[sgk-go] / regression / strategy4.tst

# numbering is continued from strategy3.tst.

loadsgf games/gwe.sgf 231
# strictly speaking tenuki is possible here. Q1 is of course bad
# tm - Indeed GNU Go has finally figured out that tenuki is best!
# tm - yes, Q1 is bad.  T5 also wastes a point.
# gf - However, policy is to restrain thrashing dragons, so T5 is correct.
151 reg_genmove white
#? [T5]

loadsgf games/gwe.sgf 241
152 reg_genmove white
#? [S8]

loadsgf games/strategy36.sgf
153 reg_genmove black
#? [Q9]*

loadsgf games/strategy37.sgf
154 reg_genmove white
#? [E8]

# Incident 190.
loadsgf games/incident189.sgf 186
155 reg_genmove black
#? [D18]

loadsgf games/ssstator.sgf 133
156 reg_genmove black
#? [B14|B16]

# This problem is interesting. First there's a combination attack
# at B16/B14/C18/B12. Second there's a combination breakin attack
# G9+E9+F10 together with the cuts at B10,B12,B14.
# The only moves defending against both combinations are B14, B13,
# B12, A14, and A12.
loadsgf games/ssstator.sgf 136
157 reg_genmove white
#? [B14|B13|B12|A14|A12]*

# E16 is completely aimless and ineffective.
loadsgf games/strategy38.sgf 51
158 restricted_genmove black E16 R18
#? [R18]

loadsgf games/strategy38.sgf 117
159 reg_genmove black
#? [R9]*

loadsgf games/strategy38.sgf 133
160 reg_genmove black
#? [K2|L2|L3]

loadsgf games/strategy38.sgf 197
161 reg_genmove black
#? [E12]

# First live then attack. This should probably be analyzed as a semeai.
loadsgf games/strategy39.sgf 156
162 reg_genmove white
#? [O7|Q7|N7]

# First live then attack. This should probably be analyzed as a semeai.
# tm - P8, though not as clean, also kills all the B stones.
loadsgf games/strategy39.sgf 158
163 reg_genmove white
#? [O7|P8]*

#### N4 is clearly superior to O5.
#### tm - N4 has it's weaknesses (at O5) too.  Both clearly kill.
###    removing this problem.
###loadsgf games/strategy39.sgf 172
###164 reg_genmove white
####? [!O5]
###

# K1 is out of the question. K3 is locally better and L6 might kill with ko.
loadsgf games/strategy39.sgf 178
165 reg_genmove white
#? [K3|L6]

loadsgf games/incident94.sgf 114
166 reg_genmove white
#? [D11]*

# Tenuki instead of urgent edge joseki move.
loadsgf games/strategy40.sgf 10
167 reg_genmove black
#? [D4]

# Safety must come first. This is almost a whole-board semeai.
loadsgf games/strategy40.sgf 30
168 reg_genmove black
#? [A3|A4]*

# Both J7 and J9 rescue the threatened stones, but J7 is the proper
# choice since it leaves a smaller ko threat behind. (That there's no
# risk for ko in this position is irrelevant. There's still no reason
# to choose J9 over J7.)
loadsgf games/strategy41.sgf
169 reg_genmove black
#? [J7]

# White can't play G19. Locally this move must be at G18.
loadsgf games/strategy42.sgf 174
170 restricted_genmove white G18 G19
#? [G18]

# T19 is out of the question. Locally T17 is better but the correct
# move is A5.
loadsgf games/strategy42.sgf 188
171 reg_genmove white
#? [A5]

# It's correct to make a ko threat, but N9 is inefficient because it
# doesn't utilize the full ko potential of the position. Locally M8 is
# superior.
loadsgf games/strategy42.sgf 214
172 restricted_genmove white M8 N9
#? [M8]*

#CATEGORY=KO
#DESCRIPTION=GNU Go lacks sufficient Ko understanding
#SEVERITY=4
# There's no way to win the ko by making atari from above. This move
# must be at A9. This is a good example showing why only having the
# difference between a ko where we make the first threat or not is
# insufficient.
loadsgf games/strategy42.sgf 218
173 reg_genmove white
#? [A9]

# Always take the ko at least once.
loadsgf games/strategy42.sgf 248
174 reg_genmove white
#? [S19]

# This example shows that we need to find additional owl attack
# points. Currently only K15 is found to owl attack N19, but clearly
# the J14 move also does this, without sacrificing K13.
loadsgf games/strategy43.sgf 271
175 reg_genmove white
#? [J14]

