After playing the sente move of S14, B can pretty much play anything safe like M16,M17 or the the most secure N17 for when B backs of high from the approach to get a very big and definite profit in the corner white's side can still be flattened and is lesser than B's corner.
And this is already bad enough assuming that the side W can expand to is empty. Imagine if B had a stone in the bottom right?

Is this punishable or is this considered equal so there's no need to punish?

Hi Shoryuu,

Sorry, I'm not sure what exactly "this" is --
there are quite a few variations and moves in your SGF.

Which move or moves are you referring to ?In your main branch, at N17 seems slow.

It's the combination of R15 and S14 moves. The approach at R15 is said to be bad and helping W to get stronger if the side is empty and also that 3-3 is still open. But with S14 played, the 3-3 weakness is covered and the only thing is making sure that you get more than what you are helping W develop in the side.

N17 is slow but I think you can get away with playing even M17 or even M16. Even if N17 is slow though, isn't the corner better with 20 points of definite territory compared to W's 5 not yet certain territory yet? And if W makes another move to expand the side to say 10-15 points of territory, B is still leading from the tradeoff and has sente.

If I understand you correctly, you think black n17 or those other nearby moves are greedy (in that they try to secure a large territory), and give a good result for black? I disagree. Although the s14 hane is sente and does a lot to eliminate 3-3 aji, there is still some aji inside and it also is a bad exchange in the sense of giving white more eyeshape* should he later capture the stone (if white can play 3-3 and black blocks to the left then taking the stone is sente, and if black doesn't block to the left then maybe white can live inside). If black can't attack/pressure/press the outside white group he has fewer strategic options in the later game, similarly with the kick (for example bye-bye centre moyo options, or splitting attacks). Kick and s14 can be a practical move to take some early territory, but it loses thickness and aji and isn't normally anything to worry about for white.

Btw There's a few suspect moves in your variations, but one important tesuji you might not know is with the n17 variation when white tries to live inside black can think about answering q18 at the s18 vital point to kill (white can plays s18 himself to make a small life).

* It might seem a minor difference, but consider the following tewari argument: if white plays 1-3 below then black would either tenuki or play atari at a (to which white would probably fight a ko, not connect; if he didn't like ko then 3 as solid connection would be better as then there is still clamp/peep aji in the corner).

Click Here To Show Diagram Code

[go]$$W
$$ ------------------+
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . X . X . 2 . |
$$ . . . . . . X 1 a |
$$ . . . . . O O . . |
$$ . . . . . . . 3 . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . , O . . |
$$ . . . . . . . . . |[/go]

But instead with the s14 exchange it is as though black plays this obviously dumb exchange:

Click Here To Show Diagram Code

[go]$$W
$$ ------------------+
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . X . X . 2 . |
$$ . . . . . . X 1 . |
$$ . . . . . O O 4 5 |
$$ . . . . . . . 3 . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . , O . . |
$$ . . . . . . . . . |[/go]

Now of course there is a timing issue in that this 2nd line capture or hane-connect is probably not the biggest move for some time, so by playing s14 black can induce white into playing there earlier than he otherwise would.

Ah yes, W can be killed unconditionally with s18 which only proves my point that the corner is solid. Can't deny the aji which means lots of ko threats.

Kinda get what you mean but it's kinda hard to say that future board development will be bad simply by playing this corner sequence. But at least what I'm getting is, this is NOT greedy overplay and it's ok to play this way then I'm all for this style. Doesn't seem too bad.

Considering that one usually wouldnt play this in a diagonal fuseki, I would have a stone at A (upperleft quadrant) where I could try to make a big moyo along the whole top side or there's a stone at Q (lower right quadrant) limiting/pressurising W's side. Both seem good or at the very least, ok for B. Would I be right to say that?

But white can live with s18 (though whether you want to is another matter as it's a small life and damages the outside).


I'd call it greedy underplay I would be happy to take white and if you keep playing like this I'd expect to win by maybe 15 points (it signifies a mentality in which I will bully you into inefficiently defending your existing territory). Your opponent does need to be patient though and just leave such inefficient positions alone for now, so a greedy opponent might do something rash and then you can punish him to get a good result.

Yeah I wouldn't play it immediately either but I would get the nagging feeling to get compensation elsewhere to make up for it since I would feel B got it for cheap.

Also I'm pretty sure that's an abuse of tewari in your earlier post. In your diagram, W had the first move. If I'm not wrong, in tewari, order of first move should be the same and end result should be the same too.

Thank you for your thoughts!

I think it's a common misconception by people navigating the kyu ranks to look at a position like this and say: "B forced white with one move, and provided some support to the corner. White added a stone to the position, but it doesn't look like it did much. B made out well in this exchange."

However, unless you take gote and connect (hane on the second line, connect by black in the argued for sequence), you leave a glaring weakness in the corner that the other player can exploit. Most of the time, W is not going to simply take the one stone. He is going to invade under your shape and use it to create a live group.

In my interpretation of Uberdude's tewari analysis, black is giving a stone to white for free. It's better not to play this move if you don't intend to finish the situation/shape by connecting and solidifying the corner.

Shoryuu, in your original suggestions, you typically closed the top side with a stone to the left of your one point jump. The total black stone investment there for black is about 5 stones. In the opening, each stone should potentially be worth about 10-15 points. It seems unlikely that the corner is worth that much, especially with the weakness. That's indicative of an overinvestment of stones in the early stages.

_________________
Ko is the best solution.
With Ko, I can keep eating and drinking until I am full.

Visit >>>Koosh's Study Journal<<<

That's wrong, Koosh. After the exchange of Black's hane and White's block, there is no 3-3 invasion left.

Perhaps I didn't make it clear what positions I was comparing: it is the situation in which white plays a gote move in that area, comparing the cases whether black has made the s14 s13 exchange or not, to see if that is a bad exchange.

If black doesn't play it an white plays the hane connect we end up with the below result. This is a fairly large move which strengthens the white group, giving it some extra eyespace and territory, improves the connection to the r10 stone, and also creates some defects in black's corner (peep or clamp may be able to live inside, though a stone like n17 might change things, and at a minimum some nice 1st line yose with the famous life/ko inside if black blocks). You probably won't actually see this sequence with the 3 space extension and such a loose corner in strong player's games though because: 1. black is less likely to kick unless he continues more immediately, 2. white has better choices like 3-3. However, you do sometimes see this hane-connect when white has a 2 space extension on the outside and wants to reinforce the group (and create black corner defects for later).

Click Here To Show Diagram Code

[go]$$Wc
$$ ------------------+
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . X . X . 2 . |
$$ . . . . . . X 1 . |
$$ . . . . . O O 3 . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . , O . . |
$$ . . . . . . . . . |[/go]

If black has played s14 and tenukid then we get the below. This position is a lot better for white than the previous one: he has far more secure eyeshape and also the connection to the r10 stone is stronger thanks to the marked stone, for example a black move at a or b (which might be a nice ladder maker, developing move for a centre moyo, or useful move in an attack) no longer threatens to separate r10 from the stones above. The defects in black's corner are pretty much the same.

Click Here To Show Diagram Code

[go]$$W (3 captures)
$$ ------------------+
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . . . . . . . |
$$ . . . X . X . 2 . |
$$ . . . . . . X 1 . |
$$ . . . . . O O . 3 |
$$ . . . . . . . @ . |
$$ . . . . . a . . . |
$$ . . . . . b . . . |
$$ . . . . . , O . . |
$$ . . . . . . . . . |[/go]

If the above analysis isn't persuasive, the tewari argument could help. Imagine in the situation without the hane white played the hanging instead of solid connection. This is probably a bad move, as black can atari to fix the bad aji in the corner, but does have the upside of the slightly stronger connection to r10. However, he certainly wouldn't throw in and give white a free ponnuki! But that's the net result if he plays the s14 hane and then tenukis. Now of course the timing of if/when to play there is a global question and too early could be a bad move (a famous example being Lee Sedol in AlphaGo game 2), but in the cases white plays there at a good time s14 becomes a bad move.



So the first of Koosh's sentences above is true in the cases white plays there at a good time. However the second as a blanket statement is not because the hane does have some benefits regarding making the 3-3 harder, and is sometimes a good move to play and tenuki. But these must be judged together with the downsides described above.

@Uberdude: You got me. Blanket statements are dangerous like that. Thanks for the correction.

@Shaddy: Thanks for commenting. I was at work and fired off a quick comment. Sorry about that. Let me try to better explain my logic.

1) The shape of five stones (N17, O16, Q16, R15, S14) uses five stones to make less than 20 points. Is this a reasonable strategy in a modern opening?

