Reckless 0.9.0-dev-2a847427 is out

[Jouni]

Here https://github.com/codedeliveryservice/ ... v-2a847427 . I haven't played any game. But I noticed, that Reckless is better or equal to Stockfish 1.1. in all testsuites I have ran now. Stunning, because Stockfish 1.1. is clearly better than 17.1. in solving.

[Jouni]

First match with HERT book:

Code: Select all

Score of Stockfish17.1 vs Reckless: 9 - 9 - 182 [0.500]
...      Stockfish17.1 playing White: 8 - 5 - 87  [0.515] 100
...      Stockfish17.1 playing Black: 1 - 4 - 95  [0.485] 100
...      White vs Black: 12 - 6 - 182  [0.515] 200
Elo difference: 0.0 +/- 14.4, LOS: 50.0 %, DrawRatio: 91.0 %
200 of 200 games finished.

[pohl4711]

First match with HERT book:

Code: Select all

Score of Stockfish17.1 vs Reckless: 9 - 9 - 182 [0.500]
...      Stockfish17.1 playing White: 8 - 5 - 87  [0.515] 100
...      Stockfish17.1 playing Black: 1 - 4 - 95  [0.485] 100
...      White vs Black: 12 - 6 - 182  [0.515] 200
Elo difference: 0.0 +/- 14.4, LOS: 50.0 %, DrawRatio: 91.0 %
200 of 200 games finished.

And this, dear children, is the reason, balanced openings are nonsense for testing top-engines...

[Jouni]

BTW this was the first ever draw against latest official Stockfish version I have played .

[Jouni]

Reckless won same opening with white and black. 60 + 0.6 games. Must be very rare happening. Once in 100 000 gamepairs?
[pgn]
[White "Reckless"]
[Black "Stockfish17.1"]
[Result "1-0"]

