Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: WAC 230 pounded on a million NPS machine...

Author: Robert Hyatt

Date: 14:55:02 04/13/01

Go up one level in this thread


On April 13, 2001 at 15:25:02, Will Singleton wrote:

>[ side issue ]
>
>
>               13->   4.58  -1.69   1. cxb4 a4 2. b5+ Kxb5 3. Rb1 c3 4.
>                                    Rc1 Kc4 5. Ba3 Kxd4 6. Kf4 Kd3 7. Rd1+
>                                    Kc4 <HT>
>               14     4.91     ++   1. cxb4!!
>               14->  19.08  -1.30   1. cxb4 a4 2. b5+ Kxb5 3. Rb1 c3 4.
>                                    Rc1 Kc4 5. Ba3 Kxd4 6. Kf4 Kd3 7. Rd1+
>                                    Kc4 <HT>
>
>This is off the topic, but of what value is the PV in ply 14, since it offers no
>enlightenment beyond ply 13?  It seems that the problem is the potential memory
>required to store the PV in the hash, hence the <HT>.  Clearly the real PV is
>different.  I wonder if there's a better way to show the difference.
>
>Will

In this case, it was a fail high at the root, but then the move failed low
due to a hashing incident...  The previous (13 ply) pv is saved in these
cases to give a better chance at move ordering at depth=15.






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.