Here is the opening in question:

[FEN ""]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 Nxe4? 4. Qe2! Qe7 5. Qxe4 d6

White has the advantage no matter which pawn is moved next, 6. d4 or 6. f4 (there is no way to keep the knight, but after the exchange white is up a pawn). However, in the interest of knowledge, which pawn is better to push?

  • If white plays 6. f4, the line continues 6... dxe5 7. fxe5 and now white has two powerful central pawns while black has none. This also allows white to play the strong 8. e4 potentially creating a pawn duo in the center, although black can prevent this with 7... c5.

    However, this also ditches the f-pawn, severely weakening the kingside position and violating the "rule of thumb" to avoid pushing the f-pawn (and losing it to boot). This makes castling kingside a lot less attractive and could allow black to mount an effective attack if white does so. (It also provides black with a weakness to take advantage of before the king has castled, the e1-h4 diagonal and the king's lack of protection in general, if white decides to castle queenside.)

  • If white plays 6. d4, castling kingside becomes more feasible again. Black has to move the queen out of the way before castling (or black can just fianchetto, but this still wastes a turn moving the g-pawn), losing a tempo (black will probably want to castle soon to avoid having the king attacked down the nice and open d file).

    This move has its disadvantages too, though. The pawn on e5 is, for all practical purposes, an isolated pawn (moving the f-pawn would negate any possible advantage of playing 6. d4 instead of 6. f4). This makes it very vulnerable to attack (ex. 7... Nc5 practically forces 8. Bf4). And just as 6. f4 violates the common "rule of thumb" to avoid pushing the f-pawn, 6. d4 violates the guideline to capture towards the center with pawns.

I ended up playing 6. f4. Black subsequently blundered away the queen and resigned soon after... but I'm still curious about this position.

Am I missing anything in my analysis? Which move is better overall for white? Could white have played better in any previous moves?


4 Answers 4


White has the advantage no matter which pawn is moved next, 6. d4 or 6. f4 (there is no way to keep the knight, but after the exchange white is up a pawn).

My initial reaction is pretty much the same as yours. You'd have to be a very poor player to even think of playing this as black. However, we are both wrong! Much stronger players than me play it as black and even win sometimes.

One of these is Austrian IM Friedrich Volkmann.

Here are 3 of his wins.

First against a stronger opponent

[FEN ""]
[Event "EU-ch 4th"]
[Site "Istanbul"]
[Date "2003.06.05"]
[Round "6"]
[White "Nedev, Trajko 2518"]
[Black "Volkmann, Friedrich Karl 2435"]
[Result "0-1"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 Nxe4 4. Qe2 Qe7 5. Qxe4 d6 6. d4 dxe5 
7. dxe5 Nc6  8. Bb5 Bd7 9. Nc3 Qb4 10. Bc4 O-O-O 11. a3 Qa5 12. Bf4 Bb4 
13. O-O Bxc3 14.  bxc3 Be6 15. Bd3 Qxc3 16. Rab1 Ba2 17. Rbd1 Kb8 
18. Rfe1 h6 19. Bg3 Rd4 20. Qe2  Rhd8 21. f4 g6 22. Rc1 Bc4 23. Bf2 Rxf4
24. Be3 Nd4 25. Qd1 Rh4 26. Bf2 Rh5 27.  Re4 Be6 28. Rb1 Nc6 29. Qf3 b6 
30. Rc4 Bxc4 31. Qxc6 Rxd3 32. cxd3 Qxd3 33. Rc1  Rh4 34. h3 Rf4 
35. Re1 Rxf2 36. Kxf2 Qd2+ 37. Kf3 Bd5+ 0-1

Against a much weaker opponent

[FEN ""]
[Event "ICS Anniversary op"]
[Site "Fics Internet"]
[Date "2002.01.15"]
[Round "7"]
[White "Alaimo, Riccardo 2158"]
[Black "Volkmann, Friedrich Karl 2414"]
[Result "0-1"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 Nxe4 4. Qe2 Qe7 5. Qxe4 d6 6. d4 dxe5 
7. Qxe5 Qxe5+  8. dxe5 Bf5 9. Bd3 Bxd3 10. cxd3 Nc6 11. O-O O-O-O 
12. f4 Nb4 13. Nc3 Bc5+ 14.  Kh1 Nxd3 15. Ne4 Bd4 16. Rb1 f6 
17. exf6 gxf6 18. Bd2 Rhe8 19. Nc3 Nf2+ 20.  Rxf2 Bxf2 21. Rd1 Rxd2 0-1

Against a slightly weaker opponent

[FEN ""]
[Event "EU-chT (Men)"]
[Site "Leon"]
[Date "2001.11.14"]
[Round "8.2"]
[White "Haznedaroglu, Kivanc 2328"]
[Black "Volkmann, Friedrich Karl 2414"]
[Result "0-1"]

