Testposition

[Lyudmil Tsvetkov]

In the g5 line, SF is very much afraid of a line ending here:

[d]8/2r2pk1/P7/1B2pK2/3bP3/6P1/R7/8 w - - 0 31

150cps white edge, why on earth should SF choose this line?

As the white king and rook get more active, the advantage increases, but no much benefit from this.

Maybe the fact that on g5 white has only acceptable response, Bc7, should somehow ring a bell.

Ra4 seems to win here:

Code: Select all

info depth 46 seldepth 70 multipv 1 score cp 511 nodes 8812113811 nps 13491384 hashfull 999 tbhits 32712853 time 653166 pv a2a4 c7c8 a4c4 c8b8 b5c6 d4e3 c6d5 b8b1 c4c7 b1f1 f5e5 f1a1 c7f7 g7g6 f7f3 e3g1 d5c4 a1a4 e5d5 a4a5 d5e6 a5a4 c4d3 g6g5 f3f7 a4a3 d3c4 a3a4 f7g7 g5h6 g7c7 a4a5 c7f7 a5a4 c4d3 a4a3 d3b5 h6g5 e4e5 a3a1 f7g7 g5h6 e6f6 a1a3 g7b7 a3f3 f6e6 h6g5 e6d5 f3f8 d5e4 f8d8

At depth 47 the score is +7.42 now and keeps climbing.
It just resolved to +8.82:

Code: Select all

info depth 47 seldepth 77 multipv 1 score cp 882 nodes 13365921658 nps 13649113 hashfull 999 tbhits 54915966 time 979252 pv a2a4 c7c8 a4c4 c8b8 b5c6 d4e3 c6d5 b8b6 c4c6 b6b1 c6c7 b1f1 f5e5 f1a1 c7f7 g7g6 f7f3 e3g1 d5c4 a1a4 e5d5 a4a5 d5e6 a5a4 c4d3 g6g5 e4e5 g5g4 f3f7 g1e3 d3b5 a4a1 b5c4 g4h3 e6d6 a1d1 d6c6 d1c1 c6d5 h3h2 e5e6 c1d1 d5c6 d1d4 e6e7 d4c4 c6d6 c4a4 e7e8q a4a6 d6d5 a6a3 e8h8 h2g1 h8h3 a3a5 d5e4

Depths 48 and 49:

Code: Select all

info depth 48 seldepth 77 multipv 1 score cp 857 nodes 15661735979 nps 13622620 hashfull 999 tbhits 65067155 time 1149686 pv a2a4 c7c8 a4c4 c8h8 c4c7 d4b6 c7b7 b6e3 b5c4 h8h5 f5g4 h5h1 b7f7 g7g6 f7c7 h1g1 c7c6 g6h7 c4d5 g1a1 g4f5 e3d4 f5g5 d4e3 g5f6 a1f1 f6e5 f1a1 c6c7 h7g6 d5c4 g6g5 e5d6 g5g4 e4e5 a1d1 c4d5 d1a1 c7g7 g4f5 g7f7 f5g6 d5c4 a1d1 d6e6 d1d8 f7c7 e3b6 c7b7 d8e8 e6d6 e8d8 d6e7 b6a5 b7b5 a5c7 b5c5 d8e8 e7e8 c7b8 e5e6 b8a7

Code: Select all

info depth 49 seldepth 78 multipv 1 score cp 979 nodes 18545141867 nps 13510053 hashfull 999 tbhits 76196621 time 1372692 pv a2a4 c7c8 a4c4 c8h8 c4c7 d4b6 c7b7 b6d4 b5c4 h8h5 f5g4 h5h1 b7f7 g7g6 f7c7 g6f6 c4a2 h1a1 c7c2 a1e1 c2c6 f6g7 g4g5 d4e3 g5f5 e1f1 f5e5 f1a1 c6c7 g7h6 a2c4 a1a5 e5e6 e3g1 c7f7 a5a4 f7f1 g1d4 e4e5 d4e3 f1f6 h6g5 f6f3 e3c1 e6d5 a4a5 d5c6 g5g4 f3f7 a5a4 c4d3 a4a3 d3e2 g4h3 e5e6 a3c3 c6d5 c3g3 e6e7 g3g8 e2b5 g8g5 d5c6 g5g8 a6a7 c1g5 f7f8

And why is Rc8 an obligatory move after Ra4?