# Best to capture with snapback.
loadsgf games/strategy44.sgf
176 reg_genmove black
#? [D9]

loadsgf games/xxlin.sgf 8
177 reg_genmove black
#? [E2]

# E2 looks somewhat smaller but is still a good move. I think it
# should be added. /gf
loadsgf games/xxlin.sgf 16
178 reg_genmove black
#? [P4|P3|E2]

#DESCRIPTION=Tenuki also seems OK here.
#Added E2 (3.1.14) -trevor
loadsgf games/xxlin.sgf 20
179 reg_genmove black
#? [P3|O2|P2|O4|E2]*

loadsgf games/xxlin.sgf 30
180 restricted_genmove black M7 Q8 R8 S8
#? [Q8|R8]*

loadsgf games/xxlin.sgf 38
181 reg_genmove black
#? [F3]*

loadsgf games/xxlin.sgf 48
182 reg_genmove black
#? [E7|F8]*

loadsgf games/xxlin.sgf 76
183 reg_genmove black
#? [P10|H9]

loadsgf games/xxlin.sgf 106
184 reg_genmove black
#? [H13|F15|Q13]

# We may want to move this to a dedicated seki or semeai test suite in
# the future.
loadsgf games/seki02.sgf
185 reg_genmove black
#? [E7]

# O8 only move.
loadsgf games/owl29.sgf 134
186 reg_genmove white
#? [O8]

# B can play at C4 if we remember the tesuji B:C4 W:E4 B:B2.

loadsgf games/pooo.sgf 16
187 restricted_genmove black E4 C4 B3 B4 C1
#? [C4|E4]*

loadsgf games/pooo.sgf 24
188 reg_genmove black
#? [B6|C4]

loadsgf games/pooo.sgf 50
189 restricted_genmove black A14 C14 D14 E14 D15 E15 C19
#? [E14]*

loadsgf games/pooo.sgf 52
190 reg_genmove black
#? [D13]*

loadsgf games/pooo.sgf 56
191 reg_genmove black
#? [C12]

# A15 captures the white corner stones but does not have much
# strategical value. The black stones are safe anyway. Thus A15 is
# worth about 19 points. Other moves may well be larger.
loadsgf games/pooo.sgf 74
192 reg_genmove black
#? [A15]*

# (3.1.9) owl status of A11 is unchecked.

loadsgf games/bretz.sgf 26
193 reg_genmove black
#? [F16]*

loadsgf games/bretz.sgf 58
194 reg_genmove black
#? [A13]*

loadsgf games/bretz.sgf 88
195 restricted_genmove black S13 S14
#? [S13]

loadsgf games/bretz.sgf 100
196 reg_genmove black
#? [M17]*

#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
loadsgf games/bretz.sgf 130
197 reg_genmove black
#? [K3|S18]

#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
loadsgf games/bretz.sgf 148
198 reg_genmove black
#? [C10|S18]

#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
loadsgf games/bretz.sgf 164
199 reg_genmove black
#? [N5|S18]

#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
loadsgf games/bretz.sgf 212
200 reg_genmove black
#? [P6|P7|Q7|S18]*

loadsgf games/juka1.sgf 8
201 reg_genmove black
#? [D5]

loadsgf games/juka1.sgf 14
202 reg_genmove black
#? [B6]*

loadsgf games/juka1.sgf 18
203 reg_genmove black
#? [D5]*

loadsgf games/juka1.sgf 28
204 reg_genmove black
#? [B2]*

loadsgf games/juka1.sgf 48
205 restricted_genmove black O11 P11
#? [P11]*

#CATEGORY=SEMEAI
#DESCRIPTION=K6 does take K8 to no eyes, but...
#SEVERITY=8
#GNU Go misses that K10 (large B group) has way fewer liberties
#than O9 group, hense missing importance of cutting off K8.
loadsgf games/strategy45.sgf
206 reg_genmove black
#? [L8]*

# T6 fails badly. Locally S9 looks best.
loadsgf games/owl30.sgf 84
207 restricted_genmove black R9 S9 T6
#? [S9]*

loadsgf games/lordofpi.sgf 36
208 reg_genmove black
#? [A5|A2|B1]

# A4 allows W to make seki
loadsgf games/lordofpi.sgf 46
209 restricted_genmove black A1 A3 A4 C1 D3
#? [A3]

loadsgf games/lordofpi.sgf 52
210 reg_genmove black
#? [H7]

loadsgf games/lordofpi.sgf 56
211 reg_genmove black
#? [H8]

# Pattern LE14 gives an unreasonable value to H3.
loadsgf games/owl31.sgf 28
212 reg_genmove white
#? [G2]

loadsgf games/strategy46.sgf 172
213 reg_genmove white
#? [F1]

loadsgf games/strategy46.sgf 174
214 reg_genmove white
#? [D2]

# No longer interesting. Removed.
# # CD103a incorrectly thinks E7 defends against a combination attack.
# loadsgf games/owl33.sgf 200
# 215 reg_genmove black
# #? [!E7]

# B6 is too slow at this time.
loadsgf games/handtalk/handtalk1.sgf 45
216 reg_genmove black
#? [!B6]

loadsgf games/handtalk/handtalk1.sgf 49
217 reg_genmove black
#? [!R9]

# locally F12 is better. Similar problem at move 51
loadsgf games/handtalk/handtalk1.sgf 55
218 restricted_genmove black C14 E12 F12
#? [F12]*

# lots of good problems in this game
loadsgf games/nngs/gnugo-3.1.30-morlvera-200204041921.sgf 154
219 reg_genmove white
#? [K9]*

loadsgf games/nngs/gnugo-3.1.30-merlin-200204041428.sgf 189
220 reg_genmove white
#? [R6]
Commit	Line	Data
7eeb782e AT	1	# numbering is continued from strategy3.tst.
	2
	3	loadsgf games/gwe.sgf 231
	4	# strictly speaking tenuki is possible here. Q1 is of course bad
	5	# tm - Indeed GNU Go has finally figured out that tenuki is best!
	6	# tm - yes, Q1 is bad. T5 also wastes a point.
	7	# gf - However, policy is to restrain thrashing dragons, so T5 is correct.
	8	151 reg_genmove white
	9	#? [T5]
	10
	11	loadsgf games/gwe.sgf 241
	12	152 reg_genmove white
	13	#? [S8]
	14
	15	loadsgf games/strategy36.sgf
	16	153 reg_genmove black
	17	#? [Q9]*
	18
	19	loadsgf games/strategy37.sgf
	20	154 reg_genmove white
	21	#? [E8]
	22
	23	# Incident 190.
	24	loadsgf games/incident189.sgf 186
	25	155 reg_genmove black
	26	#? [D18]
	27
	28	loadsgf games/ssstator.sgf 133
	29	156 reg_genmove black
	30	#? [B14\|B16]
	31
	32	# This problem is interesting. First there's a combination attack
	33	# at B16/B14/C18/B12. Second there's a combination breakin attack
	34	# G9+E9+F10 together with the cuts at B10,B12,B14.
	35	# The only moves defending against both combinations are B14, B13,
	36	# B12, A14, and A12.
	37	loadsgf games/ssstator.sgf 136
	38	157 reg_genmove white
	39	#? [B14\|B13\|B12\|A14\|A12]*
	40
	41	# E16 is completely aimless and ineffective.
	42	loadsgf games/strategy38.sgf 51
	43	158 restricted_genmove black E16 R18
	44	#? [R18]
	45
	46	loadsgf games/strategy38.sgf 117
	47	159 reg_genmove black
	48	#? [R9]*
	49
	50	loadsgf games/strategy38.sgf 133
	51	160 reg_genmove black
	52	#? [K2\|L2\|L3]
	53
	54	loadsgf games/strategy38.sgf 197
	55	161 reg_genmove black
	56	#? [E12]
	57
	58	# First live then attack. This should probably be analyzed as a semeai.
	59	loadsgf games/strategy39.sgf 156
	60	162 reg_genmove white
	61	#? [O7\|Q7\|N7]
	62
	63	# First live then attack. This should probably be analyzed as a semeai.
	64	# tm - P8, though not as clean, also kills all the B stones.
65	loadsgf games/strategy39.sgf 158
66	163 reg_genmove white
67	#? [O7\|P8]*
68
69	#### N4 is clearly superior to O5.
70	#### tm - N4 has it's weaknesses (at O5) too. Both clearly kill.
71	### removing this problem.
72	###loadsgf games/strategy39.sgf 172
73	###164 reg_genmove white
74	####? [!O5]
75	###
76
77	# K1 is out of the question. K3 is locally better and L6 might kill with ko.
78	loadsgf games/strategy39.sgf 178
79	165 reg_genmove white
80	#? [K3\|L6]
81
82	loadsgf games/incident94.sgf 114
83	166 reg_genmove white
84	#? [D11]*
85
86	# Tenuki instead of urgent edge joseki move.
87	loadsgf games/strategy40.sgf 10
88	167 reg_genmove black
89	#? [D4]
90
91	# Safety must come first. This is almost a whole-board semeai.
92	loadsgf games/strategy40.sgf 30
93	168 reg_genmove black
94	#? [A3\|A4]*
95
96	# Both J7 and J9 rescue the threatened stones, but J7 is the proper
97	# choice since it leaves a smaller ko threat behind. (That there's no
98	# risk for ko in this position is irrelevant. There's still no reason
99	# to choose J9 over J7.)
100	loadsgf games/strategy41.sgf
101	169 reg_genmove black
102	#? [J7]
103
104	# White can't play G19. Locally this move must be at G18.
105	loadsgf games/strategy42.sgf 174
106	170 restricted_genmove white G18 G19
107	#? [G18]
108
109	# T19 is out of the question. Locally T17 is better but the correct
110	# move is A5.
111	loadsgf games/strategy42.sgf 188
112	171 reg_genmove white
113	#? [A5]
114
115	# It's correct to make a ko threat, but N9 is inefficient because it
116	# doesn't utilize the full ko potential of the position. Locally M8 is
117	# superior.
118	loadsgf games/strategy42.sgf 214
119	172 restricted_genmove white M8 N9
120	#? [M8]*
121
122	#CATEGORY=KO
123	#DESCRIPTION=GNU Go lacks sufficient Ko understanding
124	#SEVERITY=4
125	# There's no way to win the ko by making atari from above. This move
126	# must be at A9. This is a good example showing why only having the
127	# difference between a ko where we make the first threat or not is
128	# insufficient.
129	loadsgf games/strategy42.sgf 218
130	173 reg_genmove white
131	#? [A9]
132
133	# Always take the ko at least once.
134	loadsgf games/strategy42.sgf 248
135	174 reg_genmove white
136	#? [S19]
137
138	# This example shows that we need to find additional owl attack
139	# points. Currently only K15 is found to owl attack N19, but clearly
140	# the J14 move also does this, without sacrificing K13.
141	loadsgf games/strategy43.sgf 271
142	175 reg_genmove white
143	#? [J14]
144
145	# Best to capture with snapback.
146	loadsgf games/strategy44.sgf
147	176 reg_genmove black
148	#? [D9]
149
150	loadsgf games/xxlin.sgf 8
151	177 reg_genmove black
152	#? [E2]
153
154	# E2 looks somewhat smaller but is still a good move. I think it
155	# should be added. /gf
156	loadsgf games/xxlin.sgf 16
157	178 reg_genmove black
158	#? [P4\|P3\|E2]
159
160	#DESCRIPTION=Tenuki also seems OK here.
161	#Added E2 (3.1.14) -trevor
162	loadsgf games/xxlin.sgf 20
163	179 reg_genmove black
164	#? [P3\|O2\|P2\|O4\|E2]*
165
166	loadsgf games/xxlin.sgf 30
167	180 restricted_genmove black M7 Q8 R8 S8
168	#? [Q8\|R8]*
169
170	loadsgf games/xxlin.sgf 38
171	181 reg_genmove black
172	#? [F3]*
173
174	loadsgf games/xxlin.sgf 48
175	182 reg_genmove black
176	#? [E7\|F8]*
177
178	loadsgf games/xxlin.sgf 76
179	183 reg_genmove black
180	#? [P10\|H9]
181
182	loadsgf games/xxlin.sgf 106
183	184 reg_genmove black
184	#? [H13\|F15\|Q13]
185
186	# We may want to move this to a dedicated seki or semeai test suite in
187	# the future.
188	loadsgf games/seki02.sgf
189	185 reg_genmove black
190	#? [E7]
191
192	# O8 only move.
193	loadsgf games/owl29.sgf 134
194	186 reg_genmove white
195	#? [O8]
196
197	# B can play at C4 if we remember the tesuji B:C4 W:E4 B:B2.
198
199	loadsgf games/pooo.sgf 16
200	187 restricted_genmove black E4 C4 B3 B4 C1
201	#? [C4\|E4]*
202
203	loadsgf games/pooo.sgf 24
204	188 reg_genmove black
205	#? [B6\|C4]
206
207	loadsgf games/pooo.sgf 50
208	189 restricted_genmove black A14 C14 D14 E14 D15 E15 C19
209	#? [E14]*
210
211	loadsgf games/pooo.sgf 52
212	190 reg_genmove black
213	#? [D13]*
214
215	loadsgf games/pooo.sgf 56
216	191 reg_genmove black
217	#? [C12]
218
219	# A15 captures the white corner stones but does not have much
220	# strategical value. The black stones are safe anyway. Thus A15 is
221	# worth about 19 points. Other moves may well be larger.
222	loadsgf games/pooo.sgf 74
223	192 reg_genmove black
224	#? [A15]*
225
226	# (3.1.9) owl status of A11 is unchecked.
227
228	loadsgf games/bretz.sgf 26
229	193 reg_genmove black
230	#? [F16]*
231
232	loadsgf games/bretz.sgf 58
233	194 reg_genmove black
234	#? [A13]*
235
236	loadsgf games/bretz.sgf 88
237	195 restricted_genmove black S13 S14
238	#? [S13]
239
240	loadsgf games/bretz.sgf 100
241	196 reg_genmove black
242	#? [M17]*
243
244	#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
245	loadsgf games/bretz.sgf 130
246	197 reg_genmove black
247	#? [K3\|S18]
248
249	#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
250	loadsgf games/bretz.sgf 148
251	198 reg_genmove black
252	#? [C10\|S18]
253
254	#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
255	loadsgf games/bretz.sgf 164
256	199 reg_genmove black
257	#? [N5\|S18]
258
259	#Tenuki at S18 is not unthinkable. (3.1.14) -trevor
260	loadsgf games/bretz.sgf 212
261	200 reg_genmove black
262	#? [P6\|P7\|Q7\|S18]*
263
264	loadsgf games/juka1.sgf 8
265	201 reg_genmove black
266	#? [D5]
267
268	loadsgf games/juka1.sgf 14
269	202 reg_genmove black
270	#? [B6]*
271
272	loadsgf games/juka1.sgf 18
273	203 reg_genmove black
274	#? [D5]*
275
276	loadsgf games/juka1.sgf 28
277	204 reg_genmove black
278	#? [B2]*
279
280	loadsgf games/juka1.sgf 48
281	205 restricted_genmove black O11 P11
282	#? [P11]*
283
284	#CATEGORY=SEMEAI
285	#DESCRIPTION=K6 does take K8 to no eyes, but...
286	#SEVERITY=8
287	#GNU Go misses that K10 (large B group) has way fewer liberties
288	#than O9 group, hense missing importance of cutting off K8.
289	loadsgf games/strategy45.sgf
290	206 reg_genmove black
291	#? [L8]*
292
293	# T6 fails badly. Locally S9 looks best.
294	loadsgf games/owl30.sgf 84
295	207 restricted_genmove black R9 S9 T6
296	#? [S9]*
297
298	loadsgf games/lordofpi.sgf 36
299	208 reg_genmove black
300	#? [A5\|A2\|B1]
301
302	# A4 allows W to make seki
303	loadsgf games/lordofpi.sgf 46
304	209 restricted_genmove black A1 A3 A4 C1 D3
305	#? [A3]
306
307	loadsgf games/lordofpi.sgf 52
308	210 reg_genmove black
309	#? [H7]
310
311	loadsgf games/lordofpi.sgf 56
312	211 reg_genmove black
313	#? [H8]
314
315	# Pattern LE14 gives an unreasonable value to H3.
316	loadsgf games/owl31.sgf 28
317	212 reg_genmove white
318	#? [G2]
319
320	loadsgf games/strategy46.sgf 172
321	213 reg_genmove white
322	#? [F1]
323
324	loadsgf games/strategy46.sgf 174
325	214 reg_genmove white
326	#? [D2]
327
328	# No longer interesting. Removed.
329	# # CD103a incorrectly thinks E7 defends against a combination attack.
330	# loadsgf games/owl33.sgf 200
331	# 215 reg_genmove black
332	# #? [!E7]
333
334	# B6 is too slow at this time.
335	loadsgf games/handtalk/handtalk1.sgf 45
336	216 reg_genmove black
337	#? [!B6]
338
339	loadsgf games/handtalk/handtalk1.sgf 49
340	217 reg_genmove black
341	#? [!R9]
342
343	# locally F12 is better. Similar problem at move 51
344	loadsgf games/handtalk/handtalk1.sgf 55
345	218 restricted_genmove black C14 E12 F12
346	#? [F12]*
347
348	# lots of good problems in this game
349	loadsgf games/nngs/gnugo-3.1.30-morlvera-200204041921.sgf 154
350	219 reg_genmove white
351	#? [K9]*
352
353	loadsgf games/nngs/gnugo-3.1.30-merlin-200204041428.sgf 189
354	220 reg_genmove white
355	#? [R6]