London Chess Classic 2016

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

Moderators: hgm, Rebel, chrisw

yanquis1972
Posts: 1766
Joined: Wed Jun 03, 2009 12:14 am

Re: London Chess Classic 2016

Post by yanquis1972 »

2:05.41 13.cxb5 (depth 40, 1/47), 58.5G nodes (7.7M nps), 100.5K tbhits

2:01.42: +0.21/38 13.cxb5 Na5 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.b4 fxg5 18.bxc5 Bf5 19.Bd3 Rab8 20.a4 a6 21.Bxf5 axb5 22.Nxd4 bxa4+ 23.Ka1 Rb4 24.Bxh7+ Kf8 25.Qc3 Rxd4 26.Bc2 Nc6 27.Rxd4 Nxd4 28.Rd1 Nxc2+ 29.Qxc2 Qe5+ 30.Qb2 Qxb2+ 31.Kxb2 Re2+ 32.Ka3 Rxf2 33.g3 Rxh2 34.Kxa4 Ke7 35.c6 Rc2 36.Kb5 f5 37.Kb6 f4 38.Rd3 Rb2+ 39.Kc7 Rc2 40.Rb3 Ke6 41.gxf4 gxf4
2:05.41: +0.37/39 13.cxb5 Na5 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Be5 20.bxa5 Bxd4 21.Nxd4 Qc7 22.Kb2 Qxc4 23.Qxc4 Rxc4 24.Rd1 Rb8 25.Rd3 Bxb5 26.Rb3 a6 27.Nf5 Kh8 28.Bg7+ Kg8 29.Bxf6 Rf4 30.Rg3+ Kf8 31.Be7+ Ke8 32.Rg8+ Kd7 33.Rxb8 Rxf2+ 34.Kb3 Rxf5 35.Bb4 Rf2 36.Rh8 Rxg2 37.Rxh7 Ke6 38.Be1 Bd3 39.Rh8 Bf5 40.Rh6+ Bg6 41.Kb4
User avatar
Eelco de Groot
Posts: 4567
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: London Chess Classic 2016

Post by Eelco de Groot »

Stockfish is slow to pick Na5 as response to 13. cxb5. Only after one hour on your PC. 13. cxb5 Qxd5 leads to complications that look a bit bleak for Black. After 13... Na5 there is not much to fear I think:


[D]r1bqr1k1/p4ppp/5n2/nPbNp1B1/3p4/P3PN2/1PQ2PPP/2KR1B1R w - -

38/90 95:53 +0.32 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6
17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Be5
20.bxa5 Bxd4 21.Nxd4 Qc7 22.Kb2 Qxc4
23.Qxc4 Rxc4 24.Be3 Rb8 25.Kb3 Rxd4
26.Bxd4 Rxb5+ 27.Kc4 (58.127.938.814) 10103

But if you follow the annoted line from the game after 13... Qxd5 14. Bc4 Qd6 15. bxc6

Now it looks to equalize

[D]r1b1r1k1/p4ppp/2Pq1n2/2b1p1B1/2Bp4/P3PN2/1PQ2PPP/2KR3R b - -

Engine: Rainbow Serpent 20160903 HT (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G.

32 20:00 +1.14 15...Be6 16.Bxe6 Rxe6 17.Bxf6 gxf6
18.exd4 exd4 19.Kb1 Rc8 20.Qf5 Qxc6
21.Rd2 Bb6 22.Nxd4 Bxd4 23.Rxd4 Re5
24.Qd7 Qc2+ 25.Ka1 Rce8 26.Rg4+ Kh8
27.Qd4 Qf5 28.Rd1 a5 (12.107.068.917) 10084

32 20:00 +1.22 15...Bg4 16.exd4 Bxd4 17.Nxd4 exd4
18.f3 Be6 19.Bxe6 Rxe6 20.Kb1 Qd5
21.c7 Rc8 22.Bxf6 Rxf6 23.Rc1 Rd6
24.Rhd1 g6 25.Qc5 Qxc5 26.Rxc5 Kf8
27.b4 Ke8 28.Kb2 Kd7 (12.107.068.917) 10084
____________________________________________________

33 51:34 +0.94 15...Bg4 16.exd4 Bxd4 17.Nxd4 exd4
18.f3 Bh5 19.Kb1 Bg6 20.Bd3 Rac8
21.Bxg6 hxg6 22.Rc1 Nd5 23.Rhd1 Qe5
24.h4 Ne3 25.Bxe3 dxe3 26.Qc4 Rc7
27.Rd4 Qf6 28.Re4 Rec8 (30.977.814.059) 10010