1. d4 {book} Nf6 {book} 2. c4 {book} e6 {book} 3. Nf3 {book} Bb4+ {book} 4.
Nbd2 {book} d5 {book} 5. Qa4+ {0.57/18 0.93s} Nc6 {-0.61/22 1.6s} 6. e3 {0.51/
24 4.7s} Be7 {-0.50/26 2.9s} 7. Qc2 {0.53/21 1.0s} a5 {-0.19/21 0.50s} 8. a3 {
0.53/20 0.84s} a4 {-0.38/28 2.4s} 9. cxd5 {0.47/21 1.3s} exd5 {-0.35/21 0.74s}
10. Bb5 {0.32/22 1.5s} Bd7 {-0.30/24 1.1s} 11. Bxa4 {0.34/19 1.4s} O-O {-0.36/
23 0.67s} 12. Bxc6 {0.21/22 5.4s} Bxc6 {-0.30/23 0.94s} 13. O-O {0.32/20 0.97s}
Bd6 {-0.27/24 0.96s} 14. b3 {0.34/18 0.74s} Bd7 {-0.33/24 2.1s} 15. a4 {0.35/
17 0.76s} Qe7 {-0.36/21 0.66s} 16. Bb2 {0.50/15 0.71s} Rfe8 {-0.31/20 0.57s}
17. Ne5 {0.50/19 1.4s} Ng4 {-0.32/26 4.7s} 18. Ndf3 {0.50/20 3.3s} Bc8 {-0.35/
21 0.51s} 19. Rfc1 {0.51/17 1.5s} f6 {-0.34/24 4.9s} 20. Nd3 {0.51/17 1.0s} Bf5
{-0.44/19 0.94s} 21. Qe2 {0.47/18 0.86s} Nh6 {-0.43/22 2.3s} 22. a5 {0.47/18 1.
5s} Bg4 {-0.42/24 1.9s} 23. h3 {0.53/18 1.1s} Bh5 {-0.35/17 0.34s} 24. Bc3 {0.
47/21 2.8s} c6 {-0.39/20 1.2s} 25. b4 {0.53/15 0.71s} Nf7 {-0.57/24 8.7s} 26.
Qf1 {0.56/17 0.85s} Qd7 {-0.60/21 2.4s} 27. Nc5 {0.49/20 2.3s} Bxc5 {-0.53/16
0.26s} 28. dxc5 {0.60/18 0.55s} Bxf3 {-0.56/19 0.66s} 29. gxf3 {0.43/20 0.73s}
Ng5 {-0.47/15 0.36s} 30. f4 {0.48/21 1.1s} Nf3+ {-0.49/18 0.34s} 31. Kg2 {0.52/
17 0.46s} Nh4+ {-0.48/18 0.35s} 32. Kg3 {0.54/19 0.90s} Nf5+ {-0.52/16 0.46s}
33. Kh2 {0.52/15 0.62s} Re4 {-0.47/20 0.68s} 34. Rab1 {0.61/14 0.44s} Rc4 {-0.
48/20 1.4s} 35. Qd3 {0.33/15 1.0s} Nh4 {-0.51/19 0.57s} 36. Qd1 {0.39/16 0.81s}
Nf5 {-0.53/23 3.5s} 37. Ba1 {0.41/18 2.0s} Rxc1 {-0.41/15 0.14s} 38. Rxc1 {0.
41/16 0.38s} Re8 {-0.66/21 1.4s} 39. Qg4 {0.48/17 0.68s} Kf7 {-0.42/12 0.16s}
40. Qh5+ {0.48/17 0.57s} Kg8 {-0.43/13 0.057s} 41. Rg1 {0.55/16 0.41s} Kh8 {-0.
39/15 0.17s} 42. Qf3 {0.42/20 3.8s} Re4 {-0.32/18 0.91s} 43. Qg4 {0.49/15 0.40s
} Qe6 {-0.16/12 0.070s} 44. Rb1 {0.82/15 0.38s} Rc4 {-0.16/17 0.24s} 45. b5 {
1.30/13 0.32s} Qe4 {-0.04/16 0.22s} 46. Rg1 {1.61/18 0.90s} Rc2 {-1.70/18 0.97s
} 47. Rg2 {1.70/17 0.38s} Qe6 {-1.71/14 0.17s} 48. a6 {2.13/18 0.60s} bxa6 {-1.
81/17 0.18s} 49. b6 {2.16/18 0.37s} Rxc5 {-1.70/17 0.20s} 50. f3 {1.91/18 0.55s
} Rb5 {-1.49/18 0.53s} 51. e4 {1.99/15 0.29s} Rxb6 {-1.54/14 0.13s} 52. Bd4 {
1.91/17 0.70s} Rb4 {-1.40/17 0.24s} 53. exf5 {2.09/18 0.79s} Qf7 {-1.56/15 0.
42s} 54. Bc5 {2.11/16 0.28s} Rb5 {-1.22/17 0.26s} 55. Rc2 {2.09/16 0.37s} Qd7 {
-1.23/15 0.33s} 56. Qh5 {3.23/16 0.55s} h6 {-1.84/21 0.74s} 57. Bd4 {2.85/17 0.
46s} Qe7 {-1.69/15 0.17s} 58. Rxc6 {2.80/14 0.57s} Rb2+ {-1.91/15 0.50s} 59.
Kh1 {2.99/16 0.31s} Rb8 {-1.92/17 0.46s} 60. Re6 {3.22/15 0.31s} Qc7 {-2.01/16
0.27s} 61. Rxf6 {5.15/15 0.24s} Qc1+ {-3.92/22 1.7s} 62. Bg1 {5.59/16 0.25s}
Qxf4 {-3.93/19 0.17s} 63. Rf7 {5.82/17 0.31s} Rd8 {-4.21/19 0.82s} 64. Rxg7 {
7.08/13 0.40s} Kxg7 {-5.64/15 0.47s} 65. Qg6+ {7.28/12 0.33s} Kh8 {-6.17/17 0.
39s} 66. Qf6+ {7.47/14 0.28s} Kg8 {-6.52/14 0.13s} 67. Qxd8+ {7.73/13 0.25s}
Kf7 {-6.55/15 0.12s} 68. Qxd5+ {7.83/18 0.62s} Ke8 {-7.26/21 0.87s, White wins
by adjudication} 1-0
[/pgn]

[pgn][White "Stockfish17.1"]
[Black "Reckless"]
[Result "0-1"]

