Carlsen Blunder or Carlsen AlphaZero

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

Moderators: hgm, Rebel, chrisw

FWCC
Posts: 117
Joined: Wed Aug 22, 2007 4:39 pm

Carlsen Blunder or Carlsen AlphaZero

Post by FWCC »

In round 8 of the Tata Steel Tournament Carlsen faced Jones.In this position Carlsen gave up a knight.Was it shades of AlphaZero or a Blunder?

AlphaZero sacrificed many pieces during it's match vs Stockfish8,this game reminds one of AlphaZero's play or was it blind luck?
https://www.youtube.com/watch?v=K-A_LG5Jx3E

r1bqr1k1/p5bp/2p3p1/3np1N1/2B2pP1/4BP2/PPP4P/1K1RQ2R w - - 0 18

And this position where Jones gives up the a7 pawn and the tables turn.

1r1q2k1/pb4b1/2p1r1p1/3np1pP/2B1N3/QP3P2/P1P5/1K1R3R w - - 0 26

So with the help of Jones Carlsen wins this game.But did Carlsen Blunder the knight?
[Event "80th Tata Steel Masters 2018"]
[Site "Wijk aan Zee"]
[Date "2018.01.21"]
[Round "8"]
[White "Carlsen, Magnus"]
[Black "Jones, Gawain C B"]
[Result "1-0"]
[ECO "B76"]
[WhiteElo "2834"]
[BlackElo "2640"]
[Annotator "ChessBase"]
[PlyCount "87"]
[EventDate "2018.??.??"]
[EventType "tourn"]
[EventCountry "NED"]
[SourceTitle "playchess.com"]
[Source "ChessBase"]
[SourceQuality "1"]
[TimeControl "40/6000+30:20/3000+30:900+30"]

1. e4 {0} c5 {41} 2. Nf3 {0} d6 {4} 3. d4 {77} cxd4 {6} 4. Nxd4 {7} Nf6 {4} 5.
Nc3 {9} g6 {7} 6. Be3 {81} Bg7 {9} 7. f3 {8} Nc6 {10} 8. Qd2 {7} O-O {6} 9.
O-O-O {42} d5 {7} 10. Qe1 {9} e5 {9} 11. Nxc6 {15} bxc6 {4} 12. exd5 {8} Nxd5 {
5} 13. Bc4 {29} Be6 {44} 14. Kb1 {11} Re8 {35} 15. Ne4 {1260} f5 {15} 16. Ng5 {
217} Bc8 {76} (16... f4 17. Nxe6 Rxe6 18. Bxf4 exf4 19. Qxe6+) 17. g4 {186} f4
{225} 18. h4 {310} fxe3 {135} 19. Qxe3 {5} h6 {897} 20. Qc5 {35} Bb7 {145} (
20... hxg5 21. Qxc6 Be6 22. Bxd5 Bxd5 23. Rxd5 Qb6) 21. Ne4 {16} Re6 {413} 22.
h5 {149} Qb6 {323} (22... g5) 23. g5 {303} hxg5 {1197} 24. Qa3 {61} Rb8 {924}
25. b3 {811} Qd8 {937} 26. Qxa7 {269} gxh5 {376} 27. Rxh5 {123} Rg6 {70} 28.
Rxg5 {88} Rxg5 {26} 29. Nxg5 {19} Qc8 {65} (29... Qxg5 30. Qxb8+) 30. Rg1 {359}
Ra8 {30} 31. Qb6 {115} Ra6 {326} 32. Qc5 {34} (32. Bxa6) 32... Qd7 {192} 33.
Ne4 {86} Kh8 {138} 34. Qf2 {219} Qe7 {28} 35. Bxa6 {258} Bxa6 {1} 36. Qh2+ {42}
Kg8 {2} 37. Qh6 {13} Qa7 {262} 38. Qe6+ {116} Kf8 {7} 39. Rg5 {42} Ne3 {51} 40.
Qd6+ {0} Kf7 {34} 41. Nc5 {350} Bc8 {67} 42. Rxg7+ {8} Kxg7 43. Qxe5+ Kg6 44.
Qxe3 1-0
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Carlsen Blunder or Carlsen AlphaZero

Post by zullil »

FWCC wrote:In round 8 of the Tata Steel Tournament Carlsen faced Jones.In this position Carlsen gave up a knight.Was it shades of AlphaZero or a Blunder?

AlphaZero sacrificed many pieces during it's match vs Stockfish8,this game reminds one of AlphaZero's play or was it blind luck?
https://www.youtube.com/watch?v=K-A_LG5Jx3E

[D] r1bqr1k1/p5bp/2p3p1/3np1N1/2B2pP1/4BP2/PPP4P/1K1RQ2R w - - 0 18

And this position where Jones gives up the a7 pawn and the tables turn.

[D] 1r1q2k1/pb4b1/2p1r1p1/3np1pP/2B1N3/QP3P2/P1P5/1K1R3R w - - 0 26

So with the help of Jones Carlsen wins this game.But did Carlsen Blunder the knight?
[pgn]
[Event "80th Tata Steel Masters 2018"]
[Site "Wijk aan Zee"]
[Date "2018.01.21"]
[Round "8"]
[White "Carlsen, Magnus"]
[Black "Jones, Gawain C B"]
[Result "1-0"]
[ECO "B76"]
[WhiteElo "2834"]
[BlackElo "2640"]
[Annotator "ChessBase"]
[PlyCount "87"]
[EventDate "2018.??.??"]
[EventType "tourn"]
[EventCountry "NED"]
[SourceTitle "playchess.com"]
[Source "ChessBase"]
[SourceQuality "1"]
[TimeControl "40/6000+30:20/3000+30:900+30"]

