Author: Robert Hyatt
Date: 18:48:59 08/20/02
Go up one level in this thread
On August 20, 2002 at 18:42:07, Uri Blass wrote: >On August 20, 2002 at 17:49:34, Robert Hyatt wrote: > >>On August 20, 2002 at 17:02:52, Uri Blass wrote: >> >>>I include only gam,es from 1993 >>> >>>Deep Blue - Hamann,S >>>5R2/r5pp/3N2k1/3Ppp2/2P5/p7/P5PP/2n3K1 w - - 0 38 bm Nb5 >> >>What did DT play here? I only find Nb5 and I find it instantly and stick >>with it... score =+1.5 so apparently DT was winning here. Did it lose by >>playing something else??? > >Yes it lost by Rxf5 > >This is a typical computer move because it wins a pawn if you do not search deep >enough when Nb5 does not win a pawn(unless you search very deep and I am sure >that your +1.5 is based on poasitional factors). The problem here is that we don't know what happened in the game. IE deep thought _always_ played over a network connection. In the late 80's and early 90's the net was not as accessible and reliable as it is today, and it is certainly possible that they lost the connection, didn't know it, and after they found out, they had to move quickly. Happened to me and Cray Blitz many times. Happened to deep thought against Fritz in Hong Kong 1995 in fact... > >programs on the hardware of that time could not find Nb5 in 15 minutes > >Movei also need time to find it(almost 5 minutes on p800 to see that Rxf5 is not >winning) > >12 0 12893 23206423 f8f5 c1a2 h2h4 h7h6 f5e5 a2b4 e5e1 b4c2 e1e6 g6h5 e6e5 h5h4 >e5e4 h4g5 e4e5 g5h4 >12 1 22487 39275894 d6b5 >12 23 25219 44104915 d6b5 a7d7 f8e8 e5e4 e8e6 g6g5 d5d6 d7d8 b5a3 c1a2 e6e7 g5f6 >a3b5 d8d6 b5d6 f6e7 d6f5 e7f6 > > > >> >> 7 0.16 ++ 1. Nb5!! >> 7 0.24 1.37 1. Nb5 Rb7 2. Re8 e4 3. Nxa3 Ne2+ 4. >> Kf2 Rb2 >> 7-> 0.26 1.37 1. Nb5 Rb7 2. Re8 e4 3. Nxa3 Ne2+ 4. >> Kf2 Rb2 >> 8 0.41 1.18 1. Nb5 Rd7 2. Ra8 Nxa2 3. Ra6+ Kg5 >> 4. Rxa3 Nb4 5. d6 >> 8-> 0.48 1.18 1. Nb5 Rd7 2. Ra8 Nxa2 3. Ra6+ Kg5 >> 4. Rxa3 Nb4 5. d6 >> 9 0.63 1.22 1. Nb5 Rd7 2. Ra8 Nxa2 3. Rxa3 Nb4 >> 4. Kf2 Nxd5 5. cxd5 Rxd5 >> 9-> 0.80 1.22 1. Nb5 Rd7 2. Ra8 Nxa2 3. Rxa3 Nb4 >> 4. Kf2 Nxd5 5. cxd5 Rxd5 >> 10 1.10 1.18 1. Nb5 Rd7 2. Ra8 Nxa2 3. Ra6+ Kg5 >> 4. Rxa3 Nc1 5. Rh3 h5 6. Rc3 >> 10-> 1.58 1.18 1. Nb5 Rd7 2. Ra8 Nxa2 3. Ra6+ Kg5 >> 4. Rxa3 Nc1 5. Rh3 h5 6. Rc3 >> 11 2.85 1.38 1. Nb5 Rd7 2. d6 Ne2+ 3. Kf2 Nd4 4. >> Nxd4 exd4 5. c5 Kg5 6. Ra8 d3 7. Rxa3 >> 11-> 4.10 1.38 1. Nb5 Rd7 2. d6 Ne2+ 3. Kf2 Nd4 4. >> Nxd4 exd4 5. c5 Kg5 6. Ra8 d3 7. Rxa3 >> 12 6.41 1.36 1. Nb5 Rd7 2. d6 Nxa2 3. c5 Nb4 4. >> Nxa3 e4 5. Nb5 Nc6 6. Re8 Kg5 >> 12-> 9.37 1.36 1. Nb5 Rd7 2. d6 Nxa2 3. c5 Nb4 4. >> Nxa3 e4 5. Nb5 Nc6 6. Re8 Kg5 >> 13 20.40 1.56 1. Nb5 Rd7 2. d6 Nd3 3. Rc8 e4 4. Rc7 >> Rd8 5. c5 e3 6. Kf1 Nc1 7. c6 Nxa2 >> 13-> 26.10 1.56 1. Nb5 Rd7 2. d6 Nd3 3. Rc8 e4 4. Rc7 >> Rd8 5. c5 e3 6. Kf1 Nc1 7. c6 Nxa2 >> 14 48.10 1.49 1. Nb5 Rd7 2. d6 Nd3 3. Ra8 e4 4. Rxa3 >> f4 5. Rc3 Nc5 6. Kf2 e3+ 7. Ke2 Kf5 >> 8. h4 >> >> >>> >>>Note that at that time the position was tested on other programs and no program >>>could find