1 A list of games to put in the regression suite.
3 This list can be processed automatically:
5 sed -ne 's/^\* //p' <TODO | while read r c; do echo wget -O $(echo "$r" | sed 's#.*games/##' | tr / - | sed 's/-\([^-]\)-/-0\1-/g;s/-\([^-]\)-/-0\1-/') "$r"; done
7 * http://files.gokgs.com/games/2011/6/5/pachi2-SimpleBot.sgf ladder
8 pachi shouldn't play the broken ladders up to move 53
10 * http://files.gokgs.com/games/2011/6/5/tazaki-pachi30s.sgf semeai falseeye
11 pachi doesn't see even one move ahead in a semeai, it expects
12 an answer outside the semeai. The score drops from 0.625 to 0.337:
15 [880000] best 0.624794 komi 0.4 | seq L19 G12 F11 P11 | can L19(0.625)
16 O19(0.623) N19(0.469) P11(0.523)
17 FINEST: Got successful response to command "genmove b": = L19
18 FINEST: Got successful response to command "play w j16": =
20 FINEST: Queued command sent to engine: genmove b
21 [1026644] best 0.336825 komi 0.2 | seq P11 T8 R3 M9 | can P11(0.337)
22 O11(0.336) O19(0.334) N19(0.340)
23 FINEST: Got successful response to command "genmove b": = P11
25 The game was lost already but pachi should have known this before.
27 move 244: T13 doesn't make sense. This is due to wrong estimate of
29 [1922010] best 0.492722 | seq T13 T4 D5 M6 | can T13(0.493)
30 S18(0.493) L4(0.493) E6(0.491)
31 FINEST: Got successful response to command "genmove b": = T13
33 At move 282 pachi still thinks white's group is undecided.
34 We must find a way of not filling the S7 eye in the playouts.
36 * http://files.gokgs.com/games/2011/5/29/snowmen-pachi2.sgf
37 169 B5 seems to be the losing move. This move doesn't make sense.
38 Pachi has been wrongly evaluating both bottom group around S4 (it's just
39 hairy for simulations) and top group around A19 - that one is a bit more
40 interesting, the problem is that black is double likely to play E19
41 (since it's C19 selfatari cousin), but after E19, white will not notice
42 the need to connect at C19. Not sure how to solve this yet, perhaps some
43 way to add cousins to last_move logic...
45 * http://files.gokgs.com/games/2011/5/28/pachi2-Rubel.sgf
46 won game lost: 212 H13 should be at J12.
48 * http://files.gokgs.com/games/2011/5/27/ManyFaces1-pachi2.sgf
49 had won game, move 135 F1 very bad. at end doesn't see ko at t7. (b t10
52 * http://files.gokgs.com/games/2011/5/23/pachi2-ManyFaces1.sgf
53 leading up to 90% then lost.
54 bad evaluation of 3 areas according to pasky.
56 * http://files.gokgs.com/games/2011/5/25/pachi2-Zen19S.sgf
57 lost but was leading according to ajahuang and Zen itself
59 * http://files.gokgs.com/games/2011/2/20/ramesses2-pachi2.sgf
60 good example of losing a won game.
62 * http://files.gokgs.com/games/2011/1/11/llopl-pachi2.sgf ladder
65 * http://files.gokgs.com/games/2010/12/5/pachi-CzechBot.sgf
66 big mistake at F9 didn't see snapback. Pasky comments:
67 The "does not seen snapback" bug is nasty. The problem is that
68 there is also a "global capture heuristic" if there're no fun
69 local moves, it looks for groups to capture and if white tenukis
70 and does not capture immediately, this heuristic kicks in. But
71 there is no "global atari heuristic" that would drive black
72 similarly to play at F1 and create the snapback if I test with
73 playout=moggy:capturerate=0, it reliably destroys the snapback
74 i.e. with capturerate=90, white is much more likely to come back
75 to the situation later even when playing tenuki first.
