Discussion of computer chess matches and engine tournaments.
Moderators: hgm, Dann Corbit, Harvey Williamson
Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
-
Werner
- Posts: 2598
- Joined: Wed Mar 08, 2006 9:09 pm
Post
by Werner » Fri Dec 30, 2011 10:20 pm
Hi all,
our actual blitz rating list is online and can be found under the attached link.
40 / 4:
New games: 7100
All games now: 950.943
New Engines
22 Critter 1.4 x64 1CPU : 3189 - 1300 games (+21 to version 1.2: nr. 3 in our single list)
35 Komodo 4.0 x64 1CPU: 3163 - 1400 games (+9 to previous version; nr. 5 in our single list)
166 Chiron 1.1a w32 1CPU: 2966 - 1000 games (+4 to v. 1.0)
581 Cheng3 v1.06 x64: 2675 - 1000 games (+19 to v. 1.05)
720 Alfil 11 1CPU: 2610 - 1000 games (behind 8.11 opti)
One Update
105 Chiron 1.1a x64 1CPU: 3025 - 1100 games (-14 and +7 to version 1.0 now)
A big „Thank you“ to our Blitz-List-testers as usual!!
Links
Blitz:
http://www.husvankempen.de/nunn/blitz.htm
Tester:
http://www.husvankempen.de/nunn/testers/testers.htm
We wish you all the best in 2012!
Werner Schuele
CEGT-Team
-
Houdini
- Posts: 1471
- Joined: Mon Mar 15, 2010 11:00 pm
-
Contact:
Post
by Houdini » Fri Dec 30, 2011 10:56 pm
Hello Werner,
If Critter 1.4 is only nr. 3, I was curious to find out who was nr. 2, turns out it is Strelka 5.0.
Are the CEGT aware that Strelka 5.0 is 99% Houdini 1.5? (assuming that the code that Jury Osipov sent me really produces the binary that is distributed)
I don't mind finding several copies of Houdini in rating lists, but it's probably unfair towards other engines.
Robert
-
Osipov Jury
- Posts: 186
- Joined: Mon Jan 21, 2008 1:07 pm
- Location: Russia
Post
by Osipov Jury » Fri Dec 30, 2011 11:56 pm
Houdini wrote:
Strelka 5.0 is 99% Houdini 1.5
Would you be kind enough to prove your claim?
I went through the Rybka code forwards and backwards and took many things.
-
gerold
- Posts: 10121
- Joined: Wed Mar 08, 2006 11:57 pm
- Location: van buren,missouri
Post
by gerold » Sat Dec 31, 2011 12:33 am
LucenaTheLucid wrote:Now THIS will be interesting.
Maybe Houdini is 99% Strelke.
-
Houdini
- Posts: 1471
- Joined: Mon Mar 15, 2010 11:00 pm
-
Contact:
Post
by Houdini » Sat Dec 31, 2011 1:59 am
Osipov Jury wrote:Houdini wrote:
Strelka 5.0 is 99% Houdini 1.5
Would you be kind enough to prove your claim?
I'm not going to invest any time in this.
Consider my message a courtesy towards other engine authors at the top of the ranking lists.
Robert
-
jdart
- Posts: 4103
- Joined: Fri Mar 10, 2006 4:23 am
- Location: http://www.arasanchess.org
Post
by jdart » Sat Dec 31, 2011 2:17 am
Strelka appears to have very large eval terms for king safety (a bit like Scorpio), while Houdini doesn't, in my experience.
--Jon
-
Laskos
- Posts: 10949
- Joined: Wed Jul 26, 2006 8:21 pm
- Full name: Kai Laskos
Post
by Laskos » Sat Dec 31, 2011 2:20 am
Houdini wrote:Osipov Jury wrote:Houdini wrote:
Strelka 5.0 is 99% Houdini 1.5
Would you be kind enough to prove your claim?
I'm not going to invest any time in this.
Consider my message a courtesy towards other engine authors at the top of the ranking lists.
Robert
The first courtesy of most rating lists was to test IvanHoe more than a year ago, when it was a close second to Houdini 1.03, and ahead of anything else these rating groups tested. And, if I am not wrong, it was Houdini 1.03 the close first of IvanHoe
The second thing to consider, is that Critter 1.4 appears to be closer to Strelka 5 than to Critter 1.2 or anything else on similarity. Strelka 5 appeared before Critter 1.4.
Kai
-
Osipov Jury
- Posts: 186
- Joined: Mon Jan 21, 2008 1:07 pm
- Location: Russia
Post
by Osipov Jury » Sat Dec 31, 2011 8:31 am
Houdini wrote:Osipov Jury wrote:Houdini wrote:
Strelka 5.0 is 99% Houdini 1.5
Would you be kind enough to prove your claim?
I'm not going to invest any time in this.
Consider my message a courtesy towards other engine authors at the top of the ranking lists.
Robert
Robert! Happy New Year!
Maybe you release the source code of Houdini 1.5 and Strelka 5? Then everyone can compare them.
P.S.
I would like to see the source of only one function of Houdini. This endgame function KRPKB. It contains a mistake, but I do not know how to fix it.
I went through the Rybka code forwards and backwards and took many things.
-
Werner
- Posts: 2598
- Joined: Wed Mar 08, 2006 9:09 pm
Post
by Werner » Sat Dec 31, 2011 9:03 am
Hello Robert,
...well, we had a discussion about Strelka in the group.
There will be no more tests with it now.
I wish you all the best for 2012!
Werner
