What is Komodo Average rating at Knight odds Vs Humans

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

lkaufman
Posts: 6279
Joined: Sun Jan 10, 2010 6:15 am
Location: Maryland USA
Full name: Larry Kaufman

Re: What is Komodo Average rating at Knight odds Vs Humans

Post by lkaufman »

Chessqueen wrote: Sat Jan 22, 2022 5:55 pm A good Solid 2400 human player that play like Komodo Dragon when I set the UCI_Elo = 2400 should NOT lose a game against any strong engine like Stockfish 14,1 NNUE, but there are NOT too many solid strong Humans Fide players that play like Komodo Dragon 2.6.1 set up to Limit Playing strength = 2400


[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.01.22"]
[Round "?"]
[White "Stockfish_14.1_win_x64_avx2"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "0-1"]
[BlackElo "2400"]
[Time "11:10:18"]
[WhiteElo "3500"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKB1R w KQkq - 0 1"]
[Termination "unterminated"]
[PlyCount "1"]
[Termination "normal"]
[PlyCount "60"]
[WhiteType "program"]
[BlackType "program"]

1. d4 d5 2. c4 e6 3. cxd5 exd5 4. e4 dxe4 5. Bc4 Nc6 6. Nc3 Qxd4 7. Qb3 Ne5
8. Be2 Nf6 9. O-O c6 10. Bg5 Qb4 11. Qxb4 Bxb4 12. Bxf6 gxf6 13. Nxe4 f5
14. Nf6+ Ke7 15. Nh5 Rd8 16. Rad1 Be6 17. b3 Bd2 18. Ng7 f4 19. Nxe6 Kxe6
20. Rb1 Rd4 21. Rfd1 Rad8 22. Bf1 Kf5 23. Rb2 h5 24. Rc2 R8d5 25. Be2 h4
26. h3 a5 27. a3 b5 28. Kh1 Ke4 29. Rb1 Kf5 30. Kg1 f3 31. gxf3 Kf4 32. b4
axb4 33. axb4 Bxb4 34. Kg2 Bd2 35. Rb3 b4 36. Ba6 c5 37. Rb1 Nxf3 38. Bb7
Rg5 {White resigns} 0-1[/pgn]
Dragon Elo needs to be set to about 2700 to be competitive with full strength Dragon 2.6.1 at knight odds. If 2400 elo can beat SF 14.1 that way, I think it just shows the same thing as your other SF games, namely that SF seems to "think" that it is good to trade queens when down a piece! A human IM would crush SF in a knight odds rapid match if it keeps trading queens.
Komodo rules!
Chessqueen
Posts: 5685
Joined: Wed Sep 05, 2018 2:16 am
Location: Moving
Full name: Jorge Picado

Re: What is Komodo Average rating at Knight odds Vs Humans

Post by Chessqueen »

Actually Stockfish14.1 had a draw against Komodo Dragon 2.6.1 UCI_Elo = 2350. at T/C 15 Minutes 10 Sec Bonus, but this time it did NOT traded Queens earlier :roll:

[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.01.22"]
[Round "?"]
[White "Stockfish_14.1_win_x64_avx2"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "1/2 1/2"]
[BlackElo "2350"]
[Time "11:51:16"]
[WhiteElo "3500"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKB1R w KQkq - 0 1"]
[Termination "normal"]
[PlyCount "70"]
[WhiteType "program"]
[BlackType "program"]

1. e4 d5 2. exd5 c6 3. dxc6 Nxc6 4. Nc3 e6 5. d4 Nf6 6. Be3 Bb4 7. a3 Bxc3+
8. bxc3 O-O 9. Bd3 Re8 10. O-O e5 11. Rb1 Bg4 12. f3 Bh5 13. Rxb7 e4 14.
Be2 Nd5 15. Bd2 e3 16. Be1 a6 17. c4 Nf4 18. c3 Qa5 19. d5 Ne5 20. Bg3
Nxe2+ 21. Qxe2 Qxa3 22. Bxe5 Rxe5 23. g4 Bg6 24. f4 Re4 25. f5 Rf8 26. fxg6
hxg6 27. Qf3 f5 28. gxf5 gxf5 29. Kh1 Qxc3 30. Rg1 Rg4 31. d6 Qxc4 32. Re1
Rf4 33. Rg1 Rg4 34. Ra1 Rf4 35. Rg1 Rg4 {3-fold repetition} *[/pgn]
Chessqueen
Posts: 5685
Joined: Wed Sep 05, 2018 2:16 am
Location: Moving
Full name: Jorge Picado

Re: What is Komodo Average rating at Knight odds Vs Humans

Post by Chessqueen »

Here is another draw Stockfish 14.1 versus Komodo Dragon 2.6.1 Limit Playing Strength Elo = 2350, so it is NOT just trading Queens the main problem, it has to be the way Komodo Dragon plays that makes it difficult for Stockfish full strength to beat a 2350 FIDE rated or estimated to equal a human 2350 rated opponent at Knight Odds.

