Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rybka #1 in 50 game Blitz1/1 test

Author: Eelco de Groot

Date: 10:26:27 12/05/05

Go up one level in this thread


On December 05, 2005 at 07:34:10, Ryan B. wrote:

>On December 05, 2005 at 07:23:48, Günther Simon wrote:
>
>>On December 05, 2005 at 07:10:33, Ryan B. wrote:
>>
>>>   Engine              Score          Ry         Gf         To         Fr
>>> Gf         Sp    S-B
>>>1: Rybka v1.0 Beta.w32 32.0/50 ·········· 0111===000 1==11==110 1=1==1=110
>>
>>>Level: Blitz 1/1
>>>Hardware: AMD Athlon(tm) Processor 807 MHz with 127 MB Memory
>>
>>Blitz 1/1 and a 800mHz machine surely hurt very slow Rybka.
>>I guess it will be better with other conditions.
>>(means clearer heading for #1 ;-)
>>
>>Guenther
>
>
>In my tests the Gap between #2 and #3 on the list gets bigger at longer time
>controls also but yes Rybka is the clear #1 still in my opinion.  Also so far
>Gambit Fruit does not do better on 64 bit and Rybka does so that makes that gap
>bigger.  Anyway it is a very impressive engine, #1 for sure.
>
>Ryan

Still also a very good performance of Gambit Fruit 4bx there Ryan! For
testpositions of Gambit Fruit I use now

Rebel Reductions
Quiescense Check Plies = 2
History Treshold = 50
(200 Mb HT Athlon 2009 MHz)

But I have no idea how that wold translate to games or Blitz. Vasik does not say
to much about Rybka's QSearch, I wonder if there is much to be discovered there?
In Pro Deo I tested Quiescence Check plies of depth four and that seemed to work
well, I could not see much slow down of Pro Deo. Anyway just some vague ideas.

Ferret Gandalf still a difficult testposition for Gambit Fruit 4b:

default settings but Rebel Reduction enabled:

[D]rr4k1/5p2/3p2p1/3P3p/pnP1P2P/Bn2NR2/5PP1/qB1Q2K1 w - - 1 1

Ferret-Gandalf door Gambit Fruit Beta 4b
Rebel Reductions
200 Mb HT

00:00:00.0	-0,83	1	3		Bxb4 Rxb4
00:00:00.0	-0,97	2	392		Bxb4 Rxb4 Qd3
00:00:00.0	 0,59	3	1633		Bxb4 Nd4 Qe1 Nxf3+ gxf3
00:00:00.0	 0,42	4	2289		Bxb4 Nd4 Qe1 Nxf3+ gxf3 a3
00:00:00.0	-0,42	5	6883		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2
00:00:00.0	-0,42	6	14253		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2
00:00:00.1	-0,73	7	31699		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4
00:00:00.1	-1,03	8	62226		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 c5 dxc5
00:00:00.2	-1,03	9	125093		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 c5 dxc5
00:00:00.4	-1,35	10	277396		Bxb4 Nd4 Qe1 Qb2 Rf6 Qxb4 Qxb4 Rxb4 Bd3 Rb3 Bf1 Rd8
00:00:00.8	-1,56	11	567231		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 Rc3 Rc8 c5 Rxc5
Rxc5 dxc5
00:00:01.7	-1,71	12	1261684		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 Rc3 Rc8 Bd3 a3
Be2
00:00:03.6	-1,55	13	2629808		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 Rc3 Rc8 Bd3 a3
Rc2 Qb1
00:00:07.9	-2,08	14	6000658		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 c5 Rc4 Bd1 Rxc5
Rg3 Rc1 Qf1
00:00:20.5	-2,50	15	15453504	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 a3 Be5
Nxf3+ gxf3 a2 c5 a1Q Bxa1 Rbxa1 c6 Rc1 Kg3 Ra2
00:00:54.3	-2,68	16	40254973	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 a3 Be5
Nxf3+ gxf3 a2 c5 a1Q Bxa1 Rbxa1 c6 Rc1 Kg3 Ra2 Kf4
00:02:27.1	-2,73	17	106902752	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 a3 Be5
Nxf3+ gxf3 Rc1 Kg3 a2 d6 Ra7 f4 a1Q Bxa1 Rg1+ Kf3 Rgxa1
00:06:05.4	-2,62	18	261339750	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 a3 Be5
Nxf3+ gxf3 Rc1 d6 a2 Nd5 a1Q Bxa1 Raxa1 d7 Ra8 Kg3 Rc2 Ne7+ Kg7 Nc6
00:26:13.8	-2,76	19	1110122342	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 a3 Be5
Nxf3+ gxf3 Rc1 d6 a2 Nd5 a1Q Bxa1 Raxa1 d7 Ra8 Kg3 Rd8 Nb6 Rc2 e5
01:22:46.0	-2,27	19	3459265601	Nf5 Ra7 Nh6+ Kg7 Nf5+ Kf8 Nxd6 Qxa3 e5 Qa1 Nb5
Re7 e6 Kg7 Nc3 fxe6 Qe1 e5 Qe4 Rb6 Qe3

