Strategy of 'My System', Lesson 2

We're learning chess strategy from the classic My System by Aaron Nimzowitsch. I'm your guide, likesforests. Grab a cold one, pull up a chair, and enjoy. :)

§1.4 An exchange followed by a gain of tempo

Imagine you're running a race against a friend. Her cell phone rings and she stops to answer it, but you continue running. Chances are you'll win. In the beginning of a game of chess both sides race to develop pieces and control the center. In this lesson we'll focus on using the exchange as a weapon to force our opponents to waste time.

I. Marshall Defense

After 1.d4 d5 2.c4 Nf6 we reach the position above. 3.cxd5! Black must recapture or he's down a pawn, but 3...Nxd5 allows 4.e4! Nf6 5.Nc3. White's winning the race! He's brought out two pawns and a knight. Black's only brought out a knight.


You gain time when you exchange if your opponent recaptures with a piece, then you make a useful move, which chases his piece to a square no better than its original square.


All that applies above. Clearly, 4...Nb6 wouldn't be much different. We can also say that in the final position White achieved three useful moves: Nc3, d4, e4. Black apparently achieved only one: Nf6. In chess lingo we say that White's ahead by (3-1)=2 tempi (half-moves).

3...Qxd5 4.Nc3 Qa5 5.Nf3 (press the "Move List" button) is more interesting. Nimzowitsch says White gains a tempo... he's implying that the queen is no better on a5 than on d8!?

Question: Is Nimzowitsch correct?

The queen on a5 is double-edged. She exerts more influence over the center, but she is also more vulnerable to attack (Bd2, Nc4, b4). There are masters on both sides of the issue. In any case, if there is a difference, it's very small, so Nimzowitsch is correct.

II. A delay changes nothing.

After 5.fxe5! the Black queen is forced to move onto the e5 square, where White can gain time attacking it. But 5...Qxe5+ is check... does that change anything? No! White blocks with 6.Be2 Nc6 and then follows through on attacking the enemy queen with 7.Nf3 Qd6 8.O-O. White gained a tempo from the exchange, despite the delaying check ...Qxe5+.

III. 'My System' in Practice: Failure!

5.cxd5 Nxd5 6.e4?! I was happy to apply a My System concept in a real game, but afterwards I discovered Black could have refuted my plan with an accurate move.

Question: Do you see it?

After ...Nf6?! or ...Nb6?! White's gained time. ...Nxc3 is stronger because White will have to get in d4 to prove that the free pawn move b2-c3 was useful. But the killer move is ...Nf4! On f4 the knight immediately threatens Nd3+, winning the bishop pair and after 5.d3 Bg4! White can't even kick the knight with 6.g3? due to 6...Nd4!, winning a piece.

Black's knight ended up on a square better than its original square.

IV. 'My System' in Practice: Success!

Karpov's play is an excellent demonstration of all the strategic concepts we've discussed. He gains a superior position strategically, then finishes off his opponent tactically.

Extra Credit - The following are a couple instructive examples of how to win after gaining a lead in development:

I welcome criticism and even praise if you feel it's deserved. Lesson 3 will be out next Friday... if you haven't already, Add me as a Friend to enjoy it hot off the press!


  • 8 years ago


    Bravo, my friend..! Excellent post..! Cool

  • 8 years ago


    Thanks for posting.

  • 8 years ago


    THANKS LIKESFORESTS! This stuff has helped VERY much!

  • 8 years ago



  • 8 years ago


    Once again, good job.  I am going to get the cell phone numbers of my opponents.  :)

  • 8 years ago


    Very nice.

  • 8 years ago


    very nice chess theory, depends on how hard you study it

  • 8 years ago


    dvaud> In the Karpov game, I missed why not 7...Kxh7?

    If you click the "MOVE LIST" button you can see that variation.  :)

    If 10...Kxh7? then 11.Ng5+. Per my recollection of Art of Attack, White should have considered the three critical retreats before playing this sac...

     11...Kg8 Qh5 sets up a mating net (as per the interactive diagram).

     11...Kh6 Nxe6 wins at least a queen.

