Chess Genius 5 - first impression
Hansjoerg Zech
Hi,
Last week I got my Chess Genius 5 copy from Gambit Soft
http://www.gambitsoft.com (Germany). To get a first impression I made
some games between Genius 5 against MChess 6 and Fritz 4 (30min Games on
a Penium 133). MChess 6 seems to have no chance against Genius (Genius
won four games, 2 remis). Same impression against Fritz. But in 4 games
Genius (3 won) there was one lost (Genius came with a -2 pawn evaluation
out of book). The lost game wasn't repeatable (learning feature!).
The games were on a single machine without permanent brain and equal
hash table size. Games with longer time controls on two machines may
produce different results - but Genius 5 did much better than Genius 4
under same conditions!
Interesting Point regarding Genius 4/3 and SSDF:
In the past Genius 3 (P90) has a better SSDF ranking.
Now (SSDF list 9):
Genius 4 (P90) same ranking as Genius 3 (P90)
Genius 4 (486) better than Genius 3 (486)
Now Genius 4 seems to be no step back (some people believe that) and
Genius 5 seems to be a big step further!
Hansjoerg
Hans-Henrik Grand
>Hi,
>Last week I got my Chess Genius 5 copy from Gambit Soft
>http://www.gambitsoft.com (Germany). To get a first impression I made
>some games between Genius 5 against MChess 6 and Fritz 4 (30min Games on
>a Penium 133). MChess 6 seems to have no chance against Genius (Genius
>won four games, 2 remis). Same impression against Fritz. But in 4 games
>Genius (3 won) there was one lost (Genius came with a -2 pawn evaluation
>out of book). The lost game wasn't repeatable (learning feature!).
>The games were on a single machine without permanent brain and equal
>hash table size. Games with longer time controls on two machines may
>produce different results - but Genius 5 did much better than Genius 4
>under same conditions!
[snip]
As long as you are testing on the same machine the result is unreliable
as CPU time is not equally for both programs
Regards /hhg
--
************** Hans-Henrik Grand ****************
** ** ** Computer Science Department ** ** **
** ** ** Aarhus University ** ** **
*********** email: h...@daimi.aau.dk ****************
Hansjoerg Zech
I make intensive checks to verify that both programs have nearly same
conditions (hash tables, permanent brain out, same slowdown for both
programs, dos box configuration,..).
I also made postgame analysis (it shows me that the programs had made
same moves if running alone!)
But I also think that running the programs on two equal machines would
give more reliable results.
Hansjoerg
--
Hansjoerg Zech
Internet: zech...@sni.de
X.400: c=DE;a=DBP;p=SCN;o=SNI;ou1=PDB1;ou2=S1;s=ZECH;g=HANSJOERG
SNI ASW C8, Heinz Nixdorf Ring 1, 33106 Paderborn
Phone: +49 5251 8 14926 Fax: +49 5251 8 14902
Francesco Di Tolla
> as CPU time is not equally for both programs
depends on the type of programs: I was running crafty vs. gbuchess via
winboard, and I could also check the CPU time used with some windows
tools. It comes out that essentially the share 50% of the CPU time
(actually crafty is a bit smarter).
So I think in this case it makes a lot of sense: even a 5% difference
in the CPU time would not affect the overall result.
regards
Franz
--
Francesco Di Tolla, Center for Atomic-scale Materials Physics
Physics Departement, Build. 307, Technical Univesity of Denmark,
DK-2800 Lyngby, Denmark, Tel.: (+45) 4525 3208 Fax: (+45) 4593 2399
mailto:dit...@fysik.dtu.dk http://www.fysik.dtu.dk/persons/ditolla.html
Robert Hyatt
Francesco Di Tolla (dit...@fysik.dtu.dk) wrote:
: Hans-Henrik Grand wrote:
: >
: > In <32A6B5...@sni.de> Hansjoerg Zech <zech...@sni.de> writes:
: > As long as you are testing on the same machine the result is unreliable
: > as CPU time is not equally for both programs
:
: depends on the type of programs: I was running crafty vs. gbuchess via
: winboard, and I could also check the CPU time used with some windows
: tools. It comes out that essentially the share 50% of the CPU time
: (actually crafty is a bit smarter).
:
: So I think in this case it makes a lot of sense: even a 5% difference
: in the CPU time would not affect the overall result.
:
: regards
: Franz
: --
For crafty, yes. However, some like (apparently) CM5000 really hog the
cpu all of the time somehow. and can really screw up the game result...
pit...@aol.com
Im Artikel <32A7D4...@sni.de>, Hansjoerg Zech <zech...@sni.de>
schreibt:
>I make intensive checks to verify that both programs have nearly same
>conditions (hash tables, permanent brain out, same slowdown for both
>programs, dos box configuration,..).
>I also made postgame analysis (it shows me that the programs had made
>same moves if running alone!)
>But I also think that running the programs on two equal machines would
>give more reliable results.
That`s the only serious way ;-)
-Peter
Komputer Korner
Hansjoerg Zech wrote:
>
> Hans-Henrik Grand wrote:
> >
> > In <32A6B5...@sni.de> Hansjoerg Zech <zech...@sni.de> writes:
>
> conditions (hash tables, permanent brain out, same slowdown for both
> programs, dos box configuration,..).
> I also made postgame analysis (it shows me that the programs had made
> same moves if running alone!)
> But I also think that running the programs on two equal machines would
> give more reliable results.
>
> --
> Hansjoerg Zech
> Internet: zech...@sni.de
> X.400: c=DE;a=DBP;p=SCN;o=SNI;ou1=PDB1;ou2=S1;s=ZECH;g=HANSJOERG
> SNI ASW C8, Heinz Nixdorf Ring 1, 33106 Paderborn
> Phone: +49 5251 8 14926 Fax: +49 5251 8 14902
Make sure that that the 2 equal machines are really equal by getting
exact copies made by the same manufacturer. One Pentium 90 is not
necessarily the same as another Pentium 90 if the manufacturers
are different. Even with the same identical machine, there can be
differences.
--
Komputer Korner
The komputer that kouldn't keep a password safe from
prying eyes, kouldn't kompute the square root of 36^n,
kouldn't find the real motive in ChessBase and missed
the real learning feature of Nimzo. Long live Nimzo!!!!!!
Pappu L N Murthy
I would like to know whether there has been any specific
improvements in the chess engine itself but not any of the
bells and whistles. I have read all about the features many
times, it is not clear at all whether anything was done
specifically to the chess engine. Also, Genius-4 came with
more than 200,000 opening positions according to the manual.
Genius-5 accourding to the reports has only 220,000 positions.
This means after waiting for about two or so years we hardly
have any deeper or extra opening lines at all. ALso, most
probably the engine is same as 4 or 3 or 2.. probably no
change there at all! I wish the author comes forward and
clarifies this issue. For me the bells and whistles and 3-D
boards, coach room etc nonsense is pure nonsense and I could
not care less fot those features.
Pappu L.N. Murthy
--