Author: Robert Hyatt
Date: 14:11:59 03/03/00
Go up one level in this thread
On March 03, 2000 at 14:02:48, Pete Galati wrote: >On March 03, 2000 at 12:02:53, Ed Schröder wrote: > >>On March 03, 2000 at 06:33:59, Steve Lim wrote: >> >>>On March 03, 2000 at 00:20:35, Jeremiah Penery wrote: >>> >>>>On March 03, 2000 at 00:19:03, Jeremiah Penery wrote: >>>> >>>>>On March 02, 2000 at 19:12:55, Pete Galati wrote: >>>>> >>>>>>On March 02, 2000 at 10:11:50, Shep wrote: >>>>>> >>>>>>>Critical position: >>>>>>>[D]r1b2rk1/1pq2pbp/p2pp1p1/4n3/2P1PBn1/2N5/PPBNQPPP/R2R2K1 b - - 0 14; bm Nxf2; >>>>>>>id Shep 2000.01 >>>>>>> >>>>>>>14...Nxf2!! >>>>>>> >>>>>>>The results from the programs I tested (PIII-550): >>>>>>> >>>>>>>Junior 6.0: 0:55, +0,35 >>>>>>>Deep Junior 6.0a: 0:59, +0,32 >>>>>>>Nimzo 7.32: 6:00, +0,40 >>>>>>>Hiarcs 7.32: 8:19, +0,06 >>>>>>>Fritz 6.0a: >10:00 >>>>>>> >>>>>>>--- >>>>>>>Shep >>>>>> >>>>>>Keep in mind this is on my slow 90mhz machine, but I could not get Crafty to >>>>>>find Nxf2 in analyze mode, so I fed it the move and annotated the game (cut >>>>>>short at 14. ...Nxf2) and also below is the game log. >>>>> >>>>>On that machine, Crafty will probably take about 30 minutes to find Nxf2. At >>>>>least 15 minutes, anyway. I saw that you only searched it for one minute - of >>>>>course this will not be long enough. Even on the PIII-550, J6 barely found it >>>>>in under a minute. >>>> >>>>I missed something. You let it search for 15 minutes, but were only at depth 10. >>>> Crafty takes to around ply 13-14 to find this. >>> >>>I asked Bob (hyatt) to check this out, apparently it doesn't like Nxf2 even at >>>ply 15 on the xeons. >>> >>>Steve. >> >>I am not conviced Nxf2 is the best move in this position. >> >>Ed > >Probably not, but I think Nxf2 is a very human move to make, even though I >wouldn't have conceived of the move myself. What Jeremiah saw was me annotating >the game for 15 minutes with Nxf2 inserted to be annotated, there were no >restrictions on the depth that it searched, the restriction was being run on my >computer. > >The reason I took that approach was that I had tried to get Crafty to find Nxf2 >when analyzing (not annotating), and I used twice as much hashtable as I >normally do so that the hashtable size wouldn't bottom out the searches so >often, and I had run it at 60mvs/120minutes, so it was running long, it wasn't >finding Nxf2, I got tired of waiting. > >I was blaming it on my slow computer having trouble searching deeper, maybe >that's not the case. > >Pete A ran a deep search on the position after Nxf2 and thru 15 plies, crafty liked white, although not by a lot. The score for Nxf2 wasn't good, nor bad... around +.2 to -.2 depending on the depth... I suspect anyone that likes this (electronic program) probably is doing so on positional grounds, not based on a tactical slam.
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.