1. d4 {book} Nf6 {book} 2. c4 {book} e6 {book} 3. Nf3 {book} Bb4+ {book} 4.
Nbd2 {book} d5 {book} 5. Qa4+ {0.48/23 2.0s} Nc6 {-0.50/20 1.4s} 6. a3 {0.47/
21 0.70s} Bxd2+ {-0.46/20 1.4s} 7. Bxd2 {0.54/19 0.65s} Ne4 {-0.46/20 1.2s} 8.
e3 {0.63/21 0.71s} O-O {-0.47/19 0.92s} 9. Qc2 {0.53/24 1.4s} Nxd2 {-0.50/19 1.
1s} 10. Qxd2 {0.46/21 1.4s} Ne7 {-0.52/20 1.7s} 11. c5 {0.64/24 2.0s} a5 {-0.
41/18 1.3s} 12. Bd3 {0.68/22 1.7s} b6 {-0.36/19 0.98s} 13. Rc1 {0.52/28 3.6s}
Ba6 {-0.47/22 2.2s} 14. Bxa6 {0.53/22 0.45s} Rxa6 {-0.49/21 1.2s} 15. Qe2 {0.
64/19 0.92s} Ra8 {-0.50/20 1.6s} 16. O-O {0.63/14 0.69s} Rb8 {-0.43/19 2.8s}
17. Ne5 {0.67/23 2.7s} f6 {-0.61/18 1.8s} 18. cxb6 {0.67/20 0.47s} c5 {-0.39/
18 1.2s} 19. dxc5 {0.77/15 0.73s} fxe5 {-0.52/16 1.5s} 20. Qb5 {0.73/17 0.62s}
Nf5 {-0.36/17 1.3s} 21. Qxa5 {0.97/17 0.56s} Nh4 {-0.34/20 1.6s} 22. Qd2 {0.95/
19 0.78s} e4 {-0.32/18 0.99s} 23. Kh1 {1.00/19 0.58s} Qg5 {-0.78/21 2.5s} 24.
Rg1 {0.96/21 1.9s} Ng6 {-0.59/17 0.76s} 25. f4 {0.78/23 3.6s} exf3 {-0.49/18 1.
1s} 26. gxf3 {0.67/20 0.69s} Qh6 {-0.39/20 1.3s} 27. f4 {0.92/24 4.1s} Nh4 {-0.
46/19 1.3s} 28. Rcf1 {0.88/20 0.47s} Nf5 {-0.43/17 0.99s} 29. b4 {0.49/28 9.1s}
Kh8 {-0.18/17 0.66s} 30. Qf2 {0.93/19 1.4s} e5 {-0.23/21 3.9s} 31. fxe5 {0.84/
15 0.34s} Qe6 {-0.15/15 0.91s} 32. Re1 {0.69/19 1.9s} g6 {0.00/16 1.3s} 33. Qf3
{0.73/18 2.3s} Kg7 {0.00/21 0.63s} 34. a4 {0.17/20 3.1s} Rbd8 {0.00/18 0.46s}
35. b5 {0.81/15 0.13s} d4 {0.20/15 0.64s} 36. Qc6 {0.21/16 1.1s} Qc4 {0.00/18
1.2s} 37. Qb7+ {0.38/17 0.52s} Kh8 {0.11/16 1.5s} 38. exd4 {1.20/13 0.053s} Nh4
{1.38/15 0.49s} 39. Rg3 {-0.18/20 0.71s} Qxd4 {5.27/15 0.63s} 40. Reg1 {-3.34/
17 0.45s} Rd5 {7.16/12 0.94s} 41. Re1 {-4.53/16 0.41s} Rd7 {10.62/17 1.8s} 42.
Qc6 {-7.57/18 0.40s} Qf2 {#21/23 0.49s} 43. Reg1 {-#20/26 0.15s} Rd2 {#17/28 0.
68s} 44. R3g2 {-#16/33 0.10s} Nxg2 {#15/29 0.67s} 45. Qxg2 {-#14/37 0.13s,
Black wins by adjudication} 0-1

[/pgn]

[peter]

Reckless won same opening with white and black. 60 + 0.6 games. Must be very rare happening. Once in 100 000 gamepairs?

