Swordfish 14.3, Blue Marlin 14.3

[dorsz]

At the beginning I would like to point out that none of these engines are stronger than Stockfish, if you are looking for more Elo you will not be interested.

Swordfish and Blue Marlin are engines based on Stockfish and ideas used in Bluefish by Eelco de Groot who gave his engine more entertaining style of playing.

Swordfish has a reasonable balance between strength and tactical abilities. It still has solid style of playing and achieves good results in test suites. It can be useful for analyzing sharp positions where regular Stockfish can overlook some tactics.

Blue Marlin is more focused on king attack, it sacrifices more material what makes it a fast tactical problem solver. It is weaker than Swordfish but can provide even more attractive and interesting games.

14.3 vs 14.2:
https://www.mediafire.com/view/3xtvxphd ... 2.png/file#

[dorsz]

Each archive contains both engines:
x64 AVX2 - https://www.mediafire.com/file/ia4xta0i ... 2.zip/file
x64 BMI2 - https://www.mediafire.com/file/lrdt1qe2 ... 2.zip/file
x64 modern - https://www.mediafire.com/file/t9seze4y ... n.zip/file
x64 base - https://www.mediafire.com/file/clmpknw2 ... e.zip/file
sources - https://www.mediafire.com/file/pju3a162 ... s.zip/file

[magicianofriga]

Hi,
I am a coding beginner, but I would like to make some changes to these engines to improve their performance. I believe that it is possible to have a strong engine which also attacks like crazy.
Can you point me to the exact changes that you have made in both Swordfish and Blue Marlin? By exact, I mean specific files and specific lines of code. I will try to change these parameters in SugaR and ShashChess and see if the persistent hash increases strength.
Thanks!
magicianofriga.

[dorsz]

I don't want to clutter the topic with unnecessary details by copying the code and explaining it. I will send you private message with it.

[Chess_Dragon]

BDG Accepted, 8 main (human) variations:
1. d4 d5 2. e4 dxe4 3. Nc3 Nf6 4. f3 exf3 5. Nxf3 {1} Bg4 ( 5... e6 {2} ) ( 5... g6 6. Bc4 {3} ) ( 5... Bf5 6. Bc4 {5} ( 6. Ne5 {4} ) ) ( 5... c6 6. Bc4 {6} ) 6. h3 Bxf3 {8} ( 6... Bh5 {7} ).

G/90"+1" 2Gh Sy.6 ; R7 2700X | RTX 2060

Code: Select all

Score of Blue Marlin 14.3-avx2 vs Lc0-0.28.0_609996: 3 - 1 - 12 [0.563]
...      Blue Marlin 14.3-avx2 playing White: 0 - 1 - 7  [0.438] 8
...      Blue Marlin 14.3-avx2 playing Black: 3 - 0 - 5  [0.688] 8
...      White vs Black: 0 - 4 - 12  [0.375] 16
Elo difference: 43.7 +/- 85.4, LOS: 84.1 %, DrawRatio: 75.0 %
16 games: BlM 9.0/16, Lc0 7.0/16.

Score of Swordfish 14.3-avx2 vs Lc0-0.28.0_609996: 4 - 0 - 12 [0.625]
...      Swordfish 14.3-avx2 playing White: 0 - 0 - 8  [0.500] 8
...      Swordfish 14.3-avx2 playing Black: 4 - 0 - 4  [0.750] 8
...      White vs Black: 0 - 4 - 12  [0.375] 16
Elo difference: 88.7 +/- 80.2, LOS: 97.7 %, DrawRatio: 75.0 %
16 games: Sw 10.0/16, Lc0 6.0/16.

BlM: https://www.mediafire.com/file/ci6lifbu ... M.pgn/file
Sw: https://www.mediafire.com/file/9595c5kc ... w.pgn/file

[Chess_Dragon]

From my BDG post above:
Lc0 0-1 Blue Marlin 14.3, game #14, Round 7: after 48...Nf5:
if 49.Be1 = BlM -0.06 90 seconds, but after Lc0's 49.Be5?? Rg8 {+3.64/21 0.78s --black is winning}.

[Chess_Dragon]

From my BDG post above.

Lc0 0-1 Swordfish 14.3, game #6, Round 3, after 22...Qe8! -+ :


Lc0 0-1 Swordfish 14.3, game #12, Round 6, after 24...Qe6! /+ :