In order to resurrect an old line in the Scandinavian defense I am forced to accept doubled pawns at the side where I have pawn majority. To make things even worse, I reach queenless middlegame, and the endgame is not so far away! Still, my engine gives equality or small disadvantage in sample lines.

I don't have hardware strong enough for this type of analysis, so I ask for deep analysis of the following position:

[Title "White to move"]
[fen "rn2kb1r/ppp2ppp/8/5p2/3PN3/8/PPP2PPP/R1B2RK1 w kq - 0 1"]

I really believe that Black might be able to hold this position with equal chances, but I need strong engine analyzing on strong hardware. If required, break the analysis into sub-variations and post them one by one. You could start with 1.Ng5 as I consider it the strongest move. Move 1.Nc5 can be discarded as I intend to play 1...Bxc5 doubling the pawns, thus removing only advantage White has in this position.

I can wait, there is no need to hurry as I prefer quality over speed ( this should be the part of my opening repertoire ).

Thank you for your time and efforts.


Here is the way to reach this position, added on request by member bof :

[Title "Scandinavian defense"]
[fen ""]

1.e4 d5 2.exd5 Qxd5 3.Nc3 Qe5+ 4.Be2 Bg4 5.d4! Bxe2 6.Ngxe2 Qf5 7.O-O Nf6 $8 8.Ng3 Qa5 9.Nge4 Nxe4 10.Nxe4 e6!? 11.Qf3 Qf5 12.Qxf5 exf5

Please focus on the diagramed position in your comments / answers, I do not seek help with the opening moves. They are added here just because member bof asked for them. Thank you.


2 Answers 2


I did some analysis on it with Stockfish 5 (64 bits, SSE4.2, default parameters). I don't know how reliable these lines after the evaluation really are, better analyse a certain position again instead of trusting

First move: (Depth in plies: 32 full / 50 selective; 11 GNodes)

1. +/=  (0.40): 1.Ng5 f6 2.Re1+ Kd7 3.Nf3 Bd6 4.c4 Re8 5.Rd1 c6 6.g3 Na6 
2. +/=  (0.32): 1.Nd2 f6 2.Nc4 Kf7 3.c3 Bd6 4.g3 Re8 5.Kg2 Nd7 6.b3 Nb6 7.Nxd6+ cxd6 8.c4 d5 9.c5 Nd7 10.Be3 Nf8 11.a4 Ne6 12.Rfe1 a5 13.Rad1 g5 14.f4 Kg6 15.Kf3 h6 
3. =  (0.25): 1.Nc3 Nc6 2.Ne2 Bd6 3.c3 0-0-0 4.Re1 Rhe8 5.g3 Ne7 6.Bd2 Nd5 7.Rad1 b6 8.Nc1 f4 9.Nd3 fxg3 10.hxg3 Kb7 11.Kg2 h6 12.c4 Nf6 13.Rxe8 Rxe8 14.c5 bxc5 15.dxc5 Be7 
4. =  (0.19): 1.Ng3 g6 2.Re1+ Kd7 3.c3 Na6 4.Ne2 c6 5.Bf4 Bd6 6.g3 Rhe8 7.Kg2 Nc7 8.Bxd6 Kxd6 9.Nf4 Re4 10.f3 Rxe1 11.Rxe1 Re8 12.Rxe8 Nxe8 13.Kf2 Nc7 14.Nd3 f6 15.Nf4 b6 16.c4 Ke7 17.Ke3 g5 18.Nh5 
5. =  (0.01): 1.Nc5 Bxc5 2.dxc5 Na6 3.c6 bxc6 4.Re1+ Kd7 5.Bg5 f6 6.Rad1+ Kc8 7.Be3 Rd8 8.Rxd8+ Kxd8 9.Rd1+ Ke8 10.c3 c5 11.Kf1 Rb8 12.b3 Rb6 13.h4 Rd6 14.Rb1 c6 15.Ke2 Kf7 16.f3 Rd7 17.Re1 Rd8 18.Rd1 Rxd1 
6. -+  (-2.32): 1.f3 fxe4 2.fxe4 Nc6 3.c3 f6 4.Bf4 Bd6 5.g3 Kd7 6.Kg2 Rae8 7.Rae1 h5 8.a4 a5 9.h3 b6 10.b3 Rhf8 11.Bd2 Kc8 12.Bf4 Kb7 13.Bxd6 cxd6 14.Kf3 Rc8 15.Ke2 Ne7 16.Kd3 Rfe8 
7. -+  (-2.44): 1.Bd2 fxe4 2.Rae1 Kd7 3.Rxe4 Na6 4.d5 Bd6 5.b4 Rhe8 6.Rd4 Re2 7.Rd1 Rae8 8.a3 Be5 9.Rd3 Bf4 
8. -+  (-2.48): 1.Re1 fxe4 2.f3 f6 3.Rxe4+ Kd7 4.Bf4 Bd6 5.Bxd6 cxd6 6.Rae1 Nc6 7.c3 Rhe8 8.Kf2 Rxe4 9.fxe4 Re8 10.Re2 Ne7 11.Re1 a6 12.Kf3 Nc6 

