HasiLover_Test wrote:

ArnoHu wrote:

I just updated GoK NNUE's neural network from Smallbrain nn-380 to nn-400. You can find it at https://turbowarp.org/1033379869

Neural network data is on github: https://github.com/ArnoHue/scratch/blob/master/chess/NNUE.scratch

In an automated 40-test-games series, GoK NNUE nn-400 won 23.5 - 16.5 against GoK NNUE nn-380.

We are all doomed

ur right

chessconscarth wrote:

ArnoHu wrote:

I just updated GoK NNUE's neural network from Smallbrain nn-380 to nn-400. You can find it at https://turbowarp.org/1033379869

Neural network data is on github: https://github.com/ArnoHue/scratch/blob/master/chess/NNUE.scratch

In an automated 40-test-games series, GoK NNUE nn-400 won 23.5 - 16.5 against GoK NNUE nn-380.

how in the name of stockfish is that possible?

that is a 50+ elo increase. in scratch engine terms that is a lot usely most updates are fixing bug or the thinking of the engine so, like 5-15 elo extra
so this is huge some needs to start make and NNUE compeer for GoK NNUE or GoK classical besides delta NNUE maybe White dove NNUE, element NNUE, miniment NNUE or LA NNUE.

chessconscarth wrote:

ArnoHu wrote:

I just updated GoK NNUE's neural network from Smallbrain nn-380 to nn-400. You can find it at https://turbowarp.org/1033379869

Neural network data is on github: https://github.com/ArnoHue/scratch/blob/master/chess/NNUE.scratch

In an automated 40-test-games series, GoK NNUE nn-400 won 23.5 - 16.5 against GoK NNUE nn-380.

how in the name of stockfish is that possible?

How is what possible? Test automation? Via the UCI project, which let's engine play against each other without user intervention. Template project: https://scratch.mit.edu/projects/1054412386/

https://turbowarp.org/1061850388/fullscreen?turbo&stuck is the new official LA version for tournament… use the competition mode by putting the depth to “c” when asked.

GoK NNUE (NO TW-Ext, comp.mode, white) with another draw against SF 16 L8. There was a small chance during midgame. Again, with one exception where the right move came very late in search, not a lot of risks for mistakes this time. nn-400 pays off, given search depth was not great during the first half of the game, due to lack of piece exchanges and lots of space for sliding pieces. 98% vs 98%. The first win without TW-Ext might be getting closer. https://lichess.org/nsQQOR0t#116

ArnoHu wrote:

GoK NNUE (NO TW-Ext, comp.mode, white) with another draw against SF 16 L8. There was a small chance during midgame. Again, with one exception where the right move came very late in search, not a lot of risks for mistakes this time. nn-400 pays off, given search depth was not great during the first half of the game, due to lack of piece exchanges and lots of space for sliding pieces. 98% vs 98%. The first win without TW-Ext might be getting closer. https://lichess.org/nsQQOR0t#116

Wow, highly impressive!

ArnoHu wrote:

GoK NNUE (NO TW-Ext, comp.mode, white) with another draw against SF 16 L8. There was a small chance during midgame. Again, with one exception where the right move came very late in search, not a lot of risks for mistakes this time. nn-400 pays off, given search depth was not great during the first half of the game, due to lack of piece exchanges and lots of space for sliding pieces. 98% vs 98%. The first win without TW-Ext might be getting closer. https://lichess.org/nsQQOR0t#116

Next game, TW experimental branch, similar story. GoK white, draw at 99% vs. 99% accuracy:
https://lichess.org/CvzNsN4h#117

Two more draws with GoK as white:
https://lichess.org/IeSh7Ur9#57
https://lichess.org/V77tZzbt#62

