The accepted answer is to play at A6, and I see that it results in a Ko situation, but I don’t understand how that is advantageous. It seems like black would respond with B6, threatening the potential eye at C5. I don’t see how white would get out of that situation.
I think you are confused about what a ko is. After black takes the stone, white needs to find a ko threat elsewhere instead of continuing to respond locally. If black answers the ko threat, then white comes back to take the ko again.
No, Sam, the board does NOT look like that, it is NOT a 9x9 board, look again at your screenshot in your opening post – there it is a part of a larger board.
And on that larger board, White plays some threat somewhere outside of the partial board, which, so W speculates, is important enough for B to answer there, so that now W can take back the Ko here, which B cannot immediately retake, instead B must now look whether they can in turn threaten W somewhere outside of this local situation. Und thus a Ko fight happens
Ah, apparently I was misunderstanding the context and objective of the puzzle.
I wish the instructions made it clear that I could assume that there was a threat I could make in a hidden part of the board, to entice B away from my eyes. Otherwise, this problem is very confusing and misleading.
This is still not understanding the problem and context of tsumego. Considering this position in isolation, it is a ko. That’s the answer of the local problem. Being able to think like this is a powerful abstraction technique.
White plays one of G3,G4,G6,G7 . If Black plays again on
the left, then White plays again in G3,G4,G6,G7 , getting a
decent-size group on the right, rather than barely living on the right.
(Black still wins by a huge amount, but in terms of points
this is much better than letting the left die without ko.)
Ah! So my W group was already basically doomed, before playing A6, and the ko is basically a distraction technique?
The problem says to “try to rescue the marked stones…” So am I understanding correctly that it’s not actually possible to save them, unless B decides not to play B6? … Is there any reason B wouldn’t want to play there?
Hmmm… I think it’s starting to make sense. So the threat W is making by playing at A6 is to capture B’s A7 (by playing A8), and they’re basically saying, “you’d better stop me from rescuing the group, (and while you do, I’m going to stake my claim to a bunch of territory on the right side).” Do I have it right?
You need to understand that there are different kind of results to a tsumego. Being able to live/kill unconditionally is of course the best, but if that’s not possible, having a ko is also better than not being able to live/kill at all.
Which one is a distraction technique for the other, depends on the size of the plays elsewhere.
Even on the 9x9 you gave, although winning the ko would be a simpler way for Black
to win the game, it may well be better in points for Black to answer on the right:
If White wins the ko, then Black likely gets the rest of the board.
“So am I understanding correctly that it’s not actually
to save them, unless B decides not to play B6?”
No: Even if Black plays B6, there’s still the question of Black’s next turn.
Will Black win the ko, or respond to White’s move elsewhere?
“Is there any reason B wouldn’t want to play there?”
Yes: Black might think White started the ko too early. Temperature at Sensei's Library
(The A6-B6 exchange helps White: The difference is whether-or-not
[the next move in this region being by White] would leave Black ko-banned.)
“So the … it right?”
Black responding to White’s move elsewhere
on the 9x9 you gave, may well
and
in general, very-easily can
be better for Black than Black stopping White
from rescuing White’s group. Otherwise, yes.
The ‘with a ko’ makes the rescue conditional, so I wouldn’t say that text is actually wrong, but yes it can be misleading to a beginner unfamiliar with the phrasing and kos. A more verbose and clear wording would be:
Create a ko shape from this position, such that a ko fight would ensue in a real game were this position part of a whole board, such that if you win the ko fight you rescue the stones, and if you lose the ko fight you get 2 moves in a row elsewhere.
Your response helped me to understand it, I think.
Allow me to check my understanding:
B cannot unconditionally save the group; it can only threaten to do so by playing at A5, drawing W into a ko fight and thus buying “2 moves in a row elsewhere,” right?
The wording for the puzzle is confusing to me because it made me think that I could save the group for good by successfully forming two eyes.
Is there a term to distinguish groups that are “immortal” from ones that are just “not dead yet”?
Well, the problem says “Make the black group alive USING KO”, so it’s not really wrong. Like you said, ko is not really alive as black needs to find a ko threat elsewhere on the board. I guess trying to express it accurately would be too wordy.
It’s not really right. A ko usually doesn’t make a group alive, only a double ko does.
The problem with the wording in these two puzzles is that you have to understand the solution in order to understand the instruction. Which is far from ideal for learning. (It’s only okay for evaluating what someone has learned.)
Problem for me is that locally there is no KO threat to use. My pov is that this is something far too conceptual for a beginner (the existence of ko threats with a size enough big and in bigger quantities as for the opponent in another part no yet visible ) . I would not introduce or ask about ko fighting of this level of abstraction. At least we should see the threats and I guess there are easier positions to work on a Ko fight.
Hey samhiatt, it’s great to see you asking questions and then coming back to discuss the responses! Keep up the good work (I’d almost given up on this forum, because too many people just say “explain this to me” and then seem to think they’ll magically improve with no further participation. Most don’t even bother to say thanks for the reply. Kids these days…)
The good news: in practice, all you need most of the time is the three concepts of alive, dead and unsettled. The weird stuff and the fine distinctions don’t come up often enough to affect how you play. Even ko isn’t that common below dan level; it’s in the problems more for the mental exercise, and to make sure you’re clear on the rules, not for its practical value.
The other good news: if you do enjoy learning theory and concepts, it just gets deeper the more you look at it, and you can learn all sorts of fascinating things. Yes, you can make a distinction between “alive for most practical purposes” (if you attack me then I know how to defend) versus “totally alive even if I ignore your attack”. The latter is sometimes called pass-alive, although this isn’t a common term.
You can also make the distinction between unsettled (either alive or dead depending on who plays first) versus alive in ko (the types of problems that started this discussion). And then there’s the weird and wonderful things like:
and more. For most amateur players, these things are interesting but not useful, so don’t spend too much time on them. Keep playing games, working problems and asking good questions!
Like Jon_Ko - and maybe the other replies here, I haven’t
read them in detail - I think the problem’s wording is bad:
I think it should be either just “Black to play.” ,
or something like “Black to play. Try to live.” .
Yeah, I agree that seeing them would be very helpful for understanding what I’m supposed to be learning from these puzzles.
As hoctaph explained above, maybe there is a case to be made that claiming territory on the right (by playing G3, G4, G6, or G7, assuming a 9x9) could be considered a ko threat. But that’s really hard to see and understand, as a beginner.
Thanks xela!
I was hesitant at first to ask my follow-ups, not expecting people to engage with such detail.
It’s also quite a challenge figuring out how to word my questions; but I’m realizing that learning to speak the language of Go is a big part of learning to play. And it takes practice!