At a TC of 60"+0.6 per game, it can happen, even if it's not very frequent as for equally very strong engines, but it occurs now and then, especially if one of the engines has decisively lower part of the hardware (-time) given, e.g. if there's too much CPU- load for both of the engines running a match, so just to ask: what was the hardware and how many concurrencies did you have?
That Stockfish accidentally blunders once with White and once with Black out of same opening, can occur without other reason but coincidence of course yet too, as said.

Anyhow, I just had a quick Forward with SF dev. along the games and did let it paste single moves with at once decisively differing eval , in game of Reckless winning with White at 43...Qe6? followed by 44...Rc4? and in the one with the reversed colours at 39.Rg3?, both games in one .pgn, that both practically were already lost after these blunders, regard: eval of SF dev. pasted together with the move is from White's perspective counted, not as in orginal games from the side to move:
[pgn][Event "?"]
[Site "?"]
[Date "2026.01.10"]
[Round "?"]
[White "Reckless"]
[Black "Stockfish17.1"]
[Result "1-0"]
[Annotator "Martan,Peter"]
[ECO "E11"]
[WhiteFideId "-1"]
[BlackFideId "-1"]
[PlyCount "136"]
[GameId "2264855192527217"]

1. d4 {book} Nf6 {book} 2. c4 {book} e6 {book} 3. Nf3 {book} Bb4+ {book} 4. Nbd2 {book} d5 {book} 5. Qa4+ {0.57/18 0.93s 2.0s} Nc6 {-0.61/22 1.6s 1.4s} 6. e3 {24 0.51/ 4.7s} (6. a3 {21 0.47/ 0.70s} Bxd2+ {-0.46/20 1.4s} 7. Bxd2 {0.54/19 0.65s} Ne4 {-0.46/20 1.2s} 8. e3 {0.63/21 0.71s} O-O {-0.47/19 0.92s} 9. Qc2 {0.53/24 1.4s} Nxd2 {-0.50/19 1. 1s} 10. Qxd2 {0.46/21 1.4s} Ne7 {-0.52/20 1.7s} 11. c5 {0.64/24 2.0s} a5 {-0. 41/18 1.3s} 12. Bd3 {0.68/22 1.7s} b6 {-0.36/19 0.98s} 13. Rc1 {0.52/28 3.6s} Ba6 {-0.47/22 2.2s} 14. Bxa6 {0.53/22 0.45s} Rxa6 {-0.49/21 1.2s} 15. Qe2 {0. 64/19 0.92s} Ra8 {-0.50/20 1.6s} 16. O-O {0.63/14 0.69s} Rb8 {-0.43/19 2.8s} 17. Ne5 {0.67/23 2.7s} f6 {-0.61/18 1.8s} 18. cxb6 {0.67/20 0.47s} c5 {18 -0.39/ 1.2s} 19. dxc5 {0.77/15 0.73s} fxe5 {-0.52/16 1.5s} 20. Qb5 {0.73/17 0.62s} Nf5 {-0.36/17 1.3s} 21. Qxa5 {0.97/17 0.56s} Nh4 {-0.34/20 1.6s} 22. Qd2 {19 0.95/ 0.78s} e4 {-0.32/18 0.99s} 23. Kh1 {1.00/19 0.58s} Qg5 {-0.78/21 2.5s} 24. Rg1 {0.96/21 1.9s} Ng6 {-0.59/17 0.76s} 25. f4 {0.78/23 3.6s} exf3 {-0.49/18 1. 1s} 26. gxf3 {0.67/20 0.69s} Qh6 {-0.39/20 1.3s} 27. f4 {0.92/24 4.1s} Nh4 {-0. 