Rook Lift

Study notes from a club player rated somewhere in the 1600s, mistakes first.

King and pawn: counting squares before I move

2026-01-04 · 2 min read

The position from my last Sunday league game: white king on e4, pawn on a3; black king on e7, pawns on f6 and h6. I was White and had just traded rooks to reach this, believing it was winning because I had the outside passed pawn and he had nothing passed at all. It is a draw at best. I knew the rule about counting squares and I did not count. Here is the count I should have done.

My a-pawn needs five moves to reach a8. His king on e7 needs three moves to reach b7, where it covers both a8 and a7 and the pawn can go no further without help. If I push the pawn first, it is my move, so the pawn is on a4 when his king is on d7, a5 when it is on c7, a6 when it is on b7, and then it is stuck, with my king four files away. If instead I bring my king over first, his king walks to b7 just as mine arrives on b5, and the familiar rook-pawn fortress follows, with his king shuffling between a8 and b7 and my pawn going nowhere. Either way, a draw, and in the meantime his f and h pawns can start running down the board with nothing in front of them, which turns my winning attempt into a losing one if I am slow.

What I did in the game was push the pawn to a4, then a5, then realise his king was arriving, then run my king back towards his pawns, which lost two more tempi and came within one move of losing. We agreed a draw with bare kings and I went home and set it up.

The process I am now making myself do, on paper if necessary, before any pawn ending:

The real error was two moves earlier, when I traded rooks into this position because it looked simpler. It was simpler. It was also a draw. A rook ending with the outside passed pawn would have been genuinely better for me and I converted a clear edge into half a point by trying to be clever.