From d60ced8f2952850495b231cf045dbbc8c830fe19 Mon Sep 17 00:00:00 2001 From: Petr Baudis Date: Tue, 4 Oct 2011 23:13:28 +0200 Subject: [PATCH] t-regress/TODO: Add some categories of games --- t-regress/TODO | 54 +++++++++++++++++++++++++++++------------------------- 1 file changed, 29 insertions(+), 25 deletions(-) diff --git a/t-regress/TODO b/t-regress/TODO index e6c6ac4..203bdd6 100644 --- a/t-regress/TODO +++ b/t-regress/TODO @@ -1,17 +1,21 @@ -Here is my current list of good candidates to put in a regression suite. +A list of games to put in the regression suite. + +This list can be processed automatically: + + sed -ne 's/^\* //p' move 171 0.2600 -* http://files.gokgs.com/games/2011/2/18/pachi2-tf-3.sgf +* http://files.gokgs.com/games/2011/2/18/pachi2-tf-3.sgf seki lost because of unrecognized seki -* http://files.gokgs.com/games/2011/6/9/botkiller2-pachi30s.sgf +* http://files.gokgs.com/games/2011/6/9/botkiller2-pachi30s.sgf tsumego0 threw away a won game. pachi thought until the end that top left white group was dead, so it made silly moves elsewhere. The playouts must be prevented from filling the eye at A17. -* http://files.gokgs.com/games/2011/6/10/pachi30s-samba-2.sgf +* http://files.gokgs.com/games/2011/6/10/pachi30s-samba-2.sgf tsumego0 exactly the same problem as botkiller2-pachi30s.sgf: pachi considers the top left black group as unsettled although it is clearly alive. So it makes silly moves elsewhere. We have to fix this bug, it is too frequent. -* http://files.gokgs.com/games/2011/6/9/jinen-pachi30s-2.sgf +* http://files.gokgs.com/games/2011/6/9/jinen-pachi30s-2.sgf ko move 165 G5 again playing the ko threat before the ko. move 187 R18 is not a ko threat move 215: @@ -123,14 +127,14 @@ kills above) pachi doesn't see that bent four in top left corner is dead. Still thinks it is winning 79% at move 250 G4. -* http://files.gokgs.com/games/2011/6/14/pachi2-cdocdo-3.sgf +* http://files.gokgs.com/games/2011/6/14/pachi2-cdocdo-3.sgf semeai pachi loses every semai by a large number of liberties but still considers itself winning 57% up to move 222 G16. -* http://files.gokgs.com/games/2011/6/13/pachi30s-Jep.sgf +* http://files.gokgs.com/games/2011/6/13/pachi30s-Jep.sgf ladder moves 102..108 playing a broken ladder -* http://files.gokgs.com/games/2011/6/18/dorabon-pachi2-4.sgf +* http://files.gokgs.com/games/2011/6/18/dorabon-pachi2-4.sgf semeai tsumego0 pachi considers the bottom left semeai won by black (but black loses it) and the bottom right as undecided (white is clearly alive). At move 270 L4 it still thinks it is leading by 80% @@ -144,10 +148,10 @@ kills above) pachi was winning up to move 230 Q14 big mistake. Win ratio dropped from 60% to 25% at next move. -* http://files.gokgs.com/games/2011/6/18/akkannbee-pachi2.sgf +* http://files.gokgs.com/games/2011/6/18/akkannbee-pachi2.sgf seki pachi totally confused by very simple seki. -* http://files.gokgs.com/games/2011/7/6/xiaosugi-pachi2.sgf +* http://files.gokgs.com/games/2011/7/6/xiaosugi-pachi2.sgf seki pachi thinks until the end that it wins by 80% because it doesn't understand the seki in lower right. It should have lost by 42.5 points but white mistakenly agreed to treat the seki as dead. @@ -175,12 +179,12 @@ kills above) W 292 A17 50%, W 294 R6 37% shows that top left was not evaluated correctly. -* http://files.gokgs.com/games/2011/7/28/pachi2-pkunzip-2.sgf +* http://files.gokgs.com/games/2011/7/28/pachi2-pkunzip-2.sgf ladder W 132 C9, W 134 B8, W 136 A8 plays a broken ladder but win ratio 52% so it should not play desperate moves. Pachi thought it was winning 72% at W 198 M3 but lost the game. -* http://files.gokgs.com/games/2011/7/16/ihatebots-pachi2.sgf +* http://files.gokgs.com/games/2011/7/16/ihatebots-pachi2.sgf semeai pachi threw away the game because it didn't evaluate the semeai in bottom left correctly. @@ -188,22 +192,22 @@ kills above) B 28 H4 should be at F1 or G1 to save E2 & E3 and the game. Pachi usually finds this so I'm surprised it didn't get it here. -* http://files.gokgs.com/games/2011/8/9/pachi2-BlueSpark.sgf +* http://files.gokgs.com/games/2011/8/9/pachi2-BlueSpark.sgf ladder move 150 R19: pachi extends a losing ladder -* http://files.gokgs.com/games/2011/8/8/Dallas-pachi2-2.sgf +* http://files.gokgs.com/games/2011/8/8/Dallas-pachi2-2.sgf ladder move 121 L19: pachi extends a losing ladder -* http://files.gokgs.com/games/2011/8/9/somrak-pachi2-2.sgf +* http://files.gokgs.com/games/2011/8/9/somrak-pachi2-2.sgf ladder move 143 N19: pachi extends a losing ladder This ladder bug seems more frequent since e9df2b6 (even though this change made Pachi 40 elo stronger). -* http://files.gokgs.com/games/2011/8/10/jerryfish-pachi2.sgf +* http://files.gokgs.com/games/2011/8/10/jerryfish-pachi2.sgf semeai lost even game against 3k. Win ratio drops from 61% at 117 N1 to 49% at the next move. So pachi didn't evaluate the semai correctly. -* http://files.gokgs.com/games/2011/8/12/xiaosugi-pachi2.sgf +* http://files.gokgs.com/games/2011/8/12/xiaosugi-pachi2.sgf tsumego0 Pachi hallucinated again on bottom right corner. This is the same problem as http://files.gokgs.com/games/2011/6/18/dorabon-pachi2-4.sgf @@ -211,7 +215,7 @@ kills above) http://files.gokgs.com/games/2011/6/9/botkiller2-pachi30s.sgf This corner shape is very frequent and pachi always gets it wrong. -* http://files.gokgs.com/games/2011/8/24/StoneGrid-pachi2.sgf +* http://files.gokgs.com/games/2011/8/24/StoneGrid-pachi2.sgf ladder Move 51 onwards: pachi plays a broken ladder, which almost cost the game. This might be the same bug as http://files.gokgs.com/games/2011/8/9/somrak-pachi2-2.sgf @@ -223,7 +227,7 @@ kills above) http://files.gokgs.com/games/2011/6/5/pachi2-SimpleBot.sgf http://files.gokgs.com/games/2011/1/11/llopl-pachi2.sgf -* http://files.gokgs.com/games/2011/9/4/pachi2-stv.sgf +* http://files.gokgs.com/games/2011/9/4/pachi2-stv.sgf ladder Move 48 onwards: pachi plays a broken ladder, which cost not just the game but also the first place in the tournament. See also broken ladders in -- 2.11.4.GIT