Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: A Rybka Miracle

Author: Dann Corbit

Date: 13:37:25 12/22/05

Go up one level in this thread


On December 22, 2005 at 15:23:07, Dann Corbit wrote:

>On December 22, 2005 at 14:35:56, Drexel,Michael wrote:
>
>>On December 22, 2005 at 13:26:55, Dann Corbit wrote:
>>
>>>On December 21, 2005 at 21:20:10, Uri Blass wrote:
>>>
>>>>On December 21, 2005 at 20:30:18, Dann Corbit wrote:
>>>>>
>>>>>Generally speaking, chess engines have terrible judgement in early opening
>>>>>positions and absurd judgement in gambits.
>>>>>
>>>>>For a bit of gambit fun, try this position on your chess engine:
>>>>>
>>>>>[D]r1bqk1nr/pppp1ppp/2n5/b7/2BpP3/2P2N2/P4PPP/RNBQ1RK1 b kq - am dxc3; bm Nge7;
>>>>>
>>>>>and see how long it takes to find Nge7.  Most engines I tried cannot find it in
>>>>>an 8 hour search.  I have not tried it lately, but I guess that they still
>>>>>cannot find it at 40/2, no matter what the hardware or program (w/o book, that
>>>>>is).
>>>>
>>>>am dxc3 is easy for rybka
>>>
>>>Not for the others, though.
>>>
>>>>bm Nge7 is an harder problem and Rybka likes Nf6
>>>>Are you sure that Nge7 is the only move?
>>>
>>>It is clearly the best move.
>>
>>I wonder how you came to that definite conclusion?
>>You are aware of the alternative 7...Bb6 8.cxd4 d6 ?
>
>It appears to have been popular in the 1800's.  In my database, I have 512 games
>from that position:
>[D]r1bqk1nr/ppp2ppp/1bnp4/8/2BPP3/5N2/P4PPP/RNBQ1RK1 w kq -
>and white scores 58.8% of the points (well above average).
>
>I am analyzing the position now.

Rybka thinks that White is slightly better here.  Considering also that White
scores only 48.8% after the Nge7 move (a ten percent swing from this
alternative) it seems to me that it is not as good for black as Nge7.  Of
course, I am not a chess expert.  Maybe someone like Robin Smith can offer more
insight.

Analysis from E:\whatabout.epd
12/22/2005 11:47:10 AM Level: 7200 Seconds
Analyzing engine: Rybka v1.0 Beta.w32

1) "what about .. Bb6 c
    Avoid move:
    Best move (Rybka v1.0 Beta.w32): Nb1-c3
    Not found in: 2:00:00
      3	00:00	         840	9.150	+0.31	Nb1c3
      4	00:00	       1.829	17.026	+0.25	Nb1c3
      5	00:01	      10.056	38.711	+0.36	Nb1c3 Nc6a5
      6	00:01	      20.915	45.665	+0.26	Nb1c3 Bc8g4 Qd1b3
      7	00:01	      51.300	48.685	+0.04	Nb1c3 Ng8f6 Bc1f4 00
      7	00:03	     140.204	57.062	+0.14	Bc4b5 a7a6 Bb5a4 Bc8d7
      8	00:03	     156.793	58.362	+0.06	Bc4b5 a7a6 Bb5a4 Bc8d7 Nb1c3
      9	00:05	     280.855	59.753	+0.05	Bc4b5 Bc8d7 Nb1c3 Ng8e7 d4d5 Nc6a5
     10	00:09	     518.777	60.704	+0.05	Bc4b5 Bc8d7 Nb1c3 Ng8e7 d4d5 Nc6a5 Bc1f4
     11	00:18	   1.054.010	65.875	+0.01	Bc4b5 Bc8d7 Nb1c3 Ng8e7 d4d5 Nc6a5 Bc1f4
Ne7g6
     11	00:25	   1.461.362	63.537	+0.08	Nb1c3 Ng8f6 h2h3 Bb6a5 Qd1b3 00 Rf1e1
Bc8d7
     12	00:43	   2.470.458	60.255	+0.08	Nb1c3 Ng8f6 h2h3 Bb6a5 Qd1b3 00 Rf1e1
Bc8d7 d4d5
     13	01:33	   5.252.573	59.017	+0.11	Nb1c3 Ng8f6 h2h3 Bb6a5 Qd1b3 00 e4e5
d6xe5 d4xe5 Ba5xc3
     14	02:52	   9.859.016	59.036	+0.05	Nb1c3 Ng8f6 h2h3 Bb6a5 Qd1b3 00 d4d5
Nc6e7 Rf1e1 Ne7g6
     15	06:33	  22.641.701	59.271	+0.01	Nb1c3 Ng8f6 h2h3 Bb6a5 Qd1b3 Ba5xc3
Qb3xc3 Nf6xe4 Qc3c2 d6d5
     16	17:16	  58.230.898	57.597	+0.07	Nb1c3 Ng8f6 Bc1a3 00 e4e5 Nf6g4 h2h3
Ng4h6 Rf1e1 Nh6f5
     17	33:06	 109.814.735	56.663	+0.08	Nb1c3 Ng8f6 Bc1a3 00 e4e5 Nf6g4 h2h3
Ng4h6 e5xd6 c7xd6
     18	1:03:45	 213.857.145	57.257	+0.08	Nb1c3 Ng8f6 Bc1a3 00 e4e5 Nf6g4 h2h3
Ng4h6 e5xd6 c7xd6
   12/22/2005 12:54:20 PM, Time for this analysis: 01:07:05, Rated time: 2:00:00

0 of 1 matching moves
12/22/2005 12:54:21 PM, Total time: 1:07:11 AM
Rated time: 2:00:00 = 7200 Seconds



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.