Note: I wonder if Komodo Dragon 2.6.1 set up to UCI_ Elo = 2350 can draw against Full strength Komodo Dragon 2.6.1 using the same opening up to move 10 of this game.

[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.01.22"]
[Round "20"]
[White "Stockfish_14.1_win_x64_avx2"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "1/2 1/2"]
[BlackElo "2350"]
[WhiteElo "3500"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKB1R w KQkq - 0 1"]
[Termination "adjudication"]
[PlyCount "199"]
[WhiteType "program"]
[BlackType "program"]

1. e4 d5 2. exd5 c6 3. dxc6 Nxc6 4. Nc3 e6 5. d4 Nf6 6. Be3 Bb4 7. a3 Bxc3+
8. bxc3 O-O 9. Bd3 e5 10. O-O Bg4 11. f3 Be6 12. Rb1 Nd5 13. Bd2 Nf6 14.
Rxb7 exd4 15. Re1 Qc8 16. Rb5 Qd8 17. Qa1 Rc8 18. Kh1 Re8 19. Rbb1 Re7 20.
a4 dxc3 21. Bxc3 Ne8 22. a5 Rd7 23. Ba6 Rcc7 24. Bd3 Rd5 25. Bc4 Rd6 26.
Bd3 Rdd7 27. a6 Nf6 28. Rb5 Rc8 29. Rb7 Rd5 30. h3 Rd6 31. Kg1 h6 32. f4
Bd7 33. Be5 Rd5 34. Bc4 Ra5 35. Qd1 Nxe5 36. fxe5 Rxc4 37. exf6 Qxf6 38.
Qxd7 Rf4 39. Rxa7 Rf2 40. Qd3 Rff5 41. Ra8+ Kh7 42. Rf1 g6 43. Rxf5 Rxf5
44. c3 Ra5 45. Kh2 Ra1 46. Qd4 Rxa6 47. Rxa6 Qxa6 48. c4 Kg8 49. c5 Qe6 50.
Kg3 Qe1+ 51. Kf3 Qf1+ 52. Qf2 Qd3+ 53. Qe3 Qd5+ 54. Kg3 h5 55. Qc3 Qc6 56.
Kf4 f6 57. h4 Kf7 58. Kg3 Ke6 59. Qc2 Ke7 60. Qc4 Qe6 61. Qc3 Qg4+ 62. Kf2
Qxh4+ 63. Kf1 Qe4 64. Kg1 Qc6 65. Qe1+ Kd8 66. Qd2+ Kc8 67. Qe3 g5 68. Qe7
Kb8 69. Kf1 h4 70. Kg1 Qd5 71. Qe3 Qe5 72. Qb3+ Kc8 73. Kh1 Kd8 74. Qb4 Qd5
75. c6 Qxc6 76. Qf8+ Kc7 77. Qf7+ Kd8 78. Qf8+ Kc7 79. Qf7+ Qd7 80. Qxf6
Qd1+ 81. Kh2 Qd6+ 82. Qxd6+ Kxd6 83. g3 hxg3+ 84. Kxg3 Ke5 85. Kg4 Kf6 86.
Kf3 Kf7 87. Kg2 Kg6 88. Kf3 Kh5 89. Kg3 g4 90. Kh2 Kg5 91. Kg3 Kf5 92. Kg2
Kf6 93. Kh2 Kg5 94. Kg1 Kh5 95. Kg2 Kg5 96. Kh2 {3-fold repetition} Kf6 97.
Kg3 Kg5 98. Kh2 {3-fold repetition} Kf6 99. Kg3 Kg5 100. Kf2 {User
Adjudication} *[/pgn]
Chessqueen
Posts: 5685
Joined: Wed Sep 05, 2018 2:16 am
Location: Moving
Full name: Jorge Picado

Re: What is Komodo Average rating at Knight odds Vs Humans

Post by Chessqueen »

lkaufman wrote: Sat Jan 22, 2022 7:28 pm
Chessqueen wrote: Sat Jan 22, 2022 5:55 pm A good Solid 2400 human player that play like Komodo Dragon when I set the UCI_Elo = 2400 should NOT lose a game against any strong engine like Stockfish 14,1 NNUE, but there are NOT too many solid strong Humans Fide players that play like Komodo Dragon 2.6.1 set up to Limit Playing strength = 2400


[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.01.22"]
[Round "?"]
[White "Stockfish_14.1_win_x64_avx2"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "0-1"]
[BlackElo "2400"]
[Time "11:10:18"]
[WhiteElo "3500"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKB1R w KQkq - 0 1"]
[Termination "unterminated"]
[PlyCount "1"]
[Termination "normal"]
[PlyCount "60"]
[WhiteType "program"]
[BlackType "program"]