    11...Kg6 {best?} 12.Qd3+ f5 13.exf6+ Kxf6 14.Qf3+ wins at least a rook.

  • 8 years ago


    In the Karpov game, I missed why not 7...Kxh7?

  • 8 years ago


    forever to load

  • 8 years ago


    This Friday. It's coming, it's coming... :)

  • 8 years ago


    likesforests, the very first post there is tempo-controversy! I freaked out seeing that.

    And, sorry i omiited to say it before, but it was cheguevara who pointed to that thread first. So I am just spreading the word; I didnt discover the word :)

    Thanks here too cheguevara,for pointing out that discussion at!

  • 8 years ago


    Thanks everyone for bearing with the slight delay on Lesson 3. I won my tourney this weekend with a perfect score so the extra study time helped!  :)

    normajeanyates - Thanks, I'll check that thread out tomorrow. I bought Rybka 3 earlier this week and I think the Rybka3/Human and IDeA will be a great aid in checking these articles for accuracy.

  • 8 years ago


    btw very interesting discussion in link below [with the proviso that engines are not so good at openings - hence opening books for engines :)]

    "rybka forum - rybka v My System": click on link below:

  • 8 years ago


    That 2000+ v 2000+ filtering - i wish more databases did that automatically!

    One very general point: [important part of justification of this material] "you have to learn the rules before you know when to break them".

  • 8 years ago


    There have been some good questions. I'm learning from them, too.  :)

  • 8 years ago


    StarJock> In the ... statistics you presented, were the players rated Expert [or better]

    The data included a mix of players. I recrunched the numbers with games where both players were 2000+ elo and not much changes:

    ...Nf4 - Black scores 73%

    ...Nxc3 - Black scores 37%

    ...Nb6 - Black scores 33%*

    ...Nf6 - Black scores 12%*

    ...Nde7 - Black scores 0%*

    * Due to the small sample size, one could argue the statistical significance of these results, though they're within +/- 15% of the previous results.

  • 8 years ago


    Pereira_Gomes, you can find all the lessons here. :)

  • 8 years ago


    mauerblume> You said this would be the killing move for that line. But normal reaction would be ...   2. d4

    Yes, I overstate things by calling ...Nf4 a "killer move". It's strong and refutes e4, yes, but Black does not win on the spot. After 7.d4 (the move you suggested, which is probably the most accurate) the most common line goes 7...exd4 8. Bxf4 dxc3 9. Qxd8+ Nxd8 10. bxc3 Ne6 when White's position is undesirable (but not completely lost):

  • 8 years ago


    StarJock> After a few moves, things even out if play is accurate. IMHO none of the three black knight moves offers a long term advantage for either player.

    You raise an interesting point, but I think you're seeing the forest but not the trees. Suppose one day we prove that with perfect play both 1.e4 e5 and 1.e4 f5? lead to draws. Would that mean they're equal moves for human OTB games? Of course not! It's much harder to hold a pawn-down game than an equal game.

    Similarly, it's true that 6...Nde7 does not lead to a forced win for White... but it's still worse than 6...Nf4. It gives White an advantage. With precise play that advantage may dissipate, but Black's margin for error is smaller and with inaccuracy defeat will be swifter.

    After 6...Nde7 7.Bc4 clearly most would prefer White:

    Rybka 2 even at depth=14 believes White is better, but by depth=18 it finally concedes that Black is equal. In other words, with 7-9 strong moves Black equalizes.

    However, what about human vs human statistics for each knight move. Will they support that they're all equal, or show some are better than others?

       ...Nf4 = played in 36% of games, Black scores 71%

       ...Nxc3 = played in 30% of games, Black scores 36%

       ...Nb6 = played in 19% of games, Black scores 28%

       ...Nf6 = played in 13% of games, Black scores 25%

       ...Nde7 = played in 2% of games, Black scores 0%

    In practice...Nde7, ...Nf6, and ...Nb6 all score much worse than ...Nf4.

    A strong strategic move usually provides you with an easier game and more tactical opportunities, but it doesn't guarantee a win. :)

Back to Top

Post your reply: