Author: Ed Schröder
Date: 01:04:28 05/12/98
Go up one level in this thread
>Posted by Jouni Uski on May 12, 1998 at 02:34:58:
>I have noticed, that Rebel shows some time maximum thinking depth of
>60 plys in normal middle game situation! It's very, very hard to
>believe - can this really be true, Ed? I want to see whole variation
>before believing...
Hi Jouni,
It happens, believe it or not :)
But it doesn't mean much as in 99.9% of these cases these variations
will never show up in the main variation. Still these deep lines are
searched mainly in cases where there are lots of checking moves.
A nice example and exception to the rule is the "mate in 30" position
you once uploaded yourself:
5n2/B3K3/2p2Np1/4k3/7P/3bN1P1/2Prn1P1/1q6 w - - id Mate in 30; bm e3g4;
pv 00:00 2.01 262144.00 Ne3-g4+ Ke5-f5 Ng4-h6+ Kf5-e5 Nh6-f7+ Ke5-f5;
Rebel finds the mate in 30 (which needs the full 60 plies) already on
the second ply. So sometimes it make sense to search that deep.
About the strange "262144.00" score, this is a display bug. It means
"mate in 30" but this bug never showed up because of the lack of a
good test position like the one above. This will be fixed for Rebel10
see the below PGN output.
- Ed -
[Event "?"]
[Site "?"]
[Date "??.??.????"]
[Round "?"]
[White "Mate in 30"]
[Black "Display bug"]
[Result "*"]
[WhiteElo "?"]
[BlackElo "?"]
[FEN "5n2/B3K3/2p2Np1/4k3/7P/3bN1P1/2Prn1P1/1q6 w - - 0 1"]
{ 00:00 2.01 Mate in 30 moves Ne3-g4+ } 1. Neg4 *
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.