I do not have the possibility to look in-depth into the position right now, but my impression is the ending is a draw.

Anyway, this later position is just a side line; the main line arises after Re7 on the position Andrew analysed for hours. I would claim with bigger certainty that it is draw there.

Apart from that, if g5 does not achieve a draw, other moves would lose much easier after c5, you might want to check this. No opposite coloured bishops in the other lines. And c5, as suggested by Komodo, instead of Rd5 on the main diagram wins with certainty, so the point is if there is a move after Rd5 that saves the game for black; I think g5 is such a move, but chess is so complex, that nothing could be claimed with certainty.

Any enthusiast could check what happens after c5 instead of Rd5.

[Lyudmil Tsvetkov]

Hi Lyudmil,


Here is what Stockfish (the latest development version) says after your 27...g5 move with a couple of hours of looking at the position, and looks like the ending position is just as you called it...a draw.


New game
r3rbk1/1p3p1p/1B2qn2/PQ1Rp1p1/2P1P3/5B1P/6P1/1R5K w - - 0 1

Analysis by Stockfish 050615 64 POPCNT:


28.Bc7 Nxd5 29.cxd5 Qa6 30.Qxb7 Qxb7 31.Rxb7 Re7 32.d6 Rd7 33.Rb6 Rc8 34.Kh2 Rcxc7 35.dxc7 Rxc7 36.a6 Bc5 37.Rb7 Rxb7 38.axb7 Bd6 39.h4 gxh4 40.Kh3 Kg7 41.Kxh4 Kg6 42.Bg4 Be7+ 43.Kg3 Bd6 44.Bf5+ Kh6 45.Kh4 Bb8 46.Kg4 Bc7 47.g3 Bd6 48.Kh4 Bb8 49.Bg4 Kg6 50.Bd7 Bd6 51.Bf5+ Kh6 52.g4 f6 53.Be6 Kg6 54.Bc4 Kh6 55.Bb3 Kg6 56.Bd5 Bc7 57.Bg8 Bb8 58.Bd5 Bd6
+/- (0.81) Depth: 49/74 02:27:47 46354MN, tb=3908988



here is the position after all of the above moves are played:

[d]8/1P5p/3b1pk1/3Bp3/4P1PK/8/8/8 w - - 0 59

Hi Andrew.

Thank you very much for coming to the rescue in such a critical moment!

Why do I always like your posts and miss your not posting?

But really, are there chess players in Ohio, I should visit sometime?

[MikeGL]

In the g5 line, SF is very much afraid of a line ending here

Can you give this whole line? Thanks in advance.

whole line is obvious: Rd5? (c5! wins) g5 Bc7 (simply no other moves, therefore g5 is good) Nd5 cd5 Qa6 Qb7 Qb7 Rb7 Re7

[d]r4bk1/1RB1rp1p/8/P2Pp1p1/4P3/5B1P/6P1/7K w - - 0 6

and now try to win this with white

it seems positionally, 27...g6 of engine is stronger than your 27...g5, as 27...g5 reveals its ugly head on this 'almost' forced line. If it is 27...g6 the Bh5 below won't be possible. Although white has very slight advantage on a R+B endgame (with opposite B) but a slight chance of winning even on accurate play from both sides.

[pgn]
[Event "Talkchess"]
[Site "Talkchess"]
[Date "2015.06.07"]
[Round "?"]
[White "Postmortem_Anand"]
[Black "Postmortem_Kasparov"]
[Result "*"]
[BlackElo "2700"]
[Time "11:53:55"]
[WhiteElo "2700"]
[TimeControl "600"]
[SetUp "1"]
[FEN "r4bk1/1RB1rp1p/8/P2Pp1p1/4P3/5B1P/6P1/7K w - - 0 6"]
[Termination "unterminated"]
[PlyCount "5"]
[WhiteType "human"]
[BlackType "human"]

6. Bg4 Rxc7 7. Rxc7 Rxa5 8. Bh5 *
[/pgn]

[Eelco de Groot]

In the g5 line, SF is very much afraid of a line ending here:

[d]8/2r2pk1/P7/1B2pK2/3bP3/6P1/R7/8 w - - 0 31

150cps white edge, why on earth should SF choose this line?

