26 February 2012

Pawn Wars

Teaching chess to young children, I constantly stress the benefits of playing pawn wars. Two variations: just pawns, first to promote wins; pawns and kings from several set positions (see my 2009 "Pawn Wars"). In the last round of the 20th Dave Collyer Memorial tournament, I was on board two as one of two players with 3.5. Two players with perfect 4.0 scores were battling for first on board one, and I was battling with an expert for second alone, a massive tie for second, or a share of first (if the players on board one draw). Our game ended with kings and pawns.

White to move

White has a clear advantage, but I did not know it. Indeed, before the rooks were exchanged, I thought my opponent (Black) had found the way to win a game that I was trying to draw. He sought the pawn break from this position.

Black to move

We had reached this position twice, on moves 23 and 25. The second time, he opted for 25...f5. There were several possibilities, but we reached the position in the first diagram after 31.Rxd7 Kxd7. There, we have a position that will not be drawn. One of us will be in zugzwang when the other runs him out of pawn moves. Calculating such things with the rooks and a knights on the board is very difficult.


  1. Hi James! In the pawn ending, as you say, there will obviously be a mutual zugzwang after 1. Kd3, Ke7 2. Ke4, Kf6. Once the pawn moves run out, either White will win the e5 pawn or black will win the f5 pawn, which should decide things.

    But, without doing any calculating, I'm surprised that Black would lose, since he goes 2nd and both queenside pawn sets already have been touched twice (that's the usual way to determine the zugzwang winner in such cases). Playing h4 won't help either, since Black just plays ...h5 and keep the situation the same. And White can't pussyfoot about getting to e4, since Black has two squares to get to f6 (e7 and f7) and White has only one square (d3) to let him get to e4. In other words, Black could zugzwang White if White hedges for a move.

    So what am I missing? Shouldn't Black win that position?

  2. Oh wait, I'm missing the fact that White's a-pawn hasn't moved yet, so it can move either to a3 or a4. . It's more complicated than I thought . . .

  3. David, I counted pawn moves and saw that he had more, but then noticed that one move, b3-b4 stops two of his moves. Kd3 was immediately necessary, and then Ke4 could wait. Then, I played h3-h4--a slight inaccuracy because a6-a5 requires precision. I replayed the position against Rybka this morning and learned that b4 as my first pawn move makes matters easier.

    From the diagram, our game concluded 32.Kd3 Ke7 33.h4 Kf6 34.Ke4 h5 35.b4 b6 36.c4 a5 37.bxa5 bxa5 38.c5 and in zugzwang, Tim resigned.

  4. Yeah, b4 is the star move. That reserve option to play a3 or a4 decides it. It's fun how Black has two reserve moves of his own (with the b and h pawns), but neither helps him in this situation. If his pawn were only at ...a7, or if yours was at a3, then he would win!

  5. Very interesting endgame. Can you email me the pgn James? Seems like I can learn something from this game!

    Congrats on performing so well! Reminds me of my 2010 EWO...:)