Botvinnik mastery

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

Moderators: hgm, Rebel, chrisw

zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Botvinnik mastery

Post by zullil »

Lyudmil Tsvetkov wrote:[D]1r1r2k1/2b1qp1p/b1p3p1/p1p1p3/2P1P3/1PN1BP2/P1Q3PP/R2R2K1 b - - 0 25

The position is from a game of Botvinnik.
Could you (your engines) find the winning continuation (and there is only one, I think)?


I would appreciate engine analysis, even if some engines are unable to find the move.

Best, Lyudmil
Stockfish finds the move in a fraction of a second, but considers the resulting position as favoring White by more than a pawn:

Code: Select all

Searching: 1r1r2k1/2b1qp1p/b1p3p1/p1p1p3/2P1P3/1PN1BP2/P1Q3PP/R2R2K1 b - - 0 25
infinite: 1 ponder: 0 time: 0 increment: 0 moves to go: 0
 1   -0.63   00:00     229  Bb7 
 2   -0.53   00:00     473  Bb7 a3 
 3   -0.73   00:00    1313  Bb7 Qf2 Bd6 
 4   -0.75   00:00    1984  Bb7 Qf2 Bd6 Rac1 
 5   -0.99   00:00    7240  Bb7 Qf2 Rd4 Bxd4 cxd4 Na4 
 6   -0.85   00:00   20014  Rd4 Na4 Qd6 Rac1 Bc8 Bxd4 exd4 
 7   -1.13   00:00   28603  Rd4 Na4 Bc8 Bxd4 cxd4 Qd2 Bd7 c5 Qh4 
 8   -1.09   00:00   35775  Rd4 Na4 Bc8 Bxd4 cxd4 Qd2 Be6 Rac1 Kg7 c5 
 9   -1.03   00:00   56507  Rd4 Na4 Bc8 Bxd4 cxd4 Qd2 Kg7 c5 f5 exf5 Bxf5 Rac1 
                            Qh4 
10   -1.03   00:00   83490  Rd4 Rac1 Kg7 Na4 Rbd8 Bxd4 cxd4 Qd2 Qa3 g3 Bb7 c5 
11   -1.03   00:00  103911  Rd4 Rac1 Kg7 Na4 Rbd8 Bxd4 cxd4 Qd2 Qa3 g3 Bb7 c5 
12   -1.07   00:00  247401  Rd4 Rac1 Bd6 Na4 Qc7 Qf2 Rd8 Bxd4 cxd4 Qd2 Ba3 Rb1 
                            Bb4 
13   -1.35   00:01   1265K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 f6 Nd3 Kg7 c5 Ba6 Rac1 a4 
                            bxa4 Ba5 Rb1 Rxb1 Rxb1 Bxd3 Qxd3 Qxc5 
14   -1.35   00:01   1428K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 f6 Nd3 Kg7 c5 Ra8 Rac1 Be6 
                            g3 Rb8 Kg2 a4 b4 a3 
15   -1.31   00:01   1536K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 f6 Nd3 Kg7 c5 Ra8 Rac1 Be6 
                            Rb1 Rb8 Rf1 Bf7 Rbd1 Be6 Qd2 
16   -1.23   00:01   1670K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 Rac1 Qa3 Rd2 
                            f5 Re1 fxe4 Rxe4 Bf5 Ree2 Bd6 
17   -1.25   00:01   2594K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Qg5 Qe2 f6 
                            Rab1 Ra8 Re1 Qh6 Qc2 Rb8 Qc1 Qxc1 Rexc1 h6 
18   -1.25   00:01   2808K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Qg5 Qe2 f6 
                            Rab1 Ra8 Re1 Qh6 Qc2 Rb8 Qc1 Qxc1 Rexc1 h6 
19   -1.23   00:01   3389K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rf1 Qf7 Rfc1 Rb8 Re1 Qe7 Rbd1 Rd8 g3 a4 b4 a3 Kg2 
20   -1.23   00:02   3770K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rf1 Qf7 Rfc1 Rb8 Re1 Qe7 Rbd1 Rd8 g3 a4 b4 a3 Kg2 
21   -1.23   00:02   4386K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rf1 Qf7 Rfc1 Rb8 Re1 Qe7 Rbd1 Rd8 g3 a4 b4 a3 Kg2 
