Life In 19x19

Could someone please explain black's mistake at move 7 and how white should best respond?

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 4 5 . 2 . 3 . .
$$ | . . 6 1 . . . . . .
$$ | . . 7 . . . . . . .
$$ | . . . . . . . . . .[/go]

Apoah wrote:Could someone please explain black's mistake at move 7 and how white should best respond?

$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 4 5 . 2 . 3 . .
$$ | . . 6 1 . . . . . .
$$ | . . 7 . . . . . . .
$$ | . . . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 4 5 . 2 . 3 . .
$$ | . . 6 1 . . . . . .
$$ | . . 7 . . . . . . .
$$ | . . . . . . . . . .[/go]

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 2 3 . . . . . .
$$ | . . W B . W . B . .
$$ | . . W B . . . . . .
$$ | . . 1 4 . . . . . .
$$ | . . . . . . . . . .[/go]

First problem: Black is too aggressive. He can't descend if white descends, as this cut is too powerful. So black's original strategy of cutting white has failed, and black has holes in his position.

edit: I haven't shown why white 4 works, but I'll leave this as an exercise. As a hint, whatever black plays now, white should be able to capture something and connect 4 to the corner.

BUT:

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 2 4 . . . . . .
$$ | . . W B . W . B . .
$$ | . . W B . . . . . .
$$ | . . 1 3 . . . . . .
$$ | . . . . . . . . . .[/go]

This is fairly okay for black. His original plan was a bad one, but he has somewhat saved the day. This is quite livable for him, though there are usually better ways to accomplish the same thing (just blocking the other way in the first place, for instance). Black can still seal white in, so white has hardly destroyed black's position.

ALSO:

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 4 2 3 . . . . .
$$ | . . W B 5 W . B . .
$$ | . . W B . . . . . .
$$ | . 8 1 7 . . . . . .
$$ | . . 6 . . . . . . .[/go]

White's correct response is probably this way. It doesn't look like white has gained a lot from black's mistake, but actually he has, becuase black's push at 1 is clearly aji keshi - black didn't need or want to play here, and there might normally be a better play later. This way, white clearly makes a profit over the joseki line.

Black could also play 7 at 8 sometimes, I think, but this starts a difficult fight and is not generally a good idea.

People often think that the first way is a strong punishment for white, but as the second diagram shows, this is not really the case - black is fine if he realises his mistake. The best way for white is to calmly take advantage of black's small slip, and be happy to have removed aji later in the game.

Thanks! I spent some time trying to work this out before posting, but I don't know that I would have ever come up with those positions. This was very helpful!

Apoah wrote:Could someone please explain black's mistake at move 7 and how white should best respond?

$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 4 5 . 2 . 3 . .
$$ | . . 6 1 . . . . . .
$$ | . . 7 . . . . . . .
$$ | . . . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 4 5 . 2 . 3 . .
$$ | . . 6 1 . . . . . .
$$ | . . 7 . . . . . . .
$$ | . . . . . . . . . .[/go]

http://www.bantari.net/go/bloopers/page-002.php

Hope this helps.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 2 3 . . . . . .
$$ | . . W B . W . B . .
$$ | . . W B . . . . . .
$$ | . . 1 4 . . . . . .
$$ | . . . . . . . . . .[/go]

Is there something fundamental about this position that says 'black is screwed'? Or do you have to read it out to its end in the dozens of variations? If we look at it strictly in this position, without adding stones or doing reading, we see that white is occupying a shape point for black, and black has 4 liberties and 2 liberties and white has 4 liberties and 2 liberties. White has 4-8 points of territory in the corner. White is enclosed (weakly) in the corner, and black is sandwiched between a weak side white group and a weak central white stone. It's black's sente to move. From this description alone it sounds like a fairly even fight still. White is more stable, black has a bit of bad shape, but black has sente.

Like, if we move everything up and over a line, is this still a bad exchange for black?

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . 2 3 . . . . . .
$$ | . . . W B . W . B . .
$$ | . . . W B . . . . . .
$$ | . . . 1 4 . . . . . .
$$ | . . . . . . . . . . .[/go]

Just a note: I believe amnal's response is the current opinion. Bantari's discussion was considered to be current, but recently (well, comparatively recently) it was judged that the descent gave too good a result to black a lot of the time (amnal's post covers it).

