Author: Uri Blass
Date: 01:54:37 05/22/02
Go up one level in this thread
On May 22, 2002 at 01:56:51, Uri Blass wrote: >On May 22, 2002 at 01:39:09, Uri Blass wrote: > >>On May 21, 2002 at 23:57:58, Vincent Lejeune wrote: >> >>>On May 21, 2002 at 18:13:21, Roy Eassa wrote: >>> >>>>On May 21, 2002 at 13:48:54, Tony Hedlund wrote: >>>> >>>>>[Event "Tony Hedlund"] >>>>>[Site "SSDF"] >>>>>[Date "2002.05.21"] >>>>>[Round "10"] >>>>>[White "Chess Tiger 14.9a Palm"] >>>>>[Black "Comet A90 P200"] >>>>>[BlackElo "2253"] >>>>>[Result "1-0"] >>>>>[Opening "C88 Ruy Lopez: Closed, 7...0-0"] >>>>> >>>>>1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O >>>>>8. d4 Nxd4 9. Nxd4 exd4 10. e5 Ne8 11. Qxd4 Bb7 12. Rd1 d6 13. Qg4 c5 >>>>>14. Bd5 Bxd5 15. Rxd5 Qc8 16. Qxc8 Rxc8 17. Bf4 dxe5 18. Bxe5 Nf6 >>>>>19. Rd3 Ng4 20. Bf4 Bf6 21. Nc3 Bd4 22. Bg3 Rce8 23. Kf1 f5 24. h3 f4 >>>>>25. hxg4 fxg3 26. f3 Re6 27. Re1 Rh6 28. Ne4 Rh1+ 29. Ke2 Rh2 30. Kf1 Bxb2 >>>>>31. Nxg3 Rh4 32. Re4 Rh6 33. Rd5 c4 34. g5 Rb6 35. Re7 Rbb8 36. Ra7 Rfd8 >>>>>37. Rxd8+ Rxd8 38. Ke2 Rd6 39. Ne4 Re6 40. Ke3 Bc1+ 41. Kd4 Rg6 >>>>>42. Ra8+ Kf7 43. Kd5 Bf4 44. c3 Ke7 45. Rh8 Bxg5 46. Rxh7 Bf4 47. g4 Bh6 >>>>>48. Ke5 Kf7 49. Rh8 Bc1 50. Ra8 Ke7 51. Ra7+ Kf8 52. Rb7 Rc6 53. f4 Rg6 >>>>>54. g5 Be3 55. Rb8+ Ke7 56. f5 Rb6 57. f6+ gxf6+ 58. gxf6+ Kd7 59. Rh8 Re6+ >>>>>60. Kd5 Rxe4 61. Kxe4 Bc5 62. Kd5 Ba3 63. f7 a5 64. Ra8 Bf8 65. Rxf8 Ke7 >>>>>66. Ra8 Kxf7 67. Rxa5 1-0 >>>>> >>>>>A fantastic game by Tiger. Look at the centralized knight and king, they >>>>>dominate the board. >>>>> >>>>>Tony >>>> >>>> >>>>Opening question: >>>> >>>>doesn't this, instead of 12...d6, win material for Black? : >>>>[d] r2qnrk1/1bppbppp/p7/1p2P3/3Q4/1B6/PPP2PPP/RNBR2K1 b - - 0 11 >>>>12...c5 13.Qxd7 c4 14.Nc3 Rb8 >>> >>>Yes, you're totally right, the trap bishop is a theme in this opening (8.d4), >>>the level of CT 14.9a Palm is not nearly 2000 elo in this position :( >> >>The level of CometA90 is even worse in that position. >> >>palm tiger could not see it with white but Comet could not see it with black and >>finding it with black is easier by 1 ply. >> >>It may be interesting to test both programs to find how much time they need to >>avoid Rd1 and how much time they need to find c5 in order to see if Comet is >>better in tactics than palm tiger or the opposite. >> >>I believe that 120/40 was not enough but that they are going to find the right >>move after enough