1. e4 {0} c5 {41} 2. Nf3 {0} d6 {4} 3. d4 {77} cxd4 {6} 4. Nxd4 {7} Nf6 {4} 5.
Nc3 {9} g6 {7} 6. Be3 {81} Bg7 {9} 7. f3 {8} Nc6 {10} 8. Qd2 {7} O-O {6} 9.
O-O-O {42} d5 {7} 10. Qe1 {9} e5 {9} 11. Nxc6 {15} bxc6 {4} 12. exd5 {8} Nxd5 {
5} 13. Bc4 {29} Be6 {44} 14. Kb1 {11} Re8 {35} 15. Ne4 {1260} f5 {15} 16. Ng5 {
217} Bc8 {76} (16... f4 17. Nxe6 Rxe6 18. Bxf4 exf4 19. Qxe6+) 17. g4 {186} f4
{225} 18. h4 {310} fxe3 {135} 19. Qxe3 {5} h6 {897} 20. Qc5 {35} Bb7 {145} (
20... hxg5 21. Qxc6 Be6 22. Bxd5 Bxd5 23. Rxd5 Qb6) 21. Ne4 {16} Re6 {413} 22.
h5 {149} Qb6 {323} (22... g5) 23. g5 {303} hxg5 {1197} 24. Qa3 {61} Rb8 {924}
25. b3 {811} Qd8 {937} 26. Qxa7 {269} gxh5 {376} 27. Rxh5 {123} Rg6 {70} 28.
Rxg5 {88} Rxg5 {26} 29. Nxg5 {19} Qc8 {65} (29... Qxg5 30. Qxb8+) 30. Rg1 {359}
Ra8 {30} 31. Qb6 {115} Ra6 {326} 32. Qc5 {34} (32. Bxa6) 32... Qd7 {192} 33.
Ne4 {86} Kh8 {138} 34. Qf2 {219} Qe7 {28} 35. Bxa6 {258} Bxa6 {1} 36. Qh2+ {42}
Kg8 {2} 37. Qh6 {13} Qa7 {262} 38. Qe6+ {116} Kf8 {7} 39. Rg5 {42} Ne3 {51} 40.
Qd6+ {0} Kf7 {34} 41. Nc5 {350} Bc8 {67} 42. Rxg7+ {8} Kxg7 43. Qxe5+ Kg6 44.
Qxe3 1-0
[/pgn]
Jouni
Posts: 3279
Joined: Wed Mar 08, 2006 8:15 pm

Re: Carlsen Blunder or Carlsen AlphaZero

Post by Jouni »

Human chess is more interesting than computer: win can be loss in one move. BTW Hou Yifan has winning position this round and loses again and again and again...
Jouni
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Carlsen Blunder or Carlsen AlphaZero

Post by MikeB »

No, these fine chess players, among the best in chess in the world, are not playing anywhere near Alpha Zero when you analyze the game. For better or for worse, chess engines have moved way beyond the best humans. The best chess play right now is human and engine combined or advanced chess.

https://en.wikipedia.org/wiki/Advanced_Chess
Uri Blass
Posts: 10267
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Carlsen Blunder or Carlsen AlphaZero

Post by Uri Blass »

Of course Carlsen blundered.

22...Qb6 is not a mistake that I expect from strong humans.

22...g5 is the obvious move to block white attack.

Even after that mistake black has the advantage when the simplest way is to trade queens by 23...Bf8
jdart
Posts: 4366
Joined: Fri Mar 10, 2006 5:23 am
Location: http://www.arasanchess.org

Re: Carlsen Blunder or Carlsen AlphaZero

Post by jdart »

Carlsen was definitely worse at one point in this game. Jones is a Dragon expert and he came out of the opening better. But then he made a bunch of serious errors.

I was more impressed with the Giri-Mamedyarov game. Giri put all his pieces on wonderful squares and made the win look easy:

[pgn]
[Event "Tata Steel"]
[Site "Wijk aan Zee"]
[Date "2018.01.21"]
[EventDate "2017.01.14"]
[Round "8"]
[Result "1-0"]
[White "Anish Giri"]
[Black "Shakhriyar Mamedyarov"]

1.c4 c5 2.Nf3 Nc6 3.Nc3 g6 4.e3 Nf6 5.d4 cxd4 6.exd4 d5 7.cxd5
Nxd5 8.Qb3 e6 9.Bb5 Bg7 10.O-O O-O 11.Bxc6 bxc6 12.Re1 Qd6
13.Ne4 Qb4 14.Qc2 a5 15.a3 Qb6 16.Nc5 Re8 17.Ne5 f6 18.Nc4 Qc7
19.Bd2 e5 20.dxe5 fxe5 21.Qa4 Nb6 22.Nxb6 Qxb6 23.Qc4+ Kh8
24.Bc3 Bf5 25.g4 Bc8 26.Re4 Qb5 27.Rae1 Qxc4 28.Rxc4 Rb8 29.h3
h5 30.gxh5 gxh5 31.Rh4 Kh7 32.Ne4 1-0
[/pgn]
User avatar
Eelco de Groot
Posts: 4561
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Carlsen Blunder or Carlsen AlphaZero

Post by Eelco de Groot »

Uri Blass wrote:Of course Carlsen blundered.

22...Qb6 is not a mistake that I expect from strong humans.

22...g5 is the obvious move to block white attack.

