Author: Eelco de Groot
Date: 17:47:53 05/05/05
Go up one level in this thread
On May 05, 2005 at 18:48:54, John Merlino wrote: >I apologize in advance if Uri did not want me to post any results or interesting >information while I was Beta testing the latest Movei (v309). > >I also apologize to Uri for not starting my testing until today. However, I'm >finally now testing the engine, starting with a 20-game G/10 match against the >default Chessmaster 9000 personality, with pondering on, on an AMD 2500. > >Here's a great move from Movei: > >[D]8/4qpk1/2p2ppN/p7/2P2Q2/1Pb2N1P/b4PK1/8 w - - 0 39 > >Given that pondering was on, it is quite impressive that Movei found 39.Nf5+! in >just under nine seconds. Analysing the position with CM9_R1 takes 2:33 to find >the move that leads to a perpetual check and a draw. > >How do other engines fare here? > >Very nice work so far. > >I'll post the final results and games of each match when they are finished. > >jm Nice move from Movei 309 John! Do I understand correctly that Chess master is pondering on the same machine, when Movei has to make a move? Is that for testing purposes? Sorry that I'm coming back -inevitably- to my Pro Deo testing again; for the fans of tweaking Pro Deo: this position seems to be a case where it is better not to have the [Adjusted Selectivity] switch in your own favorite settings. In Pro Deo 1.1 Betazoid T8 there is [Adjusted Selectivity MIDG = 4] * Maximum selective depth (midgame) [Adjusted Selectivity END0 = 4] * Maximum selective depth (early endgame) [Adjusted Selectivity END1 = 4] * Maximum selective depth (endgame) [Adjusted Selectivity END2 = 3] * Maximum selective depth (simple endgame) This may actually override some other settings besides [Selective Search = 80] that I arbitrarily put in Betazoid T8. I made a slightly different version from Pro Deo 1.1 Betazoid T8, called Betazoid T90 which has default [Selective Search = 100] and none of the [Adjusted Selectivity] settings, and that is much faster in finding the drawscore! It probably plays a bit different positionally though. The choice of move for Betazoid at least may also be influenced by the amount of hashtables it seems! At least in one position I just tested there were different search decisions in Betazoid T8 with 40 Mb hashtables compared to 28 Mb hashtables? But that has to be something for further testing, I only tested that in one particular difficult position. Betazoid T8 needs more than 2 minutes to find 39.Nf5+, Pro Deo 1.1 default is a lot faster, even on a slow Celeron 500 MHz it needs just 6 seconds! Ad hoc settings Betazoid T90 are even one ply faster and get a zero drawscore sooner as well. Pro Deo 1.1 Default (40 Mb hashtables, Celeron 500 Mhz) 00:00:00.8 -2,16 1 79 Nf5+ gxf5 00:00:00.8 -1,64 1 112 Ng4 Bxb3 00:00:00.8 1,07 1 119 Nd4 00:00:00.8 -2,21 2 329 Nd4 Bxd4 Qxd4 Kxh6 00:00:00.8 -1,21 2 385 c5 Bxb3 00:00:00.9 -1,11 2 441 Kg1 Bxb3 00:00:00.9 0,48 2 1028 Qc1 Bb1 Qxc3 Kxh6 Qxa5 00:00:00.9 -0,79 3 1941 Qc1 Bb4 Qc2 Kxh6 00:00:00.9 0,30 3 2674 Nh4 Bb1 Kg1 Qe1+ 00:00:00.9 -1,07 4 7758 Nh4 Qe5 Qc1 Bxb3 00:00:00.9 -0,31 4 9097 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:01.0 -0,31 5 20474 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:01.1 -0,32 6 47146 Qc1 Qb4 Ng4 Bxb3 Qh6+ Kg8 00:00:02.0 -0,58 7 147683 Qc1 Bb4 Ng4 Bxb3 Qh6+ Kg8 Nd4 Qe4+ 00:00:03.2 -1,14 8 328121 Qc1 Qb4 Qe3 Bxb3 Qe8 Qf8 Qe3 a4 Nf5+ 00:00:06.7 -1,14 8 844073 