22   -1.25   00:02   5686K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rd2 Rd8 Qc1 Rb8 Qa3 Bc4 Re1 Bb5 Red1 Bxd3 Rxd3 
23   -1.31   00:03  10962K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Qf7 Nb2 Qe7 Qe2 Bf7 Na4 Rb8 Qa6 
24   -1.31   00:03  14150K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Qf7 Nb2 Qe7 Qe2 Bf7 Na4 Rb8 Qa6 
25   -1.31   00:04  16680K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Qf7 Nb2 Qe7 Qe2 Bf7 Na4 Rb8 Qa6 
26   -1.29   00:05  27698K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Rf8 Qf2 Ra8 Nb2 Rb8 Qe2 
27   -1.21   00:07  39432K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 Kg8 Nb2 Rb5 Qc2 Rb8 Nd3 Kg7 Qb2 a4 b4 
                            Ra8 Qa3 Qf7 Rd2 Bc4 Rdc2 Be6 Rb1 Qe8 Rd2 Bc4 
28   -1.21   00:08  44787K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 Bd8 Qc2 Bc7 g3 Ra8 Kg2 a4 b4 a3 Re1 Qf7 
29   -1.21   00:09  51523K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 Bd8 Qc2 Bc7 g3 Ra8 Kg2 a4 b4 a3 Re1 Qf7 
30   -1.21   00:11  65115K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 a4 bxa4 Qf7 Rb1 Ra8 Nb4 Rxa4 Qc2 Ra3 Rd2 
                            Rc3 
31   -1.21   00:13  81097K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Ra8 
                            Nb2 Rb8 Qe2 
32   -1.21   00:23 147958K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Qd8 
                            Rb1 Qe7 g3 Qf7 Kg2 Ra8 Qc2 a4 bxa4 Bxa2 Rb7 Be6 
33   -1.21   00:32 219039K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Qd8 
                            Rb1 Qe7 g3 Qf7 Kg2 Ra8 Qc2 a4 bxa4 Bxa2 Rb7 Be6 
34   -1.21   01:47 808203K  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Nb2 
35   -1.21   01:59 897231K  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Rb1 Qf7 Qc2 Ra8 Rdc1 Qe8 Rf1 
                            Qb8 Qd2 Bf7 Rfd1 Qb5 a4 Qb8 
36   -1.21   02:09 969903K  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Rf1 Ra8 Qc2 Bf7 Rcd1 Be6 Rf2 
                            Rb8 Rb1 Bc8 Rc1 Be6 Rd1 Ra8 Rff1 Rb8 Rb1 Ra8 Rbd1 
37   -1.21   02:41   1206M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Nd3 
38   -1.23   03:56   1803M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 g3 Rf8 Qe2 Ra8 Rf1 Qd8 Qd2 Qe7 Rf2 Rb8 Nd3 
                            Rd8 Rd1 Ra8 Qc1 a4 b4 Qd8 Rfd2 a3 Kg2 
39   -1.23   04:58   2275M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 g3 Rf8 Qe2 Ra8 Rf1 Qd8 Qd2 Qe7 Rf2 Rb8 Nd3 
                            Rd8 Qb2 Bc8 Qe2 Ba6 Qc2 Rb8 Rff1 Ra8 Rfd1 Bc8 Rf1 
                            Ba6 
40   -1.23   06:05   2801M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 g3 Rf8 Qe2 Ra8 Rf1 Qd8 Qd2 Qe7 Rf2 Rb8 Nd3 
                            Rd8 Qe2 Re8 Nb2 Ra8 Na4 Qd8 Qd2 Bh3 Nb2 Qe7 Nd3 Bc8 
                            f4 exf4 Nxf4 Qxe4 Rff1 a4 Rce1 
Nodes: 2824103479
Nodes/second: 7661037
Best move: Rd4
Ponder move: Na4
Top
User avatar
lucasart
Posts: 3232
Joined: Mon May 31, 2010 1:29 pm
Full name: lucasart

