Author: Terry Ripple
Date: 14:17:31 08/10/00
Go up one level in this thread
On August 10, 2000 at 16:26:09, Uri Blass wrote: >On August 10, 2000 at 15:53:32, Terry Ripple wrote: > >>On August 10, 2000 at 10:06:27, Uri Blass wrote: >> >>>On August 10, 2000 at 09:57:15, Ulrich Tuerke wrote: >>> >>>>On August 10, 2000 at 09:51:16, Uri Blass wrote: >>>> >>>>>The same engine can sometimes resign in engine-engine games inspite of not >>>>>seeing a forced mate and sometimes continue and not resign inspite of seeing a >>>>>forced mate in engine-engine games. >>>>> >>>>>Can somebody tell me what are the rules that engines follow when they decide to >>>>>resign? >>>>> >>>>>Uri Hi Uri, In my 60 min./games, Fritz6 Light (old) didn't lose 1 game on time out of 12 played so far, so it will be interesting if Fritz loses any on time with this time control! Terry >>>> >>>>It is not the engine which resigns but the GUI. >>>>In the GUI, you can adjust "early", "late" or "never" resign, whatever that >>>>means. >>>>I think that I have observed when both engines agree in reporting a large value >>>>for the score for the last couple of moves, the the GUI makes its decision. >>>> >>>>I guess that only Mathias Feist knows the exact criteria. >>>> >>>>Uli >>> >>>I did not tell it nothing under hiarcs7.32 GUI and >>>I found that in part of the cases the engines contimue till mate and in part of >>>the cases resign before seeing evaluation of mate. >>> >>>Uri >> >>Hi Uri, >>I also perform my matches with the Hiarcs 7.32 GUI. I have it set to "resign" >>late in the game. I don't have this problem that you mentioned! I been >>performing hundreds of test games between Hiarcs732 and Fritz6 old and it seems >>that either engine will resign when there is no hope in the position, or will >>ask for a draw when the conditions call for an obvious draw which then the draw >>offer will be accepted! >> >>Regards,Terry > >I will check my setting but I did not change my setting. > >I do not see a clear rule about resigning and it sometimes resigns in a hopeless >positionand sometimes does not resign in hopeless position. > >I also think that part of the hopeless position are not hopeless in blitz >because of the clock. >In one game Hiarcs resigned in KRPP vs KRNB when it is not clear if Fritz can >win it on time trouble. >Fritz6light also enjoyed from few draws that Hiarcs agree when hiarcs had good >chances to win on time by playing. > >The result after 400 blitz games is 169:155(76 draws) for Fritz6 light >Fritz6light lost 9 games out of 400 on time. > >Both sides won e4 f6 with black and hiarcs won e4 f5 with black and it proves >that the opening is not very important. > >I played all the possible opening of 2 plies when white begins with one of the >pawns a-e. > >Uri
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.