Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Tony Marsland and Chinese Chess in Maastricht (slightly O.T.)

Author: Pham Hong Nguyen

Date: 08:36:14 07/30/02

Go up one level in this thread


On July 30, 2002 at 05:57:38, Omid David wrote:

>Looking at the results of the Olympiad in Maastricht, I found out that Tony
>Marsland's Chinese Chess program Abyss'99, got the last place with one draw and
>11 losses out of 12 games. Marsland has published several articles based on his
>experiments with this "Updated Chinese Chess program" [quote from his article
>"Variable Depth Search" 2000]. I wonder, what caused his program to perform so
>poorly?

That is only results against 3 other programs - 12 games for Abyss and it can
not say much. I think all 4 authours are great and any one of them gaining the
first place is ok.

As you know, Chinese chess (Xiangqui) has been being played mainly by Asian
people and in Asia, specially by Chinese and in China (with roughly 1 million
players), then by neighbours such as Vietnam (my country). Even CC is closed to
(Western) Chess around 70% (in terms of games and computer chess), other
difference of 30% requires deep knowledge of CC and it is hard for non-asian nor
other continent's people to learn it. The main and permanent reason is so lack
of CC's books and documents in English (on the contrary they are redundant in
Chinese and Vietnamese).

As a result, Asian people seem to have some advantages to develop CC software.
Taiwan is a high technique country, its programmers have been developed many CC
software for very long time (I still keep some their PC software developed in
1980s) and some are usually in the top of the CC world. Thus, they are totally
correspondent to the top ranks (they gained 1rd and 2nd).

Pascal Tang, a Vietnamese origin, surely understands much about CC. He wrote a
famous CC program - XieXie, as saying in his About dialog, to memory his father
and grandfather who have been great CC players. Pascal Tang is also co-authors
of a chess program (I forgot its name?), which has just winned France this year.
His program is also very correspondent to 3th rank or higher.

I don't know much about Marsland, just some his papers. His work is very useful
because CC lacks much basic researches. I think he has much knowledge about
computer chess, sciential methodologies. I guess (I have not any game of his
program) that his program should be very strong in general but may be not good
enough at precise knowledge of CC - thus his advantage of computer chess's
techniques may be not enough to convert to any win. However, I believe he will
learn much from losing. Probably, the results will be changed next year :)
PHN




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.