As the white king and rook get more active, the advantage increases, but no much benefit from this.

Maybe the fact that on g5 white has only acceptable response, Bc7, should somehow ring a bell.

Ra4 seems to win here:

Code: Select all

info depth 46 seldepth 70 multipv 1 score cp 511 nodes 8812113811 nps 13491384 hashfull 999 tbhits 32712853 time 653166 pv a2a4 c7c8 a4c4 c8b8 b5c6 d4e3 c6d5 b8b1 c4c7 b1f1 f5e5 f1a1 c7f7 g7g6 f7f3 e3g1 d5c4 a1a4 e5d5 a4a5 d5e6 a5a4 c4d3 g6g5 f3f7 a4a3 d3c4 a3a4 f7g7 g5h6 g7c7 a4a5 c7f7 a5a4 c4d3 a4a3 d3b5 h6g5 e4e5 a3a1 f7g7 g5h6 e6f6 a1a3 g7b7 a3f3 f6e6 h6g5 e6d5 f3f8 d5e4 f8d8

At depth 47 the score is +7.42 now and keeps climbing.
It just resolved to +8.82:

Code: Select all

info depth 47 seldepth 77 multipv 1 score cp 882 nodes 13365921658 nps 13649113 hashfull 999 tbhits 54915966 time 979252 pv a2a4 c7c8 a4c4 c8b8 b5c6 d4e3 c6d5 b8b6 c4c6 b6b1 c6c7 b1f1 f5e5 f1a1 c7f7 g7g6 f7f3 e3g1 d5c4 a1a4 e5d5 a4a5 d5e6 a5a4 c4d3 g6g5 e4e5 g5g4 f3f7 g1e3 d3b5 a4a1 b5c4 g4h3 e6d6 a1d1 d6c6 d1c1 c6d5 h3h2 e5e6 c1d1 d5c6 d1d4 e6e7 d4c4 c6d6 c4a4 e7e8q a4a6 d6d5 a6a3 e8h8 h2g1 h8h3 a3a5 d5e4

Depths 48 and 49:

Code: Select all

info depth 48 seldepth 77 multipv 1 score cp 857 nodes 15661735979 nps 13622620 hashfull 999 tbhits 65067155 time 1149686 pv a2a4 c7c8 a4c4 c8h8 c4c7 d4b6 c7b7 b6e3 b5c4 h8h5 f5g4 h5h1 b7f7 g7g6 f7c7 h1g1 c7c6 g6h7 c4d5 g1a1 g4f5 e3d4 f5g5 d4e3 g5f6 a1f1 f6e5 f1a1 c6c7 h7g6 d5c4 g6g5 e5d6 g5g4 e4e5 a1d1 c4d5 d1a1 c7g7 g4f5 g7f7 f5g6 d5c4 a1d1 d6e6 d1d8 f7c7 e3b6 c7b7 d8e8 e6d6 e8d8 d6e7 b6a5 b7b5 a5c7 b5c5 d8e8 e7e8 c7b8 e5e6 b8a7

Code: Select all

info depth 49 seldepth 78 multipv 1 score cp 979 nodes 18545141867 nps 13510053 hashfull 999 tbhits 76196621 time 1372692 pv a2a4 c7c8 a4c4 c8h8 c4c7 d4b6 c7b7 b6d4 b5c4 h8h5 f5g4 h5h1 b7f7 g7g6 f7c7 g6f6 c4a2 h1a1 c7c2 a1e1 c2c6 f6g7 g4g5 d4e3 g5f5 e1f1 f5e5 f1a1 c6c7 g7h6 a2c4 a1a5 e5e6 e3g1 c7f7 a5a4 f7f1 g1d4 e4e5 d4e3 f1f6 h6g5 f6f3 e3c1 e6d5 a4a5 d5c6 g5g4 f3f7 a5a4 c4d3 a4a3 d3e2 g4h3 e5e6 a3c3 c6d5 c3g3 e6e7 g3g8 e2b5 g8g5 d5c6 g5g8 a6a7 c1g5 f7f8

And why is Rc8 an obligatory move after Ra4?

I do not have the possibility to look in-depth into the position right now, but my impression is the ending is a draw.

Anyway, this later position is just a side line; the main line arises after Re7 on the position Andrew analysed for hours. I would claim with bigger certainty that it is draw there.

Apart from that, if g5 does not achieve a draw, other moves would lose much easier after c5, you might want to check this. No opposite coloured bishops in the other lines. And c5, as suggested by Komodo, instead of Rd5 on the main diagram wins with certainty, so the point is if there is a move after Rd5 that saves the game for black; I think g5 is such a move, but chess is so complex, that nothing could be claimed with certainty.

Any enthusiast could check what happens after c5 instead of Rd5.

First you say

SF is very much afraid of this line

, then

150cps white edge, why on earth should SF choose this line?

So it was not clear how relevant this position was. From Ronald's SyZygy results of his own program with tablebases it is clear it is won. Without tablebases it maybe takes a bit longer or less high score? Claiming that Black is playing suboptimally against a 9.79 evaluation is just silly Lyudmil if not insulting of Ronald's programming skills, you know that.

Plain Stockfish without any tablebases still does find a win, not a very high score though.

[d]8/2r2pk1/P7/1B2pK2/3bP3/6P1/R7/8 w - -

Engine: SfVerification_010 MOD MP (512 MB)
by Tord Romstad, Marco Costalba and Joona Kiiski

42 36:39 +2.94 1.Ra4 Rc8 2.Rc4 Rb8 3.Bc6 Rd8 4.Bd5 Rd7
5.g4 Re7 6.g5 Rd7 7.Ra4 Re7 8.Rb4 Ba7
9.Rb3 Bc5 10.Rc3 Bd4 11.Rc6 Ra7
12.Rf6 Bc5 13.Kxe5 Be3 14.Kf5 (8.401.618.993) 3820

42 36:39 +1.93 1.Ra5 Ba7 2.Be2 Bd4 3.Bd1 Kf8 4.Bb3 Rc3
5.Bd5 Rxg3 6.a7 Rf3+ 7.Kg5 Bxa7
8.Rxa7 Rf4 9.Rb7 Ke8 10.Bc6+ Kf8
11.Rc7 Rf2 12.Bd7 Rf1 13.Bh3 Rf2
14.Bg4 (8.401.618.993) 3820

42 36:39 +1.93 1.Ra3 Ba7 2.Ra4 Rc3 3.Bc4 Rxg3
4.Kxe5 Rc3 5.Kd5 Rc1 6.Rb4 Rd1+
7.Ke5 Rd7 8.Kf5 Kf8 9.e5 Bf2 10.Rb8+ Kg7
11.Rc8 Bb6 12.Rc6 Bg1 13.Rf6 Kf8
14.Rh6 (8.401.618.993) 3820

42 36:39 +1.59 1.Be2 Rc8 2.Bf3 Rc3 3.Bh5 Rc8 4.Be2 Rc7
5.Ra5 Rc2 6.Bh5 Rf2+ 7.Kg4 Ba7
8.Rxe5 Ra2 9.Rg5+ Kh6 10.Bxf7 Rxa6
11.Rh5+ Kg7 12.Bc4 Rg6+ 13.Kf4 Rc6
14.Bd5 (8.401.618.993) 3820


The line I get after g5 is a bit different from Andrew's and Black does have to be careful here and there, some only, singular moves, some where he has two options, but it is a draw