2) Where N17 has been placed to enclose the corner, Shaddy's right; I see now that there is no 3-3 invasion when B responds to W's attack at P17 with Q17. However, with W able to capture the stone at S14 or saddle up to the B position at L17, I'm not happy with N17 as the gote move to seal up the corner for B, which brings me to a point I'd like to clarify since I think B should be trying to push further along the side if possible.

After B's initial R15 kick, hane, and then tenuki, should B decide to seal up the corner with moves that shoryuu recommended in the sgf attached (N17, M16, M17), it seems prudent to know:
A) if W can live in the corner in each of these three additional cases without major damage outside (we've already seen that with N17, W cannot live) and
B) what the best continuations are with these shapes. Also, what about O17 as opposed to O16?

I'd end up spending about 20 minutes reading the corner if it showed up in a game, so some guidance with how to think given these positions would be appreciated.

Position 1


Position 2


Position 3

_________________
Ko is the best solution.
With Ko, I can keep eating and drinking until I am full.

Visit >>>Koosh's Study Journal<<<

1) It's nearly unplayably slow.

2) After capturing the one stone, White can generally live in the corner after clamping. Black should probably defend by making a ponnuki shape, when he wants to; descent also works but leaves worse aji, so it depends on the game.

Thanks. That's more or less what I wanted to explore. How does W live in the corner after clamping? Guessing that B descends, W peeps, black connects, and then W plays a knight's move.

I came up with one idea for a sequence to live but it only results in ko. I wanted to know if the position of the stone to the left makes a difference.

_________________
Ko is the best solution.
With Ko, I can keep eating and drinking until I am full.

Visit >>>Koosh's Study Journal<<<

S19 on move 13 lives

You're right. I also see that W dies if N17 was instead a knight's move (instead of 1-space jump and kosumi). Since W cannot live inside with the knight's move, the lesson I'm getting from this is that the knight's move secures the corner better than the one space jump and kosumi (the original shape in the first post).

Click Here To Show Diagram Code

[go]$$Bc
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . 3 4 . O . |
$$ | . . . . . . . . . . . . . X O 6 7 8 1 |
$$ | . . . . . . . . . . . . X . O 5 O O 2 |
$$ | . . . , . . . . . , . . . X X X X X X |
$$ | . . . . . . . . . . . . . . . . X O . |
$$ | . . . . . . . . . . . . . . . O O . O |
$$ | . . . . . . . . . . . . . . . . . O . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

_________________
Ko is the best solution.
With Ko, I can keep eating and drinking until I am full.

Visit >>>Koosh's Study Journal<<<

For move 10 Black can push and clamp. Then white p17 and n18 will either live, make a big ko, or some second line crawl escape. So with a little extra support outside such as L17 black could kill with that line, but then white could also play move 9 instead of slide where black pushes and that should be a straight J group ko.

@Uberdude: I made another file with thoughts based on your comment below. I'm sure I missed something, but it looks like the following points are true about this situation.

1) If B has support around L17, he can kill B as shown in the file.
Conclusion: If B has support outside at L17, W should not clamp here, and should simply hane on the first line to reduce the corner.

2) If B doesn't have support around L17, the W clamp will live inside or crawl outside.
Conclusion: If W clamps and B does not have support around L17, B must respond with L17 (letting W into the corner) or R17 ponnuki shape.

3) The move you suggested (making the straight J group) doesn't seem to make a ko. It just dies.
Conclusion: The straight J with a B descent outside is dead.

Shaddy, myself, Uberdude all basically agree that OP's original move sequence (5 stones to protect the corner) is slow. W would take the stone on the 2nd line in sente as B would owe a move to either protect the corner or extend along the top; otherwise, W can clamp and make a living group there as demonstrated in the sgf below. The position after W takes the 2nd line stone and B protects is fine for W, while B was able to protect 18 (adding a stone in the corner)-20 (extending along the top) points in the corner+side.

So...I am trying to find utility for a move similar to OP's idea. I can say from this discourse that there is a lot of aji here, and the position is pretty dynamic. I can see the allure of playing this way. If someone is going to try this style, though, it would be prudent to know all of the invading variations. (I mentioned before that I'd need to spend a chunk of time reading this out). You never know what kind of knowledge is helpful in a tournament setting.

_________________
Ko is the best solution.
With Ko, I can keep eating and drinking until I am full.

Visit >>>Koosh's Study Journal<<<