Even after that mistake black has the advantage when the simplest way is to trade queens by 23...Bf8
Well, he did win the game... And 18. h4 according to Kaissa is the strongest move. Black has the advantage perhaps, but it is not sure the engines are right. It is not easy to show a win with Kaissa. But maybe with other engines it is? Engines are not Alpha Zero :lol:


[D]r1bqr1k1/p5bp/2p3p1/3np1N1/2B2pP1/4BP2/PPP4P/1K1RQ2R w - -

Engine: Kaissa (one thread, 512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

19/31 0:01 -1.55 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Rd3 Bf8 23.Qg1 Qb6
24.g5 h5 25.Nf6 Qxg1+ 26.Rxg1 Re7
27.Rgd1 Bg7 28.Nxd5 cxd5 29.Bxd5 Bxd5
30.Rxd5 e4 31.fxe4 (3.051.198) 1920

20/35 0:02 -1.79 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Rd3 Bf8 23.Qg1 a5
24.g5 h5 25.Nf6 Re6 26.Qh2 Qe7 27.a3 a4
28.Re1 Rd8 29.Ne4 Ra8 30.Qh3 Rb8
31.Ka1 (4.656.564) 1917

21/35 0:03 -1.87-- 18.h4 fxe3 (6.216.809) 1908

21/35 0:03 -1.79++ 18.h4 (6.603.659) 1908

21/35 0:03 -1.94-- 18.h4 fxe3 (7.238.329) 1907

21/35 0:04 -1.90 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Rd3 Bf8 23.Qg1 a5
24.a3 a4 25.Rd2 Qe7 26.h5 g5 27.Ng3 Ba6
28.Nf5 Qb7 29.Bxa6 Rxa6 30.Kc1 (8.547.805) 1899
.
.
.
43/66 90:04 -2.44 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Bb3 Bf8 23.Qc4 a5
24.g5 Qb6 25.Nc3 Qb4 26.Qxb4 axb4
27.Nxd5 cxd5 28.Bxd5 Bxd5 29.Rxd5 Re7
30.b3 Rea7 31.a4 (10.836.861.513) 2005

44/56 99:05 -2.52-- 18.h4 fxe3 (11.986.536.797) 2016

44/59 112:53 -2.59-- 18.h4 fxe3 (13.727.779.422) 2026

44/69 138:28 -2.52++ 18.h4 (16.890.136.944) 2032

44/69 139:56 -2.36++ 18.h4 (17.069.978.857) 2032

44/69 151:57 -2.46 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Bd3 Bf8 23.Qf2 Qb6
24.Qxb6 axb6 25.g5 Nf4 26.Nf6 Nxd3
27.Rxd3 Red8 28.Rd7 Rxd7 29.Nxd7 Rd8
30.Nxe5 Kh7 31.c3 (18.545.380.388) 2033

45/58 166:29 -2.54-- 18.h4 fxe3 (20.426.476.978) 2044

45/64 197:26 -2.61-- 18.h4 fxe3 (24.339.608.585) 2054

45/69 237:18 -2.54++ 18.h4 (29.383.748.428) 2063

45/69 279:15 -2.72-- 18.h4 fxe3 (34.674.074.361) 2069

45/74 364:50 -2.55++ 18.h4 (45.495.578.417) 2078

45/74 392:24 -2.71 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.g5 h5 23.Bd3 Qe7
24.Rhe1 Nf4 25.a4 Rad8 26.Qxe7 Rxe7
27.Bc4 Rd4 28.Bb3 Ba6 29.c3 Rxd1+
30.Rxd1 Bd3+ 31.Kc1 {[D]7k/p3r1b1/2p3p1/4p1Pp/P3Nn1P/1BPb1P2/1P6/2KR4 b - - In the final position of this line, White still has compensation for the Bishop} (48.925.932.351) 2078
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
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Carlsen Blunder or Carlsen AlphaZero

Post by MikeB »

Eelco de Groot wrote:
Uri Blass wrote:Of course Carlsen blundered.

22...Qb6 is not a mistake that I expect from strong humans.

22...g5 is the obvious move to block white attack.

Even after that mistake black has the advantage when the simplest way is to trade queens by 23...Bf8
Well, he did win the game... And 18. h4 according to Kaissa is the strongest move. Black has the advantage perhaps, but it is not sure the engines are right. It is not easy to show a win with Kaissa. But maybe with other engines it is? Engines are not Alpha Zero :lol:


[D]r1bqr1k1/p5bp/2p3p1/3np1N1/2B2pP1/4BP2/PPP4P/1K1RQ2R w - -

Engine: Kaissa (one thread, 512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

19/31 0:01 -1.55 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Rd3 Bf8 23.Qg1 Qb6
24.g5 h5 25.Nf6 Qxg1+ 26.Rxg1 Re7
27.Rgd1 Bg7 28.Nxd5 cxd5 29.Bxd5 Bxd5
30.Rxd5 e4 31.fxe4 (3.051.198) 1920

20/35 0:02 -1.79 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Rd3 Bf8 23.Qg1 a5
24.g5 h5 25.Nf6 Re6 26.Qh2 Qe7 27.a3 a4
28.Re1 Rd8 29.Ne4 Ra8 30.Qh3 Rb8
31.Ka1 (4.656.564) 1917

21/35 0:03 -1.87-- 18.h4 fxe3 (6.216.809) 1908

21/35 0:03 -1.79++ 18.h4 (6.603.659) 1908

21/35 0:03 -1.94-- 18.h4 fxe3 (7.238.329) 1907

21/35 0:04 -1.90 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Rd3 Bf8 23.Qg1 a5
24.a3 a4 25.Rd2 Qe7 26.h5 g5 27.Ng3 Ba6
28.Nf5 Qb7 29.Bxa6 Rxa6 30.Kc1 (8.547.805) 1899
.
.
.
43/66 90:04 -2.44 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Bb3 Bf8 23.Qc4 a5
24.g5 Qb6 25.Nc3 Qb4 26.Qxb4 axb4
27.Nxd5 cxd5 28.Bxd5 Bxd5 29.Rxd5 Re7
30.b3 Rea7 31.a4 (10.836.861.513) 2005

