Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka 2.3.2 strength
- - By Quapsel (****) [de] Date 2007-06-15 08:19
Hi,

hi, to get a first look at the strength of Rybka 2.3.2 I made a tournament on a Pentium4 2,6 GHz with only 1 Second per move.
Shurely this ist not the most interesting thinking time, but in this manner I received a view results more.
Together with the expieriences of other Users we will get an Overview.

participants were Rybka 2.2, Rybka 2.3.1, Rybka 2.3.2, TogaII 1.2, Spike 1.2 an Glaurung 2
I played the Nunn-Positions, with repetition with changed colours.
And today in the morning I found, that each Engine has played 200 games and the results are:

1: Rybkav2.3.2.w32  136,0/200
2: Rybkav2.2.w32    123,0/200
3: TogaII           109,5/200
4: Rybka v2.3.1.w32 106,5/200
5: Spike1.2          64,5/200
6: Glaurung_win_32   60,5/200


This would say, that Vas (big congratulations from me!) has accomplished to increase the strength of his Rybka by a very respectable and astonishing portion!
I know, that other users noticed not as I did: the 2.3.1-Version has not the strength of the elder 2.2-Version. This Statement could be supported by this tournament.

Because some say, "the games between different Versions of one Program-Line is incest!" Toga, Spike anf Glaurung took part in the tournament.

The results in detail are:

             Rybka 2.3.2       Rybka 2.2         TogaII 1.2        Rybka 2.3.1       Spike 1.2         Glaurung 2
Rybka 2.3.2  -                 23,0/40 16-10-14  28,5/40 21-4-15   24,5/40 17-8-15   28,5/40 22-5-13   31,5/40 28-5-7
Rybka 2.2    17,0/40 10-16-14  -                 25,5/40 19-8-13   21,5/40 13-10-17  26,5/40 23-10-7   32,5/40 29-4-7
TogaII 1.2   11,5/40 4-21-15   14,5/40 8-19-13   -                 20,5/40 15-14-11  33,5/40 30-3-7    29,5/40 24-5-11
Rybka 2.3.1  15,5/40 8-17-15   18,5/40 10-13-17  19,5/40 14-15-11  -                 26,0/40 21-9-10   27,0/40 22-8-10
Spike 1.2    11,5/40 5-22-13   13,5/40 10-23-7    6,5/40 3-30-7    14,0/40 9-21-10   -                 19,0/40 13-15-12
Glaurung 2    8,5/40 5-28-7     7,5/40 4-29-7    10,5/40 5-24-11   13,0/40 8-22-10   21,0/40 15-13-12  -


Rybka reaches against its non-rybka-opponents:

Rybka 2.2      84.5 Points
Rybka 2.3.1    72.5 Points
Rybka 2.3.2    88.5 Points


Rybka seems to be stronger than older Versions, and this takes effect also against non-rybka-Programs.

My first impression is, that Rybka 2.3.2 is a very succeeded Program Version!
:-)
Please post, what are your impressions. (right - wrong?)

Some 'Bugs' of 2.3.2 were presented here, and shurely this is OK and it will be helpfull vor Vas.
But, if 2.3.2 is that strong that it seems to be, this also should be pointed out clearly.

I'll repeat this tournament at the Weekend with longer thinking Time, maybe 5 Minutes + 1 Second, and I will report here.

Quap
Parent - By Quapsel (****) [de] Date 2007-06-18 06:11
During the weekend I made a tournament (Nunn-Positions, colout-change) with 5 Minutes per game + 1 Second per move on my P4 (2,6 GHz)
The result is:

1: Rybkav2.3.2.w32  70,0/110
2: Rybka v2.3.1.w32 64,5/110
3: Rybkav2.2.w32    63,5/110
4: Spike1.2         53,0/110
5: TogaII           49,5/110
6: Glaurung_win_32  29,5/110


Rybka reaches against its non-rybka-opponents:

Rybka 2.2      46.5 Points
Rybka 2.3.1    40,0 Points
Rybka 2.3.2    45.5 Points


Quap
Parent - By Quapsel (****) [de] Date 2007-06-19 09:12
Having 1 Minute per game + 1 Second per move I got

1: Rybkav2.3.2.w32  56,5/80
2: Rybka v2.3.1.w32 49,5/80
3: Rybkav2.2.w32    44,0/80
4: Spike1.2         34,0/80
5: TogaII           33,0/80
6: Glaurung_win_32  23,0/80


Rybka got against its non-rybka-opponents:

Rybka 2.2      32.0 Points
Rybka 2.3.1    34.0 Points
Rybka 2.3.2    36.0 Points


Quap
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka 2.3.2 strength

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill