Rybka vs Deep Blue

Dann Corbit · Post by **Dann Corbit** » Tue Jan 05, 2010 10:31 pm

Dann Corbit wrote:

OliverBr wrote:
M ANSARI wrote:
You know ... I wonder if that game was really winning after Nxf2! White did not play optimal moves and could have resisted much more.

After Nxf2!

[d]3rr1k1/1b1q1p1p/p2b1npB/2pP4/1p6/4N1P1/PPQ1NnBP/R2R2K1 w - - 0 1

White recaptured with bishop which is not best as it puts the Knight back in with tempo. I think if white had captured with King and given back material things would have been quite different. I end up with this after some quick analysis with Rybka 3. It is still game on after this although black has an advantage.

Hey! This is blasphemy. You are doubting an IMMORTAL Sacrifice!

Anyway you are correct, after 25...Rxe3 26.Kxe3! yields to a quite drawn position, OliThink never doubts playing Kxe3 at any depth.
30. Bf4 ??? is the next big blunder by white instead of 30. Bc1.

"Immortal Sacrifice, we put its name to the test. And it failed" (King Leonidas)

It's an interesting position. I'm not quite sure that the immortal has died yet.

Here is a precursory examination of the position on multi-pv:

Code: Select all

Analysis from C:\test\gandalf.epd   
1/4/2010 7:44:44 PM Level: 60 Seconds
Analyzing engine: Rybka 3 multi-pv

