Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rebel 10's tactical blindness....

Author: Mark Young

Date: 00:26:59 05/24/99

Go up one level in this thread


On May 24, 1999 at 03:05:12, Terry Ripple wrote:

>On May 22, 1999 at 20:59:35, Mark Young wrote:
>
>>In the game played today with GM Rohde's we have another example of Rebel 10's
>>tactical blindness with its 26th move f4. This is just the latest example of
>>Rebel 10's tendancy to miss tactics that other programs understand with ease. I
>>understand that all programs miss some tactics from time to time, but Rebel 10's
>>frequency of tactical blindness is perplexing to me. It misses stuff that other
>>programs see in seconds, even after playing a blunder its evaluation my not drop
>>for many moves before understanding the danger.
>>
>>In the latest example I gave the position to Rebel 9, it had no problem seeing
>>that both 26. f4 and Bxb7 lose quickly and playing the correct 26. Bg2 as other
>>programs play.
>>
>>On the other hand even after the move 26 f4, Rebel 10 still think it has an
>>advantage for almost 4 mins, before seeing the error of 26. f4.
>>
>>Is this tactical blindness a bug in Rebel 10, or a justifiable design choice to
>>let rebel strengthen other areas of its play?
>
>
> Well,CM6000 likes the same move and believes its an even game! But allow Chess
>Master to continue its search for 7 min.:17 sec.and it then shows a -4.57 score
>for the move f4! It only changes its move to Bg2 after thinking for 16 min.:18
>seconds! This test was on a 266 mhz.,AMD K6 with 64 mb.RAM. Hashtable at 16MB.

This is still much better then Rebel 10...even after the move f4 it still thinks
it has the advantage for almost 4 mins.



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.