Testing Selective Search - Computer Chess Fun Olympiad 2010

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Testing Selective Search - Computer Chess Fun Olympiad 2

Post by Steve B »

Spacious_Mind wrote:
ps. a tidbit of knowledge. Did you know that Almeria was the first chess computer to beat a World Champion in a game? The World Champion being Karpov ?

regards

Nick
Wasnt it the Portorose that beat Karpov?

here is Jeroen Noomen operating the Portorose against Karpov in a simul in 1990 (Karpov was by then a former World Champion)
The Portorose won that game

Image

the historic game..

http://scacchi.qnet.it/manuale/karmep01.htm

Historical Regards
Steve
User avatar
Spacious_Mind
Posts: 317
Joined: Mon Nov 02, 2009 12:05 am
Location: Alabama

Re: Testing Selective Search - Computer Chess Fun Olympiad 2

Post by Spacious_Mind »

Steve B wrote:
Spacious_Mind wrote:
ps. a tidbit of knowledge. Did you know that Almeria was the first chess computer to beat a World Champion in a game? The World Champion being Karpov ?

regards

Nick
Wasnt it the Portorose that beat Karpov?

here is Jeroen Noomen operating the Portorose against Karpov in a simul in 1990 (Karpov was by then a former World Champion)
The Portorose won that game

Image

the historic game..

http://scacchi.qnet.it/manuale/karmep01.htm

Historical Regards
Steve
Now I have to search back in some of my chess books :) I was quite sure it was Almeria.

I'll be back regards :)

Nick
User avatar
Spacious_Mind
Posts: 317
Joined: Mon Nov 02, 2009 12:05 am
Location: Alabama

Re: Testing Selective Search - Computer Chess Fun Olympiad 2

Post by Spacious_Mind »

The final Round of matches in Group A is also now complete:

Image

Some good performances here for Team DOS/Windows. Warlord 1.0 deserves a special mention here, after having picked up some points against tough opponents in 4 of the 5 rounds.

Image

[Event "Chess Computer Olympiad 2010 - Group A"]
[Site "Hoover, Alabama, USA"]
[Date "2009.12.25"]
[Round "5.1"]
[White "Warlord, 5 Ply"]
[Black "Ikarus V0.18, 5 Ply"]
[Result "1-0"]
[ECO "D02"]
[WhiteElo "1800"]
[BlackElo "1800"]
[Annotator "Mind,Spacious"]
[PlyCount "119"]
[EventDate "2009.12.25"]
[EventType "team"]
[EventRounds "5"]
[EventCountry "USA"]
[EventCategory "16"]

1. d4 d5 2. Nd2 Bf5 3. Ngf3 e6 4. c4 Nc6 5. a3 Nf6 6. e3 Be7 7. cxd5 exd5 8. Bb5 O-O 9. Bxc6 bxc6 10. Ne5 Qd6 11. b4 a5 12. bxa5 Rxa5 13. a4 Rfa8 14. g4 Be6 15. O-O c5 16. Re1 cxd4 17. exd4 Qb6 18. Rb1 Bb4 19. g5 Nd7 20. Qb3 Rxa4 21. Qg3 Nxe5 22. Qxe5 Qa7 23. Qe3 Bxd2 24. Bxd2 Rxd4 25. Rb4 Rxb4 26. Bxb4

[d]r5k1/q1p2ppp/4b3/3p2P1/1B6/4Q3/5P1P/4R1K1 b - - 0 26

From this position it looks as if Ikarus 0.18 will win this game, but Ikarus starts to lose ideas and it's game deteriorates with a lot of irresolute Queen moves.

c6 27. Qc3 Qd7 28. Qg3 Rd8 29. Ra1 Bh3 30. Rc1 Qc8 31. Rc3 Bf5 32. Be7 Re8 33. Re3 Qb7 34. Bd6 Re6 35. h4 Qb1+ 36. Kh2 Rxe3 37. fxe3 Be4 38. Qf2 Qb7 39. Kg1 Qb6 40. Qf4 c5 41. Qe5 Qb1+ 42. Kf2 Qa2+ 43. Ke1 Qa5+ 44. Kf1 Qa4 45. Ke1 Qb4+ 46. Ke2 Qc4+ 47. Ke1 Qc1+ 48. Ke2 Qc2+ 49. Ke1 h6?

