Life In 19x19

One comment:

Click Here To Show Diagram Code

[go]$$Bcm7 Looks strong, but
$$ -------------------
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ | . . . . . X . O . |
$$ | . . . . . . . . . |
$$ | . . . . X . O . . |
$$ | . . . . 1 . . . . |
$$ | . . . . W X . . . |
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ -------------------[/go]

looks strong, taking away a liberty of , but. . . .

Click Here To Show Diagram Code

[go]$$Bcm7 Leaves weakness behind
$$ -------------------
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ | . . . . . X . O . |
$$ | . . . . . . . . . |
$$ | . . . . X . O . . |
$$ | . . . . 1 a . . . |
$$ | . . . 2 W X . . . |
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ -------------------[/go]

strengthens the stone, and now Black has a cutting point at "a".

Click Here To Show Diagram Code

[go]$$Bcm7 Simple and strong
$$ -------------------
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ | . . . . . X . O . |
$$ | . . . . . . . . . |
$$ | . . . . B . O . . |
$$ | . . . . . 1 . . . |
$$ | . . . . W B . . . |
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ -------------------[/go]

here strengthens Black, connecting the stones, and weakens the stone, as well.

Click Here To Show Diagram Code

[go]$$Bcm7 Simple and strong
$$ -------------------
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ | . . . . . X . O . |
$$ | . . . . . . . . . |
$$ | . . 3 . B . O . . |
$$ | . . . . . 1 . . . |
$$ | . . 2 . W B . . . |
$$ | . . . . . . . . . |
$$ | . . . . . . . . . |
$$ -------------------[/go]

can strengthen the stone and stake out the bottom left corner, but Black can continue with, for instance, with no worries about his stones.

Hi George,

Supplement to post 16: Elephant's Eye.

Thanks everyone for your comments and suggestions.

I've been continuing with my correspondence games. They remain fairly slow, but they do give me time to really THINK about my moves.

I signed up on KGS as FuriousGrg but did not find many people on my level. I didn't have a tremendous amount of time so I played some beginner bots. It took a second for me to understand how it works, but I liked the room system. A lot easier to promote chatting while looking for games, seemed like people had known each other for a long time in there. I'll probably go back to KGS when I'm stronger.

I also bought a super-duper cheap set of (korean, I think?) glass stones from a local international grocery store. The stones are smaller than normal (about 18mm) but are more than sufficient... for now. A coworker and I have started playing correspondence 9x9 games at work. We don't have a real goban, so I made a 9x9 and printed it out on paper. We've gotten a lot of questions and interest, but no new Go players ... yet!

I finished 'Graded Go Problems for Beginners vol. 1' and I've started volume 2. I've really been enjoying the problems and I can tell it's really helping with my visualization. I found that I really prefer the paper format instead of the online tsumegos — it really forces me to do it all in my head — I can't fiddle with it until I find the answer.

I also downloaded the Leela engine and attached it to Sabaki. I've been poking at it and getting destroyed, but it's at least SEEMINGLY correctly punishing bad play -- whereas some of the other engines like GnuGo produce some weird looking moves. I don't intend for this to be my primary type of play, but Leela makes for a good on-demand strong opponent that I can suspend at any time. I'm mostly looking to it as a way of quickly practicing the opening.

I've attached a recent correspondence game. It wound up being close, but I made several mistakes and failed to keep the whole board in mind. My opponent started at 25K, but wound up as a 17K by the end (game took about 3 months, gah).

I struggle with where the big moves are and direction of play. There are several points where I think my moves were too small, or in the wrong direction.

I also think I mishandled the ko's. There weren't really any ko FIGHTS like I usually see referred to, they just kinda hung out for awhile. Especially that ko in the upper right, I lost several points there that could have tied up the game.

Takeaways:

- Keep thinking globally, not just locally.
- You need to have SOME sort of influence, not just territory.
- Think before you cut.

You need to have some sort of territory, not just influence.

Bill Spight wrote:You need to have some sort of territory, not just influence.

Well played.

Hi George,

( Re: posts 4 & 5. )

R4 is quite big.

Bad. The lower right corner is already W's --
you don't need to do anything there now.
Why did you think you needed ?
Please show an exact sequence, not just a "feeling".

The game move extend D17 is good;
the hane E17 variation is bad.
If W hanes E17, B would not turn at F17 in the variation ( W's fantasy );
B would cut at D17 ( the shared vital point ).
After B cuts at D17, B has miai to capture your E17 stone
or atari-push at D15.

About your notes here:
D17 extend is not passive; it's a very strong and big move.
E17 hane is bad shape, with weaknesses for W.

You need to reply to . W is happy to extend to D13.

No. B hanes at D13, the local shared vital point.
W cannot allow B to atari-push-through at D15, so W must reply.
Example: if W pulls back to C14, B can just connect at E13.
B has already benefited because you ignored .

