Computer Chess Club Archives




Subject: Re: Shirov strong tactical test position

Author: Thorsten Czub

Date: 14:47:16 09/23/97

Go up one level in this thread

Shirov-test-position move 24.Bg5 !!:


[Event "Shirov test-position"]

 [Site "Amsterdam Donner(6)"]

 [Date "1995.??.??"]

 [Round "?"]

 [White "Shirov, A (2695)"]

 [Black "Piket, J (2625)"]

 [Result "1-0"]


1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. Nf3 e6 5. Bg5 Nbd7 6. e3 Qa5 7. cxd5 
Nxd5 8.

 Qd2 Bb4 9. Rc1 h6 10. Bh4 O-O 11. Bd3 Re8 12. a3 Bxc3 13. bxc3 Qxa3 14. 
O-O e5

 15. Nxe5 Nxe5 16. dxe5 Qc5 17. Kh1 Be6 18. e4 Nb6 19. f4 Bc4 20. Rf3 
Bxd3 21.

 Rxd3 Qc4 22. Re1 Kh7 23. Rh3 Re6 24. Bg5 1-0


Dirk Frickenschmidt:

 Hiarcs6 didn't find the move overnight. Well, Fritz 5 found the idea in 

 acceptable analysis time!  (P133 with only 4Mb Hash after 3hours 50 



I regard the critical position as a very useful test concerning the 

 tactical abilities of a strong new new program (nothing more, nothing 

 less). Remember: Fritz had to reach ply 16 to get the point.


Please post your results here.


Enrique Irazoqui:

 I tried with Rebel 9 beta on a P120/32. It finds Bf6 almost 

 immediately, then it prefers Bg5 at ply 8.04, after 1' 25'', but it 

 fails low at ply 10 and it plays a different move. Bg5 is found again 

 at ply 12.01, 42' 45'' +1.21.


Chris Whittington:

 I tried this on the release version of CSTal. We have this game in 

 book, so CSTal would play Lf6 without any thought; so I turned the book 

 off, and put it into infinite mode (PP200).


It found Bf6 very fast, then after 25 seconds:


Lg5 !! Kg8 f5 Rxe5 Bxh6 g6 fxg6 +1.90, 7 ply, 25 secs.


The move was holding later with the score at +2.03 for 9 ply, and 

 +2..41 failing high at 10 ply, then +2.80 on the 10 ply research.


Do I get a beer for this ?  :)




Hello !

 I let my version of cstal run on my AMDk5/100 (~ P5/120 Level) output 
the sendinfo of this position.

 Also I exported it into PGN for Ed and maybe Bruce and others to 


Here the sendinfo-data:


000:00 0000n/s d1 -0.84  f4f5 e6e5 h4d8 NULL 

 000:00 0000n/s d1 -0.76  h4g5 h7g8 g5d8 NULL 

 000:00 0000n/s d1 +0.01  h3g3 a7a5 d2d1 NULL 

 000:00 0000n/s d2 -0.10  h3g3 h7h8 h4f6 NULL 

 000:00 0000n/s d3 -0.32  h3g3 a8g8 d2c2 e6e8 e1d1 NULL 

 000:01 3310n/s d3 +0.00  d2c2 a8e8 e1a1 h7h8 NULL 

 000:02 2997n/s d4 -0.02  d2c2 e6e8 h4f2 a8d8 f2b6 a7b6 NULL 

 000:04 2715n/s d4 ?      h4f6 g7f6 NULL 

 000:06 2347n/s d4 ?      h4f6 a8e8 f4f5 e6f6 e5f6 NULL 

 000:08 2465n/s d5 ?      h4f6 a8e8 NULL 

 000:11 2393n/s d5 +0.95  h4f6 g7f6 f4f5 f6e5 f5e6 c4e6 e1a1 a7a6 NULL 

 000:17 2291n/s d6 ?      h4f6 g7f6 NULL 

 000:19 2232n/s d6 +0.95  h4f6 g7f6 f4f5 f6e5 f5e6 c4e6 e1a1 a7a6 NULL 

 001:01 1987n/s d6 ?      h4g5 h7g8 f4f5 e6e5 g5h6 g7g6 f5g6 NULL 

 002:33 2017n/s d7 +2.12  h4g5 c4c5 f4f5 b6c4 d2c1 e6e5 g5h6 c5d6 f5f6 
h7g8 NULL 

 004:15 2010n/s d8 +2.13  h4g5 c4c5 f4f5 b6c4 d2f4 e6e5 g5h6 h7g8 h6g7 

 013:13 2028n/s d9 ?      h4g5 c4c5 NULL 

 023:27 2042n/s d9 ?      h4g5 c4c5 f4f5 b6c4 d2f4 e6e8 e5e6 f7e6 f5f6 

 Played h4g5


The key-move is found after 61 seconds and this makes Chris PPro200 
2.44 times faster than my AMDK5/100.

 Please recognize the NPS. The score in the 9 ply research is 3.xy pawns 
! I don't know why this is not reported 

 in the sendinfo file, instead the ? is printed. Chris - a bug ??


Thanks Dirk for this nice test-position. :-)

 Maybe one day we visit you in Wuppertal and Chris gets all the beers 
and if you would have a cup of tea for 

 me, this would be nice too.


One question: don't you have a difficult position ??

 Has anybody else a more difficult position ?

 It took more time to input the position than CSTal needed time to solve 

 That is not acceptable ! :-)

 Please more difficult tasks.


Enrique - one question: when is it possible to download the 
threads/posts OFFLINE ?


Thanks to all. I hope we will never be disturbed heavily like in 


This page took 0.02 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.