Go to the Source

Three of the past four days, endings with a knight against pawns have been the focus as I am racing through Thomas Engqvist, 300 Most Important Chess Positions (2018). One of the positions and some of the analysis is from Yuri Averbakh and Vitaly Chekhover, Knight Endings, trans. Mary Lasher (1977). As this book was one of many that I acquired the year prior to turning 64, I pulled it from the shelf and began reading it (see "64 Endgame Books"). I skimmed the first two chapters and then looked at the analysis Engqvist referenced. 

Engqvist reaches position 41 in Knight Endings via a variation from a game played in Paris in 1848. This game is not in ChessBase Mega 2024, nor on chessgames.com, nor, as near as I can tell, anywhere else in online databases. But Engqvist offers a clue to help my search for the game: "Kieseritzky remarked that the ending was very interesting" (153).

Thanks to Chess Archaeology, it only took a minute to locate La Régence: Journal des Échecs, which Lionel Kieseritzky edited. It turns out that the source of the ending was the first game in the first issue. 

The journal offers a diagram after White's move 65. Engqvist played several training games from the position after 65...Ke5 more than twenty years before writing 300 Most Important Chess Positions. His analysis of this game, which Black could have drawn and in which both players made errors is instructive.

I played this position against the computer a couple of times on Friday and then read Engqvist's analysis more carefully yesterday. I also learned to read Kieseritzky's unusual notation: 65...E65 66. E47 E66 and entered the entire endgame into my database from the journal.

Some Easier Practice

While skimming the first chapter of Knight Endings, I paused on this statement by Averbakh and Chekhover: "Starting from any point on the board, a knight on move, can stop any pawn that has not gone beyond the fourth rank" (2).

Accordingly, I created an exercise to play against the computer. Stockfish on the iPad offered minimal resistance and was not worth the effort. I could have come to my office and used Stockfish 16 resident in Fritz, but was on the couch in the living room with my dog on my lap, so I tried using Lichess. This effort resulted in a game where I had to find a few only moves, but my composition was not as challenging as I hoped.

White to move

Almost any move works here, but I opted to only move the knight.

1.Nf7 b4 2.Nd6 Kd5

White to move

This position makes a better composition as only one move works.

3.Nf5 Ke4 4.Nd6+ Kd5 5.Nf5 Kc4 6.Ne3+ Kd3

White to move

7.Nd1

Only move, as are the next three. But they are not difficult.

7...Kc2 8.Ne3+ Kd3 9.Nd1 Kc2 10.Ne3+ Kc1 11.Nc4 b3 12.Ne5 Kc2

White to move

13.Nc4 Kc3 14.Na3

14.Nb6 and 14.Ne3 also work because 14...b2 would walk into a fork. I opted for the elementary 14.Na3 because of a known pattern.

14...b2 15.Nb1

The game went on until move 28, but there are no difficulties.