Connect. Back in post 1:

The most difficult concepts for me were Ko and Ko fights

When we play a move, we want to have a reason for it
( at least one reason; the more good reasons, the better;
sometimes we have the wrong reasons, which we discover later ).
Why did you play ?
This is a local situation; not about big points or direction, etc.
You are in atari (of course, connect is not always correct), why didn't you connect ? What's the specific reason ( tactically ) for ?
( Just like , a "fuzzy feeling" is not good enough here ;
what's your exact reading that prompted you not to connect but play ?

In your variation for , you stopped short:
B's next move is Q17 atari -- it's a ko.

This time you connected (correct). Re:

E11.

I followed until , and stopped -- Re: posts 4 & 5.

Ed, thanks for taking the time to review, I really appreciate the pointers!

Bad. The lower right corner is already W's [...] Please show an exact sequence, not just a "feeling".

The situation I was concerned about was unfounded. I thought if Black cut, they'd be able to take the left side of that group, but now working through the variations this is unlikely. I should have tried to read out the situation instead of just reacting by instinct (since my Go instincts are still in their infancy and cannot be trusted)!

Click Here To Show Diagram Code

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

Click Here To Show Diagram Code

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

The game move extend D17 is good; the hane E17 variation is bad.
...
You need to reply to . W is happy to extend to D13.

I remember thinking I needed to protect the corner, but looking at it again, I can see that protected both the corner and the sides. Black would have had a hard time trying to crawl around that.

You're also right. The hane varation would have been disastrous for the corner.

What's the specific reason ( tactically ) for ?

Looking back, this seems silly now. My original thinking was that if I can set up a stone underneath the ko, like , that if he takes the stone, I can take back and break the ko from underneath. I had a game where there was a sort of ko cascade in several connected diamonds where I was able to flip a corner ­ I was trying to repeat my success. I remember thinking something like this...

Click Here To Show Diagram Code

[go]$$B Fantasy ko sequence
$$ | . . . . . . .
$$ | . . . O O . .
$$ | . . . O X X .
$$ | . . O X X . .
$$ | . O 1 O X . .
$$ | . . O X C . .
$$ | . . . 2 . . .
$$ | . . . O . . .
$$ | . . . , . . .[/go]

...and aiming at the marked point, but Black would just connect there. It's all out of sequence and painfully obvious now. I played without really thinking about how it would play out.

In your variation for , you stopped short: B's next move is Q17 atari -- it's a ko.

I actually didn't see this until you mentioned it. Another lesson in slowing down and actually reading the situation.

Re: posts 4 & 5.

I've started tracking my progress and game results in a spreadsheet (I'm very visual) and sadly my games completed chart has plateaued recently. I have 5 concurrent correspondence games going across different servers but this is another good reminder to prioritize just playing over tsumegos, reading, or lectures.

Thanks again for your comments!

Hi George,

since my Go instincts are still in their infancy and cannot be trusted!

Yea, this takes time. One step at a time. For example, variation 1:

$$B post 22 var 1
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X . O O . . |
$$ . . X O 1 4 . . |
$$ . . 5 3 2 . . . |
$$ -----------------

Click Here To Show Diagram Code

[go]$$B post 22 var 1
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X . O O . . |
$$ . . X O 1 4 . . |
$$ . . 5 3 2 . . . |
$$ -----------------[/go]

is not a cut; it's a clamp.
And your is incorrect; can you find the correct local sequence ?
( In your var, is incorrect, and is also incorrect. )

( We go through your replies one by one; one step at a time.
We deal with your var 2 next. )

Take 2 at variations

Thinking about it again, the clamp would seem to threaten above and below the marked white stone? If playing below, if Black tries to take, then connect at 4.

Click Here To Show Diagram Code

[go]$$B
$$ . . . . . . . . |
$$ . . . . X O . . |
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X 4 O O . . |
$$ . . X Q 1 . . . |
$$ . . . 2 3 . . . |
$$ -----------------[/go]

If black tries to play above after 2, then atari? And if Black keeps going, then capture.

Click Here To Show Diagram Code

[go]$$B
$$ . . . . . . . . |
$$ . . . . X O . . |
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X 3 O O . . |
$$ . . X Q 1 4 . . |
$$ . . . 2 . . . . |
$$ -----------------[/go]

Seems like that's one extra point (of territory) for White than my original variations. Not sure if that's the right sequence, but it's a little better.

Also in my original variations puts itself into atari, which would be foolish. would have captured ... and Black could have gone on to start a ko on the bottom?

Hi George,

Also in my original variations puts itself into atari, which would be foolish. would have captured ...

Click Here To Show Diagram Code

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

( I shaved off two horizontal lines on top;
in posts 22 and 24, you shifted from P6 to Q6.
We focus only on the very local reply to the clamp. )
Correct about and ( but still incorrect ).

and Black could have gone on to start a ko on the bottom?

