Moves we would all like to play (part 11)

Mark Liburkin, 1st Prize, USSR All-Union Tourney 1950

White to play and win

This has long been one of my favourite studies, which I first saw in the anthology Modern Endgame Studies for the Chessplayer, by the veteran Dutch master, Hans Bouwmeester. Although a piece up, the fact that two knights cannot mate a king means that White must preserve his pawn, if he is to win. There is also the small matter of the black pawns, which threaten to promote with 1…a2.

1. Ng5+!

The only way. 1.Ne5+? blocks the long black diagonal, and so allows 1…Ke4. 2.g7 a2!

1…Kf4 2. g7 


Here, Black has a major alternative in 2…Nf7, to which we will return below.

3. Kb3! Kxg5 4.Ka2!!

And, believe it or not, Black is in zugzwang! Any knight move allows 5.Ne7 and the pawn queens, moving the king to a white square allows the WN to come out with check, whilst 4…Kh4 loses to 5.Nh6 Ne7 6.Nf5+!, and 4…Kf4 to 5.Nf6 Ne7 6.Nd5+!, in both cases deflecting the black knight away and queening on g8.

As if that is not good enough, now let’s go back to move 2 and try the other defence 2… Nf7

Now the point is that 3.Nxf7? is no good because of 3…a2, but instead there follows

3. Ne6+! Ke5 4. Kb3! Kxe6 5. Ka2!!


and, scarcely credible though it is, this is also zugzwang: moving the king to a light square again allows a knight check, and 5…Ke5 6.Nf6! Nh6 7.Ng4+!, or 5…Kd6 6.Ne7! Nh6 7.Nf5+! again see the BN deflected away fatally!

I hope you can now see why Mark Liburkin is the favourite composer of one of today’s finest study composers, my friend Yochanan Afek!

%d bloggers like this: