Killing a large group

Just finished an interesting game:

Can anyone figure out at what point the large black group in the middle is actually dead? How and when should black have defended, and could white have killed it earlier/more efficiently?

I had been assuming through most of the game that it was going to escape and live but that I could use chasing it to get some profit, but my opponent didn’t notice the danger until too late. Incidentally, I don’t seem to have been able to get very much profit outside, but I was playing very defensively once I thought I could kill (and wanted to just finish the game rather than contest irrelevant fights).

The opening looks fancy and way more elegant than what I can afford in my games - it reminded me handicap games where a 9d does this sort of stuff to confuse a weaker 1d player. All this elegant stuff was then quickly dumped by a few dumb moves from both sides, though :slight_smile:

In the mid game it’s hard to say what both players were thinking about: it looks like a random mess to me. Perhaps both players tried to achieve local short term goals (like capturing these 2 stones or saving that 1 stone, no matter what was happening elsewhere). This is what I was personally doing when I was 15k.

The status of the group in question changed probably after every move: either player could make a proper move to save/kill the group. See my simple variation where B takes a huge W group with a simple sequence of 3 moves. Same for many other big groups: they could be killed or saved with a proper move. This is why I think that it wouldn’t be correct to say anything about the status of the group: neither player was concerned about it and the group just “happened” to be captured, probably to the surprise of both players.


Black made some big mistakes, losing many stones towards the end of the game including the center group.

Black could have played P6, instead of playing Q7 and losing it later. It would have worked well in combination with Q5 (rather than trying to save 7 points with J6).

Black’s best option to live was actually to kill the upper white group, starting with a move at K19. After white G19, I think there’s still a ko at G17. After white L19, the upper white group is alive.

Thanks, SanDiego. During the game, I didn’t feel the ko at G17 was crucial to living on the top, as if black took it, I would just connect at F15 and then losing the ko then isn’t immediately fatal, but I see now that it does reduce my eye space to allow the kill starting K19. I know it’s a problem of mine that I analyse each element separately, so I had looked at both the ko and the eyespace at the top and decided each was OK for me, but I didn’t look at the effect of losing the ko on the eyespace.

I thought that black had plenty of space to make eyes below the center if they had tried, far later than they could kill the white group at the top. In the game I was worried about black M6 until quite late on. But in any case, killing the big white group is something black should have done, or tried to do, even if its own big group didn’t depend on it.

I think the playing style in the midgame was too risky on both sides. Groups kept getting cut off and kept running aimlessly. IMO, a better style of play is to use “poor” shape to make eyes, so most of your groups have some stability, if not actual unconditional life. Instead of always running in a linear string, make some empty triangles or some diagonal connections. Also, make use of live groups as bases for extension and threats. Watch for influence and sector lines, not just for connection. Just my opinion, I’m only about 18k. If you haven’t already, read some Go books. I own about 25 and they actually help.