Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Tiger 14 cannot play endgames

Author: Eduard Nemeth

Date: 15:42:18 09/04/01

Go up one level in this thread


On September 04, 2001 at 18:20:14, Uri Blass wrote:

>On September 04, 2001 at 18:08:15, Eduard Nemeth wrote:
>
>>On September 04, 2001 at 15:39:29, Sune Larsson wrote:
>>
>>>>snip
>>>
>>>>what did you do to your tiger?
>>>>Are you sure that you use the default parameters?
>>>>I do not understand why it cannot go deep
>>>>
>>>>Here is tiger14 under the chessbase interface
>>>>Maybe it is a problem of hash tables(I used 48 Mbytes hash) but I cannot   believe
>>>>that hash tables have this big effect)
>>>
>>>
>>> No, it was not the hash tables. I had activated the feature
>>> "opponent is human". Then Tiger needed 7 sec to avoid 29.Be3?
>>> When I deactivated it I got about the same output as you.
>>> Then Tiger got deep much faster but stuck to 29.Be3? Fast but dumb
>>> in this position. You can try it out in Fritz.
>>>
>>> Sune
>>>
>>>
>>>>New game - Deep Fritz
>>>>[D]4k3/B1p4p/8/p2n1p2/5P2/5P2/1P2K2P/8 w - - 0 1
>>>>
>>>>Analysis by Chess Tiger 14.0:
>>>>
>>>>29.Be3 Nxe3 30.Kxe3 Kd7 31.Kd4 Kd6 32.Kc4 Kc6 33.Kb3 Kb5 34.Kc3 Kc6
>>>>  =  (0.00)   Depth: 12   00:00:00  146kN
>>>>29.Be3 Nxe3 30.Kxe3 c5 31.Kd3 Kd7 32.Kc4 Kc6 33.Kb3 Kb5 34.Kc3 h5 35.Kd3
>>>>  =  (-0.02)   Depth: 12   00:00:00  164kN
>>>>29.Be3 Nxe3 30.Kxe3 c5 31.Kd3 Kd7 32.Kc4 Kc6 33.b3 Kb6 34.h4 h5 35.Kd3 Kc6
>>>>  =  (-0.12)   Depth: 13   00:00:00  299kN
>>>>29.Be3 Nxe3 30.Kxe3 c5 31.Kd3 Kd7 32.Kc4 Kc6 33.h4 a4 34.h5 h6 35.Kd3 Kd6 36.Ke3
>>>>  =  (-0.14)   Depth: 14   00:00:01  591kN
>>>>29.Be3 Nxe3 30.Kxe3 c5 31.Kd3 Kd7 32.Kc4 Kc6 33.b3 Kb6 34.h4 Kc6 35.h5 h6 36.Kc3
>>>>Kd5
>>>>  =  (-0.12)   Depth: 15   00:00:03  1246kN
>>>>29.Be3 Nxe3 30.Kxe3 c5 31.Kd3 Kd7 32.Kc4 Kc6 33.h4 a4 34.h5 h6 35.Kd3 Kd5 36.Kc3
>>>>Ke6 37.Kc4
>>>>  =  (-0.14)   Depth: 16   00:00:06  2125kN
>>>>29.Be3 Nxe3 30.Kxe3 c5 31.Kd3 Kd7 32.Kc4 Kc6 33.h4 a4 34.h5 h6 35.Kd3 Kd5 36.Kc3
>>>>Ke6 37.Kc4 Kd6
>>>>  =  (-0.14)   Depth: 17   00:00:13  4407kN
>>>>29.Be3 Nxe3 30.Kxe3 c5 31.Kd3 Kd7 32.Kc4 Kc6 33.b3 Kb6 34.h3 Kc6 35.h4 Kb6
>>>>36.Kd5 h5 37.Kc4 Kc6 38.Kc3
>>>>  =  (-0.14)   Depth: 18   00:00:25  8223kN
>>>>
>>>>(Blass, Tel-aviv 04.09.2001)
>>>>
>>>>You can see that my tiger goes deeper in less nodes.
>>>>
>>>>Uri
>>
>>Yes I played NOT with option Anti-Human.
>>I ask me, which is the "true" Tiger?
>>When the true Tiger is with option Anti-Human, then i my opinion it is not
>>right to say: Anti-Human! Right is "Anti-Computer"!
>>
>>(german: Ich frage mich welcher nun der wahre Tiger ist, der mit der Option
>>Anti-Human oder der ohne diese Option? Wenn insgesamt jener Tiger mit der Option
>>Anti-Human besser spielt, dann sollte man diese ANTI-Option anders definieren,
>>nämlich auf "Anti-Computer"! Das würde aber weiter heissen dass der wahre Tiger
>>nur gegen den Menschen auftritt, während gegen Computer eine auf Computer
>>"GETUNTE" (!) Version antritt und ich würde das wenn's so ist (?) "Schummeln"
>>nennen! Dennoch vermag diese Anti-Human option den schwarzen Sieg nach Be3 nicht
>>zu sehen.)
>>
>>Nevertheless, Tiger 14 with option "Anti-Numan" cannot seen the win for black
>>against Be3!
>>
>>Chess Tiger 14.0 - Nemeth,E
>>
>>[D]4k3/2p4p/8/p2n1p2/5P2/4BP2/1P2K2P/8 b - - 0 1
>>
>>Analysis by Chess Tiger 14.0 (anti-human):
>>
>>29...Kd7 30.Ld2 a4 31.Kd3 c5 32.h4 h6 33.Kc4 Kc6 34.Lc1 Sb6+ 35.Kc3 Sd5+ 36.Kc4
>>  =  (0.00)   Tiefe: 16   00:00:16  4116kN
>>29...Kd7 30.Ld2 a4 31.Kd3 c5 32.h4 h6 33.Kc4 Kc6 34.Lc1 Sb6+ 35.Kc3 Sd5+ 36.Kc4
>>  =  (0.00)   Tiefe: 17   00:00:33  8330kN, tb=1
>>29...Kd7 30.Ld2 a4 31.Kd3 c5 32.h4 h6 33.Kc4 Kc6 34.Lc1 Sb6+ 35.Kc3 Sd5+ 36.Kc4
>>  =  (0.00)   Tiefe: 18   00:01:12  17827kN, tb=7
>>29...Kd7 30.Ld2 a4 31.Kd3 c5 32.h4 h6 33.Kc4 Kc6 34.Lc1 Sb6+ 35.Kc3 Sd5+ 36.Kc4
>>  =  (0.00)   Tiefe: 19   00:02:41  39300kN, tb=74
>>
>>Nxe3 is the win move!
>>
>>Eduard
>
>Tiger with anti-human simply dislikes pawn endgames
>I guess that it assumes that the opponent knows more about pawn endgames so it
>adds a small negative number for pawn endgames that is enough to avoid going
>into a pawn endgame in these cases.
>
>Uri

I understand now, thanks!
Ok, in future I will play openings where I can have a endgame!
In endgames + anti-human,Tiger is also not strong! ;-)

But, against other computers Tiger cannot play with this option, yes? :)

Eduard






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.