Author: Uri Blass
Date: 11:58:28 03/08/04
Go up one level in this thread
On March 08, 2004 at 14:48:16, Uri Blass wrote:
>On March 08, 2004 at 14:15:55, Kurt Utzinger wrote:
>
>>On March 08, 2004 at 12:58:58, Uri Blass wrote:
>>
>>>On March 08, 2004 at 12:26:38, Jorge Pichard wrote:
>>>
>>>>On March 08, 2004 at 11:06:44, Uri Blass wrote:
>>>>
>>>>>On March 08, 2004 at 10:27:00, Jorge Pichard wrote:
>>>>>
>>>>>>Arena tournament
>>>>>>
>>>>>>Rank Engine Score Mo Go S-B
>>>>>>1 Gothmog 11.5/16 · · · · ·1=1001=1=0111111. . 51.75
>>>>>>2 Movei00_8_178 4.5/16 0=0110=0=1000000· · · · · · · 51.75
>>>>>>
>>>>>>
>>>>>>16 of 16 games played
>>>>>>Tournament start: 2004.03.07, 20:25:23
>>>>>>Latest update: 2004.03.08, 14:20:00
>>>>>>Site/ Country: S0025128408, United States
>>>>>>Level: Blitz 15/0
>>>>>>Hardware: AMD Athlon(tm) Processor 1202 MHz
>>>>>>Operating system: Microsoft Windows 2000 Build 2195
>>>>>>PGN-File: C:\Program Files\Arena\Arena.pgn
>>>>>>Table created with: Arena 1.05
>>>>>
>>>>>
>>>>>Thanks
>>>>>
>>>>>This result cause me to suspect that there is some bug that cause 178 to perform
>>>>>weaker.
>>>>>
>>>>>In games against itself I also got less than 50% against 174 but the result is
>>>>>not statistically significant.
>>>>>
>>>>>note that 174 was tested in games before release when 178 was not tested in
>>>>>games and I only tested it in some positions to verify that I get the same
>>>>>number of nodes slightly faster.
>>>>>
>>>>>Gothmog is not a weak program but the result is worse than what I expected.
>>>>>
>>>>>Uri
>>>>
>>>>
>>>>http://www.talkchess.com/forums/1/message.html?353243
>>>
>>>There was some difference and in the first match movei used original book.
>>>
>>>The results that I got myself in the nunn match is 104:96 for the old version in
>>>4 nunn matches.
>>>
>>>It was
>>>29.5-20.5 for 8_174 at 1/2 minutes per game
>>>26.5-23.5 for 8_174 at 1 minute per game
>>>21.5-28.5 against 8_174 at 2 minutes per game
>>>26.5-23.5 for 8_174 at 4 minutes per game
>>>
>>>
>>>Uri
>>
>>
>> It seems to be very common that most chess programs are tested
>> at very very low time controls (I do not at all like) and this
>> is a pity and does not encourage me to try such programs at
>> longer time controls -:)
>> Kurt
>
>8_174 was tested at longer time control and the only difference between 8_178
>and 8_174 is probably only a small speed improvement.
>
>No change in the evaluation and no change in the search rules unless I have a
>bug.
>
>The only reason that I released 8_178 is that I did not like the fact that 8_174
>generates logfile as default that it is impossible to unable and when I asked if
>to release a new version people asked me to do it.
>
>I did some small changes between 8_174 and 8_178 but no change is in the
>evaluation or in the search.
>
>I now tested 174 against 178 in 4 of the nunn match(15 minutes per game) and
>here are the games(the first 2 are the same and in the second one 178 could
>search slightly deeper at move 29 so it chose a different move).
>
>I hope that the bad results of 174 is only a statistical illusion but I never
>can be sure based on the rule that every small change can do the program
>significantly weaker.
>
>
>[Event "Computer chess game"]
>[Site "URI-PC"]
>[Date "2004.03.08"]
>[Round "1"]
>[White "movei00_8_174"]
>[Black "Movei00_8_178"]
>[Result "1/2-1/2"]
>[TimeControl "900"]
>
>1. c4 c5 2. Nf3 Nf6 3. Nc3 e6 4. g3 b6 5. Bg2 Bb7 6. O-O Be7 7. d4 cxd4 8.
