Author: Robert Hyatt
Date: 18:34:47 11/08/03
Go up one level in this thread
On November 08, 2003 at 12:44:35, Uri Blass wrote:
>On November 08, 2003 at 12:32:10, Robert Hyatt wrote:
>
>>On November 07, 2003 at 01:43:54, Terry McCracken wrote:
>>
>>>hxg5 loses, and Hiarcs found out a few moves later.
>>>
>>>This is from a game I played against Hiarcs 9 on Sept. 6, 2003.
>>>
>>>[D]r1b2rk1/4bpp1/pqpppn1p/6B1/4P2P/1BN5/PPP1QPP1/2KR3R b - - 0 13
>>>
>>>
>>>Shredder 7 and 7.04 refuses the sac but doesn't really know the danger, as it
>>>thought after I made it play hxg5 it was a bit ahead, then a move or two later
>>>failed low.
>>>
>>>Terry
>>
>>
>>Try Crafty. It will _never_ consider that capture, even at a 1-ply
>>search. :)
>>
>>I tested it to be sure. NO hxg5 ever shows up.
>
>What is the first node that it searches if it is not hxg5?
>
>Most programs start searching winning captures if they have no move in the hash
>tables and most programs define hxg5 as a winning capture.
>
>Uri
Here is the answer. The following is my root move list, ordered by using
both SEE, _and_ the static evaluation...
In this case, crafty is not "most programs"...
36 moves at root
move score
Rb8 -38
Qc5 -53
a5 -58
Qb4 -61
d5 -62
Bb7 -63
Qb7 -64
Bd7 -65
Rd8 -71
Re8 -71
c5 -72
Ra7 -73
Qc7 -75
h5 -76
e5 -78
Kh8 -79
Qa5 -81
Qd8 -83
Kh7 -89
Qb8 -91
Qa7 -93
Bd8 -259
g6 -262
Nxe4 -413
hxg5 -453
Nd5 -506
Ng4 -532
Nd7 -547
Nh5 -552
Nh7 -552
Ne8 -572
Qxb3 -623
Qb5 -703
Qxf2 -1045
Qe3+ -1123
Qd4 -1126
As you can see, hxg5 is way down the list...
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.