Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Chess Tiger engine. Settings

Author: Wayne Lowrance

Date: 22:32:04 03/17/00

Go up one level in this thread


On March 18, 2000 at 00:30:30, Christophe Theron wrote:

>On March 17, 2000 at 16:23:30, Wayne Lowrance wrote:
>
>>On March 17, 2000 at 13:33:29, Christophe Theron wrote:
>>
>>>On March 17, 2000 at 08:14:38, Wayne Lowrance wrote:
>>>
>>>>On March 16, 2000 at 16:40:47, Christophe Theron wrote:
>>>>
>>>>>On March 16, 2000 at 10:45:54, marcelo alvarez wrote:
>>>>>
>>>>>>Yesterday, I have purchased Chess Tiger engine and CP 4.3. I want to know the
>>>>>>strongest settings for this engine at short time controls (for example G/5) and
>>>>>>rapid chess (for example G/15 and G/30)versus others programs using auto232.
>>>>>>Thanks. Marcelo Alvarez.
>>>>>
>>>>>
>>>>>What is the speed of your computer? Everything depends on it...
>>>>>
>>>>>If you have a fast computer (200MHz or more), I would simply suggest at short
>>>>>time controls (G/5 or faster):
>>>>>1) do NOT give it too much hash tables. 4 Mb will be enough. If you give more
>>>>>than that, the time spent in clearing or updating the hash tables at each move
>>>>>will weaken the program
>>>>>2) close all unnecessary windows (for example the book window) and toolbars.
>>>>>3) do not move the mouse when the program is thinking. On some computers, this
>>>>>slows down the program tremendously.
>>>>>
>>>>>You can give the program more hash tables when you play at G/15 or slower.
>>>>>
>>>>>Then go to the Rebel Home Page (www.rebel.nl) and look for the new, more
>>>>>effective, engine settings (NSEW=1 and MoreSel=1).
>>>>>
>>>>>Also it seems that changing the "contempt factor" can help Tiger a little bit.
>>>>>In the engine properties page, set the "score for a draw" to the value "-10"
>>>>>(that means a tenth of a pawn).
>>>>>
>>>>>Last thing: make sure that the "analysis brain" is OFF when you play a game.
>>>>>
>>>>>
>>>>>
>>>>>    Christophe
>>>>
>>>>One last input from you Christophe. If entries were entered by hand with game in
>>>>5 and game in 10 (no increments) what time control setting would you use for
>>>>those two examples. 400 mhz pent.? Thank you.
>>>
>>>
>>>My own experience with manual entry is limited to the text mode version of Chess
>>>Tiger, that I have operated for years against various opponents (mainly humans
>>>at my chess club).
>>>
>>>If I am in good shape, I can enter (type) all the moves of a game in 4 minutes
>>>(assuming a rather long game, you never know). But this takes into account
>>>physically moving the piece on the board and tapping the clock too.
>>>
>>>You'll be using the mouse, and will not have to think about moving the pieces or
>>>tapping the clock, so I guess you can enter all the moves in 2 minutes?
>>>
>>>So for game in 5, set it to game in 7. For game in 10, of course it means game
>>>in 10. But you have to be fast!
>>>
>>>
>>>
>>>    Christophe
>>
>>Christophe Thank you. But I don't understand. If I set tiger to game in 7 on a
>>5/0 time control game, Tiger is gonna take too long and loose on time, everytime
>>I think. Can you explain ?
>
>
>OK, apparently I did not understand the question. So it is the opposite of what
>I have described. Set it to game in 3 if you are manually operating a game in 5.
>
>
>
>>I can work out the time allotment. I think if you can reassure me that Tiger
>>plays at full strength in any of the allowed time control committments That
>>would be fine. that probobly was a better question to start with. :)
>
>
>This time I'd better simply admit that I don't understand your question.
>
>If you are asking if Tiger plays "at full strength" in short time controls, I
>would answer: yes. Tiger is not optimized for a particular time control. You can
>use it at any time control. Well, maybe it's not really true for time controls
>under 1 minute per game, because the graphical interface is maybe going to eat
>too much processor time, but for time controls longer than 1mn/game, go ahead,
>it's OK.
>
>In short: I accept the challenge against any chess program at any time control
>(provided equal hardware and equal time controls are given to both).
>
>
>
>> I only ask
>>cause it seems, and I cannot prove it or describe the conditions as yet but
>>Tiger  has given me cause for concern when honking around in the infinate time
>>control period after having set up a pre arranged position. What i have seen is
>>that on move _n_ Tiger has rated the game as _v_ value points. and after the
>>move the value of _v_ has on occasion been rather different, always on the worse
>>side.
>
>
>Maybe after you make the move it can see a deeper combination and realize it is
>not as good as it thought before. But this is a normal behaviour of any chess
>program, just forget about it...
>
>No the value changed immediately. I know about search depth and changes in eval. I dont  want to leave the impression to the reading forum that I have discovered a bug, I have not, that I can verify !
>
>    Christophe

Thanks, you have answered my question about playing strength. no further thread
required.





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.