It is currently Sat May 03, 2025 9:38 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 20 posts ] 
Author Message
Offline
 Post subject: A typical game of mine against GNU Go
Post #1 Posted: Sun Jun 12, 2016 4:10 pm 
Lives with ko

Posts: 161
Liked others: 2
Was liked: 25
This is one of the typical games I play against non-MonteCarlo engines such as GNU Go, Aya, or the webbased Cosumi. If I can't get them to fight somehow, I'll lose by 20 points or more. Mostly it's like this:

- Me building up a huge lead.
- Then I fall behind because of the huge influence the engine has
- And then I frantically try to reduce or invade, which can have two outcomes:
1. I fail the invasion and I die.
2. I succeed to reduce, but get reduced myself, and lose anyway.

This is becoming frustrating as I don't seem to be able to prevent a non-MC-engine getting either huge influence, or a huge moyo. As soon as I get a non-MC engine to fight, I start beating it soundly, which makes me look several stones stronger; if they play solidly however, I seem to be several stones weaker. I've commented what I was thinking during the game.

Feel free to comment, and thanks in advance.



Attachments:
me_gnugo.sgf [8.15 KiB]
Downloaded 602 times


Last edited by Babelardus on Sun Jun 12, 2016 4:17 pm, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject:
Post #2 Posted: Sun Jun 12, 2016 4:15 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

You can use the SGF tags to embed your game here.

Top
 Profile  
 
Offline
 Post subject: Re:
Post #3 Posted: Sun Jun 12, 2016 4:17 pm 
Lives with ko

Posts: 161
Liked others: 2
Was liked: 25
EdLee wrote:
Hi Babelardus,

You can use the SGF tags to embed your game here.


I know. Made a mistake in the URL and had to do it over. Fixed now :)

Top
 Profile  
 
Offline
 Post subject:
Post #4 Posted: Sun Jun 12, 2016 4:20 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

OK, you included the SGF now. :)

In one of your variations of :b15: -- please see screenshot below --
you played a very strange move -- D18 connect --
to improve, you must fix this kind of moves ( by yourself ) :)
Attachment:
15.jpg
15.jpg [ 59.89 KiB | Viewed 10078 times ]
Babelardus wrote:
1. I fail the invasion and I die.
2. I succeed to reduce, but get reduced myself, and lose anyway.
There are (probably) multiple reasons for the above 'typical' results in your games,
and one reason is the kind of moves like D18 connect -- you must fix this by yourself. :)

Top
 Profile  
 
Offline
 Post subject:
Post #5 Posted: Sun Jun 12, 2016 4:30 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

Earlier, in the same :b15: variation, you played an empty triangle move for W -- E16 connect --
please see screenshot:
Attachment:
15_empty_triangle.jpg
15_empty_triangle.jpg [ 58.5 KiB | Viewed 10077 times ]
This is also not right.
The correct local move for W is block at D17.
You (also) need to develop a sense or feeling for good shapes, and bad shapes ( like your E16 connect for W ).

Top
 Profile  
 
Offline
 Post subject:
Post #6 Posted: Sun Jun 12, 2016 4:35 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

About :b15: , this is a very common situation -- basic shape --
Please see game review thread, post 4.

Top
 Profile  
 
Offline
 Post subject: Re: A typical game of mine against GNU Go
Post #7 Posted: Sun Jun 12, 2016 4:43 pm 
Lives with ko

Posts: 161
Liked others: 2
Was liked: 25
First, I obviously added the variations and comments afterward; I didn't read out 4-5 variations 10-15 moves deep. I can't yet do that. The variations are what I *think* what would have happened.

Thanks for pointing out the connection at D18. The connection is probably not necessary. The reason I play those moves regularly is because I can't read out (= get confused and lose the 'mental view' of the board) the entire variation I'm thinking about. Then I don't know if leaving the connection hanging is safe or not.

One of those moments is move 121. I see several big(ger) points on the board, but dare not play them as I'm afraid that white will, at some point, revive the stone at K11, after which my entire reduction would (possibly) be killed. At move 90, I have some variations that point out that white J11 can't be used to disconnect the reducing stones, but after white 116 at N11 I became unsure, in reading out possible variations, so I felt I needed to kill white K11 now.

