Chess System Tal II & Engine Style

[cornetmike]

I own ChessGenius 7.1, Fritz 8,9, and CSTal II. ChessGenius and Fritz are node-crunching monsters with no discernable style. When I want to play an interesting game of chess, I choose CSTal II. It does seem to play a different kind of game! I might suggest it has a "style" quite possibly human-like. And with a 2300 rating, it's quite hard to beat, at least for me.

Surfing the Web the other day, I discovered CSTal is again available commercialy, for nine dollars U.S. E-mail me privately and I'll send you the URL.

Elsewhere on this list there is a topic "tough position for portables." It's a position from a game Alekhine vs. Teichmann. CSTal found the solution, Rd5, in three seconds.

regards, Mikey

[Tony Thomas]

Why dont you just post the link? Something is shady I say.

[cornetmike]

I didn't want to be accused of being a spammer! I was just being careful.

Here it is: http://www.wholesalechess.com

CSTal II is being sold under the name of "Championship Chess" for nine dollars, plus shipping.

regards, Mikey

p.s. I neither work for nor do I recieve kickbacks from Wholesalechess!

[mclane]

i am still waiting that somebody continues the program.
version 2.5 was close to finish release of version 3.

[GenoM]

<...>CSTal II. It does seem to play a different kind of game! I might suggest it has a "style" quite possibly human-like.<...>
regards, Mikey

Hi Mikey
My observations at the style Chess System Tal playing are the same as yours -- very human-like and different from WB and UCI chess engines.
Its interresting -- can someone of programmers explain this phenomena? Something in programming? What is the difference?

[Jim Walker]

I also like the CSTal II program. It's a shame Chris stopped working on chess programs. CSTal seems to be on a mission of attacking the opponents King. It was supposed to be "smart" in that respect but it seems all of it's knowledge made it too slow. It got outsearched by Fritz/Junior most of the time. I was getting 10K nps at the time it was new and Chris admitted it needed rewriting to allow it to search deeper. Now days I can get over 60K nps on my hardware but it's still lacking something.

[mclane]

i am not a programmer, but i was involved in testing CSTal for a few years.

CSTAL was designed to play unsound incorrect moves.
the idea we had was: the other chess programs are good in finding BEST moves. CSTAL has a very BIG evaluation function and trusts this evaluation more than its search.
the idea to cope with those search based programs was to bring them into FOG where they had to compute so many moves that they cannot work precisely anymore. they cannot compute accurate anymore. its foggy there, and you would need KNOWLEDGE to adjudicate the position.
CSTAL has this knowledge, so it has an advantage in the fog.

so the first thing CSTAL does when it sees fog is, driving directly into the fog.
it plays moves that are incorrect. it bluffs.

i remember only ONE program that came somehow (but different) close to this behaviour, this was Super-Constellation by Dave Kittinger.

it took long to make CSTAL play succesful with this unsound tal-function.

[cornetmike]

Warning! I just got "Championship Chess," (CSTal II), in the mail today from wholesalechess.com. The OPENING BOOK IS MISSING!! I just fired off a bunch of e-mails...probably to no effect...

However, at amazon.com, you should be able to find "Tournament Chess," or "10 Pro Board Games," new and used. These have CSTal II with the opening book.

I simply LOVE CSTal. I almost always get complicated, interesting games.

regards
Mikey

[Patatov]

Thanks for the info Mikey... But do you know if any of these softwares have version CSTal II 2.5 ? Or when you say CSTal II does that mean 2.5?

Thanks

[cornetmike]

I don't know. I assume it's the latest version. All I can find is "copyright 1999 by Purple Software." Purple was the original developer.

regards
Mikey