a move to, eg. take a ko.

But this is more pedantic than practical.

In practice, knowing that a position is *locally* unconditionally alive with optimal play by both players is very valuable when reading out a local position.

The reason is, after reading out a branch of a local problem until you see that one side is “unconditionally alive”, you have identified a terminal state for that branch. Therefore, you no longer need to read that sequence, and can simply label that branch as “alive”. Then, you back up and explore a different branch. And you can propagate that information up the game tree until you have found the overall result of the local position “black is alive” or “black is dead”, etc.

It’s true that the global situation can have an impact, and a ko situation requires thought, too.

But I find it practical to ignore this nuance for reading a local situation, and read out the global situation separately. I.e. it’s obvious that ko, or even nearby groups can impact a local situation, but you can read this as it’s own subproblem.

Kirby · Post by **Kirby** » Thu Sep 19, 2019 7:17 am

hyperpape wrote:
Bill Spight wrote:(Note: That is the traditional way of using the term unconditionally alive. In some contexts people use the term to mean that a group is alive even if its owner always passes. But in problems it means that a group is alive with correct play against any attack, except possibly in a ko fight.)
Bill, I would have thought the lower left counts as unconditionally alive by that definition? Against any attack, black can live.

They are referring to, eg., a ko elsewhere on the board. White could make a ko threat on the bottom left group. If black answers, yep, still alive. But if black doesn’t answer, white can kill the group. For the ‘a’ groups, white can’t even make a ko threat.

I personally like the term “unconditionally alive” to be used for groups like the bottom left, taking for granted that a ko elsewhere could impact the local status.

Life In 19x19

Some weird tsumego

Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re:

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego

Re: Some weird tsumego