Karjakin 17.Bxd5 !

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

User avatar
Dariusz
Posts: 364
Joined: Sat Jun 13, 2015 10:08 am
Location: Poland
Full name: Dariusz Domagała

Karjakin 17.Bxd5 !

Post by Dariusz »

Move is rather easy to play from human point of view, but engines has problem to find.

2kr2r1/pp1qbpp1/7p/2pp4/5Pb1/P1P1BB2/1PPQ2PP/2KR3R w - - 0 17
Image

[Event "World Chess Candidates 2018"]
[Site "Berlin"]
[Date "2018.03.24"]
[Round "12"]
[White "Karjakin, Sergey"]
[Black "Caruana, Fabiano"]
[Result "1-0"]
[WhiteElo "2763"]
[BlackElo "2784"]
[PlyCount "95"]
[EventDate "2018.??.??"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nf3 Nxe4 5. Nc3 Nxc3 6. dxc3 Nc6 7. Be3 Be7
8. Qd2 Be6 9. O-O-O Qd7 10. a3 h6 11. Nd4 Nxd4 12. Bxd4 Rg8 13. Be2 c5 14. Be3
d5 15. f4 O-O-O 16. Bf3 Bg4 17. Bxd5 Bxd1 18. Rxd1 Qc7 19. c4 Rge8 20. Qf2 b6
21. g4 Bf6 22. Kb1 Rd7 23. Rd3 g5 24. Ka2 Ree7 25. Qf3 Kd8 26. Bd2 Kc8 27. Qf1
Rd6 28. fxg5 Bxg5 29. Bxg5 hxg5 30. Qf5+ Rdd7 31. Qxg5 Qe5 32. Qh6 Kd8 33. g5
Qd6 34. Qh8+ Re8 35. Qh4 Qg6 36. Qg4 Re5 37. h4 Ke7 38. Rd2 b5 39. Bxf7 Qf5 40.
Rxd7+ Kxd7 41. Qxf5+ Rxf5 42. g6 Ke7 43. cxb5 Rh5 44. c4 Rxh4 45. a4 Rg4 46. a5
Kd6 47. a6 Kc7 48. Kb3 1-0
Regards, Darius
https://chessengeria.eu
User avatar
Eelco de Groot
Posts: 4561
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Karjakin 17.Bxd5 !

Post by Eelco de Groot »

Hello Dariusz,

It's a nice move! Also on Fishcooking

2kr2r1/pp1qbpp1/7p/2pp4/5Pb1/P1P1BB2/1PPQ2PP/2KR3R w - -

Engine: Kaissa III (i7 6700, 1 thread, 512 MB)
(This is still the 2018-03-04 codebase from Stockfish)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

20/28 0:01 +0.92 17.Bxd5 Bxd1 18.Rxd1 Bd6 19.c4 Rge8
20.Qf2 Qe7 21.Rd3 Bc7 22.Qf3 Kb8
23.f5 Qe5 24.g3 Rxd5 25.Qxd5 Qxd5
26.cxd5 c4 27.Rc3 (3.710.435) 2072

21/25 0:02 +0.90 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.c4 Bd6
20.Qf2 Qe7 21.Rd3 Bc7 22.Qf3 Kb8
23.Kb1 f5 24.Qh5 Rxd5 25.cxd5 c4
26.Rc3 Qd7 (4.648.781) 2076

22/32 0:03 +0.99++ 17.Bxd5 (8.007.348) 2067

22/32 0:04 +0.64-- 17.Bxd5 Bxd1 (8.891.367) 2067

22/32 0:04 +0.69 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.c4 Qc7
20.Qf2 f5 21.Rd3 b6 22.Qf3 Rd6
23.Bf2 Bf6 24.b4 Bd4 25.Bxd4 cxd4
26.Rxd4 Kb8 27.Qf2 Qe7 (9.518.456) 2063

23/34 0:06 +0.61-- 17.Bxd5 Bxd1 (12.501.388) 2049

23/34 0:06 +0.54-- 17.Bxd5 Bxd1 (13.363.475) 2050

23/34 0:07 +0.55 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.c4 b6
20.Kb1 Bf6 21.Qd3 Qe7 22.Re1 Qd7
23.Qd1 Bd4 24.Bxd4 Rxe1 25.Qxe1 cxd4
26.Ka2 Re8 27.Qg3 f5 28.Qf2 Kb8
29.Qxd4 Re1 30.Kb3 (14.410.443) 2051

