Thanks for the rapid test again Patrick! It is in line with my expectation that the new version performs about the same for 32-bit on the Pentium IV processor (with slow bitscan instructions). My hopes for strength increase are for 64-bit on a Core 2 Duo. See http://aartbik.blogspot.com/ for some of my experiences during the rewrite.
I had a quick look at the games to see if a castling bug occurred (please update to v2.01 for a bug fix) but noticed that some draw games were simply continued until an engine runs out of time.
For example, BikJump-Garbochess was KR-KR at move 106 (where BikJump claims a zero score, Garbochess a nonzero score), and played on until BikJump lost on time at move 141. Shouldn't such games be adjudicated as a draw?
it's only a draw if you can get 3 fold repetition, or reach 50 moves. Sounds like you still had 15 moves to go. It would be a good idea to play faster in drawn endgames.
abik wrote:I had a quick look at the games to see if a castling bug occurred (please update to v2.01 for a bug fix) but noticed that some draw games were simply continued until an engine runs out of time.
For example, BikJump-Garbochess was KR-KR at move 106 (where BikJump claims a zero score, Garbochess a nonzero score), and played on until BikJump lost on time at move 141. Shouldn't such games be adjudicated as a draw?
Games lost on time should be deleted and not included in the process of calculating the engines rating....In my rating list i only use clear cut results,if the engine loses frequently on time,I just drop it from the rating list
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
abik wrote:I had a quick look at the games to see if a castling bug occurred (please update to v2.01 for a bug fix) but noticed that some draw games were simply continued until an engine runs out of time.
For example, BikJump-Garbochess was KR-KR at move 106 (where BikJump claims a zero score, Garbochess a nonzero score), and played on until BikJump lost on time at move 141. Shouldn't such games be adjudicated as a draw?
Yes, sorry about that, Garbochess is not the most intelligent in endgames . It's thinking that it's winning because BikJump's king is on the edge of the board, and it's king is in the center.
I took a quick glance at the other game, and it looks like BikJump was played into a busted opening by the book.
abik wrote:I had a quick look at the games to see if a castling bug occurred (please update to v2.01 for a bug fix) but noticed that some draw games were simply continued until an engine runs out of time.
For example, BikJump-Garbochess was KR-KR at move 106 (where BikJump claims a zero score, Garbochess a nonzero score), and played on until BikJump lost on time at move 141. Shouldn't such games be adjudicated as a draw?
Yes, sorry about that, Garbochess is not the most intelligent in endgames . It's thinking that it's winning because BikJump's king is on the edge of the board, and it's king is in the center.
I took a quick glance at the other game, and it looks like BikJump was played into a busted opening by the book.
What is the opening book in use
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
gladius wrote:[Yes, sorry about that, Garbochess is not the most intelligent in endgames . It's thinking that it's winning because BikJump's king is on the edge of the board, and it's king is in the center.
No need for apologies, it is a valuable lesson learned for me. I perform most of my BikJump tournaments in Chessbase, where the GUI starts to play rapidly from the Nalimov endgame tablebases if a known position occurs at the "root" (so that either the game is won quickly or drawn by fifty move rule). This seemed more fair to other engines that do not support Nalimov endgame tablebases.
Because BikJump supports Nalimov lookup during search, but expects the GUI to lookup at root, I never tuned for this situation. An easy fix is in the making
BikJump was slauthered even before he starts to play because of a very bad opening line....
I once asked you if you check all the games played in your tournament before you post your fast results,you didn't answer me,now I know the answer....
BTWI trust only my own opening book under Arena GUI:
iDeeb_Script.abk,an opening book which have been developed for more than six years....
P.S.I hope yet again not to consider this post as a personal attack along with your friends....facts are facts....such games spoil the whole process of getting a decent rating for the engines....if I am saying somehting wrong,please do correct me....
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….