Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Please Ed, take Rebel Decade 2.0 off of your site before you go bank

Author: Jan-Frode Myklebust

Date: 09:22:18 01/18/98

Go up one level in this thread


Could you do the same test with both programs using the same amount
(512KB) of hash? I'm just curious on how different the two engines are.
I believe Ed Shroeder stated that they were different engines, but not
how different.

janfrode

On December 21, 1997 at 16:25:30, Ernst Walet wrote:

>I did a test with Rebel9 with 28MB hash and Decade2, thus with 512KB
>hash, in "rebel9 mode", to verify if the hash size is as important as i
>stated earlier.
>I took 9 mixed middle- and endgame positions and took the time to find
>the solution or give the solution a fail high.
>
>
>pos		Decade2 in R9 mode		Rebel 9
>		time	ply	score		time	ply	score
>
>1		 2:35	10	 4.57		 0:15	 9	 4.82
>2     		>60:00	13busy	 ?.??		 2:24	10	 1.95
>3		 8:07	21	 6.00		 1:19	20	 5.81
>4		 3:38	 8	 2.35		 2:50	 8	 2.48
>5		 0:43	12	-1.50		 0:26	12	-1.50
>6		 4:14	 9	 2.30		 3:22	 9	 2.16
>7		 7:13	13	 4.18		 3:40	13	 5.29
>8		19:39	12	 1.43		12:00	12	 1.43
>9		24:29	11	 M7		 8:03	11	 M7
>
>And the positions.
>
>N7/p7/8/q1p5/1p6/k7/8/1K1Q4 w - - bm d1c1;
>1Q6/5pb1/4p1k1/6N1/q5Pp/4P2P/1n3P2/2N3K1 w - - bm b8c7;
>k1N5/2p5/8/1P6/8/8/7K/8 w - - bm c8b6;
>r4rk1/pb1p1pp1/3qp1np/1p5Q/3pP3/1B1P1N2/PPPB4/R5RK w - - bm d2h6;
>8/6pp/2B5/4kp2/3r1P2/4K1PP/8/8 b - - bm d4f4;
>rn2r2k/2q3pp/p2pBb2/1p1P4/8/P3B3/1PPQ2PP/R4RK1 w - - bm f1f6;
>5r2/1P5p/2N4K/8/8/5kP1/8/8 w - - bm g3g4;
>8/p5Nk/3B2pp/8/2PP1pn1/4p3/P3P1Bq/R2R1K2 b - - bm g4f2;
>r2q1rk1/ppp2ppp/3b2n1/1N1P4/2BBP3/P4pP1/1PQ2PbP/R3R1K1 b - - bm g6f4;
>
>So although Decade2 with the "R9 feature" is even closer to Rebel9 than
>ever intended by Ed Schröder, Rebel 9 still has a lot more to offer,
>according to me.
>
>Ernst J.



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.