Author: Ricardo Gibert
Date: 18:35:39 08/18/05
Go up one level in this thread
On August 18, 2005 at 10:28:13, Robert Hyatt wrote: >On August 18, 2005 at 09:29:45, Steve Maughan wrote: > >>In this position Fruit plays Bxh6! - how long for other engines to find this >>combo? >> >>[d]4rr1q/1bp2p1k/1ppp3p/4n3/4P3/2Q3RP/PPPBB1P1/R6K w - - 0 21 > > >Fruit was probably lucky to spot this. The move may have been even better the >previous move, but Fruit could not find it although crafty had seen it (Crafty >was searching quite a bit deeper than fruit here) and was expecting it. The >unfortunate thing was that after fruit played Nxb6, there were two ways to >recapture, axb6 and cxb6. Fruit was expecting axb6 and was pondering. Crafty >was going to play cxb6 but failed low (after it saw that Bxh6 was still strong). > As a result of the fail low, it searched a _long_ time, and at the last second >changed to axb6, which is probably what gave Fruit enough time to see the move. > >Fruit played well. But it was pretty lucky to stumble into a pretty much won >game so easily, getting outsearched as it was... > >But it happens... Given that Crafty was running on an 8 CPU machine against against a 1 cpu machine, calling it "lucky" is like a heavy weight boxer calling a bantam weight boxer "lucky" to have knocked him out. It does not look good at all and well worth avoiding. My congratulations to Fabien for a splendid program playing a splendid game.
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.