1. d4 d5 2. c4 e6 3. cxd5 exd5 4. e4 dxe4 5. Bc4 Nc6 6. Nc3 Qxd4 7. Qb3 Ne5
8. Be2 Nf6 9. O-O c6 10. Bg5 Qb4 11. Qxb4 Bxb4 12. Bxf6 gxf6 13. Nxe4 f5
14. Nf6+ Ke7 15. Nh5 Rd8 16. Rad1 Be6 17. b3 Bd2 18. Ng7 f4 19. Nxe6 Kxe6
20. Rb1 Rd4 21. Rfd1 Rad8 22. Bf1 Kf5 23. Rb2 h5 24. Rc2 R8d5 25. Be2 h4
26. h3 a5 27. a3 b5 28. Kh1 Ke4 29. Rb1 Kf5 30. Kg1 f3 31. gxf3 Kf4 32. b4
axb4 33. axb4 Bxb4 34. Kg2 Bd2 35. Rb3 b4 36. Ba6 c5 37. Rb1 Nxf3 38. Bb7
Rg5 {White resigns} 0-1[/pgn]
Dragon Elo needs to be set to about 2700 to be competitive with full strength Dragon 2.6.1 at knight odds. If 2400 elo can beat SF 14.1 that way, I think it just shows the same thing as your other SF games, namely that SF seems to "think" that it is good to trade queens when down a piece! A human IM would crush SF in a knight odds rapid match if it keeps trading queens.
I just finished playing 6 games with the same openings that GM Ben Finegold used up to move 10 using 2 intel i7 computers, the other i7 computer I borrowed from my daughter. I set up the Elo to 2600 for Komodo Dragon 2.6.1 with the Full strength and at full strength the Komodo Dragon giving the Knight Odds, the score was 3.5 to 2.5 in favor of the Komodo Dragon with the full strength, T/C 15 Minutes 10 Seconds.
lkaufman
Posts: 6279
Joined: Sun Jan 10, 2010 6:15 am
Location: Maryland USA
Full name: Larry Kaufman

Re: What is Komodo Average rating at Knight odds Vs Humans

Post by lkaufman »

Chessqueen wrote: Mon Jan 24, 2022 5:24 am
lkaufman wrote: Sat Jan 22, 2022 7:28 pm
Chessqueen wrote: Sat Jan 22, 2022 5:55 pm A good Solid 2400 human player that play like Komodo Dragon when I set the UCI_Elo = 2400 should NOT lose a game against any strong engine like Stockfish 14,1 NNUE, but there are NOT too many solid strong Humans Fide players that play like Komodo Dragon 2.6.1 set up to Limit Playing strength = 2400


[pgn][Event "Computer chess game"]
[Site "MININT-UB2PIMJ"]
[Date "2022.01.22"]
[Round "?"]
[White "Stockfish_14.1_win_x64_avx2"]
[Black "Dragon-2.6.1-64bit-avx2"]
[Result "0-1"]
[BlackElo "2400"]
[Time "11:10:18"]
[WhiteElo "3500"]
[TimeControl "900+10"]
[SetUp "1"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKB1R w KQkq - 0 1"]
[Termination "unterminated"]
[PlyCount "1"]
[Termination "normal"]
[PlyCount "60"]
[WhiteType "program"]
[BlackType "program"]

1. d4 d5 2. c4 e6 3. cxd5 exd5 4. e4 dxe4 5. Bc4 Nc6 6. Nc3 Qxd4 7. Qb3 Ne5
8. Be2 Nf6 9. O-O c6 10. Bg5 Qb4 11. Qxb4 Bxb4 12. Bxf6 gxf6 13. Nxe4 f5
14. Nf6+ Ke7 15. Nh5 Rd8 16. Rad1 Be6 17. b3 Bd2 18. Ng7 f4 19. Nxe6 Kxe6
20. Rb1 Rd4 21. Rfd1 Rad8 22. Bf1 Kf5 23. Rb2 h5 24. Rc2 R8d5 25. Be2 h4
26. h3 a5 27. a3 b5 28. Kh1 Ke4 29. Rb1 Kf5 30. Kg1 f3 31. gxf3 Kf4 32. b4
axb4 33. axb4 Bxb4 34. Kg2 Bd2 35. Rb3 b4 36. Ba6 c5 37. Rb1 Nxf3 38. Bb7
Rg5 {White resigns} 0-1[/pgn]
Dragon Elo needs to be set to about 2700 to be competitive with full strength Dragon 2.6.1 at knight odds. If 2400 elo can beat SF 14.1 that way, I think it just shows the same thing as your other SF games, namely that SF seems to "think" that it is good to trade queens when down a piece! A human IM would crush SF in a knight odds rapid match if it keeps trading queens.
I just finished playing 6 games with the same openings that GM Ben Finegold used up to move 10 using 2 intel i7 computers, the other i7 computer I borrowed from my daughter. I set up the Elo to 2600 for Komodo Dragon 2.6.1 with the Full strength and at full strength the Komodo Dragon giving the Knight Odds, the score was 3.5 to 2.5 in favor of the Komodo Dragon with the full strength, T/C 15 Minutes 10 Seconds.
That's pretty consistent with my own results, but I just wonder why you bothered to use two computers, why not just play them directly on one computer like any other normal engine match? If it's because you wanted pondering, you were wrong, because 2600 plays instantly and stops thinking at the same depth whether pondering or not, so ponder on or off has no consequence for this test. The result is still valid, you just went to a lot of unnecessary trouble.
Komodo rules!