[pgn]
[Event "Ch World (match) (PCA)"]
[Site "New York (USA)"]
[Date "1995.09.25"]
[Round "9"]
[White "Viswanathan Anand"]
[Black "Garry Kasparov"]
[ECO "B84"]
[Result "1-0"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Be2
e6 7. O-O Be7 8. a4 Nc6 9. Be3 O-O 10. f4 Qc7 11. Kh1 Re8
12. Bf3 Bd7 13. Nb3 Na5 14. Nxa5 Qxa5 15. Qd3 Rad8 16. Rfd1
Bc6 17. b4 Qc7 18. b5 Bd7 19. Rab1 axb5 20. Nxb5 Bxb5
21. Qxb5 Ra8 22. c4 e5 23. Bb6 Qc8 24. fxe5 dxe5 25. a5 Bf8
26. h3 Qe6 27. Rd5 $1 Nxd5 (27... g5 28. Bc7 Nxd5 29. cxd5
Qa6 30. Qxb7 Qxb7 31. Rxb7 Re7 32. d6 Rd7 33. Rb6 Rc8
34. a6 Rcxc7 35. dxc7 Rxc7 36. Rb1 Ra7 (36... Kg7 {?!} 37. Bg4 {wins})
37. Be2 Bc5 38. Kh2 Kg7 39. Kg3 Rc7 40. Rb7 Rxb7 41. axb7
Bd6 {This is a dead draw finally although eval still 0.82}) 28. exd5 Qg6 29. c5 e4 30. Be2 Re5 31. Qd7 Rg5 32. Rg1
e3 33. d6 Rg3 34. Qxb7 Qe6 35. Kh2 1-0[/pgn]

[Event "Ch World (match) (PCA)"]
[Site "New York (USA)"]
[Date "1995.09.25"]
[Round "9"]
[White "Viswanathan Anand"]
[Black "Garry Kasparov"]
[ECO "B84"]
[Result "1-0"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Be2
e6 7. O-O Be7 8. a4 Nc6 9. Be3 O-O 10. f4 Qc7 11. Kh1 Re8
12. Bf3 Bd7 13. Nb3 Na5 14. Nxa5 Qxa5 15. Qd3 Rad8 16. Rfd1
Bc6 17. b4 Qc7 18. b5 Bd7 19. Rab1 axb5 20. Nxb5 Bxb5
21. Qxb5 Ra8 22. c4 e5 23. Bb6 Qc8 24. fxe5 dxe5 25. a5 Bf8
26. h3 Qe6 27. Rd5 $1 Nxd5 (27... g5 28. Bc7 Nxd5 29. cxd5
Qa6 30. Qxb7 Qxb7 31. Rxb7 Re7 32. d6 Rd7 33. Rb6 Rc8
34. a6 Rcxc7 35. dxc7 Rxc7 36. Rb1 Ra7 (36... Kg7 {?!} 37. Bg4 {wins})
37. Be2 Bc5 38. Kh2 Kg7 39. Kg3 Rc7 40. Rb7 Rxb7 41. axb7
Bd6 {[d]8/1P3pkp/3b4/4p1p1/4P3/6KP/4B1P1/8 w - -
This is a dead draw finally although eval still 0.82})
28. exd5 Qg6 29. c5 e4 30. Be2 Re5 31. Qd7 Rg5 32. Rg1
e3 33. d6 Rg3 34. Qxb7 Qe6 35. Kh2 1-0

[Lyudmil Tsvetkov]

So, Eelco, the important thing you confirm is Rd5 is a mistake, after which g5 probably leads to a draw, so the Rd5! move could easily be traded for Rd5? move.

Ronald's programming skills could be left for the time being aside.

Would be interesting to see a deeper analysis of the c5 continuation instead of Rd5. Users reported Komodo saw only 50cps white edge, SF is even more clueless, does not even pick up the move, but for me c5 should win very easily.

[syzygy]

In the g5 line, SF is very much afraid of a line ending here:

[d]8/2r2pk1/P7/1B2pK2/3bP3/6P1/R7/8 w - - 0 31

150cps white edge, why on earth should SF choose this line?

As the white king and rook get more active, the advantage increases, but no much benefit from this.

Maybe the fact that on g5 white has only acceptable response, Bc7, should somehow ring a bell.