Plz add the 50-move rule to Gok
because I decided I show one of Gok weakness is locked positions so I play a game and tried to get a locked position but
the game went on and on
and pasted the 50 move rule
1. d2d4 d7d5 2. Ng1f3 Ng8f6 3. Bc1g5 e7e6 4. e2e3 Nb8d7 5. Bf1d3 Bf8d6 6. O-O O-O 7. c2c4 h7h6 8. Bg5xf6 Nd7xf6 9. c4c5 Bd6e7 10. Nb1c3 c7c6 11. Qd1c2 Qd8c7 12. Ra1c1 Qc7d8 13. Qc2e2 Qd8e8 14. h2h3 Be7d8 15. e3e4 Bd8c7 16. e4e5 Nf6h7 17. Qe2c2 g7g6 18. Qc2d2 Kg8g7 19. a2a3 f7f5 20. Rc1e1 b7b6 21. b2b4 b6b5 22. Re1b1 Bc8d7 23. a3a4 a7a6 24. Rf1e1 h6h5 25. Rb1a1 Rf8h8 26. a4xb5 a6xb5 27. Ra1xa8 Qe8xa8 28. Re1d1 Rh8b8 29. Rd1e1 Qa8b7 30. Nc3e2 Qb7c8 31. Re1a1 Rb8a8 32. Ra1d1 Qc8d8 33. Ne2c3 Qd8c8 34. Rd1e1 Bc7d8 35. Qd2e3 Bd8e7 36. Re1f1 Qc8d8 37. Nc3e2 Be7g5 38. Nf3xg5 Qd8xg5 39. Qe3xg5 Nh7xg5 40. h3h4 Ng5f7 41. Ne2f4 Nf7h6 42. f2f3 Kg7f7 43. Bd3c2 Ra8a2 44. Rf1c1 Ra2a7 45. Kg1f1 Ra7a8 46. Kf1g1 Kf7g7 47. Bc2b3 Kg7f7 48. Bb3c2 Kf7g7 49. Bc2d3 Kg7f7 50. Kg1f1 Kf7g7 51. Kf1g1 Kg7f7 52. Rc1e1 Ra8a4 53. Re1b1 Bd7c8 54. Bd3c2 Ra4a8 55. Rb1e1 Bc8d7 56. Re1d1 Ra8a2 57. Rd1c1 Ra2a8 58. Kg1h1 Ra8a3 59. Bc2d1 Ra3a8 60. Bd1c2 Ra8a7 61. Kh1g1 Ra7a8 62. Bc2d3 Ra8a3 63. Rc1b1 Ra3a8 64. Bd3e2 Kf7g7 65. Be2d3 Kg7f7 66. Bd3e2 Kf7g7 67. Kg1f1 Kg7f7 68. g2g3 Kf7g7 69. Be2d3 Kg7f7 70. Kf1g1 Kf7g7 71. Bd3e2 Kg7f7 72. Be2d3 Ra8a6 73. Bd3c2 Ra6a8 74. Rb1f1 Ra8a6 75. Rf1d1 Ra6a2 76. Rd1c1 Ra2a8 77. Bc2d3 Ra8a7 78. Rc1e1 Ra7a8 79. Kg1f1 Ra8a7 80. Bd3e2 Ra7a8 81. Re1b1 Ra8a4 82. Kf1g1 Ra4a8 83. Rb1d1 Ra8a2 84. Be2d3 Ra2a3 85. Kg1f1 Ra3a8 86. Rd1e1 Ra8a2 87. Re1e2 Ra2a8 88. Bd3c2 Ra8a7 89. Re2e1 Ra7a8 90. Kf1g1 Ra8a7 91. Re1b1 Ra7a8 92. Rb1d1 Ra8a2 93. Rd1c1 Ra2a3 94. Rc1f1 Ra3a8 95. Bc2d3 Ra8a7 96. Rf1e1 Ra7a8 97. Bd3c2 Ra8a2 98. Bc2b3 Ra2a3 99. Re1b1 Ra3a8 100. Bb3c2 Ra8a2 101. Bc2d3 Ra2a8 102. Kg1f1 Ra8a7 103. Bd3e2 Ra7a8 104. Be2d3 Ra8a7 105. Bd3c2 Ra7a8 106. Rb1d1 Ra8a7 107. Rd1e1 Ra7a8 108. Bc2b3 Ra8a3 109. Bb3d1 Ra3a8 110. Bd1e2 Ra8a7 111. Kf1g2 Ra7a2 112. Kg2f1 Ra2a8 113. Re1c1 Ra8a4 114. Rc1b1 Ra4a8 115. Rb1e1 Ra8a7 116. Be2d1 Ra7a8 117. Bd1b3 Ra8a3 118. Re1b1 Ra3a8 119. Kf1g1 Ra8a3 120. Nf4h3

