Author: James T. Walker
Date: 06:42:03 01/26/00
Go up one level in this thread
On January 26, 2000 at 09:21:03, James T. Walker wrote: >On January 26, 2000 at 08:16:34, Dusan Dobes wrote: > >>The game was played yesterday. I was analyzing it with Phalanx and it needed >>50 minutes on a P6/200 to realize that it is best to draw the game with >>39. ... Kh6 instead of losing with 39. ... Kh7? 40. Qe4+ Qxe4 41. fxe4 >> >>Is this so tough for all comps? What do other programs think? >> >>Dusan >> >>[D]4Q3/p1p3p1/6k1/2P4p/1P5q/5Pp1/P7/6K1 b - - 0 39 > >Hello Dusan, >I gave this to Chess Tiger on my K6-450 and it hit 14 plys in a little over a >minute before it started to "smell a rat". Tiger did not change it's move from >Kh7 to Kh6 untill after 5:55 at 3/5 of the 14th ply. At the 10 minute mark Tiger >is still looking at this move (Kh6) in the 14th ply and the score is -0.52 but >it will play the move after 5:55. This probably means Tiger would play this >move at tournament time controls unless it was in time trouble already. >Jim Walker Wow, the above was done with 16 Meg hash tables. I had Chess Tiger set up for blitz with only 16 Meg. When I tried again with 128 Meg hash tables here is what happens. Chess Tiger starts out with a big plus score of almost 3 points which starts dropping almost instantly. Tiger is playing Kh7 like most other programs at this point. Chess Tiger hits the 3rd move in the 14th ply at about 54 seconds which is Kh6. It studies this move untill 2:14 when it changes to actually play Kh6. So in tournament time this is no problem for the Tiger. Jim Walker
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.