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] |