My test for Kayra 1.0 v6 bmi2

Discussion of computer chess matches and engine tournaments.

Moderator: Ras

User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: My test for Kayra 1.0 v6 bmi2

Post by AlexChess »

Hi Eduard,
I have noticed that there is also a free access to Playchess servers (I want only watch your online games) but I'm unable to find Kayra 1.1.

https://play.chessbase.com/en/ Is there a direct link?

Best regards, alex
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: My test for Kayra 1.0 v6 bmi2

Post by AlexChess »

First game lost by Kayra 1.1 + Kayra 1.1 bin book on my Superblitz against dorz's Swordfish 14.7 (40 moves x 150 secs + 40 moves x 150 secs + 1 sec)

[pgn][Event "SuperBlitz3+3-BanksiaGui"]
[Site "Rome"]
[Date "2022.01.08"]
[Round "1"]
[White "Kayra 1.1 PB Kayra 1.1 bin"]
[Black "Swordfish 14.7"]
[Result "0-1"]
[Board "118"]
[ECO "B90"]
[Opening "Sicilian"]
[Termination "adjudication by engines' scores"]
[Time "10:16:37"]
[TimeControl "40/150+1"]

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3
{ B90: Sicilian, Najdorf, Byrne (English) attack } 6...e5 7.Nb3 Be6 8.Be2 Be7
9.Nd5 Nxd5 10.exd5 Bf5 { End of opening } 11.Qd2 { +0.51/25 15552 7077944 }
11...O-O { +0.15/23 15381 8087411 } 12.Na5 { +0.50/25 2396 3810199 } 12...b6
{ +0.45/25 4809 8808992 } 13.Nc4 { +0.59/25 3623 7637098 } 13...Nd7
{ +0.55/26 3194 7205585 } 14.a4 { +0.56/28 11504 23619519 } 14...Bg6
{ +0.60/26 10284 21712447 } 15.O-O { +0.74/24 3496 7764370 } 15...Rc8
{ +0.51/27 9801 20041680 } 16.f3 { +0.56/25 4426 9796970 } 16...h6
{ +0.67/26 6191 13429006 } 17.c3 { +0.65/25 3522 7971550 } 17...Rb8
{ +0.72/26 6025 12984459 } 18.b4 { +0.78/24 5359 11969576 } 18...f5
{ +0.65/23 1905 4459340 } 19.Kh1 { +0.84/24 5193 11324455 } 19...Nf6
{ +0.00/25 5901 13769065 } 20.a5 { +0.13/29 27285 59644271 } 20...b5
{ +0.18/24 2316 5713357 } 21.Bb6 { +0.30/27 1847 4640534 } 21...Qd7
{ +0.34/25 3777 9336547 } 22.Na3 { +0.35/28 4719 10666709 } 22...Bd8
{ +0.00/29 8615 19665482 } 23.Bg1 { +0.39/26 3163 7777109 } 23...Nh5
{ -0.06/26 2329 5821173 } 24.c4 { +0.87/24 2742 6919400 } 24...Bh4
{ +0.00/30 3722 8704624 } 25.cxb5 { +0.50/25 6265 15624943 } 25...f4
{ -0.80/25 3770 10243301 } 26.Bf2 { -0.96/28 18672 47503801 } 26...Bxf2
{ -0.95/28 4917 13108119 } 27.Rxf2 { -0.94/25 2020 5334190 } 27...Ng3+
{ -1.32/24 3960 9941524 } 28.Kg1 { -0.93/27 3741 10198099 } 28...e4
{ -1.45/26 3923 9474482 } 29.Qe1 { -1.03/30 14151 35013768 } 29...Nxe2+
{ -1.39/24 4457 11250394 } 30.Qxe2 { -1.04/27 1818 4922882 } 30...e3
{ -1.54/29 3912 10059028 } 31.Rff1 { -0.82/26 2625 7313929 } 31...axb5
{ -1.50/29 7522 19339282 } 32.Rfd1 { -0.90/25 3980 10496131 } 32...Rfc8
{ -1.60/28 5106 13094373 } 33.Ra2 { -1.00/29 11881 28728103 } 33...Rc3
{ -1.57/29 7355 18709448 } 34.h3 { -1.35/28 3567 9376653 } 34...Rb3
{ -1.71/28 6049 15995389 } 35.Nc2 { -1.49/29 8616 22417160 } 35...Ra8
{ -2.23/27 6497 17074875 } 36.Qe1 { -1.53/25 1797 5063132 } 36...Rd3
{ -1.78/34 12091 31667198 } 37.Ra3 { -1.03/25 2405 6715446 } 37...Rxa3
{ -2.21/26 4595 12416952 } 38.Nxa3 { -0.92/21 447 1266926 } 38...Rc8
{ -2.35/27 5750 15407847 } 39.Nb1 { -1.46/22 1725 4640903 } 39...Rc2
{ -4.32/26 5445 15374712 } 40.Nc3 { -2.55/19 1142 3406830 } 40...Qe7
{ -4.57/25 4223 12525803 } 41.Ne2 { -4.28/30 30807 88935221 } 41...Qe5
{ -4.74/25 3410 10120629 } 42.Rc1 { -4.00/27 8977 26922872 } 42...Rd2
{ -5.80/27 9551 28159477 } 43.a6 { -4.21/26 6487 18936359 } 43...Ra2
{ -5.60/23 2575 7956613 } 44.Qc3 { -4.97/31 34994 98904074 } 44...Qg5
{ -6.23/25 5281 16019008 } 45.Re1 { -5.46/29 13414 37188336 } 0-1[/pgn]
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
Eduard
Posts: 1439
Joined: Sat Oct 27, 2018 12:58 am
Location: Germany
Full name: N.N.