Ng5: (Depth: 30 full / 46 selective; 3700 MNodes)

1. +/=  (0.38): 1...Bd6 2.Re1+ Kf8 3.c3 Nd7 4.c4 b6 5.b3 f6 6.Ne6+ Kf7 7.Bf4 Bxf4 8.Nxf4 Rhe8 9.g3 c6 10.d5 g5 11.Ne6 cxd5 12.cxd5 Rac8 13.Kg2 Nc5 14.Nxc5 Rxc5 15.d6 Rc2 16.Rad1 Rxe1 17.Rxe1 Rd2 18.Re7+ 
2. +/=  (0.40): 1...f6 2.Re1+ Kd7 3.Nf3 Bd6 4.c4 Re8 5.Rd1 Nc6 6.Bd2 a5 7.g3 b6 8.Kg2 Ke7 9.Rac1 g5 10.c5 bxc5 11.dxc5 Be5 12.Nxe5 Nxe5 13.Bc3 a4 14.Re1 Kf7 15.Rcd1 g4 16.Bxe5 Rxe5 17.Rxe5 fxe5 
3. +/=  (0.55): 1...Bb4 2.a3 Be7 3.Re1 Nc6 4.c3 Rd8 5.b4 Rd7 6.Nf3 f6 7.Rb1 Kf7 8.Bb2 Nd8 9.Re3 Ne6 10.Rbe1 Nf4 11.c4 c6 12.g3 Ng6 13.Bc3 f4 14.R3e2 Re8 15.Kg2 
4. +/=  (0.56): 1...Nc6 2.Re1+ Be7 3.c3 Rd8 4.Bf4 0-0 5.Nf3 Rd7 6.g3 f6 7.Kg2 Kf7 8.h4 Rfd8 9.Nd2 Bd6 10.Bxd6 cxd6 11.Nc4 Ne7 12.a3 Rc8 13.Ne3 Rdd8 14.c4 g6 15.b3 b6 16.d5 Re8 17.a4 a5 
5. +/=  (0.56): 1...Be7 2.Re1 Nc6 3.c3 Rd8 4.Bf4 0-0 5.Nf3 Rd7 6.g3 f6 7.Kg2 Kf7 8.h4 Rfd8 9.Nd2 Bd6 10.Bxd6 cxd6 11.Nc4 Ne7 12.a3 Rc8 13.Ne3 Rdd8 14.c4 g6 15.b3 b6 16.d5 Re8 17.a4 a5 
6. +/=  (0.59): 1...Na6 2.Re1+ Be7 3.c4 Kf8 4.Nf3 f6 5.d5 Kf7 6.Nd4 g6 7.Be3 Bd6 8.g3 Nc5 9.Rad1 Rhe8 10.Kg2 h5 11.a3 a5 12.Nb5 Rac8 13.b4 axb4 14.axb4 Ne4 15.c5 Be5 
7. +/-  (0.75): 1...h6 2.Nf3 Bd6 3.Re1+ Kf8 4.c4 Nd7 5.Bd2 g6 6.Bc3 Be7 7.Rad1 Bf6 8.Ne5 Rd8 9.f4 Kg7 10.b3 h5 11.Kf2 a6 12.h3 b6 13.Kf3 a5 14.g3 h4 15.g4 fxg4+ 16.Kxg4 
8. +/-  (0.79): 1...Nd7 2.Re1+ Be7 3.Bf4 Kf8 4.Re2 h6 5.Nf3 Bd6 6.Bxd6+ cxd6 7.Rae1 g6 8.Re7 Rd8 9.Nd2 b5 10.Nb3 Kg7 11.Na5 Nf6 12.Rxa7 Rde8 13.Kf1 Rxe1+ 14.Kxe1 Re8+ 15.Kd2 