Re: Botvinnik mastery

Post by lucasart »

zullil wrote: Stockfish finds the move in a fraction of a second, but considers the resulting position as favoring White by more than a pawn
Exactly. That's what I've been trying to explain to Lyudmil since the beginning. Rd4 may be a nice move, but it does not win the game.

With best play, white has at least a draw. Botvinnik's opponent completely blundered after Rd4, which is why he lost. Not becuase Rd4 is winning.

One could let strong engines play the game from the position after Rd4, where the engine plays against itself. And I bet you that all these games would be either draws, or white would win.

Also, I do not agree that Rd4 Na4 is the only right choice. I am convinced that the materialistic Rd4 Bxe4! is fine. In fact, accepting the material sacrifice to be the exchange up could be white's only chance at winning the game, *if* he can hold it. Na4 is at best playing for a draw, it seems. So in that sense, I prefer Bxe4 to Na4.
Theory and practice sometimes clash. And when that happens, theory loses. Every single time.
Top
User avatar
michiguel
Posts: 6401
Joined: Thu Mar 09, 2006 8:30 pm
Location: Chicago, Illinois, USA

Re: Botvinnik mastery

Post by michiguel »

lucasart wrote:
zullil wrote: Stockfish finds the move in a fraction of a second, but considers the resulting position as favoring White by more than a pawn
Exactly. That's what I've been trying to explain to Lyudmil since the beginning. Rd4 may be a nice move, but it does not win the game.

With best play, white has at least a draw. Botvinnik's opponent completely blundered after Rd4, which is why he lost. Not becuase Rd4 is winning.

One could let strong engines play the game from the position after Rd4, where the engine plays against itself. And I bet you that all these games would be either draws, or white would win.

Also, I do not agree that Rd4 Na4 is the only right choice. I am convinced that the materialistic Rd4 Bxe4! is fine. In fact, accepting the material sacrifice to be the exchange up could be white's only chance at winning the game, *if* he can hold it. Na4 is at best playing for a draw, it seems. So in that sense, I prefer Bxe4 to Na4.
Rd4 is not winning, but it is the move. If SF thinks white is better by 1.2 pawns, it may be one of the typical positions in which SF is completely off in its absolute evaluation.

I picked the book "One hundred selected games" by Botvinnik, and it took me a second to find the game because I realized I left a bookmark in it :-). Over the years, I went back to some games I was impressed and bookmark them to eventually test the engine. Apparently I did that with this game. I won't try to explain anything and just quote Botvinnik, whose comment is great. There is a lot to learn from it.

"Rd4!!
The only way to save the c5 pawn. It is difficult for White to exploit his material superiority when there are no open files, and Black has good chances on the K side. Best of all for White would be to take the Rook with the Bishop and then transfer the Knight to d3. We must remark that this sacrifice of the exchange (Rd4!!) is possible only because Black will still have one Rook, without which his Bishops would not be dangerous. This is why Black could not delay with this sacrifice; if White exchanged one pair of Rooks he would ensure himself the win."

Later, Botvinnik thinks that 33.g4 was a mistake

"White's nerves fail him. The threat g4 was very unpleasant, but this weakening of the K side helps black in his attack"

Miguel
Top
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: Botvinnik mastery

Post by Lyudmil Tsvetkov »

zullil wrote:
Lyudmil Tsvetkov wrote:[D]1r1r2k1/2b1qp1p/b1p3p1/p1p1p3/2P1P3/1PN1BP2/P1Q3PP/R2R2K1 b - - 0 25

The position is from a game of Botvinnik.
Could you (your engines) find the winning continuation (and there is only one, I think)?