44/56 99:05 -2.52-- 18.h4 fxe3 (11.986.536.797) 2016

44/59 112:53 -2.59-- 18.h4 fxe3 (13.727.779.422) 2026

44/69 138:28 -2.52++ 18.h4 (16.890.136.944) 2032

44/69 139:56 -2.36++ 18.h4 (17.069.978.857) 2032

44/69 151:57 -2.46 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.Bd3 Bf8 23.Qf2 Qb6
24.Qxb6 axb6 25.g5 Nf4 26.Nf6 Nxd3
27.Rxd3 Red8 28.Rd7 Rxd7 29.Nxd7 Rd8
30.Nxe5 Kh7 31.c3 (18.545.380.388) 2033

45/58 166:29 -2.54-- 18.h4 fxe3 (20.426.476.978) 2044

45/64 197:26 -2.61-- 18.h4 fxe3 (24.339.608.585) 2054

45/69 237:18 -2.54++ 18.h4 (29.383.748.428) 2063

45/69 279:15 -2.72-- 18.h4 fxe3 (34.674.074.361) 2069

45/74 364:50 -2.55++ 18.h4 (45.495.578.417) 2078

45/74 392:24 -2.71 18.h4 fxe3 19.Qxe3 h6 20.Qc5 Bb7
21.Ne4 Kh8 22.g5 h5 23.Bd3 Qe7
24.Rhe1 Nf4 25.a4 Rad8 26.Qxe7 Rxe7
27.Bc4 Rd4 28.Bb3 Ba6 29.c3 Rxd1+
30.Rxd1 Bd3+ 31.Kc1 {[D]7k/p3r1b1/2p3p1/4p1Pp/P3Nn1P/1BPb1P2/1P6/2KR4 b - - In the final position of this line, White still has compensation for the Bishop} (48.925.932.351) 2078
It does appear that black has a winning advantage - black made a series of sub optimal moves between moves 21-24.

[pgn][Event "Mac Pro x5690 3.46 Ghz 18 CPU"]
[Site "Mac-Pro.local"]
[Date "2018.01.21"]
[Round "-"]
[White "asmFishX_2018-01-21_popc"]
[Black "SF-McB v5.0 64 POPCNT"]
[Result "0-1"]
[TimeControl "180+5"]
[FEN "r1bqr1k1/p5bp/2p3p1/3np1N1/2B2pP1/4BP2/PPP4P/1K1RQ2R w - - 0 1"]
[SetUp "1"]

1. h4 {-2.30/27 +164} fxe3 {+2.10/25 +175} 2. Qxe3 {-2.20/27 +173} h6
{+2.29/26 +180} 3. Qc5 {-2.50/29 +146} Bb7 {+2.36/29 +185} 4. Ne4
{-2.50/29 +146} Kh8 {+2.49/23 +187} 5. Nd6 {-2.22/23 +144} Qe7
{+2.53/26 +179} 6. Bb3 {-2.58/25 +142} Rad8 {+2.71/25 +180} 7. Nxb7
{-2.53/24 +146} Qxb7 {+2.75/24 +178} 8. Rhe1 {-2.73/28 +134} Rf8
{+2.74/30 +183} 9. Rd3 {-2.90/29 +127} Nf4 {+2.95/29 +188} 10. Rxd8
{-3.09/30 +106} Rxd8 {+2.99/29 +193} 11. h5 {-3.22/31 +82} Qd7
{+3.15/29 +197} 12. a3 {-3.06/31 +57} Qd2 {+3.84/30 +202} 13. Qb4
{-3.78/31 +24} gxh5 {+4.12/28 +182} 14. gxh5 {-4.08/29 +29} Nxh5
{+4.16/26 +180} 15. Ka2 {-4.25/30 +12} Qxb4 {+4.65/29 +185} 16. axb4
{-4.42/27 +10} Nf4 {+4.77/25 +189} 17. Rh1 {-4.47/28 +6} h5 {+4.89/28 +186}
18. c4 {-4.75/27 +11} Rd4 {+5.12/24 +187} 19. Ka3 {-5.05/26 +9} Bf6
{+5.46/27 +181} 20. c5 {-5.40/28 +14} Be7 {+5.78/24 +178} 21. Rh2
{-5.65/31 +13} h4 {+5.88/26 +183} 22. Ba4 {-5.93/28 +5} Nd3 {+6.27/25 +175}
23. Bxc6 {-6.02/29 +10} Nxb4 {+6.35/26 +172} 24. Be8 {-6.28/27 +13} Bxc5
{+6.69/27 +164} 25. Bg6 {-6.48/28 +18} Nd3+ {+7.00/25 +159} 26. Ka2
{-6.95/30 +14} Be7 {+7.42/26 +159} 27. Bf5 {-7.35/28 +9} Nf4
{+8.12/26 +154} 28. Kb3 {-7.75/28 +5} Kg7 {+8.41/27 +147} 29. Kc3
{-8.13/27 +10} Kf6 {+8.53/25 +149} 30. Bc8 {-8.43/28 +5} Rd6
{+9.09/29 +141} 31. Kb3 {-8.27/23 +7} Rb6+ {+9.35/25 +143} 32. Ka2
{-8.95/25 +9} Kg7 {+9.59/29 +146} 33. Rh1 {-9.41/27 +7} Rh6
{+10.35/26 +150} 34. Re1 {-9.82/27 +6} Kf6 {+10.94/28 +148} 35. Rh1
{-10.29/25 +11} h3 {+11.85/27 +144} 36. Ba6 {-11.13/27 +14} h2
{+13.09/33 +139} 37. Kb3 {-12.87/29 +5} Bc5 {+15.91/28 +126} 38. Bb7
{-13.78/29 +5} Kg5 {+17.53/34 +122} 39. Bc8 {-11.23/23 +7} Rb6+
{+28.18/32 +113} 40. Ka2 {-15.73/35 +5} Bg1 {+48.52/28 +118} 41. Kb1
{-16.37/37 +5} Kh4 {+62.32/27 +116} 42. Ka2 {-20.92/26 +5} Kg3
{+1000.22/34 +114} 43. b4 {-132.65/34 +5} Rxb4 {+1000.12/34 +112} 44. Bf5
{-1000.13/46 +5} Ne2 {+1000.11/36 +115} 45. Bc2 {-1000.10/57 +7} Nc1+
{+1000.10/38 +115} 46. Ka3 {-1000.09/70 +12} Rb6 {+1000.09/42 +111} 47. Ka4
{-1000.08/77 +17} Ne2 {+1000.08/43 +108} 48. Bd3 {-1000.07/84 +22} Nc3+
{+1000.07/49 +107} 49. Ka5 {-1000.06/101 +25} Bc5 {+1000.06/58 +107} 50.
Bb5 {-1000.05/117 +26} Rxb5+ {+1000.05/82 +109} 51. Ka6 {-1000.04/127 +31}
Rb6+ {+1000.04/124 +108} 52. Kxa7 {-1000.03/127 +36} Nb5+
{+1000.03/127 +112} 53. Ka8 {-1000.02/127 +41} Nc7+ {+1000.02/127 +117} 54.
Ka7 {-1000.01/127 +46} Rb1# {+1000.01/127 +122}
{Black mates} 0-1
[/pgn]
User avatar
Eelco de Groot
Posts: 4561
Joined: Sun Mar 12, 2006 2:40 am
Full name:   