33 51:34 +0.98 15...Be6 16.Bxe6 fxe6 17.Bxf6 gxf6
18.Qe4 Red8 19.Rd2 Rac8 20.Kb1 Rxc6
21.exd4 exd4 22.Nxd4 f5 23.Qf3 Bxd4
24.Rxd4 Qxd4 25.Qxc6 Kf7 26.Re1 Rd6
27.Qb7+ Rd7 28.Qa6 Rd6 (30.977.814.059) 10010

This first line at depth 33, with 15... Bg4, is also the coice from the annotator from Chessbase(? Other thread about this position) After 19 Kb1 Black seems not out of trouble though if you look slightly deeper here. I gave White a {+/=} as an added comment But of course Black can play it safer with Na5, avoiding most of the complications mentioned in the notes.

[pgn]
[Event "?"]
[Site "?"]
[Date "2016.12.18"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]

1. d4 Nf6 2. c4 e6 3. Nf3 d5 4. Nc3 Be7 5. Bf4 O-O 6. e3 c5
{This line of the Queen's Gambit Declined was heavily
explored in the 1978 Karpov-Kortschnoi match in Bagyo city.
Nowadays it is played practically by all the top players.}
7. dxc5 Bxc5 8. a3 (8. Qc2 {at once is the other main
move.}) 8... Nc6 9. Qc2 Re8 {A move that Anand had to face
in his second match against Carlsen in Sochi, 2014.}
(9... Qa5 {reamins the main line, for example} 10. Rd1 Re8
11. Nd2 e5 12. Bg5 Nd4 13. Qa4 Qxa4 14. Nxa4 Nc2+ 15. Ke2
Nd4+ 16. Ke1 Nc2+ 17. Ke2 Nd4+ 18. Ke1 Nc2+ {1/2-1/2
Karjakin,S (2773)-Nakamura,H (2787) Bilbao 2016}) 10. O-O-O
{The aforementioned game went} (10. Bg5 Be7 11. Rd1 Qa5
12. Bd3 h6 13. Bh4 dxc4 14. Bxc4 a6 {and Black was close to
equality, Anand,V (2792) -Carlsen,M (2863) Sochi 2014})
10... e5 11. Bg5 d4 12. Nd5 {All of this was blitzed out
until Anand uncorked the strong novelty} b5 $1 {Prior to
this game Black was sacrificing a pawn for the initiative
with:} (12... Be7 13. Nxe7+ Qxe7 14. exd4 Nxd4 15. Nxd4
exd4 16. Rxd4 Qe5 {Lalith,B (2569)-Thejkumar,M (2452)
Jalgaon 2013}) 13. Bxf6 $2 {Perhaps White could try }
(13. cxb5 Na5 (13... Qxd5 14. Bc4 {the next time this
position occurs on the board. But he has to be heavily
prepared for the arising complications.} Qd6 15. bxc6 Bg4
16. exd4 Bxd4 17. Nxd4 exd4 18. f3 Bh5 19. Kb1 {$14 see diagram below})
14. exd4 (14. Bc4) 14... exd4 15. Nxf6+ gxf6) 13... gxf6
14. cxb5 Na5 $1 15. exd4 {Obviously not} (15. Qxc5 $4 Nb3+)
(15. b4 Qxd5 16. bxc5 Nb3+ {leaves the white king badly
exposed.}) 15... exd4 16. Nb4 {Pretty much the only move as
otherwise Black clears the c-file for his heavy pieces
amazingly fast:} (16. Nf4 Bd6 17. Nd3 Be6 {and Black has
strong threats.}) (16. Ne3 $5 Be6 17. Nc4 Rc8 {looks
unpleasant as well.}) 16... Bxb4 17. axb4 Be6 $1 {The
opening was huge success for Black. He opened files on the
queenside and has an obvious attack.} 18. Nxd4 {There is
hardly a better choice:} (18. bxa5 Rc8 {drops the queen.})
(18. Kb1 Qd5 19. Qa4 Nb3 {both lead to problems on the
queenside.}) 18... Rc8 19. Nc6 Nxc6 20. bxc6 Qb6 21. Qa4
{White cannot stich the position on the queenside with}
(21. b5 {due to} a6) (21. Kb1 {was relatively best and
after} Rxc6 22. Qa4 {would transpose into the game.})
21... Rxc6+ {A slip. The engines discovered the clever}
(21... Bg4 $1 {to lure the pawn on} 22. f3 {after which the
e3 square is weaken and Black wins after} Bf5 23. Bd3 Qe3+)
22. Kb1 Rd8 23. Rxd8+ Qxd8 24. Be2 {Somewhere around here
Topalov started to defend "exceptionally well" - Anand.}
Bf5+ 25. Ka2 Rc2 (25... Qd6 $1 {was better and only after}
26. Ka1 (26. Rd1 Qe6+) 26... Rc2) 26. Rd1 Qb6 {At first
Anand wanted to put the queen on c7 and he spent a lot of
time calculating the mind-boggling line:} (26... Qc7
27. Qe8+ Kg7 28. Rd8 Be6+ 29. Ka3 Qc3+ 30. Ka4 (30. bxc3
Ra2+) 30... Qb3+ 31. Ka5 Rxe2 32. Qh8+ Kh6 33. Qxf6+ Kh5
34. Rd4 {This is where he gave up calculating, but the
computer prolonged the line with} Re4 $1 {although nothing
is yet clear after} 35. f4 $3 {where both kings are under
fire.} (35. Rxe4 Qa2+ $1 {Much better than} (35... Qd5+ $6
36. b5 Qxe4 37. f3) 36. Kb5 Qd5+ {and Black wins.}))
27. Bg4 $1 Qe6+ (27... Bxg4 28. Qxc2 Bxd1 29. Qxd1 Qxb4)
28. Ka3 (28. Ka1 Qe5 {does not look healthy for White.})
28... Qe5 29. Qb3 Bg6 30. Bf3 {Perhaps Topalov was already
calculating the line} (30. h4 {to trouble the black bishop}
Rxf2 31. Rd5 Qf4 32. Bf3 {and this was the reason why he
went for h2-h4 on the next move.}) 30... Rxf2 31. h4 {or
the move that he suggested himself} (31. Rd5) (31. Rc1 {In
both cases White's position remains worrysome, but full of
life.}) 31... Bc2 $1 32. Rd8+ Kg7 33. Qc3 $2 {This loses on
the spot. White would be still in the game after} (33. Qc4
Qe3+ 34. Ka2 Qc1 35. Qg4+ Bg6 36. Qd4 Qb1+ (36... Bb1+)
37. Kb3 Bc2+ {although it is obvious who is on the
attack.}) 33... Qb5 34. Qc6 Rxf3+ $1 {The move that White
missed.} (34... Rxf3+ 35. gxf3 (35. Qxf3 Qa4+) 35... Qxc6) *
[/pgn]

[D]r3r1k1/p4ppp/2Pq1n2/6Bb/2Bp4/P4P2/1PQ3PP/1K1R3R b - -

28 2:32 +1.18 19...Rac8 20.Ka1 h6 21.Bxf6 Qxf6
22.Bd5 Bg6 23.Qc4 d3 24.Rd2 Red8
25.f4 Qf5 26.Bf3 Qa5 27.Rhd1 Qb6
28.Bd5 Rxd5 29.Qxd5 Rxc6 30.Re1 Rc2
31.Rxc2 dxc2 32.f5 Qe3 (1.464.042.170) 9590

28 2:32 +1.18 19...Bg6 20.Bd3 Rac8 21.Bxg6 hxg6
22.Qc4 Rxc6 23.Rxd4 Qc7 24.Qd3 Rb6
25.Bc1 Qb8 26.Ka1 Kh7 27.h4 Rb3
28.Qc2 Rc8 29.Rc4 Rxc4 30.Qxc4 Qb7
31.Rd1 Rb6 32.Ka2 a5 (1.464.042.170) 9590

__________________________________________________________________

29 3:41 +1.10 19...Rac8 20.Ka1 h6 21.Bxf6 Qxf6
22.Bd5 Bg6 23.Qc4 d3 24.Rd2 Rc7
25.f4 h5 26.g3 Bf5 27.h3 Re3 28.g4 Be6
29.Rxd3 Rxd3 30.Qxd3 Bxd5 31.Qxd5 Rxc6
32.Rd1 Rb6 (2.128.660.616) 9628

29 3:41 +1.12 19...Bg6 20.Bd3 Rac8 21.Bxg6 hxg6
22.Qc4 Rxc6 23.Rxd4 Qc7 24.Qd3 Rb6
25.Bc1 Qb8 26.Ka1 Rbe6 27.g4 Qb6
28.Rd1 Re1 29.Ka2 Rxd1 30.Qxd1 Re1
31.Qxe1 Qxd4 32.Qe3 Qd5+ (2.128.660.616) 9628
__________________________________________________________________

30 4:39 +1.10 19...Rac8 20.Ka1 h6 21.Bxf6 Qxf6
22.Bd5 Bg6 23.Qc4 d3 24.Rd2 Rc7
25.f4 h5 26.Rc1 Bf5 27.Bf3 g6 28.Qb4 Kh7
29.h3 h4 30.Rdd1 Kg7 31.Rc4 Re6
32.Qc5 Rd6 (2.690.979.335) 9632

