Life In 19x19

Hi all... I hope this kind of thread will not annoy you, but nobody ever explained me a single joseki other than online dictionary so I hope to get some help here. Two question in a single thread, to avoid "Game analysis" pollution

This is GAME A



Until white 22 it's joseki, but now black insead of extending the wall, he easily steals white's big corner. In the end he will not get wall+extension but if white invade in the triangled final zone black can still attack her. Why nobody plays this way as black? Is white compensation ok? Can you clarify final position?

This is GAME B


This is an old fashon joseki that had a revival thanks to AlphaZero. The final position is black sente and white territory: the old evaluation was "Black (with that infuence) is locally better. AG says that position is even, but in any case not worse for black.

My question is: how can I use that influence? Or, put it another way: What is black LOCAL direction of play? I, of course, know that there might be trillions of variation, but I ask you to show me some LOCAL LOCAL LOCAL variations using black sente. Because I can't understand how to handle this position, with white having that big corner without aji

Thanks!

Game A: Firstly, it's not so easy to steal white's corner: at move 31 this is famous tripod group with extra leg shape so locally speaking white has a ko to kill. Also white had some other choices before and while you can't really kill the 3-3 it does sometimes get close and depend on outside aji so for example a white stone around c12 or capturing f14 can make a difference. Also sometimes people (not necessarily pros) do sometimes invade the 3-3 fairly early in this sort of shape, I've seen it a few times. I also seem to recall AlphaGo suggesting similar ideas browsing its opening book, and even playing the g16 wedge rather than m17 jump iirc.

Game B: Black wouldn't play a local move, q9 high Chinese looks nice to make a moyo. Locally black e18 is a fairly large move that stops white sliding so makes sense when you are trying to make the top side territory, but white's unlikely to answer it until endgame. Maybe this is why AlphaGo thinks this slide of 16 is a good move for white, I was surprised as it seems too small to play now but maybe it thinks black e18 is a good move if it tenukis given 15 is starting to develop the top side (but it's still very open!). That we only see the win% for a single move is rather frustrating in positions like this, does AG actually think 16 is signficantly the best move on the board, or is it about the same or even worse than a/b/c?

Click Here To Show Diagram Code

[go]$$Bcm15
$$ +---------------------------------------+
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . 2 . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . 1 . . . . . |
$$ | . . . O X . X . . , . . . . . X . . . |
$$ | . . O X . X . . . . . . . . . . . . . |
$$ | . . O . X . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , b . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . a . . c . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+[/go]

Game B: I don't have any idea what the AI would be "thinking". For me the evaluation would be highly dependent on the whole board situation. From my perspective Black's thickness works very well overall with the two right side corner star point stones and, in addition, Black has sente. Clearly at this moment White is ahead in settled territory but in further development Black is ahead and White will be playing catch-up for some time.

In game 'A', on move 28, what happens if white plays B16 instead of C16? It is too early in the morning for me to think clearly, but my first impression is that the corner is dead.

Joaz Banbeck wrote:In game 'A', on move 28, what happens if white plays B16 instead of C16? It is too early in the morning for me to think clearly, but my first impression is that the corner is dead.

Indeed, which is why black should have played c18 solid connection as b18 hanging connection for a ko, as then b16 doesn't kill (if you block the push then black can live inside, and if you pull back then black can break out).

Fllecha wrote:Until white 22 it's joseki, but now black insead of extending the wall, he easily steals white's big corner. In the end he will not get wall+extension but if white invade in the triangled final zone black can still attack her. Why nobody plays this way as black? Is white compensation ok? Can you clarify final position?

Starting with the last question first, no, I cannot clarify the final position. Edit: The reason is that I did not look at it. I agree that was bad. It should have been a hanging connection.

The corner is not big. Why not? Because White is solid on the top side, and White has an extension towards the left side from the corner. Taking the corner can only make territory. That is too limited an objective this early in the game.

Now, it is not true that nobody plays this way as Black. Waltheri ( http://ps.waltheri.net ) shows two games where the corner invasion came immediately, somewhat less than ¼ of the time. (Those games match the top half of the board only.)
Why did two top pros make the 3-3 invasion, if the corner is not big? The reason is that the invasions were probes. In each case the invader forced the defender to choose which side to block on and took sente, leaving aji behind and the corner unsettled.

This is an old fashon joseki that had a revival thanks to AlphaZero. The final position is black sente and white territory: the old evaluation was "Black (with that infuence) is locally better. AG says that position is even, but in any case not worse for black.