Re: Carlsen Blunder or Carlsen AlphaZero

Post by Eelco de Groot »

I think you mean White made a series of sub optimal moves here Michael :) I mean you let asmFish have White? 22. Nd6 {?!} is not even in the top 5 moves anymore after a while. Not even top ten. A beginner's move if we are talking Alpha Zero ...:P


r2qr2k/pb4b1/2p3pp/2Qnp3/2B1N1PP/5P2/PPP5/1K1R3R w - -

Engine: Kaissa II (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G.

30 7:13 -2.57 22.h5 Bf8 23.Qf2 g5 24.Bb3 Qb6
25.Qxb6 axb6 26.a3 b5 27.Rd3 Be7
28.c4 bxc4 29.Bxc4 Rad8 30.Rc1 Rf8
31.Ng3 Rf4 32.Nf5 Bf8 33.Rb3 Bc8
34.Ng3 Be7 35.a4 (745.420.049) 1719

30 7:13 -2.60 22.Qa3 Qe7 23.Rd3 Qxa3 24.Rxa3 a5
25.h5 g5 26.Re1 Reb8 27.Rd3 a4
28.Nc5 Ra5 29.Ne4 Bc8 30.Kc1 Be6
31.a3 Kh7 32.Rh1 Re8 33.Kb1 Rf8
34.Re1 Rf4 35.Ba2 (745.420.049) 1719

30 7:13 -2.64 22.Rd2 Bf8 23.Qf2 Qb6 24.h5 g5
25.Bb3 Qxf2 26.Rxf2 Rad8 27.Rd2 Nb6
28.Rhd1 Rxd2 29.Rxd2 c5 30.Bf7 Rb8
31.Bg6 c4 32.c3 Nd5 33.Bf7 Ne3
34.Bg6 Be7 35.Kc1 (745.420.049) 1719

30 7:13 -2.66 22.Qf2 Qb6 23.Qxb6 axb6 24.Nd6 Re7
25.Bd3 Ba6 26.Bxa6 Rxa6 27.h5 Rea7
28.a3 gxh5 29.Nf5 b5 30.c3 Rd7
31.Nxg7 Kxg7 32.gxh5 Rb6 33.Rhg1+ Kh7
34.Rde1 Rg7 35.Rg4 (745.420.049) 1719

30 7:13 -2.75 22.Qg1 Qe7 23.c3 a5 24.Rh2 a4 25.a3 Reb8
26.h5 g5 27.Ka1 Bc8 28.Rhd2 Be6
29.Ba2 Bf8 30.Qg3 Nf4 31.Bxe6 Qxe6
32.Qe1 Rb7 33.Nd6 Rd7 34.Nf5 Rxd2
35.Rxd2 (745.420.049) 1719

Here 22. Nd6 is still top ten:

r2qr2k/pb4b1/2p3pp/2Qnp3/2B1N1PP/5P2/PPP5/1K1R3R w - -

Engine: Kaissa II (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G.

31 20:57 -2.42 22.g5 h5 23.Bd3 Re6 24.Bc4 Qb6
25.Bb3 Qxc5 26.Nxc5 Re7 27.Rhe1 Rf8
28.c4 Nf4 29.Rd2 Bc8 30.Bc2 Bf5
31.Bxf5 gxf5 32.Rd6 Rc8 33.Red1 Ng2
34.Rh1 Kh7 35.Nd7 (2.196.693.755) 1746

31 20:57 -2.42 22.c3 Bf8 23.Qg1 Qe7 24.Ka1 Rab8
25.Bb3 Red8 26.h5 g5 27.Qe1 Nf4
28.Rh2 Nd3 29.Qe3 Nc5 30.Rhd2 Nxb3+
31.axb3 Rd5 32.Kb1 Rbd8 33.c4 Rxd2
34.Rxd2 Rxd2 35.Qxd2 (2.196.693.755) 1746

31 20:57 -2.57 22.Nd6 Qe7 23.a3 Rad8 24.Ne4 Bf8
25.Qa5 Ne3 26.Rxd8 Rxd8 27.Bb3 c5
28.Nxc5 Bxf3 29.Re1 Nxg4 30.Ne6 Rc8
31.Nd4 Bb7 32.Ne6 Qd7 33.Nxf8 Rxf8
34.Qxa7 h5 35.Qb6 (2.196.693.755) 1746