1) Nxf2;                
    Searching move: Ne4xf2
    Best move (Rybka 3 multi-pv): Bd6-f8
    Not found in: 01:00
      2	00:00	       3.154	3.229.696	-0.08	Bd6f8
      2	00:00	       2.388	2.445.312	 0.00	g6g5
      2	00:00	       2.234	2.287.616	+0.31	Re8e5
      2	00:00	       2.730	2.795.520	+0.38	Nf6g4
   ---------------------------------------------------------------------------
      3	00:00	       4.415	265.938	-0.04	Bd6f8
      3	00:00	       4.099	246.904	+0.02	g6g5
      3	00:00	       3.942	237.447	+0.12	Re8e5
      3	00:00	       3.605	217.148	+0.35	Nf6g4
   ---------------------------------------------------------------------------
      4	00:00	       5.724	183.168	-0.06	g6g5
      4	00:00	       8.539	273.248	 0.00	Ne4xf2
      4	00:00	       5.391	324.728	+0.12	Re8e5
      4	00:00	       5.047	304.007	+0.25	Nf6g4
   ---------------------------------------------------------------------------
      5	00:00	      10.195	217.493	-0.13	Nf6g4 Ne3xg4
      5	00:00	      23.786	193.308	-0.01	Bd6f8 Bh6xf8
      5	00:00	      15.841	205.331	 0.00	Ne4xf2 Kg1xf2 Re8xe3
      5	00:00	      15.014	194.611	+0.05	Re8e5 f2f3
   ---------------------------------------------------------------------------
      6	00:00	      59.127	203.858	-0.05	Bd6f8 Bh6xf8 Kg8xf8 Ne2f4 Kf8g8
      6	00:00	      42.065	196.687	 0.00	Ne4xf2 Kg1xf2 Re8xe3 Bh6xe3 Nf6g4+ Kf2f3 Ng4xh2+ Kf3f2 Nh2g4+
      6	00:00	      35.085	207.670	 0.00	Re8e5 Bh6f4 Re5e8
      6	00:00	      76.648	200.735	+0.27	Nf6g4 Ne3xg4 Qd7xg4 f2f3 Qg4h5 Bh6f4 Ne4g5 Bf4xd6
   ---------------------------------------------------------------------------
      7	00:01	      85.280	199.376	 0.00	Re8e5 Bh6f4 Re5e8 Bf4h6
      7	00:01	     299.101	206.526	+0.06	Bd6b8 Ra1c1 Qd7b5 f2f3
      7	00:01	      80.708	203.058	+0.27	Nf6g4 Ne3xg4 Qd7xg4 f2f3 Qg4h5 Bh6f4 Ne4g5 Bf4xd6
      7	00:01	     112.933	205.405	+0.57	Ne4xf2 Kg1xf2 Re8xe3 Bh6xe3 Nf6g4+ Kf2f3 Ng4xh2+ Kf3f2 Nh2g4+ Kf2f3 Qd7e6 Be3g5
   ---------------------------------------------------------------------------
      8	00:03	     794.420	219.980	 0.00	Re8e5 Bh6f4 Re5e8 Bf4h6
      8	00:03	     756.421	221.623	+0.02	Nf6g4 Ne3xg4 Qd7xg4 Ne2f4 Qg4f5 Ra1c1 Re8e5
      8	00:03	     756.424	221.624	+0.06	Bd6b8 Ra1c1 Qd7b5 f2f3 Ne4d6 Bh6g5 Rd8c8 Bg2f1
      8	00:03	     686.274	221.825	+1.26	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Kd2c1 Rd8e8 Ne2f4
   ---------------------------------------------------------------------------
      9	00:06	   1.261.868	211.274	 0.00	Nf6g4 Ne3xg4 Qd7xg4 Ne2f4 Ne4f6 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3
      9	00:06	   1.390.375	210.737	+0.12	Bd6c7 Ra1c1 Ne4xf2 Kg1xf2 Re8xe3
      9	00:06	   1.218.235	212.083	+0.26	Bd6b8 Ra1c1 Bb8a7 a2a3 a6a5 f2f3
      9	00:06	     869.518	217.804	+1.28	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Kd2c1 Rd8e8 Ne2f4 Nh6g4
   ---------------------------------------------------------------------------
     10	00:14	   2.903.486	208.044	 0.00	Nf6g4 Ne3xg4 Qd7xg4 Ne2f4 Ne4f6 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3
     10	00:14	   2.817.960	208.752	 0.00	Bd6c7 Ra1c1 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6
     10	00:14	   2.540.056	207.103	+0.09	Bd6b8 Ne2f4 Bb8xf4 g3xf4 Rd8c8 d5d6 Rc8d8 a2a3 b4xa3 Ra1xa3
     10	00:14	   1.759.945	208.128	+1.33	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Qc2d3 Bd6e5 Kd2c1 c5c4 Qd3d2
   ---------------------------------------------------------------------------
     11	00:20	   4.196.071	207.233	 0.00	Nf6g4 Ne3xg4 Qd7xg4 Ne2f4 Ne4f6 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3
     11	00:20	   4.098.956	207.757	 0.00	Bd6c7 Ra1c1 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6
     11	00:20	   3.872.840	205.662	+0.17	Bd6b8 Ne2f4 Bb8xf4 g3xf4 Rd8c8 Ne3c4 Qd7f5 d5d6 Rc8d8 Bh6g5
     11	00:20	   3.492.530	203.943	+1.24	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Bd6f8 Ra1d1 Nh6g4 h2h3 Bb7xd5 h3xg4 Bd5xg2+ Kd2c1
   ---------------------------------------------------------------------------
     12	00:40	   8.170.078	208.668	 0.00	Nf6g4 Ne3xg4 Qd7xg4 Ne2f4 Ne4f6 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3 Bh3g2 Qf3g4 Bg2h3 Qg4f3
     12	00:40	   7.816.235	209.491	 0.00	Bd6c7 Ra1c1 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6 Ne3c4 Bb6c7 Nc4e3 Bc7b6
     12	00:40	   7.315.631	207.696	+0.08	Bd6b8 Ne2f4 Bb8xf4 g3xf4 Qd7c8 Rd1e1 c5c4 Ra1c1 c4c3 f4f5 g6xf5 b2xc3
     12	00:40	   6.515.481	207.703	+1.40	Ne4xf2 Kg1xf2 Re8xe3 Bh6xe3 Nf6g4+ Kf2f3 Ng4xh2+ Kf3f2 Nh2g4+ Kf2f3 Qd7e6 Be3c1 c5c4 Qc2e4 Qe6xe4+ Kf3xe4 Ng4f2+ Ke4d4 Nf2xd1 Bc1g5 Rd8e8 Ra1xd1
   1/4/2010 7:45:49 PM, Time for this analysis: 00:01:00, Rated time: 01:00

I will do a much deeper analysis later with several engines.
Positions like this are why I really love chess.