Moi, I don't see how the influence makes Black "locally" better. For one thing, influence is not local, and besides that, I like influence and I still don't see that it is better in general than the underneath connection. {shrug}

My question is: how can I use that influence? Or, put it another way: What is black LOCAL direction of play?

I think you can answer the second question yourself. And get "LOCAL" out of your head. Influence in the opening is not local.

As for the first question, that is perhaps a bit difficult to explain. It is true that a lot of amateurs find it difficult to utilize thickness. The standard answer is to use it to attack, but attacking is not easy. {shrug}

In playing over ancient games from a few centuries ago, I discovered something that may help explain how to handle this kind of thickness. Players of that era were more territorial than players today, but they did not shy away from making influence (thickness) and letting their opponent make territory. One thing that surprised me was how often they did not seem to make use of that thickness. Now, by the time I was studying those games I already knew how to use influence on one part of the board to make territory on another part of the board, or to take away territory elsewhere. But these guys just seemed not to do much with the thickness. They did not use it to make territory, OC, but they did not use it to attack, either. It just sat there.

This looks like the kind of thickness that you can let sit. How come? First, its particular shape is primed to make eyes, so it is fairly robust. Second, White has the large slide one point further than the one Uberdude shows, so building a good framework on the top side is not easy for Black to do. At the same time, White is not going to accomplish much on the top side, either. Given all that, what is there to do? In the immortal words of the Beatles, Let It Be.

What about the White slide that Uberdude shows? As he points out, that eliminates Black E-18, which would then threaten to make a moyo. It also undercuts the Black thickness, which makes it more difficult for it to live. White may then be able to make hay by attacking Black. But all that is down the road.

On a somewhat related note of the corner invasion in first game, here is an old OGS game with some comments between me and another mid dan amateur in which, after the g16 wedge he didn't f16 push through but made an iron pillar, presumably because he didn't like the prospect of me taking the corner later. See move 33. I don't think it was particularly good or particularly bad: I managed to use the aji to more easily invade the right side later.

...as usual than you for the replies.

Yes, I made a mistake on taking the corner that easy and I missed B16 that basically kills the corner or at best the defending side has a ko.

Fllecha wrote:nobody ever explained me a single joseki

i vaguely remember that the Ishii Press book "39 basic Joseki" did a pretty good job of explanation. i remember reading Ishida's dictionary book too, and thinking that it was all too complicated for me.

Bill Spight wrote: get "LOCAL" out of your head.

best advice i ever saw on L19

Bill Spight wrote:ancient games ... these guys just seemed not to do much with the thickness. They did not use it to make territory, OC, but they did not use it to attack, either. It just sat there.

even without having studied any of them, i'm sure that's not quite fair; thickness is like a guard dog - stay away from it or it will bite you. So just by sitting there it's doing a lot. That's the meaning of influence.here, Swim sees that whereas the top left corner isn't white territory, white does have some influence in it so black better be careful about jumping in.

But more to the point, global considerations suggest that a black (or white!) move along or across the green line is where the real action is, from the point of view of moyo expansion and reduction.

I disagree with djhbrown/swim that in the last diagram at move 22 the left side green line is the important place: it's the right side which is most important here * as black has more potential there (so a black move to develop or white move to hinder it biggest), whilst the left side is comparably smaller and white can't build so much there with the black thickness at g14 looming above. Checking pro games in waltheri this position with 2 3-4s at the bottom hasn't occurred, but with 2 4-4s it has and in both cases black played the star point on the right side to develop the moyo (though it is true that with the 3-4s here approaching them is relatively larger).

* in terms of largest area, there is also a somewhat hot local question of whether black should play n17, would white answer, is it a good exchange if answered and if not is it ok to spend a move (and possibly soon another) on?

Click Here To Show Diagram Code

[go]$$B
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . d . . X . O O O . O . . . . . . . |
$$ | . . . O . O O O X X X . . . . X . . . |
$$ | . . . . . X O X . . . . X . . . . . . |
$$ | . . . O . . X X . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . b . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . a c . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . O . . . . . . . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Zen7 (Winrate black):
a 49%
b 48%
c 48%
d 47%


(By the way, I am already not a fan of move 5 and 6 in game A (In fact I dont like 3 also, but that is a completly different matter ). Before learning Joseki it is nice to decide which moves are the best in fuseki, then you only have to deal with a lot less Joseki.)

