In 38 Basic Joseki it says there is no way for white (invader) to live with the 3-3 invasion when black has the 4-4 and a small knight's extension, except fighting a ko.

I have tried many variations to effectively kill off the white invader when he doesn't follow the ko-joseki but attempts to play the usual joseki of en open corner (only 4-4 stone there).

Can anyone point me in the right direction? So next time my opponent invades my enclosure and doesn't go for ko, I can finally punish him for it

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

Visualize whirled peas.

Everything with love. Stay safe.

I was confused by the same thing at your level. I won't claim to fully understand it now, but I am much less confused (So, hopefully, someone else will correct me if I'm wrong, and we can both learn from this!)

Usually, what is said in this situation is that with the 3-3 invasion, "The best white can do is ko". I think beginners often interpret this as "it's impossible to live in there unconditionally." The logic being that if you can live in there, then living unconditionally would be better than having to fight a ko to live? What they actually mean is that, while white can live, living unconditionally is actually worse than the ko. As Bill shows above, when white chooses the option that leads to unconditional life, white gets very few points in the corner, black builds up quite a bit of influence, and white sacrifices sente. In thinking about this, I also looked up the 3-3 invasion in the GoGoD database of pro games. Even in situation in which black declines the ko, I was stunned at how often white would just tenuki. The pros seem to place very little value on just living in that corner until much later in the game. This main issue is that white gives up sente and at the stage in the game where this is usually played, sente is much more valuable than enclosed, small life in gote.

That said, if my experience is any guide, you won't see people playing the ko for a while. I'm in the mid-SDK range and I still don't think I've had a similarly ranked opponent play it.

Two other things that were not obvious to me at the time I was thinking about this:

1. In this position, why is it necessary for black to protect the cut with A?

Click Here To Show Diagram Code

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

If black did not have to protect the cut, black could kill white by limiting white to one eye with B. However, without something like A, white will then play the clamp at C, get out of the corner, and get a second eye. Two things to take from this -- (1.) normally, you have to protect the cut but (2.) if you have a situation in which you have support around say D, then it may be possible forgo A for B to kill. This is only possible if the extra support allows black to contain white after the clamp and prevent a second eye.

2. The logic above does NOT necessarily apply to the large knights enclosure. The reason emphasizes the important of sente in the assessment of the small knight invasion. With the large knight enclosure, white can live unconditionally in the corner while retaining sente. The normal sequence is the following:

Click Here To Show Diagram Code

[go]$$cB
$$ +---------------------------------------+
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . 5 4 . . . . |
$$ | . . . . . . . . . . . . X 7 3 2 O . . |
$$ | . . . , . . . . . , . . . . 6 X 1 8 . |
$$ | . . . . . . . . . . . . . . . . . 9 . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ +---------------------------------------+[/go]

And continues:

Click Here To Show Diagram Code

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

If black does not come back and connect at 5, white cuts his stones which can cause some significant problems. So, as long as the threat of being cut is sufficiently bad, this invasion is considered viable because white retains sente. In fact, some people will immediately play the 3-3 invasion after an opponent creates the large knight's enclosure. That said, black has other options to respond to the 3-3 invasion. So, you should look at those as well.

Hi BG,

My understanding is that if I ask a pro "what's the status of this corner?" and they reply "It's a ko." it means locally the invader cannot live unconditionally with perfect play by both sides. In this example, if chooses to let live, it means doesn't play perfectly locally, because presumably letting live is better globally.

So we have the be clear when we ask the question: is it local (say, an empty board), or is it global. The answerer should also make it clear; example, "Locally, can force to a ko, but for this particular whole board position, may choose to let live.", etc.

I think I'm missing something here. If white wants to live rather than fight the ko and we only consider local moves, even with perfect play, white either lives unconditionally in the corner (as Bill's sgf shows) or using the clamp described above, escapes into the right side with one eye in the corner.

I thought that was the gist of Solomon's explanation in this thread: https://www.lifein19x19.com/viewtopic.php?f=15&t=14273.

Where am I going wrong here?