I would appreciate engine analysis, even if some engines are unable to find the move.

Best, Lyudmil
Stockfish finds the move in a fraction of a second, but considers the resulting position as favoring White by more than a pawn:

Code: Select all

Searching: 1r1r2k1/2b1qp1p/b1p3p1/p1p1p3/2P1P3/1PN1BP2/P1Q3PP/R2R2K1 b - - 0 25
infinite: 1 ponder: 0 time: 0 increment: 0 moves to go: 0
 1   -0.63   00:00     229  Bb7 
 2   -0.53   00:00     473  Bb7 a3 
 3   -0.73   00:00    1313  Bb7 Qf2 Bd6 
 4   -0.75   00:00    1984  Bb7 Qf2 Bd6 Rac1 
 5   -0.99   00:00    7240  Bb7 Qf2 Rd4 Bxd4 cxd4 Na4 
 6   -0.85   00:00   20014  Rd4 Na4 Qd6 Rac1 Bc8 Bxd4 exd4 
 7   -1.13   00:00   28603  Rd4 Na4 Bc8 Bxd4 cxd4 Qd2 Bd7 c5 Qh4 
 8   -1.09   00:00   35775  Rd4 Na4 Bc8 Bxd4 cxd4 Qd2 Be6 Rac1 Kg7 c5 
 9   -1.03   00:00   56507  Rd4 Na4 Bc8 Bxd4 cxd4 Qd2 Kg7 c5 f5 exf5 Bxf5 Rac1 
                            Qh4 
10   -1.03   00:00   83490  Rd4 Rac1 Kg7 Na4 Rbd8 Bxd4 cxd4 Qd2 Qa3 g3 Bb7 c5 
11   -1.03   00:00  103911  Rd4 Rac1 Kg7 Na4 Rbd8 Bxd4 cxd4 Qd2 Qa3 g3 Bb7 c5 
12   -1.07   00:00  247401  Rd4 Rac1 Bd6 Na4 Qc7 Qf2 Rd8 Bxd4 cxd4 Qd2 Ba3 Rb1 
                            Bb4 
13   -1.35   00:01   1265K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 f6 Nd3 Kg7 c5 Ba6 Rac1 a4 
                            bxa4 Ba5 Rb1 Rxb1 Rxb1 Bxd3 Qxd3 Qxc5 
14   -1.35   00:01   1428K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 f6 Nd3 Kg7 c5 Ra8 Rac1 Be6 
                            g3 Rb8 Kg2 a4 b4 a3 
15   -1.31   00:01   1536K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 f6 Nd3 Kg7 c5 Ra8 Rac1 Be6 
                            Rb1 Rb8 Rf1 Bf7 Rbd1 Be6 Qd2 
16   -1.23   00:01   1670K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 Rac1 Qa3 Rd2 
                            f5 Re1 fxe4 Rxe4 Bf5 Ree2 Bd6 
17   -1.25   00:01   2594K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Qg5 Qe2 f6 
                            Rab1 Ra8 Re1 Qh6 Qc2 Rb8 Qc1 Qxc1 Rexc1 h6 
18   -1.25   00:01   2808K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Qg5 Qe2 f6 
                            Rab1 Ra8 Re1 Qh6 Qc2 Rb8 Qc1 Qxc1 Rexc1 h6 
19   -1.23   00:01   3389K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rf1 Qf7 Rfc1 Rb8 Re1 Qe7 Rbd1 Rd8 g3 a4 b4 a3 Kg2 
20   -1.23   00:02   3770K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rf1 Qf7 Rfc1 Rb8 Re1 Qe7 Rbd1 Rd8 g3 a4 b4 a3 Kg2 
21   -1.23   00:02   4386K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rf1 Qf7 Rfc1 Rb8 Re1 Qe7 Rbd1 Rd8 g3 a4 b4 a3 Kg2 
22   -1.25   00:02   5686K  Rd4 Na4 Bc8 Bxd4 cxd4 Nb2 Kg7 Nd3 Be6 c5 Ra8 Rab1 f6 
                            Rd2 Rd8 Qc1 Rb8 Qa3 Bc4 Re1 Bb5 Red1 Bxd3 Rxd3 
23   -1.31   00:03  10962K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Qf7 Nb2 Qe7 Qe2 Bf7 Na4 Rb8 Qa6 
24   -1.31   00:03  14150K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Qf7 Nb2 Qe7 Qe2 Bf7 Na4 Rb8 Qa6 
25   -1.31   00:04  16680K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Qf7 Nb2 Qe7 Qe2 Bf7 Na4 Rb8 Qa6 
26   -1.29   00:05  27698K  Rd4 Na4 Bc8 Rac1 Be6 Qe2 Kg7 Nb2 f6 Bxd4 cxd4 c5 Rb5 
                            Nd3 Rb8 Qd2 Ra8 Qc2 Rf8 Qf2 Ra8 Nb2 Rb8 Qe2 
27   -1.21   00:07  39432K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 Kg8 Nb2 Rb5 Qc2 Rb8 Nd3 Kg7 Qb2 a4 b4 
                            Ra8 Qa3 Qf7 Rd2 Bc4 Rdc2 Be6 Rb1 Qe8 Rd2 Bc4 
28   -1.21   00:08  44787K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 Bd8 Qc2 Bc7 g3 Ra8 Kg2 a4 b4 a3 Re1 Qf7 
29   -1.21   00:09  51523K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 Bd8 Qc2 Bc7 g3 Ra8 Kg2 a4 b4 a3 Re1 Qf7 
30   -1.21   00:11  65115K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Kg8 
                            Qf2 Kg7 Qe2 a4 bxa4 Qf7 Rb1 Ra8 Nb4 Rxa4 Qc2 Ra3 Rd2 
                            Rc3 
31   -1.21   00:13  81097K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Ra8 
                            Nb2 Rb8 Qe2 
32   -1.21   00:23 147958K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Qd8 
                            Rb1 Qe7 g3 Qf7 Kg2 Ra8 Qc2 a4 bxa4 Bxa2 Rb7 Be6 
33   -1.21   00:32 219039K  Rd4 Na4 Bc8 Rac1 Be6 Bxd4 cxd4 c5 Bd8 Nb2 Kg7 Nd3 
                            Bc7 Qd2 f6 Qc2 Ra8 Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Qd2 Qd8 
                            Rb1 Qe7 g3 Qf7 Kg2 Ra8 Qc2 a4 bxa4 Bxa2 Rb7 Be6 
34   -1.21   01:47 808203K  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Nb2 
35   -1.21   01:59 897231K  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Rb1 Qf7 Qc2 Ra8 Rdc1 Qe8 Rf1 
                            Qb8 Qd2 Bf7 Rfd1 Qb5 a4 Qb8 
36   -1.21   02:09 969903K  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Qe2 Rb5 Nd3 Rb8 Rf1 Ra8 Qc2 Bf7 Rcd1 Be6 Rf2 
                            Rb8 Rb1 Bc8 Rc1 Be6 Rd1 Ra8 Rff1 Rb8 Rb1 Ra8 Rbd1 
37   -1.21   02:41   1206M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 Nd3 
38   -1.23   03:56   1803M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 g3 Rf8 Qe2 Ra8 Rf1 Qd8 Qd2 Qe7 Rf2 Rb8 Nd3 
                            Rd8 Rd1 Ra8 Qc1 a4 b4 Qd8 Rfd2 a3 Kg2 
39   -1.23   04:58   2275M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 g3 Rf8 Qe2 Ra8 Rf1 Qd8 Qd2 Qe7 Rf2 Rb8 Nd3 
                            Rd8 Qb2 Bc8 Qe2 Ba6 Qc2 Rb8 Rff1 Ra8 Rfd1 Bc8 Rf1 
                            Ba6 
40   -1.23   06:05   2801M  Rd4 Na4 Kg7 Nb2 Bc8 Nd3 Be6 Rac1 f6 Bxd4 cxd4 c5 Ra8 
                            Nb2 Rb8 g3 Rf8 Qe2 Ra8 Rf1 Qd8 Qd2 Qe7 Rf2 Rb8 Nd3 
                            Rd8 Qe2 Re8 Nb2 Ra8 Na4 Qd8 Qd2 Bh3 Nb2 Qe7 Nd3 Bc8 
                            f4 exf4 Nxf4 Qxe4 Rff1 a4 Rce1 
Nodes: 2824103479
Nodes/second: 7661037
Best move: Rd4
Ponder move: Na4
I see that in some variations Stockfish finds Bc8 and f5 for black, but for some reason it refrains from pushing that pawn further on, which would be natural. A pawn occupying a more advanced rank, if there are no other threats, would always be natural.
Top
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: Botvinnik mastery