[d]6k1/5pp1/3B3p/2ppQ1P1/4b2P/4P3/2q5/4K3 w - - 0 50

Ikarus who has posted some good win's in this Tournament, in this game struggled to find the winning moves. Meanwhile, Warlord who had fought hard just to survive manages to place it's Bishop and Queen into a threatening position that seems to have disconcerted Ikarus. After 49. ... h6?, Ikarus's game now totally falls apart.

50. gxh6 Qc1+ 51. Ke2 Qc4+ 52. Kf2 Qa2+ 53. Ke1 Qa5+ 54. Kf1 Qb5+ 55. Kf2 gxh6 56. Qf6 Bg6?? 57. Be5 Kf8 58. Qd8+ Qe8 59. Bd6+ Kg7 60. Qxe8 {Ikarus V0.18 resigns} 1-0

Warlord is a nice looking and very capable DOS chess program written by William H. Rogers (a frequent flyer at Talk Chess :) ) in 2002.

Image

Neo Classic Engines needed a very big win in this match to have a chance for qualification to the next stage. This was certainly a big win again the Atari Software but will the 10 points scored in this match be enough?

Anaconda 2.0.1 has been playing extremely good chess at 5 Ply in this Tournament . Here is an important (and a little lucky) win against Atari Colossus Chess 4.1:

[Event "Chess Computer Olympiad 2010 - Group A"]
[Site "Hoover, Alabama, USA"]
[Date "2009.12.26"]
[Round "5.1"]
[White "Atari 800 - Colossus 4.1, 60/7542S."]
[Black "Anaconda 2.0.1, 5 Ply"]
[Result "0-1"]
[ECO "A45"]
[WhiteElo "1800"]
[BlackElo "1800"]
[Annotator "Mind,Spacious"]
[PlyCount "142"]
[EventDate "2009.12.26"]
[EventType "team"]
[EventRounds "5"]
[EventCountry "USA"]
[EventCategory "16"]

1. d4 d6 {Atari 800 - Colossus 4.1 out of book} 2. Bf4 {Anaconda 2.0.1 out of book} Nf6 3. Nc3 Nh5 4. Bg5 Bg4 5. h3 Bd7 6. e4 Nf6 7. Nf3 e6 8. Bc4 Nc6 9. O-O Be7 10. e5 Nd5 11. Bxd5 Bxg5 12. Nxg5 Qxg5 13. f4 Qd8
14. Bf3 O-O 15. d5 exd5 16. exd6 cxd6 17. Nxd5 Be6 18. c4 Re8 19. Qd3 Bxd5 20. Qxd5 Qf6 21. Qd2 Re6 22. Bd5 Re7 23. Rad1 Rae8 24. Bxc6 bxc6 25. Qxd6 Re6 26. Qd4 Qe7 27. b3 Qa3 28. Rf2 c5 29. Qd7 a6 30. Rfd2 h5 31. Kh2 Qa5 32. Qb7 Re4 33. Rd8 Rxd8 34. Rxd8+ Qxd8 35. Qxe4 h4 36. Qd5 Qa5 37. Qa8+ Kh7 38. Qe4+ g6 39. Qe7 Kg8 40. Qe8+ Kg7 41. Qe5+ Kg8 42. a4 Qb6 43. Qe3 Qd6 44. Qe8+ Kg7

[d]4Q3/5pk1/p2q2p1/2p5/P1P2P1p/1P5P/6PK/8 w - - 0 45

For 44 moves Colossus 4.1 outplays Anaconda 2.0.1 whose defence held making it very difficult for Colossus to find the final breakthrough. In this position Colossus 4.1 mistakenly thinks that exchanging the Queen's is the key to success. Afterall, a pawn ahead in the endgame. What could possibly go wrong? Right?