time. >> >>The position is not so easy for computers and I bet that a lot of amatuers are >>not going to find c5 on p200 at time control of 120/40. >> >>movei seems to be one of them. >> >>Uri > >I tested movei0.07a on p200(16 Mbytes hash) at this moment. > >The debug file gives the following information but the time management from >setup position seem to be broken. > >I do not think that movei could get depth 10 on p200 in a real tournament. > >I also believe that the speed gain of movei from p800 is bigger than the speed >gain of most chess program because I know that usually movei search at least 150 >kn/second on p800 when here it searched less than 25 kn/second on p200 >so it seems that the p800 help it to be 6 times faster than the p200. > >I am not sure because it is possible that my p200 was slowed down by some factor >because other applications were also open. > > >9 7 10666 2495681 d7d6 e5e6 f7f6 d4e3 c7c5 c2c4 b5c4 b3c4 d6d5 >9 7 13083 3104704 d7d6 e5e6 f7f6 d4e3 c7c5 c2c4 b5c4 b3c4 d6d5 >10 -6 32098 7759346 d7d6 e5e6 f7e6 b3e6 g8h8 b1c3 c7c5 d4e3 e8c7 c1d2 c7e6 e3e6 >10 -5 38250 9169704 c7c5 >10 43 64504 15363702 c7c5 d4d7 c5c4 b1c3 a8b8 d7g4 d8a5 d1d7 e7c5 b3c4 b5c4 g4c4 >10 43 66570 15892165 c7c5 d4d7 c5c4 b1c3 a8b8 d7g4 d8a5 d1d7 e7c5 b3c4 b5c4 g4c4 > >Uri Movei007e seems to be slightly better Here is analysis on p800 with the same size of hash tables 9 2 1637 2397965 d7d6 e5e6 f7f6 d4e3 c7c5 c2c4 b5c4 b3c4 d6d5 9 2 2058 3085976 d7d6 e5e6 f7f6 d4e3 c7c5 c2c4 b5c4 b3c4 d6d5 10 -6 5000 7837489 d7d6 e5e6 f7e6 b3e6 g8h8 b1c3 c7c5 d4e3 e8c7 c1d2 c7e6 e3e6 10 -5 5770 9124329 c7c5 10 43 7958 12459809 c7c5 d4d7 c5c4 b1c3 a8b8 d7g4 d8a5 d1d7 e7c5 b3c4 b5c4 g4c4 10 43 8252 12886138 c7c5 d4d7 c5c4 b1c3 a8b8 d7g4 d8a5 d1d7 e7c5 b3c4 b5c4 g4c4 11 93 12463 19395693 c7c5 d4d7 c5c4 b1c3 a8b8 d7g4 d8a5 d1d7 b7c8 g4d1 c8d7 d1d7 c4b3 d7e7 b3c2 c3b5 11 93 13643 21303324 c7c5 d4d7 c5c4 b1c3 a8b8 d7g4 d8a5 d1d7 b7c8 g4d1 c8d7 d1d7 c4b3 d7e7 b3c2 c3b5 12 120 23479 37409082 c7c5 d4d7 c5c4 b1c3 a8b8 b3c4 b5c4 c1f4 d8c7 a1c1 c7d7 d1d7 12 120 26030 41365800 c7c5 d4d7 c5c4 b1c3 a8b8 b3c4 b5c4 c1f4 d8c7 a1c1 c7d7 d1d7 Comet shows similiar speed on the same p800 in finding c5 New game - Deep Fritz r2qnrk1/1bppbppp/p7/1p2P3/3Q4/1B6/PPP2PPP/RNBR2K1 b - - 0 1 Analysis by Comet B31: 12...Bxg2 13.Kxg2 +- (3.06) Depth: 1/3 00:00:00 12...Bg5 = (0.22) Depth: 1/6 00:00:00 12...d6 = (0.13) Depth: 1/14 00:00:00 12...d6 13.Qc3 ² (0.40) Depth: 2/14 00:00:00 12...Bg5 13.Qb4 Bxc1 14.Rxc1 ² (0.36) Depth: 2/14 00:00:00 12...Bg5 13.Nc3 ² (0.70) Depth: 2/14 00:00:00 12...d6 13.Nc3 ² (0.61) Depth: 2/14 00:00:00 12...d6 13.Bf4 dxe5 14.Bxf7+ Rxf7 15.Qxd8 Bxd8 16.Rxd8 ² (0.48) Depth: 3/14 00:00:00 12...d6 13.Nc3 Qd7 ² (0.30) Depth: 3/14 00:00:00 12...d6 13.Nc3 h6 14.Bf4 dxe5 15.Qxd8 Bxd8 16.Bxe5 Bxg2 17.Rxd8 Rxd8 18.Kxg2 ² (0.50) Depth: 4/16 00:00:00 12...Bc8 ² (0.44) Depth: 4/16 00:00:00 12...Bc8 13.Qc3 Bb7 14.Bf4 ± (0.78) Depth: 4/16 00:00:00 12...d6 13.Nc3 Kh8 14.Bf4 dxe5 15.Bxe5 Qxd4 16.Bxd4 ± (0.71) Depth: 4/16 00:00:00 12...d6 13.Nc3 ± (0.71) Depth: 4/16 00:00:00 12...d6 13.Nc3 dxe5 14.Qxe5 Bd6 15.Nd5 ² (0.58) Depth: 5/18 00:00:00 25kN 12...d6 13.Nc3 c5 14.Qf4 c4 ² (0.32) Depth: 5/21 00:00:00 43kN 12...d6 13.Qc3 Qd7 14.f3 Qf5 15.exd6 cxd6 ² (0.35) Depth: 6/21 00:00:00 130kN 12...d6 13.Qc3 Qd7 14.f3 Qf5 15.Kf2 dxe5 16.Qxe5 = (0.22) Depth: 7/22 00:00:02 415kN 12...d6 13.Qg4 c5 14.Bd5 Bxd5 15.Rxd5 Nc7 16.Rd3 = (0.17) Depth: 7/23 00:00:04 827kN 12...d6 13.Qg4 c5 14.Bd5 Bxd5 15.Rxd5 f6 16.Qe6+ Rf7 17.Bf4 fxe5 18.Bxe5 = (0.19) Depth: 8/27 00:00:08 1592kN 12...d6 13.Qg4 c5 14.Bd5 Bxd5 15.Rxd5 f6 16.Qe6+ Rf7 17.Bf4 fxe5 18.Bxe5 = (0.19) Depth: 9/27 00:00:18 3670kN 12...d6 13.Qg4 c5 14.c3 Qc7 15.Bc2 Rd8 16.Qh3 h6 17.Qg4 ² (0.26) Depth: 10/33 00:00:49 10334kN 12...c5 13.Qxd7 c4 14.Nc3 Rb8 15.Qxe7 Qxe7 16.Nxb5 = (0.06) Depth: 10/33 00:00:57 12092kN 12...c5 13.Qxc5 Bxc5 14.Bxf7+ Rxf7 15.Rxd7 = (-0.21) Depth: 10/33 00:00:59 12574kN 12...c5 13.Qxc5 Bxc5 14.Bxf7+ Rxf7 15.Rxd7 ³ (-0.48) Depth: 10/33 00:01:01 13028kN 12...c5 13.Qxd7 c4 14.Nc3 Rb8 15.Qf5 Qa5 16.Be3 cxb3 17.axb3 ³ (-0.64) Depth: 10/33 00:01:03 13507kN 12...c5 13.Qxd7 c4 14.Nc3 Qc7 15.Qxc7 Nxc7 16.Nxb5 axb5 17.Be3 µ (-0.77) Depth: 11/33 00:01:22 18414kN 12...c5 13.Qxd7 c4 14.Nc3 Qc7 15.Qxc7 Nxc7 16.Rd7 Bxg2 µ (-0.88) Depth: 11/33 00:01:32 20644kN (Blass, Tel-aviv 22.05.2002) avoiding Rd1 is harder problem and movei may need fast hardware in order to do it at tournament time control I guess that better algorithms should also help and today hash tables are not used in the best way and movei does not extend moves like c5 that threats the queen. movei00_7e(p800 16 mbytes hash tables) 10 -2 6037 9424247 e1d1 d7d6 e5e6 f7f6 d4e3 c7c5 c2c4 b5c4 b3c4 d6d5 10 -2 19020 28874193 e1d1 d7d6 e5e6 f7f6 d4e3 c7c5 c2c4 b5c4 b3c4 d6d5 11 -43 26476 40853445 e1d1 c7c5 d4d7 c5c4 b1c3 a8b8 d7g4 d8a5 d1d7 e7c5 b3c4 b5c4 g4c4 11 -42 27879 43174194 c2c3 11 -19 41557 63839305 c2c3 c7c5 d4e3 d7d6 b1d2 d6e5 e3e5 e7d6 e5f5 e8f6 f2f4 11 -18 43283 66298135 c2c4 11 -2 51350 78213292 c2c4 c7c5 d4e3 b5c4 b3c4 d7d5 e1d1 e8c7 b1c3 d5d4 e3f4 11 -2 89815 139631458 c2c4 c7c5 d4e3 b5c4 b3c4 d7d5 e1d1 e8c7 b1c3 d5d4 e3f4 Comet seems to be clearly better here It does not avoid Rd1 for the right reason but it avoids Rd1 at depth 9 in 40 seconds and does not change it's mind at depth 10(I expect it also not to change it's mind at depth 11 because at depth 11 it should see it for the tactical reasons). New game - Deep Fritz r2qnrk1/1bppbppp/p7/1p2P3/3Q4/1B6/PPP2PPP/RNB1R1K1 w - - 0 1 Analysis by Comet B31: 12.a3 ² (0.34) Depth: 1/7 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Qd3 ± (0.85) Depth: 3/11 00:00:00 12.Nc3 h6 13.Be3 b4 ± (0.71) Depth: 4/14 00:00:00 12.Nc3 c5 13.Qd1 c4 14.a4 ² (0.43) Depth: 4/18 00:00:00 12.Na3 c5 13.Qg4 Bh4 ² (0.44) Depth: 4/18 00:00:00 12.Na3 c5 13.Qd2 c4 14.Qd6 = (0.15) Depth: 4/22 00:00:00 12.Nc3 c5 13.Qd3 c4 14.Qf5 = (0.17) Depth: 4/22 00:00:00 12.Bd5 c6 13.Bf3 Nc7 = (0.22) Depth: 4/22 00:00:00 12.Rd1 d6 13.Nc3 Qd7 ² (0.30) Depth: 4/22 00:00:00 12.Qd3 Kh8 13.Be3 Bb4 ² (0.35) Depth: 4/22 00:00:00 12.Qd3 c5 ² (0.48) Depth: 5/22 00:00:00 44kN 12.Qd3 c5 13.c4 Nc7 14.Nc3 ² (0.56) Depth: 5/22 00:00:00 50kN 12.Rd1 d6 13.Nc3 Qd7 14.Be3 ± (0.72) Depth: 5/22 00:00:00 57kN 12.Rd1 d6 13.Nc3 dxe5 14.Qxe5 Bd6 15.Nd5 ² (0.58) Depth: 6/22 00:00:00 90kN 12.Rd1 d6 13.Nc3 c5 14.Qf4 c4 ² (0.32) Depth: 6/22 00:00:01 202kN 12.Qd3 d5 13.Nc3 d4 14.Ne2 Bb4 ² (0.38) Depth: 6/22 00:00:01 263kN 12.Qd3 c5 13.c4 bxc4 14.Qxh7+ Kxh7 15.Bxc4 = (0.24) Depth: 7/23 00:00:02 401kN 12.Rd1 d6 13.Qc3 Qc8 14.f3 Qf5 15.exd6 cxd6 ² (0.35) Depth: 7/26 00:00:03 546kN 12.Rd1 d6 13.Qc3 Qc8 14.exd6 Nxd6 15.Qxg7+ Kxg7 16.Rxd6 = (0.21) Depth: 8/26 00:00:08 1346kN 12.Rd1 d6 13.Qg4 c5 14.Bd5 Bxd5 15.Rxd5 Nc7 16.Rd3 = (0.17) Depth: 8/26 00:00:15 2678kN 12.Rd1 d6 13.Qg4 c5 14.Bd5 Bxd5 15.Rxd5 f6 16.Qe6+ Rf7 17.Bf4 fxe5 18.Bxe5 = (0.19) Depth: 9/26 00:00:22 3933kN 12.Qg4 c5 13.c3 c4 14.Bc2 Ba3 ² (0.30) Depth: 9/30 00:00:40 7438kN 12.Qg4 c5 13.c3 d6 14.Nd2 dxe5 15.Qh3 Nf6 ² (0.35) Depth: 9/30 00:00:45 8266kN 12.Qg4 d5 13.c3 c5 14.Bf4 c4 15.Bc2 d4 16.Be4 Bxe4 17.Rxe4 ² (0.27) Depth: 10/30 00:01:31 17158kN 12.c4 c5 13.Qg4 d5 ² (0.35) Depth: 10/33 00:03:09 35926kN (Blass, Tel-aviv 22.05.2002) 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.