@Numsgil: in your first diagram black can't get a good result - I (think I) can read out all variations and none work for black. If you move it diagonally I can't read out everything anymore (I think white is still fine, but not sure).

The reason is mostly due to moving one line towards the direction of black hane. If it is as below then black still can't get a good result as far as I can tell.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . 2 3 . . . . .
$$ | . . . W B . W . B .
$$ | . . . W B . . . . .
$$ | . . . 1 4 . . . . .
$$ | . . . . . . . . . .[/go]

Numsgil wrote:

$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 2 3 . . . . . .
$$ | . . W B . W . B . .
$$ | . . W B 6 . . . . .
$$ | . . 1 4 . . . . . .
$$ | . . 5 . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 2 3 . . . . . .
$$ | . . W B . W . B . .
$$ | . . W B 6 . . . . .
$$ | . . 1 4 . . . . . .
$$ | . . 5 . . . . . . .[/go]

Is there something fundamental about this position that says 'black is screwed'? Or do you have to read it out to its end in the dozens of variations? If we look at it strictly in this position, without adding stones or doing reading, we see that white is occupying a shape point for black, and black has 4 liberties and 2 liberties and white has 4 liberties and 2 liberties. White has 4-8 points of territory in the corner. White is enclosed (weakly) in the corner, and black is sandwiched between a weak side white group and a weak central white stone. It's black's sente to move. From this description alone it sounds like a fairly even fight still. White is more stable, black has a bit of bad shape, but black has sente.

No matter what black does, white will capture something - the fight isn't even in this sense. If black saves 1, white captures the 3 stones with the move I added at 6. If black saves the three, white clearly captures stone 1.

EDIT: Bantari's site has the sequences showing how black is doomed, there are not very many variations.

Like, if we move everything up and over a line, is this still a bad exchange for black?

$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . 2 3 . . . . . .
$$ | . . . W B . W . B . .
$$ | . . . W B . . . . . .
$$ | . . . 1 4 . . . . . .
$$ | . . . . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . 2 3 . . . . . .
$$ | . . . W B . W . B . .
$$ | . . . W B . . . . . .
$$ | . . . 1 4 . . . . . .
$$ | . . . . . . . . . . .[/go]

Now it is not obviously so bad for black, because white has no move to immediately capture something. For instance:

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . 2 3 . 5 . . . .
$$ | . . . W B . W . B . .
$$ | . . . W B . . . . . .
$$ | . . . 1 4 . . . . . .
$$ | . . . 6 . . . . . . .[/go]

This 6 doesn't smash 1 against the side in an easy way like before, so it isn't clear what will happen.



Also, apparently I've subsumed and put forward relatively new knowledge on the position. I'd like to stress that the important thing is knowing why the various moves are wrong, and that either 'punishment' is going to realistically be equally good in a real game.

@numsgil: black and white each have stones with two liberties, but only one of them requires an immediate defense in order to live. Not sure if that requires too much reading to be fundamental the way you're using the term.

Numsgil wrote:

$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 2 3 . . . . . .
$$ | . . W B . W . B . .
$$ | . . W B . . . . . .
$$ | . . 1 4 . . . . . .
$$ | . . . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . 2 3 . . . . . .
$$ | . . W B . W . B . .
$$ | . . W B . . . . . .
$$ | . . 1 4 . . . . . .
$$ | . . . . . . . . . .[/go]

Is there something fundamental about this position that says 'black is screwed'? ...

Almost. A 3-stone wall wants to go to 'a' or 'b' or nearby.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . X . . . . . .
$$ | . a . X . b . . . .
$$ | . . . X . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]

When white has 'a' completely out of reach with his three stones, black now wants 'b' twice as much:

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . W . . . .
$$ | . . O X . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]

...then the marked white stone screws black's chances of decent shape.

Okay, I see.

White can capture black's original hane in a Cathedral (also called an 'edge crush' in Wilcox's stuff. Not sure if it has other names). So black's stone is in a sort of 'virtual' atari.

Likewise, there's a similar edge pattern for black's 3 stones amnal pointed out.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . O . . . .
$$ | . . O X W . . . . .
$$ | . . X O . . . . . .
$$ | . . X . . . . . . .[/go]

So black's 3 stones are also in a sort of virtual atari also. If white gets to play the circled stone, black can't escape dying.