Ra4 seems to win here:

Code: Select all

info depth 46 seldepth 70 multipv 1 score cp 511 nodes 8812113811 nps 13491384 hashfull 999 tbhits 32712853 time 653166 pv a2a4 c7c8 a4c4 c8b8 b5c6 d4e3 c6d5 b8b1 c4c7 b1f1 f5e5 f1a1 c7f7 g7g6 f7f3 e3g1 d5c4 a1a4 e5d5 a4a5 d5e6 a5a4 c4d3 g6g5 f3f7 a4a3 d3c4 a3a4 f7g7 g5h6 g7c7 a4a5 c7f7 a5a4 c4d3 a4a3 d3b5 h6g5 e4e5 a3a1 f7g7 g5h6 e6f6 a1a3 g7b7 a3f3 f6e6 h6g5 e6d5 f3f8 d5e4 f8d8

At depth 47 the score is +7.42 now and keeps climbing.
It just resolved to +8.82:

Code: Select all

info depth 47 seldepth 77 multipv 1 score cp 882 nodes 13365921658 nps 13649113 hashfull 999 tbhits 54915966 time 979252 pv a2a4 c7c8 a4c4 c8b8 b5c6 d4e3 c6d5 b8b6 c4c6 b6b1 c6c7 b1f1 f5e5 f1a1 c7f7 g7g6 f7f3 e3g1 d5c4 a1a4 e5d5 a4a5 d5e6 a5a4 c4d3 g6g5 e4e5 g5g4 f3f7 g1e3 d3b5 a4a1 b5c4 g4h3 e6d6 a1d1 d6c6 d1c1 c6d5 h3h2 e5e6 c1d1 d5c6 d1d4 e6e7 d4c4 c6d6 c4a4 e7e8q a4a6 d6d5 a6a3 e8h8 h2g1 h8h3 a3a5 d5e4

Depths 48 and 49:

Code: Select all

info depth 48 seldepth 77 multipv 1 score cp 857 nodes 15661735979 nps 13622620 hashfull 999 tbhits 65067155 time 1149686 pv a2a4 c7c8 a4c4 c8h8 c4c7 d4b6 c7b7 b6e3 b5c4 h8h5 f5g4 h5h1 b7f7 g7g6 f7c7 h1g1 c7c6 g6h7 c4d5 g1a1 g4f5 e3d4 f5g5 d4e3 g5f6 a1f1 f6e5 f1a1 c6c7 h7g6 d5c4 g6g5 e5d6 g5g4 e4e5 a1d1 c4d5 d1a1 c7g7 g4f5 g7f7 f5g6 d5c4 a1d1 d6e6 d1d8 f7c7 e3b6 c7b7 d8e8 e6d6 e8d8 d6e7 b6a5 b7b5 a5c7 b5c5 d8e8 e7e8 c7b8 e5e6 b8a7

Code: Select all

info depth 49 seldepth 78 multipv 1 score cp 979 nodes 18545141867 nps 13510053 hashfull 999 tbhits 76196621 time 1372692 pv a2a4 c7c8 a4c4 c8h8 c4c7 d4b6 c7b7 b6d4 b5c4 h8h5 f5g4 h5h1 b7f7 g7g6 f7c7 g6f6 c4a2 h1a1 c7c2 a1e1 c2c6 f6g7 g4g5 d4e3 g5f5 e1f1 f5e5 f1a1 c6c7 g7h6 a2c4 a1a5 e5e6 e3g1 c7f7 a5a4 f7f1 g1d4 e4e5 d4e3 f1f6 h6g5 f6f3 e3c1 e6d5 a4a5 d5c6 g5g4 f3f7 a5a4 c4d3 a4a3 d3e2 g4h3 e5e6 a3c3 c6d5 c3g3 e6e7 g3g8 e2b5 g8g5 d5c6 g5g8 a6a7 c1g5 f7f8

And why is Rc8 an obligatory move after Ra4?

I do not have the possibility to look in-depth into the position right now, but my impression is the ending is a draw.

Yes, and so much for your impression I'm afraid.

Anyway, this later position is just a side line; the main line arises after Re7 on the position Andrew analysed for hours. I would claim with bigger certainty that it is draw there.

Andrew's line is simply the pv starting from the position after g5. The end position of a pv is of no relevance.

I'm not claiming g5 loses, but it should be clear that "your impression" is indeed just a personal impression.