Still looks best after one hour Rybka time (4 CPUs at 3 GHz, 2 GB hash):

Code: Select all

Analysis from C:\test\gandalf.epd   
1/5/2010 12:22:54 PM Level: 3600 Seconds
Analyzing engine: Rybka 3

1) Nxf2;                
    Searching move: Ne4xf2
    Best move (Rybka 3): Ne4xf2
    identical moves! Found in: 00:01
      2	00:00	         983	62.912	+0.31	Re8e5
      2	00:00	       1.414	90.496	+0.38	Nf6g4
      3	00:00	       2.038	65.216	+0.35	Nf6g4
      4	00:00	       2.495	79.840	+0.25	Nf6g4
      5	00:00	       3.535	113.120	+0.07	Nf6g4 Ne3xg4
      6+	00:00	      19.599	160.555	+0.27	Nf6g4
      6	00:00	      21.607	177.004	+0.10	Nf6g4 Ne3xg4 Qd7xg4 f2f3 Qg4h5 Bh6f4 Ne4f6 Bf4xd6 Rd8xd6 Qc2xc5 Rd6xd5 Qc5xb4
      7	00:00	      25.760	168.014	+0.10	Nf6g4 Ne3xg4 Qd7xg4 f2f3 Qg4h5 Bh6f4 Ne4f6 Bf4xd6 Rd8xd6 Qc2xc5 Rd6xd5 Qc5xb4
      8	00:00	      48.939	168.732	+0.10	Nf6g4 Ne3xg4 Qd7xg4 f2f3 Qg4h5 Bh6f4 Ne4f6 Bf4xd6 Rd8xd6 Qc2xc5 Rd6xd5 Qc5xb4 Rd5xd1+ Ra1xd1 Bb7xf3
      9	00:01	     102.164	176.418	+0.17	Nf6g4 Ne3xg4 Qd7xg4 f2f3 Qg4h5 Bh6f4 Ne4f6 Bf4xd6 Rd8xd6 Qc2xc5 Rd6xd5 Qc5xb4 Rd5xd1+ Ra1xd1 Bb7xf3
      9+	00:01	     126.397	176.335	+0.30	Ne4xf2
      9+	00:01	     162.086	186.489	+0.90	Ne4xf2
      9	00:02	     262.528	189.315	+1.31	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Qc2d3 Nh6g4 Bg2h3 Bd6e5 Kd2c1
     10	00:03	     563.775	186.890	+1.13	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Nh6f5 Ra1e1 Bd6e5 Qc2b3 Nf5e7
     11+	00:05	     918.664	193.244	+1.33	Ne4xf2
     11	00:07	   1.247.799	194.068	+1.38	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1e1 Nh6g4 h2h3 Ng4f6 Ra1d1 Nf6xd5 Bg2xd5 Bb7xd5 Kd2c1
     12	00:08	   1.522.623	195.948	+1.41	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Qc2d3 Bd6e5 Kd2c1 Qd7e7 Qd3d2 Nh6g4
     13	00:12	   2.317.587	200.152	+1.40	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Qc2d3 Bd6e5 Kd2c1 Nh6f5 Rd1e1 c5c4 Qd3d2 Be5g7 Bg2h1
     14	00:31	   6.330.812	208.502	+1.54	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Qc2d3 Qd7e7 Qd3e3 Qe7f8 Kd2c1
     15	01:06	  13.442.743	207.425	+1.51	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Qc2b1 Qa2a5 Re1d1 Rd8xd1+ Kc1xd1 Nh6f5 Kd1e1 Qa5a4 Ke1f2 Qa4b3 Rf4e4
     16	01:22	  16.421.731	206.499	+1.43	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Qc2b1 Qa2a5 Re1d1 Rd8xd1+ Kc1xd1 Nh6f5 Kd1e1 Qa5a4 Ke1f2 Qa4b3 Rf4e4
     17	01:55	  22.829.941	204.246	+1.43	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Qc2b1 Qa2a5 Re1d1 Rd8xd1+ Kc1xd1 Nh6f5 Kd1e1 Qa5a4 Ke1f2 Qa4b3 Rf4e4
     18	03:10	  37.727.633	203.706	+1.43	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Qc2b1 Qa2a5 Re1d1 Rd8xd1+ Kc1xd1 Nh6f5 Kd1e1 Qa5a4 Ke1f2 Qa4b3 Rf4f3 Qb3d5
     19+	16:04	 202.448.722	215.158	+1.63	Ne4xf2
     19+	41:32	 538.807.391	221.415	+1.83	Ne4xf2

M ANSARI · Post by **M ANSARI** » Tue Jan 05, 2010 11:21 pm

Of course I do think that black is better after Nxf2, but it is not the knock out blow that the author would want us to see. A knight + queen can be very strong and most likely black has a technical win somewhere, but also the position is very dynamic with many drawing chances. I played it out to a black win, but I am sure white can offer even more resistance to what I played as white ... hell it could even be drawn. I have seen much more promising positions proven to be drawn.