30 4:39 +1.12 19...Bg6 20.Bd3 Rac8 21.Bxg6 hxg6
22.Qc4 Rxc6 23.Rxd4 Qc7 24.Qd3 Rb6
25.Bc1 Qb8 26.Ka1 Rbe6 27.g4 Qb7
28.g5 Nh5 29.Bd2 Re2 30.Rc1 R2e7
31.f4 a6 32.f5 gxf5 (2.690.979.335) 9632
__________________________________________________________________

31 8:31 +1.14 19...Rac8 20.Ka1 h6 21.Bxf6 Qxf6
22.Bd5 Bg6 23.Qc4 d3 24.Rd2 Rc7
25.g3 Qd6 26.f4 Bf5 27.Rc1 Kf8 28.h3 Re2
29.Rcd1 Rxd2 30.Rxd2 Be6 31.Rxd3 Bxd5
32.Rxd5 Qxc6 (4.977.628.989) 9726

31 8:31 +1.20 19...Bg6 20.Bd3 Rac8 21.Bxg6 hxg6
22.Qc4 Rxc6 23.Rxd4 Qc7 24.Qd3 Rb6
25.Bc1 Qb7 26.Rd8 Rbe6 27.Rxe8+ Rxe8
28.Rd1 Qb6 29.Qc4 Qf2 30.Rd2 Qb6
31.h3 Rd8 32.Rxd8+ Qxd8 (4.977.628.989) 9726
_________________________________________________________________

32 13:46 +1.20 19...Bg6 20.Bd3 Rac8 21.Bxg6 hxg6
22.Qc4 Rxc6 23.Rxd4 Qc7 24.Qd3 Rb6
25.Bc1 Qb7 26.Rd8 Rbe6 27.Rxe8+ Rxe8
28.Rd1 Qb6 29.Qc4 Rd8 30.Rxd8+ Qxd8
31.b3 Qd1 32.Qc2 Qd6 (8.090.643.569) 9793

32 13:46 +1.27 19...Rac8 20.Ka1 h6 21.Bxf6 Qxf6
22.Qa4 Red8 23.Bb5 Rd5 24.Ba6 Rc7
25.Bb7 Qd6 26.Qxa7 Bg6 27.Rd2 d3
28.Qf2 Re5 29.Rhd1 Rce7 30.Qg3 Qd5
31.Ba6 Qxc6 32.Bxd3 Bxd3 (8.090.643.569) 9793
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
yanquis1972
Posts: 1766
Joined: Wed Jun 03, 2009 12:14 am

Re: London Chess Classic 2016

Post by yanquis1972 »

i agree Na5 is best, but its hardly losing for white. H5 eval w/ some interaction is also around the ~.25 mark (though dropping; i'll try to leave it going till the next round starts)

