Author: Robert Hyatt
Date: 08:38:23 09/08/04
Go up one level in this thread
On September 07, 2004 at 22:26:26, Jeremiah Penery wrote:
>On September 07, 2004 at 19:32:43, Robert Hyatt wrote:
>
>>On September 07, 2004 at 16:46:03, Albert Silver wrote:
>>
>>>Using the new switch [Time Control = 125] and [Time Control Status = ALWAYS] to
>>>help Pro Deo use its time better, Pro Deo scored 20.5 - 19.5 in the Nunn 2 set
>>>against Ruffian 1.05. It had previously lost, though I must admit I don't
>>>remember the score.
>>>
>>>I then tested my new settings and it too scored 20.5 - 19.5. It squandered a
>>>number of winning positions, which was probably due to the time control, but
>>>most annoying of all, was to see it lose a very basic opposite colored bishop
>>>ending despite being a pawn up! This was in the last game. It played a terrible
>>>blunder and managed to lose. I checked to see whether this was due to my own
>>>settings, but Pro Deo default also plays this move and needs 3 minutes 34
>>>seconds to find the right move. Still, to be fair, even Junior 8 takes over 2
>>>minutes to see this, whereas Shredder 7.04 takes a mere 6 seconds. It's one of
>>>those issues where 2 distant passed pawns overcome the bishop.
>>>
>>>How does your engine do? Make sure it thinks long enough to consider the
>>>maneuver Kf5 bringing around the king.
>>
>>I let Crafty run for 10+ minutes. It liked Bc2 with a score of 0.0 instantly,
>>and never wavered from that move through 10+ minutes on my dual 2.8ghz xeon...
>
>If you tell Crafty to search Kf5, does it also see a drawscore?
Actually it does. In fact, it sees an EGTB hit at about depth-15 or so that
says this is a forced draw. Perhaps the losing move occurred a few moves
further in the game.
19-> 1:00 -0.01 1. ... Kf5 2. Kxa4 Ke4 3. Bd2 d4 4.
Kb4 c3 5. Bxc3 dxc3 6. Kxc3 <EGTB>
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.