77 * http://files.gokgs.com/games/2011/1/15/pachi2-rollingon.sgf semeai
78 big semeai with easy counting of liberties misjuged
79 move 169 0.7322 -> move 171 0.2600
81 * http://files.gokgs.com/games/2011/2/18/pachi2-tf-3.sgf seki
82 lost because of unrecognized seki
84 * http://files.gokgs.com/games/2011/6/9/botkiller2-pachi30s.sgf falseeye
85 threw away a won game. pachi thought until the end that top left
86 white group was dead, so it made silly moves elsewhere.
87 The playouts must be prevented from filling the eye at A17.
89 * http://files.gokgs.com/games/2011/6/10/pachi30s-samba-2.sgf falseeye
90 exactly the same problem as botkiller2-pachi30s.sgf:
91 pachi considers the top left black group as unsettled
92 although it is clearly alive. So it makes silly moves
93 elsewhere. We have to fix this bug, it is too frequent.
95 * http://files.gokgs.com/games/2011/6/9/jinen-pachi30s-2.sgf ko
96 move 165 G5 again playing the ko threat before the ko.
97 move 187 R18 is not a ko threat
99 At this point pachi still considers the top right as
100 unsetled, although it is is clearly lost for B. So pachi's
101 estimate of winning by 61% is way too optimistic.
103 * http://files.gokgs.com/games/2011/6/14/pachi2-Regenbogen.sgf
104 pachi doesn't see that bent four in top left corner is dead.
105 Still thinks it is winning 79% at move 250 G4.
107 * http://files.gokgs.com/games/2011/6/14/pachi2-cdocdo-3.sgf semeai
108 pachi loses every semai by a large number of liberties
109 but still considers itself winning 57% up to move 222 G16.
111 * http://files.gokgs.com/games/2011/6/13/pachi30s-Jep.sgf ladder
112 moves 102..108 playing a broken ladder
114 * http://files.gokgs.com/games/2011/6/18/dorabon-pachi2-4.sgf semeai falseeye
115 pachi considers the bottom left semeai won by black (but black
116 loses it) and the bottom right as undecided (white is clearly
117 alive). At move 270 L4 it still thinks it is leading by 80%
118 although the game is lost.
119 The bottom right is the same problem as
120 http://files.gokgs.com/games/2011/6/10/pachi30s-samba-2.sgf
121 and http://files.gokgs.com/games/2011/6/9/botkiller2-pachi30s.sgf
122 This corner shape is very frequent and pachi always gets it wrong.
124 * http://files.gokgs.com/games/2011/6/18/pachi2-ultimakang-4.sgf
125 pachi was winning up to move 230 Q14 big mistake. Win ratio dropped
126 from 60% to 25% at next move.
128 * http://files.gokgs.com/games/2011/6/18/akkannbee-pachi2.sgf seki
129 pachi totally confused by very simple seki.
131 * http://files.gokgs.com/games/2011/7/6/xiaosugi-pachi2.sgf seki
132 pachi thinks until the end that it wins by 80% because it
133 doesn't understand the seki in lower right. It should have lost
134 by 42.5 points but white mistakenly agreed to treat the seki as dead.
136 * http://files.gokgs.com/games/2011/7/5/pachi2-azlan7-2.sgf
137 move 84 E10 & 86 F10 big mistakes, leave white group dead.
138 (amazingly pachi recovered and azlan7 escaped in the end).
140 * http://files.gokgs.com/games/2011/7/5/fafahhhh-pachi2.sgf
141 threw away won game, probably because of complex corner situations
142 won by time but game was lost on board
144 * http://files.gokgs.com/games/2011/7/3/pachi2-oasis7.sgf
145 Passed even though a dead opponent group was marked alive.
147 * http://files.gokgs.com/games/2011/7/28/pachi2-Novicer.sgf falseeye
148 Pachi makes many stupid moves because it thinks the top left
149 corner is unsettled although it is clearly alive. This shape
150 (4 in a row) is so simple that it should get it right.
153 W 264 E19 doesn't make sense, but still 69%
154 W 266 H19 doesn't make sense, but still 61%
155 W 268 T9 should be T8
156 W 292 A17 50%, W 294 R6 37% shows that top left was not
158 Also false eye evaluation problem in bottom left.
160 * http://files.gokgs.com/games/2011/7/28/pachi2-pkunzip-2.sgf ladder
161 W 132 C9, W 134 B8, W 136 A8 plays a broken ladder but win ratio 52%
162 so it should not play desperate moves.