chessconscarth wrote:

Plz add the 50-move rule to Gok
because I decided I show one of Gok weakness is locked positions so I play a game and tried to get a locked position but
the game went on and on
and pasted the 50 move rule
1. d2d4 d7d5 2. Ng1f3 Ng8f6 3. Bc1g5 e7e6 4. e2e3 Nb8d7 5. Bf1d3 Bf8d6 6. O-O O-O 7. c2c4 h7h6 8. Bg5xf6 Nd7xf6 9. c4c5 Bd6e7 10. Nb1c3 c7c6 11. Qd1c2 Qd8c7 12. Ra1c1 Qc7d8 13. Qc2e2 Qd8e8 14. h2h3 Be7d8 15. e3e4 Bd8c7 16. e4e5 Nf6h7 17. Qe2c2 g7g6 18. Qc2d2 Kg8g7 19. a2a3 f7f5 20. Rc1e1 b7b6 21. b2b4 b6b5 22. Re1b1 Bc8d7 23. a3a4 a7a6 24. Rf1e1 h6h5 25. Rb1a1 Rf8h8 26. a4xb5 a6xb5 27. Ra1xa8 Qe8xa8 28. Re1d1 Rh8b8 29. Rd1e1 Qa8b7 30. Nc3e2 Qb7c8 31. Re1a1 Rb8a8 32. Ra1d1 Qc8d8 33. Ne2c3 Qd8c8 34. Rd1e1 Bc7d8 35. Qd2e3 Bd8e7 36. Re1f1 Qc8d8 37. Nc3e2 Be7g5 38. Nf3xg5 Qd8xg5 39. Qe3xg5 Nh7xg5 40. h3h4 Ng5f7 41. Ne2f4 Nf7h6 42. f2f3 Kg7f7 43. Bd3c2 Ra8a2 44. Rf1c1 Ra2a7 45. Kg1f1 Ra7a8 46. Kf1g1 Kf7g7 47. Bc2b3 Kg7f7 48. Bb3c2 Kf7g7 49. Bc2d3 Kg7f7 50. Kg1f1 Kf7g7 51. Kf1g1 Kg7f7 52. Rc1e1 Ra8a4 53. Re1b1 Bd7c8 54. Bd3c2 Ra4a8 55. Rb1e1 Bc8d7 56. Re1d1 Ra8a2 57. Rd1c1 Ra2a8 58. Kg1h1 Ra8a3 59. Bc2d1 Ra3a8 60. Bd1c2 Ra8a7 61. Kh1g1 Ra7a8 62. Bc2d3 Ra8a3 63. Rc1b1 Ra3a8 64. Bd3e2 Kf7g7 65. Be2d3 Kg7f7 66. Bd3e2 Kf7g7 67. Kg1f1 Kg7f7 68. g2g3 Kf7g7 69. Be2d3 Kg7f7 70. Kf1g1 Kf7g7 71. Bd3e2 Kg7f7 72. Be2d3 Ra8a6 73. Bd3c2 Ra6a8 74. Rb1f1 Ra8a6 75. Rf1d1 Ra6a2 76. Rd1c1 Ra2a8 77. Bc2d3 Ra8a7 78. Rc1e1 Ra7a8 79. Kg1f1 Ra8a7 80. Bd3e2 Ra7a8 81. Re1b1 Ra8a4 82. Kf1g1 Ra4a8 83. Rb1d1 Ra8a2 84. Be2d3 Ra2a3 85. Kg1f1 Ra3a8 86. Rd1e1 Ra8a2 87. Re1e2 Ra2a8 88. Bd3c2 Ra8a7 89. Re2e1 Ra7a8 90. Kf1g1 Ra8a7 91. Re1b1 Ra7a8 92. Rb1d1 Ra8a2 93. Rd1c1 Ra2a3 94. Rc1f1 Ra3a8 95. Bc2d3 Ra8a7 96. Rf1e1 Ra7a8 97. Bd3c2 Ra8a2 98. Bc2b3 Ra2a3 99. Re1b1 Ra3a8 100. Bb3c2 Ra8a2 101. Bc2d3 Ra2a8 102. Kg1f1 Ra8a7 103. Bd3e2 Ra7a8 104. Be2d3 Ra8a7 105. Bd3c2 Ra7a8 106. Rb1d1 Ra8a7 107. Rd1e1 Ra7a8 108. Bc2b3 Ra8a3 109. Bb3d1 Ra3a8 110. Bd1e2 Ra8a7 111. Kf1g2 Ra7a2 112. Kg2f1 Ra2a8 113. Re1c1 Ra8a4 114. Rc1b1 Ra4a8 115. Rb1e1 Ra8a7 116. Be2d1 Ra7a8 117. Bd1b3 Ra8a3 118. Re1b1 Ra3a8 119. Kf1g1 Ra8a3 120. Nf4h3

