9

One of the illustrative games featured in Fred Reinfeld's The Complete Chessplayer is a game between A. Muffang - P. Devos, France-Belgium 1948:

[FEN ""]

1. d4 d5 2. c4 dxc4 3. Nf3 Nf6 4. e3 e6 5. Bxc4 c5 6. O-O a6 7. a3 Be7 8. Qe2 b5
9. Ba2 Bb7 10. dxc5 Bxc5 11. b4 Ba7 12. Bb2 O-O 13. Nbd2 Ne4? 14. Rfd1 Qe7 15. Rac1 Nxd2
16. Qxd2! Bb6 17. Qc3! f6 18. Nd4! Bxd4 19. Qxd4 Nc6 20. Qc5 Rfe8 21. Rd6 Nd8
22. Qd4! Bc8 23. h4 Kh8 24. Bb1! Rb8 25. Qe4 f5 26. Qe5 Ra8 27. Rc7!! 1-0

The point in question is Reinfeld's commentary on move 16, after 16. Qxd2!; he says:

A pretty variation is 16 ... N-B3; 17 RxN! BxR; 18 Q-B3 (double attack) threatening mate and therefore winning the Bishop on QB6, with the victorious material advantage of two minor pieces for a Rook.

Translated into algebraic, that is 16... Nc6 17. Rxc6! Bxc6 18. Qc3:

[FEN "rn3rk1/bb2qppp/p3p3/1p6/1P6/P3PN2/BB1nQPPP/2RR2K1 w - - 0 16"]

1. Qxd2! Nc6 2. Rxc6! Bxc6 3. Qc3

But what about 18... Qf6? That not only parries the mate threat, but if now 19. Qxc6, then 19... Qxb2, bishop for bishop, and White cannot even continue with 20. Qxa6? (taking a vital unprotected pawn), for Black can then take White's other bishop - 20... Qxa2. In fact, any line of play makes it hard for White to rescue the bishop after 19... Qxb2. 20. Rd2??, the only possible move to protect the bishop results in 20... Qa1+! forcing mate in four. And 20. Bb1, the only possible move of the bishop to safety results in the loss of his QR pawn - 20... Qxa3.

[FEN "r4rk1/b3qppp/p1b1p3/1p6/1P6/P1Q1PN2/BB3PPP/3R2K1 b - - 1 18"]

1... Qf6 2. Qxc6 Qxb2 3. Bb1 Qxa3 (3. Qxa6? Qxa2) (3. Rd2?? Qa1+! 4. Bb1 Qxb1+ 5. Qc1 Qxc1+ 6. Rd1 Qxd1+ 7. Ne1 Qxe1#)

I don't see any way for White to justify the sacrifice of the Exchange with 17. Rxc6, since after 18... Qf6 it looks as if Black can hold his own. (Yes, there is 19. Qxf6 gxf6 20. Bxf6 winning a Pawn, but that doesn't make up for the loss of the Exchange. And Black can also double White's pawns and steer toward the endgame with 20... Bxf3 21. gxf3.)

Am I right in saying that Reinfeld made a mistake, or am I missing something?

Daniel
  • 6,212
  • 5
  • 39
  • 74
  • Phalanx-Scid shows white ahead by after `17. Qc3 f6 18. Ng5 Kh8 19. Nxe6 Rfc8 20. Nc5 Ne5 21. Qc2 a5 22. Bd4 Bb6 (+1.55)` – Tony Ennis Jun 15 '12 at 04:23
  • After `17. Rxc6`, Phalanx-Scid says: `17. ... Bxc6 18. Qc3 Qf6 19. Qc1 Bxf3 20. Bxf6 Bxd1 21. Bxg7 Rfd8 22. Bd4 Bh5 23. Qd2 e5 (+.73)` so not as good as the straight-forward variation. I'm not sure Reinfeld made a mistake as such since his move still yields an improvement. – Tony Ennis Jun 15 '12 at 04:37
  • Well, the incontrovertible "mistake" is that Reinfeld said "therefore winning the Bishop on QB6" after `16... Nc6 17. Rxc6! Bxc6 18. Qc3`. – Daniel Jun 19 '12 at 00:13

1 Answers1

10

Yes, it does look like Reinfeld made a mistake, but white should still be doing well after 18... Qf6.