Nb5 in 15 minutes but today chess programs are better and have no >>>problem to find Nb5 >>> >>>8/r5pp/3N2k1/3PpR2/2P5/p7/n5PP/6K1 w - - 0 39 am Rxe5 bm Rf3 (I am not sure if >>>h4 is enough for a draw because of h5) >> >> >>I get Rf3 instantly here with a draw score. Score 0.00 from the 1 second >>mark on. Did DT play Re5? I could see where it might miss the draw since >>it had a repetition flaw... > >Yes it played Rxe5 and lost. OK this could be a place where Crafty thinks white is losing and it goes for the perpetual. It is possible another program might think white is winning and go for something else that avoids the draw... Again, we don't know what happened here, so conclusions are hard to draw... > >> 9 2.30 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> (2) 9-> 2.65 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> 10 3.00 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> (3) 10-> 5.40 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> (2) 11 6.40 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> (2) 11-> 12.10 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> 12 14.93 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> (2) 12-> 26.11 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> 13 30.78 0.00 1. Rf3 Nc1 2. Rg3+ Kh5 3. Rh3+ Kg4 >> 4. Rg3+ Kh5 >> >> >>> >>>Deep Blue - Kristensen,B >>>r1b2rk1/ppppnppp/2n3q1/1N6/2B1P3/2P1b3/PP3PPP/RN1Q1RK1 w - - 0 1 bm fxe3 >>> >> >>This one is harder (for me/crafty). Crafty likes Nxc7 for a good while. >>I ran these on a quad 550 and after 1:44, finishing depth=12, it had not >>changed... I might run this longer later tonite... > >I agree that this is not simple >even movei that does not like knight in the corner wants to play Nxc7 > >Here is some analysis of movei > >11 -38 11637 14297599 b5c7 d7d5 e4d5 c6e5 f2e3 e5c4 d1d4 c8h3 f1f2 b7b5 c7a8 >f8a8 >12 -59 20619 25755380 b5c7 d7d5 e4d5 c6e5 f2e3 e5c4 d1d4 e7f5 d4c4 f5e3 c4e2 >e3f1 c7a8 >12 -59 42033 52206112 b5c7 d7d5 e4d5 c6e5 f2e3 e5c4 d1d4 e7f5 d4c4 f5e3 c4e2 >e3f1 c7a8 > > >My opinion is that white position is already not good and white also did a >mistake earlier in the game. > >The previous mistake was Nb5 that should lose a pawn. > >[D]r1b2rk1/ppppnppp/2n3q1/2b5/2BNP3/2P1B3/PP3PPP/RN1Q1RK1 w - - 0 9 > >> >> >> >>> >>> >>> >>>Kristiansen,J - Deep Blue >>>1r5r/5ppk/ppppqP1p/4p1PR/P1P1P2P/1P4Q1/2P3R1/7K b - - 0 35 bm Kg8 >> >> >>What does this matter? Black is dead lost already it seems. My score starts >>out at +1.5, goes to +2.7 within seconds... Thru depth=11, score is +2.77 for >>g6. g6 fails low to +3.5 at depth=12, Kg8 becomes best with a score of +2.81. >>I hardly see the difference between the two moves and it looks hopeless >>already... > >I see only a pawn down for Kg8 when g6 lose the game by simple tactics. >I guess that you have +1.81 positional score because of black's positional >problems. > >There are other programs that see only +1.xx for white after kg8. > >Uri
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.