Re: My test for Kayra 1.0 v6 bmi2

Post by Eduard »

AlexChess wrote: Fri Jan 07, 2022 7:59 pm Hi Eduard,
I have noticed that there is also a free access to Playchess servers (I want only watch your online games) but I'm unable to find Kayra 1.1.

https://play.chessbase.com/en/ Is there a direct link?

Best regards, alex
Hello Alex,
PlayChess is not lichess. Online in the browser you can only play and watch human games. To watch machine engine games, you need the Fritz access software. I explained everything in the General Forum in the Kayra thread.
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: My test for Kayra 1.0 v6 bmi2

Post by AlexChess »

Eduard wrote: Sat Jan 08, 2022 11:02 am
AlexChess wrote: Fri Jan 07, 2022 7:59 pm Hi Eduard,
I have noticed that there is also a free access to Playchess servers (I want only watch your online games) but I'm unable to find Kayra 1.1.

https://play.chessbase.com/en/ Is there a direct link?

Best regards, alex
Hello Alex,
PlayChess is not lichess. Online in the browser you can only play and watch human games. To watch machine engine games, you need the Fritz access software. I explained everything in the General Forum in the Kayra thread.
Thank you! I study your tutorial. I'm very interested because engines challenge online is becoming every day better :)

Kayra 1.1 only loss analyzed by Stockfish 14.1 (mac version)

[pgn][Event "SuperBlitz3+3-BanksiaGui"]
[Site "Rome"]
[Date "2022.01.08"]
[Round "1"]
[White "Kayra 1.1 PB Kayra 1.1 bin"]
[Black "Swordfish 14.7"]
[Result "0-1"]
[Board "118"]
[ECO "B90"]
[Opening "Sicilian"]
[Termination "adjudication by engines' scores"]
[Time "10:16:37"]
[TimeControl "40/150+1"]

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3
{ B90: Sicilian, Najdorf, Byrne (English) attack } 6...e5 7.Nb3 Be6 8.Be2 Be7
9.Nd5 Nxd5 10.exd5 Bf5 { End of opening } 11.Qd2 { +0.51/25 15552 7077944 }
11...O-O { +0.15/23 15381 8087411 } 12.Na5 { +0.50/25 2396 3810199 } 12...b6
{ +0.45/25 4809 8808992 } 13.Nc4 { +0.59/25 3623 7637098 } 13...Nd7
{ +0.55/26 3194 7205585 } 14.a4 { +0.56/28 11504 23619519 } 14...Bg6
{ +0.60/26 10284 21712447 } 15.O-O { +0.74/24 3496 7764370 } 15...Rc8
{ +0.51/27 9801 20041680 } 16.f3 { +0.56/25 4426 9796970 } 16...h6
{ +0.67/26 6191 13429006 } 17.c3 { +0.65/25 3522 7971550 } 17...Rb8
{ +0.72/26 6025 12984459 } 18.b4 { +0.78/24 5359 11969576 } 18...f5
{ +0.65/23 1905 4459340 } 19.Kh1 { +0.84/24 5193 11324455 } 19...Nf6
{ +0.00/25 5901 13769065 } 20.a5 { +0.13/29 27285 59644271 } 20...b5
{ +0.18/24 2316 5713357 } 21.Bb6 { +0.30/27 1847 4640534 } 21...Qd7
{ +0.34/25 3777 9336547 } 22.Na3 { +0.35/28 4719 10666709 } 22...Bd8
{ +0.00/29 8615 19665482 } 23.Bg1 { +0.39/26 3163 7777109 } 23...Nh5
{ -0.06/26 2329 5821173 } 24.c4 { +0.87/24 2742 6919400 } 24...Bh4
{ +0.00/30 3722 8704624 } 25.cxb5 { +0.50/25 6265 15624943 } 25...f4
{ -0.80/25 3770 10243301 } 26.Bf2 { -0.96/28 18672 47503801 } 26...Bxf2
{ -0.95/28 4917 13108119 } 27.Rxf2 { -0.94/25 2020 5334190 } 27...Ng3+
{ -1.32/24 3960 9941524 } 28.Kg1 { -0.93/27 3741 10198099 } 28...e4
{ -1.45/26 3923 9474482 } 29.Qe1 { -1.03/30 14151 35013768 } 29...Nxe2+
{ -1.39/24 4457 11250394 } 30.Qxe2 { -1.04/27 1818 4922882 } 30...e3
{ -1.54/29 3912 10059028 } 31.Rff1 { -0.82/26 2625 7313929 } 31...axb5
{ -1.50/29 7522 19339282 } 32.Rfd1 { -0.90/25 3980 10496131 } 32...Rfc8
{ -1.60/28 5106 13094373 } 33.Ra2 { -1.00/29 11881 28728103 } 33...Rc3
{ -1.57/29 7355 18709448 } 34.h3 { -1.35/28 3567 9376653 } 34...Rb3
{ -1.71/28 6049 15995389 } 35.Nc2 { -1.49/29 8616 22417160 } 35...Ra8
{ -2.23/27 6497 17074875 } 36.Qe1 { -1.53/25 1797 5063132 } 36...Rd3
{ -1.78/34 12091 31667198 } 37.Ra3 { -1.03/25 2405 6715446 } 37...Rxa3
{ -2.21/26 4595 12416952 } 38.Nxa3 { -0.92/21 447 1266926 } 38...Rc8
{ -2.35/27 5750 15407847 } 39.Nb1 { -1.46/22 1725 4640903 } 39...Rc2
{ -4.32/26 5445 15374712 } 40.Nc3 { -2.55/19 1142 3406830 } 40...Qe7
{ -4.57/25 4223 12525803 } 41.Ne2 { -4.28/30 30807 88935221 } 41...Qe5
{ -4.74/25 3410 10120629 } 42.Rc1 { -4.00/27 8977 26922872 } 42...Rd2
{ -5.80/27 9551 28159477 } 43.a6 { -4.21/26 6487 18936359 } 43...Ra2
{ -5.60/23 2575 7956613 } 44.Qc3 { -4.97/31 34994 98904074 } 44...Qg5
{ -6.23/25 5281 16019008 } 45.Re1 { -5.46/29 13414 37188336 } 0-1[/pgn]
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
Eduard
Posts: 1439
Joined: Sat Oct 27, 2018 12:58 am
Location: Germany
Full name: N.N.