Wrong! White's King is badly placed and the exchange here is fatal.

45. Qe5+? Qxe5 46. fxe5 Kh6?

[d]8/5p2/p5pk/2p1P3/P1P4p/1P5P/6PK/8 w - - 0 47

Here Anaconda 2.0.1 makes a mistake that gave Colossus 4.1 one final opportunity to salvage the draw if 48. b4! is played instantly. Unfortunately Colossus 4.1 misses the move and loses. This is just one of the many examples of why this 5 Ply Tournament is so much fun to play. No, Brute Force long calculations here. The programs who instinctively find the right chess positions/moves invariably win the game. If compared to humans the programs playing at 5 Ply play approximately between 1700 and 2100 ELO, which makes them perfect Practice Partners for 99% of all humans.

47. a5? Kg5 48. Kg1 Kf4 49. Kf2 g5 50. Ke2 Ke4 51. Kd2
Kxe5 52. Ke3 f5 53. Kd3 f4 54. Ke2 Ke4 55. Kf2 Kd4 56. Kf3 Kc3 57. Kg4 Kxb3 58. Kxg5 Kxc4 59. Kg4 Kd3 60. Kxh4 c4 61. Kg5 c3 62. Kxf4 c2 63. Kf5 c1=Q 64. g4 Qf1+ 65. Kg6 Qxh3 66. g5 Ke4 67. Kg7 Qc3+ 68. Kh7 Qxa5 69. g6 Qh5+ 70. Kg7 Kf5 71. Kf8 Qxg6 0-1

Atari Mychess II also deserves a showcase after having played a very good tournament. Here is a game as White against Viper 1.0:

[Event "Chess Computer Olympiad 2010 - Group A"]
[Site "Hoover, Alabama, USA"]
[Date "2009.12.26"]
[Round "5.1"]
[White "Atari 800 - Mychess II, LV 7."]
[Black "Viper 0.1, 5 Ply"]
[Result "1-0"]
[ECO "B00"]
[WhiteElo "1800"]
[BlackElo "1800"]
[Annotator "Mind,Spacious"]
[PlyCount "145"]
[EventDate "2009.12.26"]
[EventType "team"]
[EventRounds "5"]
[EventCountry "USA"]
[EventCategory "16"]

1. e4 Nc6 2. d4 e5 3. dxe5 Nxe5 4. Nf3 {Viper 0.1 out of book} Nxf3+ {Atari 800 - Mychess II out of book} 5. Qxf3 Nf6 6. Bc4 Bc5 7. Nc3 O-O 8. Bg5 Bd4 9. O-O d6 10. Qd3 c5 11. Rad1 Bg4 12. Bxf6 Qxf6 13. Rd2 Rfe8?! 14. Nd5!

[d]r3r1k1/pp3ppp/3p1q2/2pN4/2BbP1b1/3Q4/PPPR1PPP/5RK1 b - - 0 14

Viper's last move 13. ... Rfe8?! was not a good choice because it gives Mychess II the opportunity to place the Knight on d5 giving White many threats. Viper's next move 14. ... Qh4? was also bad and is the root cause for this loss. 14. ... Qd8 would have been the best and safest choice. After this Viper almost clawed it's way back in to the game, but the damage had been done.