163 Pachi thought it was winning 72% at W 198 M3 but lost the game.
165 * http://files.gokgs.com/games/2011/7/16/ihatebots-pachi2.sgf semeai
166 pachi threw away the game because it didn't evaluate the
167 semeai in bottom left correctly.
169 * http://files.gokgs.com/games/2011/7/31/pachi2-Drogba987.sgf
170 B 28 H4 should be at F1 or G1 to save E2 & E3 and the game.
171 Pachi usually finds this so I'm surprised it didn't get it here.
173 * http://files.gokgs.com/games/2011/8/9/pachi2-BlueSpark.sgf ladder
174 move 150 R19: pachi extends a losing ladder
176 * http://files.gokgs.com/games/2011/8/8/Dallas-pachi2-2.sgf ladder
177 move 121 L19: pachi extends a losing ladder
179 * http://files.gokgs.com/games/2011/8/9/somrak-pachi2-2.sgf ladder
180 move 143 N19: pachi extends a losing ladder
181 This ladder bug seems more frequent since e9df2b6 (even though
182 this change made Pachi 40 elo stronger).
184 * http://files.gokgs.com/games/2011/8/10/jerryfish-pachi2.sgf semeai
185 lost even game against 3k. Win ratio drops from 61% at 117 N1
186 to 49% at the next move. So pachi didn't evaluate the semai correctly.
188 * http://files.gokgs.com/games/2011/8/12/xiaosugi-pachi2.sgf falseeye
189 Pachi hallucinated again on bottom right corner.
190 This is the same problem as
191 http://files.gokgs.com/games/2011/6/18/dorabon-pachi2-4.sgf
192 http://files.gokgs.com/games/2011/6/10/pachi30s-samba-2.sgf
193 http://files.gokgs.com/games/2011/6/9/botkiller2-pachi30s.sgf
194 This corner shape is very frequent and pachi always gets it wrong.
196 * http://files.gokgs.com/games/2011/8/24/StoneGrid-pachi2.sgf ladder
197 Move 51 onwards: pachi plays a broken ladder, which almost cost the game.
198 This might be the same bug as
199 http://files.gokgs.com/games/2011/8/9/somrak-pachi2-2.sgf
200 http://files.gokgs.com/games/2011/8/9/pachi2-BlueSpark.sgf
201 http://files.gokgs.com/games/2011/8/8/Dallas-pachi2-2.sgf
202 http://files.gokgs.com/games/2011/7/28/pachi2-pkunzip-2.sgf
203 http://files.gokgs.com/games/2011/6/13/pachi30s-Jep.sgf
204 http://files.gokgs.com/games/2011/6/6/tyzef-pachi30s.sgf
205 http://files.gokgs.com/games/2011/6/5/pachi2-SimpleBot.sgf
206 http://files.gokgs.com/games/2011/1/11/llopl-pachi2.sgf
208 * http://files.gokgs.com/games/2011/9/4/pachi2-stv.sgf ladder
209 Move 48 onwards: pachi plays a broken ladder, which cost not
210 just the game but also the first place in the tournament.
211 See also broken ladders in
212 http://files.gokgs.com/games/2011/8/24/StoneGrid-pachi2.sgf
213 http://files.gokgs.com/games/2011/8/9/somrak-pachi2-2.sgf
214 http://files.gokgs.com/games/2011/8/9/pachi2-BlueSpark.sgf
215 http://files.gokgs.com/games/2011/8/8/Dallas-pachi2-2.sgf
216 http://files.gokgs.com/games/2011/7/28/pachi2-pkunzip-2.sgf
217 http://files.gokgs.com/games/2011/6/13/pachi30s-Jep.sgf
218 http://files.gokgs.com/games/2011/6/6/tyzef-pachi30s.sgf
219 http://files.gokgs.com/games/2011/6/5/pachi2-SimpleBot.sgf
220 http://files.gokgs.com/games/2011/1/11/llopl-pachi2.sgf