Author: Robert Hyatt
Date: 20:21:56 03/31/00
Go up one level in this thread
On March 31, 2000 at 19:21:12, blass uri wrote: >On March 31, 2000 at 15:40:29, Dann Corbit wrote: > >>On March 31, 2000 at 15:24:30, Stephen Ham wrote: >> >>>Dear Dan and Uri, >>> >>>Thanks for the valuable input. This is good iformation and I appreciate your >>>kind help. >>> >>>I guess, in my ignorance, I didn't believe there'd be much difference between 16 >>>ply and 17 ply. But since you gentlemen think that the difference is >>>substantial, then we'll continue without having the chess engines reveal their >>>2nd and 3rd choices. >> >>There is (according to two studies) about a 15% chance that a computer program >>will make a better choice by going one ply deeper. It is possible that by the >>time we are out to 16 plies, the percentage is less than that, but it would >>certainly be tangible. > >There is about 15% chance that it is going to do a different choice and not a >better choice. > >The different choice is more times a better choice than a worse choice but >is not always a better choice. > >Uri It is _always_ a better choice according to the program's search code and the evaluation it uses. Of course, sometimes the search misses a tactic, or the evaluation misses a positional nuance... but in general, deeper is better, always...
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.