271 lines
5.9 KiB
Scilab
271 lines
5.9 KiB
Scilab
|
#--------------------------------------------------------------
|
||
|
# Not sure it's the best move on the board, but it should be
|
||
|
# considered. Unless I missed something, it looks like a good
|
||
|
# reducing move.
|
||
|
#CATEGORY=STRATEGY
|
||
|
loadsgf games/nando/auto001.sgf
|
||
|
1 reg_genmove white
|
||
|
#? [M12|N11]*
|
||
|
|
||
|
# Uh ?!
|
||
|
#CATEGORY=OWL
|
||
|
loadsgf games/nando/auto002.sgf
|
||
|
2 owl_does_defend A14 B13
|
||
|
#? [0]*
|
||
|
|
||
|
#CATEGORY=OPTICS
|
||
|
loadsgf games/nando/auto003.sgf
|
||
|
3 owl_defend B11
|
||
|
#? [1 A12]
|
||
|
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/auto004.sgf
|
||
|
4 disconnect S7 S11
|
||
|
#? [1 (S10|T9)]
|
||
|
5 same_dragon S7 S11
|
||
|
#? [0]*
|
||
|
|
||
|
# Problem with multiple lunches
|
||
|
#CATEGORY=OWL
|
||
|
loadsgf games/nando/auto005.sgf
|
||
|
6 owl_does_defend S2 Q2
|
||
|
#? [0]
|
||
|
|
||
|
#CATEGORY=ATARI_ATARI
|
||
|
loadsgf games/nando/auto006.sgf
|
||
|
7 reg_genmove black
|
||
|
#? [N16]
|
||
|
8 reg_genmove white
|
||
|
#? [M16|N16]
|
||
|
|
||
|
# Beware of the double snapback.
|
||
|
#CATEGORY=ATARI_ATARI
|
||
|
loadsgf games/nando/auto007.sgf
|
||
|
white A1
|
||
|
black B5
|
||
|
9 reg_genmove white
|
||
|
#? [E1|E2|F1|G1|G2]
|
||
|
|
||
|
# GG (as of 3.3.15) doesn't even try a single move. On the other
|
||
|
# side, GG doesn't have (yet) any mechanism for detecting a
|
||
|
# "connect either" combination.
|
||
|
#CATEGORY=OWL
|
||
|
loadsgf games/nando/auto008.sgf
|
||
|
10 owl_defend D12
|
||
|
#? [1 (C11|F11|C12)]*
|
||
|
|
||
|
#CATEGORY=OWL/OPTICS
|
||
|
loadsgf games/nando/auto009.sgf
|
||
|
11 owl_attack S8
|
||
|
#? [1 (S2|T2|S1|T4|T3)]
|
||
|
|
||
|
#CATEGORY=OWL/OPTICS
|
||
|
# Note that K12 connects out (sacrificing L11 if necessary).
|
||
|
loadsgf games/nando/auto010.sgf
|
||
|
12 owl_attack L17
|
||
|
#? [0]*
|
||
|
|
||
|
#CATEGORY=OWL/OPTICS
|
||
|
# (If white B2 in reply to B1, Black A1 kills cleanly. /ab)
|
||
|
loadsgf games/nando/auto011.sgf
|
||
|
13 owl_attack A4
|
||
|
#? [1 B1]*
|
||
|
|
||
|
# W should better counter-attack
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/auto012.sgf 90
|
||
|
14 reg_genmove white
|
||
|
#? [P7]*
|
||
|
|
||
|
# Still the better choice, S7 only strengthens B.
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/auto012.sgf 92
|
||
|
15 reg_genmove white
|
||
|
#? [P7]*
|
||
|
|
||
|
# D5 is just stupid
|
||
|
# (but A5 doesn't work. /ab)
|
||
|
#CATEGORY=OWL
|
||
|
loadsgf games/nando/auto012.sgf 124
|
||
|
17 reg_genmove white
|
||
|
#? [B4|B5]*
|
||
|
|
||
|
# Non-transitivity problem.
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/auto013.sgf
|
||
|
18 disconnect D6 H6
|
||
|
#? [1 E6]
|
||
|
19 same_dragon D6 H6
|
||
|
#? [0]*
|
||
|
|
||
|
# S8, T12 and a lack of liberties are too many weaknesses.
|
||
|
# W must defend.
|
||
|
#CATEGORY=ATARI_ATARI
|
||
|
loadsgf games/nando/auto014.sgf
|
||
|
black S16
|
||
|
20 reg_genmove white
|
||
|
#? [S8|S9|T12]
|
||
|
# With another move ordering
|
||
|
loadsgf games/nando/auto014.sgf
|
||
|
black S8
|
||
|
white T9
|
||
|
black S16
|
||
|
21 reg_genmove white
|
||
|
#? [S9|T12]
|
||
|
|
||
|
# Another non-transitivity problem.
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/auto015.sgf
|
||
|
22 disconnect L3 N2
|
||
|
#? [1 M3]
|
||
|
23 same_dragon L3 N2
|
||
|
#? [0]
|
||
|
|
||
|
#CATEGORY=OWL
|
||
|
loadsgf games/nando/auto016.sgf
|
||
|
24 owl_defend P17
|
||
|
#? [1 M13]
|
||
|
|
||
|
#CATEGORY=BLUNDER
|
||
|
loadsgf games/nando/auto017.sgf
|
||
|
white A8
|
||
|
black A10
|
||
|
25 reg_genmove white
|
||
|
#? [!P16]
|
||
|
|
||
|
# Another lunch "mirage" at N16/O15.
|
||
|
#CATEGORY=OWL
|
||
|
loadsgf games/nando/auto018.sgf
|
||
|
26 owl_does_defend P16 R18
|
||
|
#? [0]
|
||
|
|
||
|
# Move valuation looks way off. No idea what is happening here.
|
||
|
#CATEGORY=?
|
||
|
loadsgf games/nando/auto019.sgf
|
||
|
27 reg_genmove white
|
||
|
#? [D6]*
|
||
|
|
||
|
# Non-transitivity problem (or is it ok to amalgamate by ko ?)
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/auto020.sgf
|
||
|
28 disconnect S3 S7
|
||
|
#? [2 S6]
|
||
|
29 same_dragon S3 S7
|
||
|
#? [0]*
|
||
|
|
||
|
# W still can either connect or make 2 eyes
|
||
|
#CATEGORY=OWL
|
||
|
loadsgf games/nando/auto020.sgf
|
||
|
black S6
|
||
|
30 owl_defend S11
|
||
|
#? [1 T6]*
|
||
|
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/auto021.sgf
|
||
|
31 connect E18 H19
|
||
|
#? [0]
|
||
|
|
||
|
# O12 and Q14 are amalgamated per pattern CC502. The OWL engine
|
||
|
# thinks that the dragon can simply escape by playing O13 in
|
||
|
# response to black P13. And unfortunately, no other safeguards
|
||
|
# prevent the blunder (as of 3.5.5)
|
||
|
#CATEGORY=ATARI_ATARI (?)
|
||
|
loadsgf games/nando/auto022.sgf 206
|
||
|
32 reg_genmove white
|
||
|
#? [N13|O13]
|
||
|
|
||
|
# After W:H11 B:K13, White can save the day with K12, but GNU Go 3.7.5
|
||
|
# thinks that the move is unsafe (see reading:220)
|
||
|
loadsgf games/nando/auto023.sgf 170
|
||
|
33 reg_genmove white
|
||
|
#? [H11]
|
||
|
|
||
|
loadsgf games/nando/auto026.sgf 118
|
||
|
34 restricted_genmove white S12 S8
|
||
|
#? [S8]*
|
||
|
|
||
|
# Apparently, a ko depth issue (3.7.7)
|
||
|
play white T8
|
||
|
35 attack T8
|
||
|
#? [3 S8]*
|
||
|
|
||
|
loadsgf games/nando/auto027.sgf 165
|
||
|
36 disconnect L8 N7
|
||
|
#? [1 L7]*
|
||
|
|
||
|
|
||
|
#--------------------------------------------------------------
|
||
|
# Cascade of chain breakings
|
||
|
#CATEGORY=TACTICAL_READING
|
||
|
loadsgf games/nando/meijin27_g1.sgf 105
|
||
|
110 owl_defend B16
|
||
|
#? [0]
|
||
|
111 attack C18
|
||
|
#? [0]
|
||
|
|
||
|
# Connection overvaluation. This dragon has miai options to
|
||
|
# connect to strong dragons, so it isn't urgent to do it now.
|
||
|
#CATEGORY=STRATEGY
|
||
|
loadsgf games/nando/meijin27_g1.sgf 186
|
||
|
112 reg_genmove white
|
||
|
#? [!P7]
|
||
|
|
||
|
#--------------------------------------------------------------
|
||
|
# After B:P2 W:Q2, the connection code (as of 3.3.15) overvalues
|
||
|
# B:P3 (-1.95) and consequently rejects the correct move B:Q1
|
||
|
# at candidate selection stage
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/meijin27_g4.sgf 23
|
||
|
140 disconnect O2 Q3
|
||
|
#? [1 P2]*
|
||
|
|
||
|
# More an amalgamation problem. GG knows how to cut, but still
|
||
|
# thinks it's the same dragon.
|
||
|
# (Non-transitivity problem. /ab)
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/meijin27_g4.sgf 128
|
||
|
141 disconnect P2 P4
|
||
|
#? [1 P3]
|
||
|
142 same_dragon P2 P4
|
||
|
#? [0]*
|
||
|
|
||
|
#--------------------------------------------------------------
|
||
|
# Not really a connection problem, rather a tactical reading one
|
||
|
# see test 151
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/meijin27_g5.sgf 48
|
||
|
150 disconnect P3 Q4
|
||
|
#? [1 P4]*
|
||
|
|
||
|
#CATEGORY=TACTICAL_READING
|
||
|
loadsgf games/nando/meijin27_g5.sgf 48
|
||
|
black P4
|
||
|
151 attack P4
|
||
|
#? [0]*
|
||
|
|
||
|
#--------------------------------------------------------------
|
||
|
# C4 should be played only when W already has a stone around D10
|
||
|
#CATEGORY=FUSEKI
|
||
|
loadsgf games/nando/kisei27_g1.sgf 14
|
||
|
210 reg_genmove white
|
||
|
#? [D3]
|
||
|
|
||
|
# D5 is exceedingly urgent. GG should _never_ tenuki !
|
||
|
#CATEGORY=FUSEKI
|
||
|
loadsgf games/nando/kisei27_g1.sgf 16
|
||
|
211 reg_genmove white
|
||
|
#? [D5]
|
||
|
|
||
|
# Connection is mandatory.
|
||
|
#CATEGORY=FUSEKI
|
||
|
loadsgf games/nando/kisei27_g1.sgf 20
|
||
|
212 reg_genmove white
|
||
|
#? [E3]
|
||
|
|
||
|
#CATEGORY=CONNECTION
|
||
|
loadsgf games/nando/kisei27_g1.sgf 26
|
||
|
213 disconnect C4 C6
|
||
|
#? [0]*
|
||
|
|
||
|
#---------------------- End of Tests --------------------------
|