Russian computer chess games?

[Arthur S Loh]

I have heard of many successful computer chess games from the US. Are
there any Russian computer chess programs? I know back in the 70s they had
a few good chess programs, but lately I haven't heard of anything. Please
fill me in on the details.

Thanks,
Art
al...@en.ecn.purdue.edu

[M Valvo]

Yes, there are Russian Chess Programs. Jonathon went to Russia a couple of
years ago to witness the then current state of their machines. The last
World Championship in Edmonton had a Russian entry. Judging from the machine's
abilities, they are way behind: the program was having trouble making legal
moves.

On the Chess Data Base side of things, there is a new Russian entry from
KarpovSoft called Chess Assistant. Its distinctive feature is its ability
to find all games containing any position you set up.

--Mike Valvo

[haf...@ifi.unizh.ch]

In article <12...@granjon.UUCP> m...@granjon.UUCP (M Valvo) writes:
>Yes, there are Russian Chess Programs. Jonathon went to Russia a couple of
>years ago to witness the then current state of their machines. The last
>World Championship in Edmonton had a Russian entry. Judging from the machine's
>abilities, they are way behind: the program was having trouble making legal
>moves.

Former world champion M. Botwinnik is working on a chess program called
"Pioneer". He is doing this since the seventies. His goal is to directly
simulate the reasoning of a chess master. He rejects the brute force approach.
The program tries to construct the gametree a master builds in his head, which
is very small but can be rather deep. So far "Pioneer" is not able to play
a whole game, is full of bugs and runs on a slow russian computer. The potential
of the approach is not obvious at the moment. The last thing I heard is that
Botwinnik tries to get a IBM-PC 386 to speed up execution of "Pioneer" and de-
bugging! Does anyone know more?
M. Hafner

[Jonathan Schaeffer]

In article <1990Jul27.1...@ifi.unizh.ch>, haf...@ifi.unizh.ch writes:
>
> Former world champion M. Botwinnik is working on a chess program called
> "Pioneer". He is doing this since the seventies. His goal is to directly
> simulate the reasoning of a chess master. He rejects the brute force approach.
> The program tries to construct the gametree a master builds in his head, which
> is very small but can be rather deep. So far "Pioneer" is not able to play
> a whole game, is full of bugs and runs on a slow russian computer. The potential
> of the approach is not obvious at the moment. The last thing I heard is that
> Botwinnik tries to get a IBM-PC 386 to speed up execution of "Pioneer" and de-
> bugging! Does anyone know more?
> M. Hafner

Pioneer does not really exist as a complete program. Yes, it has been
around for years and Botvinnik has published a great deal about it.
Unfortunately, his ideas look nice on paper but lack the concreteness
necessary for an implementation. In Moscow in 1988, I spoke with some
of the people who knew of the program (including one person who had
worked on the program). Their comments were that the project was:
- too vague; Botvinnik described things in human terms and left the
programmers the responsibility of trying to find ways to solve
this on a machine
- not enough machine time; they had access to a PC duing non-prime-time
hours. This clearly is not enough!
- program size; appartently the program is huge and is full of bugs.
Extensive effort, however, has been devoted to "tuning" the program
to solve a couple of famous positions from Botvinnik's games.
Of course, these positions appear in all Botvinnik's articles
on Pioneer (e.g. the famous Botvinnik-Capablanca game).

One of the things that has bothered me personally about Pioneer is
Botvinnik's grandiose claims. For example, I have a Botvinnik quote
from 1988 in which he states that if he only had a MC68000 chip,
he would be World Computer Chess Champion. I guess Botvinnik knows
something that the rest of the world does not.

For the 1989 World Computer Chess Championship in Edmonton, Tony Marsland
and I extended an offer to Botvinnik to try and entice him to compete.
We offerd him a 10 mip machine (a generous upgrade from the 68000 chip
he wanted) and access to a programmer analyst here to help port his code
to our machine. Unfortunately, Dr. Botvinnik was not amused at our offer
and it was rejected out of hand!?

[Tony Marsland]

In article <1990Jul23.1...@ecn.purdue.edu>, al...@en.ecn.purdue.edu (Arthur S Loh) writes:
> I have heard of many successful computer chess games from the US. Are
> there any Russian computer chess programs? I know back in the 70s they had
> a few good chess programs, but lately I haven't heard of anything. Please
> fill me in on the details.
>

There is again active interest in computer chess in the Soviet Union.
In recent years a few national micro-computer chess championships have
been held. The most popular systems are based on MS-DOS (on IBM clones).
Last year Victor Vikhrev with his program Centuar competed in the 6th
World Computer Chess Championships (without success), see Sept. 1989
issue of the ICCA Journal or the upcoming book "Computers, Chess, and
Cognition", published by Springer-Verlag Fall 1990.
I believe there is also a "joint-venture" program (possibly with the
Mikro-Kontour cooperative, through whom Gary Kasparov provided school
children access to 100 (?) home computers) being developed by Michail
Donskoy (of Kaissa fame).

[jupak...@cc.helsinki.fi]

In article <12...@granjon.UUCP>, m...@granjon.UUCP (M Valvo) writes:
>
> On the Chess Data Base side of things, there is a new Russian entry from
> KarpovSoft called Chess Assistant. Its distinctive feature is its ability
> to find all games containing any position you set up.
>
> --Mike Valvo

Not so distinctive, most of the Chess database programs have that
feature, Nicbase or Stavenbase example.

Jukka