chess 960 engines vs humans

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

Moderators: Harvey Williamson, bob, hgm

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
stavros
Posts: 160
Joined: Tue Dec 02, 2014 12:29 am

chess 960 engines vs humans

Post by stavros » Fri Jul 29, 2016 10:13 am

i would like to know if the top engines are even better than humans in 960 chess than
clasical chess,any opinion?

Modern Times
Posts: 2604
Joined: Thu Jun 07, 2012 9:02 pm

Re: chess 960 engines vs humans

Post by Modern Times » Fri Jul 29, 2016 12:00 pm

Logically I would have thought so, because the human can't rely on memorised opening theory, but if that is true in practice I don't know.

Perhaps the next Komodo vs human match can be chess960.... not that that would answer the question, but it would be good nevertheless :)
.

Opinions expressed here are my own, and not necessarily those of the CCRL Group.

User avatar
yurikvelo
Posts: 583
Joined: Sat Dec 06, 2014 12:53 pm

Re: chess 960 engines vs humans

Post by yurikvelo » Fri Jul 29, 2016 12:35 pm

stavros wrote:i would like to know if the top engines are even better than humans in 960 chess than
clasical chess,any opinion?
it might depend on rules of classical tournament: what kind of books to allow for machine team.


If define "engine vs human superiority" as minimal amount of CPU resources (speed * time-control) one need to get equivalent ELO strength, or to say "which is the slowest CPU and fastest time control needed to play on par vs human" = than it is highly dependent on book used.
Book is infinite amount of knowledge at zero resource cost, more moves engine play from book - less speed*time-control it need after to play on its own.

whereagles
Posts: 562
Joined: Thu Nov 13, 2014 11:03 am

Re: chess 960 engines vs humans

Post by whereagles » Fri Jul 29, 2016 12:40 pm

Probably engines will pull off further.

Just look at BrainFish vs other engines. Currently it tops the rankings by ~80 ELO, which is solely due to book prep.

Vinvin
Posts: 4765
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: chess 960 engines vs humans

Post by Vinvin » Fri Jul 29, 2016 1:00 pm

stavros wrote:i would like to know if the top engines are even better than humans in 960 chess than
clasical chess,any opinion?
In 2006, programs was already stronger than average GMs : https://en.wikipedia.org/wiki/Chess960#Computers .
Today they probably are way stronger than top GMs as in regular chess.

User avatar
yurikvelo
Posts: 583
Joined: Sat Dec 06, 2014 12:53 pm

Re: chess 960 engines vs humans

Post by yurikvelo » Fri Jul 29, 2016 1:11 pm

Vinvin wrote:
stavros wrote:i would like to know if the top engines are even better than humans in 960 chess than
clasical chess,any opinion?
In 2006, programs was already stronger than average GMs : https://en.wikipedia.org/wiki/Chess960#Computers .
Today they probably are way stronger than top GMs as in regular chess.
http://www.chesstigers.de/index_news.ph ... rubrik=100
For the first time a computer program managed to win a Chess960 game against a strong grandmaster.
Peter Svidler easily won his match against the Dutch program The Baron from Richard Pijl 1,5-0,5.
In chess960, engines reached the same "human vs machine" 10 years later

User avatar
George
Posts: 682
Joined: Sun Oct 13, 2013 11:44 pm

Re: chess 960 engines vs humans

Post by George » Wed Aug 10, 2016 3:17 pm

Modern Times wrote:Logically I would have thought so, because the human can't rely on memorised opening theory, but if that is true in practice I don't know.

Perhaps the next Komodo vs human match can be chess960.... not that that would answer the question, but it would be good nevertheless :)
When we compare standard chess to Chess960 there is only one good logical thing that we have to keep in mind when we use computer versus human, and that is that computer in standard chess uses large opening databases that it is impossible for any super GM like Carlsen or GM Nakamura to memorize ALL the latest opening variations; therefore, computer have a large advantage over humans. In Chess960 Super GM like GM Nakamura that was the last Chess960 Champion have a better chance against Komodo than in standard chess simply relying on vast chess opening principles.

rcmaddox
Posts: 180
Joined: Mon Oct 14, 2013 6:59 pm
Location: Winder, GA
Full name: Robert C. Maddox

Re: chess 960 engines vs humans

Post by rcmaddox » Wed Aug 10, 2016 7:29 pm

George wrote:
Modern Times wrote:Logically I would have thought so, because the human can't rely on memorised opening theory, but if that is true in practice I don't know.

Perhaps the next Komodo vs human match can be chess960.... not that that would answer the question, but it would be good nevertheless :)
When we compare standard chess to Chess960 there is only one good logical thing that we have to keep in mind when we use computer versus human, and that is that computer in standard chess uses large opening databases that it is impossible for any super GM like Carlsen or GM Nakamura to memorize ALL the latest opening variations; therefore, computer have a large advantage over humans. In Chess960 Super GM like GM Nakamura that was the last Chess960 Champion have a better chance against Komodo than in standard chess simply relying on vast chess opening principles.
I think the exact opposite is true. The GM suffers when his knowledge of opening theory is thrown out the window. In fact, I think Deep Blue would have mauled Kasparov in '96 if Chess 960 had been played.

Just my opinion, of course, which obviously I can't prove.

Dirt
Posts: 2851
Joined: Wed Mar 08, 2006 9:01 pm
Location: Irvine, CA, USA

FRC engines vs humans

Post by Dirt » Wed Aug 10, 2016 10:13 pm

George wrote:When we compare standard chess to Chess960 there is only one good logical thing that we have to keep in mind when we use computer versus human, and that is that computer in standard chess uses large opening databases that it is impossible for any super GM like Carlsen or GM Nakamura to memorize ALL the latest opening variations; therefore, computer have a large advantage over humans. In Chess960 Super GM like GM Nakamura that was the last Chess960 Champion have a better chance against Komodo than in standard chess simply relying on vast chess opening principles.
The FRC openings may not be known as deeply as those of standard chess, but there are already large databases of them. Engines can access that data in an instant while GMs struggle to learn a small part of them.
Deasil is the right way to go.

User avatar
George
Posts: 682
Joined: Sun Oct 13, 2013 11:44 pm

Re: FRC engines vs humans

Post by George » Thu Aug 11, 2016 10:39 pm

Dirt wrote:
George wrote:When we compare standard chess to Chess960 there is only one good logical thing that we have to keep in mind when we use computer versus human, and that is that computer in standard chess uses large opening databases that it is impossible for any super GM like Carlsen or GM Nakamura to memorize ALL the latest opening variations; therefore, computer have a large advantage over humans. In Chess960 Super GM like GM Nakamura that was the last Chess960 Champion have a better chance against Komodo than in standard chess simply relying on vast chess opening principles.
The FRC openings may not be known as deeply as those of standard chess, but there are already large databases of them. Engines can access that data in an instant while GMs struggle to learn a small part of them.
When there is only one set position as it is in standard chess the opening databases is more complete for almost all kinds of openings even the 1.h3 Clemenz opening, whereas with so many positions in FRC the openings is NOT even 10% explored with all different variations that can be created by humans GM. Therefore, humans GM have a greater chance of coming up with and even chance from the very opening without having to memorize all the possibilities.

Post Reply