Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fruitt X Beta results on the Playchess server

Author: Uri Blass

Date: 09:30:55 09/20/05

Go up one level in this thread


On September 20, 2005 at 11:49:08, Djordje Vidanovic wrote:

>For the past couple of days I've been running a copy of Fruit X Beta (the
>pre-WCCC competition dev-compile, running about 30% slower than the WCCC
>version) on the Playchess server.  The machine it's been running on is an FX-53,
>with 512 MB hash, mostly at 15 min + 5 sec increment, or game in 16 min (slow
>time controls on the server).
>
>These are the scores after 40 games:
>
>
>  1 Mans,Fruit X 05/08/09                  26.0/40
>  2 Turklion,Deep Fritz 8            2609   1.5/4
>  3 Murx,Fritz 8                     2517   1.5/3
>  4 Merepawn,Deep Shredder 9         2479   1.5/3
>  5 Great Ozzie,Shredder 9           2444   1.5/3
>  6 Szemcsó,Toga II 1.0              2399   1.5/2
>  7 Murx,Deep Shredder 9             2409   1.0/2
>  8 Cattoo                           2405   1.0/2
>  9 Katzenmaier,Toga II 1.0 Beta2k8  2324   1.0/1
> 10 Felix 2,Deep Shredder 9                 1.0/1
> 11 The Tromp                               0.5/3
> 12 Auryn                            2490   0.5/1
> 13 Great Ozzie,Toga II 1.0          2479   0.5/1
> 14 Wronskij,Toga II 1.0             2464   0.5/1
> 15 K_A_R_,Toga II 1.0               2415   0.5/1
> 16 Felix 2                          2547   0.0/2
> 17 Pizza Margherita,Toga II 1.0     2447   0.0/2
> 18 ETHANOL,Deep Shredder 9          2372   0.0/2
> 19 Auryn,Shredder 9                 2508   0.0/1
> 20 Tondern12,Deep Fritz 8           2425   0.0/1
> 21 Tondern12,Deep Shredder 9        2407   0.0/1
> 22 Alansacount                      2404   0.0/1
> 23 Takker                           2394   0.0/1
> 24 Szemcsó                                 0.0/1
>
>
>Fruit X managed to score 65% against very strong opponents.  Many of them ran on
>machines much faster than Fruit X but lost their mini matches.  For example,
>Turk Lion, running Deep Fritz 8 on a dual Opteron lost by 2.5-1.5, and the same
>applies to Tondern12 which tried both Deep Fritz 8 and Deep Shredder 9 on his
>dual Xeon 3.2 Ghz, and lost both games.
>
>Fruit X (Beta of the WCCC version) is an impressive program that seems to be at
>least 60 elo points stronger than Fruit 2.1.  It uses heavier pruning than Fruit
>2.1 and thus is able to reach greater depths.  It is also much more aggressive
>than its predecessor and, seemingly, has more endgame knowledge.

I read that no additional knowledge was added from fruit2.1 to FruitWCCC except
bonus for the side to move.

Bonus for the side to move may be productive in the endgame but I cannot call it
"endgame knowledge".

Uri



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.