Author: Dann Corbit
Date: 15:50:33 04/05/99
Go up one level in this thread
On April 04, 1999 at 14:25:16, Lawrence S. Tamarkin wrote: >I don't know if this is important at all, especially when you are a computer >chess software junkie like myself! - All the chess programs crush me like a bug >anyway (on mere 200mHz machines, with 24-32 meg ram And even though I'm USCF >Life Master, my USCF is 2049 for those with morbid curiousity:)). Even Rebel >1.2 or Fritz2 are too strong for me at 5 minute chess, which is the only level I >play as I'm too impatient. And I'm sure even a 'Master' level program will beat >me silly too:) - No, mostly I use the 'serious' programs like Genius6 (5), >Rebel10, and especialy Fritz5.32 (Junior5), for analysis and study. CSTal >interest me, for it supposedly, radically different implemantation of its >algorythms (Programmer's - Is this true, or just hype?). It is a slow searcher like Diep and Hiarcs. It implements chess knowledge rather than super fast 'less thought consuming' sort of searching. Whether it is better or not depends on what you want. CS-Tal is famous for being a pirates-pouring-over-the-wall, all-out attack monster. I think you will see by examining the games it plays that it is not as liable to boring chess as some others. Sometimes the style it plays costs it in comp verses comp matches. For example, what percentage of sacrifices (not just deep combinations) are truly sound? Nobody knows, otherwise they would not be sacrifices. So programs that try to play positional chess take significant risks. >And this new feature >of being chess server compatible, I find facinating. I'd like to know a lot >more about its tutorials (I'm into that kind of stuff), but so far as I know, no >one has elaborated on this, other than a blurb on the Oxford Works page about >'Quick ELO grading - improve and grade your play using over 6000 solved chess >puzzles.' What do they mean & how does this work, and is it included in the >Windows package? I suspect that this is nothing but a big pile of solved, publicly know EPD test suites. Beware, as they are buggy. However, by throwing out the worst 1500 offenders, he may be providing a very solid test suite. So, it is probably about the right size (I have about 7700 chess problems and about 1500 have a book supplied answer that is just plain wrong). I would be very interested in a thorough, formal review of this product.
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.