24/30 0:08 +0.63++ 17.Bxd5 (18.365.294) 2053

24/30 0:09 +0.70++ 17.Bxd5 (20.088.086) 2056

24/30 0:10 +0.69 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.c4 b6
20.Qf2 Bf6 21.Qf3 Qe7 22.Re1 Kc7
23.Kb1 Bd4 24.Bd2 Qd7 25.f5 Rxe1+
26.Bxe1 Re8 27.Bg3+ Be5 28.b3 f6
29.Ka2 Bxg3 (22.070.440) 2057

25/34 0:12 +0.70 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.c4 f5
20.Qf2 b6 21.Kb1 Bf6 22.Rd3 Qe7
23.Qf3 Kb8 24.Ka2 g6 25.Rb3 Bd4
26.Bxd4 cxd4 27.Bc6 Qe2 28.Bxe8 Qxf3
29.Rxf3 (26.274.700) 2059

26/28 0:16 +0.70 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.c4 f5
20.Qf2 b6 21.Kb1 Bf6 22.Rd3 Qe7
23.Qf3 Kb8 24.Ka2 g6 25.Rb3 Bd4
26.Bxd4 cxd4 27.Bc6 Qe2 28.Bxe8 (33.770.812) 2062

27/30 0:22 +0.62-- 17.Bxd5 Bxd1 (46.886.664) 2056

27/30 0:22 +0.62 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.c4 f5
20.Qf2 b6 21.Rd3 Bf6 22.Kb1 Qc7
23.h3 Qe7 24.g4 Kb8 25.gxf5 Rxd5
26.cxd5 Qe4 27.d6 (47.277.257) 2057

28/36 0:35 +0.70++ 17.Bxd5 (73.262.480) 2052

28/36 0:46 +0.68 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.Qf2 f5
20.Rd3 Qc7 21.Kb1 b6 22.c4 Bf6 23.h3 g6
24.g4 Qg7 25.Bc1 Bd4 26.Qd2 Qd7
27.g5 hxg5 28.fxg5 Qe7 29.c3 Bg7
30.Qf4 (95.218.140) 2067

29/37 1:14 +0.75++ 17.Bxd5 (154.429.694) 2072

29/37 1:32 +0.76 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.Qf2 Bf6
20.c4 b6 21.Qf3 Qe7 22.Re1 Kc7
23.Kb1 Bd4 24.Bd2 Qd7 25.f5 Rxe1+
26.Bxe1 Be5 27.Ka2 Re8 28.Bg3 f6
29.b3 Rb8 30.h3 (194.836.614) 2100

30/37 1:34 +0.76 17.Bxd5 Bxd1 18.Rxd1 Rge8 19.Qf2 Bf6
20.c4 b6 21.Qf3 Qe7 22.Re1 Kc7
23.Kb1 Bd4 24.Bd2 Qd7 25.f5 Rxe1+
26.Bxe1 Be5 27.Ka2 Re8 28.Bg3 f6
29.b3 Rb8 30.h3 (199.252.740) 2101


best move: Bf3xd5 time: 1:34.844 min n/s: 2.101.445 nodes: 199.252.740
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
User avatar
Dariusz
Posts: 364
Joined: Sat Jun 13, 2015 10:08 am
Location: Poland
Full name: Dariusz Domagała

Re: Karjakin 17.Bxd5 !

Post by Dariusz »

Hello Eelco

Thx for a link to fishcooking.
Kaissa pretty fast found this move :D
Regards, Darius
https://chessengeria.eu
User avatar
Thomas Lagershausen
Posts: 328
Joined: Mon Jun 11, 2007 6:59 pm

Re: Karjakin 17.Bxd5 !

Post by Thomas Lagershausen »

This is one of the most important testpositions i have seen on CCC.

The reason is it test higher chessunderstanding.

White has with this outpost on d5 a clear strategic advantage.

Today no 3300 elo-engine understands this brilliant move.

17.Bxd5 deserves two exclamation marks.

