Vitruvius- New Beta Version v Rybka 4.1

Discussion of computer chess matches and engine tournaments.

Moderators: hgm, Rebel, chrisw

User avatar
geots
Posts: 4790
Joined: Sat Mar 11, 2006 12:42 am

Vitruvius- New Beta Version v Rybka 4.1

Post by geots »

Vitruvius_1.1C_x32 vs. Rybka 4.1 w32


First results with Vitruvius 1.1, which I expect may be released sometime this week as an official version. Certain issues still being resolved.




1CPU/32bit
128MB hash
Bases=NONE (Rybka 4.1 used Nalimov)
Ponder_Learning=OFF
Perfect 12.32 book w/12-move limit
40/3 Repeating
Match=50 games


Code: Select all

1   Vitruvius_1.1C_x32   +28    +13/-9/=28   54.00%   27.0/50
2   Rybka 4.1 w32        -28    +9/-13/=28   46.00%   23.0/50





When my worldly duties are over, and my time on this earth has passed- I want to be buried face down, so the...................... I better stop here. I don't have a clue why I even started.

g
e
o
r
g
e
ernest
Posts: 2041
Joined: Wed Mar 08, 2006 8:30 pm

Re: Vitruvius- New Beta Version v Rybka 4.1

Post by ernest »

geots wrote:

Code: Select all

1   Vitruvius_1.1C_x32   +28    +13/-9/=28   54.00%   27.0/50
2   Rybka 4.1 w32        -28    +9/-13/=28   46.00%   23.0/50
Well, how did you finally manage to get that +28 -28 Elo in your table? :)
User avatar
geots
Posts: 4790
Joined: Sat Mar 11, 2006 12:42 am

Re: Vitruvius- New Beta Version v Rybka 4.1

Post by geots »

ernest wrote:
geots wrote:

Code: Select all

1   Vitruvius_1.1C_x32   +28    +13/-9/=28   54.00%   27.0/50
2   Rybka 4.1 w32        -28    +9/-13/=28   46.00%   23.0/50
Well, how did you finally manage to get that +28 -28 Elo in your table? :)


Had to put the damn figures in manually- then align manually. I checked a bit earlier, and my Fritz 13 gui will include them, but it sticks them at the end. Unless I dreamed that. I have been working on this new 64 bit quad trying to get every file and folder over to it, and it is about to drive me nuts- if I wasn't nuts already. I have an external hard drive that holds probably 750gbs- and I could have done it all in a short time if it was xp. It is this Windows 7 system. Whoever thought up the idea for it should be chained to a stovepipe in hell. But back to your observation. I don't think it is possible in Fritz 10 or 11 for them to show.


Best regards,

george
User avatar
geots
Posts: 4790
Joined: Sat Mar 11, 2006 12:42 am

Re: Vitruvius- New Beta Version v Rybka 4.1

Post by geots »

SzG wrote:
geots wrote:Vitruvius_1.1C_x32 vs. Rybka 4.1 w32


First results with Vitruvius 1.1, which I expect may be released sometime this week as an official version. Certain issues still being resolved.




1CPU/32bit
128MB hash
Bases=NONE (Rybka 4.1 used Nalimov)
Ponder_Learning=OFF
Perfect 12.32 book w/12-move limit
40/3 Repeating
Match=50 games


Code: Select all

1   Vitruvius_1.1C_x32   +28    +13/-9/=28   54.00%   27.0/50
2   Rybka 4.1 w32        -28    +9/-13/=28   46.00%   23.0/50





When my worldly duties are over, and my time on this earth has passed- I want to be buried face down, so the...................... I better stop here. I don't have a clue why I even started.

g
e
o
r
g
e
In my view rather poor result from an Ippolit-based engine against the practically 2-year-old one which has almost sunk in oblivion.


Hi Gabor- nice to hear from you. I was thinking a bit like that, and then I see Rybka beating Stockfish and Komodo quite a bit on 64-bit. And Rybka seems to still be holding firmly in the top 5 or 6 in 1CPU 32bit in your rating list in 40/40. Rybka just seems tough for him to beat- always has. Then Vitruvius will turn around and beat Komodo 20 - 8 and beat Fritz 13 26 - 3. And it just decapitated Chiron 1.1a. Rybka after 2 or 2-1/2 years is still a tough old girl. The first released version, 1.0 beat Rybka by 6 games. And that is the largest margin of victory it has had against Rybka- tho it has never lost. I imagine some might say being ippolit based in a sense- it and Rybka are too much alike. But that makes no sense either, because the top 8 or so Ivanhoe versions literally demolish Rybka. Styles make boxing matches at times- maybe it has a playing style that gives Vitruvius problems. I'm just not really sure. But I am sure that I'm glad to hear from you.


Best regards,

george