Post by Lyudmil Tsvetkov »

lucasart wrote:
zullil wrote: Stockfish finds the move in a fraction of a second, but considers the resulting position as favoring White by more than a pawn
Exactly. That's what I've been trying to explain to Lyudmil since the beginning. Rd4 may be a nice move, but it does not win the game.

With best play, white has at least a draw. Botvinnik's opponent completely blundered after Rd4, which is why he lost. Not becuase Rd4 is winning.

One could let strong engines play the game from the position after Rd4, where the engine plays against itself. And I bet you that all these games would be either draws, or white would win.

Also, I do not agree that Rd4 Na4 is the only right choice. I am convinced that the materialistic Rd4 Bxe4! is fine. In fact, accepting the material sacrifice to be the exchange up could be white's only chance at winning the game, *if* he can hold it. Na4 is at best playing for a draw, it seems. So in that sense, I prefer Bxe4 to Na4.
Lucas,
Rd4 is the best move, after other continuations black is worse. I thought that the move was winning by force, but now I have changed my opinion that it only draws (but only if white transfers its king to safety in the closed center, which many might not recognize). Still, it is the best possible move.

[D][D]1r4k1/2bbq2p/8/p1p1p1p1/2PpPp2/1P3P2/P1Q2BPP/3R1RK1 w - - 0 33

You might wish to continue the game from the above position, which is about forced (I do not think it matters much if there is a knight on d3 or a bishop on f2). Do engines still think that white is better? Based on what?

For me, white is severely on the defense, and that is why it plays g4.

- again, the b3 white pawn does not exist, the b4 square is overcontrolled by black pawns and pieces, so it can not advance, consequently, 2 black pawns on the queen side stop 3 white
- black has a protected passer, 2 bishops, that already makes the position materially equal.
- but black is still attacking, storming with pawns, how could white prevent this attack, unless it plays g4, like in the game, or is blessed with finding the transfer of the king to the center (with safety code relying more on PST and enemy attacks, rather than on pawn shelter and closed files, that might be difficult to do)

Again, in terms of material, please take a look, the b3 pawn DOES NOT EXIST (a severe penalty for it, and you solve the positional mystery, you do not need any other deeper positional understanding, but to have the correct chess eval), black is a full pawn up. How do we know that pawn is backward? Well, if white plays a4, b3 will be backward-fated (-70 outright); if it plays a3, to prepare b4, black will still have predominant pawn and piece control of b4, so it still can not advance, and the penalty might be the same. You need pawns, only if they can do something, if they can advance to take part into the struggle, and a pawn that can not do that is not a full pawn.