Dann Corbit · Post by **Dann Corbit** » Wed Jan 06, 2010 1:30 am

M ANSARI wrote:Of course I do think that black is better after Nxf2, but it is not the knock out blow that the author would want us to see. A knight + queen can be very strong and most likely black has a technical win somewhere, but also the position is very dynamic with many drawing chances. I played it out to a black win, but I am sure white can offer even more resistance to what I played as white ... hell it could even be drawn. I have seen much more promising positions proven to be drawn.

A bit deeper search (not fully decisive by any means):

Code: Select all

Analysis from C:\test\gandalf.epd   
1/5/2010 3:16:32 PM Level: 3600 Seconds
Analyzing engine: Stockfish64

1) Nxf2;                
    Searching move: Ne4xf2
    Best move (Stockfish64): Ne4xf2
    identical moves! Found in: 00:01
      2	00:01	         268	142	-0.08	Bb7c8 Ne2f4
      2	00:01	         357	189	+0.24	Bd6c7 Ne2f4
      2	00:01	         470	249	+0.40	g6g5 Ra1c1
      3	00:01	       1.092	579	+0.16	g6g5 f2f4 Ne4xg3 Ne2xg3 Re8xe3 Bh6xg5
      4	00:01	       3.278	1.735	+0.16	g6g5 f2f4 Ne4xg3 Ne2xg3 Re8xe3 Bh6xg5
      5	00:01	      12.632	6.637	-0.08	g6g5 Ra1c1 Qd7b5 Rd1e1 Qb5d7
      6	00:01	      22.701	11.854	 0.00	g6g5 Ra1c1 a6a5 f2f3 Ne4xg3 h2xg3 Re8xe3 Bh6xg5 Nf6xd5 Bg5xd8 Qd7xd8
      7	00:01	      62.913	32.131	-0.28	g6g5 Ra1c1 Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 b2b3
      7+	00:01	      66.455	33.888	+0.40	Ne4xf2 Kg1xf2 Re8xe3 Bh6xe3 Nf6g4+ Kf2f3 Ng4xh2+ Kf3f2 Nh2g4+
      8	00:01	      75.482	38.374	+0.88	Ne4xf2 Kg1xf2 Re8xe3 Bh6xe3 Nf6g4+ Kf2f3 Ng4xh2+ Kf3f2 Nh2g4+ Kf2f3 Rd8e8 Be3xc5 Qd7c8
      9	00:02	     151.886	75.678	+0.56	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 a2a3 Nh6g4 h2h3 Ng4f2 Rd1f1 Nf2xh3 a3xb4 c5xb4
     10	00:02	     388.736	186.444	+0.72	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 a2a3 Nh6g4 h2h3 Ng4f2 Rd1f1 Nf2xh3 a3xb4 c5xb4 Qc2a4 Qd7g4
     11	00:02	     616.615	285.337	+1.29	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 a2a3 Rd8e8 Rd1e1 Nh6f5 Ne2f4 Nf5d4 Qc2d3 a6a5 Re1xe8+ Qd7xe8 Ra1e1
     12	00:02	   1.460.060	604.579	+0.92	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Ra1c1 Nh6g4 h2h3 Ng4f2 Rd1f1 Nf2xh3 b2b3 Rd8e8 Bg2xh3 Qd7xh3
     13	00:02	   3.050.387	1.080.547	+1.13	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Qc2c4 Nh6g4 Ne2f4 Ng4xh2 Nf4d3 Bd6xg3 Qc4xc5 Bb7xd5 Bg2xd5 Qd7xd5 Qc5xd5 Rd8xd5
     14	00:03	   6.289.358	1.743.170	+1.17	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 a2a3 Rd8e8 Rd1e1 Qd7g4 Kd2c1 Nh6f5 h2h3 Qg4h5 Qc2d2 Nf5e3 g3g4
     15+	00:04	  10.273.882	2.248.606	+1.45	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 a2a3 Rd8e8 Rd1e1 Qd7g4 Kd2c1 Nh6f5 h2h3 Qg4g5+ Qc2d2 Qg5e7 a3xb4 Bd6xg3
