17th WCCC and the 14th Computer Olympiad

[swami]

in 1993 e.g. in munich pandix did a very good job:
http://www.grappa.univ-lille3.fr/icga/t ... .php?id=57

7th behind quest (fritz).

Excellent score from the program that's unknown to many of us. Even in that tourney ages ago, its performance was about 2300-2400. I'd hope that it would get released after this championship!

[mclane]

its not an UNKNOWN program. it also made a title (maybe amateur-title or stuff like this) but that was long time ago. long before the idea of UCI was born. therefore it is maybe not remembered in OUR days.
as far as i remember it was quite good in positional play, played often closed and passive variations. and was not so good in tactical play.

but ... that was long time ago. maybe today it has all been changed.
i remember a year ago or so it also played good in the hungarian championships.

the pandix family was also at the AEGON tournament in the hague.

Its IMO an interesting program and i would like to buy/test it.

therefore i am looking forward for the games.

I guess trapped pieces code is the next thing the programmer will implement

Description given in 1995:
Gyula Horvath started writing chess programs in 1985. His program won the Amateur World Chess Championship in 1987 and the Personal Computer Chess Champion title in 1988 and 1989. His wife, Szuzsa, joined the development in 1986. She is mainly active in testing the program and in designing and programming the graphics of the commercial versions of the program. Both of them pursue chess programming as a hobby - Gyula works as a marketing researcher and Zsuzsa works as a telemarketing assistant. They have participated in various computer chess events since 1986. In 1993 their team increased to three members when Csaba Szues began to implement a new 100,000 moves opening book. The program is written in C and uses a 400KB hash table. It measures the move interestingness and incrementally updates the attack map. The program uses principal variation search, advanced time control and special limited quiescence search.
Description given in 1988:
Besides the WORLD PC Champion title, Pandix won the Amateur Micro World Championtitle as well.

source:
http://www.grappa.univ-lille3.fr/icga/p ... php?id=190

the program is as far as i remember in the CXG Legend dedicated chess computer and maybe in few others too.

[Mike S.]

Against Junior, it looks like Pandix' book ended after 3...cxd4. At least, the rest of the opening looks very strange...

After 9 moves (wtm.):

[d]rnbqkbnr/pp1p1ppp/4p3/1N6/1P6/4B3/P1PN1PPP/R2QKB1R w KQkq - 0 10

In Joker-Rybka, Joker played a surprising pawn sac (I guess it was calculated, not from book?).

[d]r2qkb1r/p2bpppp/2pp1n2/8/3NP3/8/PPP2PPP/RNBQ1RK1 w kq - 0 8

8.e5

I thought it wasn't bad, or at least not very bad, and interesting. White gains the square c5 for a knight (Nd4-b3-c5). Rybka won, but the game lasted until the 47th move.

[mclane]

hi - thank you for the information.

i do nevertheless hope that we one day can buy or test pandix again,
as it was in those old days with "chess friend" alias pandix.

[mclane]

Against Junior, it looks like Pandix' book ended after 3...cxd4. At least, the rest of the opening looks very strange...

it seems pandix has no malus for moving the queen too often in the opening.

[Dr.Wael Deeb]

As for Pandix:

1. It is a private engine and I don't think it will ever be commercial (even if it was in the past).

2. I didn't see the video but if the man in red-white-green T-shirt has a thick dark moustache it is certainly Gyula. Please note the spelling. For English speakers it is pronunciated somehow like dew-lah, for German speakers it is Djula.

3. It seems to me Pandix wastes many moves in the opening phase for queen moves, at the cost of development. AFAIK it is a rather raw, experimental version and I wouldn't be surprised if Gyula changed it during the tournament.

You got that right Gabor....it's playing like a weak club player developing it's queen so early and ignoring the development of the minor pieces....
Dr.D

[mclane]

pandix is maybe a new rewrite. and new programs often lack many things. but computerchess is about DEVELOPMENT. and without new beginning you cannot make much progress IMO.

[Dr.Wael Deeb]

pandix is maybe a new rewrite. and new programs often lack many things. but computerchess is about DEVELOPMENT. and without new beginning you cannot make much progress IMO.

Yes,I agree....but still....it's playing weak
Dr.D

[Mike S.]

I assume HG generously offered a draw before Pandix exceeded time, or?

I don't think such a "gift of 0.5 points" would be appropriate as it may affect the ranking of others. Maybe Pandix didn't stop but just couldn't find a win. I can imagine that the delay was due to an arbiter involvement, but I'm not sure and we need to wait for the official result on the ICGA page, which is not yet up at the moment.

[mclane]

yes i saw the game. great. he will find the time to fix the bugs. these tournament are made for finding bugs like this

golden rule is:
if there is a bug in a program it will appear in an important game in championship. at home on autoplayer it never appears. but in an important tournament or championship it WILL appear.

we had many cases in history of computerchess where this golden rule
was true

for the programmer this is often sad thing to stand. but on the other side
he will make progress AFTER the tournament.