GoK already has that.

ScratchChessChampion wrote:

chessconscarth wrote:

Plz add the 50-move rule to Gok
because I decided I show one of Gok weakness is locked positions so I play a game and tried to get a locked position but
the game went on and on
and pasted the 50 move rule
1. d2d4 d7d5 2. Ng1f3 Ng8f6 3. Bc1g5 e7e6 4. e2e3 Nb8d7 5. Bf1d3 Bf8d6 6. O-O O-O 7. c2c4 h7h6 8. Bg5xf6 Nd7xf6 9. c4c5 Bd6e7 10. Nb1c3 c7c6 11. Qd1c2 Qd8c7 12. Ra1c1 Qc7d8 13. Qc2e2 Qd8e8 14. h2h3 Be7d8 15. e3e4 Bd8c7 16. e4e5 Nf6h7 17. Qe2c2 g7g6 18. Qc2d2 Kg8g7 19. a2a3 f7f5 20. Rc1e1 b7b6 21. b2b4 b6b5 22. Re1b1 Bc8d7 23. a3a4 a7a6 24. Rf1e1 h6h5 25. Rb1a1 Rf8h8 26. a4xb5 a6xb5 27. Ra1xa8 Qe8xa8 28. Re1d1 Rh8b8 29. Rd1e1 Qa8b7 30. Nc3e2 Qb7c8 31. Re1a1 Rb8a8 32. Ra1d1 Qc8d8 33. Ne2c3 Qd8c8 34. Rd1e1 Bc7d8 35. Qd2e3 Bd8e7 36. Re1f1 Qc8d8 37. Nc3e2 Be7g5 38. Nf3xg5 Qd8xg5 39. Qe3xg5 Nh7xg5 40. h3h4 Ng5f7 41. Ne2f4 Nf7h6 42. f2f3 Kg7f7 43. Bd3c2 Ra8a2 44. Rf1c1 Ra2a7 45. Kg1f1 Ra7a8 46. Kf1g1 Kf7g7 47. Bc2b3 Kg7f7 48. Bb3c2 Kf7g7 49. Bc2d3 Kg7f7 50. Kg1f1 Kf7g7 51. Kf1g1 Kg7f7 52. Rc1e1 Ra8a4 53. Re1b1 Bd7c8 54. Bd3c2 Ra4a8 55. Rb1e1 Bc8d7 56. Re1d1 Ra8a2 57. Rd1c1 Ra2a8 58. Kg1h1 Ra8a3 59. Bc2d1 Ra3a8 60. Bd1c2 Ra8a7 61. Kh1g1 Ra7a8 62. Bc2d3 Ra8a3 63. Rc1b1 Ra3a8 64. Bd3e2 Kf7g7 65. Be2d3 Kg7f7 66. Bd3e2 Kf7g7 67. Kg1f1 Kg7f7 68. g2g3 Kf7g7 69. Be2d3 Kg7f7 70. Kf1g1 Kf7g7 71. Bd3e2 Kg7f7 72. Be2d3 Ra8a6 73. Bd3c2 Ra6a8 74. Rb1f1 Ra8a6 75. Rf1d1 Ra6a2 76. Rd1c1 Ra2a8 77. Bc2d3 Ra8a7 78. Rc1e1 Ra7a8 79. Kg1f1 Ra8a7 80. Bd3e2 Ra7a8 81. Re1b1 Ra8a4 82. Kf1g1 Ra4a8 83. Rb1d1 Ra8a2 84. Be2d3 Ra2a3 85. Kg1f1 Ra3a8 86. Rd1e1 Ra8a2 87. Re1e2 Ra2a8 88. Bd3c2 Ra8a7 89. Re2e1 Ra7a8 90. Kf1g1 Ra8a7 91. Re1b1 Ra7a8 92. Rb1d1 Ra8a2 93. Rd1c1 Ra2a3 94. Rc1f1 Ra3a8 