G is less strong but can make life on the lower side if White adds a move somewhere between C and G.
B has no eyes but is strong compared to A since it’s one move ahead, so can live either by attacking A or by connecting to C.
A and H have no eyes and are surrounded by strong white groups.
This situation is bad for Black but not hopeless, players around 16k make sufficiently big mistakes so that a big disadvantage like that can be overcome. But Black needs to make a good strategic decision. Which group to defend? A or H?
It’s better to defend the group H, which is the biggest. Try to make eyes towards the lower side, keeping in mind that the stone G can be cut. Perhaps cut with L4, or play a move around H5 which makes shape and at the same time protects against the cut J7 and starts to threaten stone G.
The group A is less important. Not only because it’s smaller (2 stones instead of 7), but also because it separates two groups, one of which is very strong and the other not really weak. The stones A just represent a few points but don’t affect much life and death of other groups, and probably a move later on the right side is worth at least as much as saving the stones A.
Oh another idea I remember from the EGC in Toulouse, but I don’t think it was recorded, was I think Antti talking about the opening and how adding a stone to the weakest group on the board (yours or the opponents) can’t be a bad move.
Then he tried to explain how you can tell if a group is weak, by approximately counting how many tenukis you can do against reasonable moves.
If the move is unreasonable, like playing a stone in a bad spot that can be captured for free or if it’s that the opponent wouldn’t ignore an atari or throw in and easily captures something important, then you can assume they’ll answer it.
So you’ve to count how many “reasonable” looking moves you can play to surround or approach or take away eyespace and then use that as a measure for how weak the group is.
Probably if it takes three, four, five+ moves to put the group in danger, it might not be that weak. You might even be able to save it at any stage possibly, or at least help it get to safety in some way.
You might not even need to live, if those 5 tenukis secured more than ~30 points, could just be a trade. And then you also get all the aji on the outside threatening to connect and revive the group later.
I like Antti’s approach too (it’s even my current development goal to keep that always in mind).
Following @shinuito’s reasoning I’d think: Okay, so this group can withstand at least four moves, maybe five. What about the other groups on the board? The 3-3 stones can withstand three moves each (double low approach and 5-5, then they need an answer). White’s lower lefg is in danger after three to four moves.
So I’d play at two 3-3 corners, if white attacks my left group in the meantime as shown. It is then time to defend the group, because it can only withstand two to three moves now (four to five minus two). But defending can also mean attacking the white stones or groups in the vicinity.
Remembered this thread while playing a game just now. I love killing groups, and 9x9 generates a lot of opportunities! Here was my 7k thought process (as Black)…
First, trying to split White apart makes it more likely that something will be killable:
Then, attacking the weaker side:
At this point, I had no idea that the top group was killable and was about to switch my attention to the bottom one. My first instinct was to play A to simplify that area, but I read out B and realized that even though it seemed risky, I actually wouldn’t get killed right away because I could trap the single white stone.
At this point I finally saw a path to kill. White will eventually play A and get an L-group. At my level I know that’s “dead” but am not very confident in actually being able to kill it, especially under time pressure. This one also has a kind of escape route to the left that looks dangerous.
Anyway, we fought over that escape route, then a few moves later the L-group appeared as planned, and I knew I should be able to kill.
White tried to live, and it took me more than a minute to read out the correct response, but it really helps to know that there is one.