Qh4? 15. Nc7 Qh6 16. h3 Bd7 17. c3 Be5 18. Nxa8 Rxa8 19. Bd5 Rb8 20. Rdd1 b5 21. Qf3 Qg6 22. Qe3 b4 23. f4 Bf6 24. Bb3 Bb5 25. c4 Bc6 26. f5 Qg5 27. Qxg5 Bxg5 28. Rxd6 Bxe4 29. g4 Be3+ 30. Kh2 Bd4 31. Rd7 Be5+ 32. Kg1 Bd4+ 33. Rf2 a5 34. f6 gxf6 35. Ra7 Ra8 36. Rxa8+ Bxa8 37. Bc2 Bf3 38. Kf1 Bxf2 39. Kxf2 Bc6 40. h4 a4 41. Ke3 Bd7 42. Ke4 Bxg4 43. Kd5 b3 44. axb3 axb3 45. Bxb3 f5 46. Kxc5 f4 47. Kd4 f3 48. Ke3 Kg7 49. c5 Kf6 50. Bd5 f2 51. Kxf2 Be6 52. Bf3 Bd7 53. c6 Bh3
54. c7 Bc8 55. b4 Ke7 56. b5 Kd7 57. b6 Kd6 58. b7 Bxb7 59. Bxb7 Kxc7 60. Be4 h6 61. Ke3 Kd6 62. Kf4 Ke6 63. Bf5+ Kd6 64. Bg4 Ke7 65. Kf5 Kd6 66. Kf6 Kd5 67. Kxf7 Ke5 68. Kg6 Kf4 69. Kxh6 Kxg4 70. h5 Kf5 71. Kg7 Ke5 72. h6 Kd4 73. h7 {Viper 0.1 resigns} 1-0

Image

Team Mephisto wins the long awaited Battle of the Giants in this Group. The first four matches were so close with Mephisto narrowly leading 4.1/2 to 3.1/2. Everyone of these games could be posted here, they were all so good. But here are a few, starting with Ktulu 8 and Milano Pro. Milano Pro is looking to keep it's 100% record:

WHAT A FIGHT!

[Event "Chess Computer Olympiad 2010 - Group A"]
[Site "Hoover, Alabama, USA"]
[Date "2009.12.27"]
[Round "5.1"]
[White "Ktulu 8.0, 5 Ply"]
[Black "Mephsto Milano Pro, 10s"]
[Result "1/2-1/2"]
[ECO "A03"]
[WhiteElo "1800"]
[BlackElo "1800"]
[Annotator "Mind,Spacious"]
[PlyCount "308"]
[EventDate "2009.12.27"]
[EventType "team"]
[EventRounds "5"]
[EventCountry "USA"]
[EventCategory "16"]

1. f4 d5 2. c4 {Mephsto Milano Pro out of book} dxc4 3. e3 e5 4. Bxc4 exf4 5. exf4 Nc6 {Ktulu 8.0 out of book} 6. Nc3 Nf6 7. Qb3 Nd4 8. Qa4+ c6 9. Nge2 Bc5 10. b4 Bf5 11. Nxd4 Bxd4 12. Qa5 Qxa5 13. bxa5 Nd5 14. Bxd5 cxd5 15. Kd1 Kd7 16. Ba3 Bxc3 17. dxc3 Rhe8 18. Re1 Rxe1+ 19. Kxe1 Re8+ 20. Kd2 Kc6 21. Re1 Rxe1 22. Kxe1 Kb5 23. Bf8 g6

[d]5B2/pp3p1p/6p1/Pk1p1b2/5P2/2P5/P5PP/4K3 w - - 0 24

With bishops of different colors the game looks like a draw. But, because of Milano Pro's 100% record, I wanted the game to continue to the end.

