Author: Mogens Larsen
Date: 04:21:20 10/23/00
Go up one level in this thread
On October 23, 2000 at 07:04:19, Uri Blass wrote: >I do not say that it is easy to tune the evaluation function. >I do not say that it is only changing numbers in the evaluations but I believe >that it is only changing the evaluation. Yes, that is correct. Simple ideas can be complicated to implement, but they're still simple. >I agree that Gambittiger is better. >I did not say that gambittiger is not a good program. >I did not say that it is not the best. Neither have I. The only objection is the ludicrous 'new paradigm' boloney that have no basis in the real world. That and the flawed comparison with CW's attempts with CS-Tal. The sudden implementation of planning is an especially entertaining mirage. Gambit Tiger appears to be the strongest program at the moment. No problem for me at all as long as the religious rhetoric is kept at a minimum. Fortunately very few people believe a word TC utters and trusts the results instead. Good idea IMHO. >I only did not call it a new paradigm because the idea of big positional score >is not new and gambittiger only generalized it for more cases. Correct again. It's just calculation with changed parameters. Might very well be the correct way to go, but 'new'? No. Mogens.
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.