Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Computers-underpromotion

Author: blass uri

Date: 08:06:16 04/16/00

Go up one level in this thread


On April 16, 2000 at 09:53:15, Robert Hyatt wrote:

>On April 16, 2000 at 09:25:33, Daniel Clausen wrote:
>
>>Hi
>>
>>On April 16, 2000 at 09:11:08, Charlie GOLD wrote:
>>>     Crafty analysed the position and showed # in one (Pf8-N #), BUT IT DID NOT
>>>MAKE THE MOVE UNTIL I FORCED IT TO. The other machines made the move in less
>>>than 5 seconds.   charlie
>>
>>Some engine continue with the search when they found a mate-in-N on a
>>certain ply in the hope to find a shorter mate. Of course in the case
>>of a mate-in-1 this is a bit silly. :) Maybe you saw this effect in
>>Crafty?
>>
>>Kind regards,
>> -sargon
>
>
>Crafty stops at the right point, _unless_ given a sd=N command.  It uses the
>rule of "search to 2*N=1" when it finds a mate in N, just to be sure there
>isn't a shorter mate.    In the position you gave, it stopped after 4 plies
>as that is the minimum depth it will normally search to, regardless of the
>mate found...
>
>I don't see how it could continue to search unless it is in infinite mode or
>a specific depth has been given to it...

I know that under chessbase crafty always use time to play even when it can see
mate in 1 and the only cases that it does not use time are because of the
permanent brain.

It does not change the result of the game but I saw crafty using 246 seconds to
do mate in 1 against Fritz6 in the ssdf games when it was under chessbase
interface.
It has nothing to do with underpromotions.

Under infinite time control it will never play because it "believes" that it has
to use time.

Uri



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.