White should now play 19. Qc1! hitting the queen and the bishop. Now black's best continuation is to sacrifice the queen with 19... Bxf3 20. Bxf6 Bxd1

[FEN "r4rk1/b4ppp/p1b1pq2/1p6/1P6/P1Q1PN2/BB3PPP/3R2K1 w - - 1 18"]

1. Qc1 Bxf3 2. Bxf6 Bxd1

One possible line from here is 

[FEN "r4rk1/b4ppp/p3pB2/1p6/1P6/P3P3/B4PPP/2Qb2K1 w - - 1 18"]

1. Bxg7 Kxg7 2. Qxd1 Rad8 3. Qg4+ Kh8

and white has a queen and a pawn for the two rooks, but a strong attack because black's king is so weak.

To summarize, it does appear that Reinfeld made a mistake because after 16... Nc6? white can simply play 17. Qc3! and win easily, but even the line he gives should be okay for white.

Brian Towers
  • 92,895
  • 10
  • 227
  • 372
Andrew
  • 15,186
  • 4
  • 65
  • 105
  • OK, so we both (Reinfeld and I, that is) missed some things. Still a little question - How does `17. Qc3!` win? He can't take the c6 Knight, and to my mind, Black can easily parry the mate threat (for the time being, anyhow) with `17... f6` – Daniel Jun 14 '12 at 15:28
  • `17... f6` runs into `18. Ng5!!` and white is winning material. Black's pieces are cramped so white will invade along the 7th next move. – Andrew Jun 14 '12 at 15:35
  • Then what about `18... fxg5`? – Daniel Jun 14 '12 at 15:37
  • Ah yes - then `19 Bxe6+!` I get it. Then if `19... Kh8` (trying to avoid losing the Exchange), `20. Rd7!` Very fun. Thanks! – Daniel Jun 14 '12 at 15:39
  • Even better than `19. Bxe6+` is `19. Rd7!!` and black has to resign, but both moves win quite easily for white. – Andrew Jun 14 '12 at 15:44
  • After `19. Rd7!!` Black can play `19... Qf6`. Then White has `20. Bxe6+ Kh8`, right? Just a transposed `19. Bxe6+ Kh8 20 Rd7! Qf6`. (Or does White have a better move than `20. Rd7!`) Then what? `21. Rxb7` regains the material for White, but it hands Black the initiative, doesn't it? And an exchange of queens (`21. Qxf6 gxf6`) seems bad for White's attack. – Daniel Jun 14 '12 at 15:51
  • Your line is correct, but at the end of that line white will be up a pawn, attacking f6, and black's rooks cannot move. The reason that `19. Rd7` is better than `19. Bxe6+` is because after `19. Bxe6+ Kh8 20. Rd7` black plays `20... Bd4!` and black can struggle on for a while (although white is still winning). By reversing the move order, black must lose immediately. – Andrew Jun 14 '12 at 15:55
  • I see. So the upshot is that while `16... Nc6?` loses for Black, the possible continuation is more complicated than Reinfeld pointed out. Thanks for clearing that up for me! – Daniel Jun 14 '12 at 16:02
  • After 8 hours of grinding, Phalanx-Scid does not not report a slam-dunk win for White after `16. ... Nc6?`. I think it's a bad move, but it isn't fatal. `17. ... Bxc6 18. Qc3 Qf6 19. Qc1 Bxf3 20. Bxf6 Bxd1 21. Bxg7 Rfd8 22. Qb2 Ba4 23. h3 Rd1 24. Kh2 Bb8 25. Be5 (+.60)` – Tony Ennis Jun 15 '12 at 11:15