Updated README: Equal sign not required with `--mode` flag.
[sgk-go] / regression / tactics1.tst
CommitLineData
7eeb782e
AT
1# Apparently, T9 gets overvalued because the effect on S7 and T6 is
2# counted twice (territorially and strategically)
3loadsgf games/tactics05.sgf
410 restricted_genmove white T9 T8 S9 S8
5#? [S8|S9]
6
7loadsgf games/tactics06.sgf
820 owl_defend C19
9#? [1 A17]*
10
11# J1 is worth 6 points
12loadsgf games/seki03.sgf
1330 reg_genmove white
14#? [J1]*
15
16
17# attack_either and defend_both tests could go in a separate test suite,
18# once there is more of them
19
20loadsgf games/nngs/Lazarus-gnugo-3.1.34-200204280120.sgf 32
21trymove black D12
22trymove white C13
23trymove white C12
24trymove black C11
25101 attack_either C11 D12
26#? [1]
27popgo
28popgo
29popgo
30popgo
31
32# This is subtle. White N3 works as defense to L4.
33loadsgf games/mertin13x13/katsunari-gnugo2.W+4.sgf 126
34trymove black L4
35102 defend_both N7 K4
36#? [0]*
37popgo
38
39loadsgf games/nngs/Lazarus-gnugo-3.2-200205011927.sgf 86
40trymove black G4
41trymove white G3
42103 attack_either E5 G3
43#? [1]
44104 attack_either G3 E5
45#? [1]
46popgo
47popgo
48
49# C9 is errouneously thought to be an owl-attack against G13,
50# which could indicate a semeai reading problem. On the other
51# hand, since G13 is judged owl-defenseless, it seems to me
52# that C9 is strangely overvalued.
53loadsgf games/tactics07.sgf
54105 reg_genmove white
55#? [F7]
56
57loadsgf games/tactics08.sgf
58106 reg_genmove white
59#? [F8|F6]
60
61loadsgf games/edge_defense1.sgf
62107 reg_genmove white
63#? [B13|B14]
64
65# Obviously P7 is effective.
66loadsgf games/gifu2005/mfg-gnugo.sgf 204
67trymove white O6
68108 defend_both O9 P6
69#? [1]*
70popgo