24. Bb4 d4 25. Kd 2 Kc4 26. a3 d3 27. Ke3 b5 28. axb6 axb6 29. h3 b5 30. g4 Be6 31. Kd2 Bc8 32. Bd6 h5 33. gxh5 gxh5 34. h4 Kd5 35. Be5 Ke4 36. Bd4 Bf5 37. Be5 Kf3 38. Bd6 Kg4 39. Be7 Kxf4 40. Bc5 Ke4 41. Bd4 Be6 42. Bf6 Bb3 43. Ke1 Ke3 44. Bd4+ Kf4 45. Kd2 Ke4 46. Bf6 Be6 47. Kd1 Ke3 48. Bd4+ Kf4 49. Kd2 Bc4 50. Bf6 Kf5 51. Be7 Ke4 52. Bf6 Ba2 53. Kd1 Bd5 54. Kd2 Bc4 55. Kd1 Ba2 56. Kd2 Bd5 57. Kd1 Ke3 58. Bd4+ Kf4 59. Kd2 Be4 60. Bf6 Bf5 61. Bd4 Ke4 62. Bg7 Bg4 63. Bf6 Bh3 64. Ke1 Bc8 65. Kd2 Bd7 66. Kd1 Bf5 67. Kd2 Bg4 68. Ke1 Ke3 69. Bd4+ Kf4 70. Kd2 Be2 71. Bf6 Ke4 72. Ke1 Ke3 73. Bd4+ Kf4 74. Bf6 Ke4 75. Kd2 Bf3 76. Ke1 Bg2 77. Kd2 Bf1 78. Ke1 Bh3 79. Kd2 Bc8 80. Ke1 Ke3 81. Bd4+ Kf4 82. Kd2 Ba6 83. Kxd3 f5 84. Be3+ Kg3 85. Kd2 f4 86. Bc5 Kxh4 87. Ke1 Kg4 88. Kf2 h4 89. Bd4 Bb7 90. Kg1 Kh3 91. Kf2 Be4 92. Kg1 Bd5 93. Bf2 Bc4 94. Kh1 Be6 95. Kg1 Bd5 96. Bd4 Bf3 97. Bf2 Bb7 98. Bc5 Bc6 99. Bf2 Ba8 100. Bd4 Be4 101. Bf2 Bc6 102. Bd4 Ba8 103. Bb6 Bg2 104. Bd4 f3 105. Kf2 Kg4 106. Be5 h3 107. Ke3 Bf1 108. Bd6 Bc4 109. Be5 Kf5 110. Bd6 Bd5 111. Bc7 Bc6 112. Bd6 Bb7 113. Bg3 Be4 114. Bc7 Kg4 115. Be5 Bd5 116. Bc7 Kf5 117. Bd6 Kg5 118. Be5 Kg6 119. Bf4 Kh5 120. Bg3 Be4 121. Bc7 Kh4 122. Bf4 Bd5 123. Bd6 Bc6 124. Bf4 Kg4 125. Bc7 Bd5 126. Bd6 Bb7 127. Be5 Be4 128. Bc7 Kg5 129. Bg3 Bb7 130. Bd6 Kh4 131. Be5 Kg4 132. Bc7 Ba8 133. Be5 Kf5 134. Bc7 Be4 135. Bg3 Bb7 136. Bc7 Ba8 137. Bg3 Kg5 138. Bf4+ Kh5 139. Bd6 Kh4 140. Bc7 h2 141. Bxh2 Kh3 142. Bg1 Bd5 143. Kf2 Bb7 144. Kf1 Kg4 145. Bd4 Kf4 146. Kf2 Ke4 147. Kg3 Kd3 148. Kf2 Kc4 149. Ke3 Kb3 150. Be5 Kxa3 151. Bg3 Kb3 152. Kd4 Kc2 153. Kc5?!

[d]8/1b6/8/1pK5/8/2P2pB1/2k5/8 b - - 0 153

I could have ended the game, but when the King moved to c5 there it looked like Ktulu 8 after having played such great defense, might still slip up in the end. So I waited with great interest the next move from Milano Pro. Argh! it missed a possible win. 153. ... Ba6!? would have been very interesting because I had a feeling that Ktulu might not retreat the King. MP's 153. ... Kd3 is a draw. After the game I replayed the position with Ktulu 8 and as I thought after 153. ... Ba6 came 154. Kb4 Kd3 155. Ka5?? 156. Kxc3! and wins for MP. It would have been a shame to lose in this way after 153 moves :) therefore I am very happy that Ktulu got the draw.

