When reviewing my games i try to understand why a move is considered important (opposed to my own failure…). Most times i can explain, mostly simply territory or a good follow up.
In situations like in the game attached (around move 30 on onwards) i remain puzzled why AI considers crawling for the top left group more important than “big points”. I see that this denies White a sente, but will end in white getting a big point anyway. In addition i see no whopping follow up. On the other hand, when white gets this point the corner still lives wo problem.
You mean move H18?
I don’t think that playing other big moves around move 30 was bad, you made bigger mistakes than that in the game.
Crawling is a reverse-sente move. In general, reverse-sente moves are bigger than they appear.
Once White has played hane and connect (B17, B18, B16), crawling is a good move in any case. But even more so a few moves after because if Black crawls and White doesn’t respond, then Black can hane at J17 and start to attack the two stones O17.
I’ll comment at move 32. where it is most clear
Crawling denies White a sente which is also a big move, because it is a sente where White closes a territory that is already closed on the other side. The last move to finish a territory is the largest. Crawling also has a big follow-up which is double hane to attack this other white stone.
It would be way less interesting without the other White stone (top right), so it is a matter of context.
At this point, I don’t know what AI says but playing somewhere around D10 would also look reasonable, and your sequence is not my style but looks fine. I wouldn’t call crawling an only move and other moves mistakes. Does Katago strongly favor it? If it is less than 2-3 points I would not think of it as a mistake, rather as opening your eyes to another possibility.
PS: for what it’s worth, I don’t undertand the comments about it not being joseki. What Black played in the top left looked perfectly joseki (the modern computer version) to me.
Note that by the time the AI is recommending H18, you’re well out of joseki territory.
Yes, after three stones that stops for black - that was the question.
The move itself seems not to be “important” in itself (it would have been proposed earlier). It is not until white has haned in the corner and a white stone in the upper right would help white enclose the top otherwise. This i understand now.
But still this feeling i can not judge why this “local point” is considered about equal to extending the influence on the left.
thanks for the pointer to “josekis at OGS” i use this a lot.
thanks for your analysis - that is about what i wanted to understand.
C10 is considered about equal to h18 and the left influence seemed to me “bigger” than the crawl. The question about joseki was only when it stops (why h18 is not considered part of it)
To state it a bit more strongly: in this sequence H18 is not only not joseki, it is bad in general (in a random position, if that makes sense). For every exchange (crawl - extend) the infuence outside is more valuable than the single point of territory you are taking.
OJE only has extend, not hane.
I see that I misunderstood what you meant. When I read “being out of joseki” I usually think that one of the player playes a move that is different from the classical ones. In this case they finished the joseki then other moves have been added. I can see the ambiguity now.
I think there are plenty positions where this is false. If blocking on the second line is sente for the player with the outside wall (due to life/death status), then crawling on the second line is often important, to prevent such a block in sente.
This is a tricky question to answer, because whether it’s considered Joseki or not, may depend on who you ask.
In this case I personally would answer “Because the Hane at B17 is already not part of the main Joseki, rather it’s one possible Joseki follow-up. After B17, B18, B16, then H18 is Joseki, in the sense that it’s often optimal.”
You’re right to think the left is big. But Black seems to get a decent amount in this trade: the lone stone at M17 gets backup, and now the white group in the top right needs to start taking care of itself. The variation recommended by AI shows white takes the left, but then the top right becomes area of active play.
):