[d]r1bqr1k1/p4ppp/5n2/nPbNp1B1/3p4/P3PN2/1PQ2PPP/2KR1B1R w - - 1 14


4.24 14.exd4 (depth 30, 1/50), 3.8G nodes (14.6M nps), 246 tbhits
0.00: +0.26/18 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.Bxf6 Be6 18.b3 Qxb3+ 19.Qxb3 Bxb3 20.Rc1 Bxa3 21.Rc7 a6 22.b6 Bd5 23.Bxd4 Rab8 24.h4 Nc6 25.Be3 Rxe3 26.fxe3 Rxb6+ 27.Kc2 Kg7
0.00: +0.41/19 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Nxc4 20.Rxc4 Qb6 21.Nd4 a6 22.Kb1 Rxc4 23.Qxc4 Be5 24.Rd1 axb5 25.Qd3 Qd6 26.h3 Rd8 27.f4 Bxf4 28.Bxf4 Qxf4 29.Nc2
0.00: +0.58/20 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Nxc4 20.Rxc4 Qb6 21.Nd4 a6 22.Kb1 Rxc4 23.Qxc4 Be5 24.Rd1 axb5 25.Qd3 Rd8 26.f4 Bg4 27.Qg3 f5 28.fxe5 Rxd4 29.Rxd4 Qxd4
0.01: +0.53/21 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Be5 20.bxa5 Bxd4 21.Nxd4 Qc7 22.Kb2 Qxc4 23.Qxc4 Rxc4 24.Be3 Rb8 25.Kb3 Rc5 26.a4 Rd5 27.Rc1 a6 28.Bf4 Re8
0.02: +0.60/22 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Be5 20.bxa5 Bxd4 21.Nxd4 Qc7 22.Kb2 Qxc4 23.Qxc4 Rxc4 24.Be3 Rb8 25.Kb3 Rbc8 26.b6 Ra4 27.Ka2 axb6 28.axb6 Rb8 29.Rb1 Rb7
0.04: +0.60/23 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Be6 18.Kb2 Rc8 19.Qe4 Nc4+ 20.Bxc4 Rxc4 21.Rc1 f5 22.Qd3 Rxc1 23.Kxc1 Qf6 24.Bd2 Be5 25.Nxe5 Qxe5 26.Qg3+ Qxg3 27.hxg3 Bb3 28.Rh4 Rd8 29.Bf4
0.10: +0.26/24 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.Bxf6 Be6 18.b3 Qxb3+ 19.Qxb3 Bxb3 20.Rd2 Rac8 21.Bd3 Nc4 22.Re2 Rxe2 23.Bxe2 Nxa3+ 24.Kb2 Ba4 25.Nxd4 Nxb5 26.Nxb5 Bxb5 27.Bxb5 Rb8 28.Kc2 Rxb5 29.Re1 Kf8 30.f4 Rb6 31.Bc3 Bb4 32.Bxb4+ Rxb4 33.g3 Rc4+ 34.Kd3 Ra4 35.Re2 Kg7 36.Kc3 Ra6 37.Kd4
0.12: +0.32/25 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.b4 Bb6 18.Bd3 fxg5 19.bxa5 g4 20.axb6 gxf3 21.gxf3 axb6 22.Bxh7+ Kf8 23.Be4 Qxb5+ 24.Qb2 Qxb2+ 25.Kxb2 Ra4 26.Rc1 b5 27.h4 b4 28.Kb3 Rxa3+ 29.Kxb4 Ra2 30.h5 Rd8 31.h6 Kg8 32.Rhd1 Rxf2 33.Kc5 Be6 34.Ra1
0.14: +0.31/26 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.b4 Bb6 18.Bd3 fxg5 19.bxa5 g4 20.axb6 gxf3 21.gxf3 axb6 22.Bxh7+ Kf8 23.Be4 Qxb5+ 24.Qb2 Qxb2+ 25.Kxb2 Ra4 26.Rc1 b5 27.h4 b4 28.Kb3 Rxa3+ 29.Kxb4 Ra2 30.h5 Rd8 31.h6 Kg8 32.Rhd1 Rxf2 33.Kc5 Be6 34.Ra1 f5 35.Ra8 Rxa8 36.Bxa8 Rh2
0.21: +0.25/27 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.b4 Bb6 18.Bd3 fxg5 19.bxa5 g4 20.axb6 gxf3 21.gxf3 axb6 22.Bxh7+ Kf8 23.Be4 Qxb5+ 24.Qb2 Qxb2+ 25.Kxb2 Ra4 26.Rc1 b5 27.h4 b4 28.axb4 Rxb4+ 29.Kc2 Bb7 30.Bxb7 Rxb7 31.h5 Kg7 32.h6+ Kh7 33.Rcg1 d3+ 34.Kd2 f6 35.Kxd3 Rd7+ 36.Kc4 Rc8+ 37.Kb5 Rc2 38.f4 Rd5+ 39.Ka6
0.26: +0.49/28 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.b4 Bb6 18.Bd3 fxg5 19.bxa5 g4 20.axb6 gxf3 21.gxf3 axb6 22.Bxh7+ Kf8 23.Be4 Qxb5+ 24.Qb2 Qxb2+ 25.Kxb2 Ra4 26.Rc1 b5 27.h4 b4 28.Bc6 Re2+ 29.Kb3 Rxa3+ 30.Kxb4 Ra7 31.Rhd1 Bd7 32.Rxd4 Bxc6 33.Rxc6 Rxf2 34.Rf4 Rb2+ 35.Kc3 Rb5 36.Kd3 Ra8 37.Rc3
1.08: +0.54/29 14.exd4 exd4 15.Nxf6+ gxf6 16.Kb1 Qd5 17.b4 Bb6 18.Bd3 fxg5 19.bxa5 g4 20.axb6 gxf3 21.gxf3 axb6 22.Bxh7+ Kf8 23.Be4 Qxb5+ 24.Qb2 Qxb2+ 25.Kxb2 Ra4 26.Rc1 b5 27.h4 b4 28.Bc6 Re2+ 29.Kb3 Rxa3+ 30.Kxb4 Ra7 31.Be4 Be6 32.h5 Rxf2 33.Rc6 Rb7+ 34.Kc5 d3 35.h6 Rb8 36.h7 Kg7 37.Rd6 Kh8 38.Kd4 Ra2 39.Ke3 d2 40.Rxd2 Rxd2 41.Kxd2
3.42: +0.23/30 14.exd4 exd4
yanquis1972
Posts: 1766
Joined: Wed Jun 03, 2009 12:14 am