>Qxd4 d6 9. Rd1 a6 10. b3 Nbd7 11. e4 Qc7 12. Bb2 O-O 13. Qe3 Ng4 14. Qe2
>Nde5 15. Nd4 Nf6 16. Rac1 Rab8 17. Ba3 Rbd8 18. Nc2 Ned7 19. f4 e5 20. Ne3
>exf4 21. Nf5 Ne5 22. gxf4 Ng6 23. Qd2 Ne8 24. Nd5 Bxd5 25. cxd5 Qb7 26. Rc6
>b5 27. Bb4 Ra8 28. e5 a5 29. Nxe7+ Qxe7 30. exd6 Qd7 31. Bc5 a4 32. Re1
>axb3 33. axb3 Nf6 34. Rf1 Qg4 35. h3 Qg3 36. Qe3 Nh5 37. f5 Nh4 38. Qxg3
>Nxg3 39. d7 Rfd8 40. Be7 Rxd7 41. Bxh4 Nxf1 42. Kxf1 f6 43. Rc5 Ra1+ 44.
>Kf2 Ra2+ 45. Kg1 Re2 46. Rxb5 Re5 47. Bf3 Kf8 48. Rb8+ Ke7 49. Rg8 Kf7 50.
>Rh8 Rxf5 51. Kg2 Ra7 52. Rxh7 Ra2+ 53. Kg3 Kg8 54. Rh5 Rxh5 55. Bxh5 g5 56.
>Bxg5 fxg5 57. Kg4 Rd2 58. Kxg5 Rxd5+ 59. Kg6 Rd3 60. Bg4 Kf8 61. h4 Rxb3
>62. Bf5 Ke7 63. Kg5 Rg3+ 64. Kf4 Rg2 65. h5 Kf6 66. h6 Rf2+ 67. Kg4 Rxf5
>68. h7 Rg5+ 69. Kh4 Rh5+ 70. Kxh5 Kg7 71. h8=Q+ Kxh8
>{Movei->Draw by insufficient material} 1/2-1/2
>
>[Event "Computer chess game"]
>[Site "URI-PC"]
>[Date "2004.03.08"]
>[Round "2"]
>[White "Movei00_8_178"]
>[Black "movei00_8_174"]
>[Result "1/2-1/2"]
>[TimeControl "900"]
>
>1. c4 c5 2. Nf3 Nf6 3. Nc3 e6 4. g3 b6 5. Bg2 Bb7 6. O-O Be7 7. d4 cxd4 8.
>Qxd4 d6 9. Rd1 a6 10. b3 Nbd7 11. e4 Qc7 12. Bb2 O-O 13. Qe3 Ng4 14. Qe2
>Nde5 15. Nd4 Nf6 16. Rac1 Rab8 17. Ba3 Rbd8 18. Nc2 Ned7 19. f4 e5 20. Ne3
>exf4 21. Nf5 Ne5 22. gxf4 Ng6 23. Qd2 Ne8 24. Nd5 Bxd5 25. cxd5 Qb7 26. Rc6
>b5 27. Bb4 Ra8 28. e5 a5 29. Nxe7+ Qxe7 30. exd6 Qd7 31. Bc5 a4 32. Re1
>axb3 33. axb3 Nf6 34. Rf1 Qg4 35. h3 Qg3 36. Qe3 Nh5 37. f5 Nh4 38. Qxg3
>Nxg3 39. d7 Rfd8 40. Be7 Rxd7 41. Bxh4 Nxf1 42. Kxf1 f6 43. Rc5 Ra1+ 44.
>Kf2 Ra2+ 45. Kg1 Re2 46. Rxb5 Re5 47. Bf3 Kf8 48. Rb8+ Ke7 49. Rg8 Kf7 50.
>Rh8 Rxf5 51. Kg2 Ra7 52. Rxh7 Ra2+ 53. Kg3 Kg8 54. Rh5 Rxh5 55. Bxh5 g5 56.
>Bxg5 fxg5 57. Kg4 Rd2 58. Kxg5 Rxd5+ 59. Kg6 Rd3 60. Bg4 Kf8 61. h4 Rxb3
>62. Bf5 Ke7 63. Kg5 Rg3+ 64. Kf4 Rg2 65. h5 Kf6 66. h6 Rf2+ 67. Kg4 Rxf5
>68. h7 Rg5+ 69. Kh4 Rh5+ 70. Kxh5 Kg7 71. h8=Q+ Kxh8
>{Movei->Draw by insufficient material} 1/2-1/2
>
>[Event "Computer chess game"]
>[Site "URI-PC"]
>[Date "2004.03.08"]
>[Round "1"]
>[White "movei00_8_174"]
>[Black "Movei00_8_178"]
>[Result "1-0"]
>[TimeControl "900"]
>
>1. d4 Nf6 2. c4 e6 3. Nc3 c5 4. d5 exd5 5. cxd5 d6 6. Nf3 g6 7. e4 Bg7 8.
