Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: I still don't get it: time increment, why?

Author: Robert Hyatt

Date: 11:33:57 01/14/04

Go up one level in this thread


On January 14, 2004 at 14:25:07, Jeroen van Dorp wrote:

>On January 14, 2004 at 12:59:37, Robert Hyatt wrote:
>
>
>>If you don't, and the two programs are nearly equal, the result is almost a
>>coin-toss, because when time is low, searches change.  adding an increment
>>says that you can _always_ search for that long, plus whatever fraction of
>>the original time block you want to invest...  you never have to play a move
>>in a fraction of a second where tactics might smoke you...
>
>I wonder.... with that, are you _explaining_ the definition of "best chess
>engine" or are you just _redefining_ it?
>
>
>J.


I'm not sure what you mean.  I've _always_ maintained that the best long-game
engine is not necessarily the best short-game engine.  It is _possible_ that
is true, but it is not guaranteed.  So taking a long game and reducing it to
milliseconds-per-move at the end can change the result and skew the overall
outcome.  Or just turn it into a coin-flip as to who makes the first mistake
and actually suffers for it.



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.