Re: [gnugo-devel] 100 game series

[Daniel Bump]

> Particularly annoying was the tendency to make "two space diagonal
> jumps", leaving a big gap in the position. The worst example of this
> is P6 but also K14 and N10 are similar. I guess one way of solving it
> is to make the pattern Intrusion21 more powerful, but I leave it to
> Arend to try to tune this.

Perhaps we should tune to reduce this propensity but not to
eliminate it.

In some cases this is good behavior. In fact, I would argue that
P6 is a very good move. The capping move is like this is good if
the capped stone is not alive and the capping move either swells
a moyo or makes possible a splitting attack. If B goes through
the opening, the intention is to fall back with a keima, thus:

. . . . .
. . . . .
. . . * .
. O . . .
. . X . .
. X . O .

and this is exactly what happened in the game. For examples
of this see Kato's ATTACK AND KILL, p.80 dia.1 and p.124 dia.1.

In the game in question, W plays this way with move 50.
Owing to the strong shape in the corner, there is little
B can do. The only way to cut is

B:Q6 W:Q7 B:S5

after which W captures one stone in the corner and B has
accomplished nothing. So in my opinion P6 is a good move
but this depends on characteristics of the position that
GNU Go doesn't currently understand.

Dan