STS comparison of IPPOLIT, Rybka 3 and Stockfish

Code, algorithms, languages, construction...
Post Reply
Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

STS comparison of IPPOLIT, Rybka 3 and Stockfish

Post by Jeremy Bernstein » Wed Jun 09, 2010 2:39 pm

Attached to this post is a PDF summary of Strategic Test Suite (http://sites.google.com/site/strategictestsuite/) results from IPPOLIT, Rybka 3 and Stockfish. My initial idea was to attempt to demonstrate that IPPOLIT and Rybka 3 have similar weaknesses via the suite, using Stockfish 1.7.1 as a 'control'. The results are far from conclusive, which is to say, I can't identify any correspondences between IPPOLIT and Rybka based on time spent or inability to solve certain positions. I've marked wrong answers in blue and "hard problems" (> 10 seconds of thinking) in orange. All engines were 1-core, 32-bit. Each engine had up to 420 seconds to solve each problem.

Jeremy
STS.pdf
STS 1-9 IPP/RYB/STO
(86.07 KiB) Downloaded 793 times

User avatar
Angel
Posts: 103
Joined: Thu Jun 10, 2010 12:57 am
Contact:

Re: STS comparison of IPPOLIT, Rybka 3 and Stockfish

Post by Angel » Thu Jun 10, 2010 3:02 am

:o :) :idea: An EXELLENT post
Nothing is ever truly lost , just miss placed and awaiting us

gaard
Posts: 127
Joined: Thu Jun 10, 2010 1:39 am
Real Name: Martin Wyngaarden
Location: Holland, Michigan

Re: STS comparison of IPPOLIT, Rybka 3 and Stockfish

Post by gaard » Thu Jun 10, 2010 5:26 am

Jeremy Bernstein wrote:Attached to this post is a PDF summary of Strategic Test Suite (http://sites.google.com/site/strategictestsuite/) results from IPPOLIT, Rybka 3 and Stockfish. My initial idea was to attempt to demonstrate that IPPOLIT and Rybka 3 have similar weaknesses via the suite, using Stockfish 1.7.1 as a 'control'. The results are far from conclusive, which is to say, I can't identify any correspondences between IPPOLIT and Rybka based on time spent or inability to solve certain positions. I've marked wrong answers in blue and "hard problems" (> 10 seconds of thinking) in orange. All engines were 1-core, 32-bit. Each engine had up to 420 seconds to solve each problem.

Jeremy
STS.pdf
When only looking at move choice, there seemed to be significant correlations between the Ippos and Rybka. I saw an analysis by Miguel on CCC and it left me with little doubt that the two play markedly similar moves. I can't speak to the origins of Ippo, or the code, but as far as playing style goes... the two are incredibly close.

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: STS comparison of IPPOLIT, Rybka 3 and Stockfish

Post by Jeremy Bernstein » Thu Jun 10, 2010 8:51 pm

The next batch I'll hit is a 3-way comparison of IPPOLIT, RobboLito and Houdini. I'm mostly just curious about how well STS works as a relationship indicator.

Richard Vida
Posts: 50
Joined: Thu Jun 10, 2010 12:48 am

Re: STS comparison of IPPOLIT, Rybka 3 and Stockfish

Post by Richard Vida » Thu Jun 10, 2010 9:58 pm

BTW, what happened with the material Zach Wegner mentioned in this post?

Richard

mcostalba
Posts: 91
Joined: Thu Jun 10, 2010 11:45 pm
Real Name: Marco Costalba

Re: STS comparison of IPPOLIT, Rybka 3 and Stockfish

Post by mcostalba » Sat Jun 12, 2010 12:10 pm

Richard Vida wrote:BTW, what happened with the material Zach Wegner mentioned in this post?
C'mon Richard, lets throw some good your idea on the table :-) don't just ask for other's :-)

P.S: Take with a bit of joking, not too serious, I say this because from the few posts where you expressed some idea/suggestion I've made up my mind you have a good vision of how an engine works.

User avatar
kingliveson
Posts: 1388
Joined: Thu Jun 10, 2010 1:22 am
Real Name: Franklin Titus
Location: 28°32'1"N 81°22'33"W

Re: STS comparison of IPPOLIT, Rybka 3 and Stockfish

Post by kingliveson » Sat Jun 12, 2010 12:36 pm

Richard Vida wrote:BTW, what happened with the material Zach Wegner mentioned in this post?

Richard

It's in a hidden forum. :o
PAWN : Knight >> Bishop >> Rook >>Queen

Post Reply