Saturday, December 3, 2011

Pawn Tension part 1 (a wee lesson)

Well, I've played a lot of chess recently, and I seem to have a lot of it bubbling out of me, so hopefully this is the start of some more frequent posts ("more frequent" being a standard that is not hard to exceed). However, I do have a lot of other stuff bubbling out of me also, so we'll see which upwellings win the battle for my time.

This post (and hopefully a related one to follow) are explorations of the issues of pawn tension. It’s funny the way sometimes you see something basic in chess and then you start seeing it all over. I played a friendly game with a co-worker yesterday, and afterwards he asked me where he went wrong. I drew his attention to this position:
1. d4 d5 2. Nf3 Nf6 3. e3 Bf5 4. c4 c6 5. Bd3 Bxd3 6. Qxd3 e6

At this point, White pushed his c-pawn another square forward, which is something you should never, ever do. In fact, my computer Fritz’s evaluation of the position instantly goes from completely equal in the diagram to .4 pawns advantage for Black after White pushes the c-pawn.

The main concept here is pawn tension. In the diagram, there is tension between the black pawn at d5 and the white pawn at c4, which are mutually attacking each other. This means that Black can’t take action in the center without taking into account what might happen if White cashes in on the tension by capturing on d5. By pushing on past, White gets rid of the tension, thus giving Black a free hand for his own operations. In fact, Fritz’s preferred method of exploiting the push to c5 is for Black to play b6, immediately re-establishing pawn tension on a pair of squares that White has less control over:

This shows another reason why the push of the c-pawn is bad: it makes White’s bishop a worse piece. At this point, all of White’s center pawns are on dark squares, hemming in his remaining bishop. This is called a “bad bishop”. By contrast Black has a “good” bishop: its pawns are on squares of the opposite color. Perhaps White pushed the c-pawn in part to restrict the action of Black’s bishop, but you can see that this “restraining” pawn has become a target. In a way, the overall dynamic of this contrast is the same: White’s bad bishop can’t attack Black pawns (which are on the other color) so Black’s hand is free to do things like attacking pawns with his bishop.

In the actual game, after a few more developing moves,

Black got to play 12…e5. If Black had still needed to consider the consequences of a capture cxd4, this would have needed a whole lot more calculation. As it was, I just needed to see that there’s a potential pawn fork if the e-pawn continues forward; and after an exchange on e5, the c-pawn may also be vulnerable. After a few more exchanges in the center,

I got to play my knight to c4 (the square that White refused to contest at the very beginning), attacking both of White’s minor pieces. After 19…Nc4 20.Nb1 cxb5 21.Rxd5 b4 22.Bc1 Rfd8 I exchanged all the rooks on the d-file and my queenside pawns made a touchdown – precisely on the side of the board where White had originally advanced his own pawns.

After the game, I told my friend that his strategy had been discredited in the 1880s. I was thinking of a couple of games from the world championship match between Steinitz and Zukertort, and when I went and looked them up, it was clear why I associated them with this sort of pawn-play. They’re very cool games:

Johannes Hermann Zukertort - Wilhelm Steinitz
World Championship Match game 1, 1886

1.d4 d5 2.c4 c6 3.e3 Bf5 4.Nc3 e6 5.Nf3 Nd7 6.a3 Bd6

Look familiar? White now pushes he c-pawn forward; and on the next move, Black achieves the …e5 push
7.c5 Bc7 8.b4 e5

As in our stem game, White’s Bc1 is colossally “bad”, and Black is in control of the central tension. Black’s Bf5 is also bad theoretically; but since it’s outside the chain of black pawns, it’s a much better piece than White’s QB.
9.Be2 Ngf6 10.Bb2 e4 11.Nd2 h5 12.h3 Nf8 13.a4 Ng6 14.b5

Over the next past moves, Black has brought his knights over to the kingside and killed the pawn tension in the center, thus making sure that nothing significant is happening anywhere on the board except the sector where he plans to attack. He loved maneuvers like his Nf8-g6 here, building up a slow attack with his king still in the center. Now the knight continues his mission:
The available ways of defending the g-pawn all involve some sort of positional concession, so Zukertort just pushes it, and then…
15.g3 Ng2+!? 16.Kf1 Nxe3+ 17.fxe3 Bxg3 18.Kg2 Bc7

