Author: Harald Faber
Date: 03:13:13 04/21/99
Go up one level in this thread
On April 21, 1999 at 05:18:49, blass uri wrote: >>What is the best positional test for chess programs? I don't know. The problem is the best move. Who tells and confirms the best POSITIONAL move? I hear many guys crying "but there is (at least) another good move which may not be worse than your best move" A serious problem. >>What is more important Tactic Abilities or Positional Knowledgment ??? >> >>I think that Hiarcs7 is the most beautiful strong program in all the world. >> >>I think that Hiarcs7 in a faster machine could win an important top human >>tournament. >> >>What do you think? > >I know that hiarcs7 is a tactical monster. >It is good at fast time control. And slow time controls too! >It won Junior5 and Fritz5 at blitz in enrique's tournament. And it played and performed in his 40moves/40minutes on P2-400 very well too! >I am not impressed by the positional understanding of Hiarcs7. >It lost some games against Fritz5.32 because of bad positional understanding. Oh, in the meantime there are SOME? You always mentioned "but there is ONE game where Hiarcs lost because of positional lack of knowledge" >I hope that junior5 is stronger than Hiarcs7 Wishful thinking. :-) (Because you can't be objective, no problem) I admit sometimes I am impressed about the way Junior plays dynamically and starts to create chances and possibilites to get out of a bad position. At the moment I think Hiarcs7 is the best program concerning results against humans and programs. MCP and Rebel, maybe the King may be better against humans but they are weaker against programs. On the other hand there is Fritz and Nimzo who perform very well against other programs but results against humans are missing. And my subjective view says the style they play does not as good against humans as the above mentioned programs. Just my personal opinion, based on some games I saw. >The ssdf tells me that Junior5 lost against Hiarcs7 28.5:11.5 when all the top >programs did better result against it(for example Hiarcs6-Hiarcs7 on P2000MMX >20:20). >I need to see the games to be convinced that the tester really played the games >and did not invent the results. If they don't provide the games you have the choice to a) believe it b) don't believe it and c) test on your own. >Junior5 did clearly better results in public games. >Uri Which games? ICC with short time controls? How many? Which Junior-version? Your experimental one?
This page took 0.01 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.