I’m reading Alexander Kotov’s excellent book Think Like a Grandmaster, and it occurred to me that a lot of Go players also don’t know how to analyze variations systematically. Kotov describes a method where he would set up a complicated position, analyze all the variations, write them down, and then establish which of them deserved consideration within the demanding conditions of timed play, and which could be left out to save time. Normally a master decides this by intuition, but most of us have to develop our intuition. He developed his by practicing this method and then successfully tried out the same method in study groups, so I thought it would be fun to try it here.
Here are the rules:
If this sounds interesting to you, no matter what level you are, please practice with us on the position(s) below:
My initial analysis was pretty wrong. I added some joseki lines that I looked up when I was done studying on my own, as well as the game line with some comments (I looked it up when I was done with my self-review.)
that was really hard… and i dont doubt my analysis is quite wrong .
i gave myself 25 mins, but found i was thinking about the resulting positions more than actually reading different variations . in the analysis i stuck to my preselected moves (you could call it intuition) and didnt have time left to try others.
anyway, i found that 2 of my 4 initial ideas probably werent worth the trouble and i discarded a third after a bit more thought and wouldnt want to leave it out of consideration.
great idea btw
count me in for the future, if there will be more.
Spoiler. Don’t read if you haven’t participated yet and want to.
One thing I noticed so far is that no one found the pro move n3. And no one found LeelaZero’s move j4. And no one found Zen’s move k4, which was also CrazyStone’s move. None of these got any consideration whatever. Everyone found l3, a move that LeelaZero thought was better than Zen’s k4, but everyone discarded it, with the possible exception of violaine who didn’t say one way or another.
Spoiler. Don’t read if you haven’t participated yet and want to.
L3 did look possible to me, but a bit passive. I didn’t think of the bots moves, which put more pressure on b but same time leave more weaknesses. I got that direction to push b to the wall.
Here’s my analysis, but consider it more of a sample answer since I checked the answers beforehand and that affected my process. I’ll do my analysis first for any future positions.
The unhelpful answer is: intuition. The truer answer is that attaching would kill the aji there. I can’t exploit the weak connection later if I induce Black to fix it by attaching now.
I only found two candidate moves that I really cared for. I’m feeling more lost on this one than the first, probably a good indicator that I need to work my midgame skills, which I’ve known for a while…
All I would really care about at this point is the balance of power in the lower left.
C9 is tempting as it prevents white eyes on the side, but then white can counter attack around F6. So I’d rather have it the other way around, pressure white from the outside and force a reply at C10.
I’d love to play G4 but I don’t see how black can resist a double hane. So I would take some distance and play H5.
Park Junghwan has a very interesting and instructive style. As opposed to super sharp, super aggressive players like Mi Yuting, he tends to play very solid shapes, without many weaknesses. Unlike Lee Changho’s “play solidly and then win the endgame,” though, he takes his solid shapes and uses them to cut up his opponent’s groups, often starting big ko fights and winning them. His ko-fighting successes, of course, are in no small part due to the lack of ko-threatable weaknesses he tends to leave behind, compared to some of his opponents.
.
. in the analysis i stuck to my preselected moves (you could call it intuition) and didnt have time left to try others.