It did cost me a move, and such moves could obviously be used to strengthen an invasion or reduction.

What do you mean with "you must fix this by yourself" ?

Top
 Profile  
 
Offline
 Post subject:
Post #8 Posted: Sun Jun 12, 2016 4:44 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

As we see in the above 3 examples, basic shapes are a recurring problem.

Here's another one:

:b9: This is wrong feeling, and bad shape for Black.
( Diagram rotated for space ):
Click Here To Show Diagram Code
[go]$$B Wrong feeling.
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . X 1 O . . . . ,
$$ | . . O . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ ---------------------[/go]

The result is good shape for W, and bad shape for B:
Click Here To Show Diagram Code
[go]$$B
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . X X W . . . . ,
$$ | . . O 2 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ ---------------------[/go]

Proverb: "Hane at the head of (2) enemy stones." -- you make bad shape for B, and good shape for W.
Please see also Hane Head of 2 stones.

Top
 Profile  
 
Offline
 Post subject: Re:
Post #9 Posted: Sun Jun 12, 2016 4:46 pm 
Lives with ko

Posts: 161
Liked others: 2
Was liked: 25
EdLee wrote:
Hi Babelardus,

Earlier, in the same :b15: variation, you played an empty triangle move for W -- E16 connect --
please see screenshot:
Attachment:
15_empty_triangle.jpg
This is also not right.
The correct local move for W is block at D17.
You (also) need to develop a sense or feeling for good shapes, and bad shapes ( like your E16 connect for W ).


Hi, EdLee,

Thanks for the help. I know the shape you point out, but I don't understand why it works.

After white D17, doesn't B-E16, W-E15, B-D15, W-C15, B-D14 sequence disconnect those stones?

EdLee wrote:
Hi Babelardus,

As we see in the above 3 examples, basic shapes are a recurring problem.

Here's another one:

:b9: This is wrong feeling, and bad shape for Black.
( Diagram rotated for space ):
Click Here To Show Diagram Code
[go]$$B Wrong feeling.
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . X 1 O . . . . ,
$$ | . . O . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ ---------------------[/go]


Why is this, with the two or three space extension as a follow-up, a bad shape? What would be the correct answer in that situation?

Top
 Profile  
 
Offline
 Post subject:
Post #10 Posted: Sun Jun 12, 2016 4:57 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Babelardus wrote:
I didn't read out 4-5 variations 10-15 moves deep. I can't yet do that.
...
Thanks for pointing out the connection at D18. The connection is probably not necessary. The reason I play those moves regularly is...
...
What do you mean with "you must fix this by yourself" ?
Hi Babelardus,

About your D18 connect in your :b15: variation:
This does not require 15 move reading.
It requires 3 moves, perhaps:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . 1 X 2 3 . . . . .
$$ | . O O X X X . . . .
$$ | . . . O O O X . . ,
$$ | . . . . . O . . . .
$$ | . . . . . . . . . .[/go]
It means rather than waiting for other people to fix this kind of moves for you, to point them out for you,
it's much better if you learn to see this kind of mistakes, and the better moves, by yourself.

Top
 Profile  
 
Offline
 Post subject:
Post #11 Posted: Sun Jun 12, 2016 5:05 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

About :b9: , one continuation is the hane:
Click Here To Show Diagram Code
[go]$$B
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . X . O . . . . ,
$$ | . 1 O . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ ---------------------[/go]
Continuing:
Click Here To Show Diagram Code
[go]$$B
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . 3 . . . . . . . .
$$ | . . X . O . . . . ,
$$ | . 1 O 2 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ ---------------------[/go]
There are other variations.
It may be a good time to start to read up on some basic joseki sequences.

Top
 Profile  
 
Offline
 Post subject:
Post #12 Posted: Sun Jun 12, 2016 5:22 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Babelardus wrote:
After white D17, doesn't B-E16, W-E15, B-D15, W-C15, B-D14 sequence disconnect those stones?
Hi Babelardus,
Click Here To Show Diagram Code
[go]$$Wc
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . 1 X X . . . .
$$ | . . . O 2 O X . ,
$$ | . . 5 4 3 O . . . .
$$ | . . . 6 . . . . . .
$$ | . . . . . . . . . .[/go]
We look at some more basic shapes here.
Your exchange of ( :w5: atari - :b6: extend ) is a kind of 'common mistake' --
you must ask yourself: why atari ? Why do you send Black out so nicely for him ?

Instead, W simply descends:
Click Here To Show Diagram Code
[go]$$W
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . 1 . . . . . .
$$ | . . . O X X a . . .
$$ | . . . O X O X . ,
$$ | . . . X O O . . . .
$$ | . . . b . . . . . .
$$ | . . . . . . . . . .[/go]
Now, you see W has a problem with the (a) cut.

Suppose B also descends; W cuts with :w2: -- now the 3 :bc: stones are in trouble, with only 2 liberties left:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . O . . . . . .
$$ | . . . O B B 2 . . .
$$ | . . . O B O X . ,
$$ | . . 1 X O O . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]
B can try to resist a bit; B must connect with :b3: ; after :w4: , B has only 3 liberties left --
2 liberties + 1 shared liberty with W;
W has 4 outside liberties and 1 shared liberty:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . O . 3 1 4 . .
$$ | . . . O X X O 2 . .
$$ | . . . O X O X . ,
$$ | . . X X O O . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]

Top
 Profile  
 
Offline
 Post subject:
Post #13 Posted: Sun Jun 12, 2016 5:42 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
If B connects the (a) cut, here's one continuation:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . 7 . O . . . . . .
$$ | . 8 . O X X 1 . . .
$$ | . 5 . O X O X . ,
$$ | . . 3 X O O . . . .
$$ | . 6 4 2 . . . . . .
$$ | . . . . . . . . . .[/go]
You can play around with the variations here and see what you discover. :)

One basic shape lesson here: notice the atari W wants is this ( :w2: - :b3: ) exchange,
where W takes the outside and forces B to take the inside.
Compare this with your exchange of ( :w5: - :b6: ) in post 12.

Top
 Profile  
 
Offline
 Post subject:
Post #14 Posted: Sun Jun 12, 2016 5:47 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

We see that after :w1: drops, B has two problems -- (a) and (b) --
B cannot fix both of them. White has miai of (a) and (b):
Click Here To Show Diagram Code
[go]$$W
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . 1 . . . . . .
$$ | . . . O X X a . . .
$$ | . . . O B O X . ,
$$ | . . . X W O . . . .
$$ | . . . b . . . . . .
$$ | . . . . . . . . . .[/go]
One reason is B made the exchange of ( :bc: - :wc: ),
which has reduced B's own liberties -- B now has a shortage of liberties on top.

Top
 Profile  
 
Offline
 Post subject:
Post #15 Posted: Sun Jun 12, 2016 5:51 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Something else interesting about this :w1: descend:
Click Here To Show Diagram Code
[go]$$W
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . 1 . . . . . .
$$ | . . . O X X . . . .
$$ | . . . O X O X . ,
$$ | . . . X O O . . . .
$$ | . . . . . . . . . .[/go]
Suppose W missed this move, and B gets it first instead:
Does this shape remind you of anything we discussed so far:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . 1 . . . . . .
$$ | . . . O X X . . . .
$$ | . . . O X O X . ,
$$ | . . . X O O . . . .
$$ | . . . . . . . . . .[/go]
Basic shape (proverb): Hane Head of 2 stones.

Please see also Post 8, above.
Did you notice that this proverb has occurred twice on the 2 :white: stones:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . B . . . . . .
$$ | . . . O X ? . . . .
$$ | . . . O X ? ? . ,
$$ | . . . B ? ? . . . .
$$ | . . . . . . . . . .[/go]
It's a hint to help us see the :w1: descend.
( It doesn't always work like this; but sometimes it does. :) )

Top
 Profile  
 
Offline
 Post subject:
Post #16 Posted: Sun Jun 12, 2016 6:03 pm 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Hi Babelardus,