After 17.Bxd5 !! the leader of the candidates tournament F. Caruana was helpless against the white play.

Maybe it was the game of the year.

I hope topchessprograms will understand this move in the near future.
TL
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Karjakin 17.Bxd5 !

Post by zullil »

Thomas Lagershausen wrote:This is one of the most important testpositions i have seen on CCC.

The reason is it test higher chessunderstanding.

White has with this outpost on d5 a clear strategic advantage.

Today no 3300 elo-engine understands this brilliant move.

17.Bxd5 deserves two exclamation marks.

After 17.Bxd5 !! the leader of the candidates tournament F. Caruana was helpless against the white play.

Maybe it was the game of the year.

I hope topchessprograms will understand this move in the near future.
White has the outpost, but has paid for it in material. Does Bxd5 actually win? Human chess is full of blunders; did Caruana defend correctly? Has anyone checked? (I haven't.)
User avatar
Thomas Lagershausen
Posts: 328
Joined: Mon Jun 11, 2007 6:59 pm

Re: Karjakin 17.Bxd5 !

Post by Thomas Lagershausen »

Who shows us the right defence?

Nobody did and no engine did.

Till then i believe in the white advantage.
TL
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Karjakin 17.Bxd5 !

Post by zullil »

Thomas Lagershausen wrote:Who shows us the right defence?

Nobody did and no engine did.

Till then i believe in the white advantage.
White has an advantage without Bxd5. The question is: Does Bxd5 give a winning advantage?

God might show us the right defense. But anyone with a strong computer and a strong engine can determine if Caruana's was a wrong defense. :wink:
yanquis1972
Posts: 1766
Joined: Wed Jun 03, 2009 12:14 am

Re: Karjakin 17.Bxd5 !

Post by yanquis1972 »

have you checked 20...f5? that was the obvious move so i'm assuming you have; i haven't been able to check any professional analysis yet myself.
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Karjakin 17.Bxd5 !

Post by zullil »

Dariusz wrote:Move is rather easy to play from human point of view, but engines has problem to find.

[D] 2kr2r1/pp1qbpp1/7p/2pp4/5Pb1/P1P1BB2/1PPQ2PP/2KR3R w - - 0 17

[pgn]
[Event "World Chess Candidates 2018"]
[Site "Berlin"]
[Date "2018.03.24"]
[Round "12"]
[White "Karjakin, Sergey"]
[Black "Caruana, Fabiano"]
[Result "1-0"]
[WhiteElo "2763"]
[BlackElo "2784"]
[PlyCount "95"]
[EventDate "2018.??.??"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 d6 4. Nf3 Nxe4 5. Nc3 Nxc3 6. dxc3 Nc6 7. Be3 Be7
8. Qd2 Be6 9. O-O-O Qd7 10. a3 h6 11. Nd4 Nxd4 12. Bxd4 Rg8 13. Be2 c5 14. Be3
d5 15. f4 O-O-O 16. Bf3 Bg4 17. Bxd5 Bxd1 18. Rxd1 Qc7 19. c4 Rge8 20. Qf2 b6
21. g4 Bf6 22. Kb1 Rd7 23. Rd3 g5 24. Ka2 Ree7 25. Qf3 Kd8 26. Bd2 Kc8 27. Qf1
Rd6 28. fxg5 Bxg5 29. Bxg5 hxg5 30. Qf5+ Rdd7 31. Qxg5 Qe5 32. Qh6 Kd8 33. g5
Qd6 34. Qh8+ Re8 35. Qh4 Qg6 36. Qg4 Re5 37. h4 Ke7 38. Rd2 b5 39. Bxf7 Qf5 40.
Rxd7+ Kxd7 41. Qxf5+ Rxf5 42. g6 Ke7 43. cxb5 Rh5 44. c4 Rxh4 45. a4 Rg4 46. a5
Kd6 47. a6 Kc7 48. Kb3 1-0[/pgn]
yanquis1972
Posts: 1766
Joined: Wed Jun 03, 2009 12:14 am

Re: Karjakin 17.Bxd5 !