Nd2: (Depth: 30 full / 46 selective; 8100 MNodes)

1. +/=  (0.31): 1...f6 2.Nc4 Kf7 3.g3 Bd6 4.Kg2 Nc6 5.c3 a6 6.a4 Rhe8 7.b3 h5 8.Bd2 b5 9.Nxd6+ cxd6 10.Bf4 d5 11.axb5 axb5 12.Bd6 g5 13.Rfd1 Kg6 14.Bc5 Na5 15.Rdb1 Nc6 16.Re1 Rxe1 17.Rxa8 
2. +/=  (0.34): 1...Nc6 2.Re1+ Kd7 3.Nf3 Bd6 4.c4 Rhe8 5.Rd1 a5 6.Bd2 g6 7.g3 Bb4 8.Be3 Bf8 9.Kg2 Bg7 10.a4 b6 11.d5 Nb4 12.Bd4 Bxd4 13.Nxd4 Re4 14.Rab1 Rae8 15.b3 h5 16.h3 
3. +/=  (0.34): 1...Bd6 2.Re1+ Kd7 3.Nf3 Nc6 4.c4 Rhe8 5.Rd1 a5 6.Bd2 g6 7.g3 Bb4 8.Be3 Bf8 9.Kg2 Bg7 10.a4 b6 11.d5 Nb4 12.Bd4 Bxd4 13.Nxd4 Re4 14.Rab1 Rae8 15.b3 h5 16.h3 
4. +/=  (0.41): 1...Na6 2.c3 Bd6 3.Re1+ Kd7 4.Nc4 Rhe8 5.Be3 c6 6.g3 Nc7 7.Nxd6 Kxd6 8.c4 f6 9.Kg2 h6 10.b3 b5 11.h4 bxc4 12.bxc4 Rab8 13.Bf4+ Kd7 14.Rxe8 Rxe8 15.Rb1 g5 16.hxg5 hxg5 17.Be3 
5. +/=  (0.41): 1...b6 2.Re1+ Kd7 3.Nf3 Bd6 4.c4 Na6 5.a3 Rhe8 6.Be3 f6 7.g3 c6 8.h4 Nc7 
6. +/=  (0.42): 1...Be7 2.c3 Nc6 3.g3 0-0-0 4.a4 Rhe8 5.Nc4 Bf6 6.Kg2 Ne7 7.Bd2 Nd5 8.Rfe1 Kd7 9.Kf3 h6 10.h4 a6 11.Bf4 Nxf4 12.Kxf4 g6 13.Rxe8 Rxe8 14.Ne3 Kc6 15.a5 
7. +/=  (0.43): 1...Bb4 2.c3 Bd6 3.g3 Nc6 4.Kg2 f6 5.Nc4 Kf7 6.a4 a6 7.Re1 Rhe8 8.Be3 Rad8 9.b3 Ne7 10.Rad1 Nd5 11.Nxd6+ cxd6 12.Bd2 Ne7 13.c4 Nc6 14.Bc3 Rxe1 15.Rxe1 Re8 16.Rxe8 Kxe8 
8. +/=  (0.48): 1...Rg8 2.Re1+ Kd7 3.a3 g6 4.Nf3 Bg7 5.Be3 Nc6 6.g3 h6 7.Rab1 Rae8 8.h4 g5 9.hxg5 hxg5 10.Bxg5 Nxd4 11.Red1 Kc8 12.Nxd4 Bxd4 13.Rxd4 Rxg5 14.Rbd1 b6 15.b4 f6 16.Kf1 

For validating purposes, i ran another test with Houdini 4 Pro x64 for the first move (Depth: 26 full / 61 selective; 5000 MNodes):