31 20:57 -2.57 22.h5 Bf8 23.Qf2 g5 24.Rhe1 Qb6
25.Rd3 Qxf2 26.Nxf2 Rad8 27.Ne4 Be7
28.Ng3 Bc8 29.a3 Nf4 30.Rxd8 Bxd8
31.Bb3 Bc7 32.Ba4 Rd8 33.Bxc6 Be6
34.Be4 Ng2 35.Rh1 (2.196.693.755) 1746

31 20:57 -2.64 22.Rd2 Bf8 23.Qf2 Qb6 24.h5 g5
25.Bb3 Qxf2 26.Rxf2 Rad8 27.Rd2 Kg7
28.Rhd1 Be7 29.Kc1 Rf8 30.c4 Nb6
31.Rxd8 Rxd8 32.Rxd8 Bxd8 33.c5 Nd5
34.Nd6 Ba6 35.Nf5+ (2.196.693.755) 1746

31 20:57 -2.65 22.Qf2 Qb6 23.Qxb6 axb6 24.Nd6 Re7
25.Nxb7 Rxb7 26.Rxd5 cxd5 27.Bxd5 Rab8
28.Bxb7 Rxb7 29.Re1 Rf7 30.Re3 Kh7
31.c3 h5 32.Kc2 Rf4 33.Kd3 b5 34.b3 Kh6
35.a4 (2.196.693.755) 1746

31 20:57 -2.65 22.Bd3 Bf8 23.Qf2 Qb6 24.Qxb6 axb6
25.Nd2 Nf4 26.Be4 Rab8 27.c3 b5
28.g5 h5 29.Rhf1 Be7 30.Nb3 c5
31.Rd7 Bc8 32.Rc7 Bh3 33.Rd1 c4
34.Nc5 Rbc8 35.Na6 (2.196.693.755) 1746

31 20:57 -2.66 22.Bb3 Bf8 23.Qf2 Qb6 24.Qxb6 axb6
25.h5 g5 26.a3 b5 27.Rd3 Rad8
28.Rhd1 Be7 29.c4 bxc4 30.Bxc4 Kg7
31.Rc1 Nf6 32.Rb3 Nxe4 33.Rxb7 Nd2+
34.Ka2 Nxf3 35.Be2 (2.196.693.755) 1746

31 20:57 -2.68 22.Qa3 Qe7 23.Rd3 Qxa3 24.Rxa3 a5
25.h5 g5 26.Re1 Reb8 27.Kc1 a4
28.Rd1 Bf8 29.Rad3 Be7 30.Re1 a3
31.b3 Rf8 32.Kb1 Rad8 33.c3 Bc8
34.Kc2 Be6 35.Nd2 (2.196.693.755) 1746

31 20:57 -2.70 22.Kc1 Bf8 23.Qf2 Qb6 24.Rd3 Rad8
25.h5 g5 26.Rhd1 Be7 27.a3 Qxf2
28.Nxf2 Kg7 29.Ne4 Rf8 30.b4 Bc8
31.R1d2 Nb6 32.Bb3 Rxd3 33.Rxd3 Nd5
34.Kb2 Nf6 35.Nd6 (2.196.693.755) 1746

But here not anymore:

[D]r2qr2k/pb4b1/2p3pp/2Qnp3/2B1N1PP/5P2/PPP5/1K1R3R w - -

Engine: Kaissa II (512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

33 56:45 -2.58 22.h5 Bf8 23.Qf2 g5 24.Rh2 Qb6
25.Rd3 Rad8 26.a3 Qxf2 27.Rxf2 Nb6
28.Rfd2 Rxd3 29.Bxd3 Rd8 30.Kc1 c5
31.Bb5 Rxd2 32.Kxd2 Kg7 33.c3 Nd5
34.Bd3 Nf6 35.c4 (5.957.270.385) 1749

33 56:45 -2.59 22.g5 h5 23.Bd3 Qe7 24.Kc1 Bf8
25.Qxe7 Bxe7 26.Nf6 Nxf6 27.gxf6 Bxf6
28.Bxg6 Red8 29.Bxh5 Rxd1+ 30.Rxd1 Bxh4
31.Bg6 Bg5+ 32.Kb1 Rd8 33.Rxd8+ Bxd8
34.Be4 Be7 35.a3 (5.957.270.385) 1749

33 56:45 -2.61 22.Bd3 Bf8 23.Qf2 Qb6 24.Qxb6 axb6
25.g5 Nf4 26.Nf6 Nxd3 27.Rxd3 Red8
28.Rd7 Rxd7 29.Nxd7 Rd8 30.Nxe5 Kh7
31.c3 Bg7 32.Nc4 Ba6 33.Nxb6 Bd3+
34.Ka1 Rf8 35.a4 (5.957.270.385) 1749

33 56:45 -2.66 22.Bb3 Bf8 23.Qf2 Qb6 24.Rhe1 Qxf2
25.Nxf2 a5 26.a3 Nf4 27.Rd7 Ba6
28.h5 gxh5 29.gxh5 a4 30.Ba2 Red8
31.Rc7 Be2 32.Rg1 Bd6 33.Rf7 Rg8
34.Ng4 Bxf3 35.Rf6 (5.957.270.385) 1749

33 56:45 -2.72 22.c3 Bf8 23.Qf2 Qb6 24.Bb3 Qxf2
25.Nxf2 Ba6 26.Rhe1 Nf4 27.Rd7 Be2
28.Bd1 Bc4 29.g5 Red8 30.Rxd8 Rxd8
31.Rxe5 hxg5 32.hxg5 Rd2 33.Re4 Nd3
34.Nxd3 Bxd3+ 35.Kc1 (5.957.270.385) 1749

