Author: Dieter Buerssner
Date: 15:37:18 02/25/01
Go up one level in this thread
On February 21, 2001 at 15:53:25, Uri Blass wrote: I think, Yace did not outsearch SOS. I could see Rudolf Huber's monitor, and if I remember correctly, both programs reached similar search depths through the game. Also, I think, that both programs had for a long time similar scores, when you take into account, that Yace only has a material value of 0.8 for a pawn. But you are right, that Yace detected the perpetual check much earlier than SOS. I of course was happy to see the 0.0 score on my screen. But I also have to say, that, while I try to be very careful about perpetual checks, some luck may be needed to really detect them. This is because of hash tables. The situation can be the following: In the search, there occures a transposition of the position, that is found in the hash. The value stored in the hash may show a non draw score. But if you would look at the mainline after the hash move (which I think no program does while searching), you may detect, that there is a repetion of the position now, which was not there, when you searched the position the last time and stored the score in the hash. Regards, Dieter "just back from Paderborn" Bürßner
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.