95. Bc2d3 Ra8a7 96. Rf1e1 Ra7a8 97. Bd3c2 Ra8a2 98. Bc2b3 Ra2a3 99. Re1b1 Ra3a8 100. Bb3c2 Ra8a2 101. Bc2d3 Ra2a8 102. Kg1f1 Ra8a7 103. Bd3e2 Ra7a8 104. Be2d3 Ra8a7 105. Bd3c2 Ra7a8 106. Rb1d1 Ra8a7 107. Rd1e1 Ra7a8 108. Bc2b3 Ra8a3 109. Bb3d1 Ra3a8 110. Bd1e2 Ra8a7 111. Kf1g2 Ra7a2 112. Kg2f1 Ra2a8 113. Re1c1 Ra8a4 114. Rc1b1 Ra4a8 115. Rb1e1 Ra8a7 116. Be2d1 Ra7a8 117. Bd1b3 Ra8a3 118. Re1b1 Ra3a8 119. Kf1g1 Ra8a3 120. Nf4h3

GoK already has that.

Not really. I will add it once I find time. OK, I have been saying that for 3+ years, this feature has low priority for me compared to others.

ArnoHu wrote:

ArnoHu wrote:

GoK NNUE (NO TW-Ext, comp.mode, white) with another draw against SF 16 L8. There was a small chance during midgame. Again, with one exception where the right move came very late in search, not a lot of risks for mistakes this time. nn-400 pays off, given search depth was not great during the first half of the game, due to lack of piece exchanges and lots of space for sliding pieces. 98% vs 98%. The first win without TW-Ext might be getting closer. https://lichess.org/nsQQOR0t#116

Next game, TW experimental branch, similar story. GoK white, draw at 99% vs. 99% accuracy:
https://lichess.org/CvzNsN4h#117

Two more draws with GoK as white:
https://lichess.org/IeSh7Ur9#57
https://lichess.org/V77tZzbt#62

Finally, GoK NNUE for the first time wins against Stockfish 16 level 8 without using TurboWarp extensions (96% vs. 94% accuracy):

https://lichess.org/broadcast/gok-nnue-vs-stockfish/round-1/Yi0Qj1sB/kddgp48n#181

Challenge Idea: What if everyone tried making a chess engine in 7 days (In a generalized Board with move generation prefixed and UCI) and then we let them play against each other? Like some sort of competition?

HasiLover_Test wrote:

Challange Idea: What if everyone tried making a chess engine in 7 days (In a generalized Board with move generation prefixed and UCI) and then we let them play against each other? Like some sort of competition?

Cool, but there may be cheating…
Stupid thing, I lost D2 DEV… (the USB drive on which it was)…

ababoin07 wrote:

HasiLover_Test wrote:

Challange Idea: What if everyone tried making a chess engine in 7 days (In a generalized Board with move generation prefixed and UCI) and then we let them play against each other? Like some sort of competition?

Cool, but there may be cheating…
Stupid thing, I lost D2 DEV… (the USB drive on which it was)…

That doesnt matter, the Cheater is the true looser

HasiLover_Test wrote:

ababoin07 wrote:

HasiLover_Test wrote:

Challange Idea: What if everyone tried making a chess engine in 7 days (In a generalized Board with move generation prefixed and UCI) and then we let them play against each other? Like some sort of competition?

Cool, but there may be cheating…
Stupid thing, I lost D2 DEV… (the USB drive on which it was)…

That doesnt matter, the Cheater is the true looser

you're right

ababoin07 wrote:

HasiLover_Test wrote:

ababoin07 wrote:

HasiLover_Test wrote:

Challange Idea: What if everyone tried making a chess engine in 7 days (In a generalized Board with move generation prefixed and UCI) and then we let them play against each other? Like some sort of competition?

Cool, but there may be cheating…
Stupid thing, I lost D2 DEV… (the USB drive on which it was)…

That doesnt matter, the Cheater is the true looser

you're right

I was thinking we could do a different game like Connect 4 or Checkers.

BirdracerTrois wrote:

ababoin07 wrote:

HasiLover_Test wrote:

ababoin07 wrote:

HasiLover_Test wrote:

Challange Idea: What if everyone tried making a chess engine in 7 days (In a generalized Board with move generation prefixed and UCI) and then we let them play against each other? Like some sort of competition?

Cool, but there may be cheating…
Stupid thing, I lost D2 DEV… (the USB drive on which it was)…

That doesnt matter, the Cheater is the true looser

you're right

I was thinking we could do a different game like Connect 4 or Checkers.

Yes, I agree completely, even a collab with bloc**ive, we could make an ultra-optimized minimax algorithm, with all the prunings that we know, for example for the game of Checkers, and then we could make a collaboration of a chess engine with this functional minimax, We could then play against goK (so the collab without @arnohu)

ArnoHu wrote:

ScratchChessChampion wrote:

chessconscarth wrote:

