Life In 19x19

Jika wrote:
I thought (reading) like this

$$Bc
$$ +------------
$$ | 2 3 1 O X .
$$ | . . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | 2 3 1 O X .
$$ | . . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

But on my board this at it's best creates seki.

What about at B18 ?

Maybe this?

$$Bc
$$ +------------
$$ | . 2 . O X .
$$ | 1 . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | . 2 . O X .
$$ | 1 . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

What about at B18?

Bill Spight wrote:
Sorry for being obscure.

$$Bc Approach ko
$$ +------------
$$ | . 3 9 O X .
$$ | 1 2 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc Approach ko
$$ +------------
$$ | . 3 9 O X .
$$ | 1 2 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

takes ko, plays ko threat, answers ko threat, takes ko back, plays elsewhere (perhaps a ko threat that Black ignores)

makes a direct ko.

----

I hope that is clearer.

It isn't, but that's my lack of knowledge:
I know the ko-rule, and I think a ko-threat is when someone creates a position that (if not connecting etc) will lead to a ko.
But I'm not familiar with the whole terms of ko-language.
so I don't get what your exlanation means

Could you make me a full-sgf of that?? Please?

And does that mean that while the outer stones are considered worry-free, it is not generally assumed in a tsumego that there are no other ko situations on the board?

jlt wrote:

Jika wrote:
I thought (reading) like this

$$Bc
$$ +------------
$$ | 2 3 1 O X .
$$ | . . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | 2 3 1 O X .
$$ | . . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

But on my board this at it's best creates seki.

What about at B18 ?

Uhm, yes... black kills white. Thanks!

Maybe this?

$$Bc
$$ +------------
$$ | . 2 . O X .
$$ | 1 . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | . 2 . O X .
$$ | 1 . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

What about at B18?

Seki??

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | . 2 4 O X .
$$ | 1 3 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

Jika wrote: Seki??

$$Bc
$$ +------------
$$ | . 2 4 O X .
$$ | 1 3 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | . 2 4 O X .
$$ | 1 3 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | 6 2 4 O X .
$$ | 1 3 5 O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | O O O O X .
$$ | . 7 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

Dead.

Jika wrote:

Bill Spight wrote:
Sorry for being obscure.

$$Bc Approach ko
$$ +------------
$$ | . 3 9 O X .
$$ | 1 2 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc Approach ko
$$ +------------
$$ | . 3 9 O X .
$$ | 1 2 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

takes ko, plays ko threat, answers ko threat, takes ko back, plays elsewhere (perhaps a ko threat that Black ignores)

makes a direct ko.

----

I hope that is clearer.

It isn't, but that's my lack of knowledge:
I know the ko-rule, and I think a ko-threat is when someone creates a position that (if not connecting etc) will lead to a ko.

OK, so you know that the ko rule prevents the immediate recapture of a ko. Suppose that you want to recapture the ko, which will usually be the case. The if you can make a play that your opponent will answer instead of winning the ko, after she answers you can take the ko back. That play is called a ko threat. Then your opponent might make a ko threat that you must answer, and then she retakes the ko. This is called a ko fight, in which one or both players make ko threats in order to retake a ko. The ko fight ends when one player wins the ko, either because the opponent failed to make a ko threat, or because the player ignores a threat.

The following SGF file shows a part of the fight for the approach ko. Just the two corners are relevant. It is not intended to show the full board.

And does that mean that while the outer stones are considered worry-free, it is not generally assumed in a tsumego that there are no other ko situations on the board?

In tsumego a ko for the life of a player's group is considered to be worse for that player than living and better than dying. Since seki is a kind of life, a ko is treated as worse, but in real life it may be better than seki, depending on the ko threat situation.

In modern tsumego the rest of the board is generally unknown, but escaping is one way to live. In some ancient problems the play could go anywhere.

BTW:

Click Here To Show Diagram Code

[go]$$Bc Black to play
$$ +------------
$$ | 2 3 1 O X .
$$ | . 4 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

makes ko.

Thank you!

Obviously I did not think far enough to see that sacrificing those stones is worth it.

jlt wrote:

Jika wrote: Seki??

$$Bc
$$ +------------
$$ | . 2 4 O X .
$$ | 1 3 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | . 2 4 O X .
$$ | 1 3 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

$$Bc
$$ +------------
$$ | 6 2 4 O X .
$$ | 1 3 5 O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | 6 2 4 O X .
$$ | 1 3 5 O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

$$Bc
$$ +------------
$$ | O O O O X .
$$ | . 7 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .

Click Here To Show Diagram Code

[go]$$Bc
$$ +------------
$$ | O O O O X .
$$ | . 7 . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

Dead.

Plus, I thought in this, black would be lost (square 4), if white takes A18.
Didn't think of black's throw in-atari.
Obviously, sacrificing stones for a bigger goal is not on my mind (yet, I hope).

Click Here To Show Diagram Code

[go]$$Bc Black to play
$$ +------------
$$ | 2 . 1 O X .
$$ | . . . O X .
$$ | O O O O X .
$$ | X X X X X .
$$ | . . . . . .[/go]

@Bill:

Thank you for making full-SGF, I overlooked (and hopefully learned) much there!

