Author: Robert Hyatt
Date: 11:10:28 08/16/01
Go up one level in this thread
On August 16, 2001 at 04:42:20, Scott Gasch wrote: >This is a position that can come up in the sicilian dragon. My engine does not >see it until it's way too late, even in standard time controls. > >[D] 3qr1k1/pp1bppb1/3p2p1/2rN2Bn/2nNPPp1/1B6/PPP1Q3/2KR3R b - - 0 1 > >The idea is to avoid Ng3 because it leads to something like this: > >[D] 3qr1k1/pp1bppb1/3p2p1/2rN2B1/2nNPPp1/1B6/PPP4Q/2K4R b - - 0 3 > >I bet that the strong amateurs and pros will avoid this move quickly. I need >about 6M nodes and 10 ply to see its bad... even then my engine still wants to >play it as it thinks the king can sneak out. It can't, according to opening >book literature. > >Scott This is not easy for Crafty. Takes 1:54 to change away from Ng3. Ng3 fails low at 18 seconds, depth=11 (quad 700), which means that it would not play this move in any reasonable time control since once it fails low, it has tha ability to use a bunch of time to find something better. at 1:54 Na5 fails high. I didn't let it search longer as an ICC game started up and I had to stop the test...
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.