We're getting a bit off course; we can discuss this later. But first:

$$B post 24 var 1
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X 4 O O . . |
$$ . . X O 1 . . . |
$$ . . . 2 3 . . . |
$$ -----------------

Click Here To Show Diagram Code

[go]$$B post 24 var 1
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X 4 O O . . |
$$ . . X O 1 . . . |
$$ . . . 2 3 . . . |
$$ -----------------[/go]

$$B post 24 var 2
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X 3 O O . . |
$$ . . X O 1 4 . . |
$$ . . . 2 . . . . |
$$ -----------------

Click Here To Show Diagram Code

[go]$$B post 24 var 2
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X 3 O O . . |
$$ . . X O 1 4 . . |
$$ . . . 2 . . . . |
$$ -----------------[/go]

Both of these vars are still incorrect. The correct is still tricky for you.

The correct is still tricky for you.

Yes, I often struggle with how my opponent will reply to my move when I try to visualize a sequence. Thank you for your patience.

Click Here To Show Diagram Code

[go]$$B Take 3
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X b O O . . |
$$ . a X O 1 . . . |
$$ . 3 2 4 . . . . |
$$ -----------------[/go]

I started here, thinking that moving out one stone on the first line still gives the same space, but creates a weakness at A. However, I noticed that Black can play at B, and eat those White stones. I cannot capture fast enough to prevent or defend. So this cannot be the correct variation.

Click Here To Show Diagram Code

[go]$$B Take 4
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X . O O . . |
$$ . . X O 1 2 . . |
$$ . . . . a . . . |
$$ -----------------[/go]

I think I might be overthinking it. In my original variation I atari from underneath, allowing Black to run in. If I atari from the right, even if Black tries to save at A, there's nowhere to go.

I think the variations you provided are more than good enough to prove that the black clamp at is just bad.
I think Ed wants to play at b in your last diagram(right?),which is even better, because this threatens the cut follow up. But even without that cut, black 1 would be just bad, as most of the moves you provided show(descent and either atari seem all fine to me in that respect).

I had made an earlier diagramming mistake that would affect the cut, context you didn't have. Adding it back in for reference.

I had considered connecting after the hypothetical clamp, but it seemed risky to me. If Black plays underneath:

Click Here To Show Diagram Code

[go]$$B
$$ . . . X . O . . |
$$ . . . 5 . . . . |
$$ . . 4 X X O . . |
$$ . . X 2 O O . . |
$$ . . X O 1 . . . |
$$ . . b 3 a . . . |
$$ -----------------[/go]

I can't play at A or B. If I take the cut then Black could connect at 5. Defending the lower right corner might be difficult. Unless there's something I've overlooked.

I would agree that the clamp is bad for Black and that my reasoning in-game for was faulty, but this is a good puzzle for me. Where is the path to MAXIMUM PUNISHMENT!?

I'm not sure the cut is so scary right away, its just something additional to worry black, but you need to realize that your black 3 is a small gote endgame. yes he takes some point away from you, but very few. If you want to cut right away, the continuation would be exend at 6

Click Here To Show Diagram Code

[go]$$B
$$ . . . X . O . . |
$$ . . . 5 . . . . |
$$ . 6 4 X X O . . |
$$ . . X 2 O O . . |
$$ . . X O 1 . . . |
$$ . . b 3 a . . . |
$$ -----------------[/go]

to see that his connection at 3 is small:

Click Here To Show Diagram Code

[go]$$B
$$ . . . X . O . . |
$$ . . . . . . . . |
$$ . . . X X O . . |
$$ . . X 2 O O . . |
$$ . . X O 1 5 6 . |
$$ . . b 3 a . . . |
$$ -----------------[/go]

If you play away with 4, then his push in at 5 doesnt do to much, as he needs to come back protect at a after 6.


Also, come to think of it, maybe Ed was opting for tenuki as (which gives up 1 stones). That might be even better, but this was about showing that is not to be feared, so I wouldnt show that with tenuki.

Hi George,

$$B Take 4
$$ . . . . . . . . |
$$ . . x X X O . . |
$$ . . X . O O . . |
$$ . . X O 1 2 . . |
$$ . . . . a . . . |
$$ -----------------

Click Here To Show Diagram Code

[go]$$B Take 4
$$ . . . . . . . . |
$$ . . x X X O . . |
$$ . . X . O O . . |
$$ . . X O 1 2 . . |
$$ . . . . a . . . |
$$ -----------------[/go]

In my original variation I atari from underneath, allowing Black to run in. If I atari from the right, even if Black tries to save at A, there's nowhere to go.

Thank you for the patience; I was looking for this atari in take 4, but because of other local properties (say, (x)), this wasn't so clear.
I debated for some time whether to tweak the local shape a bit; example:

Click Here To Show Diagram Code

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