Plz add the 50-move rule to Gok
because I decided I show one of Gok weakness is locked positions so I play a game and tried to get a locked position but
the game went on and on
and pasted the 50 move rule
1. d2d4 d7d5 2. Ng1f3 Ng8f6 3. Bc1g5 e7e6 4. e2e3 Nb8d7 5. Bf1d3 Bf8d6 6. O-O O-O 7. c2c4 h7h6 8. Bg5xf6 Nd7xf6 9. c4c5 Bd6e7 10. Nb1c3 c7c6 11. Qd1c2 Qd8c7 12. Ra1c1 Qc7d8 13. Qc2e2 Qd8e8 14. h2h3 Be7d8 15. e3e4 Bd8c7 16. e4e5 Nf6h7 17. Qe2c2 g7g6 18. Qc2d2 Kg8g7 19. a2a3 f7f5 20. Rc1e1 b7b6 21. b2b4 b6b5 22. Re1b1 Bc8d7 23. a3a4 a7a6 24. Rf1e1 h6h5 25. Rb1a1 Rf8h8 26. a4xb5 a6xb5 27. Ra1xa8 Qe8xa8 28. Re1d1 Rh8b8 29. Rd1e1 Qa8b7 30. Nc3e2 Qb7c8 31. Re1a1 Rb8a8 32. Ra1d1 Qc8d8 33. Ne2c3 Qd8c8 34. Rd1e1 Bc7d8 35. Qd2e3 Bd8e7 36. Re1f1 Qc8d8 37. Nc3e2 Be7g5 38. Nf3xg5 Qd8xg5 39. Qe3xg5 Nh7xg5 40. h3h4 Ng5f7 41. Ne2f4 Nf7h6 42. f2f3 Kg7f7 43. Bd3c2 Ra8a2 44. Rf1c1 Ra2a7 45. Kg1f1 Ra7a8 46. Kf1g1 Kf7g7 47. Bc2b3 Kg7f7 48. Bb3c2 Kf7g7 49. Bc2d3 Kg7f7 50. Kg1f1 Kf7g7 51. Kf1g1 Kg7f7 52. Rc1e1 Ra8a4 53. Re1b1 Bd7c8 54. Bd3c2 Ra4a8 55. Rb1e1 Bc8d7 56. Re1d1 Ra8a2 57. Rd1c1 Ra2a8 58. Kg1h1 Ra8a3 59. Bc2d1 Ra3a8 60. Bd1c2 Ra8a7 61. Kh1g1 Ra7a8 62. Bc2d3 Ra8a3 63. Rc1b1 Ra3a8 64. Bd3e2 Kf7g7 65. Be2d3 Kg7f7 66. Bd3e2 Kf7g7 67. Kg1f1 Kg7f7 68. g2g3 Kf7g7 69. Be2d3 Kg7f7 70. Kf1g1 Kf7g7 71. Bd3e2 Kg7f7 72. Be2d3 Ra8a6 73. Bd3c2 Ra6a8 74. Rb1f1 Ra8a6 75. Rf1d1 Ra6a2 76. Rd1c1 Ra2a8 77. Bc2d3 Ra8a7 78. Rc1e1 Ra7a8 79. Kg1f1 Ra8a7 80. Bd3e2 Ra7a8 81. Re1b1 Ra8a4 82. Kf1g1 Ra4a8 83. Rb1d1 Ra8a2 84. Be2d3 Ra2a3 85. Kg1f1 Ra3a8 86. Rd1e1 Ra8a2 87. Re1e2 Ra2a8 88. Bd3c2 Ra8a7 89. Re2e1 Ra7a8 90. Kf1g1 Ra8a7 91. Re1b1 Ra7a8 92. Rb1d1 Ra8a2 93. Rd1c1 Ra2a3 94. Rc1f1 Ra3a8 95. Bc2d3 Ra8a7 96. Rf1e1 Ra7a8 97. Bd3c2 Ra8a2 98. Bc2b3 Ra2a3 99. Re1b1 Ra3a8 100. Bb3c2 Ra8a2 101. Bc2d3 Ra2a8 102. Kg1f1 Ra8a7 103. Bd3e2 Ra7a8 104. Be2d3 Ra8a7 105. Bd3c2 Ra7a8 106. Rb1d1 Ra8a7 107. Rd1e1 Ra7a8 108. Bc2b3 Ra8a3 109. Bb3d1 Ra3a8 110. Bd1e2 Ra8a7 111. Kf1g2 Ra7a2 112. Kg2f1 Ra2a8 113. Re1c1 Ra8a4 114. Rc1b1 Ra4a8 115. Rb1e1 Ra8a7 116. Be2d1 Ra7a8 117. Bd1b3 Ra8a3 118. Re1b1 Ra3a8 119. Kf1g1 Ra8a3 120. Nf4h3

GoK already has that.

Not really. I will add it once I find time. OK, I have been saying that for 3+ years, this feature has low priority for me compared to others.

make sense

Hello, how can I implement Manhatten Distance to my Engine?

HasiLover_Test wrote:

Hello, how can I implement Manhatten Distance to my Engine?

Lookup table