Thanks also for explaining the ko terminology in such simple words.
I had heard of a move the opponent has to answer, but did not know this was a ko-threat (I thought a ko-threat was threatening to make a ko situation).

Jika wrote:@Bill:

Thank you for making full-SGF, I overlooked (and hopefully learned) much there!

Thanks also for explaining the ko terminology in such simple words.
I had heard of a move the opponent has to answer, but did not know this was a ko-threat (I thought a ko-threat was threatening to make a ko situation).

I misspoke slightly. A player can ignore a ko threat, and sometimes that's the right thing to do.

So, what is exactly our community's distinction on reading v. visualization?

I definitely "see" the stones on the board when I am reading through variations.. but only up to a certain point. Once I'm reading deeper than 15 stones or so, and it becomes more global, I lose focus on "seeing" the stones, and the reading becomes more logical based on where I know stones should be, knowing where I should play globally based on what I expect the context of the board to be. For example, say I read a variation where I get a corner and my opponent gets the outside, then I won't focus on memorizing the exact stone placements, but understanding that in that variation the opponent has 'thickness' and visualize another pattern in a different area of the board. Is this sort of the distinction we have come up with?

On a side note, how deep can you guys read? I'm around 1D and I have been able to count a little over 20 moves deep in a large life and death fight expanding from a corner into the side/center. Makes me wonder what our resident 3/4/5D and higher can do.. and by extension a pro.

Hades12 wrote:I'm around 1D and I have been able to count a little over 20 moves deep in a large life and death fight.

I am 5k on some servers (KGS, Fox) and 8k EGF, and I've never been able to read more than 10 moves deep (except special situations like ladders). I regularly misread move 2. No wonder I am so far from dan level.

Hades12 wrote: On a side note, how deep can you guys read? I'm around 1D and I have been able to count a little over 20 moves deep in a large life and death fight expanding from a corner into the side/center. Makes me wonder what our resident 3/4/5D and higher can do.. and by extension a pro.

Here's a 1 hour online tournament game I annotated with some of my in-game reading: http://eidogo.com/#300G4navp. I'd recommend Lee Sedol's commented games books to see what a top pro reads during a game. Accuracy in breadth is generally more important than high-depth (though for something like a fight ending with semeai you sometimes need the big depth too): no point reading 40 moves deep when you missed opponent's good move 4. Pros are not immune to such reading failures, I've seen plenty of commentaries with things like "I didn't expect that reply" ie didn't read enough choices for move 2.

jlt wrote:

Hades12 wrote:I'm around 1D and I have been able to count a little over 20 moves deep in a large life and death fight.

I am 5k on some servers (KGS, Fox) and 8k EGF, and I've never been able to read more than 10 moves deep (except special situations like ladders). I regularly misread move 2. No wonder I am so far from dan level.

I try not to get in a situation where I HAVE to read that deep. Most of the time if you get that far into reading, you are in some type of trouble. It's at the point where win or lose that fight, and you win or lose the game. If you watch dwyrin's YouTube basics, you will see that he focuses a lot on direction of play/sente and he can regular dispose of mid-level dans easily. I would say you can get to dan level only being able to read five-ten moves deep, as long as you are perfectly correct in your reading.

Uberdude wrote:

Hades12 wrote: On a side note, how deep can you guys read? I'm around 1D and I have been able to count a little over 20 moves deep in a large life and death fight expanding from a corner into the side/center. Makes me wonder what our resident 3/4/5D and higher can do.. and by extension a pro.

Here's a 1 hour online tournament game I annotated with some of my in-game reading: http://eidogo.com/#300G4navp. I'd recommend Lee Sedol's commented games books to see what a top pro reads during a game. Accuracy in breadth is generally more important than high-depth (though for something like a fight ending with semeai you sometimes need the big depth too): no point reading 40 moves deep when you missed opponent's good move 4. Pros are not immune to such reading failures, I've seen plenty of commentaries with things like "I didn't expect that reply" ie didn't read enough choices for move 2.

Thank you for posting this game. I appreciated your comments and it was insightful. It encourages me to continue to work on reading seriously/playing slower games, rather than normal time settings and just playing based off of intuition/shape (which is my normal playing mentality).

IMO, reading depth and breadth are difficult to quantify objectively. The reason is because all moves are not equally likely to be correct. As a result, through training life and death problems, playing lots of games, and generally getting experience with a variety of shapes and situations, the initial selection of moves becomes better.

Someone who just learned the rules of go, having a strong capacity for holding variations in their mind, might very well be able to read a good number of moves wide and deep. But the moves they are choosing to read are unlikely to be very good, and it's unlikely that they'll end up with a good result, even if they read, say 15~20 sequences - they were just all the wrong sequences.

In contrast, a pro player has the intuition and instinct to play a most-likely-to-be-correct move almost instantly, without really "reading" deep or wide.

In some sense, a player's "reading" also includes the training they've already done, which allows them to choose better moves - and as a result, they don't have to really iterate as many sequences in their mind while they're actually playing a game in order to get to the right move.