Test no.1 (DiscoCheck w32 comparative)

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Rebel, chrisw

User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: Gaviota 1.0 - mon amour - test !

Post by Sylwy »

Sylwy wrote:INNA (Romania) - Cola Song !

:lol: for Maestro Ballicora & the Team ! :lol:

The newest song of INNA !

==================================
http://www.youtube.com/watch?v=Yz2658gzOuM
==================================

SilvianR :wink:

:lol:

Like Mr.Ruxy promised ! Versus a classic; Shredder 11 UCI of Mr.Stefan Meyer-Kahlen (from Düsseldorf). So, a German produkt !
(BTW, I remembered - they are immense chances that the next President of Romania -elections in November, 2014 - to be Mr.Klaus Werner Iohannis - the mayor of Sibiu/Hermannstadt. And I know very well what I have said http://en.wikipedia.org/wiki/Klaus_Iohannis ) !

Gaviota 1.0 is great ! I have never seen a such nice play to this engine !

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

Gaviota 1.0 was linked to Gaviota TBs (gtb).
Shredder 11 UCI was linked to Shredder bbs & Nalimov TBs.

Regards,
SilvianR :wink:

Image
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: An Arasan never seen in Milky Way !!! 17.0 !!!

Post by Sylwy »

Strong & ellegant play !

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

Arasan 17.0 was linked to Gaviota TBs (gtb).
Deep Sjeng 2.7 was linked to Nalimov TBs.

Regards,
SilvianR :wink:


Image

Image
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: A perfect chess engine - Protector 1.6.0

Post by Sylwy »

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

Protector 1.6.0 was linked to Nalimov TBs.

Very balanced in all the phases of a game ! A fine chess engine !

75-80 ELO points increase over Protector 1.5.0 (my estimation) !

Have - you all - a nice weekend !

SilvianR :wink:

Image
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re:A very interesting comparative test (Nirvanachess)

Post by Sylwy »

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

SmarThink 1.40 was linked to Nalimov TBs.

===================================================
Like a Sunday bonus: a superb traditional melody from Southern Moldavia:

http://www.youtube.com/watch?v=H647ZrNNojo

===================================================

Cheers,
SilvianR :wink:

Image

Image

Image

Image
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: A double test (Gull 3)

Post by Sylwy »

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

All the next month (until the 2-nd of June 2014) Mr. Ruxy will be far from IASI, Romania & .....CCC forum ! :lol:

===========
So, something very special - just for you - exactly from IASI:

http://www.youtube.com/watch?v=6zzfmus1HNk

===========

Regards,
SilvianR :wink:

Image
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: EXchess 7.26b (w32) weaker than 7.18b & 7.01b versio

Post by Sylwy »

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

EXchess (all versions) were linked to Nalimov TBS.
Deuterium was linked to Scorpio egbbs.

Cheers,
Silvian :wink: R

Image

Image

Image
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: A test for strong characters (Strelka 6)

Post by Sylwy »

:lol:

Well, after the new bolshevik & imperial resurrection - the Russian science hits again: Strelka 6 of Herr Jury von O$ipov !!!

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

Image

My conclusion:

Image

Have a nice Saturday & Sunday !

Silvian :wink: R
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: Arasan 17.1w32 < Arasan 17.0w32

Post by Sylwy »

Conditions:

-TC = 4'+2"
-Hash = 128 MB
-Book = Shredder12.ctg
-GUI = Fritz 11
-CPU = Athlon 64 3800+

Arasan 17.0 & 17.1 were linked to Gaviota TBs (gtb).
Deep Sjeng 2.7 was linked to Nalimov TBs.

Image

Image

Silvian :wink: R
jdart
Posts: 4366
Joined: Fri Mar 10, 2006 5:23 am
Location: http://www.arasanchess.org

Re: Arasan 17.1w32 < Arasan 17.0w32

Post by jdart »

Over 32,000 fast time control games, 17.1 is ahead of 17.0 by about 10 ELO (on an arbitrary base of 2400), which is outside the error bars. Not a large different but significant.

--Jon
User avatar
Sylwy
Posts: 4464
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: My assertion is true !

Post by Sylwy »

jdart wrote:Over 32,000 fast time control games, 17.1 is ahead of 17.0 by about 10 ELO (on an arbitrary base of 2400), which is outside the error bars. Not a large different but significant.

--Jon
The chess engines behave differently at different TC ! Fast (ultra fast TC) # blitz 4'+2" (around 15' per game ) !
And 120 games at 4'+2" is enough to shows me the correct trend. I was never in fault. Maybe 5-10 ELO points.
Lets show CCRL & CEGT. At long TC Arasan 17.1 < 17.0 (for the 32-bit version). That's my prediction !
The same for EXchess 7.26bw32 ( < EXchess 7.18bw32).

Hyper-bullet time control it is not an foolproof method. Means near nothing for normal TCs - just my opinion !

Do you remember our old discussion ? Here it is:

http://talkchess.com/forum/viewtopic.ph ... 19&t=49519

"My own tests (38,000 games, hyper-bullet time control) show 16.0 stronger than 15.6..."

Wrong !

CCRL 40/40:

Arasan 15.6 = 2732 ELO.............+20 points
Arasan 15.1 = 2721 ELO
Arasan 16.0 = 2712 ELO

http://www.computerchess.org.uk/ccrl/40 ... t_all.html

Otherwise - often - a new Arasan version is weaker than the previous ! Nothing wrong ! It's normal !
And Arasan is a nice piece of computer chess !

Regards,

Silvian :wink: R