Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Crafty not that strong (2)

Author: Robert Hyatt

Date: 18:24:42 04/24/00

Go up one level in this thread


On April 24, 2000 at 19:08:21, blass uri wrote:

>On April 24, 2000 at 16:56:05, Robert Hyatt wrote:
>
>>On April 24, 2000 at 16:50:01, Djordje Vidanovic wrote:
>>
>>>On April 24, 2000 at 16:06:55, Robert Hyatt wrote:
>>>
>>>>On April 24, 2000 at 15:01:22, Christophe Theron wrote:
>>>>>
>>>>>I'm not sure I am a respectable chess programmer, but I want to ask something:
>>>>>
>>>>>Why is Crafty's management of pondering supposed to be superior to Fritz'?
>>>>>
>>>>>Why is pondering=off supposed to handicap Crafty more than Fritz?
>>>>>
>>>>>Who can seriously believe that Frans Morsch is so lousy that he cannot take
>>>>>advantage of pondering as well as Bob does?
>>>>>
>>>>>
>>>>>    Christophe
>>>>>
>>>>Your question is _bass-ackward_ in it's phrasing.  It should be "who would not
>>>>believe that Frans has spent more time testing with ponder=off than bob has,
>>>>so that it isn't a surprise that fritz probably does a better job of allocating
>>>>time in that mode than crafty does?"
>>>>
>>>>_that_ is the point.  Not that I am better with ponder=on.  That I am _worse_
>>>>with ponder = off.  How hard is that to understand???
>>>>
>>>>
>>>
>>>
>>>What makes you think that Frans Morsch would spend his time on a futile thing
>>>such as testing his program with ponder=off?  You yourself stated that testing
>>>your program with ponder=off is a waste of time, so I simply can't see why Frans
>>>would care to waste _his_ time.  And, if you are 'not better with ponder=on, and
>>>are worse with ponder=off', what is the inference to be drawn?
>>
>>
>>
>>
>>trivial to answer, when you think about it.  Why would he spend any time on
>>_anything_ other than the 'engine' itself?  Perhaps because he is the author of
>>a commercial program?  Why have the commercial programmers spent so much time
>>tweaking and tuning for SSDF play?  To make their program look better than the
>>others, for a marketing edge?  If you knew that lots of people wre going to be
>>playing games on one computer, using your program, would _you_ spend time to
>>make it play as strongly as possible to keep that marketing edge?
>
>I do not think that Frans Morsch cares about engine-engine games otherwise he
>could prevent fritz to lose on time by telling it not to use big hash tables on
>time trouble.
>
>The fact that Frans Morsch did not do it convince me that he does not care much
>about blitz results that is the most common engine-engine games.
>


The same problem holds true in standard engine-engine games.  Not just blitz.




>The fact that many users do many engine-engine games on one computer does not
>say that they buy programs based on the results of engine-engine games.
>They do it usually because they like to watch comp-comp games and not to tell
>friends which program to buy.
>
>Uri



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.