Kd3? 154. Kxb5 Kxc3 1/2-1/2

This next match brings two 100% players to the board:

[Event "Chess Computer Olympiad 2010 - Group A"]
[Site "Hoover, Alabama, USA"]
[Date "2009.12.27"]
[Round "5.1"]
[White "Mephisto Dallas 68020, 16s"]
[Black "SmarThink 1.0, 5 Ply"]
[Result "1-0"]
[ECO "D14"]
[WhiteElo "1800"]
[BlackElo "1800"]
[Annotator "Mind,Spacious"]
[PlyCount "88"]
[EventDate "2009.12.27"]
[EventType "team"]
[EventRounds "5"]
[EventCountry "USA"]
[EventCategory "16"]

1. e4 e5 2. Nf3 Nc6 3. Bc4 h6 {Mephisto Dallas 68020 out of book} 4. d4 exd4 5. Nxd4 Qf6 6. Nxc6 {SmarThink 1.0 out of book} dxc6 7. O-O Be6 8. Bxe6 fxe6 9. Qh5+ g6 10. Qa5 Bd6 11. Nc3 Ne7 12. Be3 Qe5 13. Qxe5 Bxe5 14. Rad1 Kf7 15. f3 Rad8 16. Rxd8 Rxd8 17. Bxh6

[d]3r4/ppp1nk2/2p1p1pB/8/3bP3/2N2P2/PPP3PP/5RK1 w - - 0 18

Dallas wins a pawn. But this could be short lived if SmartThink moves the Rook back to h8. Unforturnately, SmartThink misses this move and let's Dallas keep the pawn advantage. Watch how nicely Dallas creates a Kingside attack to win the game.

Bd4+? 18. Kh1 Bxc3 19. bxc3 Kf6 20. h4 b5 21. Bg5+ Kf7 22. Kh2 c5 23. Kg3 c4 24. Ra1 c6 25. Kg4 Rd7 26. a4 Nc8 27. a5 Ne7 28. Rf1 Ng8 29. Be3 e5 30. Kg5 Nf6 31. g4 a6 32. h5 gxh5 33. gxh5 Ke6 34. h6 Rd8 35. Kh4 Rd7 36. Rg1 Rb7 37. Rg6 Rf7 38. Kg5 b4 39. cxb4 c3 40. Bc5 Kd7 41. Rxf6 Rh7 42. Kg6 Rxh6+ 43. Kxh6 Kc7 44. Bd6+ Kc8 {SmarThink 1.0 resigns} 1-0

The Final Table in Group A is as follows:

Image

Team Neo Classic beats Team Modern Engines to the final qualification spot. The 7:5 Head-To-Head win in their earlier round encounter clinched it.

Here are the final Team Performances in Group A:

Image

And, here are the Individual Performances in Group A:

Image

Best Regards

Nick
User avatar
Spacious_Mind
Posts: 317
Joined: Mon Nov 02, 2009 12:05 am
Location: Alabama

Re: Testing Selective Search - Computer Chess Fun Olympiad

Post by Spacious_Mind »

Group B in the Selective Search 5 Ply Olympiad has started.

The Teams for Group B are as follows:

Image

Here is the schedule for Round 1

Image

I will update the above chart with the exact times/levels used by the Dedicated Chess Computers and Atari ST Programs once I get to their games.

Best regards

Nick
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Testing Selective Search - Computer Chess Fun Olympiad

Post by Steve B »

A quick question regarding the Risc 2500
how many MB?
probably dosent make much of a difference with fast time controls but im just curious

i have a match coming up this Saturday in the Play Chess Engine Room where i will be operating the Scisys Mark VI(Philidor) against another player who will be operating an over clocked Novag Star Opal(24 Mhz)
TC ..30 Sec/Avg
best of 3
any predictions?