1. =  (0.22): 1.Nd2 Nc6 2.c3 Bd6 3.g3 Kd7 4.Kg2 Bae8 5.Bd1 f4 6.Nc4 fxg3 7.hxg3 h5 8.Bh1 Ne7 9.Bd2 Ng6 10.Bh3 h4 11.Bah1 hxg3 12.Bxh8 Bxh8 13.Bxh8 Nxh8 14.fxg3 Ng6 15.Kf3 Ne7 16.Nxd6 Kxd6 17.Bf4+ Kc6 18.c4 Nf5 19.Ke4 
2. =  (0.21): 1.Ng3 g6 2.Be1+ Kd7 3.Nf1 Nc6 4.c3 Be8 5.Bxe8 Kxe8 6.g3 Kd7 7.Kg2 Ne7 8.Bg5 Nd5 9.Be1 Bg7 10.Kf3 h6 11.Bd2 a5 12.h3 Nf6 13.Bf4 Ne4 14.Ne3 Be8 15.Kg2 
3. =  (0.18): 1.Ng5 f6 2.Be1+ Kd7 3.Ne6 Na6 4.Bf4 Be8 5.d5 g5 6.Bd2 Nc5 7.Bad1 Nxe6 8.dxe6+ Bxe6 9.Bxg5+ Bd6 10.Bxe6 Kxe6 11.Bc1 f4 12.g3 fxg3 13.hxg3 h5 14.Kg2 Kf5 15.Be3 Be8 16.Kf3 b6 17.c3 a6 18.a3 Kg6 19.a4 Kf5 20.a5 b5 
4. =  (0.12): 1.Nc3 Nc6 2.Be1+ Kd7 3.a3 Bd6 4.Ne2 Bae8 5.Be3 Bhg8 6.h4 f6 7.b4 f4 8.Nxf4 Nxd4 9.Bxd4 Bxf4 10.c4 b6 11.g3 Bd6 12.Bed1 Ke6 13.c5 Be5 14.Bxe5 fxe5 15.Kg2 Bd8 16.Bac1 Bgf8 17.Bxd8 Bxd8 18.cxb6 cxb6 19.Bc7 
5. =  (0.00): 1.Nc5 Bxc5 2.dxc5 f6 3.Bf4 Na6 4.Bad1 Bd8 5.b4 g5 6.Bxd8+ Kxd8 7.Bd2 Ke7 8.Bc3 c6 9.Be1+ Kf7 10.Bd1 Be8 11.Bd7+ Be7 12.Bd6 Be6 13.Bd7+ Be7 14.Bd6 
6. -+  (-2.37): 1.f3 fxe4 2.fxe4 Nc6 3.c3 f6 4.g3 Bd6 5.Kg2 Kd7 6.Bd2 Rae8 7.Rae1 h6 8.Rf5 a6 9.Re2 Re7 10.Rf1 Rhe8 11.Rfe1 b6 
7. -+  (-2.40): 1.Re1 fxe4 2.Rxe4+ Kd7 3.Be3 Nc6 4.Rd1 Bd6 5.c4 f5 6.Rh4 Be7 7.Rf4 Rhf8 8.Rf3 h6 9.g3 Rae8 10.Kg2 g5 11.h3 Rf6 12.Bd2 Bf8 13.Bc3 f4 14.a3 Re2 15.d5 Ne5 
8. -+  (-2.45): 1.Bd2 fxe4 2.Rae1 f5 3.f3 Nc6 4.c3 e3 5.Bxe3 Kf7 6.Rf2 Bd6 7.Rfe2 Rae8 8.g3 h5 9.Kg2 h4 10.f4 Re4 11.Kf3 Ne7 12.Bd2 Rxe2 13.Rxe2 b5 14.b3 a6 15.c4 Nc6 16.Bc3 hxg3 17.hxg3 Bb4 18.cxb5 axb5 19.Bb2 Rd8 20.Rc2 

There are some differences, but they agree with the fact, that Nc5 is the worst reply.

Here is a game played by Stockfish vs itself in a 40 minutes per 40 moves time control in this position on a Core i7-3770: (the {0.42/27 41} in the PGN denote evaluation, depth and time for the move)