>Be2 O-O 9. O-O Re8 10. Nd2 Nbd7 11. Qb3 Qe7 12. a4 b6 13. Re1 Ne5 14. h3 g5
>15. Nf1 h6 16. Bd2 Rd8 17. Ng3 Bd7 18. Rac1 Ng6 19. Bb5 a6 20. Bxd7 Qxd7
>21. Nf5 Ne5 22. Ra1 Re8 23. f4 gxf4 24. Bxf4 h5 25. Bxe5 Rxe5 26. Qxb6 Bf8
>27. a5 Kh8 28. Ne3 Rg5 29. Kh1 Bg7 30. Ra3 h4 31. Nc4 Ne8 32. e5 dxe5 33.
>Qxc5 Rd8 34. Nb6 Qc7 35. Qf2 Rh5 36. Ra4 Nd6 37. Rxh4 Rxh4 38. Qxh4+ Kg8
>39. Qg3 Qe7 40. Qd3 Qh4 41. Rf1 e4 42. Qxa6 Qg3 43. Ne2 Qb3 44. Qa7 Qxb2
>45. Qe7 Rf8 46. Ng3 Qe5 47. Qh4 Qe8 48. Nh5 f5 49. Nxg7 Kxg7 50. a6 Qb5 51.
>Qg5+ Kf7 52. Nc4 Qxa6 53. Ne5+ Ke8 54. Qg6+ Kd8 55. Rb1 Re8 56. Nc6+ Qxc6
>57. dxc6 Kc7 58. Qg7+ Kxc6 59. Rc1+ Kd5 60. Rd1+ Kc6 61. Qf6 Rd8 62. Qxd8
>Nb5 63. Qd5+ Kb6 64. Rb1 Kc7 65. Rxb5 Kc8 66. Rb7 f4 67. Qc6+ Kd8 68. Rd7+
>Ke8 69. Rh7+ Kf8 70. Qa8#
>{Movei->White mates} 1-0
>
>[Event "Computer chess game"]
>[Site "URI-PC"]
>[Date "2004.03.08"]
>[Round "2"]
>[White "Movei00_8_178"]
>[Black "movei00_8_174"]
>[Result "1-0"]
>[TimeControl "900"]
>
>1. d4 Nf6 2. c4 e6 3. Nc3 c5 4. d5 exd5 5. cxd5 d6 6. Nf3 g6 7. e4 Bg7 8.
>Be2 O-O 9. O-O Re8 10. Nd2 Nbd7 11. Qb3 Qe7 12. a4 b6 13. Re1 Ne5 14. h3 g5
>15. Nf1 h6 16. Bd2 Rd8 17. Ng3 Bd7 18. Rac1 Ng6 19. Bb5 a6 20. Bxd7 Qxd7
>21. Nf5 Ne5 22. Ra1 Re8 23. f4 gxf4 24. Bxf4 h5 25. Bxe5 Rxe5 26. Qxb6 Bf8
>27. a5 Kh8 28. Ne3 Rg5 29. Re2 Qxh3 30. Qb7 Qc8 31. Qxf7 Bg7 32. Nc4 Ne8
>33. e5 Qg4 34. Ne3 Qg3 35. Ne4 Qxe5 36. Nxg5 Qxg5 37. Rf1 Nf6 38. Nf5 Rf8
>39. Qe7 Nh7 40. Nxg7 Qxg7 41. Qxd6 Qd4+ 42. Rff2 Rf6 43. Qe7 Rf5 44. d6 Nf8
>45. Rd2 Qg4 46. Rxf5 Qxf5 47. d7 Nxd7 48. Rxd7 Qxd7 49. Qxd7 c4 50. Qc8+
>Kg7 51. Qxa6 c3 52. bxc3 Kf7 53. Qd6 h4 54. a6 h3 55. a7 h2+ 56. Kxh2 Kg7
>57. a8=Q Kh7 58. Qb7+ Kh8 59. Qdb8#
>{Movei->White mates} 1-0
>
>
>
>Uri
Note also that there is another way to test if there are bugs in new movei
1)run movei as a console application
2)Write to the screen
help
3)after getting some list of commands write
new
4)write
auto
5)choose a number(for example 4)
The program is going to play a game against itself and count the nodes.
At the end of the game it is going to print number of total nodes
If the number of nodes of 174 is different then the number of nodes in 178 then
it is clear that I have some bug.
In this case both 174 and 178 give 622770 nodes when 16 mbytes are used.
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.