     16	00:05	  15.613.991	2.654.537	+1.37	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 a2a3 Rd8e8 Rd1e1 Qd7g4 a3xb4 Qg4g5+ Kd2d1 c5xb4 Ne2f4 Qg5g4+ Re1e2 Bd6xf4 g3xf4 Bb7xd5 Bg2xd5 Qg4g1+ Kd1d2 Re8xe2+ Kd2xe2 Qg1xa1 Qc2d2 Qa1a4
     17	00:11	  41.386.082	3.456.906	+1.49	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Kd2c1 Qd7e7 Ne2f4 Nh6g4 a2a3 g6g5 Nf4h5 Ng4e3 Rd1e1 Ne3xc2 Re1xe7 Bd6xe7 Kc1xc2 Bb7xd5 Bg2xd5 Rd8xd5 a3xb4 c5xb4 Ra1xa6
     18	00:18	  70.732.297	3.820.063	+1.37	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Kd2c1 Qd7e7 Ne2f4 Nh6g4 a2a3 g6g5 Nf4h5 Ng4e3 Rd1e1 Ne3xc2 Re1xe7 Bd6xe7 Kc1xc2 Kg8f8 a3xb4 c5xb4 g3g4 Be7d6 Kc2b3 Bd6xh2 Kb3xb4 Bb7xd5 Bg2xd5 Rd8xd5 Ra1xa6
     19	00:33	 135.240.471	4.081.006	+1.53	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Kd2c1 Qd7e7 Ne2f4 Nh6g4 a2a3 g6g5 Nf4h5 Ng4e3 Rd1e1 Ne3xc2 Re1xe7 Bd6xe7 Kc1xc2 Kg8f8 a3xb4 c5xb4 Bg2e4 Rd8d6 b2b3 Rd6d8 h2h3 Bb7xd5 Be4xd5 Rd8xd5 Ra1xa6
     20	00:38	 157.406.168	4.130.094	+1.53	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Kd2c1 Qd7e7 Ne2f4 Nh6g4 a2a3 g6g5 Nf4h5 Ng4e3 Rd1e1 Ne3xc2 Re1xe7 Bd6xe7 Kc1xc2 Kg8f8 a3xb4 c5xb4 Bg2e4 Rd8d6 b2b3 Rd6d8 Ra1d1 Rd8c8+ Kc2b2 Rc8c5 Rd1c1 f7f5 Rc1xc5 Be7xc5 Be4xf5 Bb7xd5 Bf5xh7
     21	00:54	 231.848.478	4.242.812	+1.53	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Kd2c1 Qd7e7 a2a3 Nh6g4 Ne2f4 g6g5 Nf4h5 Ng4e3 Rd1e1 Ne3xc2 Re1xe7 Bd6xe7 Kc1xc2 Kg8f8 a3xb4 c5xb4 Bg2e4 Rd8d6 b2b3 Rd6d8 Ra1d1 Rd8c8+ Kc2b2 Rc8c5 Rd1c1 f7f5 Rc1xc5 Be7xc5 Be4xf5 Bb7xd5 Bf5xh7
     22-	02:24	 620.038.388	4.284.548	+1.33	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Rf4e4 b4b3 Qc2b1 Qa2a5 Re4e8+ Kg8g7 Re1e2 Nh6f5 Re8xd8 Qa5xd8 Qb1e4 Nf5d4 Re2f2
     23	03:11	 810.487.309	4.247.236	+1.21	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Rf4e4 b4b3 Qc2b1 Qa2a5 Re4e8+ Kg8g7 Re1e2 Nh6f5 Re8xd8 Qa5xd8 Qb1e4 Nf5d4 Re2f2 Qd8d6 Kc1d2
     24	06:47	1.743.086.181	4.276.305	+1.41	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Rf4e4 b4b3 Qc2b1 Qa2a5 Re4e8+ Kg8g7 Re1e2 Nh6f5 Re8xd8 Qa5xd8 Qb1e4 Qd8a5 Re2e1 Nf5d4 Qe4e5+ f7f6 Qe5e7+ Kg7h6 Qe7e4 Qa5a1+ Qe4b1 Qa1xb1+ Kc1xb1
     25+	28:49	7.479.789,039	4.324.443	+1.89	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 Rf4e4 b4b3 Qc2b1 Qa2a5 Re4e8+
     26	33:48	8.788.989,709	4.332.719	+1.65	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 g3g4 c5c4 Rf4e4 Qa2a1+ Qc2b1 Qa1a5 Re4e5 Qa5a4 Qb1e4 Nh6xg4 Re5e8+ Rd8xe8 Qe4xe8+ Qa4xe8 Re1xe8+ Kg8g7 Re8e4 Ng4xh2 Re4xc4 b4b3 Kc1d2 g6g5 Rc4a4 f7f5 Ra4xa6 h7h5 Kd2c3 h5h4 Kc3xb3
     27	59:29	15.644.572,742	4.383.427	+1.57	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 g3g4 c5c4 Rf4e4 Qa2a1+ Qc2b1 Qa1a5 Re4e5 Qa5a4 Qb1e4 Nh6xg4 Re5e8+ Rd8xe8 Qe4xe8+ Qa4xe8 Re1xe8+ Kg8g7 Re8e4 Ng4xh2 Re4xc4 a6a5 Rc4c5 Nh2f3 Rc5xa5 g6g5 Kc1d1 g5g4 Ra5b5 Kg7f6 Rb5xb4 h7h5 Kd1e2
   1/5/2010 4:16:37 PM, Time for this analysis: 01:00:00, Rated time: 00:01