Re: My test for Kayra 1.0 v6 bmi2

Post by Eduard »

AlexChess wrote: Sat Jan 08, 2022 10:23 am First game lost by Kayra 1.1 + Kayra 1.1 bin book on my Superblitz against dorz's Swordfish 14.7 (40 moves x 150 secs + 40 moves x 150 secs + 1 sec)

[pgn][Event "SuperBlitz3+3-BanksiaGui"]
[Site "Rome"]
[Date "2022.01.08"]
[Round "1"]
[White "Kayra 1.1 PB Kayra 1.1 bin"]
[Black "Swordfish 14.7"]
[Result "0-1"]
[Board "118"]
[ECO "B90"]
[Opening "Sicilian"]
[Termination "adjudication by engines' scores"]
[Time "10:16:37"]
[TimeControl "40/150+1"]

1.e4 c5 2.Nf3 d6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 a6 6.Be3
{ B90: Sicilian, Najdorf, Byrne (English) attack } 6...e5 7.Nb3 Be6 8.Be2 Be7
9.Nd5 Nxd5 10.exd5 Bf5 { End of opening } 11.Qd2 { +0.51/25 15552 7077944 }
11...O-O { +0.15/23 15381 8087411 } 12.Na5 { +0.50/25 2396 3810199 } 12...b6
{ +0.45/25 4809 8808992 } 13.Nc4 { +0.59/25 3623 7637098 } 13...Nd7
{ +0.55/26 3194 7205585 } 14.a4 { +0.56/28 11504 23619519 } 14...Bg6
{ +0.60/26 10284 21712447 } 15.O-O { +0.74/24 3496 7764370 } 15...Rc8
{ +0.51/27 9801 20041680 } 16.f3 { +0.56/25 4426 9796970 } 16...h6
{ +0.67/26 6191 13429006 } 17.c3 { +0.65/25 3522 7971550 } 17...Rb8
{ +0.72/26 6025 12984459 } 18.b4 { +0.78/24 5359 11969576 } 18...f5
{ +0.65/23 1905 4459340 } 19.Kh1 { +0.84/24 5193 11324455 } 19...Nf6
{ +0.00/25 5901 13769065 } 20.a5 { +0.13/29 27285 59644271 } 20...b5
{ +0.18/24 2316 5713357 } 21.Bb6 { +0.30/27 1847 4640534 } 21...Qd7
{ +0.34/25 3777 9336547 } 22.Na3 { +0.35/28 4719 10666709 } 22...Bd8
{ +0.00/29 8615 19665482 } 23.Bg1 { +0.39/26 3163 7777109 } 23...Nh5
{ -0.06/26 2329 5821173 } 24.c4 { +0.87/24 2742 6919400 } 24...Bh4
{ +0.00/30 3722 8704624 } 25.cxb5 { +0.50/25 6265 15624943 } 25...f4
{ -0.80/25 3770 10243301 } 26.Bf2 { -0.96/28 18672 47503801 } 26...Bxf2
{ -0.95/28 4917 13108119 } 27.Rxf2 { -0.94/25 2020 5334190 } 27...Ng3+
{ -1.32/24 3960 9941524 } 28.Kg1 { -0.93/27 3741 10198099 } 28...e4
{ -1.45/26 3923 9474482 } 29.Qe1 { -1.03/30 14151 35013768 } 29...Nxe2+
{ -1.39/24 4457 11250394 } 30.Qxe2 { -1.04/27 1818 4922882 } 30...e3
{ -1.54/29 3912 10059028 } 31.Rff1 { -0.82/26 2625 7313929 } 31...axb5
{ -1.50/29 7522 19339282 } 32.Rfd1 { -0.90/25 3980 10496131 } 32...Rfc8
{ -1.60/28 5106 13094373 } 33.Ra2 { -1.00/29 11881 28728103 } 33...Rc3
{ -1.57/29 7355 18709448 } 34.h3 { -1.35/28 3567 9376653 } 34...Rb3
{ -1.71/28 6049 15995389 } 35.Nc2 { -1.49/29 8616 22417160 } 35...Ra8
{ -2.23/27 6497 17074875 } 36.Qe1 { -1.53/25 1797 5063132 } 36...Rd3
{ -1.78/34 12091 31667198 } 37.Ra3 { -1.03/25 2405 6715446 } 37...Rxa3
{ -2.21/26 4595 12416952 } 38.Nxa3 { -0.92/21 447 1266926 } 38...Rc8
{ -2.35/27 5750 15407847 } 39.Nb1 { -1.46/22 1725 4640903 } 39...Rc2
{ -4.32/26 5445 15374712 } 40.Nc3 { -2.55/19 1142 3406830 } 40...Qe7
{ -4.57/25 4223 12525803 } 41.Ne2 { -4.28/30 30807 88935221 } 41...Qe5
{ -4.74/25 3410 10120629 } 42.Rc1 { -4.00/27 8977 26922872 } 42...Rd2
{ -5.80/27 9551 28159477 } 43.a6 { -4.21/26 6487 18936359 } 43...Ra2
{ -5.60/23 2575 7956613 } 44.Qc3 { -4.97/31 34994 98904074 } 44...Qg5
{ -6.23/25 5281 16019008 } 45.Re1 { -5.46/29 13414 37188336 } 0-1[/pgn]
Hello Alex,
do you testers wonder why the engines lost or won? Do you also analyze?