Since black can't defend both 'virtual' ataris at once, he's in a bad position. This is a complicated version of a double atari basically.

I don't mind thinking of the edge crush ('cathedral' feels like a weird term) as 'fundamental' because it turns up fairly often and it's an easy pattern to recognize and read out (though I admit I missed it in this context). Same with the 1st and 2nd line versions of the edge patterns. Like a ladder it doesn't branch into meaningful variations, so I don't consider it reading necessarily. It either works on the main line or it doesn't at all. And really the only 'gotcha' is to make sure that the 'roof' has at least 3 liberties. Otherwise, unless there's random stones thrown in, the result is a foregone conclusion.

Is the pattern that captures the 3 stone black string a similar recurring pattern? It's not immediately obvious (to me) that black is captured unless you read it out. When people see this, do they go 'ah, I've seen this before, black should die, let's see if it works here...' or do they go 'bleh, what is this? Can black escape? Let me read out a few lines of resistance'?

Depending on how others think about go this may or may not be a meaningful question But basically there are short patterns that I memorize with known outcomes. The edge crush is one, same as the other edge patterns on sensei's. Is there a similar common 4th line pattern? It seems to have similar characteristics.

Numsgil wrote:Okay, I see.

White can capture black's original hane in a Cathedral (also called an 'edge crush' in Wilcox's stuff. Not sure if it has other names). So black's stone is in a sort of 'virtual' atari.

Likewise, there's a similar edge pattern for black's 3 stones amnal pointed out.

$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . O . . . .
$$ | . . O X W . . . . .
$$ | . . X O . . . . . .
$$ | . . X . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . O . . . .
$$ | . . O X W . . . . .
$$ | . . X O . . . . . .
$$ | . . X . . . . . . .[/go]

So black's 3 stones are also in a sort of virtual atari also. If white gets to play the circled stone, black can't escape dying.

Since black can't defend both 'virtual' ataris at once, he's in a bad position. This is a complicated version of a double atari basically.

I don't mind thinking of the edge crush ('cathedral' feels like a weird term) as 'fundamental' because it turns up fairly often and it's an easy pattern to recognize and read out (though I admit I missed it in this context). Same with the 1st and 2nd line versions of the edge patterns. Like a ladder it doesn't branch into meaningful variations, so I don't consider it reading necessarily. It either works on the main line or it doesn't at all. And really the only 'gotcha' is to make sure that the 'roof' has at least 3 liberties. Otherwise, unless there's random stones thrown in, the result is a foregone conclusion.

Is the pattern that captures the 3 stone black string a similar recurring pattern? It's not immediately obvious (to me) that black is captured unless you read it out. When people see this, do they go 'ah, I've seen this before, black should die, let's see if it works here...' or do they go 'bleh, what is this? Can black escape? Let me read out a few lines of resistance'?

Depending on how others think about go this may or may not be a meaningful question But basically there are short patterns that I memorize with known outcomes. The edge crush is one, same as the other edge patterns on sensei's. Is there a similar common 4th line pattern? It seems to have similar characteristics.

There is no similar 4th line pattern, in that it is much harder to say 'black's stone cannot escape. But it may be possible to kill it.

As an example, there is this well known joseki:

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . 1 8 2 . . 3 . . .
$$ | . . . . 5 6 . . . . .
$$ | . . . . 7 . . . . . .
$$ | . . . 4 . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . 6 2 . . . . . . . .
$$ | . . B W W 7 . B . . .
$$ | . . 3 . B W . . . . .
$$ | . . 1 . B 4 . . . . .
$$ | . . . W 5 . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

Here, white's 6 depends upon a ladder. If you don't know this sequence, it's interesting to look it up.

Black's 7 is a complex way to maximise white's aji. It is clearly an example of the 4th line smash down, if such a thing exists. Here, it is clearly heavily dependent on the surrounding conditions, but this is my point - it always is.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . . 5 . . . . . .
$$ | . W W . . 3 4 . . . .
$$ | . . B W W 1 2 B . . .
$$ | . . B . B W . . . . .
$$ | . . B . B W . . . . .
$$ | . . . W B . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

5 is a tesuji to maximise liberties. It is often seen in a 4th line smash.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . 5 1 3 4 . 8 . .
$$ | . W W 6 2 B W . . . .
$$ | . . B W W B W B . . .
$$ | . . B . B W 7 . . . .
$$ | . . B . B W . . . . .
$$ | . . . W B . . 9 . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

The sequence concludes something like this. If black has the ladder to the bottom right *or* middle left, he will now save the smashed-down stones by capturing white's two (can you see why?). If he doesn't, white captures his stones, but he gets some aji out of it.