Game B:

Click Here To Show Diagram Code

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

Zen7 (Winrate black):

a 53%
b 53%
c 53%


But move 6 is already a mistake for white:

Click Here To Show Diagram Code

[go]$$W
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . a . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . b . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . O . . . . . . . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Zen7 (Winrate white):

a is already a mistake for white (49%)
b is much better (53%) (And therefore move 5 for black not good as well )

Gomoto wrote:

$$B
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . d . . X . O O O . O . . . . . . . |
$$ | . . . O . O O O X X X . . . . X . . . |
$$ | . . . . . X O X . . . . X . . . . . . |
$$ | . . . O . . X X . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . b . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . a c . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . O . . . . . . . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------

Click Here To Show Diagram Code

[go]$$B
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . d . . X . O O O . O . . . . . . . |
$$ | . . . O . O O O X X X . . . . X . . . |
$$ | . . . . . X O X . . . . X . . . . . . |
$$ | . . . O . . X X . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . b . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . a c . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . O . . . . . . . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

...Before learning Joseki it is nice to decide which moves are the best in fuseki

Yes, that endorses Bill's point but maybe goes a touch too far past it, in that i don't think any mere mortal can decided what's "best"! ... although, from their blurb, it does sound like Fan Hui et al do think Alfie can...

I am unconvinced that the sequence to 22 is an even result in the context of the rest of the board - it looks top-heavy. i walked through the moves one by one, and got the impression that white 18 (wedge at G15) was premature, since white was already settled on the top, and 3/4 of the board was still wide open. I guess that white was trying to kill something black, right from the get go, but i see the wedge as aji-keshi, only solidifying black's thickness and thereby widening black's centre influence.

Gomoto wrote: then you only have to deal with a lot less Joseki.

i'm not so sure about that - i think it means you have to study more! i took a look at what Kogo has to say about white 10 .. but Sabaki won't let me copy its text so you'll have to ask Kogo yourself to see what a difference (in Kogo's eyes) the exact placement of the black stone in the top right makes.

Gomoto wrote:..I am already not a fan of move 5 and 6 ..

black 5 looks ok to me - but what do i know?!

white 6 looks like wrong direction, because of the black stone on Q16 which limits white at the top, and the white stone at D3, which must be feeling like a neglected lover after white H17. In contrast, black 7 looks to me like a good move, thumbing its nose at white 6.

After black 7, black has a solo presence in 2 corners, and has reduced white's grip on a third, and has split white's top 2 stones creating plenty of aji on the left as well as staring down from both sides at the lonely white stone on the top. so if i were black, i would feel pretty happy at that stage. Given all that, i wouldn't call the sequence an even result, even if it is joseki out of context.

Sir H: "Everything is connected..." etc

Gomoto wrote:But move 6 is already a mistake for white:

$$W
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . a . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . b . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . O . . . . . . . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------

Click Here To Show Diagram Code

[go]$$W
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . a . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . b . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . O . . . . . . . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Zen7 (Winrate white):

a is already a mistake for white (49%)

Yes, AlphaGo devalues pincers. Interesting that Zen7 does, too, at least in this case.

In playing over ancient games from a few centuries ago, I discovered something that may help explain how to handle this kind of thickness. Players of that era were more territorial than players today, but they did not shy away from making influence (thickness) and letting their opponent make territory. One thing that surprised me was how often they did not seem to make use of that thickness. Now, by the time I was studying those games I already knew how to use influence on one part of the board to make territory on another part of the board, or to take away territory elsewhere. But these guys just seemed not to do much with the thickness. They did not use it to make territory, OC, but they did not use it to attack, either. It just sat there.

I just came across a game (and pro comment) that seems to illustrate exactly what Bill was talking about above. Chitoku is playing Genjo and has just played the triangled move instead of the very common A.



Iwamoto explained that the reason was that White has no reason to worry about the Black force with Black B, White C, Black D, White A (as was eventually played much later) because he has thickness above. Now, as it happens, in this game White, for the quite small investment we see here, got a rather large territory of 25 points on the lower side and Black, not having any forcing moves on the lower side got only the left hand corner. On the right side White did not use his thickness to attack but still made just 17 points with Black erasing from afar, starting in the vicinity of P8. But that erasure got him little extra benefit because White was so strong below. In other words, the value of White's thickness n the upper right must include something for what he made on the lower side.

This thickness was powerful but Genjo seemingly cooperated in allowing it. This cooperation was in fact common at the time. Iwamoto disapproved strongly in this case and made a fascinating comment: "It is intriguing, given the simplicity of the position, to think that in those days they must have used a quite different means of assessment, especially considering that it is hard to find games won by the enclosed side."

It seems we may be witnessing Chitoku's thinking just starting to evolve beyond the theory of the time.