In the game here, Kayra only comes to depth 24.

Kayra 1.1 PB Kayra 1.1 bin - Swordfish 14.7 0-1 1.118, SuperBlitz3+3-BanksiaGui 2022
1r1b1rk1/3q2p1/p2p2bp/Pp1Ppp1n/1P6/N1P2P2/3QB1PP/R4RBK w - - 0 1

Analysis by Kayra 1.1 avx2:

24.c4 Bh4 25.Bf2 Bxf2 26.Rxf2 f4 27.Qe1 Qe7 28.cxb5 Ng3+ 29.Kg1 Nxe2+ 30.Qxe2 axb5 31.Rc1 Rfc8 32.Rc6 Be8 33.Rxc8 Rxc8 34.Qd2 Qa7 35.h3 Bd7 36.Kh2 Rc3 37.Nxb5 Bxb5 38.Qxc3 Qxf2 39.Qc8+ Kh7 40.Qf5+ Kh8 41.Qc8+
The position is equal: = (0.00) Depth: 35/51 00:00:10 195MN, tb=30

With me on 22 threads the engine comes to depth 35 after 10s! When I play on PlayChess, such plies are normal. With a depth of 24 you will very often lose, so you can take whatever books you want. Without sufficient search depth, the engines will always make bad mistakes.

You also have to look for where the error occurs. I do this every day and analyze hundreds of games.

Kayra 1.1 PB Kayra 1.1 bin - Swordfish 14.7 0-1 1.118, SuperBlitz3+3-BanksiaGui 2022
[fen]1r3rk1/3q2p1/p2p2bp/Pp1Ppp1n/1PP4b/N4P2/3QB1PP/R4RBK w - - 0 1[/fen]


Analysis by Kayra 1.1 avx2:

25.cxb5 axb5 26.Bxb5 Rxb5 27.Nxb5 Qxb5 28.Qa2 Ra8 29.a6 Be8 30.a7 Bf6 31.Qa6 e4 32.Ra3 Qxb4 33.fxe4 Kh7 34.Rxf5 Qxe4 35.Rff3 Qxd5 36.Ra5 Qe4 37.Qd3 Qxd3 38.Rxd3
White is better: +/- (0.72) Depth: 24/38 00:00:01 36945kN, tb=12
25.Bf2 Bxf2 26.Rxf2 Qa7 27.Bf1 f4 28.cxb5 Ng3+ 29.Kg1 Nxf1 30.Raxf1 axb5 31.Rc1 Rfe8 32.Kh1 Qf7 33.h3 Kh7 34.Kh2 e4 35.fxe4 Rxe4 36.Re1 Rxe1 37.Qxe1 Qxd5 38.Rxf4 Qc6 39.Qd2 d5 40.Kg1 Rc8
The position is equal: = (-0.01) Depth: 25/43 00:00:05 122MN, tb=50

(---)
25.Bf2 Bxf2 26.Rxf2 f4 27.Qe1 Qe7 28.cxb5 Ng3+ 29.Kg1 Nxe2+ 30.Qxe2 axb5 31.Rc1 Rfc8 32.Rc6 Be8 33.Rxc8 Rxc8 34.Rf1 Qb7 35.Qd2 Qa7+ 36.Qf2 Qf7 37.Qd2
The position is equal: = (0.00) Depth: 34/26 00:00:06 133MN, tb=51

Kayra 1.1 made a mistake with cxb5? On my computer, Kayra needs 5s in order not to play this move. After 6s Kayra is on my computer at depth 34!

