Position A (de Groot)

Position A: White to move and Win

deGroot Position A

Chess master and psychologist de Groot gave this position to Alekhine, Keres, Euwe and other GMs at the 1938 AVRO tournament, as well as to a number of weaker, but still expert, players.

All the GMs except Flohr quickly found the win with 1. Bxd5 exd5 2. Qf3 Qd8 3. Rfe1 Kg7 4. Ng4 Nxg4 5. Bxe7.

de Groot has written much about "Position A" in his studies of how humans determine the best move in a chess game. Let's see what Rybka thinks of this position:

Rybka finds the winning move immediately:

1. Bxd5 exd5 (1...Bxd5 2.Bxf6 Bxf6 3.Nxd5 or 3.Nd7 +- ; 1...Nxd5 2.Nxd5 exd5 3.Bxe7 +- )

2. Qf3 Qd8 (2...Kg7 3.Ng4 +-)

and now:

Analysis by Rybka 3 : <25-ply>

1. � (1.28): 3.Bh6 Re8 4.Rfe1 Bd6 5.Bg5 Bxe5 6.dxe5 d4 7.Qxf6 Qxf6 8.Bxf6 dxc3 9.Rxc3 Bd7 10.Rd3 Be6 11.Red1 b5 12.h4 a5 13.Rd4 h6 14.Kh2 Rb8 15.Kg3 b4 16.axb4 Rxb4 17.Rxb4 axb4 18.Rd4

2. � (1.25): 3.Rfe1 Ne4 4.Nxc6 bxc6 5.Bxe7 Qxe7 6.Nxe4 dxe4 7.Rxe4 Qd8 8.Qc3 a5 9.Re5 Rb8 10.Rc5 Re8 11.Rxc6 Re2

3. � (1.13): 3.Rce1 Ne4 4.Bh6 Bf6 5.Nxc6 bxc6 6.Nxe4 dxe4 7.Rxe4 Bg7 8.Bxg7 Kxg7 9.Qc3 Qf6 10.g3 c5 11.Rd1 cxd4 12.Qxd4 Rfd8 13.Qxf6+

3. Rfe1 and now: 23-ply (older version of Rybka)

1. (+0.97): 3...Ne4 4.Nxc6 bxc6 5.Bxe7 Qxe7 6.Nxe4 dxe4 7.Rxe4

2. (+1.04): 3...Rc7 4.Bh6 Re8 5.g3 Qd6 6.Bf4 Qd8 7.Bg5 Rf8 8.Bh6 Re8 9.Kg2

Position B: Black to move

deGroot Position B

DeGroot says "It is impossible, however, to produce a complete analysis. No really advantageous solution for Black can be pointed out. On the contrary: Black should be happy with a draw."

Degroot gives one possible line: 1...Rb8 2.b3 Rb5 3.c4 Rg5 4.f3 h5 5.gxh5 Rxh5+ 6.Kg1 Be5 7.Rad1 Bd4+ 8.Kf1 Rh1+ 9.Ke2 Rh2 etc. But 3.d4! Rg5 4.Rad1 Rxg4 5.d5 is probably better for white.

(24-ply)

1. (+0.51): 1...Kg7 2.f3 Rb8 3.Rab1 Kf6 4.Re2 g5 5.b3 Rd8 6.b4 Ba4 7.b5 Rb8 8.Rb4

2. (+0.57): 1...Kf8 2.Rad1 a5 3.Kg1 a4 4.c4 Bd7 5.f3 Bg3 6.Re4 Rb8 7.Rd2 f5 8.Rd4

3. (+0.58): 1...a5 2.f3 a4 3.Rad1 Bg3 4.Re2 Bd7 5.c4 Rc8 6.b4 axb3 7.axb3 Rb8 8.Rb2

4. (+0.58): 1...Rc8 2.f3 Bd7 3.Re2 Kf8 4.Kg1 Bg3 5.c4 f5 6.gxf5 Bxf5 7.Rd1 a5 8.b3

5. (+0.69): 1...Rd8 2.Rad1 a5 3.b3 a4 4.Kg1 Kg7 5.Re2 Bd6 6.d4 h6 7.d5 Bd7 8.f3

6. (+0.69): 1...Rb8 2.b3 Kf8 3.Rad1 a5 4.f3 a4 5.Re2 Bg3 6.d4 Bb5 7.Re4 axb3 8.axb3

7. (+0.70): 1...Bd6 2.Rad1 Rd8 3.f3 a6 4.b3 Bb4 5.Re5 Bd6 6.Re2 Bc5 7.c3 Bb5 8.Re5

Position C: Black to move

deGroot Position C

deGroot thinks that Black can escape with the sequence 1...e5 2.Bxc8 Raxc8 3.dxe5 Qe4 (3...Rce8 or 3...Nd7 might be better for Black but white has a small advantage) 4.Rhe1 Qxe2 5.Rxe2 Ng4 6.exd6 Rxf3 but both 7.d7! Rd8 8.Re8+ Rf8 9.Nxc7 Nxf2 10.Rd4! and 7.Na7! Nxf2 8.Nxc8 Rc3+ 9.Kd2 Rd3+ 10.Ke1 Rxd1+ 11.Kxf2 cxd6 12.Re8+ offer advantage to White.

Also interesting is the quiet 1...h6 which might also equalize for Black.

(26-ply)

1. (0.06): 1...Ne4 2.Bg2 Nf6 3.Rhg1 Qb6 4.Ng5 Bd7 5.Bh3 Rae8 6.Nxe6 Rf7 7.Rge1 Bxb5 8.axb5

2. (0.12): 1...e5 2.Bxc8 Raxc8 3.dxe5 Rce8 4.Nfd4 Qe4 5.exf6 Qxe2 6.Nxe2 Rxe2 7.Rd2 Rxd2 8.Kxd2

3. (0.13): 1...h6 2.Rhe1 Bd7 3.Nh4 e5 4.Bxd7 Nxd7 5.f4 Rae8 6.Ng6 Rf6 7.Qh5 Nf8 8.Nh4

4. (0.22): 1...Bd7 2.Ng5 Rae8 3.Rhe1 Re7 4.d5 Nbxd5 5.Bxe6+ Bxe6 6.Nxe6 Nb4 7.Nbd4 Qd7 8.f4

5. (0.20): 1...Re8 2.Rhe1 h6 3.Kb2 Bd7 4.Nh4 Qb6 5.Ng6 Rad8 6.Qe3 Bxb5 7.axb5 e5 8.dxe5

6. (0.30): 1...Rf7 2.Ng5 Re7 3.Rhe1 d5 4.f4 Bd7 5.f5 e5 6.dxe5 Ne4 7.Nxe4 Rxe5 8.Kb1

Enter supporting content here