Life In 19x19
http://www.lifein19x19.com/

two question about 4-4 point approach
http://www.lifein19x19.com/viewtopic.php?f=4&t=16877
Page 1 of 1

Author:  Fllecha [ Sat Aug 24, 2019 2:36 pm ]
Post subject:  two question about 4-4 point approach

Hi all.

I am struggling about grasping the ideas behind this two related topic about 4-4 point joseki. I'm 4k btw.

After LeelaZero now this approach it's pattern, which usually leads to a pushing battle (according to my analysis with bots) that leaves me kinda confused.



Isn't simpler the following josekis, which lead to simpler position for white, without pushing battle and stuff like that? Why this isn't pattern anymore? Why the attachment seems more "urgent" than the slide (locally of course)




Second question. The famous 3-3 invasion.

With LeelaZero help we play this fuseki


The bots says that here black is perfectly placed and is around 56% winning chances. Since black is around 43% on the first move that means that black having "sente" in the final position (and having few points on the corner) this is more than sufficient for having a +13% improvement even if white has nice territory at the top and komi. It seems to me that white position at the top is big and hardly invadable. Ok M17 is an invasion possibility but after the usual joseki white is very thick I think and how to reduce that? And in the top left under the wall that is pure 20 points territory.... it seems to me too big.

Here what is black's usual plan against that position on the top left? How Black usually attacks the wall that white build after he faces the 3-3 invasion? Can you please show me some plans against the wall of the 3-3 invasion or here to study it?

Thanks.

Author:  Kirby [ Sat Aug 24, 2019 4:20 pm ]
Post subject:  Re: two question about 4-4 point approach

In the second position, I'm not all that worried about white's points on the top, yet. What does he have in terms of definite points? You already mentioned the black invasion possibility, so after that, white has like 10 or 15 points of definite territory, maybe. Yeah, white has thickness then, but it's give and take... If black doesn't invade, white has 25 or 30 points maybe, with 10 stones to do it. So on average, getting 2.5 or 3 points per move, maybe... I'm not too worried about it.

For the first position, white is playing tighter - has the option of getting the corner by capturing R3, too.

That being said, I don't think you have to disregard your idea from the second sgf. If you're playing something reasonable, especially in the opening, I think it's better to play to get into a position you're comfortable with than to strictly follow the AI. Like before AI, if pros were playing some complicated avalanche variation all of the time, it didn't mean that we had to follow that trend.

AIs can give you new ideas, and think about interesting possibilities.

But at the end of the day, you are the commander of your stones - so I think you should play in a way that you feel comfortable with.

Author:  ez4u [ Sat Aug 24, 2019 6:42 pm ]
Post subject:  Re: two question about 4-4 point approach

Fllecha wrote:
Hi all.

I am struggling about grasping the ideas behind this two related topic about 4-4 point joseki. I'm 4k btw.

...
Second question. The famous 3-3 invasion.

With LeelaZero help we play this fuseki


The bots says that here black is perfectly placed and is around 56% winning chances. Since black is around 43% on the first move that means that black having "sente" in the final position (and having few points on the corner) this is more than sufficient for having a +13% improvement even if white has nice territory at the top and komi. It seems to me that white position at the top is big and hardly invadable. Ok M17 is an invasion possibility but after the usual joseki white is very thick I think and how to reduce that? And in the top left under the wall that is pure 20 points territory.... it seems to me too big.

Here what is black's usual plan against that position on the top left? How Black usually attacks the wall that white build after he faces the 3-3 invasion? Can you please show me some plans against the wall of the 3-3 invasion or here to study it?

Thanks.


If you used LZ to create this fuseki, you should have some feeling already for where your thinking is going wrong. White should continue extending at :w3: below instead of playing "a" in your example. LZ considers that "a" loses 6% versus 3 and the game is instantly equal when Black grabs the key point with a play at 3. In old-school terms as well, "a" is too close to White's wall and it allows Black to "hane at the head of three stones". At 4k the rules to live by include:
* "Always allow your opponent to push from behind on the second line!"
* "Don't tenuki and allow your opponent to hane at the head of either your two stones ( :w1: below) or your three stones ( :w3: )."
By the time you get to 4d you will not have found many exceptions to these rules!


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


By the end of the example LZ considers :w1: below a -8% play. It is way too close to the wall (now something like seven stones tall but with lots of defects. LZ considers that White should just play away to "a" in the bottom right for example with a roughly equal game. :w1: does pop up frequently in LZ analyses - as a possible response to Black "b". LZ will often sacrifice the marked White stones as apparently it does not consider the effort to fix up the defects worth it!
Click Here To Show Diagram Code
[go]$$Wc
$$ +---------------------------------------+
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X O . . . . . . . . . . . . . . . . |
$$ | . . X O . . . . . . . . . O . . . . . |
$$ | . . X O . . 1 . . O . . . . . X . . . |
$$ | . X @ . . . . . . . . . . . . . . . . |
$$ | . X @ . b . . . . . . . . . . . X . . |
$$ | . X @ . . . . . . . . . . . . . . . . |
$$ | . . X @ . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . a . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+[/go]

Author:  Gomoto [ Sun Aug 25, 2019 3:07 am ]
Post subject:  Re: two question about 4-4 point approach

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


recent pro games, white to move:

move a: 1396
move b: 1179
move c: 360

I am always happy if white chooses c. You may still call move c joseki if YOU like to.

Author:  Uberdude [ Sun Aug 25, 2019 3:44 am ]
Post subject:  Re: two question about 4-4 point approach

Worth noting that the vast majority of those 360 instances of 'c' in recent games will be followed by the double hane rather than extend.

Page 1 of 1 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/