Don · Post by **Don** » Wed Jan 06, 2010 8:52 pm

Mike S. wrote:To explore this question, the logs of the match from 1997 may be useful:

http://researchweb.watson.ibm.com/deepb ... tml/c.html

(the "cleaned" versions are easier to read)

There we find calculation depths which are very small compared to what we see from today's PCs at tournament time, IIRC 12 or 13 plies being typical. In 2010, that is blitz. On the other hand, I think Bob Hyatt explained once that D.B. made huge extensions. - Also, I can guess that the search tree of that system with 256 parallel chess cpus is very differerent even from todays mp engines on quads or octos, but I don't know what that means for playing strength, in principle.

Also, from all the discussion about it it seems that these six games only, unfortunately do not provide good test positions for comparisons.

The question about Rybka vs. Deep Blue of 1997 was also part of the recent video interview with Vas Rajlich, and he thinks that "that probably Rybka 1 could beat that version of Deep Blue."

http://rybkaforum.net/cgi-bin/rybkaforu ... ?tid=14546
http://rybkaforum.net/cgi-bin/rybkaforu ... pid=214321
(with a transcript / .doc file attached)

I think he was being modest. Rybka would easily win such a match. Deep Blue was fortunate to get the win after an early uncharacteristic blunder and in a very short match. But when Rybka plays against the top players it must give odds or the match is pretty much a forgone conclusion.

Still, his answer was safe as Deep Blue was never really tested in a publicly in a statistically significant way. So we don't have an easy way to really say how good it was.

OliverBr · Post by **OliverBr** » Mon Jan 11, 2010 6:07 am

Dann Corbit wrote:

Code: Select all

     27	59:29	15.644.572,742	4.383.427	+1.57	Ne4xf2 Kg1xf2 Re8xe3 Kf2xe3 Nf6g4+ Ke3d2 Ng4xh6 Rd1f1 Qd7e7 Ra1e1 Bb7xd5 Bg2xd5 Qe7g5+ Ne2f4 Bd6xf4+ Rf1xf4 Qg5xd5+ Kd2c1 Qd5xa2 g3g4 c5c4 Rf4e4 Qa2a1+ Qc2b1 Qa1a5 Re4e5 Qa5a4 Qb1e4 Nh6xg4 Re5e8+ Rd8xe8 Qe4xe8+ Qa4xe8 Re1xe8+ Kg8g7 Re8e4 Ng4xh2 Re4xc4 a6a5 Rc4c5 Nh2f3 Rc5xa5 g6g5 Kc1d1 g5g4 Ra5b5 Kg7f6 Rb5xb4 h7h5 Kd1e2
   1/5/2010 4:16:37 PM, Time for this analysis: 01:00:00, Rated time: 00:01

So Stockfish searches in a tactical, open position with almost every piece on board 27 plies within an hour?!

Incredible!

Rybka vs Deep Blue

Re: Rybka vs Deep Blue

Re: Rybka vs Deep Blue

Re: Rybka vs Deep Blue

Re: Rybka vs Deep Blue

Re: Rybka vs Deep Blue