Unheralded Zappa wins 2005 World Computer Chess Championship

[Guest]

Archived from groups: rec.games.chess.computer (More info?)

website: http://www.ru.is/wccc05/

Little-known program Zappa (it had won the CCT7 event earlier in the
year, which I helped broadcast on www.chess.fm) was a surprising and
overwhelming 10.5/11 victor at the 2005 World Computer Chess
Championship. The competition included past winners and well-known
programs Deep Junior and Shredder as well as Crafty. Up and coming
"Fruit" (the only contender to run on a single CPU!) was second.
Zappa's programmer is Anthony Cozzie. Zappa ran on two dual processor
PC's (4 CPUs total). Fritz and Hydra did not play but Shredder usually
comes in ahead of Fritz at these events, so Zappa's victory is not
hollow. Of course Hydra was eligible and, as the only hardware machine,
would have been the heavy favorite, but that takes nothing from Zappa's
superb showing. Zappa 1.0 is free to download from its website but this
is version 2...

Congratulations Anthony!

- Dan Heisman
www.danheisman.com

 

[Guest]

Archived from groups: rec.games.chess.computer (More info?)

That was very impressive. What is the author willing to reveal about
what he did to achieve such an impressive result?

 

[Guest]

Archived from groups: rec.games.chess.computer (More info?)

I downloaded the 3 files and unzipped them, what do I do with them now
to make the book and other file interact with the zappa engine? I put
all 3 in a Zappa folder. The authors email didn't work, thanks

 

[Guest]

Archived from groups: rec.games.chess.computer (More info?)

danheisman@comcast.net wrote:

>Up and coming "Fruit" (the only contender to run on a single CPU!)
>was second.

This is false. At least The Crazy Bishop, The Baron and Fute also ran
single CPU.

> Zappa's programmer is Anthony Cozzie. Zappa ran on two dual processor
> PC's (4 CPUs total).

One PC with 2 dual core processors.

> Of course Hydra was eligible and, as the only hardware machine,
> would have been the heavy favorite,

I see no reason why you would expect Hydra to do particularly well in
this event (and apparently, neither did the Hydra team!)

Quite a few participants ran on very fast large multiprocessors machines
which would have voided quite a bit of Hydra's hardware 'advantage'.

--
GCP

 

[Guest]

Archived from groups: rec.games.chess.computer (More info?)

Gian-Carlo Pascutto wrote:

> I see no reason why you would expect Hydra to do particularly well in
> this event (and apparently, neither did the Hydra team!)
>
> Quite a few participants ran on very fast large multiprocessors machines
> which would have voided quite a bit of Hydra's hardware 'advantage'.

Deep Sjeng came in 3rd, tied with Shredder I believe, another very
impressive result! It seems like it was a very interesting tournament,
do you have any comments you can share about it? It seems like some
great progress must have been made increasing the playing strength
of some of the engines, how was that done (if the secrets can be given
away)?

 

[Chuck]

Archived from groups: rec.games.chess.computer (More info?)

I believe it has a lot to do with what I commented before about the program
being able to identify, and focus on the key pieces in play, and not waste
processor time on such things as checking and calculating all the moves
available to a newly castled King that is currently not under any type of
threat. It can understand and assign CPU priority on what is important.
Prior to each response, it may do something like run a risk assessment
algorithm, and then assigns CPU priority accordingly, tries to eliminate any
and all non threats from the task list, then after those checks have all
been made, the core program then takes over to work on what is now a much
smaller job.


<Akorps@aol.com> wrote in message
news:1124958287.969315.264760@g43g2000cwa.googlegroups.com...

Gian-Carlo Pascutto wrote:

> I see no reason why you would expect Hydra to do particularly well in
> this event (and apparently, neither did the Hydra team!)
>
> Quite a few participants ran on very fast large multiprocessors machines
> which would have voided quite a bit of Hydra's hardware 'advantage'.

Deep Sjeng came in 3rd, tied with Shredder I believe, another very
impressive result! It seems like it was a very interesting tournament,
do you have any comments you can share about it? It seems like some
great progress must have been made increasing the playing strength
of some of the engines, how was that done (if the secrets can be given
away)?