Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Serious blunder by Junior 8 or excellent attack by CM? You decide.

Author: Slater Wold

Date: 23:22:21 08/02/03

Go up one level in this thread


On August 03, 2003 at 00:50:02, Mike Byrne wrote:

>On August 03, 2003 at 00:33:57, John Merlino wrote:
>
>>On August 02, 2003 at 22:05:27, Anson T J wrote:
>>
>>>[d]2r2rk1/p3ppbp/6p1/8/8/QPqBBP2/P1P3PP/2KR3R w - - 0 20
>>>
>>>Junior 8 played 20. Bf2
>>>=  (0.00)  Depth: 16   0:03:29
>>>
>>>
>>>The King 3.23 SKR then played
>>>20... Rfd8!
>>>-+  (-8.87)  Depth: 13   0:10:47
>>>
>>>and won.
>>>
>>>Does any other engine also want to play 20. Bf2 ? I'm not sure if White's
>>>position lost before move 20. or if it was a blunder on move 20. that cost
>>>Junior the game.
>>>
>>>Please post your evals and comments:
>>
>>Chessmaster 9000 (default personality), after a reasonably long search on a
>>P4-2.4G, seems to think that White's best chance is Qa4, with only a slight eval
>>advantage to Black (-0.18). After 20.Bf2, a 1-minute search gives an eval of
>>-1.37 for 20...Rfd8.
>>
>>But my vote is for "excellent attack by CM", starting with 18...Qc3!! How many
>>engines would not play this move and not gxf5?
>>
>>[D]1r3rk1/p3ppbp/2q3p1/5B2/8/QP2BP2/P1P3PP/2KR3R b - - 0 18
>>
>>Majestic Chess, for example, needs 18 seconds on my P4-2.4 to find Qc3....
>>
>>jm
>
>excellent attack by CM gets my vote...
>
>3:00 for Crafty to find Qc3 on my "slow" dual 1.7 Ghz P4 -
>
>New game
>1r3rk1/p3ppbp/2q3p1/5B2/8/QP2BP2/P1P3PP/2KR3R b - - 0 1
>
>Analysis by Crafty 19.04 ST:
>
>1...gxf5 2.Qxe7 Rfc8 3.Rd2
>  +-  (1.89)   Depth: 4   00:00:00
>1...gxf5 2.Qc5 Qa6 3.Rd7 Qxa2 4.Qxe7
>  +-  (1.72)   Depth: 5   00:00:00
>1...gxf5 2.Qc5 Qa6 3.Rd7 Qxa2 4.Qxe7
>  +-  (1.72)   Depth: 5   00:00:00
>1...gxf5 2.Qxe7 Rfc8 3.c4 Rc7 4.Rd8+ Rxd8 5.Qxd8+
>  +-  (1.58)   Depth: 6   00:00:00  105kN
>1...gxf5 2.Qxe7 Rfc8 3.c4 Rc7 4.Rd8+ Rxd8 5.Qxd8+
>  +-  (1.58)   Depth: 6   00:00:00  175kN
>1...gxf5
>  ±  (1.25)   Depth: 7   00:00:00  381kN
>1...gxf5 2.Qxa7 Rfc8 3.c4 Rb7 4.Qa5 e6 5.Rd8+ Bf8 6.Rxc8 Qxc8
>  ±  (1.16)   Depth: 7   00:00:01  737kN
>1...gxf5 2.Qxa7 Rfc8 3.c4 Rb7 4.Qa5 e6 5.Rd8+ Bf8 6.Rxc8 Qxc8
>  ±  (1.16)   Depth: 7   00:00:01  1089kN
>1...gxf5 2.Qc5 Qf6 3.Qd4 Qe6 4.Qd3 Qc8 5.Kb1
>  +-  (1.61)   Depth: 8   00:00:02  1909kN
>1...gxf5 2.Qc5 Qf6 3.Qd4 Qe6 4.Qd3 Qc8 5.Kb1
>  +-  (1.61)   Depth: 8   00:00:05  4351kN
>1...gxf5 2.Qc5 Qe6 3.Bf4 Rbc8 4.Qd5 Qe2 5.Rd2 Qa6
>  +-  (1.53)   Depth: 9   00:00:06  5033kN
>1...gxf5 2.Qc5 Qe6 3.Bf4 Rbc8 4.Qd5 Qe2 5.Rd2 Qa6
>  +-  (1.53)   Depth: 9   00:00:07  6344kN
>1...gxf5 2.Qc5 Qf6 3.Qd4 Qc6 4.Qd3 Bc3 5.Qxf5
>  ±  (1.32)   Depth: 10   00:00:11  9702kN
>1...gxf5 2.Qc5 Qf6 3.Qd4 Qc6 4.Qd3 Bc3 5.Qxf5
>  ±  (1.32)   Depth: 10   00:00:23  20422kN
>1...gxf5 2.Qc5 Qf6 3.Qd4 Qc6 4.Qd7 Qf6 5.Rd4 Qe5 6.Re1 Rfc8
>  ±  (1.20)   Depth: 11   00:00:37  33055kN
>1...gxf5 2.Qc5 Qf6 3.Qd4 Qc6 4.Qd7 Qf6 5.Rd4 Qe5 6.Re1 Rfc8
>  ±  (1.20)   Depth: 11   00:01:00  53334kN
>1...gxf5 2.Qc5 Qf6 3.Bd4 e5 4.Bb2 Qa6 5.Bxe5 Rb5 6.Qd6 Qxa2 7.Rhe1
>  ±  (1.34)   Depth: 12   00:01:20  69887kN
>1...Qc3
>  ±  (0.87)   Depth: 12   00:03:00  155326kN
>
>(Byrne, Secane, PA USA 03.08.2003)