Re: London Chess Classic 2016

Post by yanquis1972 »

stopped trending down so ill flick it off, last line
9.27: +0.74/30 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Nxc4 20.Rxc4 Qb6 21.Nd4 a6 22.Kb1 Rxc4 23.Qxc4 Be5 24.Rd1 axb5 25.Qd3 Bg4 26.Rd2 Rd8 27.Qe4 Bh5 28.Nf5 Rxd2 29.Bxd2 Bg6 30.f4 Bd6 31.Kb2 Bf8 32.g3 Qe6


pretty confident whites just fine
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: London Chess Classic 2016

Post by Lyudmil Tsvetkov »

yanquis1972 wrote:stopped trending down so ill flick it off, last line
9.27: +0.74/30 14.exd4 exd4 15.Nxf6+ gxf6 16.b4 Bd6 17.Bh6 Bd7 18.Rxd4 Rc8 19.Bc4 Nxc4 20.Rxc4 Qb6 21.Nd4 a6 22.Kb1 Rxc4 23.Qxc4 Be5 24.Rd1 axb5 25.Qd3 Bg4 26.Rd2 Rd8 27.Qe4 Bh5 28.Nf5 Rxd2 29.Bxd2 Bg6 30.f4 Bd6 31.Kb2 Bf8 32.g3 Qe6


pretty confident whites just fine
try this one: 16.b4 fg5 17. bc5 Bb7:

[d]r2qr1k1/pb3p1p/8/nPP3p1/3p4/P4N2/2Q2PPP/2KR1B1R w - - 0 5

and white should get worse, though very complicated

black proceeds with Rc8 and a6, leaving the white king completely bare
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: London Chess Classic 2016

Post by Lyudmil Tsvetkov »

current tops can not quite judge the moves made by top GMs.

they can go all the way to judging the moves of some 2300-2400 elo player, but not higher.

arithmetics is very simple:

current tops: 2300 elo positional play 4300 elo tactical play for an overall rating of 3300
top humans: 2900 elo positional play 2900 elo tactical play for an overall rating of 2900

as most moves are quiet positional, engines can not judge them; still, they will possibly prevail in a match, due to better tactics, machine-related vs human factors, etc.

15 years ago, Fritz used to be 1800 elo positionally and 3800 tactically for an overall elo of 2800, matching the best humans overall.

so top engines made a nice progress positionally, some 500 elo from 1800 to 2300, roughly half of the 50 elo yearly progress, though this 2300 positional mark is still very low to understand top GMs. and positional play means, of course, besides deeper search, first and foremost knowledge, which current tops do not quite possess.

above, Houdini misses key move Na5, possibly an only GM choice, does not see fg5, etc.

what concerns SF, it would static-evaluate above node with white edge, due to also scoring advanced b5 and c5 white pawns as king shelter, but what shelter can be pawns on the 5th rank be? A strong player would eaily see white is in trouble; even if white manages to draw this line and that, some 70% of all lines should be black wins.
yanquis1972
Posts: 1766
Joined: Wed Jun 03, 2009 12:14 am

Re: London Chess Classic 2016

Post by yanquis1972 »

engines don't miss Na5, or any of the following. SF takes some time to come around to Na5, but it still evaluates white as slightly better. similarly the line you propose is just evaluated as inferior by houdini:

[d]r2qr1k1/pb3p1p/8/nPP3p1/3p4/P4N2/2Q2PPP/2KR1B1R w - - 1 18


2.28 18.Nxd4 (depth 28, 1/39), 1.6G nodes (11.2M nps), 149 tbhits
0.00: +0.87/18 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rc3 25.hxg5 Nc4 26.Bxc4 Rxc4 27.Rc1 Rac8 28.Rxc4 Rxc4 29.Rh6
0.00: +0.99/19 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rh8 25.hxg5 Rxh1 26.Rxh1 Rxg5 27.Rh7+ Kg8 28.Rh4 Nb3 29.Kc2 Rc5+ 30.Kxb3 Rc3+ 31.Kb4 Rxd3
0.01: +1.02/20 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rc3 25.hxg5 Nc4 26.Bxc4 Rxc4 27.Rc1 Rac8 28.Rxc4 Rxc4 29.Rh6 Rc3 30.Kb2 Rxf3 31.Ra6 Rxf2+ 32.Kb3
0.02: +1.03/21 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rc3 25.hxg5 Nc4 26.Bxc4 Rxc4 27.Rc1 Rac8 28.Rxc4 Rxc4 29.Rh6 Rc5 30.a4 d3 31.Rd6 Rc4 32.Kb2
0.23: +0.73/22 18.Bd3 Bxf3 19.gxf3 a6 20.f4 Kh8 21.Rhe1 axb5 22.Rxe8+ Qxe8 23.Qb2 Kg8 24.Bc2 Nc6 25.Qb3 Kf8 26.Qh3 Qe6 27.Qxe6 fxe6 28.Be4 Rc8 29.fxg5 Na5 30.Kb1 Rxc5 31.h4 Nc4 32.Rxd4
0.32: +0.85/23 18.Bd3 Bxf3 19.gxf3 a6 20.f4 Qc7 21.b6 Qxc5 22.Qxc5 Nb3+ 23.Kc2 Nxc5 24.Rb1 Rac8 25.Kd2 gxf4 26.Rhg1+ Kh8 27.Rg4 Rb8 28.Rxf4 Nxd3 29.Kxd3 Re6 30.Rxf7 Rexb6 31.Rxb6 Rxb6 32.Kxd4 Rb3 33.Ra7
0.38: +0.78/24 18.Bd3 Bxf3
1.00: +0.91/24 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Rd4 Rd8 25.Rxg4+ Kf8 26.Nxb7 Rxb7 27.Bc4 a6 28.Rc1 axb5 29.Bf1 Nd3+ 30.Bxd3 Rxd3 31.Rc3 Rd2+ 32.Rc2 Rd3 33.Rb4 Ra7
1.09: +0.96/25 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Nxb7 Rxb7 25.Rd4 g3 26.fxg3 Re8 27.Bc4 Re3 28.Rh3 Ne6 29.Rd5 Kg7 30.Bd3 Nc5 31.Rg5+ Kf8 32.Rxc5 Rxd3
1.28: +0.96/26 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Rd4 Rd8 25.Rxg4+ Kf8 26.Nxb7 Nxb7 27.Ra4 Nc5 28.Rc4 Ne4 29.Rc2 Rd1 30.g3 Re5 31.Bg2 Rd4 32.Ka2 Rxb5 33.Rb1 Rxb1 34.Kxb1 Rd1+ 35.Kb2
2.07: +1.04/27 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Nxb7 Rxb7 25.Rd4 g3 26.fxg3 Re8 27.Bc4 Re3 28.g4 Rg3 29.Rh3 Rxg2+ 30.Kc3 Rc7 31.Kb4 Rb2+ 32.Bb3 Ne6 33.Rd6 Rf2 34.Bxe6 Rf4+
2.21: +1.11/28 18.Nxd4
kgburcham
Posts: 2016
Joined: Sun Feb 17, 2008 4:19 pm

Re: London Chess Classic 2016

Post by kgburcham »

arithmetics is very simple:

current tops: 2300 elo positional play 4300 elo tactical play for an overall rating of 3300
top humans: 2900 elo positional play 2900 elo tactical play for an overall rating of 2900
you forgot to deduct for human blunders.

current programs: 3300 elo positional play 3300 elo tactical play for an overall rating of 3300
top humans: 2700 elo positional play 2700 elo tactical play for an overall rating of 2700 minus blunder factor = overall 2500

humans are way over rated at the top level
no chess program was born totally from one mind. all chess programs have many ideas from many minds.
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: London Chess Classic 2016

Post by Lyudmil Tsvetkov »

kgburcham wrote:you forgot to deduct for human blunders.

current programs: 3300 elo positional play 3300 elo tactical play for an overall rating of 3300
top humans: 2700 elo positional play 2700 elo tactical play for an overall rating of 2700 minus blunder factor = overall 2500

humans are way over rated at the top level
:) :) :)

great, Kim, great.

top engines at 3300 positionally? when they reach 3300 positionally, their cumulative rating will be somewhere in the range of 5500-6000 elo, that is for certain.

you can not have 3300 elo positionally with less than 3000 eval terms, that is also quite certain. current tops are reaching 300-500 eval terms at most, and that considering sub-terms/psqt, etc. no knowledge, you understand?

where you are correct is that humans blunder a lot, way lot than machines, but that is due mainly to human factors, not so much the low level of play. 2500 is a bit low for Carlsen, do not you think?
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: London Chess Classic 2016

Post by Lyudmil Tsvetkov »

yanquis1972 wrote:engines don't miss Na5, or any of the following. SF takes some time to come around to Na5, but it still evaluates white as slightly better. similarly the line you propose is just evaluated as inferior by houdini:

[d]r2qr1k1/pb3p1p/8/nPP3p1/3p4/P4N2/2Q2PPP/2KR1B1R w - - 1 18


2.28 18.Nxd4 (depth 28, 1/39), 1.6G nodes (11.2M nps), 149 tbhits
0.00: +0.87/18 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rc3 25.hxg5 Nc4 26.Bxc4 Rxc4 27.Rc1 Rac8 28.Rxc4 Rxc4 29.Rh6
0.00: +0.99/19 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rh8 25.hxg5 Rxh1 26.Rxh1 Rxg5 27.Rh7+ Kg8 28.Rh4 Nb3 29.Kc2 Rc5+ 30.Kxb3 Rc3+ 31.Kb4 Rxd3
0.01: +1.02/20 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rc3 25.hxg5 Nc4 26.Bxc4 Rxc4 27.Rc1 Rac8 28.Rxc4 Rxc4 29.Rh6 Rc3 30.Kb2 Rxf3 31.Ra6 Rxf2+ 32.Kb3
0.02: +1.03/21 18.Bd3 Bxf3 19.gxf3 Re5 20.Kb1 Qe7 21.Bxh7+ Kg7 22.Bd3 Qxc5 23.Qxc5 Rxc5 24.h4 Rc3 25.hxg5 Nc4 26.Bxc4 Rxc4 27.Rc1 Rac8 28.Rxc4 Rxc4 29.Rh6 Rc5 30.a4 d3 31.Rd6 Rc4 32.Kb2
0.23: +0.73/22 18.Bd3 Bxf3 19.gxf3 a6 20.f4 Kh8 21.Rhe1 axb5 22.Rxe8+ Qxe8 23.Qb2 Kg8 24.Bc2 Nc6 25.Qb3 Kf8 26.Qh3 Qe6 27.Qxe6 fxe6 28.Be4 Rc8 29.fxg5 Na5 30.Kb1 Rxc5 31.h4 Nc4 32.Rxd4
0.32: +0.85/23 18.Bd3 Bxf3 19.gxf3 a6 20.f4 Qc7 21.b6 Qxc5 22.Qxc5 Nb3+ 23.Kc2 Nxc5 24.Rb1 Rac8 25.Kd2 gxf4 26.Rhg1+ Kh8 27.Rg4 Rb8 28.Rxf4 Nxd3 29.Kxd3 Re6 30.Rxf7 Rexb6 31.Rxb6 Rxb6 32.Kxd4 Rb3 33.Ra7
0.38: +0.78/24 18.Bd3 Bxf3
1.00: +0.91/24 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Rd4 Rd8 25.Rxg4+ Kf8 26.Nxb7 Rxb7 27.Bc4 a6 28.Rc1 axb5 29.Bf1 Nd3+ 30.Bxd3 Rxd3 31.Rc3 Rd2+ 32.Rc2 Rd3 33.Rb4 Ra7
1.09: +0.96/25 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Nxb7 Rxb7 25.Rd4 g3 26.fxg3 Re8 27.Bc4 Re3 28.Rh3 Ne6 29.Rd5 Kg7 30.Bd3 Nc5 31.Rg5+ Kf8 32.Rxc5 Rxd3
1.28: +0.96/26 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Rd4 Rd8 25.Rxg4+ Kf8 26.Nxb7 Nxb7 27.Ra4 Nc5 28.Rc4 Ne4 29.Rc2 Rd1 30.g3 Re5 31.Bg2 Rd4 32.Ka2 Rxb5 33.Rb1 Rxb1 34.Kxb1 Rd1+ 35.Kb2
2.07: +1.04/27 18.Nxd4 Qf6 19.h4 g4 20.Qf5 Qxf5 21.Nxf5 Nb3+ 22.Kb2 Nxc5 23.Nd6 Re7 24.Nxb7 Rxb7 25.Rd4 g3 26.fxg3 Re8 27.Bc4 Re3 28.g4 Rg3 29.Rh3 Rxg2+ 30.Kc3 Rc7 31.Kb4 Rb2+ 32.Bb3 Ne6 33.Rd6 Rf2 34.Bxe6 Rf4+
2.21: +1.11/28 18.Nxd4
I just checked this briefly with SF, but maybe will have to take a deeper look, when possible, at the main diagram after b5.

Nd4 above might be drawing, but, at least in the lines I have checked (most main ones), black wins about 70%, with 30% draws. well, the position migth be objectively drawn with best play from both sides, but, in terms of stats, if white has 2 drawing moves at ply 1 with 4 losing, 3 drawing with 5 losing at ply 3, 1 drawing with 2 losing at ply 5, etc., then the winning probability for black is quite good.

no time for thorough analysis now, but, if I spot something, might be back in due time. from a human point of view, the white position is simply shaky, with that almost complete lack of pawn shelter (just the a3 pawn).