You might want to run some checks on the above position with engines. I am sure that, even in fast games, black will be winning more often, because there is simply not a way to lose here. Still, the engines might be showing big advantage for white.
Top
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: Botvinnik mastery

Post by Lyudmil Tsvetkov »

michiguel wrote:
lucasart wrote:
zullil wrote: Stockfish finds the move in a fraction of a second, but considers the resulting position as favoring White by more than a pawn
Exactly. That's what I've been trying to explain to Lyudmil since the beginning. Rd4 may be a nice move, but it does not win the game.

With best play, white has at least a draw. Botvinnik's opponent completely blundered after Rd4, which is why he lost. Not becuase Rd4 is winning.

One could let strong engines play the game from the position after Rd4, where the engine plays against itself. And I bet you that all these games would be either draws, or white would win.

Also, I do not agree that Rd4 Na4 is the only right choice. I am convinced that the materialistic Rd4 Bxe4! is fine. In fact, accepting the material sacrifice to be the exchange up could be white's only chance at winning the game, *if* he can hold it. Na4 is at best playing for a draw, it seems. So in that sense, I prefer Bxe4 to Na4.
Rd4 is not winning, but it is the move. If SF thinks white is better by 1.2 pawns, it may be one of the typical positions in which SF is completely off in its absolute evaluation.

I picked the book "One hundred selected games" by Botvinnik, and it took me a second to find the game because I realized I left a bookmark in it :-). Over the years, I went back to some games I was impressed and bookmark them to eventually test the engine. Apparently I did that with this game. I won't try to explain anything and just quote Botvinnik, whose comment is great. There is a lot to learn from it.

"Rd4!!
The only way to save the c5 pawn. It is difficult for White to exploit his material superiority when there are no open files, and Black has good chances on the K side. Best of all for White would be to take the Rook with the Bishop and then transfer the Knight to d3. We must remark that this sacrifice of the exchange (Rd4!!) is possible only because Black will still have one Rook, without which his Bishops would not be dangerous. This is why Black could not delay with this sacrifice; if White exchanged one pair of Rooks he would ensure himself the win."

Later, Botvinnik thinks that 33.g4 was a mistake

"White's nerves fail him. The threat g4 was very unpleasant, but this weakening of the K side helps black in his attack"

Miguel
Thanks Miguel again, for the commentary!

I understand Botvinnik words in the following way:
- he puts 2 exclamation marks for the move, meaning it is exceptional
- Botvinnik was a modest guy, so he did not want to state outright 'white is on the defense and can only hope to draw', but he says instead 'that black is favoured by still having one rook'.
Top
User avatar
emadsen
Posts: 434
Joined: Thu Apr 26, 2012 1:51 am
Location: Oak Park, IL, USA
Full name: Erik Madsen

Re: Botvinnik mastery

Post by emadsen »

One could let strong engines play the game from the position after Rd4, where the engine plays against itself. And I bet you that all these games would be either draws, or white would win.
I ran a shootout from the position after ... Rd4 with nine strong engines, plus my own modest engine. White never lost.

Code: Select all

128 MB hash
Intel i7 960 @ 3.20 GHz
1 CPU

Shoot out:
----- Time = Blitz 2m -----
Critter 1.6a 64-bit   ½-½
HIARCS 14 WCSC   1-0
Junior 12.5.0.3   ½-½
Komodo 5.1r2 64-bit   1-0
MadChess 1.3.04   1-0
Rybka 4.1 SSE42 x64   ½-½
Shredder 12 UCI   1-0
Spike 1.4   1-0
Stockfish 2.2.2 SSE42   1-0
Zappa Mexico II x64   1-0
 
Score White: 85%, 10 Games
My C# chess engine: https://www.madchess.net
Top
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: Botvinnik mastery

Post by Lyudmil Tsvetkov »

emadsen wrote:
One could let strong engines play the game from the position after Rd4, where the engine plays against itself. And I bet you that all these games would be either draws, or white would win.
I ran a shootout from the position after ... Rd4 with nine strong engines, plus my own modest engine. White never lost.

