Home > Steve > Go > Problem of the month

January 1997's Go Answer of the Month


Dia 1 & 2 - Is this valid? What should have happened.

Dia 1   Dia 2

The following situation arose in a handicap game I was playing at Reading a week or so ago. The sequence shown here is the sequence played.
In this sequence 14 (at 6) took the ko (11); 15 was a ko-threat; 16 a ko-threat-answer; 17 (at 11) retook the ko; 18 another ko-threat; 19 at the marked stone captured the black string.

Is this a valid sequence? Could Black have done better? If so, how?
What happens to the remaining Black stones in the corner?

Note that as played it involved a ko - the bane of all kyu players!


When reading situations, you have to be methodical about it. I try to be, but like everyone I make mistakes sometimes and my method isn't exhaustive - just exhausting.
To that end I reckon the way to do it is to choose a likely sequence and see what happens. If it has Black doing badly, then you look at Black's last move and try another one and see what comes of that. If all the sequences from that Black move turn out badly for him, you try a different move for the previous Black play, etc etc.
When you find a Black move with a sequence that white can't counter then you try a different White move for the one before that. Effectively this means that you start at the end of the sequence and work your way backwards through it justifying (or not) each move.
As I say, it can be exhausting, even with a computer.

To that end, lets start by looking at the continuation from Diagram 2 (whether or not we should have gotten there).


Dia 3 - A Repeat of Dia 2 with captured stones removed.

Dia 3

Dia 3 is what needs discussing. My "exhaustive" methodical approach started by studying Black 20 at A, B and C which lead to either a seki or a ko. Only much later did I think to try D and E which let Black live with eyes (E is bigger).

Dia 4 - How Black lives.

Dia 4

If White attacks with 2 at 3, after Black 3 at 2 White has no further attack. The latter is better for White than Dia 4.

Dia 5 - One of the sekis.

Dia 5

I've included this diagram because it surprised me and I found it an interesting seki.

Note that this seki relies on Black not connecting the two halves of his group.


Now back to the main question - should this situation have arisen in the first place?
A repeat of Dia 2   Dia 2   14 at 6; 17 at 11; 19 at triangle.

As it stands, with move 19 in Dia 2, Black has lost a lot. So we need to see if there are any better moves for Black 18.
Being a ko threat outside the locality we can't answer this one, so lets assume it was OK and look for a better 16.
16 was a ko-threat-answer, again outside the locality. So let's try a better 14.
13 was atari so 14 was forced. Try 12 - 11 was atari as well. Try 10.

There are some possible alternatives to 10. 12 is one, above 7 is another but it soon becomes apparent that they are all near enough equivalent or just shuffle move order. Try 8.
Other 8's are rebutted by White at the 2-2 point. Try 6.

Dia 6 - A better Black 6.

Dia 6

This Black 6 turns out to be a good move (the original 6 was at 7 here). There is nothing White can do about the shortage of liberties that he finds himself in.


So now it is White's turn to seek a better 5.

Dia 7 - A seki from this White 5.

Dia 7 13 at 6

There are also variations here which let Black choose to play this as a ko.


So now we look at Black 4. And yes - a better one that I have found no white counter to.

Dia 8 - The best Black 4 - White's attack fails.

Dia 8

So as it turns out, White got away with a rip-off!


Home > Steve > Go > Problem of the month
British Go Association
Last updated 2004-08-10
This page is part of http://www.stocton.org/
Email: webmaster@stocton.org