When I play on PlayChess, the engine sometimes even calculates significantly longer than just 5s. Sometimes it is calculated for up to 60s in critical positions. The engine sometimes even comes to depth 55!

So what did the engine do wrong?

I know why I don't like this kind of games (sorry) and also the strange book tournaments, where the engines often make mistakes and not the books. I am undefeated with Kayra 1.1 on PlayChess for 1561 games.
Last edited by Eduard on Sat Jan 08, 2022 12:06 pm, edited 3 times in total.
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: My test for Kayra 1.0 v6 bmi2

Post by AlexChess »

Eduard wrote: Sat Jan 08, 2022 11:51 am
do you testers wonder why the engines lost or won? Do you also analyze?

In the game here, Kayra only comes to depth 24.

Kayra 1.1 PB Kayra 1.1 bin - Swordfish 14.7 0-1 1.118, SuperBlitz3+3-BanksiaGui 2022
1r1b1rk1/3q2p1/p2p2bp/Pp1Ppp1n/1P6/N1P2P2/3QB1PP/R4RBK w - - 0 1

Analysis by Kayra 1.1 avx2:

24.c4 Bh4 25.Bf2 Bxf2 26.Rxf2 f4 27.Qe1 Qe7 28.cxb5 Ng3+ 29.Kg1 Nxe2+ 30.Qxe2 axb5 31.Rc1 Rfc8 32.Rc6 Be8 33.Rxc8 Rxc8 34.Qd2 Qa7 35.h3 Bd7 36.Kh2 Rc3 37.Nxb5 Bxb5 38.Qxc3 Qxf2 39.Qc8+ Kh7 40.Qf5+ Kh8 41.Qc8+
The position is equal: = (0.00) Depth: 35/51 00:00:10 195MN, tb=30

With me on 22 threads the engine comes to depth 35 after 10s! When I play on PlayChess, such plies are normal. With a depth of 24 you will very often lose, so you can take whatever books you want. Without sufficient search depth, the engines will always make bad mistakes.

You also have to look for where the error occurs. I do this every day and analyze hundreds of games.

Kayra 1.1 PB Kayra 1.1 bin - Swordfish 14.7 0-1 1.118, SuperBlitz3+3-BanksiaGui 2022
[fen]1r3rk1/3q2p1/p2p2bp/Pp1Ppp1n/1PP4b/N4P2/3QB1PP/R4RBK w - - 0 1[/fen]


So what did the engine do wrong?
I know that my hardware isn't powerful, but I've sent to you this game because maybe you could avoid to enter this position early already on opening. :)
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
Eduard
Posts: 1439
Joined: Sat Oct 27, 2018 12:58 am
Location: Germany
Full name: N.N.

Re: My test for Kayra 1.0 v6 bmi2

Post by Eduard »

Should I make the book better because of mistakes that happen? You can not be serious. The book made no mistakes, but the engine, long after the last book move. However, the book is very small, because I had only just started with it. That would not have happened with more books moves. However, I played all of these variants on PlayChess myself, and Kayra didn't make a mistake.

Have a nice day.
Eduard
Posts: 1439
Joined: Sat Oct 27, 2018 12:58 am
Location: Germany
Full name: N.N.

Re: My test for Kayra 1.0 v6 bmi2

Post by Eduard »

Folks, I closed my book projects. I hope you continue to enjoy testing.

I don't feel up to the computer chess community at the moment.

I had always said that I ONLY make my books for server games (Lichess, PlayChess and InfinityChess). These are my playgrounds.

Tuning books for book tournaments or rankings is not my thing.

Example is a game from Kayra posted by Alex. I do a lot of analysis, but I also expect the testers to ask themselves why engine X won and not engine Y?


[fen]1r1q1rk1/3nbpp1/pp1p2bp/3Pp3/P1N5/2P1BP2/1P1QB1PP/R4RK1 w - - 0 18[/fen]


This position above is balanced and the better one wins!