rn2kb1r/ppp2ppp/8/5p2/3PN3/8/PPP2PPP/R1B2RK1 w kq - 0 1

1. Ng5 {0.42/27 41} f6 {0.33/29 28} 2. Re1+ {0.39/29 64} Kd7 {0.32/31 28} 3.
Nf3 {0.39/31 28} Bd6 {0.33/30 29} 4. c4 {0.35/30 39} Re8 {0.34/31 29} 5. Rd1 {
0.34/31 29} a5 {0.43/30 142} 6. g3 {0.36/32 188} Nc6 {0.36/32 29} 7. Bd2 {0.29/
31 313} b6 {0.29/31 29 (g6)} 8. Kg2 {0.18/30 133 (c5)} Ke7 {0.25/30 41} 9. b3 {
0.26/32 23} Kf7 {0.29/33 99} 10. d5 {0.18/33 78} Nb4 {0.17/35 41} 11. Bc3 {0.
17/35 24} Re2 {0.17/34 30} 12. Nd4 {0.09/34 39} Rxa2 {0.10/36 45} 13. Nxf5 {0.
10/36 24} Bf8 {0.10/35 31} 14. Rxa2 {0.06/33 94} Nxa2 {0.17/35 58} 15. Bb2 {0.
15/35 40} Nb4 {0.15/35 32} 16. Nd4 {0.15/33 36} Bc5 {0.12/34 33} 17. Kf3 {0.12/
34 57} Bxd4 {0.19/35 97} 18. Bxd4 {0.19/37 24} Ke7 {0.13/34 41 (a4)} 19. Ke2 {
0.27/32 59 (Te1+)} Kd6 {0.25/36 34} 20. Kd2 {0.29/36 25} Re8 {0.25/38 108} 21.
Be3 {0.31/40 25} Na6 {0.31/39 34} 22. Re1 {0.27/35 26} Kd7 {0.25/35 35} 23. Ra1
{0.24/34 40} Kc8 {0.17/36 50} 24. Rd1 {0.13/36 97} Kb7 {0.14/38 37} 25. Kc3 {
0.20/37 47} c6 {0.20/37 38} 26. h4 {0.22/36 116} g6 {0.15/36 78 (Sb4)} 27. d6 {
0.15/34 45} Nb8 {0.19/38 92} 28. d7 {0.19/38 24} Rd8 {0.19/38 39} 29. Rd6 {0.
19/37 25} Nxd7 {0.28/36 78} 30. Kc2 {0.26/37 84} Kc7 {0.20/37 41} 31. Re6 {0.
18/33 34 (Lf4)} Rh8 {0.22/37 96} 32. Kb2 {0.22/37 26} h5 {0.21/35 57} 33. Bf4+
{0.21/36 27} Kb7 {0.20/39 43} 34. Re7 {0.21/40 107} Kc8 {0.21/40 46} 35. Rg7 {
0.21/39 25} Ne5 {0.15/38 50} 36. Bxe5 {0.05/35 50} fxe5 {0.10/39 95} 37. Rxg6 {
0.10/39 29} Kc7 {0.10/38 57} 38. Rf6 {0.05/38 77} e4 {0.03/39 68} 39. Rf5 {0.
00/38 87} Kd6 {0.00/41 87} 40. Kc3 {0.00/41 33} Ke6 {0.00/42 146} 41. Rg5 {0.
00/42 48} 1/2-1/2
  • Thank you for your help, I have upvoted your post. I will thoroughly analyze the lines you posted. Best regards. Aug 5, 2014 at 7:52
  • You can omit 1.Nc5 from analysis, see my edited post. Thanks again for pointing that out. Aug 5, 2014 at 8:01
  • Your results are consistent with my engine's analysis and my own evaluation. It seems that best Black can get is small disadvantage. I doubt any improvement will be found that gives Black equal chances, so I have decided to accept officially your answer ( I have already upvoted it ). Thank you for your help. Best regards. Aug 7, 2014 at 9:57
  • Thank you. Though I wasn't sure if you want me to also analyse the lines behind Nc3 and Ng3 since Ng3 is Houdinis second choice.
    – user3598
    Aug 7, 2014 at 10:35
  • *Though I wasn't sure if you want me to also analyse the lines behind Nc3 and Ng3 * I believe Ng5 is the strongest, and this was confirmed by my engine and with your multiple analysis. The doubled pawns will not disappear so White will always have permanent minimal advantage. I hoped for a miracle... Aug 7, 2014 at 11:17

...and I'll give you some very superficial knowledge off the top of my head based on a quick glance at the position.

  1. This is very much like a Ruy exchange in that white can create a passed pawn on the queenside but black can't on the kingside. Unlike the Ruy exchange though black doesn't have the two bishops as compensation.

  2. That e-file should be a little scary for black. Black is behind in development and it could turn bad very quickly if black doesn't play accurately.

White is clearly better here. Black is hoping for a draw. White has the better position AND the initiative. White can play pretty much anything and reach a decent position but black really needs to play accurately just to reach a draw.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.