46/19 1.3s} 28. Rcf1 {0.88/20 0.47s} Nf5 {-0.43/17 0.99s} 29. b4 {0.49/28 9.1s} Kh8 {-0.18/17 0.66s} 30. Qf2 {0.93/19 1.4s} e5 {-0.23/21 3.9s} 31. fxe5 {15 0.84/ 0.34s} Qe6 {-0.15/15 0.91s} 32. Re1 {0.69/19 1.9s} g6 {0.00/16 1.3s} 33. Qf3 {0.73/18 2.3s} Kg7 {0.00/21 0.63s} 34. a4 {0.17/20 3.1s} Rbd8 {0.00/18 0.46s} 35. b5 {0.81/15 0.13s} d4 {0.20/15 0.64s} 36. Qc6 {0.21/16 1.1s} Qc4 {0.00/18 1.2s} 37. Qb7+ {0.38/17 0.52s} Kh8 {0.11/16 1.5s} 38. exd4 {1.20/13 0.053s} Nh4 {1.38/15 0.49s} 39. Rg3 $2 {-0.18/20 0.71s} (39. e6 {0.00/49 4}) 39... Qxd4 {5.27/15 0.63s} 40. Reg1 {17 -3.34/ 0.45s} Rd5 {7.16/12 0.94s} 41. Re1 {-4.53/16 0.41s} Rd7 {10.62/17 1.8s} 42. Qc6 {-7.57/18 0.40s} Qf2 {#21/23 0.49s} 43. Reg1 {-#20/26 0.15s} Rd2 {#17/28 0. 68s} 44. R3g2 {-#16/33 0.10s} Nxg2 {#15/29 0.67s} 45. Qxg2 {-#14/37;0-1 Stockfish17.1-Reckless 0.13s, Black wins by adjudication}) 6... Be7 {-0.50/26 2.9s} 7. Qc2 {0.53/21 1.0s} a5 {-0.19/21 0.50s} 8. a3 {0.53/20 0.84s} a4 {-0.38/28 2.4s} 9. cxd5 {0.47/21 1.3s} exd5 {-0.35/21 0.74s} 10. Bb5 {0.32/22 1.5s} Bd7 {-0.30/24 1.1s} 11. Bxa4 {0.34/19 1.4s} O-O {23 -0.36/ 0.67s} 12. Bxc6 {0.21/22 5.4s} Bxc6 {-0.30/23 0.94s} 13. O-O {0.32/20 0.97s} Bd6 {-0.27/24 0.96s} 14. b3 {0.34/18 0.74s} Bd7 {-0.33/24 2.1s} 15. a4 {17 0.35/ 0.76s} Qe7 {-0.36/21 0.66s} 16. Bb2 {0.50/15 0.71s} Rfe8 {-0.31/20 0.57s} 17. Ne5 {0.50/19 1.4s} Ng4 {-0.32/26 4.7s} 18. Ndf3 {0.50/20 3.3s} Bc8 {21 -0.35/ 0.51s} 19. Rfc1 {0.51/17 1.5s} f6 {-0.34/24 4.9s} 20. Nd3 {0.51/17 1.0s} Bf5 {-0.44/19 0.94s} 21. Qe2 {0.47/18 0.86s} Nh6 {-0.43/22 2.3s} 22. a5 {0.47/18 1. 5s} Bg4 {-0.42/24 1.9s} 23. h3 {0.53/18 1.1s} Bh5 {-0.35/17 0.34s} 24. Bc3 {0. 47/21 2.8s} c6 {-0.39/20 1.2s} 25. b4 {0.53/15 0.71s} Nf7 {-0.57/24 8.7s} 26. Qf1 {0.56/17 0.85s} Qd7 {-0.60/21 2.4s} 27. Nc5 {0.49/20 2.3s} Bxc5 {-0.53/16 0.26s} 28. dxc5 {0.60/18 0.55s} Bxf3 {-0.56/19 0.66s} 29. gxf3 {0.43/20 0.73s} Ng5 {-0.47/15 0.36s} 30. f4 {0.48/21 1.1s} Nf3+ {-0.49/18 0.34s} 31. Kg2 {17 0.52/ 0.46s} Nh4+ {-0.48/18 0.35s} 32. Kg3 {0.54/19 0.90s} Nf5+ {-0.52/16 0.46s} 33. Kh2 {0.52/15 0.62s} Re4 {-0.47/20 0.68s} 34. Rab1 {0.61/14 0.44s} Rc4 {-0. 48/20 1.4s} 35. Qd3 {0.33/15 1.0s} Nh4 {-0.51/19 0.57s} 36. Qd1 {0.39/16 0.81s} Nf5 {-0.53/23 3.5s} 37. Ba1 {0.41/18 2.0s} Rxc1 {-0.41/15 0.14s} 38. Rxc1 {0. 41/16 0.38s} Re8 {-0.66/21 1.4s} 39. Qg4 {0.48/17 0.68s} Kf7 {-0.42/12 0.16s} 40. Qh5+ {0.48/17 0.57s} Kg8 {-0.43/13 0.057s} 41. Rg1 {0.55/16 0.41s} Kh8 {-0. 