What's new? I tried to make the engine a bit more aggressive. The engine should also quickly reach the search depths. This is an advantage, especially for blitz games, but it doesn't mean that the engine plays better. This is good for slow computers but for Blitz games. For testing I played on PlayChess with 3 cores (6 threads) of the Ryzen 3900x and 7000 kns (startposition).
This engine uses a slightly older NNUE net (nn-3c0aa92af1da.nnue), which I consider to be tactically stronger than the new nets since July 1, 2022. This engine also uses a different scaling, and will not play well with new nets. Therefore please do not use any other network. Kayra 1.5 was also optimized for older networks when it was released.
Options:
---------------------------------------
MinimumThinkingTime:
I ONLY use this option for online games, not offline. The default setting 0 is best for offline games.
Online, I set the value to 100 to 200, but ONLY if the engine moves in depth 1 in time trouble. If you have a bad LAG, this can happen. Even with extremely closed positions and well over 100 moves played, the engine sometimes moves to depth 1.
---------------------------------------
1) Materialistic Evaluation Strategy: Minimum = -12, Maximum = +12, Default = 0. Lower values will cause the engine assign less value to material differences between the sides. More values will cause the engine to assign more value to the material difference.
2) Positional Evaluation Strategy: Minimum = -12, Maximum = +12, Default = 0. Lower values will cause the engine assign less value to positional differences between the sides. More values will cause the engine to assign more value to the positional difference.
The default value for both options (0 = zero) is equivalent to the default evaluation strategy of Stockfish.
Example (Eduard with Ryzen 3900x):
Default Setting:
8/3P3k/n2K3p/2p3n1/1b4N1/2p1p1P1/8/3B4 w - - 0 1
Analysis by Kayra 1.5 Attack-avx2:
1.Nxe3 Ba5 2.Be2 Nb8 3.Ke7 Nc6+ 4.Kd6 Nd4 5.Bd3+ Kg8 6.Kxc5 Nge6+ 7.Kc4 Kg7 8.Bb1 Kf7 9.d8Q Bxd8 10.Kxc3 Ne2+ 11.Kd2 Nxg3 12.Nd5 Bg5+ 13.Ke1 Nd4 14.Kf2 Ngf5 15.Nc3 Bh4+ 16.Kg1 Ne3 17.Bd3 h5 18.Ne4 Ke6 19.Kh2 Ke5 20.Kh3 Be1 21.Nc5 h4 22.Ba6 Kf6
-+ (-5.14 --) Depth: 29/60 00:00:15 204MN, tb=1632600
Black is clearly winning
With
Materialistic Evaluation Strategy -12
Positional Evaluation Strategy +12
Analysis by Kayra 1.5 Attack-avx2:
1.Nf6+ Kg7 2.Nh5+ Kg6 3.Bc2+ Kxh5 4.d8Q Kg4 5.Qf6 Kxg3 6.Qe5+ Kf3 7.Bd1+ Kg2 8.Qxe3 c4+ 9.Kd5 Bc5 10.Qxc3 Nc7+ 11.Kxc4 Nce6 12.Qe5 Bf2 13.Bg4 Nd4 14.Bc8 Nf7 15.Bb7+ Kf1 16.Qh2 Ne2 17.Qg2+ Ke1 18.Qd5 Ng5 19.Qh1+ Kd2
+- (5.56 ++) Depth: 24/42 00:00:06 34125kN, tb=351448
White is clearly winning
Downloads Kayra 1.5 Attack (AVX2, BMI2, SSE41, Source Code):
https://filehorst.de/d/efHbdbFF
Eduard Nemeth