Analysis by Kayra 1.1 avx2:

18.b4 f5 19.Rad1 f4 20.Bf2 Qc7 21.Qa2 Rfc8 22.Rc1 a5 23.Qb3 Bf7 24.Rfd1 Nf6 25.Na3 axb4 26.Nb5 Qb7 27.c4 Nd7 28.Qxb4 Bg6 29.Bd3 Bxd3 30.Rxd3 Nc5 31.Ra3 e4 32.Bd4 Rf8 33.Qb1 Bf6 34.fxe4 Qe7 35.Bf2 Be5 36.Nd4 Qf6 37.Nf3 Bb2 38.e5 Bxe5 39.Re1 Rbe8 40.Nxe5 Rxe5 41.Rxe5 dxe5
White is slightly better: +/= (0.42 ++) Depth: 41/51 00:04:58 5616MN, tb=1432

Of course, I could implement moves here to fix Kayra's later bug. Only: The mistake happened on slow hardware, and Kayra wouldn't make the mistake on my hardware. Whoever wins that position was better on the test hardware! And in this case it wasn't Kayra. That has to be accepted.

The other reason for my closure is Hurnavich's cautious. He followed me everywhere to provoke me. On PlayChess he renamed engine names like no one else does! That's not forbidden, but bizzare. Especially when the names of an engine are renamed so that they sound similar to the original:

Image

This is a puppet show! I put someone like that on Ignore, so I'm glad I don't see such people on the server.

He also followed me in another forum and translated a text addressed to me from French into English. I hadn't asked him, I can handle it myself. I see it as a provocation.

I just wanted to make some friends in the forums and have fun with others, and share things like books, tests, etc. But I also don't want to have anything more to do with some people, and that's why I am now partially closing my homepage. I will also withdraw from the forum. If you want to be in Kontact with me, you should write to me by email (it is on my homepage under Contact), and only if it is important.
User avatar
AlexChess
Posts: 1562
Joined: Sat Feb 06, 2021 8:06 am
Full name: Alex Morales

Re: My test for Kayra 1.0 v6 bmi2

Post by AlexChess »

Eduard wrote: Sun Jan 09, 2022 4:39 am Folks, I closed my book projects. I hope you continue to enjoy testing.

I don't feel up to the computer chess community at the moment.

I had always said that I ONLY make my books for server games (Lichess, PlayChess and InfinityChess). These are my playgrounds.
Sorry for that :) Thinking that was what you were asking, I was giving you the only game lost by Kayra 1.1 on my tests to let you to change the variant leading to this position that the engine cannot handle with my weak hardware.

This is a competitive playground, we must expect unfair behavior by some members and we must be resilient.

On my side, I want to reduce time spent updating SuperBlitz (I'm thinking to change completely TC to long tournaments playing few games between TOP 10 engines) and back to develop and PLAY chess.
Chess engines and dedicated chess computers fan since 1981 :D macOS Sequoia 16GB-512GB, Windows 11 & Ubuntu ARM64.
ProteusSF Dev Forum
Eduard
Posts: 1439
Joined: Sat Oct 27, 2018 12:58 am
Location: Germany
Full name: N.N.

Re: My test for Kayra 1.0 v6 bmi2

Post by Eduard »

To Alex: You should accept that engines make mistakes. In the defeat of Kayra, the opponent played with the same book lines. He played better than Kayra. I can't do anything about it.

Info (from my homepage): Although I have successfully participated in a few book tournaments (1st and 3rd place), I am not particularly interested in that. Because the bigger mistakes are made by the engines, not the books! This is due to weak hardware, low time control (less than 5 minutes) and ponder OFF. You can take the best book and the engines will still fail. My record on PlayChess is 1561 undefeated games in a row! And the game that was lost only came about because of a timeout. I play on the PlayChess, Lichess and InfinityChess servers. And that's why I make my books exclusively. It is fought with strong hardware and with Ponder ON. If you want books that should be consistently successful elsewhere, you've come to the wrong place. I regularly improve my books, but only for server games! In the future, I will therefore only give my books privately to friends who play on servers and want to have fun there. Friends and prospects can contact me by email. I have stopped public downloads, also because some people are only interested in provocation and confrontation out of envy. I do not need that.

Bye all.