Hi BG,

Sorry, I probably misunderstood this particular corner situation.
In general, my understanding of "It's a ko" is decribed in post 4.
Also, like your impression previously, "The best W can do is ko"
to me also means W cannot live unconditionally with perfect play by both sides.
However, it seems this doesn't apply in this case.
I feel this is an exception, but I wonder maybe it's more common than I thought ?
( I still maintain it's important for both people in the discussion to be very clear about what they mean. Mis-communication is quite easy. )

Some interesting replies. Indeed, the "best white can do" maybe doesn't mean he can't live unconditionally.

I've also been thinking about the following variation, is this seen as sub-optimal?

- White invades black's enclosed corner. Black decides to block the other way and use his 6-3 stone as the final move in the joseki in order to come out of it with sente instead of gote.




Of course, it depends on the board situation. But that way you "traded" your encloser stone as a way to gain sente after the invasion, which usually ends in gote.
On the other hand, you say the invasion of the enclosure ends in gote anyway for the invader? But that's only if you play optimally. The above variation is standard joseki, so it's less complicated.

If white lives in gote anyway, couldn't he exchange P18 for O18 first, and then come back to play Q18 to live in gote? Or is taking that point away from black not worth is because you make black a bit stronger with that extra stone? Or this probably removes the aji, too, of a later undercut depending on the situation?

After the White descent to Q-18, Black P-18 is sente against the corner. By comparison, after the hane-connect, W P-18, B O-18, W-Q18, Black P-19 is also sente against the corner. White's expected corner territory is the same in either case, but the effect of Black on the top side is different. When the pros do invade, I showed the more popular choice, but in some cases the hane-connect may well be preferable. It would be interesting to see what the strong bots say.

Edit: Oh, yes. After the White descent to Q-18, Black P-19 is also sente against the corner, and that may be preferable to P-18.

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

Visualize whirled peas.

Everything with love. Stay safe.

Hello?

Instead of taking advantage of our opponent, let's transpose to a standard joseki so that it's less complicated???

Please!

Seriously, take another look at the board after Black Q-17. Don't you wish the O-17 stone were somewhere else? White's invasion loses around one whole stone. Do you want to give that back? Or most of it?

Edit: Besides, should be another crawl on the right side to S-13. Then after Black extends to R-12 White can take sente.

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

Visualize whirled peas.

Everything with love. Stay safe.

So even if the whole board situation would almost require of black to build a wall facing the other way, you'd still block at R16?

I didn't mean to always do it like I suggested (with the enclosure stone as the final move in standard joseki), but it could be used that way, that was my idea/question.

The easy one first. White's hane-connect is bad precisely because it makes the O-17 stone good. OC, White could make a bad play.

Edit: Our new strong bots have shown that the hane-connect is not so good, anyway. But with the O-17 stone in place it is horrible.



I feel sure that it would be possible to construct a whole board position where Black should block at Q-17 instead of R-16. It would even be possible to construct a position where Q-17 is the only play. Let's take a look at the two plays.

Click Here To Show Diagram Code

[go]$$Bc
$$ --------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . W . . B . . . .
$$ | . . 1 X . . . . . ,
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]

exerts influence down the left side. It also nearly kills , enclosing it in the corner (or chasing it out in a low posture).

Click Here To Show Diagram Code

[go]$$Bc
$$ --------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . W 1 a B . . . .
$$ | . . . X . . . . . ,
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]

does not enclose , but it nearly does. It does not exert very much influence along the top side, since is already there. For not to be superfluous, Black probably needs an eye at "a", which would mean that Black is under severe attack, so that living is more important than exerting influence. I have trouble picturing such a position without a number of White stones in the vicinity.

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

Visualize whirled peas.

Everything with love. Stay safe.

I had a quick look and it seems there are a few professional examples. In some cases black blocks the 'wrong' way for the opposite reason to trying to live - if black is strong locally, it's one way to try to kill white!

Here's one of the games where the corner was most isolated, there are no white stones very nearby but black still chooses the unusual direction. The position appears at move 112. Black does win, for whatever that's worth.

Thanks! Very interesting. I wondered about a large White wall, but in this case there's more to it than that.

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

Visualize whirled peas.

Everything with love. Stay safe.