Nf5+ 00:00:06.8 -0,26 8 877597 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg7 Qg3+ Kh7 Qh4+ 00:00:07.2 -0,26 9 932584 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg7 Qg3+ Kh7 Qh4+ 00:00:11.0 0,00 10 1596292 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg7 Qg3+ Kh7 Qh4+ 00:00:17.7 0,00 11 2820909 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg7 Qg3+ Kh7 Qh4+ 00:00:31.7 0,00 12 5470792 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg7 Qg3+ Kh7 Qh4+ 00:01:04.6 0,00 13 11583884 Nf5+ gxf5 Qg3+ Kh7 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg7 Qg3+ Kh7 Qh4+ 00:02:08.8 0,00 14 23160171 Nf5+ gxf5 Qg3+ Kh7 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Pro Deo 1.1 Betazoid T8 00:00:00.5 1,29 1 116 Nd4 00:00:00.5 -2,07 2 307 Nd4 Bxd4 Qxd4 Kxh6 00:00:00.5 -1,18 2 363 c5 Bxb3 00:00:00.5 -1,05 2 419 Kg1 Bxb3 00:00:00.6 0,53 2 1020 Qc1 Bb1 Qxc3 Kxh6 Qxa5 00:00:00.6 0,53 3 2103 Qc1 Bb1 Qxc3 00:00:00.6 -0,18 4 5782 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:00.7 -0,18 5 15408 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:00.8 -0,18 6 35803 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:01.4 -0,75 7 118112 Qc1 Qb4 Ng4 Bxb3 Qh6+ Kg8 Nd2 a4 00:00:04.4 -1,12 8 577105 Qc1 Qb4 Qe3 Bxb3 Qe8 Qf8 Qe3 Qb8 Qxc3 Kxh6 00:00:06.1 -1,11 8 878960 Kh1 Qb4 Ng4 Bxb3 Qh6+ Kg8 c5 Qxc5 00:00:17.5 -1,80 9 2811736 Kh1 f5 Qc1 Bf6 Qd1 Qb4 Nxf5+ 00:00:18.8 -1,22 9 3032648 Qc1 Bb4 Ng4 Bxb3 Qf4 Qd6 Qh6+ Kg8 Kh1 00:00:33.7 -1,30 10 5616701 Qc1 Bb4 Ng4 Bxb3 Qf4 a4 Nh4 00:01:11.6 -1,41 11 12391834 Qc1 Bb4 Ng4 Bxb3 Qf4 a4 Nd4 Bc3 Nxc6 Bd2 00:02:21.8 -1,41 11 24927987 Nf5+ 00:02:28.0 0,00 11 26068360 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh7 Qh4+ Kg6 Qg3+ Kh6 Qg8 Qe4 00:02:44.7 0,00 12 29201950 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ 00:03:49.6 0,00 13 41287854 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh7 Qh4+ Kg8 Qg3+ Kh7 Qh4+ 00:06:51.5 0,00 14 75582782 Nf5+ gxf5 Qg3+ Kf8 Qb8+ Kg7 Qg3+ Kf8 Qb8+ Kg7 Pro Deo 1.1 Betazoid T90 00:00:00.6 -1,85 1 121 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ 00:00:00.6 -1,65 1 159 Ng4 Bxb3 00:00:00.6 1,29 1 166 Nd4 00:00:00.6 -2,07 2 382 Nd4 Bxd4 Qxd4 Kxh6 00:00:00.6 -1,18 2 592 c5 Bxb3 00:00:00.6 -1,05 2 764 Kg1 Bxb3 00:00:00.7 0,53 2 1555 Qc1 Bb1 Qxc3 Kxh6 Qxa5 00:00:00.7 0,58 2 2026 Nh4 Kf8 Qb8+ Qe8 00:00:00.7 0,29 3 2772 Nh4 Bb1 Kg3 00:00:00.7 0,53 3 3149 Qc1 Bb1 Qxc3 Kxh6 00:00:00.8 -0,18 4 7201 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:00.8 -0,18 5 18698 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:01.1 -0,49 6 60374 Qc1 Bb4 Ng4 Bxb3 Qh6+ 00:00:02.5 -1,03 7 220087 Qc1 Qb4 Ng4 Bxb3 Qh6+ Kg8 c5 Qxc5 Nd2 00:00:02.9 -1,00 7 288507 Kg1 Qb4 Ng4 Bxb3 Qh6+ Kg8 c5 Qxc5 00:00:03.1 -1,00 7 335753 Nf5+ 00:00:03.4 0,00 7 348028 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8 00:00:03.9 0,00 8 415664 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8 00:00:05.9 0,00 9 724965 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8 00:00:09.6 0,00 10 1368679 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8 00:00:18.7 0,00 11 3026497 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8 00:00:33.4 0,00 12 5710453 Nf5+ gxf5 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8 00:01:20.5 0,00 13 14438401 Nf5+ gxf5 Qg3+ Kh7 Qh4+ Kg8 Qg3+ Kf8 Qb8+ Qe8 Qd6+ Kg8 Qg3+ Kh8 Qh4+ Kg8 Qg3+ Kh8 Qh4+ Kg8
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.