Author: Bernhard Bauer
Date: 23:24:49 01/08/04
Go up one level in this thread
On January 08, 2004 at 15:08:28, Drexel,Michael wrote:
>List doesn't use tablebases in the search (Ruffian does).
>That might be the reason.
>
>Michael
I don't think the problem is related to tablebases.
It's a simple nullmove problem. May be the authors of ruffian and list
trusted too much in crafty.
On of the nice things with crafty is that you can change the source.
Here is what I get with crafty 19.09
clearing hash tables
time surplus 0.00 time limit 1:00 (1:00)
depth time score variation (1)
16 0.64 0.01 1. Bg6 Rxg6+ <EGTB>
16-> 0.64 0.01 1. Bg6 Rxg6+ <EGTB>
17 1.06 0.01 1. Bg6 Rxg6+ <EGTB>
17-> 1.06 0.01 1. Bg6 Rxg6+ <EGTB>
18 1.90 0.01 1. Bg6 Rxg6+ <EGTB>
18-> 1.90 0.01 1. Bg6 Rxg6+ <EGTB>
19 3.98 0.01 1. Bg6 Rxg6+ <EGTB>
19-> 3.98 0.01 1. Bg6 Rxg6+ <EGTB>
20 9.42 0.01 1. Bg6 Rxg6+ <EGTB>
20-> 9.42 0.01 1. Bg6 Rxg6+ <EGTB>
21 17.18 0.01 1. Bg6 Rxg6+ <EGTB>
21-> 17.18 0.01 1. Bg6 Rxg6+ <EGTB>
22 28.95 0.01 1. Bg6 Rxg6+ <EGTB>
22-> 28.95 0.01 1. Bg6 Rxg6+ <EGTB>
23 49.65 0.01 1. Bg6 Rxg6+ <EGTB>
23-> 49.65 0.01 1. Bg6 Rxg6+ <EGTB>
time=1:00 cpu=19% mat=0 n=10909374 fh=83% nps=181K
ext-> chk=504488 cap=3 pp=41706 1rep=6000 mate=1492
predicted=0 nodes=10909374 evals=4788146
endgame tablebase-> probes=107650 hits=107650
And with a modified search, no nullmove is done in case of sum_of_pieces/2 < 8
I get:
clearing hash tables
time surplus 0.00 time limit 1:00 (1:00)
depth time score variation (1)
10 0.39 -Mat20 1. Bg6 Ke4 2. Ne2 Rxg6+ <EGTB>
10 0.51 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
12 1.78 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
12-> 1.78 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
13 3.34 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
13-> 3.36 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
14 7.26 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
14-> 7.26 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
15 17.14 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
15-> 17.14 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
16 30.05 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
16-> 30.05 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
17 50.70 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
17-> 50.70 0.01 1. Ng6 Kd2 2. Kh4 Rxg6 <EGTB>
time=1:00 cpu=32% mat=0 n=22066897 fh=84% nps=367K
ext-> chk=940927 cap=28 pp=96132 1rep=6085 mate=692
predicted=0 nodes=22066897 evals=8177379
endgame tablebase-> probes=71028 hits=71028
Note that I get only 17 plys in a minute instead of 23 plys in the original
version. So there is a price to pay.
regards
Bernhard
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.