Code: Select all

128 MB hash
Intel i7 960 @ 3.20 GHz
1 CPU

Shoot out:
----- Time = Blitz 2m -----
Critter 1.6a 64-bit   ½-½
HIARCS 14 WCSC   1-0
Junior 12.5.0.3   ½-½
Komodo 5.1r2 64-bit   1-0
MadChess 1.3.04   1-0
Rybka 4.1 SSE42 x64   ½-½
Shredder 12 UCI   1-0
Spike 1.4   1-0
Stockfish 2.2.2 SSE42   1-0
Zappa Mexico II x64   1-0
 
Score White: 85%, 10 Games
Hi Erik.
This is a Lucas quote.

Instead I suggested that white will never win in a shootout, starting from the diagram below:

[D]1r4k1/2bbq2p/8/p1p1p1p1/2PpPp2/1P3P2/P1Q2BPP/3R1RK1 w - - 0 33

Space advantage and storming pawns already confer on this second position a quality, when it is already difficult even imagining winning with white (not so in the initial diagram).

Could you, please, if at all interested, run a second shootout with the same engines, but starting from the above diagram.

Many thanks.

Best, Lyudmil
Top
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: Botvinnik mastery

Post by Lyudmil Tsvetkov »

[D]1r4k1/2bbq2p/8/p1p1p1p1/2PpPp2/1P3P2/P1Q2BPP/3R1RK1 w - - 0 33

Some additional eval suggestions that might put black in a better light in the above position:

- 2 very strong d4 and f4 apex pawns (meaning that the structure is flexible, for example, black could play g5-g4, and the f4 pawn would still be protected)
- the direction of the e5-f4 chain of pawns, as Carl Langan rightly emphasizes; it is directed towards the enemy king on g1. I think this could be done very easily in programs, whenever you have a chain, and the most advanced pawn (which should get the bonus) is closer in terms of files to the file the enemy king is than the least advanced pawn of the chain, then you aare willing to assign some bonus; 10-15 cps might be fine. I think similar considerations (you do not have to look any further, it is all in the eval) might resolve difficulties computers have with some types of closed positions, like the King's Indian, the ARB, different hippopotami, etc.

In this sense, the f4 black pawn in the above position is very strong. It might have at least the following bonuses:
- bonus for space advantage, 15cps on the king side
- bonus for an apex pawn, 5-10cps additionally
- bonus for a peak (lead) pawn of a chain (e5-f4), maybe some 20cps, considering the closeness to the enemy king, the aquare where it is and the number of pawns being led
- bonus for the direction of the chain towards the enemy king, another 10cps

Overall, in the above position, whether you are sceptical or not, the f4 black pawn might be worth some 50cps additionally. That already changes things much.

In contrast, when the white king is transferred to d3, most of these factors cease to exist, and white has drawing chances:
- space advantage and closeness to the enemy king already would have a different meaning with a king on d3, they would be less influential
- the direction of the chain would also change, etc.

Just my 2 cents.
Top
Lyudmil Tsvetkov
Posts: 6052
Joined: Tue Jun 12, 2012 12:41 pm

Re: Botvinnik mastery

Post by Lyudmil Tsvetkov »

Just a small digression, referring to the possibility to transfer the king into the center, but with all files around it closed.

[D]1k6/8/5p2/4pPp1/3pP1Pp/3P3P/5K2/8 w - - 0 1

King on a fully-closed side. This is not a position, but just a representation.

The f2 white king is the safest there could be. All files on the king side are closed. The white king does not need pawn shelter around it, it can change squares on the side without any real threat, as the closed side eliminates in itself all or at least most enemy attacking opportunities. Almost any shelter would by definition be weaker than a fully-closed side. So such a king would be due some very nice bonus without any further considerations.
Top

Return to “Computer Chess Club: General Topics”