Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Heeeeeelllllloooooooo Johan de Koning!!!!!!!!!!

Author: Ed Schröder

Date: 06:01:06 08/16/03

Go up one level in this thread


On August 16, 2003 at 04:25:59, Johan de Koning wrote:

>On August 15, 2003 at 10:52:20, Ed Schröder wrote:
>
>>On August 15, 2003 at 02:46:32, Johan de Koning wrote:
>>
>>>On August 14, 2003 at 15:10:59, John Merlino wrote:
>>>
>>>>On August 14, 2003 at 14:42:46, Uri Blass wrote:
>>>>
>>>>>On August 14, 2003 at 14:34:06, Andreas St. wrote:
>>>>>
>>>>>>On August 14, 2003 at 14:20:20, Uri Blass wrote:
>>>>>>
>>>>>>>On August 14, 2003 at 13:51:20, John Merlino wrote:
>>>>>>>
>>>>>>>>On August 14, 2003 at 05:53:43, Andreas St. wrote:
>>>>>>>>
>>>>>>>>>Hi,
>>>>>>>>>
>>>>>>>>>is there any chance, that there will be out a patch for Chessmaster, that helps
>>>>>>>>>in time consuming???
>>>>>>>>>Chessmaster moves much too fast. Only with such patch chessmaster could play
>>>>>>>>>much stronger!!
>>>>>>>>>
>>>>>>>>>Heeeelllllloooo Johan :-(
>>>>>>>>>
>>>>>>>>>Greetings
>>>>>>>>>
>>>>>>>>>AS
>>>>>>>>
>>>>>>>>I can't speak for Johan, but I can GUESS that he would disagree with you when
>>>>>>>>you say that the engine "moves much too fast" and that it "could play much
>>>>>>>>stronger" if it took more time.
>>>>>>>>
>>>>>>>>Johan sticks by his clock management algorithm, and told me many times that he
>>>>>>>>feels that it is appropriate for The King.
>>>>>>>>
>>>>>>>>jm
>>>>>>>
>>>>>>>He may decide that the time management of today is the default option but give
>>>>>>>the engine also options to play slower in the first moves by defining a
>>>>>>>different personality(using time 100 is the default and using time when you
>>>>>>>believe that you have twice the time that you really have can have parameter of
>>>>>>>200).
>>>>>>>
>>>>>>>I do not think that chessmaster is going to lose on time with parameter 200
>>>>>>>because I doubt if there are cases when it is using more than half of the
>>>>>>>remaining time for a single move.
>>>>>>>
>>>>>>>In any case it is possible to decide about limit in the personality(never use
>>>>>>>more than x% of the remaining time for a move).
>>>>>>>
>>>>>>>Uri
>>>>>>
>>>>>>
>>>>>>Hi Uri
>>>>>>
>>>>>>of what time parameter are you talking about?
>>>>>>Where can i find it. Is it possible to set any time parameter by using the king
>>>>>>with UCI?
>>>>>>
>>>>>>Greetings
>>>>>>
>>>>>>AS
>>>>>
>>>>>I am not talking about existing parameter.
>>>>>
>>>>>When I said to define a different personality I meant also to add a new
>>>>>parameter for all the personalities of chessmaster that means to play slower or
>>>>>to play faster.
>>>>>
>>>>>Uri
>>>>
>>>>We did discuss adding such a "time management" personality parameter for CM9000.
>>>>We eventually decided against it for several reasons. At this point, if Johan
>>>>wants to clarify, he can do so.
>>>
>>>We actually decided pretty immediately against it. :-)
>>>
>>>Looking back we may have discarded it too quickly, because I do agree with
>>>others that the engine does play relatively fast early in the game.
>>>
>>>However, I do not agree with others that an agressive time schedule would make
>>>the engine play *much* stronger. Saved time will be fed back and will be very
>>>valuable when panic strikes or when the game goes beyond 40 moves. Used time on
>>>the other hand can never be restored.
>>>
>>>... Johan
>>
>>
>>Addup that playing faster then your computer opponent gives you an advantage in
>>the permanent brain as it will narrow the time of your computer opponent. I am
>>sure this is a consideration of yours to keep your fast time control as is.

>Well, now that you mention it ...
>I'll keep this in mind when (if) reredesigning the time manager.
>
>I am sure this is an attempt of yours to pull my leg.
>Well done. :-)

I did not realize it was that obvious :-)

But seriously, the technique has been practized by the Nimzo team at a
tournament against Rebel. I noticed the time control behaviour, when I asked why
Nimzo was playing so fast I got a muttering as response which made me even more
suspicious. Since I saw the advantages it made me think and afterwards I
implemented such an algorithm also in Rebel which I from that moment on used
against GM's, it works great. Against comps I got about equal results but I was
too lazy to test the algorithm extensive.

I haven't set it to default for reasons you experience now, the impossibility to
explain it to the average user :)

Ed


>... Johan



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.