Phalanx XXIV

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

User avatar
fern
Posts: 8755
Joined: Sun Feb 26, 2006 4:07 pm

Re: Phalanx XXIV

Post by fern »

Hi Susan:
Just played a game at 30 minutes and got a draw after having an extra pawn of advantage for me, but I got bored, Xmas diner was approaching and besides I had the feeling the extra pawn could not be used for good. Even so I am almost sure there was a way to get the full point, but I am lazy.
I can send you the game if you feel with the desire to examine how to improve even more the program so never again a patzer like me can get half the point.
Fern
carldaman
Posts: 2283
Joined: Sat Jun 02, 2012 2:13 am

Re: Phalanx XXIV

Post by carldaman »

dusando wrote:This is a perfect date to release the XXIV version. Here it is, the Phalanx Christmas packages are available from https://sourceforge.net/projects/phalan ... on%20XXIV/

Changes are:
- added late move reductions and pruning
- static evaluation - better passed pawn and mobility bonuses
- hashtable size can now be set via the GUI
- xboard automatic engine load support
- other minor fixes

Merry Christmas!
Big thanks, Dusan, and Merry Chrismas to you! :D

Best wishes,
CL
dusando
Posts: 27
Joined: Thu May 01, 2014 12:58 pm

Re: Phalanx XXIV

Post by dusando »

Just found CEGT has already 40/4 rating for XXIV, at 2691 it's 93 points higher than XXIII. CEGT uses a different rating system than CCRL, the differences are usually larger in CCRL.

http://www.husvankempen.de/nunn/40_4_Ra ... n/889.html
Wolfgang
Posts: 971
Joined: Sat May 13, 2006 1:08 am

Re: Phalanx XXIV

Post by Wolfgang »

dusando wrote:Just found CEGT has already 40/4 rating for XXIV, at 2691 it's 93 points higher than XXIII. CEGT uses a different rating system than CCRL, the differences are usually larger in CCRL.

http://www.husvankempen.de/nunn/40_4_Ra ... n/889.html
Hi Dusan,

there seems to be a problem with a new ORDO version we used for the latest update, so ratings and differences might not be correct.

Gerhard will hopefully manage to make a new update with a fixed ORDO version this evening or tomorrow.

Claculated with ELOStat the difference was +126, as you can see here, http://cegt.forumieren.com/t241-testing ... d-xxiv-w32
Best
Wolfgang
CEGT-Team
www.cegt.net
www.cegt.forumieren.com
dusando
Posts: 27
Joined: Thu May 01, 2014 12:58 pm

Re: Phalanx XXIV

Post by dusando »

Thanks Wolfgang, this has been fixed and now the CEGT 40/4 pages show 139 points increase, XXIII at 2231 and XXIV at 2370.
http://www.husvankempen.de/nunn/40_4_Ra ... /1050.html
Wolfgang wrote:
dusando wrote:Just found CEGT has already 40/4 rating for XXIV, at 2691 it's 93 points higher than XXIII. CEGT uses a different rating system than CCRL, the differences are usually larger in CCRL.

http://www.husvankempen.de/nunn/40_4_Ra ... n/889.html
Hi Dusan,

there seems to be a problem with a new ORDO version we used for the latest update, so ratings and differences might not be correct.

Gerhard will hopefully manage to make a new update with a fixed ORDO version this evening or tomorrow.

Claculated with ELOStat the difference was +126, as you can see here, http://cegt.forumieren.com/t241-testing ... d-xxiv-w32
Wolfgang
Posts: 971
Joined: Sat May 13, 2006 1:08 am

Re: Phalanx XXIV

Post by Wolfgang »

dusando wrote:Thanks Wolfgang, this has been fixed and now the CEGT 40/4 pages show 139 points increase, XXIII at 2231 and XXIV at 2370.
http://www.husvankempen.de/nunn/40_4_Ra ... /1050.html

...
Congratulations! :D
Best
Wolfgang
CEGT-Team
www.cegt.net
www.cegt.forumieren.com