33 56:45 -2.72 22.Kc1 Bf8 23.Qf2 Qb6 24.Rd3 Rad8
25.h5 g5 26.Rhd1 Be7 27.Qxb6 axb6
28.c3 b5 29.Bb3 Rf8 30.Ng3 Bc8 31.a4 bxa4
32.Bxa4 Bd7 33.c4 Nb6 34.Ne4 Kg7
35.Kc2 (5.957.270.385) 1749

33 56:45 -2.74 22.Rd2 Bf8 23.Qf2 Qb6 24.h5 g5
25.Bb3 Qxf2 26.Rxf2 c5 27.Re1 Nf4
28.Rd2 Rad8 29.Red1 Be7 30.Rxd8 Rxd8
31.Rxd8+ Bxd8 32.Bc4 Bxe4 33.fxe4 Nh3
34.Be6 Kg7 35.Kc1 (5.957.270.385) 1749

33 56:45 -2.75 22.Qa3 Qe7 23.Rd3 Qxa3 24.Rxa3 a5
25.h5 g5 26.Re1 Reb8 27.Kc1 a4
28.Rd1 Bf8 29.Rad3 Be7 30.Nc3 a3
31.b3 Nxc3 32.Rxc3 Rd8 33.Rcd3 Rxd3
34.Bxd3 Rd8 35.c3 (5.957.270.385) 1749

33 56:45 -2.76 22.Qf2 Qb6 23.Qxb6 axb6 24.Nd6 Re7
25.Bd3 Ba6 26.Bxa6 Rxa6 27.a3 b5
28.h5 gxh5 29.Nf5 Rea7 30.c3 Rd7
31.gxh5 Rf7 32.Nxg7 Rxg7 33.Rde1 Rg5
34.Reg1 Rxg1+ 35.Rxg1 (5.957.270.385) 1749

33 56:45 -2.76 22.Qg1 Qe7 23.Re1 Rf8 24.h5 g5
25.Rh3 Bc8 26.Qc5 Qxc5 27.Nxc5 Rb8
28.Bb3 a5 29.c3 e4 30.Rxe4 Bxc3
31.bxc3 Nxc3+ 32.Kc1 Nxe4 33.Nxe4 Ba6
34.Nd6 Rf4 35.Nf5 (5.957.270.385) 1749
______________________________________________

34 71:28 -2.60 22.g5 h5 23.Bd3 Qe7 24.Kc1 Bf8
25.Qxe7 Bxe7 26.Nf6 Nxf6 27.gxf6 Bxf6
28.Bxg6 Red8 29.Bxh5 Rxd1+ 30.Rxd1 Bxh4
31.Bg6 Bg5+ 32.Kb1 Rd8 33.Rxd8+ Bxd8
34.Be4 Be7 35.c3 (7.552.860.585) 1761

34 71:28 -2.60 22.Qa3 Qe7 23.Qxe7 Rxe7 24.Bd3 c5
25.Nxc5 Ne3 26.Nxb7 Nxd1 27.Rxd1 Rxb7
28.Be4 Rab8 29.Bxb7 Rxb7 30.c3 e4
31.fxe4 Bxc3 32.b3 Re7 33.Kc2 Rc7
34.h5 Kg7 35.Rd5 (7.552.860.585) 1761

34 71:28 -2.63 22.h5 Bf8 23.Qf2 g5 24.Bb3 Qb6
25.Rh2 Rad8 26.Qxb6 Nxb6 27.Rhd2 Rxd2
28.Rxd2 c5 29.Bf7 Rb8 30.c4 Kg7
31.Bg6 Bc6 32.Rc2 Be7 33.Rc3 Bxe4+
34.Bxe4 Rd8 35.Ra3 (7.552.860.585) 1761

34 71:28 -2.65 22.Qf2 Qb6 23.Qxb6 axb6 24.Nd6 Re7
25.Nxb7 Rxb7 26.Rxd5 cxd5 27.Bxd5 Raa7
28.Bxb7 Rxb7 29.Kc1 Rf7 30.Rd1 Rxf3
31.Rd6 Rf6 32.Rxf6 Bxf6 33.h5 Bg5+
34.Kd1 Kg7 35.Ke2 (7.552.860.585) 1761

34 71:28 -2.69 22.Rhe1 Bf8 23.Qf2 Qb6 24.Qe2 a5
25.Ka1 Nf4 26.Qf1 Reb8 27.Rb1 Bc8
28.Red1 Ra7 29.h5 gxh5 30.Nf6 Qb4
31.c3 Qe7 32.Qg1 Rab7 33.Nxh5 Nxh5
34.gxh5 Be6 35.Bxe6 (7.552.860.585) 1761

34 71:28 -2.70 22.Bd3 Bf8 23.Qf2 Qb6 24.Qxb6 axb6
25.g5 Nf4 26.Nf6 Nxd3 27.Rxd3 Red8
28.Rd7 Rxd7 29.Nxd7 Rd8 30.Nxe5 Kh7
31.c3 Bg7 32.Nc4 b5 33.Na5 Ba8
34.Nb3 Rd5 35.f4 (7.552.860.585) 1761

34 71:28 -2.70 22.c3 Bf8 23.Qf2 Qb6 24.Bb3 Qxf2
25.Nxf2 Ba6 26.Rhe1 Nf4 27.Rd7 Be2
28.Bd1 Bc4 29.h5 Red8 30.Rxd8 Rxd8
31.Rxe5 Rd2 32.Re4 Nd3 33.Nxd3 Rxd1+
34.Kc2 Bxd3+ 35.Kxd1 (7.552.860.585) 1761

34 71:28 -2.72 22.Bb3 Bf8 23.Qf2 Qb6 24.Rhe1 Qxf2
25.Nxf2 Rad8 26.c3 Kg7 27.h5 Nf6
28.Rxd8 Rxd8 29.Rxe5 gxh5 30.gxh5 c5
31.Rf5 Bc8 32.Re5 Rd2 33.Ne4 Nxe4
34.Kc1 Ng5 35.Re8 (7.552.860.585) 1761

