Is Naum 4 a serious challenge for Rybka

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

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
bigo

Is Naum 4 a serious challenge for Rybka

Post by bigo » Thu Jan 01, 2009 8:32 am

Naum is a good positional player?

User avatar
Werner
Posts: 2595
Joined: Wed Mar 08, 2006 9:09 pm

Re: Is Naum 4 a serious challenge for Rybka

Post by Werner » Thu Jan 01, 2009 11:32 am

Hi,
have a look in our CEGT Forum:
Hi all ,

another update after 159 games. Currently the difference between Rybka 3 x64 and Naum 4 x64 would only be 60 points in the quad list and Naum with a lead of 30 points over Rybka 2.3.2a

Code: Select all

CEGT 40/120 rep.  
1   Rybka 3 x64 4CPU  ½½½½1½½½½½0½½½½½110½½1½½½½½½½½01111½½1½½11 25.0/42 
2   Naum 4 x64 4CPU   ½½½½0½½½½½1½½½½½001½½0½½½½½½½½10000½½0½½00 17.0/42
Werner

ernst
Posts: 336
Joined: Thu Mar 09, 2006 5:00 pm

Re: Is Naum 4 a serious challenge for Rybka

Post by ernst » Thu Jan 01, 2009 12:59 pm

I am running a tournament between Naum4 and Rybka3, which you can follow here.
It will only have 60 games and therefore has no statistical meaning, but then again, it is a free service. :D
You can also download the games played.

Nimzovik
Posts: 1831
Joined: Sat Jan 06, 2007 10:08 pm

Re: Is Naum 4 a serious challenge for Rybka

Post by Nimzovik » Thu Jan 01, 2009 4:14 pm

I enjoy the way your site has the optin to replay the current game! I really wish you would do the same for all the games. I just do not want to download the file of the games you supply. (I am a security afficionado) 8-)

User avatar
M ANSARI
Posts: 3500
Joined: Thu Mar 16, 2006 6:10 pm

Re: Is Naum 4 a serious challenge for Rybka

Post by M ANSARI » Thu Jan 01, 2009 8:02 pm

N4 is still a little green to mount a strong challenge to R3. But I think once a few things get worked on then chances are that will get a huge ELO boost. For some reason N4 has very slow search, not sure if it is due to excess redundant code or due to lack of efficient search tricks. It also has weak evaluation of passed pawns and loses way too many games that way. Another thing I think is weaker than Rybka is how much it underestimates the strength of Queen vs. 2 rooks in some positions. Having said that N4 plays very well positionally and will opt for the bishop pair much more than R3 and has changed my view of how strong a bishop pair is. Once its few weaknesses are fixed (which should be not too difficult) it will be a really strong engine. It already is strong enough to be the strongest non Rybka 3 engine.

ernst
Posts: 336
Joined: Thu Mar 09, 2006 5:00 pm

Re: Is Naum 4 a serious challenge for Rybka

Post by ernst » Thu Jan 01, 2009 8:23 pm

The replay function is provided by the CB interface and only supports the current game.

User avatar
Laskos
Posts: 10949
Joined: Wed Jul 26, 2006 8:21 pm
Full name: Kai Laskos

Re: Is Naum 4 a serious challenge for Rybka

Post by Laskos » Thu Jan 01, 2009 8:34 pm

Werner wrote:Hi,
have a look in our CEGT Forum:
Hi all ,

another update after 159 games. Currently the difference between Rybka 3 x64 and Naum 4 x64 would only be 60 points in the quad list and Naum with a lead of 30 points over Rybka 2.3.2a

Code: Select all

CEGT 40/120 rep.  
1   Rybka 3 x64 4CPU  ½½½½1½½½½½0½½½½½110½½1½½½½½½½½01111½½1½½11 25.0/42 
2   Naum 4 x64 4CPU   ½½½½0½½½½½1½½½½½001½½0½½½½½½½½10000½½0½½00 17.0/42
Naum 4 is improving on Rybka 3 on longer time controls and faster hardware. Yes, it IS a challenge for Rybka on say, 40/240 octal games, I don't even know who will win. These positional/tactical talks are beyond me, as they are intermingled for such deep searchers, and, besides that, one has to observe a lot of games.

Kai

User avatar
Mike S.
Posts: 1480
Joined: Thu Mar 09, 2006 4:33 am

Re: Is Naum 4 a serious challenge for Rybka

Post by Mike S. » Fri Jan 02, 2009 3:36 am

M ANSARI wrote: It already is strong enough to be the strongest non Rybka 3 engine.
Yes, that is the point! I consider Naum still a quite "new" engine, but it is already stronger than all the other commercials incl. super heroes like Shredder, Fritz, etc. CEGT 40/20m has Naum 4 x64 on 2 CPU (3069) ahead of Deep Fritz 11 with 4 CPU (3026).

Currently, only Rybka is stronger (but Rybka shouldn't be the only yardstick).

What is it that makes Naum 4 stronger than all the rest? Are there certain types of positions which Naum handles significantly better?
Regards, Mike

Uri Blass
Posts: 8945
Joined: Wed Mar 08, 2006 11:37 pm
Location: Tel-Aviv Israel

Re: Is Naum 4 a serious challenge for Rybka

Post by Uri Blass » Fri Jan 02, 2009 4:29 am

Laskos wrote:
Werner wrote:Hi,
have a look in our CEGT Forum:
Hi all ,

another update after 159 games. Currently the difference between Rybka 3 x64 and Naum 4 x64 would only be 60 points in the quad list and Naum with a lead of 30 points over Rybka 2.3.2a

Code: Select all

CEGT 40/120 rep.  
1   Rybka 3 x64 4CPU  ½½½½1½½½½½0½½½½½110½½1½½½½½½½½01111½½1½½11 25.0/42 
2   Naum 4 x64 4CPU   ½½½½0½½½½½1½½½½½001½½0½½½½½½½½10000½½0½½00 17.0/42
Naum 4 is improving on Rybka 3 on longer time controls and faster hardware. Yes, it IS a challenge for Rybka on say, 40/240 octal games, I don't even know who will win. These positional/tactical talks are beyond me, as they are intermingled for such deep searchers, and, besides that, one has to observe a lot of games.

Kai
The result here suggest no evidence that Naum is better relative to rybka at long time control.

There are more draws in long time control so the right comparison to see if Naum improves is simply to count only wins

Rybka won 10-3 if you do not count draw games.
Can Naum do better in decisive games at faster time control?

Uri

swami
Posts: 6546
Joined: Thu Mar 09, 2006 3:21 am

Re: Is Naum 4 a serious challenge for Rybka

Post by swami » Fri Jan 02, 2009 6:21 am

I have read a lot of posts about the test between Naum and Rybka, I think Naum is not yet at Rybka's level, It's some 60-80 elo points behind but It would perform even better with the next update :)

Post Reply