Deep Rybka 3 Q6600 - Shredder 11 A1200 13½-½

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

krazyken

Re: Deep Rybka 3 Q6600 - Shredder 11 A1200, 3½-½ now 22½-1½

Post by krazyken »

Dr.Ex wrote:In all games with Rybka 3 playing white I see one book loss after the next.
Sorry, your testing method is outdated.

Of course Shredder would also receive a beating if equal opening positions had to be played with both colours.
Just not so drastically...
You seem to fail to understand what the SSDF tests. Their testing method is correct for what they are testing. Go read their website foe more info. I appreciate the SSDF as they test how well chess products play chess. Other testers such as the CCRL focus on what they deem to be engine strength, the ability to play well from any postion, while the results from that are likely analogous, it is not the same as a chess rating.
User avatar
Bill Rogers
Posts: 3562
Joined: Thu Mar 09, 2006 3:54 am
Location: San Jose, California

Re: Deep Rybka 3 Q6600 - Shredder 11 A1200, 18½-2½ now 41-4

Post by Bill Rogers »

Tony
Is there a place where we can download the pgn's for all these games?
Thanks
Bill
User avatar
icander
Posts: 921
Joined: Fri Mar 10, 2006 11:21 am
Location: Lulea, Sweden

Re: Deep Rybka 3 Q6600 - Shredder 11 A1200, 18½-2½ now 41-4

Post by icander »

Bill Rogers wrote:Tony
Is there a place where we can download the pgn's for all these games?
Thanks
Bill
http://home.interact.se/~w100107/welcome.htm

Do you still have problem with copy and paste the games?
Tony, SSDF
Dr.Ex
Posts: 202
Joined: Sun Jul 08, 2007 4:10 am

Re: Deep Rybka 3 Q6600 - Shredder 11 A1200, 3½-½ now 22½-1½

Post by Dr.Ex »

krazyken wrote:
Dr.Ex wrote:In all games with Rybka 3 playing white I see one book loss after the next.
Sorry, your testing method is outdated.

Of course Shredder would also receive a beating if equal opening positions had to be played with both colours.
Just not so drastically...
You seem to fail to understand what the SSDF tests. Their testing method is correct for what they are testing. Go read their website foe more info. I appreciate the SSDF as they test how well chess products play chess. Other testers such as the CCRL focus on what they deem to be engine strength, the ability to play well from any postion, while the results from that are likely analogous, it is not the same as a chess rating.
You seem to fail to understand a couple of things.

It is not interesting at all for the majority of users how well chess engine and provided opening book plays against a chess engine with an older book. Especially since most books from Chessbase are Human Chess player oriented and not suitable at all for engine games.

A minority is mostly interested in results from computer vs. computer games. These "nerds" can easily download a better opening book from various sites on the net.

The SSDF rating list is unfortunately often cited for advertising purposes.
Testing the package, above all on different hardware is misleading.
There is no doubt that testing approximately even opening positions with alternate colours gives much more reliable results.
User avatar
Bill Rogers
Posts: 3562
Joined: Thu Mar 09, 2006 3:54 am
Location: San Jose, California

Re: Deep Rybka 3 Q6600 - Shredder 11 A1200, 18½-2½ now 41-4

Post by Bill Rogers »

Tony
Do you think that when you finish this tournement that you could email me the PGN's for the games? I really would appreciate it.
Thanks
Bill
User avatar
icander
Posts: 921
Joined: Fri Mar 10, 2006 11:21 am
Location: Lulea, Sweden

Re: Deep Rybka 3 Q6600 - Shredder 11 A1200, 18½-2½ now 41-4

Post by icander »

Bill Rogers wrote:Tony
Do you think that when you finish this tournement that you could email me the PGN's for the games? I really would appreciate it.
Thanks
Bill
All the games are here in pgn-format. Just copy and paste them to a textfile called bill.pgn or whatever. Then every chessprogram can read it. Everyone else does it.
Tony, SSDF
krazyken

Re: Deep Rybka 3 Q6600 - Shredder 11 A1200, 3½-½ now 22½-1½

Post by krazyken »

Dr.Ex wrote: There is no doubt that testing approximately even opening positions with alternate colours gives much more reliable results.
More reliable for what? Please be more specific.