Things to note are:
- White could only capture by leaving cutting points
- White could only capture because of the exact arrangement of neighbouring stones, it is far more dependent on these than the 3rd line version
- Black could spread out a long way on the second line before being captured. In the corner, this makes shortage of liberties for white's group here a real problem, such as in the positions suggested above where black might be able to slide into the corner or similar even if the smash down works..

So, if white is strong, in general he can have a good go at capturing black's cutting stone and there isn't much black can do about it. But it is not very clear cut like the 3rd line version.


EDIT: To use an example position from earlier in this thread:

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . . .
$$ | . . . x . . . . . . .
$$ | . x . W B . B . . . .
$$ | . . . W B . W . B . .
$$ | . 5 . W B . . . . . .
$$ | . . 3 1 W . . . . . .
$$ | . x 4 2 x . . . . . .[/go]

What do you think will happen now? Clearly there are several vital points (the xs) which could be important, and white cannot play them all. So there is a lot of reading to see if the capture will work, even in a fairly simple position.

Joaz Banbeck wrote:
Almost. A 3-stone wall wants to go to 'a' or 'b' or nearby.

$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . X . . . . . .
$$ | . a . X . b . . . .
$$ | . . . X . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . . X . . . . . .
$$ | . a . X . b . . . .
$$ | . . . X . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]

Really? How come?

When white has 'a' completely out of reach with his three stones, black now wants 'b' twice as much:

$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . W . . . .
$$ | . . O X . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . W . . . .
$$ | . . O X . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]

...then the marked white stone screws black's chances of decent shape.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . W . . . .
$$ | . . O X . . . . . .
$$ | . . . 1 . . . . . .
$$ | . . . . . . . . . .[/go]

Numsgil wrote:Is the pattern that captures the 3 stone black string a similar recurring pattern? It's not immediately obvious (to me) that black is captured unless you read it out. When people see this, do they go 'ah, I've seen this before, black should die, let's see if it works here...' or do they go 'bleh, what is this? Can black escape? Let me read out a few lines of resistance'?

In my experience, as one gets stronger, they are able to say the former to more and more shapes. For instance, to a total beginner, this shape can be rather baffling:

Click Here To Show Diagram Code

[go]$$B White to play.
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . X X O . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]

But then, pros can recognize very complicated (for us) life and death issues in a split second, and get it correct. So it all depends on your individual skill level / experience.

Click Here To Show Diagram Code

[go]$$B
$$ ----------------
$$ | . . . . . . . . . .
$$ | . . O X . . . . . .
$$ | . . O X . O . . . .
$$ | . . O X O . . . . .
$$ | . . X O . . . . . .
$$ | . . X . . . . . . .[/go]

For me, black's stones seem dead rather obviously, and I would only take a couple of seconds in a real game to declare black dead and look away. But to others, this position seems like something straight out of the Igo Hatsuyoron, and they probably won't be able to read it out no matter how much time you give them. I assume you are somewhere in the middle .

Dusk Eagle wrote: For me, black's stones seem dead rather obviously, and I would only take a couple of seconds in a real game to declare black dead and look away. But to others, this position seems like something straight out of the Igo Hatsuyoron, and they probably won't be able to read it out no matter how much time you give them. I assume you are somewhere in the middle .

Yeah...

Just to be very explicit, this is my exact thought process. I'm literally leaving nothing out, dead ends and all. Every stone and sequence is exactly as far as I read, and what I read. I'd be interested in how this compares to others' thought processes about this position. Specifically if stronger players read this out less far, or don't read out some of the later variations, or if there are other tricky plays to think through, and in what order they're thought through. (I guess this is a sort of stream-of-consciousness Malkovich for a specific position).

Well, I don't like in your first diagram, but I guess that's slightly off topic. I would say I read (past tense) all that out, but that it takes only a few seconds to see it all, while I imagine you spent quite a bit of time just getting the diagrams to work . The first time I saw the shape, I spent a bit longer looking at it, whereas if I were to see that in a game now I really don't even need to read anything to declare black dead (but I quickly read (present tense) it anyway).