Rebel Reductions
Quiescense Check Plies = 2
History Treshold = 50
200 Mb HT

00:00:00.0	-0,83	1	3		Bxb4 Rxb4
00:00:00.0	-0,97	2	471		Bxb4 Rxb4 Qd3
00:00:00.0	 0,59	3	1724		Bxb4 Nd4 Qe1 Nxf3+ gxf3
00:00:00.0	 0,42	4	2405		Bxb4 Nd4 Qe1 Nxf3+ gxf3 a3
00:00:00.0	-0,42	5	7043		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2
00:00:00.0	-0,42	6	14703		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2
00:00:00.0	-0,73	7	33712		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4
00:00:00.1	-1,03	8	68203		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 c5 dxc5
00:00:00.2	-1,03	9	142140		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 c5 dxc5
00:00:00.5	-1,45	10	344837		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 Bd3 a3 c5 dxc5
00:00:01.1	-1,56	11	806236		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 Rc3 Rc8 c5 Rxc5
Rxc5 dxc5
00:00:02.2	-1,71	12	1612975		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 Rc3 Rc8 Bd3 a3
Be2
00:00:05.5	-1,55	13	4144589		Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 Rc3 Rc8 Bd3 a3
Rc2 Qb1
00:00:14.1	-1,97	14	10791287	Bxb4 Nd4 Qe1 Qb2 Nc2 Nxc2 Bxc2 Rxb4 c5 Rc4 Bd1 Rxc5
Qe3 a3 Qh6
00:00:52.7	-2,50	15	40444521	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 Nxf3+
gxf3 a3 Be5 a2 c5 a1Q Bxa1 Rbxa1 c6 Rc1 Kg3 Ra2
00:02:10.9	-2,68	16	100492929	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 Nxf3+
gxf3 a3 Be5 a2 c5 a1Q Bxa1 Rbxa1 c6 Rc1 Kg3 Ra2 Kf4
00:18:16.9	-2,76	17	812107973	Bxb4 Nd4 Qe1 Qb2 Bxd6 Qxb1 Qxb1 Rxb1+ Kh2 Nxf3+
gxf3 a3 Be5 Rc1 d6 a2 Kg3 Ra7 Kf4 Kf8 Bb2 a1Q Bxa1 Rcxa1
00:26:17.6	-2,28	17	1165898761	e5 dxe5 Rg3 Nd2 Kh2 Nxb1 Bxb4 Rxb4 Nc2 Qb2 Nxb4
Qxb4 Qxh5 Qxc4 Rxg6+ Kf8 Rg5 Qg4

But this is a very speculative testposition, as White is lost no matter what if
Black plays well. Haven't tried it with Rybka yet!

 Eelco





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.