Knight Sacrifices on d4

The Tata Steel tournament is more than halfway through, and currently Anand and Nakamura share first place in the A group with 5.5/8. One of Anand’s wins earlier in the event caught my eye, and I thought I’d show a couple connections I made in m head when I saw that game.

(FEN: rn3rk1/ppq2ppp/4b3/nBP1p3/1Q2P3/P1P1BP2/4N1PP/R4RK1 w - - 0 16)

In the above position (from a 4.f3 Nimzo), Anand uncorked 16.Nd4!!, sacrificing a knight for a powerful phalanx of central pawns. After 16…exd4 17.cxd4 Nbc6 18.Qc3, Wang Hao tried to stop the march of the pawns with 18…Ne7. After 19.Rfd1, Black played 19…Rad8. The main point behind 19.Rfd1, as Anand said himself, was to draw a rook to d8, thereby taking that square away from the queen.

With d8 occupied, the transfer of the bishop to g3 becomes that much more powerful, and so 20.Bf2 not only reintroduces d5 as a threat (because after Black takes, White will still have the advantage of the two bishops), but threatens Bg3 to hit the queen that defends the Na5. White is already much better and Anand went on to win in smooth fashion.

The first thing I realized was that this was amazingly similar to what I had just read a day before on ChessCafe. There, a book review of Moskalenko’s Revolutionize Your Chess was published and I had skimmed it until one diagram caught my eye:

(FEN: k1rn3r/ppqb4/3p1npp/P1pPp3/2PNP3/Q1PBBP2/1R1K2P1/1R6 b - - 0 26)

White had just played 26.Nd4!!, and this apparently featured in an entire chapter on Nd4 sacrifices in the Nimzo! There was one further example given in the book review (from Moskalenko’s own practice), which is actually from a similar 4.f3 line to the Anand game above.

This got me thinking – did I have any similar Nd4 sacrifices? I couldn’t think of any, but I did remember being on the receiving end of a Nd4 move against then-FM Vladimir Mezentsev. I didn’t remember too much else about the game, but I remembered I lost and that Nd4 was the stunner. This game was from 1999, when both of us were about 2400 FIDE. I had annotated the game soon after the event, but I hadn’t touched the notes since then.

(FEN: 2r5/1pq3kp/p2p1npb/3Pp3/P1p1P3/2P3PP/1PN1Q1K1/5RB1 w - - 0 38)

I had just played 37…Nd7-f6, oblivious to the danger on the kingside dark squares. In my notes, I wrote that “I was choosing between two moves, 37…Nf6? and 37…Rf8 but wasn’t sure which one to play.” Frankly, 37…Rf8 is a bad move because of 38.Rxf8 Kxf8 39.Na3, winning the c4-pawn.

Mezentsev then played 38.Nd4!! (my exclamations then), and since Black can’t tolerate a knight landing on e6, I had to play 38…exd4 39.Bxd4 Rf8. I’d have to find the scorebook that has this game to know the times, but maybe we were in some time pressure. Mezentsev erred with 40.h4?, when 40.Qg4 was definitely better, aiming for e6. With 40.h4, Mezentsev wanted to prepare g4 I think.

(FEN: 5r2/1pq3kp/p2p1npb/3P4/P1pBP2P/2P3P1/1P2Q1K1/5R2 b - - 0 40)

I played 40…Qe7, and gave my move a question mark, concluding that 40…Qd7 was the right square. The point is that after 41.g4, Black has the ridiculous 41….Be3! to break the pin. White doesn’t want to take with his queen and allow …Qxg4+, so he takes with the bishop when Black exchanges queens on g4 and seems to survive. In reality, though, White is still better in that endgame after 42.Bxe3 Qxg4+ 43.Qxg4 Nxg4 44.Bd4+ Kg8 45.Rb1!, with 46.b3 to follow. Black’s queenside is rather weak.

The real problem with 40…Qd7 turns out to be 41.Qf3! Qe7 42.g4!, when the …Be3 resource doesn’t save Black. So, it turns out 40…Qd7 isn’t the saving move. Instead, the saving move comes after 40…Qe7 41.Qf3 (otherwise 41…Kg8 leaves Black a piece up), but I didn’t find the one and only way to avoid a worse position. The diagram is below.

(FEN: 5r2/1p2q1kp/p2p1npb/3P4/P1pBP2P/2P2QP1/1P4K1/5R2 b - - 0 41)

White’s threat is to play 42.g4, and then after 42…g5 continue with 43.Qf5!. The pressure is then unbearable as a White pawn will end up g5 rather soon. The bishop on h6 is rather silly.

That should have clued me in to the saving move – 41…Bd2!. Black’s bishop doesn’t do much on d2 except get out of the way of the kingside pawns. After 42.g4 g5 43.Qf5, Black now has 43…h6 and White can’t improve. Meanwhile, 41…Bd2 prepares 42…h5 to take g4 off the table. With best play, the game can end in a draw in a number of ways, as White just has to chase Black’s bishop back and forth. For example, 41…Bd2 42.Rf2 Bc1 43.Rf1 and a draw is called for.

As I didn’t find 41…Bd2, I ended up returning my extra piece with interest and went on to lose the game. Meanwhile, Mezentsev’s Nd4 move was impressive at the time, but not quite as impressive when I look at it now.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s