134 lines
3.0 KiB
Scilab
134 lines
3.0 KiB
Scilab
|
# This regression file covers mistakes made in the Computer Go Ladder
|
||
|
# game between GNU Go 2.7.168 (white) and NeuroGo II (black) on
|
||
|
# September 9, 2000.
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 20
|
||
|
1 reg_genmove white
|
||
|
#? [Q3]
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 22
|
||
|
2 reg_genmove white
|
||
|
#? [R3]
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 34
|
||
|
3 reg_genmove white
|
||
|
#? [H3|K3|K4]*
|
||
|
|
||
|
# Tenuki is acceptable. In the local area the game move is bad.
|
||
|
loadsgf games/FSF-neurogo.sgf 40
|
||
|
4 reg_genmove white
|
||
|
#? [D6]*
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 50
|
||
|
5 restricted_genmove white F8 F9
|
||
|
#? [F9]
|
||
|
|
||
|
# Tenuki is perhaps acceptable.
|
||
|
loadsgf games/FSF-neurogo.sgf 86
|
||
|
6 reg_genmove white
|
||
|
#? [C14]*
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 90
|
||
|
7 reg_genmove white
|
||
|
#? [O2]*
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 92
|
||
|
8 reg_genmove white
|
||
|
#? [O2]
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 124
|
||
|
9 reg_genmove white
|
||
|
#? [A18]*
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 136
|
||
|
10 restricted_genmove white A3 B4
|
||
|
#? [B4]
|
||
|
|
||
|
# Semeai problem.
|
||
|
loadsgf games/FSF-neurogo.sgf 148
|
||
|
11 reg_genmove white
|
||
|
#? [N5]*
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 152
|
||
|
12 reg_genmove white
|
||
|
#? [B10]
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 176
|
||
|
# This may be hard to fix at this point but since we
|
||
|
# can't win the ko we shouldn't start it.
|
||
|
13 restricted_genmove white P7 M7 M8
|
||
|
#? [M7|M8]*
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 178
|
||
|
14 reg_genmove white
|
||
|
#? [Q4]*
|
||
|
|
||
|
# White A16 is about three points reverse sente. G19 is two points
|
||
|
# double sente, but black can get in A16 in sente before answering
|
||
|
# G19. We accept both moves as being correct.
|
||
|
loadsgf games/FSF-neurogo.sgf 230
|
||
|
15 reg_genmove white
|
||
|
#? [A16|G19]
|
||
|
|
||
|
# It doesn't look like black can live, so this would only be a
|
||
|
# reinforcing move. It may not be worth regressing over.
|
||
|
loadsgf games/FSF-neurogo.sgf 276
|
||
|
16 reg_genmove white
|
||
|
#? [H17|K17|H15|L16]
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 286
|
||
|
# K19 makes ko. H18 kills right out.
|
||
|
17 reg_genmove white
|
||
|
#? [H18]
|
||
|
|
||
|
loadsgf games/FSF-neurogo.sgf 288
|
||
|
18 reg_genmove white
|
||
|
#? [K19|H18|K16]
|
||
|
|
||
|
# Move 12
|
||
|
# M17 is an odd move. The local good shape and
|
||
|
# connection is at O16, but in order to really kill the black corner,
|
||
|
# a move at S16 is called for.
|
||
|
loadsgf games/FSF-neurogo.sgf 12
|
||
|
19 owl_attack Q17
|
||
|
#? [1 S16]
|
||
|
20 owl_defend Q17
|
||
|
#? [1 S16]
|
||
|
21 reg_genmove white
|
||
|
#? [S16]
|
||
|
|
||
|
# Move 18
|
||
|
# H16 looks strange. Locally better to make a wider extension towards
|
||
|
# the corner stone. Globally a move like R7 is more urgent,
|
||
|
# strengthening the lone lower right corner stone and keeping the
|
||
|
# black stones on the right side weak.
|
||
|
# (no actual test case yet)
|
||
|
|
||
|
# Move 34
|
||
|
# J3 is considered owl-dead and is therefore abandoned. This seems too
|
||
|
# pessismistic, however.
|
||
|
loadsgf games/FSF-neurogo.sgf 34
|
||
|
22 owl_attack J3
|
||
|
#? [1 K3]
|
||
|
23 owl_defend J3
|
||
|
#? [1 (K3|H3)]
|
||
|
|
||
|
# Move 88
|
||
|
# The C15 stones can be saved by a move at F16.
|
||
|
loadsgf games/FSF-neurogo.sgf 88
|
||
|
24 owl_attack C15
|
||
|
#? [1 (F16|E16|F14)]
|
||
|
25 owl_defend C15
|
||
|
#? [1 F16]
|
||
|
|
||
|
# Move 96
|
||
|
# The B16 stones can no longer be saved. Better to make connection at
|
||
|
# O2.
|
||
|
loadsgf games/FSF-neurogo.sgf 96
|
||
|
26 owl_attack B16
|
||
|
#? [1 (PASS|B17)]
|
||
|
27 owl_defend B16
|
||
|
#? [0]
|
||
|
28 reg_genmove white
|
||
|
#? [O2]*
|