1. e4 e5 2. Nf3 Nf6 3. Nxe5 Nxe4 4. Qe2 Qe7 5. Qxe4 d6 6. f4 dxe5 
7. fxe5 f6 8.  Bc4 Qxe5 9. Qxe5+ fxe5 10. d3 Nc6 11. Be3 Nd4 
12. Kd2 Nf5 13. Bf2 Bd7 14. Nc3  O-O-O 15. Rae1 Re8 16. Bf7 Re7 
17. Bd5 c6 18. Be4 b6 19. Rhf1 g6 20. Kc1 Bh6+  21. Kb1 Rf8 
22. Bf3 Ref7 23. Ne4 Kc7 24. c3 Nd6 25. Ng3 Bh3 26. Bg1 Nf5 
27. Be4  Nxg3 28. Rxf7+ Rxf7 29. hxg3 Bf5 30. Rf1 Kd6 31. Kc2 Ke6 
32. Rf3 Bxe4 33. dxe4  Rxf3 34. gxf3 Bf8 35. b4 h5 36. Kd3 g5 
37. Bf2 Be7 38. a4 h4 39. g4 c5 40. Kc4  cxb4 41. cxb4 h3 
42. Bg3 a6 43. b5 a5 44. Kd3 Bc5 45. Ke2 Kd6 46. Kf1 Be3 
47. Ke2 Bf4 48. Bxf4 exf4 0-1

After those games I'm not so sure that it is a blunder. The statistics still favor white by something like 40% white wins to about 20% black wins, so it looks like a better opening for white but psychologically I can see how it would be good for black. Black gets the kind of position he knows and is comfortable with while white thinks he is playing an idiot and so is maybe more likely to play badly.


As white, you first need to acknowledge that black has some compensation for the pawn: he will get a minute lead in development, the extra pawn on e5 can become weak, and the queen's placement on e4 is suspect. So if you hold on to the pawn too strongly (or wrongly), you can be overrun quickly. On the other extreme, if you give it back too easily, black would come out ahead.

This line has been employed by Volkmann and Afromeev rather frequently, as well as Ipatov, all of whom are highly rated, and they have obtained good results with black, albeit usually against relatively weaker opponents (which means players in the 2400-2500 range...). In almost all of these games, white played 6.d4 except for Romain-Ipatov, where white played 6.f4, and queens were exchanged.

Strange-looking, rarely played sidelines like these are usually not worth overstudying, because your opponent will know a lot about it if you choose to enter a complicated variation. One sub-optimal idea (but incredibly simple) can be found in Yu-Ipatov from the 2014 World Blitz Championship: white's solution was to play 6.d4 dxe5 7.dxe5 Nc6 8.Nc3 Qxe5 9.Qxe5+ Nxe5 10.Bf4 Bd6 11. Bd3 Bd7 12.Ne4 O-O-O 13. Nxd6 cxd6. White has the bishop pair and black's d6 pawn is isolated (but not immediately weak). It's probably not enough to win if black doesn't make any mistakes, but then again, you can't really win a game without them. The beauty of this simple idea is that black can't throw you off with his preparation while you get a small edge, but of course, if you are faced with a must-win scenario, then perhaps not this line.

This variation does demand some respect; 3...Nxe4 is not a blunder; perhaps it is inferior and gives up a pawn for insufficient compensation (which is the point you have to prove as white), but it is certainly no more of a mistake than playing the King's Gambit as white!

Summary and final point: white can easily get an edge in this line; black's only task is to demonstrate that it's not sufficient for a win, which can be difficult in practice. As white, the easiest way to lose in this line is to believe that you have a big advantage, or worse, to think that you are simply a pawn up.


At 14 ply Fritz 6 prefers 6. d4 (+0.59) to 6. f4 (+0.50), so they are nearly the same good.


It depends what your goal is.
Play for win or play for best average result?

6d4 will win about 33% and draw 48% for a score of about 57% vs 43% for black.

6f4 will win about 50% and draw 25% for a score of about 62.5% vs 37.5%

The difference is that trying for the higher winning score increases you chances of losing in that particulr game a net 6%.

In theory you play f4 but in practice for a single game, especially a tournament that only has a small number of rounds compared to your entire lifetime of playing, it might make sense to try to avoid the loss more than to try to win a bit more.

Fischer went all out to win every game. Will that strategy work for you too?

Depending on your style you might look ahead to the endgame and see which pawn move might result in the preferable position. If you plan to slash and burn attack and take your chances in the end game then you may play f4 to maximise your chance of outright winning.

To directly answer the OP: Yes.
White should play one of those moves as they get better results than the alternatives.

No top GM (2700+) plays that variation per chess tempo DB. And I would prefer to deviate much sooner. It all comes down to other factors like style, preference, and what your objective is.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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