I'd have to agree.  After looking & looking, I'm impressed any engine would have
that kind of positional know-how.

Rebel 12.0 beta 1 & beta 9 don't get it in ~3 minutes (13 ply or so).  And even
the personality I'm working on now (which has scored about 20% better than the
default or beta 9 against other engines) doesn't see it in 5 minutes.

However...

Fritz 7
1r3rk1/p3ppbp/2q3p1/5B2/8/QP2BP2/P1P3PP/2KR3R b - - 0 1

Analysis by Fritz 7:

1...gxf5 2.Qc5 Qf6 3.Qd4 Qc6 4.Qc4 Qxc4 5.bxc4 Rb2
  ±  (0.78)   Depth: 8/26   00:00:00  537kN
1...gxf5 2.Qc5 Qf6 3.Qd4 Qc6 4.Qd3 e6 5.Qc4 Bb2+ 6.Kb1
  ²  (0.59)   Depth: 9/26   00:00:00  950kN
1...gxf5 2.Qc5 Qf6 3.Qd4 Qe6 4.Qd3 Rfc8 5.Bxa7 Rb4 6.Be3
  ²  (0.62)   Depth: 10/26   00:00:02  2558kN
1...Qc3!
  ²  (0.59)   Depth: 10/28   00:00:03  3533kN
1...Qc3!
  ²  (0.44)   Depth: 10/28   00:00:03  3813kN
1...Qc3!
  =  (0.12)   Depth: 10/30   00:00:04  4506kN
1...Qc3 2.Bd3 Rfd8 3.Rhf1 Rxd3 4.Rxd3 Qa1+ 5.Kd2 Qxf1 6.Qxe7 Qxg2+ 7.Kc1 Qh1+
8.Rd1
  =  (-0.22)   Depth: 10/30   00:00:05  6451kN
1...Qc3 2.Bd3 Rfd8 3.Bf2 Rxd3 4.Rxd3 Qa1+ 5.Kd2 Qxh1 6.Qxe7 Rf8 7.Rd8
  =  (-0.16)   Depth: 11/32   00:00:09  10405kN
1...Qc3 2.Bd3 Rfd8 3.Bf2 Rxd3 4.Rxd3 Qa1+ 5.Kd2 Qxh1 6.Qxe7 Bh6+ 7.Ke2 a6 8.Rd8+
Rxd8
  =  (-0.03)   Depth: 12/36   00:00:16  17875kN
1...Qc3 2.Bd3 Rfd8 3.Bf2 Rxd3 4.Rxd3 Qa1+ 5.Kd2 Qxh1 6.Qxe7 Bh6+ 7.Ke2 a6 8.Rd8+
Rxd8
  =  (-0.03)   Depth: 12/36   00:00:20  23248kN

(,  03.08.2003)


I'd like to see an eval from CM9k, but I don't own it.  Maybe I should.  ;)



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.