39/15 0.17s} 42. Qf3 {0.42/20 3.8s} Re4 {-0.32/18 0.91s} 43. Qg4 {0.49/15 0.40s} Qe6 $2 {-0.16/12 0.070s} (43... Re8 {0.31/28 9}) 44. Rb1 {0.82/15 0.38s} Rc4 $2 {-0.16/17 0.24s} (44... Qd7 {1.35/33 9}) 45. b5 {1.30/13 0.32s} Qe4 {-0.04/16 0.22s} 46. Rg1 {1.61/18 0.90s} Rc2 {-1.70/18 0.97s} 47. Rg2 {1.70/17 0.38s} Qe6 {-1.71/14 0.17s} 48. a6 {2.13/18 0.60s} bxa6 {-1. 81/17 0.18s} 49. b6 {2.16/18 0.37s} Rxc5 {-1.70/17 0.20s} 50. f3 {1.91/18 0.55s} Rb5 {-1.49/18 0.53s} 51. e4 {1.99/15 0.29s} Rxb6 {-1.54/14 0.13s} 52. Bd4 {1.91/17 0.70s} Rb4 {-1.40/17 0.24s} 53. exf5 {2.09/18 0.79s} Qf7 {-1.56/15 0. 42s} 54. Bc5 {2.11/16 0.28s} Rb5 {-1.22/17 0.26s} 55. Rc2 {2.09/16 0.37s} Qd7 {-1.23/15 0.33s} 56. Qh5 {3.23/16 0.55s} h6 {-1.84/21 0.74s} 57. Bd4 {2.85/17 0. 46s} Qe7 {-1.69/15 0.17s} 58. Rxc6 {2.80/14 0.57s} Rb2+ {-1.91/15 0.50s} 59. Kh1 {2.99/16 0.31s} Rb8 {-1.92/17 0.46s} 60. Re6 {3.22/15 0.31s} Qc7 {-2.01/16 0.27s} 61. Rxf6 {5.15/15 0.24s} Qc1+ {-3.92/22 1.7s} 62. Bg1 {5.59/16 0.25s} Qxf4 {-3.93/19 0.17s} 63. Rf7 {5.82/17 0.31s} Rd8 {-4.21/19 0.82s} 64. Rxg7 {7.08/13 0.40s} Kxg7 {-5.64/15 0.47s} 65. Qg6+ {7.28/12 0.33s} Kh8 {-6.17/17 0. 39s} 66. Qf6+ {7.47/14 0.28s} Kg8 {-6.52/14 0.13s} 67. Qxd8+ {7.73/13 0.25s} Kf7 {-6.55/15 0.12s} 68. Qxd5+ {7.83/18 0.62s} Ke8 {-7.26/21 0.87s, White wins by adjudication} 1-0
[/pgn]

[Jouni]

Yes SF blundered in both games. I was running 5 games in 6 core i5 at 4 Ghz. No hyperthreading.

[peter]

Yes SF blundered in both games. I was running 5 games in 6 core i5 at 4 Ghz. No hyperthreading.

So 5 of 6 cores busy with 5 games concurrently, correct?
Did you allocate each concurrency (game) to its specified core?

[Jouni]

Did you allocate each concurrency (game) to its specified core?

No. How is that done?

[peter]

Did you allocate each concurrency (game) to its specified core?

No. How is that done?

E.g. in Task Manager under Details after choosing the process and opening its context- menu- Set affinity and there with -Processor affinity
choosing a certain CPU- core for each single concurrency.
I don't do so with matches of at least 3'+1" and different settings or branches of the same engine neither, yet with VSTC and not so near to each other engines it may be the most secure way to deny unequal distribution of usage of threads (cores).
As said, need not have any other reasons to get 1-0-0-1 pairs now and then with VSTC others but pure coincidence at all anyhow, regards