Steinitz has give up his charging knight for two pawns in order to expose the White king. Because the rest of the board is all locked up, he can continue feeding pieces over there without much else happening. White has to find some really precise defensive moves to neutralize this very nebulous attacking setup, and Zukertort isn’t up to the task:

He’s getting the queen out of the way so that his king can run back to the center (and hopefully even the queenside) but this is the wrong way to do it. He should move the queen to f1, overprotecting the h3 pawn and gaining a tempo by attacking the undefended Bf5. After the bishop retreats, White can play Kf2 and the king will continue his run to safety. Then the game will go on; Fritz thinks that White has a slight advantage, but there isn't any clear plan on the board for him to undertake.

In response to Qf1, Black must retreat the Bf5, since after
19.Qf1 Qd7? 20.bxc6 bxc6 21.Nb5!

White is suddenly doing very worthwhile things on the queenside. Black can’t capture the knight since the bishop recapture pins his queen. And Black’s queen can’t defend both bishops at the same time, so after 21…Be6 22.Nd6+! Black doesn’t want to leave the knight sitting there in the heart of his position, blocking the attacking Bc7; but after 22…Bxd6 23.cxd6 Qxd6 24.Ba3! suddenly White’s “bad” bishop is very actively placed, and after 24…Qd7 25.Kf2 the White king continues his trek to safety.
As it is, after
19.Qg1? Rh6 (threatening to win the queen with Rg6+) 20.Kf1 Rg6 21.Qf2 Qd7 22.bxc6 bxc6

with Black’s rook already in attacking position at g6, White has no time for Nb5 here, because after 23.Nb5 Bxh3+ 24.Ke1 Bg3 White’s queen is pinned and lost. Black will just move his king in response to Nd6+ here. So White must defend g3 with another piece:
23.Rg1 Bxh3+ 24.Ke1 Ng4 25.Bxg4 Bxg4 26.Ne2 Qe7 27.Nf4 Rh6 28.Bc3 g5 29.Ne2 Rf6 30.Qg2 Rf3!

Steinitz must have enjoyed making this move: he always did say that you should attack a pawn chain at its base. If you compare this diagram and the previous one, you’ll see that Steinitz’s pieces have been moving forward while White’s are basically just shuffling around. That trend will continue.
If 31.Nxf3 exf3 32.Qf2 fxe2 then Black has just gotten two knights for his rook; factoring in the knight that he originally sacrificed, and the pawns he’s picked up along the way, that leaves him with three pawns for the exchange, a massive material advantage.
Steinitz adds insult to injury by taking control of the open file in White’s preferred sector of the board.
32.Kd2 f5 33.a5 f4 34.Rh1 Qf7

Black has just kept building up. It’s now impossible for White to avoid the loss of major material.
35.Re1 fxe3+ 36.Nxe3 Rf2

The White queen’s only flight square is g1; but then Qf3 comes and the Ne2 falls, then the king behind it. So….
37.Qxf2 Qxf2 38.Nxg4 Bf4+ 39.Kc2 hxg4 40.Bd2 e3

This seems to have been a tremendous time scramble, but even so I think I would have been embarrassed into resigning at this point. Note, however, that Rh8+ doesn’t win Black’s rook, as it is protected by his bishop. Given what happens later, that might have been a good try, though.
41.Bc1 Qg2 42.Kc3 Kd7 43.Rh7+ Ke6 44.Rh6+ Kf5 45.Bxe3 Bxe3 46.Rf1+

Yes, apparently this really happened in a world championship game: White gives check with an unprotected rook next to the black queen, and Black doesn’t take it. One for the ages.
46…Bf4 0–1

Okay, this is enough for the moment; hopefully I’ll be posting the second Zukertort-Steinitz encounter and another relevant game anon.


  1. 46 Rf1+ is not such a bad cheapo in a time scramble since black might lose on time after 46...Qxf1 47 Ng3+ Kf4 48 Nxf1 if he pauses to think about how to save his bishop.

  2. Ummm... yeah! That's right!
    I saw the fork.
    Even though it was really late...