Positional Pawn Sac

My computer issues are resolved, so I can go back to showing you games with diagrams.

This one is an amusing riff on “how to beat the computer.” To be honest I really don’t know how to beat the computer, but sometimes I get lucky. Rule number one is to get a blocked position. First, because computers really don’t understand blocked positions very well, and second, because they understand open positions really well. (They’ll kill you with tactics.)

Position after 18. … Na5. White to move.

FEN: 2r2rk1/4qpbp/pn1p4/npp1pPp1/4P1P1/2PPN1QP/PP3BB1/R4RK1 w – – 0 19

Here Shredder (with its rating set at 2321) has just played 18. … Na5. I was not very impressed with this move, because I couldn’t see that it creates any real threats. The two ideas for White that I had been setting up were 19. Nd5 and 19. h4, and I think either one of those moves would be okay here.

But one other thing caught my attention in this position: that “pocket” in Black’s position at f6. I didn’t know how long Shredder would allow it to stay there without closing it with … f6. Why not take the opportunity to move my own pawn there, freeing up the f5 square for my knight? Although I’m sacrificing a pawn, it seems pretty unlikely that Black’s backward f-pawn will ever amount to anything, and meanwhile I’ve gained a really great knight outpost. So I played 19. f6!?

To be honest I don’t know if this is a good move or not. Both Shredder and Rybka think that it reduces White’s advantage from about half a pawn to zero. However, I think that I can say with some confidence that I wouldn’t have won the game without this move. You’ll see how the knight turns into a colossus over the next 15 moves.

The game continued 19. … Qxf6 20. Nf5 Na4?!

Already the computer’s play looks a little bit dubious to me. If I were playing Black, I would try to get my knights over to the kingside as quickly as possible. I would play 20. … Nc6 with the idea of … Ne7 and then either … Nxf5 or setting up an equivalent post on f4 with … Ng6-f4.

However, it turns out that this doesn’t quite work, so Black is in a little bit more trouble than I realized. If 20. … Nc6 21. Be3 Ne7 White has the very strong move 22. h4! If 22. … gh 23. Nxh4 Black’s queen has only one square, 23. … Qe6, and then 24. g5! wins at least an exchange because of the threat of Bh3. Alternatively, if 22. … h6 White now has the cool move 23. Nxh6+! when 23. … Qxh6 24. Bxg5 bags two pawns.

Shredder’s style always tends to be sharp and aggressive, but its insistence on trying to play on the queenside allowed me to develop my kingside initiative unimpeded. But first there was the issue of what to do about the b-pawn. I played the simple and ugly 21. Rab1. Against the computer, you’ve just got to defend your loose buttons sometimes. The game continued 21. … Qe6 22. Be3 Bf6 23. Qf2 Bd8 24. h4 h6 25. hg hg 26. Qd2 (diagram).

Position after 26. Qd2. Black to move.

FEN: 2rb1rk1/5p2/p2pq3/npp1pNp1/n3P1P1/2PPB3/PP1Q2B1/1R3RK1 b – – 0 26

So far I’ve kept Black from making any progress on the queenside because of my pressure on the d- and g-pawns. Here Shredder makes what I thought was the losing move, 26. … Qg6? I thought it had to play 26. … f6, with play similar to the game except the Queen is still defending the d-pawn. However, Rybka says that Black stands 0.95 pawns worse after 26. … f6, so Shredder’s mistake must have come earlier.

After 26. … Qg6 I won my pawn back with 27. Kf2 Nb6 28. Rh1 f6 (Finally! It’s interesting that this is the point where Shredder finally “realized” that it was losing. Of course my threat was to trap its queen with 29. Rh6.) After 29. Nxd6 it’s clearly White’s game, but I’d like to go forward just a few more moves so that you can see my troops marching into the enemy capital. 29. … Rc6 30. Nf5 Rc7 31. Rh6 Qe8 32. Rbh1 Qc6 (diagram)

Position after 32. … Qc6. White to move.

FEN: 3b1rk1/2r5/pnq2p1R/npp1pNp1/4P1P1/2PPB3/PP1Q1KB1/7R w – – 0 33

Here I played 33. Bxg5 and Shredder told me in its insufferable way that I had made a mistake, because my advantage went down from +4.9 pawns to +3.3 pawns! I don’t know what it saw, but 33. Bxg5 is a completely overwhelming move that allows White’s remaining pieces, the queen and bishop and even the g-pawn, to join in the attack. In fact, that’s how I won the game, by marching the pawn to g5-g6-g7.

What’s striking to me about this game is that the pawn sac on move 19 looks a bit speculative, but by playing completely natural moves White soon had an overwhelming position. Is it because Shredder’s defense was so bad, or because the move 19. f6 is so good? The game definitely shows a lack of “feel for the position” on the computer’s part. The antidote to the knight on f5 should be a knight on f4, but instead the computer moved both of its knights way over to the queenside where they were basically irrelevant to the main action.

A game like this makes it look so easy to beat the computer. You think, “Ha! I’ve got it down to a science!” Then you lose six or seven in a row, and you realize that you don’t understand anything at all.