Post by yanquis1972 »

unassisted analysis of 20...f5 (same move order as in game)

48: Karjakin,S - Caruana,F 1-0, World Chess Candidates 2018
[d]2krr3/ppq1b1p1/7p/2pB1p2/2P2P2/P3B3/1PP2QPP/2KR4 w - - 0 1

Analysis by Stockfish 210318 64 BMI2:

21.h3 b6 22.g4 g6 23.Rd3 Bf6 24.b4 Be7 25.Rb3 Bf8 26.Kb1 cxb4 27.axb4 Rxd5 28.cxd5 Re4 29.gxf5 gxf5 30.Qg2 Qf7 31.Qd2 Qf6 32.b5 Bd6 33.Qd3 Kd7 34.c4 Bxf4 35.Bxf4 Rxf4 36.Kc2 Re4
White has an edge: = (0.30) Depth: 13/13 00:00:00 60kN
The position is equal: = (0.12) Depth: 26/42 00:00:03 35221kN, tb=2
21.Kb1 b6 22.Qf3 Kb8 23.Rd3 Bf6 24.Qh5 Rf8 25.g3 Qc8 26.Rb3 Rd6 27.Qf3 Rfd8 28.a4 a5 29.Rb5 Ka7 30.Bxc5 bxc5 31.Rb7+ Qxb7 32.Bxb7 Rb6 33.Bd5 Bxb2 34.Ka2 Re8 35.c3 Re7 36.Qd3 g6 37.Qf3
The position is equal: = (0.20 ++) Depth: 27/36 00:00:05 52409kN, tb=2
The position is equal: = (0.20 ++) Depth: 29/41 00:00:16 192MN, tb=3
21.h3 b6 22.g4 g6 23.Qf3 Bf6 24.Kb1 Rd6 25.gxf5 gxf5 26.Bf2 Kb8 27.Rg1 Red8 28.Qb3 Kc8 29.Rd1 Re8 30.Qd3 Qd7 31.Rg1 Rd8 32.Qf3 Rxd5 33.cxd5 Qxd5 34.Qxd5 Rxd5 35.Rg6 Be7 36.Rc6+ Kd7 37.Rxh6 Rd6
The position is equal: = (0.22) Depth: 29/48 00:00:25 294MN, tb=82
21.Kb1 b6 22.Rd3 Bf6 23.Qf3 Kb8 24.h3 a6 25.g4 fxg4 26.hxg4 Be7 27.Ka2 Ka7 28.Bf2 Bf6 29.b4 Qe7 30.g5 hxg5 31.bxc5 bxc5 32.fxg5 Be5 33.Re3 Qd6 34.Bg3 Bxg3 35.Qf7+ Qc7 36.Rxe8 Qxf7 37.Bxf7 Rxe8 38.Bxe8 Bd6 39.Bd7 g6 40.Be8 Bf4 41.Bxg6 Bxg5 42.Kb3 Kb6 43.Be4
White has an edge: = (0.30 ++) Depth: 30/48 00:00:30 355MN, tb=84
The position is equal: = (0.16 --) Depth: 39/51 00:03:05 2313MN, tb=1553
21.Rd3 b6 22.Kb1 Kb8 23.h3 a6 24.g4 fxg4 25.hxg4 Ka7 26.Qg2 Bh4 27.Ka2 Bf6 28.Qf3 Qe7 29.Qd1 Qc7 30.Qg1 Re7 31.Qf1 Ree8 32.Qf2 Rb8 33.Qg1 Rbc8 34.Qh1 Rf8 35.Qg2 Rce8 36.Qg1 Rd8 37.Qe1 Rfe8 38.Qd1 Qe7 39.Qf3 Rb8 40.Rb3 Rbc8 41.Qg2 Rcd8 42.Bc6 Rf8 43.a4 Bd4 44.Bd2 Qd6 45.Qf3
White has an edge: = (0.24 ++) Depth: 39/52 00:03:19 2510MN, tb=2187
White is slightly better: +/= (0.31) Depth: 48/75 01:02:13 50833MN, tb=596606

iirc that's only ~10 centipawns higher than its eval at the root at shallow depths in mpv; don't know that it's enough to put Bxd5 first.

imo contempt should give Bxd5 enough of a bonus to be the selected move (if it doesn't already) but objectively not sure it's best, just based on that output...