2-0 for the Mark VI Regards
Steve
User avatar
Spacious_Mind
Posts: 317
Joined: Mon Nov 02, 2009 12:05 am
Location: Alabama

Re: Testing Selective Search - Computer Chess Fun Olympiad

Post by Spacious_Mind »

Steve B wrote:A quick question regarding the Risc 2500
how many MB?
probably dosent make much of a difference with fast time controls but im just curious

i have a match coming up this Saturday in the Play Chess Engine Room where i will be operating the Scisys Mark VI(Philidor) against another player who will be operating an over clocked Novag Star Opal(24 Mhz)
TC ..30 Sec/Avg
best of 3
any predictions?

2-0 for the Mark VI Regards
Steve
Happy New Year !

My Risc has 2 MB RAM. I am not sure if the 2 MB RAM makes much difference. I have not detected any difference to other 512K Risc games often seen at Info, but I am sure there has to be an improvement somewhere, especially in endgames. Nice thing about the Risc since it runs at 14 MHz is that it will get to play at the same time setting as Dallas which should make it quite interesting.

Mark VI are rare (Steve's Trademark = RARE :P ) and I don't have one. So it is difficult for me to know how much better if at all it is compared to Mark V. I know a Mark V is only slightly better than an Opal, Star Opal, Opal II they all seem to be about the same strength. Mark V can lose to Opal. Therefore with your Mark VI anything could happen in that match. The Software in the Opal is quite weak therefore I don't know how much improvement there would be with the tripled speed increase but I dont think that it would be too much. I don't know how much search improvement you can get using 768 bytes of RAM? I doubt it will improve more than 100 ELO.. Therefore ELO 1450-1550 now becomes 1550 - 1650. Compared to you at 1550 - 1600?

Ok since both are unknowns for me, I will take a guess 2-1 to Star Opal (because of it's massive 24 MHZ). :P

I would actually even though this might sound funny (2 MHz vs. 24 MHz)favor you in a longer game like 40/2 Hrs.

I look forward to hearing about the outcome of your match!

I have never tried visitng Playchess.

Best regards

Nick
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Testing Selective Search - Computer Chess Fun Olympiad

Post by Steve B »

Happy New Year To You And Yours Nick

2MB Risc 2500..nice computer!

the Mark VI is rated only slightly better then the Mark V
Selective Search shows it at 1428 and the Mark V at 1419 ..
not even 10 Elo better

when it was first released i recall that the consensus was about 25 Elo but i dont recall if that was based on any significant number of games

Either way it should be an interesting live match

Best Regards
Steve
User avatar
Spacious_Mind
Posts: 317
Joined: Mon Nov 02, 2009 12:05 am
Location: Alabama

Re: Testing Selective Search - Computer Chess Fun Olympiad

Post by Spacious_Mind »

Steve B wrote:Happy New Year To You And Yours Nick

2MB Risc 2500..nice computer!

the Mark VI is rated only slightly better then the Mark V
Selective Search shows it at 1428 and the Mark V at 1419 ..
not even 10 Elo better

when it was first released i recall that the consensus was about 25 Elo but i dont recall if that was based on any significant number of games

Either way it should be an interesting live match

Best Regards
Steve
Hi Steve,

Yes I know Info shows it at Active Level at around 1350 but I think in comparison on the chart the low end computers are much too low or the ones above them are too high. That's why I tend to think based on Info that the Mark V is above 1500 plus your improvement for Mach VI makes me think in comparison 1550. But thats all irrelevant I think you should have a pretty close match.

The problem with these lists is that the Top End Computers like Resurrection and Co drag the ratings too far down at the bottom end and not enough games are played between bottom and middle.

My Cosmic for example beats Mephisto III so can an Excalibur Chess Station yet the gap shown is always much larger. Almost like some computers get favoritism.

Good Luck!