Depending on the global situation ( the whole board ),
suppose the 2 :wc: stones are not important to White:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . O X X . . . .
$$ | . . . O 1 W X . ,
$$ | . . . . . W . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]
Then White even has the option to give up the 2 stones,
and take a lot of territory on the left:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . O X X . . . .
$$ | . . . O X O X . ,
$$ | . . . 2 . O . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]
I'm not saying that's the situation in your variation.
I'm saying for certain board positions, this is a possibility for White to consider.
The key word here is the global perspective:
Click Here To Show Diagram Code
[go]$$B
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . O X X . . . .
$$ | . . . O X O X . ,
$$ | . . . O 3 O . . . .
$$ | . . . 4 . . . . . .
$$ | . . . . . . . . . .[/go]

Top
 Profile  
 
Offline
 Post subject: Re: A typical game of mine against GNU Go
Post #17 Posted: Sun Jun 12, 2016 8:56 pm 
Honinbo

Posts: 10905
Liked others: 3651
Was liked: 3374
While I am sure that there are many plays that could use improvement, I tried to find where the game was lost. I think that the game was close at around move 100, but White quickly took the lead. Go to move 101.



Main focus: Do not follow your opponent around. Look for moves that threaten the life of groups or secure life while threatening to take away potential territory. Look for double sente.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.


Last edited by Bill Spight on Mon Jun 13, 2016 7:06 am, edited 2 times in total.
Top
 Profile  
 
Offline
 Post subject: Re: A typical game of mine against GNU Go
Post #18 Posted: Mon Jun 13, 2016 4:20 am 
Lives with ko

Posts: 161
Liked others: 2
Was liked: 25
Bill and EdLee: Thanks for taking the time to take a look at the game.

I'll definitely have to pick up a book on shapes (and joseki as well, as I don't know enough of them either). Mostly, in lectures, I see people saying that 'this is bad shape' or 'good shape', without really explaining why. So now, I often make a shape because I know it's a good shape, but don't understand why. I also play Joseki's without really understanding why they are equal.

I see that I indeed missed quite some opportunities around move 100 (and at other points in the game as well, probably). I also see that I seem to be way too afraid to play things like the hane or knight's jump, for fear of cutting and/or disconnecting; I'm playing too passive.

Top
 Profile  
 
Offline
 Post subject:
Post #19 Posted: Mon Jun 13, 2016 5:23 am 
Honinbo
User avatar

Posts: 8859
Location: Santa Barbara, CA
Liked others: 349
Was liked: 2076
GD Posts: 312
Babelardus wrote:
I see people saying that 'this is bad shape' or 'good shape', without really explaining why. So now, I often make a shape because I know it's a good shape, but don't understand why.
Hi Babelardus,

Yes; this is a common experience.

On the topic of good v. bad shapes,
going back to post 4 -- as discussed,
your empty triangle :white: (a) is bad shape --
the natural move is block with :w2: :
Click Here To Show Diagram Code
[go]$$B Reference shape
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . 2 1 X . . . .
$$ | . . . O a O X . . ,
$$ | . . . . . O . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]
However, even if you're worried about the :black: (a) push,
you can still do better than (a) -- this fixes it:
Click Here To Show Diagram Code
[go]$$B Var 1
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 1 X . . . .
$$ | . . 2 O . O X . . ,
$$ | . . . . . O . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]
Or:
Click Here To Show Diagram Code
[go]$$B Var 2
$$ ---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 1 X . . . .
$$ | . . . O . O X . . ,
$$ | . . . . . O . . . .
$$ | . . . 2 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .[/go]
Of course, :w2: block in the Reference diagram is still the best,
but at least you could consider Vars 1 & 2,
instead of the empty triangle connect.

Top
 Profile  
 
Offline
 Post subject: Re: A typical game of mine against GNU Go
Post #20 Posted: Mon Jun 13, 2016 7:10 am 
Honinbo

Posts: 10905
Liked others: 3651
Was liked: 3374
BTW, best not to use [] in comments. Eidogo seems to have trouble with that. In the SGF file I uploaded White had lost four stones on the board by move 100. I corrected [] to O, and the file looks OK now.

Here is what happened. In your uploaded file, [] in the comments looks like this: [\]. Eidogo handles that fine. But GoWrite does not, because there are too many "["s in the file. So GoWrite converts that to \[\]. Eidogo responds by not showing the play that the comment goes with. ;)

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 20 posts ] 

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group