yet even better ko?  

How about this? is an interesting move. White has to be very careful in order to kill Black; looks like the only move.

puts White on the spot. Either let Black live or fight expensive ko. If White decides to go for the kill with then Black even gets the first ko capture at a. Can anyone find something better?

nando: If at , life still doesn't look obvious to me. What is the sequence please ?

here fails  

After Black answers at . Best White can get is a (sente) seki up to .

nando: thanks Arno, but I still have a doubt: what if at now ? (sorry to bug you, I must be missing something again)

Arno: you are right, at kills Black.

Notochord: (Shouldn't ever ever have to be ko. Black has lots of eyespace and eyeshape, so it would be surpising to see him forced dead, and it is also rare to see such a relatively large eyespace made seki. I think should be played differently. Black only needs 1 eye on the edge.)

Still works this way, I think.  

unkx80: at b lives. But...

Notochord: Sorry: you are correct, I misread this variation. 'b' for seems to be the only move. is a killing move here.

And this way, I think.  

I'm very bad with calculational mistakes since I tend to operate a bit lopsidedly on the more intuitive end of things, so feel free to pipe up if I am starkly wrong.-

Just connecting directly should also pretty clearly work, since its practically the same thing, and the most vital point.

Does not work.  

unkx80: Here , and kill by the usual eyespace reduction method.

Notochord: I'm not sure that needs to be obeyed. needs to go at the best place for eyeshape.

Noto: Does not work?  

unkx80: Sorry but I can't help feeling a bit tired explaining this again and again. (But really, it's not your fault.) I shall leave it as a reading exercise this time. If you can solve Beginner Exercise 196, then you should be able to solve this as well.

Notochord: I see. Thanks.

Alive? (3rd variation)  

nando: What if at a ?

Arno: yes a a kills Black.

Starting with the connection?  

Dave: Quite and interesting position. In such a cases, I think the first question to be answered is what if Black simply connects? This threatens to force one eye along the edge. Also, as we see above a White throw in at a is frequently a very damaging play against Black's eye space. However, it is important to realize that Black has resources here as well. An early Black play at b, threatens to force a large ko at c while preventing the throw in. I haven't found outright life for Black yet but there are some interesting variations along the way. Let's see what people can find. :-)

Black lives.  

unkx80: This is frying my brain when I am supposed to be working on my thesis. ;)

I don't gurantee the correctness yet though. ;P

The long sequence. (2 at marked stone)  

Dave: What a great idea! I had only considered the case where playing 5 as soon as possible starts a large ko at 8 where Black has a huge number of internal ko threats.

The long sequence. (6 at 1)  

Dave: So White has to give way and play at something like a allowing Black to live with ?

Loses the fight in the end.  

Dave: Black ends up short of liberties so...

Beyond the edge 1.  

Dave: Black first turns aside at threatening to capture the center stones. After , Black can not cut with because White a will create a snapback. However, is a real threat that White must answer. Then the question comes up whether White has to play (in which case wins the larger fight). If would not be atari, White can plan herself and win.

Beyond the edge 2.  

Dave: Here also forces and then we need to know whether is atari or not.

Arno: as I have now added the whole board you can see that is not atari so the sequence does not work. Sorry. But I found a new line:

The long sequence: alternative  

looks like winning it. It threatens to capture the two stones, so White's best guess is . threatens the cut off the large group so White has to . But then makes miai of a and b.

The long sequence: B dead  

As the previuos line fails White plays at here. Black is short of liberties and cannot prevent the large White dragon from connecting to the outside. So Black is dead. Backtracking ...

The long sequence: finally alive?  

Instead of playing a immediatly, Black plays . White has to play , but then makes miai of a and b so Black finally lives, or doesn't he?

Arno: I looked at the position some more ...

Black is dead  

Because of the cutting point at Black can neither make a second eye here, nor capture the White dragon. I looked at the moves a - e (the last one with followups f and g. In these variations the stone provides the crucial liberty for the white top group in the semeai).

So I think the the long sequence (in all variations) fails.