Regards

Nick
User avatar
Spacious_Mind
Posts: 317
Joined: Mon Nov 02, 2009 12:05 am
Location: Alabama

Re: Testing Selective Search - Computer Chess Fun Olympiad

Post by Spacious_Mind »

I thought you might find this interesting:

Here is a Tournament where everyone played 50 games. 650 in total:

Image

See how well Morphy did :)

Now here is my very same Morphy in another Tournament (U1500 tournament) in which Mach V played with exactly the same game settings:

http://www.schach-computer.info/wiki/in ... rnier_2009

A huge difference in performance :)

Best regards

Nick
Steve B
Posts: 3697
Joined: Tue Jul 31, 2007 4:26 pm

Re: Testing Selective Search - Computer Chess Fun Olympiad

Post by Steve B »

Spacious_Mind wrote:
See how well Morphy did :)

Now here is my very same Morphy in another Tournament (U1500 tournament) in which Mach V played with exactly the same game settings:

http://www.schach-computer.info/wiki/in ... rnier_2009

A huge difference in performance :)

Best regards
Nick
Yes Huge difference
:P

With a few Hrs. to go to game time i put the Mark VI(Philidor) through its paces against a Chess Master 9000 personality
Liz ..rated 1422
I am also using the Sensor board attachment

Phil has White
TC..30 Seconds Avg

[Event "Training Game For Big Match"]
[Site "Here "]
[Date "2010.1.2"]
[Round "1"]
[White "Scisys Mark VI(Phil)"]
[Black "Liz"]
[Time Control "30 Sec Avg?"]
[Result "1-0"]

1.e4 e6 2.d4 d5 3.Nd2 Nf6 4.e5 Nfd7 5.Bd3 c5 6.c3 Nc6 7.Ne2 cxd4 8.cxd4
f6 9.Nf4 Nxd4 10.Qh5+ Ke7 11.Ng6+ hxg6 12.exf6+ Nxf6 13.Qxh8 Qb6 14.O-O
e5 15.Bxg6 Ne2+ 16.Kh1 Ng4 17.Qh4+ Qf6 18.Qxf6+ gxf6 19.f3 Bf5 20.Bxf5
Ne3 21.Re1 Nxf5 22.Rxe2 Rc8 23.g4 Nd4 24.Rg2 a5 25.a4 Rc2 26.b3 Ke6
27.Rb1 Kf7 28.h4 Bb4 29.h5 Kg7 30.Rh2 Kh8 31.Rg2 Kg7 32.Rh2 Kh8 33.Nf1
Rc6 34.f4 Nf3 35.Rf2 Rc3 36.fxe5 Nxe5 37.Rxf6 Nxg4 38.Rf4 Rh3+ 39.Kg2
Rh4 40.Bb2+ Kg8 41.Kg3 Rxh5 42.Kxg4 Rh6 43.Rd1 Be7 44.Rxd5 Rh7 45.Rxa5
Rh4+ 46.Kf3 Rxf4+ 47.Kxf4 Kf7 48.Rb5 b6 49.Rxb6 Bc5 50.Rf6+ Ke7 51.Kg5
Ke8 52.Ra6 Kd8 53.Bf6+ Kc7 54.Be5+ Kb7 55.Re6 Bb4 56.Kf5 Kc8 57.Rb6 Ba5
58.Rb5 Kd7 59.Rxa5 Kc6 60.Ke6 Kb6 61.Rd5 Ka7 62.Bd4+ Kb7 63.Kd7 Ka6
64.Nd2 Kb7 65.a5 Ka6 66.Rd6+ Kb7 67.a6+ Ka8


[d] k7/3K4/P2R4/8/3B4/1P6/3N4/8 w - - 0 68
Here Phil announces a Mate in 2

68.Rh6 Kb8 69.Rh8# 1-0

As we can see Phil made mincemeat out of Lizzy
based upon this game i am comfortable guesstimating Phil to be about 2700+ Elo
:P

Regards From Phil and Lizzy:
Image

Steve