34 71:28 -2.74 22.Rd2 Bf8 23.Qf2 Qb6 24.h5 g5
25.Bb3 Qxf2 26.Rxf2 c5 27.Re1 Nf4
28.Rd2 Rad8 29.Red1 Be7 30.Rxd8 Rxd8
31.Rxd8+ Bxd8 32.Bc4 Kg7 33.Kc1 Bxe4
34.fxe4 Nh3 35.Be6 (7.552.860.585) 1761

34 71:28 -2.78 22.Kc1 Bf8 23.Qf2 Qb6 24.Rd3 Rad8
25.h5 g5 26.Rhd1 Qxf2 27.Nxf2 Be7
28.Ne4 Rf8 29.a3 Kg7 30.b4 Bc8
31.Ba2 Nf4 32.Rxd8 Rxd8 33.Rxd8 Bxd8
34.Nd6 Bd7 35.Kd2 (7.552.860.585) 1761
______________________________________________

35 93:11 -2.63 22.h5 Bf8 23.Qf2 g5 24.Bb3 Qb6
25.Rh2 Rad8 26.Qxb6 Nxb6 27.Rhd2 Rxd2
28.Rxd2 c5 29.Bf7 Rb8 30.c4 Kg7
31.Bg6 Bc6 32.Rc2 Be7 33.Rc3 Na4
34.Rc2 Rd8 35.Kc1 (9.826.581.778) 1757

35 93:11 -2.68 22.Bb3 Bf8 23.Qf2 Qb6 24.c3 Qxf2
25.Nxf2 Ba6 26.Rhe1 Nf4 27.Rd7 Be2
28.Bd1 Bc4 29.h5 gxh5 30.gxh5 Red8
31.Rxd8 Rxd8 32.Rxe5 Rd2 33.Re4 Nd3
34.Nxd3 Bxd3+ 35.Kc1 (9.826.581.778) 1757

35 93:11 -2.69 22.c3 Bf8 23.Qf2 Qb6 24.Bb3 Qxf2
25.Nxf2 Ba6 26.Rhe1 Nf4 27.Rd7 Be2
28.Bd1 Bc4 29.h5 gxh5 30.gxh5 Red8
31.Rxd8 Rxd8 32.Rxe5 Rd2 33.Re4 Nd3
34.Nxd3 Bxd3+ 35.Kc1 (9.826.581.778) 1757

35 93:11 -2.72 22.g5 h5 23.Bd3 Qe7 24.Ka1 Rab8
25.Qxe7 Rxe7 26.Nc5 Ne3 27.Rd2 Nf5
28.Kb1 Kh7 29.Be4 Rf7 30.b3 Bc8
31.Rd8 Rb5 32.Nd3 Rc7 33.Kb2 Be6
34.Re8 Bf7 35.Ra8 (9.826.581.778) 1757

35 93:11 -2.72 22.Bd3 Bf8 23.Qf2 Qb6 24.Qxb6 axb6
25.g5 Nf4 26.Nf6 Nxd3 27.Rxd3 Red8
28.Rd7 Rxd7 29.Nxd7 Rd8 30.Nxe5 Kh7
31.c3 Bg7 32.Nc4 b5 33.Na5 Ba8
34.Nb3 Rd5 35.f4 (9.826.581.778) 1757

35 93:11 -2.77 22.Qf2 Qb6 23.Qxb6 axb6 24.Nd6 Re7
25.Nxb7 Rxb7 26.Rxd5 cxd5 27.Bxd5 Raa7
28.Bxb7 Rxb7 29.Rh3 Rf7 30.Kc1 e4
31.fxe4 Rf1+ 32.Kd2 Bxb2 33.Rd3 Bc1+
34.Kc3 Rf4 35.Kd4 (9.826.581.778) 1757

35 93:11 -2.78 22.Qa3 Qe7 23.Qxe7 Rxe7 24.Bd3 Rb8
25.Rd2 c5 26.Re1 Rc7 27.Ba6 Bxa6
28.Rxd5 Bb7 29.Rd6 Rf8 30.Rxg6 Rxf3
31.Nd6 Kh7 32.h5 Rf4 33.a3 Bc6
34.Nf5 Be4 35.Nxg7 (9.826.581.778) 1757

35 93:11 -2.80 22.Rhe1 Bf8 23.Qf2 Qb6 24.Rd3 Qxf2
25.Nxf2 Rad8 26.h5 g5 27.c3 Bc5
28.Ne4 Bb6 29.a4 Kg7 30.Red1 Rf8
31.b4 Bc8 32.a5 Bc7 33.Bb3 Nf6
34.Rxd8 Rxd8 35.Rxd8 (9.826.581.778) 1757

35 93:11 -2.82 22.Rd2 Bf8 23.Qf2 Qb6 24.Bb3 Qxf2
25.Rxf2 Rad8 26.g5 Nf4 27.Nf6 Re7
28.c3 Bc8 29.Rff1 Bf5+ 30.Ka1 c5
31.Bc4 Be6 32.Bxe6 Rxe6 33.Rd1 Rxd1+
34.Rxd1 Be7 35.Ne4 (9.826.581.778) 1757

35 93:11 -2.85 22.Kc1 Bf8 23.Qf2 Qb6 24.Rd3 Rad8
25.h5 g5 26.Qxb6 axb6 27.Rhd1 Be7
28.Bb3 Ba6 29.R3d2 Ne3 30.Re1 Nc4
31.Rxd8 Rxd8 32.Ng3 Bf6 33.Nf5 e4
34.Nxh6 Bxb2+ 35.Kb1 (9.826.581.778) 1757
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