The following game, Samisch-Nimzowitsch, Copenhagen 1923, contains an example of a piece sacrifice that is actually quite understandable. Look at the following, and ask yourself if you would have sacrificed the knight given the same position, or if you would have fallen into the trap of "point-counting" and tried to save it.
White had just played 20.e2-e4, with a discovered attack on the knight.
20... fxe4!
You don't need Fritz to tell you that this is good. All you have to do is visualize what the position will look like after 21. Qxh5 Rxf2:
Black has seized control of the f-file (note that White can't play a rook to f1 because of the bishop on b5) and will probably double up on it. White's pieces are all tangled up defensively on the back ranks. Black's bishops are like sharks with frickin' laser beams, pewpewpew, whereas White's just seem to be in the way. All this, and two pawns, for the sacrificed knight. Then, consider that White's extra piece is the undeveloped knight on b1 that's stuck protecting the d2 bishop.
Just for kicks, let's also apply the space-counting technique mentioned in Best Lessons of a Chess Coach. The rules are a bit unclear, but here's how I apply them:
- For every Black piece, count the number of squares it attacks on the opponent's side. For example, the bishop on b5 attacks 5 squares on White's side of the board.
- If there's a piece blocking the way, you can't count the squares beyond. So for the bishop on d6 I count g3 and not h2.
- Occupying a square on the opponent's side doesn't count, but protecting a piece on the other side does. So I count d2, e2, g2, f1, f3 and f4 but not f2 for the rook on f2. I count b4 for the bishop on d6 but not for the b4 pawn itself. We're counting applied force, not occupation.
- You can count a square more than once if more than one piece attacks it, because that means you're projecting extra force. The square f3 is counted twice for Black because it's attacked by both pawn and rook.
If you don't like these rules, relax. This is just one way of illustrating whether one side has a space advantage. Here, it's pretty clear just looking at the board who has the advantage, but I thought I 'd introduce the concept here.
Black out-attacks White by 20 to 17. However, it's only that close because of the relative queen activity (9 for White vs. 0 for Black), and it's clear that the White queen can't do much all by herself. Should the a8 rook finds an open file and the queen placement change, White will be absolutely smothered.
I want to re-emphasize that this is more of a party trick than meaningful analysis, but it gives you another way of thinking about space advantage. Don't do this over the board, for Pete's sake!
White lost fairly quickly after
22. Qg5 Raf8 23. Kh1 R8f5 24.Qe3 Bd3 25. Rce1 h6 0-1. Fritz finds some inaccuracies on both sides but nothing that would save White.
Counting the point values of the pieces is helpful, but shouldn't be adhered to dogmatically. You should be
willing to risk some of the "safer" sacrifices, like minor piece for two pawns, or Rook for minor piece and (usually) a pawn, if you feel there's adequate compensation. I think it's important to get practice exercising such judgements. For example, such sacrifices can be used defensively, as well as offensively, to help defuse an attack.
Nimzo's sacrifice here feels perfectly natural to me, and I would like to think that I'd find it over the board. I know that I've made similar judgement calls in my own games. Sometimes I've been wrong, but at least I'm confident that I'm capable of thinking outside of the point-counting box and considering other factors.
